Lectrodiagnostic Medicine: Second Edition

LECTRODIAGNOSTIC
MEDICINE
Second Edition
DANIEL DIJMITRU, M D , P H D
Professgpand Deputy Chairman
Department of Rehabilitation Medicine
University of Texas Health Science Center at San Antonio
San Antonio,Texas
ANTHONY A. AMATO, MD
Associate Professor
Department of Neurology
Harvard Medical School
Chief, Neuromuscular Division
Director, Clinical Neurophysiology Laboratory
Vice-Chairman, Department of Neurology
Brigham and Women's Hospital
Associate Neurologist and Neuromuscular Consultant
Massachusetts General Hospital
Boston, Massachusetts
MACHIEL ZWARTS, M D , P H D
Professor
Department of Clinical Neurophysiology
Institute of Neurology
University Medical Centre
Nijmegen,The Netherlands
H A N LEY & BELFUS, I N C , / Philadelphia

Publisher: HANLEY & BELFUS, INC.
Medical Publishers
210 South 13th Street
Philadelphia, PA 19107
(215) 546-7293; 800-962-1892
FAX (215) 790-9330
Web site: http://www.hanleyandbelfus.com
Note to the reader: Although the information in this book has been carefully reviewed for correctness of dosage and in-
dications, neither the authors nor the publisher can accept any legal responsibility for any errors or omissions that may
be made. Neither the publisher nor the authors make any warranty, expressed or implied, with respect to the material
contained hetein. Before prescribing any drug, the reader must review the manufacturer's current product information
(package inserts) for accepted indications, absolute dosage recommendations, and other information pertinent to the safe
and effective use of the product described.
Library of Congress Cataloging-in-Publication Data
Electrodiagnostic medicine / edited by Daniel Dumitru, Machiel J. Zwarts, Anthony A.

Amato.—2nd ed.
p. ; cm.
Includes bibliographical references and index.
ISBN 1-56053-433-8 (alk. paper)
1. Electrodiagnosis. I. Dumitru, Daniel. II. Zwarts, Machiel J., 1953- III. Amato,
Anthony A., 1960-
[DNLM: 1. Electrodiagnosis—methods. 2. Nervous System Physiology. 3.
Neuromuscular Diseases—diagnosis. WB 141 E374 2001]
RC77.D87 2001
616.07'547—dc21
2001026407
Electrodiagnostic Medicine, 2nd edition ISBN 1-56053-433-8
© 2002 by Hanley & Belfus, Inc. All rights reserved. No part of this book may be reproduced, reused, republished, or
transmitted in any form, or stored in a data base or retrieval system, without written permission of the publisher.
Last digit is the print number: 9 8 7 6 5 4 3 2 1

Acknowledgments
I am indebted to the support provided by my chairman, Nicolas E. Walsh M.D., and the tireless
efforts of my secretary, Sharon Stowe.
DD
I am extremely grateful to my mentors and good friends, Jerry Mendell, M.D., John Kissel,
M.D., Zarife Sahenk, M.D., and Richard Barohn M.D., who taught me the art of evaluating and
treating patients with neuromuscular disorders.
AAA
I owe gratitude to my colleagues, and the technicians and secretaries of the department of
Clinical Neurophysiology and the physics-technical group for their support and inspiration. The
cooperation of the neurologists and other members of the Neuromuscular Centre Nijmegen is
greatly appreciated.
MZ
Dedication
To my wife, Cyra Sweet Dumitru, and children,
Cyra Alexandra Dumitru and Daniel Amadeus Dumitru.
DD
To my wife, Mary, and children, Joseph, Erin, Michael,

and Katie.
AAA
To my wife, Anita, and our children, Jelte, Iris, and Irene.

MZ
Someday When I A m an Old Woman The Mysterious Cats

and My Heart Is a Movie Theatre of Memories,
Let Me Replay This One A cat is a cat.
It prowls through the night
You never know when happiness might rise, mysterious and black.
an outburst of quiet fire catching beneath Two big yellow eyes
skin. Sometimes all it takes is a town staring into the moonlight.
called Comfort, a small park with a white Then his sister comes along.
gazebo and circular path just right She meows to him,
for the children to ride bikes, now he meows to her.
that it's almost spring. A bit breezy Two cats now sit beneath
but warm enough. Green knots glimmer the golden moon.
in certain trees, three stone churches
stand side by side, a see-saw beckons as a launchpad Cyra Alexandra Dumitru
that all four of us finally astride:
everything comes together and we ride the rise.
Each moment flows from the fingers

of the one before: a soft creek of joy.
The children pedal down church lane, Flying People
remember to slow at the intersection
and look, sun glinting off their helmets. I wish I were a bird.
Your husband easy in the drive, uncertainty I wish I had wings
of what stopping place we'll find. Then there it is. to fly all the way
Gazebo and steeples. You park, to the Statue of Liberty
open the back of the station wagon, and back.
pull out the bikes, feel sunlight dusting And I wish to go to Australia
uncovered arms, and blue sky falling to see the big-footed kangaroo
forever around you— and speed at 2000 miles an hour.
Cyra S. Dumitru Daniel Amadeus Dumitru

Contents
Part 1 F U N D A M E N T A L PRINCIPLES
1 N e r v e and M u s c l e A n a t o m y and Physiology 3
Daniel Dumitru, M.D., Ph.D., and Andrew J. Gitter, M.D.
2 Electrical Sources and Volume C o n d u c t i o n 27

Daniel Dumitru, M.D., Ph.D., Dick F. Stegeman, Ph.D.,
and Machiel Zwarts, M.D., Ph.D.
Appendix T h e Leading/Trailing Dipole M o d e l a n d Near-Field/Far-Field 54

Waveforms
Daniel Dumitru, M.D., Ph.D., Dick F. Stegeman, Ph.D.,
and Machiel Zwarts, M.D., Ph.D.
3 Instrumentation 69
Daniel Dumitru, M.D., Ph.D., and Machiel J. Zwarts, M.D., Ph.D.
Appendix Basic Electricity Primer 98

John C. King, B.S.E.E., M.D.
4 Peripheral N e r v o u s S y s t e m ' s Reaction to Injury 115

Daniel Dumitru, M.D., Ph.D., Machiel J. Zwarts, M.D., Ph.D.,
and Anthony A. Amato, M.D.
Part II BASIC A N D A D V A N C E D T E C H N I Q U E S
5 Nerve Conduction Studies 159

Daniel Dumitru, M.D., Ph.D., Anthony A. Amato, M.D., and
Machiel Zwarts, M.D., Ph.D.
6 Special N e r v e Conduction Techniques 225

7 Needle Electromyography 257

8 Quantitative E M G 293
Sanjeev D. Nandedkar, Ph.D., Erik V Stdlberg, M.D., Ph.D,
and Donald Sanders, M.D.
9 S o m a t o s e n s o r y Evoked Potentials 357

Daniel Dumitru, M.D., Ph.D., Lawrence R. Robinson, M.D.,
and Machiel J. Zwarts, M.D., Ph.D.
vii
viii — CONTENTS
10 Magnetic Stimulation of the Central and Peripheral Nervous Systems . . 415

Lawrence R. Robinson, M.D
11 Quantitative Sensory Testing: Basic Principles and Clinical Applications 429

Gill Wolfe, M.D.
12 Intraoperative N e u r o p h y s i o l o g y M o n i t o r i n g 439
John C. King, B.S.E.E., M.D., Jaime R. Lopez, M.D.,
and Tod B. Sloan, M.D., Ph.D
13 C h e m i c a l Denervation 479
Joyce R. Grissom, M.D.
Part III P A T I E N T CARE-RELATED ISSUES
14 T h e Electrodiagnostic M e d i c i n e Consultation: Approach and Report .. 515

Generation
15 Electrodiagnostic M e d i c i n e Pitfalls 541

Part IV C L I N I C A L APPLICATIONS
16 Disorders Affecting M o t o r Neurons 581

Daniel Dumitru, M.D., Ph.D., and Anthony A. Amato, M.D.
17 F o c a l Cranial Neuropathies 653

18 Radiculopathies 713
19 Brachial Plexopathies and Proximal M o n o n e u r o p a t h i e s 777

20 L u m b o s a c r a l Plexopathies and Proximal M o n o n e u r o p a t h i e s 837

21 A p p r o a c h to Peripheral Neuropathy 885

Anthony A. Amato, M.D., and Daniel Dumitru, M.D., Ph.D.
22 Hereditary Neuropathies 899

23 Acquired Neuropathies 937

24 Focal Peripheral Neuropathies 1043

25 N e u r o m u s c u l a r Junction Disorders 1127

CONTENTS — ix
26 Introduction to M y o p a t h i e s and M u s c l e Tissue's Reaction to Injury . . . 1229

27 Hereditary Myopathies 1265

28 Acquired M y o p a t h i e s 1371
29 Electrodiagnostic Medicine Evaluation of Children 1433

Maureen R. Nelson, M.D.
30 A A E M Glossary of Terms in Electrodiagnostic Medicine 1449
and Illustrations of Selected Waveforms
Index 1489
C D - R O M : Needle Electromyography Inside back c o v e r

Contributors
A n d r e w Gitter, M . D .
Associate Professor, Department of Rehabilitation Medicine, University of Texas Health Science Center,
San Antonio, Texas
Joyce Grissom, M . D .
Chief of Movement Disorders and Chemodenervation Clinic, Wilford Medical Center, San Antonio, Texas
J o h n C. K i n g , M . D .
Associate Professor, Department of Rehabilitation Medicine, University of Texas Health Science Center,
San Antonio, Texas
J a i m e R. L o p e z , M . D .
Assistant Professor, Department of Neurology and Neurological Sciences, Stanford University School of
Medicine, Stanford University Medical Center, Stanford, California
Sanjeev D . N a n d e d k a r , P h . D .
Clinical Applications Manager, Oxford Instruments, Hawthorne, New York
Maureen Nelson, M . D .
Associate Professor, Department of Physical Medicine and Rehabilitation, Baylor College of Medicine;
Chief, Physical Medicine and Rehabilitation, Texas Children's Hospital, Houston, Texas
L a w r e n c e R. R o b i n s o n , M . D .
Professor and Chair, Department of Rehabilitation Medicine, University of Washington School of
Medicine, Harborview Medical Center, Seattle, Washington
D o n a l d B. S a n d e r s , M . D
Professor, Department of Neurology, Duke University Medical Center, Durham, North Carolina
T o d B . Sloan, M . D . , P h . D .
Professor, Department of Anesthesiology, University of Texas Health Science Center, San Antonio, Texas
Erik V. Stalberg, M . D . , P h . D .
Professor and Chairman, Department of Clinical Neurophysiology, Uppsala University Hospital, Uppsala,
Sweden
D i c k F. S t e g e m a n , P h . D .
Professor, Department of Neurophysiology, Institute of Neurology, University Medical Center Nijmegen,
Nijmegen, The Netherlands
Gil I. Wolfe, M . D .
Associate Professor, Department of Neurology, University of Texas Southwestern Medical Center, Parkland
Hospital, Zale Lipshy University Hospital, Texas Scottish Rite Hospital for Children, Dallas, Texas
x
Preface
W e l c o m e to the second edition of Electrodiagnostic Medicine. I am gratified that the first e d i -
tion was very well received by the electrodiagnostic medicine community w o r l d w i d e . Every effort
h a s been m a d e for the second edition to continue in the s a m e tradition of excellence.
Perhaps one of the m o s t significant c h a n g e s in the second edition is the addition of an interac-
tive cross-platform ( P C or Mac) C D R O M . T h e C D contains the Needle E l e c t r o m y o g r a p h y c h a p t e r
found in this textbook with links to specific waveforms described in the text. After reading a b o u t
t h e waveform, o n e can simply click on the linked waveform to play it. It is also possible to p l a y
e a c h of the w a v e f o r m s in any s e q u e n c e i n d e p e n d e n t of the text. T h e w a v e f o r m s are needle e l e c -
tromyographic recordings from patients e x a m i n e d by the editors with a narrated description of t h e
w a v e f o r m s ' p e r t i n e n t e l e c t r o p h y s i o l o g i c findings. E s s e n t i a l l y any w a v e f o r m , n o r m a l or p a t h o -
l o g i c , likely to be e n c o u n t e r e d in clinical p r a c t i c e is r e p r e s e n t e d , in a d d i t i o n to several artifact
recordings. If the practitioner has a PC-based electrodiagnostic instrument, it is possible to d o w n -
load the C D onto the instrument and c o m p a r e those waveforms recorded from o n e ' s patients w i t h
those on the C D .
A n o t h e r major addition to the t e x t b o o k is a m o r e detailed a s s e s s m e n t of the individual d i s o r -
d e r s ' histopathology, molecular biology, and genetics. T h e m o s t current information available r e -
g a r d i n g the m u l t i t u d e of d i s e a s e s d e s c r i b e d within t h e b o o k is included r e p l e t e with n u m e r o u s
references. As in the first edition, an e m p h a s i s is placed on corresponding figures and tables to a i d
in the understanding of disease pathophysiology. Every chapter in the text w a s continually u p d a t e d
u p to the point of publication.
In this edition, t w o additional editors h a v e j o i n e d m e to b r i n g a more c o m p l e t e perspective t o
electrodiagnostic medicine. Anthony A. A m a t o , M . D . , contributed his expertise in n e u r o m u s c u l a r
disorders to ensure that the genetic and molecular biology aspects of the diseases discussed are a c -
curate and up to date. Machiel Zwarts, M . D . , Ph.D., ensured that the information contained in t h e
text w a s c o m p r e h e n s i v e l y a d d r e s s e d from a clinical n e u r o p h y s i o l o g y s t a n d p o i n t and a d d e d a
E u r o p e a n p e r s p e c t i v e on the e l e c t r o p h y s i o l o g i c d i a g n o s i s of n e u r o m u s c u l a r d i s o r d e r s . F u r t h e r ,
several new c h a p t e r a u t h o r s have been a d d e d to e n s u r e that a broad a p p r o a c h to n e u r o m u s c u l a r
disease is presented to the practitioner.
C o n s i d e r a b l e effort has been m a d e to e n s u r e that the textbook, while b r o a d in scope, r e m a i n s
practical and r e a d a b l e . A l t h o u g h the b o o k has clearly d e v e l o p e d into a major reference text, it is
still possible to read it from cover to cover. This is due in large part to significant editing to e n s u r e
a uniform writing style irrespective of multiple authors. A s with the first edition, an emphasis r e -
m a i n s on the c o m p l e t e description of a disorder in plain l a n g u a g e supported by figures and infor-
m a t i v e tables. It is m y c o n t e n t i o n that the patient c a n n o t truly be served u n l e s s the p r a c t i t i o n e r
acquires a thorough understanding of a disorder. T h e second edition of Electrodiagnostic Medicine
continues in the tradition of the first edition in providing the practitioner with the m o s t c o m p r e h e n -
sive information and literature citations available in the field.
Daniel Dumitru, M D , P h D
xi
PART
I
FUNDAMENTAL
PRINCIPLES
Chapter I
N e r v e a n d Muscle A n a t o m y
and Physiology
Daniel Dumitru, M.D., Ph.D.
Andrew J. Gitter, M.D.
CHAPTER OUTLINE
Nervous Tissue Neuromuscular Transmission

Plasma Membrane • Transport Through the Cell Electrochemical Conduction • Acetylcholine Recycling
Membrane
Muscle Tissue
The Membrane Potential
Membrane Potential Generation • Nerve Cells • Action Muscular Components • Electrical Activity
Potential • Action Potential Propagation
Conclusion
The practice of electrodiagnostic medicine requires knowl- substance has a (1) polar or hydrophilic ("water-loving") aspect
edge of nerve and muscle physiology. During the course of an and a (2) nonpolar or hydrophobic ("water-fearing") part. Most
electrodiagnostic medicine examination, the practitioner inves- plasma membrane phospholipids have a polar head group attached
tigates either directly or indirectly nerve and muscle action po- to two nonpolar tail moieties (Fig. 1-1). If amphipathic lipid mole-
tentials. This chapter reviews the basic principles of cell cules are placed in a water solution, they spontaneously coalesce
membranes, membrane potentials, action potential generation so that the nonpolar tail groups form an interior compartment ex-
and propagation, nerve-muscle interaction, and muscle physiol- cluding water, thereby exposing the polar head groups to the water
ogy. Understanding normal function allows greater appreciation environment. Two self-sealing lipid arrangements may occur.
43
and insight into the pathophysiologic processes and electrodiag- The first is a spherical micelle, which forms when the lipid tails
nostic manifestations of nerve and muscle disease. congregate within a sphere and the polar groups face outward. The
second spontaneous lipid aggregate is a bilayer.
vv
Lipid Bilaver. The outer cell membrane is a lipid bilayer, a
NERVOUS TISSUE dual sheet of phospholipid molecules aligned so that the hy-
drophobic tails contact each other forming a nonpolar region
PLASMA MEMBRANE free of water. The polar head groups are exposed to the sur-
rounding aqueous environment (Fig. 1-1B). The usual width of
The plasma membrane forms the outer boundaries of the cell this structure in most cells approaches 75 angstroms (1 A =
thus defining intracellular aêxtracellular contents. It serves to: 1CH meters {m}) and is known as a unit membrane. T h e
0 62
(1) maintain the cell's iruegntv. (2) produce and sustain ion con- lipid bilayer possesses a high degree of fluidity and may be
centration gradients and electric charge differences arrngg t h P
thought of as a two-dimensional liquid in which the lipid mole-
cell, (3) control nutrition intake and waste disposal and (4) assist cules have constrained mobility characteristics. First, a lipid
in the c.eM\ interaction with its environment All plasma mem- molecule can rotate about its long axis very rapidly (Fig. 1-1C).
branes share three basic structural molecules: lipids, proteins, Secondly, the lipid molecule can diffuse in a lateral direction. A
and carbohydrates. Lipid and protein each constitute approxi- third, and rare occurrence, is for a lipid molecule to "flip-flop"
mately 45-49% with carbohydrate accounting for the remaining from one side of the bilayer to the other (Fig. 1-1C). Inter-
15
2-10% of the cell membrane's molecular composition. 13

spersed within the bilayer are cholesterol molecules that are be-
There are three different types of lipids forming the plasma lieved to add stability to the bilayer. Glycolipids exist only on
membrane: phospholipids, cholesterol, and glycolipids. These 62
the outer surface of the bilayer covalently bound to carbohydrate;
three lipid molecules are a m p h i p a t h i c .
1323
An amphipathic their true function is not fully understood. Cholesterol, present
3
4 — PART I FUNDAMENTAL PRINCIPLES
Phospholipid Fluid mosaic structure passing through the hydrophobic region. Two types of trans-
membrane or transport proteins are used to facilitate move-
ment of these substances through the plasma membrane. - 3 49
glycerol Channel proteins extend across the lipid bilayer and have a
water-filled central tunnel or pore that allows the passage of spe-
cific ions or molecules (Fig. 1-2). A second type of transport pro-
tein, carrier protein, binds the substance to be transported and
undergoes a conformational change to effect transmembrane
crossing. These protein transport mechanisms are selective for the
particular ion or molecule to be delivered across the membrane.
Lipid mobility Substances can pass through the cell membrane by eitheriaif-
Iranslational movement flip-flop movement
juUSjfcmôPto the movement of in-
dividual molecules driven by thejJiejTriaiand kinetic energy of
matter's random movement, either directly through the inter-
molecular spaces of the plasma membrane's. J ipid portion, or in
conjunction with a channel or carrier protew4^^|Jr^njn)OTt
requires an energy source other than kinetic energy to move a
Figure I-I. The Plasma Membrane. A, Molecular model of a
substance across the plasma membrane "uphill" against an
phospholipid molecule (phosphatidylcholine) consisting of a hy-
eêjr^JXidieiit, e.g., from low to high ion concentration. 22
drophilic "head" group (choline and glycerol) and a hydrophobic "tail"

portion (fatty acid). Phosphatidylcholine is only one of many possible iffusion
phospholipids. B, Representation of t h e fluid mosaic model of t h e
An ion or molecule may cross the cellular membrane by
plasma m e m b r a n e . A lipid bilayer is pictured with m e m b r a n o u s and
either simple diffusion or facilitated diffusion (Fig. 1-2). In
t r a n s m e m b r a n o u s proteins. C T r a n s l a t i o n a l and flip-flop degrees of
simple diffusion a substance crosses the plasma membrane rely-
freedom available t o t h e phospholipid molecule in the bilayer. (From
ing solely upon thermal-induced random kinetic motion without
Barchi RL: Excitation and conduction in nerve. In Sumner AJ (ed):The
the need for carrier proteins. The rate of simple diffusion de-
Physiology of Peripheral Nerve Disease. Philadelphia, W.B. Saunders,
pends on the (1) kinetic energy present in the system, (2)
1980, pp I - 4 0 , with permission.)
number of openings (channel proteins) in the membrane avai 1 -
able to the ion or molecule, and (3) hjBy^mucJL^
within the bilayer increases the plasma membrane's stability as is present, i.e., the concentration gradient across the membrane"
well as reduces permeability to small water-soluble molecules. , for a particular substance. Facilitated diffusion, however, re-
22
quires that the molecule or ion first bind a carrier protein prior
TRANSPORT THROUGH THE CELL MEMBRANE to passing through the membrane.
pie Diffusion. Simple diffusion allows substances to cross
While the lipid bilayer provides structure for the cell mem- through the cell m§ pbrane by eitherfoassing through the lipid bi-
brane, proteins imbedded in, or projecting through, the bilayer layer directly or by Jtis"
carry out the active processes necessary for the cell to function. 62
possible to measure the transport properties of the lipid bilayer by
In order for the cell to function, a wide variety of substances rang- removing the membrane proteins. The primary factor that deter-
ing from simple ions to complex glycoproteins must pass through mines the ease with which a substance moves through the bilayer
the plasma membrane. Lipid-soluble molecules can enter or exit is its lipid solubility. Molecules such as oxygen, nitrogen, and
the cell directly across the cell membrane. The lipid bilayer, how- carbon dioxide are highly lipid-soluble and pass directly through
ever, prevents relatively large water-soluble molecules from the cell membrane without difficulty. Although water is not
highly lipid-soluble, it penetrates die lipid bilayer relatively easily
Carrier because of its small size and high kinetic energy. It is possible,
20
proteins therefore, for small molecules with a low lipid solubility to cross
Channel the cell membrane. As a molecule's size increases, its ability to
J protein diffuse in or out of the cell diminishes markedly (Table 1-1 ). 22
In addition to relatively large molecules, charged substances
mm or ions also have difficulty crossing the cell membrane, even

when small. Small ions such as hydrogen, sodium, and potas-
sium penetrate the lipid bilayer approximately 10 times less
6
readily than water. The reason ions demonstrate great diffi-

22
culty passing through the lipid bilayer is twofold. First, the ion's
Simple inherent charge attracts and binds multiple water molecules pro-
diffusion ducing hydrated ions in aqueous solutions. The size of the hy-
drated ion is significantly larger than the ion itself. The larger
size of a substance decreases its ability to penetrate a mem-
Q? Diffusion ©Âctive Transport
brane. Secondly, the electric charge of the ion interacts with the
Figure \-2. Transport through the cell membrane. Simple lipid charge of the polar portion of the lipid bilayer, reducing its abil-
bilayer with several t r a n s m e m b r a n e proteins capable of performing ity to cross the membrane. As a result, if these ions are to cross
either diffusion o r active transport. (From Guyton AC:Textbook of the membrane, some type of protein channel is needed. Large
Medical Physiology, 9th ed. Philadelphia,W.B. Saunders, I996,pp 4 3 - 5 5 , nonpolar substances also require protein channel assistance, but
with permission.) for our purposes only ions will be discussed.
Chapter I NERVE A N D MUSCLE ANATOMY A N D PHYSIOLOGY — 5
Table I -1. Relationship of Effective Diameters of Different Transported

Substances t o Their Lipid Bilayer Permeabilities 22
molecule
Substance Diameter (nm) Relative Permeability \
Binding
W a t e r molecule 0.30 1.0 f / point
Urea molecule 0.36 0.0006

Hydrated chloride ion 0.386 0.00000001
Hydrated potassium ion 0.396 0.0000000006 Carrier protein
and
Hydrated sodium ion 0.512 0.0000000002
conformational
Glycerol 0.62 0.0006 change
Glucose 0.86 0.000009
Protein Channels. Protein ion channels are water-filled pro-

teiiTtunme1 sTITaTspan the cell membrane phospholipid bilayer Release
and control ion movement across the membrane. Two key char- of binding
acteristics of ion channels determine their function: (1) selective
permeabilit^t^^ ion and (2) channel opening and clos-
Figure 1-3. Facilitated diffusion. A possible mechanism through
m
jLlJ3£QH&!2^^ The majority of protein which a carrier protein must undergo a conformation change t o a c -
channels only allow specific ions to pass. This selectivity is con- complish t h e process of facilitated diffusion. (From Guyton A C :
trolled by incompletely understood molecular features of the Textbook of Medical Physiology, 9th ed. Philadelphia, W.B. Saunders,
channel but appears to be related to an interaction between the 1996, pp 43-55, with permission.)
ion's electric charge and size, and the associated channel charge
and morphology. Ion channels exist in either an open or closed
state that is controlled by a "gate." The "gate" is believed to be a concentration. This situation requires a process other than diffu-
portion of the protein channel macromolecule that is capable of sion because diffusion permits unequal concentrations of vari-
movement on a molecular scale. Under special circumstances, ous substances across a membrane to eventually equilibrate.
the protein channel undergoes a conformational change moving Cells have developed a unique mechanism, active transport, to
the gate away from the channel tunnel. Control of the opening accumulate substances in the direction opposite to their concen-
or closing of the protein channel can occur through ligand tration gradients. The process of active transport depends on a
^^^SKJU^^^J^ÎM^\^ ligand gating a molecule, i.e., a transmembrane carrier protein that uses energy in the form of
neurotransmitter binds the channel protein resulting in a confor- adenosine triphosphate (ATP). The ATP causes the carrier pro-
mational change that opens the gate. An example of this kind of tein to undergo a conformational change permitting ions to b e
gate is the acetylcholine receptor at the neuromuscular junction "pumped uphill" against their concentration gradient.
that opens in response to the binding of two acetylcholine mole-
cules. In voltage-gated channels such as the sodium or potas-
sium channel, changes in the transmembrane voltage initiate the
conformational change and open the gate. When the internal
voltage of the cell is at the resting cell membrane potential, the All living cells have an electrical potential across their mem-
voltage-gated sodium and potassium protein channels are branes. Using microelectrode techniques, the intracellular p o -
34
closed. In some excitable tissues such as nerve axons and tential or voltage can be measured with respect to the
muscle fibers, as the cell's internal voltage becomes less nega- extracellular space. Depending on the species and particular cell
tive, the sodium gates open and are followed somewhat later by type examined, the voltage difference across the cell membrane
potassium gates, permitting the respective ions to move freely is approximately 60-90 millivolts (mV), with the inside of the
and an action potential to be generated. This concept will be dis- cell negative. This voltage or potential difference across the
37
cussed in detail in subsequent sections. cell membrane is known as the resting membrane potential or
Facilitated Diffusion. Facilitated diffusion also depends tx^WI^^J^^^^S?^^- During the process of action poten-
upojLa^transmembrane protein to facilitate the passage of specific tial generation, the transmembrane potential changes. To avoid
substances. UaMJife^^^^dyffcion, facilitated diffusion requires ambiguity in describing changes in the transmembrane poten-
that a substance first binds a protein carrier, which then under- tial, the terms depolarization and hyperpolarization are used.
goes a conformational change transporting the substance through Membrane ( k g o l i u ^ ^
the membrane (Fig. 1-3). This binding and conformational alter- tial becomes less negative, e.g., when the normal intracellular
ation takes some time before a second molecule can be carried. potential of - 7 5 mV changes to - 5 0 mV or +20 mV. Membrane
The speed of transport is rate-limited and dependent upon the to^lU^telllilteP- means the intracellular potential becomes
conformational change of the carrier protein. Simple diffusion, more negative, e.g., when the resting membrane potential
however, is not rate-limited and depends primarily upon the con- changes from - 7 5 mV to -100 mV.
centration gradient of the substance crossing the membrane.
MEMBRANE POTENTIAL GENERATION
Active Transport
The cell membrane is capable of maintaining a high trans- The resting transmembrane potential is generated by an un-
membrane concentration gradient, e.g., a high intracellular con- equal distribution of ions across the cell membrane. Generating
centration of a substance that exists in a low extracellular and maintaining a steady resting transmembrane potential involves
0 0 there is no movement of CI" ions because the membrane is not

permeable to this ion species. With the selective movement of
K cations out of the region of high concentration, there is a cor-
+
responding loss of positive charges. This creates an electrical

voltage difference between the two sides of the beaker.
Relatively more negatively charged CI" ions exist on the high
concentration side of the barrier because of the loss of the posi-
tive K ions. This generates an electrical force that opposes fur-
+
ther K ion loss (negative Cl~ charges attract positive K

+ +
charges). Eventually equilibrium is reached when the move-

ment of potassium "down" its concentration gradient is bal-
anced by the electrical force generated from the unequal electric
charge separation. In equilibrium, individual K ions may con- +
tinue to move across the barrier, but any movement across the
barrier from the concentration difference is matched by move-
ment in the opposite direction from electrical forces. A volt-
meter placed across the two solutions now measures a voltage
difference with the original higher-concentration side of potas-
sium chloride having a negative potential relative to the lower-
concentration solution.
A B Equilibrium Potential. Similar to the beaker example, glial
cells are believed to have only potassium channels in their cell
Figure 1-4. Membrane potential generation. A, A beaker con-
membrane with a high intracellular concentration of K and or- +
taining t w o solutions of potassium chloride (K CI") with a greater

+
ganic anions. In the resting state, K flows into the extracellu-

39 +
concentration (100 millimolar) on the right half compared t o the left

lar space (region of low concentration) until the unbalanced
half (50 millimolar).The partition separating t h e t w o solutions does
intracellular anions and accumulating extracellular positive
not allow any mixing of t h e two solutions (impermeable). A voltmeter
potassium ions impede further net K efflux. The chemical con- +
does n o t measure a potential difference. B.The partition is replaced

centration gradient favoring K efflux is then balanced by the +
with a divider that has many K pores only (semipermeable). A volt-

+
difference in electrical potential opposing K efflux. The mag- +
meter now records a potential difference with the higher concentra-

nitude of the transmembrane voltage in this balanced state is the
tion portion being negative compared t o t h e less concentrated side.
equilibrium potential for potassium.
In 1888, the physical chemist Walter Nernst used basic ther- 48
ion diffusion effects, electrostatic forces, and ion transport modynamic principles to describe the transmembrane work ex-
pumps. To understand the development of the membrane poten- erted on cellular ionic species. In short, the electrical work
tial it is useful to start by considering the equilibrium state that ( W ) required to move a fixed quantity (e.g., 1 mole) of an ion
elec
occurs as an ion distributes itself across a membrane. (I) against an electrical potential difference across the membrane
A beaker containing two aqueous solutions of potassium (E ) is equal and opposite (hence the negative sign in the equa-
m
chloride of different concentrations separated by an imperme- tion below) to the work of the concentration gradient ( W ) driv- conc
able barrier may be used to explain some aspects of the cell's ing the ion from a high to low concentration: W =- W elec c o n c
resting membrane potential (Fig. 1-4). In an aqueous environ- The work required to create the electrical potential depends
ment, a potassium chloride solution exists as potassium (K ) +
on the ion's charge or valence (Z), the number of charges sepa-
and chloride (CI ) ions. Positive ions are called cations and neg-
-
rated (expressed by Faraday's constant (F), which is the number
ative ions are known as anions. The individual cations and of coulombs for each mole of charge), and the magnitude of the
anions are in continuous random motion on both sides of the electrical potential ( E ) . Electrical work can be expressed as:
m
barrier due to inherent thermal energy. The randomly moving W = Zj F E . The work created by the unequal ion concentra-
elec m
ions collide with the barrier dividing the beaker, with more col- tion gradient tending to move the ion across the membrane de-
lisions occurring on the side with higher ionic concentration pends on the logarithmic concentration difference of the ion in
simply because there are more ions per unit volume. Because the intracellular ([I]j and extracellular ([I] ) spaces and the tem- e
the barrier is initially impermeable, no flow or redistribution of perature (T) that affects the random movement of ions. The con-
ions can occur and the number of cations and anions on each centration work is expressed as:
side of the barrier remains balanced. Using a voltmeter, no volt-
age or potential difference exists within the beaker (Fig. 1-4A). W c o n c = -RT(ln[I] -ln[I] ) i e
If the barrier is altered by adding potassium channels, move- where R is the universal gas constant. These values may be sub-
ment of K ions across the membrane is now possible (Fig. 1-
+
stituted into the equation, W = - W , to arrive at a formula
e l e c c o n c
4B). The barrier is considered semipermeable to K . In this +

that relates the electrical and chemical transmembrane forces
environment, ions tend to move from the high-concentration to exerted upon an ion:
low-concentration side of the barrier down the K concentration
+
gradient. Because more K ions exist in the concentrated solu-

+ Z FE = -RT(ln[I] -ln[I] )
i m i e
tion, there is a greater likelihood of random encounters with the This equation when solved for E is referred to as the Nernst m
potassium channels compared to the less concentrated side of equation and may be rewritten as:
the beaker. As a result, more potassium ions from the concen-
trated solution flow into the less concentrated solution, redis-
tributing K ions down their concentration gradient. In contrast,
+
E = _ (In—j
m
As an example, consider the glial cell described previously. Table 1-2. Cellular Ionic Concentrations and Equilibrium
For an ion such as K with a charge of 1 , the valance value Z =
+ +
Potentials
1 . At a temp of 20°C, the quantity RT/ZF reduces to 26 milli-
+
Intracellular Extracellular Equilibrium

volts (mV). In vivo, the glial cell exists in an environment with (mmol/L) (mmol/L) Potential (mV)
intracellular K concentration approximately 20 times greater
+
than the extracellular K concentration. Under these conditions,

+ Squid Axon
the Nernst equation would predict a transmembrane equilibrium Na +
50 440 +55
potential of: K +
400 20 -76
ci- 52 560 -66
-(26 mV) 76 mV Organic anions 385
Mammalian Axon
In this example, maintenance of the equilibrium potential de- Na +
10 145 +56
pends only upon the passive electrical and diffusion forces and K +
160 4 -102
does not require additional energy other than thermal energy a- 3 1 14 -76
driving diffusion. This can be appreciated by considering an ex- Organic anions 163 34
periment in which a microelectrode is inserted into the glial cell
so that it does not disrupt the plasma membrane or other cellular
mechanisms. Suppose a small positive current is injected into the cell membrane at equilibrium is much more permeable to
the cell partially depolarizing it, i.e., the intracellular space be- K than N a with an intermediate permeability to CI".
+ + 30
comes less negative than the resting potential. The negative The consequences of having multiple ions contribute to the
electrical force preventing potassium ion efflux is reduced and equilibrium potential is more complex but the same basic prin-
potassium ions diffuse out of the cell toward the low K concen- +
ciples outlined by the Nerst equation apply. As previously
30,31
tration extracellular space. Potassium leaves the cell until described for K ions, each ion species has its own unique equi-
+
enough positive potassium cations are lost to balance the posi- librium potential that depends on that ion's concentration inside
tive charge injected and return the transmembrane potential to and outside the cell. A nerve axon's equilibrium potential is pro-
- 7 6 mV. Now, suppose the cell is hyperpolarized (i.e. the intra- portional to not only K , but also N a and CI".
+ +
cellular space becomes more negative than the resting potential If the cell membrane were permeable only to N a ions (this +
of - 76mV) by injecting negative charges into the cell. The in- can be accomplished by blocking other ion channels with neu-
crease in electrical negativity inside the cell creates an electrical rotoxins) a resting equilibrium potential of + 55 mV would be
force that attracts potassium ions from the extracellular space. generated. This reflects the diffusion and electrical forces acting
Potassium ions diffuse into the cell through the ion channels on the sodium ion in a manner analogous to the potassium ion.
until transmembrane potential is once again restored to the equi- Because of the steep concentration gradient of N a from a +
librium level of - 76mV. The equilibrium potential is the electri-

cal potential (voltage) across the cell membrane at which the
force driving potassium ions from a high intracellular to low ex-
tracellular concentration is just balanced by an inward electrical
force acting on the positive potassium ion.
NERVE CELLS
Using the Nernst equation to predict the resting membrane
potential of nerve cells, Julius Bernstein in 1902 suggested a
nerve's resting membrane potential depended upon the selec-
tive permeability of K . This prediction would require that the
+ 8
transmembrane potential of neural tissue varies directly with

changes in the intra/extracellular concentration of K . For glial
+
cells, the predicted and measured transmembrane potentials as

K concentrations are varied agree quite well. However, this
+
effect was not observed with other nerve cells and axons. At
high extracellular concentrations of K , the predicted and ob-
+
served transmembrane potentials match. At low extracellular 2 5 10 20 50 100 2 0 0

K concentrations, the measured transmembrane potential de-
+
External Potassium Cone. (mM)

viates from the predicted values (Fig. 1-5). This observation
implied that the axon membrane is permeable to ions other than Figure J-5. Membrane dependence on potassium ion.
potassium. Demonstration of t h e relationship between t h e membrane potential
Further studies, in which a giant squid axon was bathed in an of a nerve and t h e external K concentration.The dashed line is t h e
+
ion solution containing radioactively labeled Na , K , and Cl~,

+ +
predicted m e m b r a n e potential using t h e N e r n s t equation, while t h e
demonstrated that all three ions enter the intracellular space. 13 3 9
solid line is the actual experimentally recorded membrane potential.
Because charged ions do not diffuse through the lipid bilayer of N o t e how the t w o lines deviate at low K concentrations, suggesting
+
the membrane, the cell membrane must contain protein ion o t h e r ions contribute t o t h e membrane potential in addition t o K*.
channels for these three ions. The intra/extracellular concentra- (From Barchi RL: Excitation and conduction in nerve. In Sumner AJ
tions of the three ions and their associated equilibrium poten- (ed):The Physiology of Peripheral N e r v e Disease. Philadelphia, W.B.
tials are quite different (Table 1-2). Studies have revealed that Saunders, 1980, pp I-40, with permission.)
higher extracellular to a low intracellular concentration, diffu- phosphate bonds of ATP to maintain the ion concentration gra-
sion forces drive N a into the cell (Table 1-2). As the N a
+ +
dients across the plasma membrane. The Na - K -ATP pump
+ +
cations diffuse into the cell and accumulate, the inside of the imports 2 potassium ions for every 3 sodium ions it exports to
cell becomes positively charged, repelling further N a entry. +
the extracellular region. Due to the unequal transport of ions
The competing effects of electrical and diffusion forces on N a +
across the cell membrane (more positive charges exported than
ion flow are balanced at the transmembrane potential of + 55 imported), the pump generates a small electrogenic offset that
mV. The opposite polarity of the transmembrane potential for lowers the resting membrane potential several millivolts below
N a compared to K reflects the opposite concentration gradi-
+ +
that predicted by the Goldman-Hodgkin-Katz equation.
ents for these ions. Unlike the K and Na ions, the chloride anion is not subject
+ +
Typically nerve cells have a resting potential that is approxi- to active transport, but is free to be distributed by purely passive
mately 6 0 - 7 0 mV negative intracellularly, placing the actual forces. Chloride anions are by default in equilibrium with re-
membrane potential in between the individual ion membrane spect to the axon's membrane. The actual transmembrane con-
potentials. At this resting membrane potential, the potassium centrations of Cl" are determined by the resting membrane
ion is close to its equilibrium voltage, while the sodium ion is potential and are adjusted accordingly. As a result, N a and K+ +
more than 100 mV from achieving its equilibrium potential. As are actively distributed, whereas Cl" is passively distributed.
a result, there remains a strong electrical gradient, in addition to
the concentration gradient, that favors N a movement into the ACTION POTENTIAL
+
cell. The N a that diffuses into the cell under these conditions
+
has the same effect as the injected current in the above-noted One of the fundamental purposes of the nervous system is
experiment, i.e., the membrane potential deviates from the communication, both within an organism as well as between the
potassium equilibrium potential toward the sodium ion's equi- organism and its environment. The nervous system has devel-
librium potential (a less negative value). The loss of some of the oped a relatively simple yet highly efficient means of relaying
negative intracellular potential allows some potassium to exit information and reacting to stimuli by means of electrical im-
the cell. This exiting positive potassium charge acts to balance pulses. These self-sustaining impulses are based on ion perme-
the influx of positive sodium ions. A new equilibrium potential ability shifts and are known as action potentials. In 1849 the
(e.g., - 60 mV) is achieved at which the K efflux equals the +
German physiologist Emil DuBois-Reymond first described the
Na influx. The resting membrane potential is not halfway be-
+
production of action potentials by axons. It has taken approxi-
40
tween the two ions' equilibrium potential because the mem- mately 130 years to begin to appreciate the molecular mecha-
brane is not equally permeable to both ions; it is much more nisms responsible for the generation and propagation of action
permeable to potassium, and therefore closer to the potassium potentials.
equilibrium potential. Permeability depends on the number of Ionic Hypothesis. K. S. Cole and H. J. Curtis in 1938 pro-
channels available for a particular ion. The resting membrane duced action potentials in squid axons while performing intra-
potential determined by the simultaneous effects of K , N a , + +
cellular recordings. The squid axon was used because it was
12
and Cl" ions was first described by Goldman and depends on the large enough to allow the intra-axonal placement of the rela-
ion's permeability (p) and the transmembrane ion concentra- tively large recording electrodes available at the time. These
tions. The relationship dictating the resting membrane potential
is referred to as the Goldman-Hodgkin-Katz equation: 21
INSIDE Membrane OUTSIDE
-RT /P [K ]j-f P [ N a ]
K
+
Na
+
i + P [Cl-3 \
cl e
fcm
F n
\P [K ]
K P [Na ]
+
e N a
+
e PdCl-V
From this equation, it can be seen that the membrane poten-
tial is influenced heavily by the most permeable ion. For nerve
and muscle, the resting membrane potential is closest to K , the +
most permeable ion.

Because the resting membrane potential of the cell is not the
same as either the Na or K equilibrium potentials, a small N a
+ + +
ion influx balanced by K ion efflux is constantly occurring.

+
Without a cellular mechanism to compensate for this, the intra-

cellular pools of sodium and potassium ions would ultimately
become depleted, preventing the cell from maintaining a con-
stant resting potential. To prevent such an occurrence, it is nec-
essary to actively import potassium ions into the cell (against its
concentration gradient) and export sodium ions out of the cell
(also against its concentration gradient). Moving ions "up" their
concentration gradient requires energy. The cellular mechanism
that performs this task is a Na -K -ATP-dependent pump. The + +
Figure 1-6. Sodium-potassium ATP pump. The electrogenic
Na -K -ATP-dependent pump is believed to exist within the
+ +
Na -K -ATP pump located within t h e membrane maintains the resting
+ +
plasma membrane and actively transport Na and K across the + +

membrane potential despite passive diffusions of N a into and K out
+ +
membrane to prevent the dissipation of their concentration gra- of the cell.The steepness of the arrows indicates the magnitude of the
dients (Fig. 1-6). The passive N a and K leakage caused by
60 + +
electrochemical forces driving t h e various ions. (From Barchi RL:
the difference between the resting potential and individual ion Excitation and conduction in nerve. In Sumner AJ (ed):The Physiology
equilibrium potentials is balanced by this active pump. A steady of Peripheral N e r v e Disease. Philadelphia, W.B. Saunders, 1980, pp
state occurs that uses metabolic energy from the high-energy 1-40, with permission.)
investigators found that during the course of an action potential,

-91-
the cell membrane increased its ionic conductance. This finding
established that the action potential is directly dependent upon n > 56 m V D e p o l a r i z a t i o n
E e E
the transmembrane movement of ions. Determining the exact
- 6 5 \~
ion(s) involved in action potential generation began with the
work of Hodgkin and Katz, who noted a reduction in the squid
axon's action potential magnitude if the N a concentration of
+
the solution bathing the axon was reduced. Combining the in-
25
formation of Cole and Curtis with their findings, Hodgkin and

Katz proposed that the action potential is generated by depo-
25
larization of the cell caused by a transient increase in N a con- +
ductance. Using an experimental technique known as the

voltage clamp, Hodgkin and Huxley further elucidated the role
of ion conductance in action potential generation. 13
The voltage clamp is a technique that can be used to exter-

nally stabilize the membrane at a particular voltage, making it
possible to measure the conductance of various ions in the
surrounding solution. Two electrodes are inserted longitudinally
down the long axis of excitable tissue, e.g., a squid axon. One of
the electrodes measures the transmembrane potential of the
axon. The other electrode injects a known quantity of either
positive or negative charges (electric current) into the axon's in- No Na" " in1
terior to maintain the transmembrane voltage at a predetermined external

solution
level (Fig. 1-7). Using feedback amplifier techniques, the
amount of current injected can be regulated to match the current
flow caused by ion movement across the cell membrane. The
end result is the ability to directly record the amount of current
passing through the membrane. Initially, the axon's membrane
was abruptly changed from its resting potential of - 65 mV to
- 9 mV and held there with the voltage clamp (Fig. 1-8A). To
maintain the transmembrane potential at - 9 mV, the voltage
clamp injected a Diphasic current into the cell. During the first Calculated
several milliseconds following the change in membrane poten- difference
between
tial, the voltage clamp needed to inject negative charges into the
B and C
cell. This was immediately followed by the need to remove neg-
ative charges from the cell. From the perspective of the cell
membrane, this behavior of the voltage clamp would indicate
that following a depolarizing shift in the membrane potential
Addition
of TEA
1 2 3 * 5
F i m o (ma)
Figure I -8. Voltage clamp current flows. Current flows across an

axon using a voltage clamp method (see Figure I -6). A, In this example
the axon's membrane is changed from its resting potential of - 65 mV t o
Membrane current
- 9 mV and maintained o r "clamped" at that level. B.The axon produced
an initial inward followed by an outward positive current flow in r e -
Figure 1-7. Voltage clamp. Schematic representation of the volt- sponse t o depolarizing t h e cell membrane. N o t e that the outward cur-
age clamp apparatus used t o determine the ionic flows in a squid axon. rent is delayed and lasts for as long as t h e membrane is clamped. C,
T h e transmembrane voltage is determined between an intracellular Repeating the experiment in a surrounding bath devoid of Na eliminates +
and extracellular electrode (V) and compared t o t h e desired voltage the initial inward current flow, suggesting that the inward current is d u e
set by the e x p e r i m e n t e r (control voltage). Any difference between t o N a . D.The difference between the curves in B and C equals the N a
+ +
t h e s e t w o values is removed by t h e passage of a current (i) through current alone, that is short-lived. E, Blockage of the K channels with +
the second set of electrodes across t h e axons cell membrane. (From tetraethylammonium chloride eliminates the outward current confirming
McComas AJ: Neuromuscular Function and Disorders. Boston, its K nature. (From Darnell J, Lodish H, Baltimore D: Molecular Cell
+
Butterworth, 1977, with permission.) Biology. New York, Scientific American, 1986, with permission.)
channels and investigate the properties of the sodium current in

isolation (Fig. 1-8E). These methods of selective pharmaco-
62
logic inhibition of the voltage-sensitive ion channels combined

with voltage clamping techniques confirmed the original hy-
potheses of Hodgkin and Huxley. - 26 29
Voltage-Gated Channels
Through these and other studies, convincing evidence now
exists that the action potential is mediated by abrupt changes in
ion flow controlled by specific membrane channels for N a and +
K . The opening and closing of these N a and K channels

+ + +
depend on the transmembrane voltage, i.e., the channels are

voltage-gated. Experimental studies using patch-clamp tech-
niques that isolate small patches of the cell membrane surface
have demonstrated that an unmyelinated nerve has approxi-
mately 5-500 N a channels per square micrometer ( p m ) of
+ 2
membrane, which are open for approximately 0.7 ms during the

course of an action potential. - - The channel is either fully
50 51 60
opened or fully closed with no intermediate states of partial

opening. Radioactively labeling tetrodotoxin and applying it to
a nerve membrane has confirmed the density of Na channels. + 58
The voltage-gated N a and K channels are separate from

+ +
each other and different from the passive Na" and K "leak" +
channels that are responsible for the resting steady-state mem-

25 50 brane potential. Although individual voltage-gated channels
DISTANCE (mm) have not been directly visualized, their functional subcompo-
nents have been identified and their role in the process of gener-
Figure 1-9. Action potential. Action potential produced by t h e
ating an action potential has been well established. The
depolarization of a giant squid axon. N o t e t h a t t h e N a channels are
+
potassium ion channel appears to exist in two voltage-depen-

first open for less than I ms (see time scale a t bottom) and produce
dent conformational states, an open or activated state and a
the initial rise of the action potential.The N a channels then close and
+
closed or deactivated state. The physical site of the activation

the delayed opening of the K channels influences the action potential
+
gate remains uncertain, but it is believed to be located within the

and produces its decline back toward t h e resting membrane potential.
ion channel pore near the extracellular opening. Sodium chan-
Because t h e K channels stay open slightly longer than that required
+
nels have a similar activation gate, but display greater electro-

to reach t h e previous resting level, t h e cell is slightly hyperpolarized.
physiologic complexity that indicates they also exist in a third
The local currents (inward flow of only N a ) a r e shown and extend
+
conformational state in which the channel is inactivated and

over a distance close t o 30 mm. (From Hille B: Introduction t o physiol-
not permeable even though the activation gate may be open.
ogy of excitable cells. In Patton HD, Fuchs AF, Hille B, e t al (eds):
This can be accomplished by the inclusion of a separate inacti-
Textbook of Physiology, 21st ed. Philadelphia, W.B. Saunders, 1989, pp
vation gate, typically conceptualized as a ball and chain or
I-80, with permission.)
hinged lid that blocks ion flow by occluding the inner opening
of the channel.
there is a brief initial inward flow of positive ions followed by a Variations in conformation states and the temporal character-
delayed sustained outward flow of positive ions (Fig. 1-8B). istics of channel opening and closing underlie the generation of
If the solution bathing the cell is replaced with one devoid of the action potential. Initially, the nerve is in a resting state with
Na but does contain an impermeable cation such as choline, the
+
a membrane potential of - 90 mV (mammalian nerves). N a and +
inward-directed positive current no longer occurs (Fig. 1-8C). K ions are passively diffusing across the membrane in their 3:2
+
The outward positive current, however, remains unchanged. This ratio balanced by the electrogenic Na -K -ATP pump creating a
+ +
result implies that N a mediates the inward but not the outward
+
resting steady state (individual ions moving across the cell
current. By manipulating ion concentrations and substituting membrane, but no net change in charge). The voltage-gated N a +
ions species in the bathing solutions, Hodgkin and Huxley and K channels are closed (Fig. 1-10). If the transmembrane
+
demonstrated that depolarization triggers an initial short-lived voltage is reduced by 15-20 mV toward a threshold value of
inward N a current and a somewhat delayed and prolonged out-
+
approximately - 65 to - 70 mV, the voltage-gated N a channel +
ward K current (Fig. 1-8C, 1-8D). The action potential, then, is

+
senses the threshold voltage change and rapidly undergoes a
a transient reversal of transmembrane potential producing an ini- conformational change that activates and opens the N a channel +
tial depolarization through an inward-directed Na ion flow, fol-

+
(Fig. 1-10). The membrane permeability to N a is increased by +
lowed by a repolarizing outward-directed K current (Fig. 1-9).

+
a factor of 5,000 times allowing rapid sodium entry into the
Tetrodotoxin, a purified poison from the puffer fish has been cell. Approximately 1 ms later the N a gate becomes inacti-
22 +
employed to confirm the results of Hodgkin and Huxley. 58

vated and the rapid sodium influx is halted. During the interval
Tetrodotoxin specifically binds to the voltage-gated N a chan- +
between activation and inactivation, the N a channel allows for +
nel and blocks it from opening. This blockade allows investiga- a transient flow of ions that shifts the cell membrane toward the
tors to eliminate the sodium current and directly investigate the Na equilibrium potential, depolarizing the membrane to a volt-
+
various parameters of the potassium current. One may also age of approximately + 40 mV. Inactivation occurs through a
apply tetraethylammonium (TEA) to selectively block the K +
gating mechanism that is distinct from the action of the activation
Chapter I NERVE AND MUSCLE ANATOMY AND PHYSIOLOGY — I I
Outside depolarizing current is needed to initiate an action potential.

The relative refractory period lasts several milliseconds and re-
Activation N a + â +
Na +
sults from the need for extra positive ion influx during the hy-
perpolarized state generated by the prolonged K channel +
opening and the associated outward K ion flow. +
The total number of intracellular potassium ions that leave

the cell during the generation of an action potential is small and
only reduces the intracellular concentration by approximately
0.03 to 0.0003 percent. This concentration difference is most
34
likely not experimentally detectable. As the voltage-dependent

{- 90 mV) ( - 90 to + 35 mV) delayed) gates open and close, ion permeability is dramatically altered
for a brief period but only a small number of ions pass through
the membrane relative to the size of the intracellular and extra-
cellular ion pools. Relating this to the Goldman-Hodgkin-Katz
equation, it is the alterations in ion permeability or conductance
that cause the dramatic shifts in the transmembrane potential
rather than any significant ion concentration change (Fig. 1-9).
Molecular Structure of Voltage-Gated Channels

Although the basic functional features of the voltage-gated
K +
K +
N a and K ion channels have been demonstrated by the work

+ +
Resting Slow activation of Hodgkin, Huxley, and others, relating individual channel
Inside (-90mV) (-90to+35mV) functions to their molecular mechanisms has been an ongoing
challenge to researchers. Through the use of neurotoxins to
Figure f-iO. Voltage-gated channels.The properties of the volt- block components of ion channels, gene cloning, and mutagen-
age-gated Na and K protein channels are depicted. Note the posi-
+ + esis studies the major protein subunits of voltage-gated ion
tion of the activation and inactivation gate for the Na channel at the + channels have been identified. • - - Using this information,
1 9 33 55
different membrane voltages.The K channel demonstrates a similar

+ conceptual models, some confirmed experimentally, have
but relatively more simple mechanism. (From Guyton AQTextbook of emerged that attempt to link channel function to specific protein
Medical Physiology, 9th ed. Philadelphia,W.B. Saunders, I996,pp 43-55, locations and to show proposed physical models of how channel
with permission.) conformational changes may create ion selection (filtering) and
gating effects.
The voltage-gated N a and K channels (along with C a ion
+ + ++
gate that initially opened the channel. Consequently, N a ions +

channel) appear to share substantial structural similarity.
are allowed to flow only when the channel is activated but not Voltage-gated ion channels consist of three subunits designated
inactivated. as alpha, b e t a and beta (Fig. 1-1l).'- - The main structural
h 2
9 55
At the same time that the Na channel senses threshold, the

+
and functional component of the channel is the alpha subunit, a
voltage-gated K channel, sensing the same threshold voltage,
+
large macromolecular protein that creates the ion pore and con-
initiates a conformational change that activates and opens the tains the gating mechanisms that controls ion flow. Current un-
potassium gate. The conformational change that activates the derstanding of the alpha subunit suggests it consists of six
potassium gate takes longer then the activation of the Na gate +
alpha-helical proteins (S1-S6) that span the cell phospholipid
and thus appears delayed in time (Fig. 1-9). This delayed open- bilayer (Fig. 1-11C). Alternating intracellular and extracellular
ing of the K gate occurs around the time that the sodium gate
+
protein loops connect the transmembrane protein helices. In the
inactivates (Fig. 1-10). Opening the K channels increases the +
N a and C a channels this basic structure is repeated four
+ ++
cell permeability to K ions and the resulting rapid efflux of K

+ +
times, creating a single large molecular protein that folds into a
cations returns the cell membrane to its resting transmembrane circular structure with the ion pore or tunnel located centrally
potential. In actuality, the K gate remains open for a period
+
(Fig. 1-1 ID). - - - The potassium channel structure is similar,
1 9 33 55
slightly longer than required to restore the membrane potential although the alpha subunit is assembled from four separate mol-
to exactly its pre-depolarization level. This results in a slightly ecular subunits, each consisting of a single protein entity with
more negative or hyperpolarized state that subsequently returns six helical transmembrane domains. Connecting the S 1 - S 6
to baseline as potassium equilibrates through the potassium leak transmembrane segments are protein loops that are hypothe-
channels (Fig. 1-9). sized to have several functions. Loops attached to the extracel-
The return of the membrane back to its resting potential is lular ends of the transmembrane segments appear to fold to
needed to "reset" the gating mechanisms and allow depolariza- form the inner wall of the pore, control ion selectivity, and con-
tion in response to the next threshold crossing. The N a inacti- +
tribute to the formation of the activation gate. Attached to the
vation gate will not reopen until the transmembrane potential intracellular ends of the transmembrane segments are protein
approaches the resting level. sequences that form the inactivation gates present in the N a +
The period during which the inactivation gate cannot be re- and C a ion channels. The inactivation gates are commonly
+ +
opened even with a strong depolarizing current perturbation is conceptualized to function as a "ball and chain" or hinged lid
referred to as the absolute refractory period. During this inter- that moves to occlude the intracellular opening of the channel
val, the membrane is incapable of conducting action potentials. pore. In general, the similarity in amino acid sequencing is
A relative refractory period follows the absolute refractory greatest in the transmembrane segments, S1-S6, while greater
period and is a time interval during which stronger than usual heterogeneity is seen in the connecting loops. This difference
A B the S4 transmembrane protein segment moves upward and out-

ward from the center of the channel perhaps by as much as 20
closed inactivated
angstroms. - This movement is coupled in an as yet unidenti-
9 46
fied manner to the activation gate. The actual site and mecha-
nism of action of the activation gate are unknown, although it
l
' open
appears to be located in the outer portion of the channel pore,
possibly in the region of the selectivity filter. Following move-
merit of the S4 segment, there appears to be secondary move-
C „ D ment of positively charged residues away from the center of the
I II III IV
channel, thus opening the pore. The actual gate may be located
on the S5 linker protein or the S6 segment. - 9 53
Inactivation gating of the N a channel occurs shortly after de-

+
polarization and terminates the inward sodium current. The pu-

tative location of the inactivation channel is an intracellular
protein link containing an apparently essential isoleucine,
phenylalanine, and methione (IFM) fragment that connects the
S3 and S4 segment. - - This protein loop is hypothesized to
9 46 55
function as a hinged lid, moving to occlude the inner opening of

the pore (Fig 1-12). This location for the inactivation gate fits
Figure I-I f. The structural and functional properties of the with experimental data showing disruption of inactivation by in-
voltage-gated sodium channels. A, Sodium current recorded from ternal proteases. Despite the putative role of the IFM fragment
a Xenopus oocyte expressing rat brain sodium channels. B, Alpha and as the inactivation channel, inactivation is likely a more com-
beta subunits of t h e sodium channel folded t o form an hourglass- plex process as gene manipulation studies have shown that mu-
shaped central pore.The narrowed central region is the putative loca- tations spread widely throughout the channel can affect the
tion of t h e ion selectivity filter. C, Proposed membrane structure of inactivation process. K channel inactivation appears to also in-
+
the alpha and beta subunits. Note the alpha subunit consists of four re- volve intracellular molecular conformational changes through a
peating groups each consisting of six t r a n s m e m b r a n e protein helices polypeptide located on the amino-terminus of the a subunit.
designated S I - S 6 . D, Arrangement of t h e four alpha subunit domains Conceptually gate inactivation is thought to occur by a "ball and
around t h e central pore. (From Ragsdale DS,Avoli M: Sodium channels chain" model. A protein fragment, the "inactivation ball" moves
as molecular targets for antiepileptic drugs. Brain Res Rev 1998; and interacts with a receptor at the intracellular mouth of the
26:16-28, with permission.) channel opening and closing the ion pore. 9
may help to determine ion selectivity and differences in gating ACTION POTENTIAL PROPAGATION
mechanisms between the various members of the voltage-gated
channel family. 9
Passive Current Potentials
Gene cloning and mutagenesis studies are used to map the The ability of a membrane to abruptly depolarize in response
key function of ion selectivity, voltage sensing, activation, and to a threshold voltage change is an essential feature that allows
inactivation gating to structural subunits of the ion channel. for propagation of action potentials along a nerve or muscle
Although considerable understanding of ion channels exists, de- fiber. Experimental studies using unmyelinated squid axons and
tails of the molecular shape, movement, and action that create microelectrode injection of depolarizing currents have been
the various functions are incompletely understood. The alpha used to define the events that lead to action potential conduc-
subunit alone appears to be capable of forming a functional ion tion. When microelectrodes have been placed within the axon
channel displaying both selectivity and gating properties.
Connecting the S5 and S6 transmembrane segment is a peptide
chain named the H5 or P loop (Fig 1-11). - The P loop is 9 46
RESTING OPEN INACTIVATED
thought to project inward and line the water-filled central pore
of the ion channel. Resembling an hourglass, the central pore
has relatively inert hydrophobic characteristics except in the
narrow portion where the ion selectivity filter may be located
(Fig. 1-1 IB). While the precise mechanism of ion selection is
unknown, filter characteristics include a physical size that
favors one particular ion over others and electrically charged
residues within the channel pore that stabilize appropriately
sized hydrated cations allowing them to pass through an other-
wise hydrophobic c h a n n e l . 91646
In generating an action potential, the voltage-gating mecha-

nism creates a physical conformational change in response to a Figure 1-12. The model for sodium channel activation and
threshold alteration in the transmembrane potential. This thresh- inactivation. In response t o membrane depolarization, S4 segments
old potential needs to be "sensed" by the ion channel. Experi- move outward, resulting in opening of t h e channel pore. Inactivation
mental data indicate that voltage sensing is a function of the S4 occurs when the intracellular linker segment between domains 3 and 4
transmembrane segment, which has a unique repeating structure closes over t h e intracellular mouth of t h e pore, blocking t h e flow of
of lysine or arginine amino acid residues that are electrically ions. (From Ragsdale DS, Avoli M: Sodium channels as molecular targets
charged (Fig 1-11). In response to a threshold depolarization, for antiepileptic drugs. Brain Res Rev 1998;26:16-28, with permission.)
Figure 1-13. Electronic potential distribution. A,

Experimental design of three intracellularly located recording
electrodes (RI-R3) sequentially placed from an electrode (I)
injecting a current into t h e cell.The three electrodes record
t h e membrane potential at their locations (E ). B,The time
m
course of the membrane voltage changes. N o t e that the volt-

age change at each location further from the injection site is
less and takes longer t o reach its maximum. C, A graph show-
ing the declining magnitudes of the membrane voltage at t w o
times (8 ms and 105 ms) as o n e moves further form t h e in-
jection site. (From Brown W R T h e Physiological and Technical
Basis of Electromyography. Boston, Butterworth, 1984, pp
I - 3 5 , with permission.)
DISTANCE ( m m )
along its length, the magnitude of the resting membrane poten- space. This process of current flow is known as a local circuit
tial as one moves away from the site of stimulation can be current. This type of current flow is also seen in electrical cir-
recorded. If a brief subthreshold positive current is injected into cuits that use capacitors (discussed below).
the axon, the sequentially located microelectrodes record an ex- In addition to this capacitive local current flow, - a small
24 56
ponential decline in the transmembrane voltage with distance outward-directed current flows through open ion (mainly K ) +
(Fig. 1-13). This decline in potential is known as an electro-

24
channels. Because only a limited number of potassium channels
tonic potential and results from a diminishing positive current are available, the membrane offers some resistance to this path
density the further one gets from the point of current injec- of current flow. In unmyelinated axons, most of the current that
tion. - As the injected current spreads away from the entry
24 29
site, it attempts to flow along the path of least resistance. In the

case of the axon, the current essentially has two choices, to flow
either down the length of the axon or through the plasma mem-
brane into the extracellular space.
Positive charge introduced into a localized segment of an
axon, whether injected experimentally or due to N a influx from
+
the opening of an ion channel, creates a region that is more pos-

itive or depolarized relative to adjacent segments. This extra
positive charge will flow longitudinally (moving charge is
called a current) into the more negative surrounding portions of
the axon through both intra-axonal and transmembrane routes.
Transmembrane current flow begins as the positive charge
spreading from the site of injection into more negative adjacent
regions balances (or neutralizes) an equivalent number of the
normally present intracellular negative charges that are main-
taining the resting transmembrane potential. As these negative
charges are balanced or neutralized, their ability to attract the
positive sodium ions aligned on the extracellular side of the
membrane is reduced (Fig. 1-14). The extracellular positive N a +
Figure 1-14. Local circuit current. Local circuit current d e m o n -
ions are then free to move away from the membrane. The result strating movement of positive charges into t h e excitable tissue a n d
is current flow without the need for ions to actually pass through then depolarizing adjacent regions of t h e fiber. (From Guyton A C :
the membrane. Positively charged ions moving intracellularly Textbook of Medical Physiology, 9th ed. Philadelphia, W.B. Saunders,
induce the flow of different positive ions in the extracellular 1996, pp 43-55, with permission.)
is injected is carried away by local current transmembrane plate. Charge accumulates until the two plates have a voltage
mechanisms. The remaining current that did not exit though the difference equal to the battery, at which point equilibrium
membrane is free to proceed further down the axon into the next exists. If the voltage across the plates is suddenly changed, for
adjacent segment. The same process of local circuit currents is example if the negative plate is suddenly made more positive,
repeated as intracellular anions are balanced and extracellular the loss of the negativity will reduce the electrical forces that at-
N a is no longer as tightly bound to the immediate vicinity of
+
tract and hold positive charges onto the other plate, freeing
the membrane and flows away. At each subsequent segment of some of them to flow away (Fig. 1-15). This is very similar to
membrane less current is available, making the magnitude of what occurs in axon membranes. Indeed the cell membrane ef-
the transmembrane voltage change smaller and smaller as it fectively acts as a capacitor by separating intracellular negative
moves down the axon. The process repeats until there is no charges from extracellular positive charges. The inner portion of
longer any appreciable current. the lipid bilayer is hydrophobic and nonpolar, which means that
The value of local current flow is that it allows transmem- it does not conduct or hold a charge well, thereby serving as a
brane current flow without the need for open voltage-gated ion good insulator. The outer and inner surfaces of the membrane,
channels. Current flows across the membrane without ions actu- however, are polar and charged, which means that they can act
ally passing through the membrane because the membrane acts as the plates of a capacitor to hold charge. An injection of posi-
like an electrical device known as a capacitor. A capacitor is an tive charges into the cell allows current to flow by discharging
electronic device capable of storing and separating opposite po- the membrane capacitor. This discharge occurs when positive
larity charges (Fig. 1-15). Simplified, a capacitor consists of
37
charges are added to the inside of the membrane, allowing posi-
two metal plates separated by an insulating material. If these tive charges to be removed from the outer membrane (Fig. 1¬
two plates are connected to a voltage source such as a battery, 13). The relative electrical resistance of the various pathways
current initially flows onto the two plates. Positive charges ac- for current flow dictates the preferential path of current flow:
cumulate on one of the plates and negative charges on the other transmembrane versus longitudinal flow. In unmyelinated axons
the intracellular resistance of the axoplasm is high because the
axon diameter is very small. As a result, current preferentially
discharges the membrane capacitor or flows through the mem-
brane's aqueous "resistor" channels (a small effect). The net
result of current loss through adjacent sequential portions of
the membrane is a transmembrane potential difference that di-
minishes exponentially down the length of the axon. This is
known as an electrotonically declining potential difference
(Fig. 1-13).
Active Currents in Unmyelinated N e r v e

Suppose adequate positive current is injected into a nerve to
raise the transmembrane potential to threshold. Once threshold
is reached, the voltage-gated N a channels open. The open
+
sodium channels allow positive N a ion charges to enter and

+
spread into adjacent segments of the axon's interior. This influx

of N a creates a capacitive local current flow through the adja-
+
cent segments of membrane that alters the transmembrane volt-

age enough to reach threshold in these neighboring regions.
New voltage-gated N a channels open in the neighboring mem-
+
brane allowing the process to repeat. This repetitive process re-

sults in a continuous and complete depolarization of the entire
axon membrane as the action potential propagates away from
the stimulus site (Fig. 1-14). The sequential membrane activa-
Figure 1-15. Capacitance and resistance. A, Simple circuit dia- tion for each small region of axon takes a finite amount of time
gram depicting a capacitor connected t o a c u r r e n t source. Positive as the depolarizing current: (1) first spreads to adjacent portions
charges build up on t h e capacitor connected t o the wire labeled "in," of membrane, (2) discharges the capacitor portion of the adja-
producing a net positive charge on the plate. An equal number of posi- cent membrane, (3) newly activated voltage-gated N a channels
+
tive charges are withdrawn (same as adding negative charges) from the undergo a conformational change, and ( 4 ) additional N a ions+
capacitor labeled "out," creating a n e t negative charge on t h e plate. enter the axon. In unmyelinated axons, the cumulative time for
Although t h e t w o capacitor plates are separated by a nonconducting action potential generation by this process results in the slow
medium, c u r r e n t effectively flows across t h e capacitor because t h e conduction velocity of 10 - 1 5 meters/second for fibers for
charges added t o t h e "in" plate are r e m o v e d from the " o u t " plate. axons with a diameter of 10 micrometers (u.). 6
Charges flow until both plates equal the voltage of the current gener- Each action potential generated by a particular membrane
ator. This process of charging the capacitor takes some time. B.The region is identical to the action potential from which it was gen-
membrane also has passive channels through which current may flow erated; this phenomenon is called the all-or-none principle. 37
and effectively act as resistor.The cell m e m b r a n e may be considered In unmyelinated axons, action potential propagation occurs as
t o act as a capacitor and resistor in parallel. (From Koester J: Resting each small area of nerve activates its neighbor in a continuous
membrane potential and action potential. In Kandel ER, Schwartz JH fashion. These currents are active in the sense that they produce
(eds): Principles of Neural Science, 2nd ed. N e w York, Elsevier, 1985, pp a self-sustaining alteration in the transmembrane voltage as op-
4 9 - 5 7 , with permission.) posed to the passive spread over relatively small distances as in
the case of the electrotonic potential. In the process of action

potential generation, sodium channel inactivation results in the
refractory period and renders the nerve unexcitable for a brief
time. This limits physiologically generated action potentials to
unidirectional propagation because the region immediately
behind the presently active portion of membrane is in the ab-
solute refractory period and cannot be activated again.
Depolarizing currents of slowly increasing magnitude can be
injected into the axon and actually raise the membrane potential
above the threshold voltage without inducing an action poten- Schwann cell
tial. This is possible because not all of the voltage-gated N a +
cytoplasm
chwann cell
channels open at the same identical threshold voltage. If a nucleus
slowly increasing depolarization current opens a few N a chan-+
Node of Ranvier
nels at a time, the initially open channels enter into the refrac-
tory state before others actually open. This results in less
synchronous opening and closing of the voltage-gated channels
preventing the rapid opening of a large number of N a channels
+
needed to create an action potential. The process of a depolariz-

ing current exceeding the nerve's threshold voltage without in-
ducing an action potential is known as accommodation. A 37
Unmyelinated axons
relatively slow increase in a depolarizing current may induce an
action potential at a greater than normal threshold voltage, or
Schwann cell nucleus
not result in an action potential at all, even if the transmembrane
potential is completely reversed. Schwann cell cytoplasm
Active Currents in Myelinated N e r v e

In unmyelinated axons, the current spreads from the site of
depolarization by three mechanisms: (1) capacitive current
through the membrane, (2) current through ion channels, and
(3) longitudinally down the nerve to activate the adjacent mem-
Figure 1-16. Myelination. A, A myelinated axon demonstrating t h e
brane region. Continuous conduction in small-diameter un-
circumferential wrapping of t h e myelin sheath insulating the n e r v e
myelinated axons results in unacceptably slow conduction
along its length except for that region at t h e node of Ranvier.There is
velocities for organisms that need to respond and react quickly
one Schwann cell per internode. B, O n e Schwann cell may provide a
to external stimuli. One means of increasing impulse conduc-
modicum of myelination for multiple axons in unmyelinated n e r v e s .
tion is to speed current movement down the axon's interior. This
(From Guyton AC: Textbook of Medical Physiology, 9th e d .
may be accomplished by decreasing the longitudinal resistance
Philadelphia,W.B. Saunders, 1996, pp 43—55, with permission.)
to current flow. Longitudinal resistance is reduced as the size
(cross-sectional area) of the axon increases. The concept of in-
creasing unmyelinated axon diameter has been taken to an ex- channels. - Myelinated segments between the nodes of
59 66
treme in the giant axon of the squid. The squid's axonal Ranvier are referred to as the internode regions. The axon
diameter may reach up to 1 millimeter (mm). - Although this
24 29
membrane beneath the myelinated sheath is devoid of any type
extreme axon size may be appropriate for marine mollusks, it is of N a channel but does contain a number of K c h a n n e l s .
+ + 101167
impractical for complex mammalian nervous systems that have If an action potential is initiated at a node of Ranvier, the in-
nerves containing hundreds of axons. wardly directed sodium current passively proceeds longitudi-
A second mechanism of increasing velocity is to avoid the nally down the axon's interior similar to the longitudinal current
continuous depolarization of the entire axon membrane as flow in unmyelinated axons. In the myelinated axon, however,
occurs in unmyelinated fibers. Myelination of nerve fibers is an the myelin sheath insulates the axonal membrane and allows
evolutionary adaptation to this problem. All mammalian periph- only a very small portion of the total current to exit through the
eral axons are surrounded by the plasma membrane of Schwann internode region (Fig. 1-17A). No insulator is perfect; a small
cells. - In myelinated nerve fibers, a Schwann cell envelops a
6 37
amount of current can still pass through the myelin sheath but is
portion of a single axon and wraps itself around the axon multi- reduced by about 5,000 times compared to an unmyelinated
ple times sequentially laying down layers of its cellular mem- nerve. Myelination of the internode impedes transmembrane
22
brane. A second Schwann cell can then envelop an adjacent current flow through three mechanisms. First, the thick myelin
segment of axon leaving a small portion of bare axon between covering significantly increases the distance between the nega-
them. This process is repeated along the entire length of the tive intracellular charges and the positive extracellular charges,
axon. The Schwann cell membrane contains the lipid sphin- thereby minimizing attraction for each other. The larger dis-
gomyelin, which is a superb insulating material. The myelinated tance between the charges reduces the capacitance of the axonal
nerve consists of an outer myelin sheath and the axon's plasma membrane approximately 50-fold markedly reducing the
membrane (axolemma) which encases the axon's cellular mate- amount of current needed to discharge the membrane's capaci-
rial or axoplasm. The myelinated axon is completely sur- tance (Fig. 1-17A). Second, the internodal axon has few trans-
22
rounded by myelin except where adjacent Schwann cells abut membrane ion channels through which ion efflux can
each other (Fig. 1-16). This region of the myelinated axon is occur. Third, the impermeability of the sphingolipid sub-
10 n 6 7
known as a node of Ranvier and contains voltage-gated N a +

strate and its thickness offers a substantial resistance to ion flow
Distance along axon
Figure f-/7. Saltatory conduction. A, In a myelinated nerve, the local circuit currents primarily flow into o n e node of Ranvier and extend
longitudinally down the axon, exiting at t h e next node of Ranvier.There is a small amount of current lost in the internode region but n o t enough
t o reduce t h e amount of current below necessary t o depolarize t h e next node in line. B,A demonstration of the saltatory nature of conduction in
myelinated nerve. (From Koester J; Voltage-gated channels and t h e generation of the action potential. In Kandel ER, Schwartz JH (eds): Principles
of Neural Science, 2nd ed. New York, Elsevier, 1985, pp 75-86, with permission.)
even if ions could penetrate the axon membrane. Because of close together slow conduction by requiring unnecessary action
these electrical characteristics of the internodal axon membrane, potential generation at unneeded nodes. Investigators have
the inward N a current through a node of Ranvier spreads longi-
+
found that a nearly linear relationship between axonal diameter
tudinally toward the next node without the internodal membrane and internodal length is present. The apparent optimal inter-
depolarizing. The nodal membrane contains a large number of nodal length is about 100 (75-175) times the axon's diame-
N a ion channels compared to a similarly sized region of un-
+
ter. - An axon with a diameter of 10 urn has an internode
52 37
myelinated membrane, which enhances sodium influx during length of approximately 1 mm. For a typical action potential ve-
depolarization. The process is repeated until the end of the axon locity of 50 m/s and duration of 0.5 milliseconds (ms), the
is reached (Fig. 1-17A). action potential extends over 25 internodes (nerve conduction
Myelination significantly increases the impulse propagation velocity = distance/time; 50 m/s = D/0.5 ms = 25 mm). The in-
velocity allowing speeds of up to 100 m/s or more. - - The 22 37 66
ternodes serving the entire action potential over this span are in
myelin eliminates the need to depolarize the internodal mem- various stages of depolarization and repolarization (Fig. 1-18).
brane. The intracellular current quickly proceeds to the next Myelinated nerve consists of two components, myelin and
node where an action potential is generated (Fig. 1-17B). The axon, with an optimal ratio between them. For a given diameter
action potential essentially jumps from one node to the next. of axon, increases in myelin thickness create two opposing ef-
This process is known as saltatory conduction (from Latin fects on conduction velocity. Increased thickness of myelin re-
saltare, to jump or leap). - In some mammalian nerves (includ-
6 22
duces internodal membrane capacitance, lessening local current
ing human), the nodal membrane does not appear to have volt- leak, and increases propagation velocity. At the same time, axon
age-gated K channels. Repolarization is theorized to occur
+
diameter is reduced, resulting in increased internal resistance
through a sodium back-leak or handled by the Na -K -ATP + +
and decreased conduction velocity. Theoretical calculations pre-
pump. - - Recall that depolarization and repolarization can
63 66 67
dict that optimal conduction velocity occurs with a ratio of ap-
occur through the alteration of sodium permeability according proximately 60% axon and 40% myelin. This agrees closely
to the Goldman-Hodgkin-Katz equation. Sodium channel inac- with experimental measurements that have shown the ratio of
tivation decreases sodium permeability and allows the mem- axon to myelin in various species is approximately O.6. 61
brane to return to its resting state by the passive equilibration of Myelination effectively solves the problem of increasing con-
sodium and potassium ions through the protein "leak" channels duction velocity without excessively increasing axon diameter.
of the membrane. The giant squid axon has a diameter of about 1000 urn and con-
ducts impulses at 25 m/s. - - A myelinated nerve with a diame-
6 29 37
Optimizing t h e Myelinated N e r v e ter of 20 fim can have conduction velocities approaching 100
The distance between adjacent nodes of Ranvier influences m/s. The result of myelination is a 50-fold reduction in diameter
conduction velocity. If the nodes are spaced too far apart, prop- with a 4-fold increase in conduction velocity. -
6 22
agation may cease because the slowly declining intra-axonal Neural Classification. Two classification schemes have
current may be of insufficient magnitude to traverse an inter- been developed to categorize the relationship between nerve
node and depolarize the next node in line. Nodes that are too conduction velocity and fiber diameter. Erlanger and Gasser
NEUROMUSCULAR TRANSMISSION
DIRECTION OF
PROPAGATION
Skeletal muscles must first be activated by an electrical im-

pulse prior to performing their essential function of moving our
joints. Large myelinated motor axons divide into a variable
number of fine terminal nerve branches, depending upon the
muscle, prior to synapsing with a muscle fiber. Large muscles
that perform gross movements, e.g., gastrocnemius, have a high
number of individual muscle fibers (greater than several hun-
dred) innervated by each axon. Conversely, muscles requiring
fine motor control have a low terminal innervation ratio (ratio
of muscle fibers innervated by one axon) approaching one-to-
one in the extraocular and vocal muscles. Each terminal branch
attaches to a single muscle fiber at a specific region approximat-
-60 J
ing its mid-point referred to as the end-plate or neuromuscular

P O S m O N S OF NODES OF RANVIER
junction. Only about 2% of muscle fibers have more than one
48
Figure 1-18. Longitudinal expansion of the action potential. neuromuscular junction and these are found in especially long
The action potential's depolarization and repolarization is not an in-
muscles like the sartorius. 19
stantaneous event, but e x t e n d s over a finite period of time during Neuromuscular Junction Anatomy. When a large myeli-
i which the impulse continues t o propagate. As a result of this propaga- nated motor axon approaches a muscle fiber, it divides into mul-
1
tion, some portion of t h e nerve is depolarized while t h e r e are o t h e r tiple small nerve twigs that run along the muscle's surface for a
aspects of the nerve repolarizing over a somewhat longer time course. short distance before ending. The terminal portion of each axon
This has the effect of distributing the entire action potential over mul- contains neurotubules and neurofilaments from the axon, multi-
tiple internodes. T h e dashed lined r e p r e s e n t t h e internodes over ple mitochondria for metabolic energy production, and large
which an action potential extends (about 25 internodes). (From Brown numbers of membrane-bound spheres 3 0 0 - 5 0 0 A in diameter
WF:The Physiological and Technical Basis of Electromyography. Boston, called synaptic vesicles, containing the neurotransmitter
Butterworth, I984»pp I-35, with permission.) acetylcholine ( A C h ) . The region of the muscle fiber lying
19
under the nerve twig is called the motor end-plate (Fig. 1-19
and 1-20). Within this region of muscle are several muscle fiber
developed the first system. It considers both sensory and
18
nuclei, mitochondria, ribosomes, and pinocytotic vesicles. 48
motor fibers, and uses a combination of capital letters and lower Covering the entire end-plate are cytoplasmic extensions of the
case Greek letters. The second classification, by Lloyd and Schwann cell associated with the nerve twig innervating the
Hunt, uses Roman numerals I-IV and just considers sensory particular muscle fiber. The nerve's end portion, axon terminal,
fibers (Table 1-3). Nerve fibers belonging to groups A - B and
44
is not actually in contact with the muscle cell membrane but
I—III are myelinated, while categories C and IV designate un- separated from it by a distance of about 50-75 nm. - This sep- 19 48
myelinated nerve fibers. Larger fibers have greater conduction aration, the primary synaptic cleft, appears as an irregular
velocities than smaller fibers and myelinated axons conduct im- region on electron microscopy and consists of multiple invagi-
pulses faster than unmyelinated axons. nations into the muscle fiber creating secondary synaptic clefts
Table 1-3. Nerve Fiber Classification

Sensory and Sensory
Motor Fibers Fibers Diameter Velocity Function
A-a la 10-20 0-120 Motor: alpha m o t o r neurons
Sensory: muscle spindle afferents
A-a lb 10-20 50-120 Sensory: Golgi tendon organs,
touch, and pressure
A-(3 II 4-12 25-70 Motor: motoneurons t o intra/extra-
fusal muscle fibers
Sensory: secondary muscle spindle
afferents, touch, pressure, vibration
A-Y 2-8 10-50 Motor: small gamma motoneurons t o
intrafusal muscle fibers
A-a III 1-5 3-30 Sensory: small touch, pain, and
temperature fibers
B 1-3 3-15 Motor: small unmyelinated pre-
ganglionic autonomic fibers
C IV <1 < 2 Motor: all postganglionic autonomic
fibers
Sensory: pain and temperature
The table represents a combination of the t w o nerve classification systems. Nerve fiber diameter is in micrometers and conduction velocity is in meters/second."
Figure / - / 9 . Neuromuscular junction.A neuromuscular junction involving skeletal muscle. A,A longitudinal section through the end-plate of
t w o terminal axons showing the presynaptic and postsynaptic regions overlying skeletal muscle. B, Overview of the same region in A. C, Close-up
of t h e neuromuscular junction detailing t h e synaptic vesicles, terminal axon lying in t h e synaptic trough and t h e subneural clefts. (From Fawcett
D W : Bloom and Fawcett: A Textbook of Histology. Philadelphia, W.B. Saunders, 1986, with permission.)
or subneural clefts 0.5-1.0 urn deep (Fig. 1-19). The postsy-

48
remarkable accomplishments. An action potential is propagated
naptic membrane corrugation serves to increase the end-plate along the axon into each of the multiple terminal axon twigs.
surface area. The ACh receptors are located on the summits of The action potential's conduction velocity slows considerably
the postsynaptic folds and are optimally positioned opposite the in the axon terminals to about 10-20 m/s from the greater than
ACh release sites of the presynaptic membrane. 50 m/s in the main axon. The smaller diameter of the axon
35
twigs and loss of myelination near the neuromuscular junction

ELECTROCHEMICAL CONDUCTION causes conduction velocity to decrease. When the action poten-
tial depolarizes the terminal axon, sodium conductance is in-
The process of transferring electrical impulses from the increased. In addition, C a conductance also dramatically
++
nervating axon to its muscle fiber is one of the body's truly increases secondary to the presence and opening of C a channels,
++
Figure 1-20. Neuromuscular junction. Photomicro-

graph of t h e region shown in Figure 1-17. N o t e t h e nerve
branching into several terminal axons each innervating
one muscle fiber. (From Fawcett D W : Bloom and Fawcett:
A Textbook of Histology. Philadelphia, W.B. Saunders,
1986, with permission.)
Chapter 1 NERVE AND MUSCLE ANATOMY A N D PHYSIOLOGY — 19
permitting calcium ions to enter the terminal portion of the axon. ACETYLCHOLINE RECYCLING
Calcium ion entry is critical to the process of neuromuscular
transmission; when C a is removed from the extracellular
+ +
The neurotransmitter ACh is confined within synaptic vesi-
space, transmission ceases. Magnesium ions ( M g ) have been ++
cles contained in the terminal axon. Confining acetylcholine in
shown to compete with C a entry and can block the effects of
+ +
vesicles likely (1) protects the molecules from breakdown by
Ca . - Although the exact mechanism is unknown, Ca""" facili-
++ 13 14
the small amount of an acetylcholinesterase (AchE) enzyme
tates fusion of the ACh-containing vesicles with the presynaptic present in the axon terminal, and (2) optimizes the necessary
membrane of the terminal axon. Once the vesicles have fused amount of transmitter released relative to the finite number of
with the nerve terminal membrane, ACh discharges into the vesicle binding sites. Approximately 10,000 molecules of
22
synaptic cleft. An action potential releases between 100 and 200 acetylcholine are contained within a single synaptic vesicle and
vesicles at each axon terminal in amphibians compared to less are referred to as a quantum of acetylcholine. ' 4
than 100 vesicles in mammals. - The C a persists in the axon 37 48 ++

At rest, there is spontaneous random release of synaptic vesi-
terminal for approximately 200 ms keeping the axon terminal in cles or quanta. Spontaneous release is believed to reflect the
a "hyperexcitable state," enhancing the release of ACh if a presence of an intracellular resting level of C a in the axon ter-
++
second action potential depolarizes the axon within this time minal necessary for the normal physiologic functioning of the
frame. 15
mitochondria. With release, ACh from one vesicle diffuses
4
The ACh quickly diffuses across the synaptic cleft (0.2-0.8 across the synaptic cleft heading for their receptors on the post-
ms) - - to bind with ACh receptors. These receptors are large
36 54 56
synaptic membrane of the end-plate (Fig. 1-21). Acetylcholin-
transmembrane proteins that contain both a binding site for esterase enzyme molecules, in addition to their location in the
ACh and an ion channel. The ACh receptors are located primar- axon terminal, are also evenly distributed throughout the pri-
ily but not exclusively on the summits of the secondary synap- mary and secondary synaptic clefts, weakly attached to the sar-
tic clefts at an approximate density of 20,000 to 25,000 colemma. It is estimated that there are between 3,000 and
48
receptors per Jim of membrane. - These receptors are li-

2 32 54
6,000 AChE molecules per jam in the postsynaptic mammalian
2
gand-activated rather than voltage-gated. The surface portion membrane. Hydrolysis of the acetylcholine molecules into
7
of the protein acts as a receptor for the ACh molecule. When choline and acetate begins soon after they are released into the
the ACh ligand fuses with its receptor, a conformational change cleft region. The surviving acetylcholine is free to bind to its re-
occurs in the channel portion of the receptor that opens the ion ceptor, resulting in a small amount of postsynaptic membrane
gate for about 1 m s . 22
depolarization (Fig. 1-21). The magnitude of the intracellularly
The channel opening is large enough to allow transmembrane recorded depolarization resulting from the remaining acetyl-
flow of N a , K , and C a cations. When open, the channel
+ + + +
choline molecules is about 1 mV; it is referred to as a miniature
allows nonspecific cation influx; however, anions are not per- end-plate potential ( M E P P ) and occurs approximately once
18
mitted to pass as the inner wall of the channel is negatively

charged and repels anions such as Cl". Similar to the nerve fiber
membranes, the major ion influx causing depolarization is N a . +
Although the K ion can easily fit through the open channel, the
+
resting membrane potential of - 90 mV to - 80 mV for muscle

is close to the equilibrium potential of K , which restricts large +
potassium effluxes. Calcium ions are present outside of the

muscle and do enter the muscle to some extent, but the concen-
tration is 50 times less than that of Na . +
The N a normally entering the end-plate region is sufficient

+
to depolarize the adjacent muscle fiber membrane. The quantity

of N a that normally enters through the ion channels reverses
+
the transmembrane potential of the end-plate region by as much

as 75 mV. This greatly exceeds the threshold voltage of approx-
imately 15 mV needed to initiate the positive feedback loop of
sodium activation. The voltage difference between the threshold
level and the final magnitude of the end-plate potential is re-
ferred to as the neuromuscular junction's safety factor. This
large safety factor ensures that even under normal physiologic
repetitive nerve action potential generation during which the
amount of ACh released from the axon terminal decreases,
enough ACh continues to be emitted to exceed threshold.
Localized non-propagating reversals (depolarization) of the
muscle end-plate's transmembrane potential by any amount is
referred to as the end-plate potential (EPP). The spread of this
potential is purely by electrotonic means over the end-plate Figure / - 2 / . ACh release and synthesis. Sequential steps in-
region. If enough ACh is released to reach or exceed threshold, volved in the ACh cycle in t h e electrochemical transmission in muscle
the EPP is quickly overtaken by the generation of an action po- activation. ACh, acetylcholine, AChE, acetylcholine esterase, AChR,
tential. This action potential then results in depolarization of the acetylcholine receptor, CoA, coenzyme A, Ch, choline. See text for e x -
muscle membrane surrounding the end-plate and an all-or-noth- planation of process and Figure 1-20. (From McComas AJ:
ing propagating impulse is initiated in the muscle fiber (see Neuromuscular Function and Disorders. Boston, Butterworth, 1977,
Muscle Tissue). with permission.)
necessary constituents for synthesis have also been demon-

— Action potential strated in the cell body. 45
c J f Acetate
I Calcium entry (Z) / © Structure of the Acetylcholine Receptor. The AChR is an
. .— } Packaging / ionotropic ligand-gated transmembrane receptor channel that
§ Vesicle rupture (5) -a of ACh -» Synthesis of ACh © shares structural similarity with a family of protein receptors
< f in vesicle h that include the glycine, gamma-aminobutyric acid (GABA),
Release of ACh © (3) Uptake or choline (§)
and serotonin 5-HT receptors. - Substantial structure details
33 65
1 •- '• 3
Cleft Diffusion of ACh (f) of the ACh receptor exist aided by the early ease of isolating the
1 channel from the electric organ of the Torpedo ray. The AChR
: Combination Hydrolysis of ACh ^ consists of five protein subunits: a , p, y, and o with each recep-
£ with receptors © *" (ACh esterase) ^
| } tor having two copies of the a subunit. These protein subunits
E Increased permeability,-. congregate to form a ring-shaped structure with a central pore
£ to Na* and K" &
^ Depolarization (Fig. 1-23). The extracellular portion of the receptor extends
,2 (end-plate potential) beyond the cell membrane by approximately 100 A and creates
T
Action potential a funnel that is thought to direct ions into the pore opening. The
Figure 1-22. Neuromuscular transmission. Summary of neuro- funnel merges with the channel pore and narrows in its midpor-
muscular transmission processing of ACh. T h e numbers above corre- tion to form the ion filter and gate mechanism. Each of the
spond t o t h o s e shown in Figure 1-19. (From McComas AJ: major subunits of the receptor consists of four transmembrane
Neuromuscular Function and Disorders. Boston, Butterworth, 1977, helical proteins termed TM1-TM4. These subunits are arranged
with permission.) so that the TM2 regions are oriented to form the lining of the
central pore. The central pore is hydrophobic and narrows to a
diameter of approximately 6-10 A. The amino acids that com-
64
every 5 seconds. Following invasion of the terminal axon by

17
pose the TM2 subunit are organized so that a ring of negatively
an action potential, multiple ACh-containing vesicles are re- charged residues encircles the pore near its narrowest portion.
leased. The summated effect of multiple MEPPs produces the The combination of the hydrophobic lining, negative charge
previously described EPP. If the E P P reaches threshold, a ring, and physical size appears to provide the ion selectivity of
muscle action potential ensues. the ACh receptor. When open, the channel allows small cations
Acetylcholine is released from its receptor binding site back including Na , K , and C a to enter (though monovalent ions
+ + ++
into the synaptic cleft following closure of the receptor channel are selectively prefered). Anions and large cations do not pass
complex and is subsequently hydrolyzed by AChE to choline through the channel. During the process of neuromuscular junc-
and acetate. The choline is taken up by the axon terminal. tion synaptic activity, the opening of the ACh receptor channel
Within the terminal, the enzyme choline acetyltransferase cat- allows N a influx down its concentration gradient while simul-
+
alyzes the resynthesis of ACh from choline and acetate. The ac- taneously allowing K efflux along its opposite concentration
+
etate is derived from the hydrolysis of acetyl-coenzyme A. The 48

gradient.
acetylcholine is then incorporated into synaptic vesicles ready Each receptor contains two binding sites for the ACh neuro-
to repeat the process (Figs. 1-21 and 1-22). The ACh is manu- transmitter molecules located on the a subunits. Cooperation
factured and packaged primarily but not exclusively in the ter- exists between the binding sites so that the binding of the first
minal portion of the axon. A small quantity of ACh and the ACh molecule enhances the binding of the second. Following
successful binding of two ACh molecules the central channel
opens almost immediately (within 20 ps). The current experi-
mental model of channel opening indicates that the gating
mechanism depends on a conformation change in the T M 2
member of the subunits. The T M 2 transmembrane proteins
forming the wall of the channel pore are alpha-helical and have
a kink in their structure that creates a tight ring that narrows the
channel and prevents ion passage. With binding of ACh, through i
an unknown mechanism, the TM2 segments rotate. In doing so,
the kinked portion of the molecule is rotated out of the central
pore, increasing its diameter and allowing ion flow. 64
MUSCLE TISSUE
The primary function of muscle tissue is to create forces that
allow for efficient movement. This section covers the mecha-
Figure I-23. Model of the nicotinic acetylcholine receptor. nisms by which muscular tissue contracts and generates tension.
The model includes t h e quaternary structure (clockwise arranged sub- Basic muscle histology and architecture, the fundamental elec-
units a , o, p, and a; t h e y subunit in front is removed), the helix TM2 trical properties of muscle membranes, the neuromuscular
forming t h e channel wall, the negative charges arranged in three rings, synapse, and excitation/contraction coupling are reviewed.
which determine channel conductivity and selectivity, and t h e binding Since the basic concepts of nerve and muscle membrane physi-
sites for t h e a - n e u r o t o x i n s at t h e subunit interfaces (Tx). (From ology are relatively similar, only the specific details of muscle
Fawcett D W : Bloom and Fawcett: A Textbook of Histology. electrophysiology are noted. Additional details of muscle elec-
Philadelphia,W.B. Saunders, 1986, with permission.) trophysiology can be found in standard physiology texts. - - I9 22 23
MUSCULAR COMPONENTS Extrafusal Fibers

Voluntary activation is initiated through electrochemical con-
Skeletal Muscle Types duction at the end-plate or neuromuscular junction. The adult
Skeletal or voluntary muscles consist of two main cate- human extrafusal skeletal muscle fibers extend from one end of
gories: extrafusal and intrafusal fibers. Extrafusal muscle
22
the muscle to the other and are 4 0 - 1 0 0 urn in diameter. A 19
tissue constitutes the bulk of a skeletal muscle and is responsi- single muscle fiber is a multinucleated cell surrounded by a
ble for force generation during muscle contraction. These are plasma membrane (sarcolemma). The sarcolemma is very sim-
the muscles fibers attached to bone through tendon and will be ilar to the plasma membrane of the nerve fiber; it is a lipid bi-
the main focus of this discussion. Intrafusal fibers consist of two layer with hydrophobic tails forming the interior and
subtypes referred to as nuclear chain and nuclear bag fibers, hydrophilic heads constituting the inner and outer surfaces.
also collectively referred to as the muscle spindle. The muscle Transmembrane proteins are also present in the sarcolemma for
spindle has its own nerve supply and lies in parallel with the ex- the passage of various ions. It is interesting to note that the
trafusal fiber, thereby sensing when the extrafusal fiber muscle membrane is rather convoluted along its length. These
stretches or contracts. The intrafusal fibers primarily act as a
22
folds disappear when the muscle is stretched and most likely
servo-control or feedback mechanism to monitor and assist the represent the necessary slack to accommodate relaxation of the
extrafusal fibers in performing various tasks in a controlled muscle. Extensions of the sarcolemma penetrate into the sub-
48
manner. stance of the muscle fiber forming tunnels. These tunnels are
Myofibrils
Sarcolemma
Transverse tubule
I band Transverse tubule
Terminal
cisternae
Sarcotubules
Figure 1-24. Muscle tissue. Section through skeletal muscle demonstrating t h e various subcomponents of muscle tissue. In mammalian muscle
t h e T tubule would align with the A/I bands as opposed t o the Z line as shown above.The mammalian skeletal muscle, therefore, would have t w o T
tubules per sarcomere (Z line t o Z line). (From Fawcett DW: Bloom and Fawcett: A Textbook of Histology. Philadelphia, W.B. Saunders, 1986, with
permission.)
22 — PART 1 FUNDAMENTAL PRINCIPLES
SKELETAL MUSCLE
Muscle
"A wmmmmmmmm^ •*—- •

Muscle Fasciculus
2—Sarcomere—Z \ s Myofibril
G-Actin M o l e c u l e s
oo
Mvof Moments
' F-Actin Filament
Myosin F i l a m e n t
Myosin M o l e c u l e
s i
T ) l
Light Heavy
Meromyosin Meromyosin
Figure 1-25. Skeletal muscle subcomponents. Increasing magnification of the skeletal muscle's subcomponents. (From Fawcett DW: Bloom
and Fawcett: A Textbook of Histology. Philadelphia, W.B. Saunders, 1986, with permission.)
called T (transverse) tubules and allow the extracellular space out from the tail through an arm segment. One myosin filament
to extend all the way across the muscle fiber from one side to contains about 200 of these single-tail/two-head molecules,
the other. - The T tubules contain extracellular fluid and form
19 22
with the heads projecting outward from the central conglomera-
an intricate channel system throughout the muscle's interior tion of tails forming the main filament. The head and its project-
(Fig. 1 -24). The sarcolemma is covered externally by a 50-nm ing part are referred to as a cross-bridge; it has two flexible
thick basement membrane, which is composed of polysaccha- hinges, one at the head/arm and the other at the arm/filament in-
rides and protein. The basement membrane provides support
48
terface (Fig. 1-26). Each myosin head acts as an ATPase to
for the muscle fiber and may act as a catchment area to prevent cleave a molecule of ATP. The energy liberated by this process
22
effluxing ions from diffusing too far from the sarcolemma. Just is used to maintain the cross-bridge in the extended or "cocked"
on the interior of the sarcolemma and suspended from it are the position. The entire myosin filament is twisted about a central
muscle cell's nuclei. Multiple nuclei are dispersed along the axis allowing the cross-bridges to extend longitudinally and cir-
length of the muscle fiber. cumferentially 360°.
Contained within the muscle cell are myofibrils, the struc- The actin filament is composed of three subcomponents:
tures responsible for muscle contraction (Fig. 1-25). Myofibrils actin, tropomyosin, and troponin (Fig. 1-26). Two helical
consist of two large polymerized protein molecules forming the strands of F-actin, composed of polymerized G-actin molecules,
myosin and actin filaments. A myosin molecule consists of a form the primary structure of the actin filament. Each G-actin
single tail attached to two head portions, each of which extends molecule contains one molecule of ADP. Wound loosely within
Chapter I NERVE AND MUSCLE ANATOMY A N D PHYSIOLOGY — 23
Myosin molecule
Myosin filament
D E
Figure 1-26. Myofilament structure. Proposed composition of the myofilaments involved in muscle contraction and the molecular basis of
contraction. A, Subcomponents of the myosin molecule. B, Foreshortened arrangement of the myosin filament. C,The actin filament's double heli-
cal composition with the troponin complex attached. D-E, Hypothesized mechanism of C a binding t o t h e troponin/tropomyosin complex r e -
+ +
sulting in a conformational change exposing actin's active site.The myosin head attaches t o t h e actin and moves this filament along. (From Fawcett
DW: Bloom and Fawcett: A Textbook of Histology. Philadelphia, W.B. Saunders, 1986, with permission.)
the helical structure of the F-actin are two strands of a tropo- The larger myosin and smaller actin filaments interdigitate,
myosin molecule. The tropomyosin molecules are believed to giving the myofibrils their characteristic alternate light and dark
overlie "active sites" on the actin molecules. It is these "active bands (Fig. 1-25). The light bands consist only of actin filaments
sites" that are the linkage sites for the myosin cross bridges. and are "isotropic" to polarized light, hence they are called I
Troponin, the third subcomponent of the actin filament, is at- bands. One end of the actin filaments is firmly anchored to the Z
tached two thirds down the tropomyosin molecule and consists disc and the other end projects out between myosin filaments.
of three globular proteins: troponin I, T, and C (Fig. 1-26). These Z discs extend from myofibril to myofibril across the di-
Troponin I binds strongly to actin, troponin T is attached to ameter of a muscle fiber. The region of muscle or myofibril be-
tropomyosin, while troponin C has a large affinity for C a . The++
tween two Z discs is called a sarcomere (Fig. 1-25). The light or
troponin complex attaches the tropomyosin molecules to the I band is flanked by two dark bands composed of overlapping
actin molecules, thereby forming the complete actin filament. actin and myosin filaments that are "anisotropic" to polarized
light, hence A bands. The interaction between the myosin cross- of the neuromuscular junction area. An induced action potential
bridges and actin filaments causes the muscle fiber to shorten or then propagates along the sarcolemma. Voltage-gated N a chan- +
contract because the above-noted filaments slide past each other. nels open, further altering the transmembrane potential and
Down the middle of an A band lies the H zone, which is merely opening additional voltage-sensitive N a channels. The in-
+
a lighter-colored region where actin filaments do not abut each wardly directed N a current then spreads longitudinally down
+
other in a muscle longer than its resting length (Fig. 1-25). 22

the muscle fiber and completes a local circuit current by dis-
Sarcoplasm is the intracellular fluid surrounding the myofib- charging the capacitance of the membrane. In unmyelinated
ril. This fluid contains the aqueous ions of potassium, magne- nerve, discharging the regional membrane capacitance takes
sium, phosphate, sodium, and multiple enzymes. Large 48
time, contributing to the slow conduction velocities of 2 5 - 3 0
numbers of mitochondria are suspended in the sarcoplasm and m/s. Muscle fibers conduct action potentials similarly to un-
lie in and about the myofibrils (Fig. 1-24). Mitochondria sug- myelinated nerves; they do not possess myelin and, therefore,
gest that the muscle cell requires substantial energy for the con- must rely upon each segment of membrane depolarizing the ad-
traction process. The myofibrils are also surrounded by an jacent regions. Of interest is the observation that muscle fibers
intricate series of channels called the sarcoplasmic reticulum. of comparable size to unmyelinated nerve have slower action
Longitudinal sarcoplasmic reticulum channels end in a rela- potential conduction velocities. The reason for this is the in-
22
tively large common terminal cisternae at either end of the sar- creased surface area of muscle membrane due to the T tubule
comere (Fig. 1-24). Within the muscle fiber, T tubules are system. With more membrane to depolarize, a larger capacitor
closely associated with terminal cisternae at each demarcation must be discharged. Discharging a larger capacitor, or bringing
between the A and I bands. Two terminal cisternae are in close more membrane to threshold,, takes more time. Slowed muscle
association with one T tubule to form a so-called triad. fiber depolarization results in an action potential duration about
Although these three structures are in very close association, 5 times greater than nerve, and conduction velocities on the
there are no known channels connecting them. The triad is criti- order of 3-5 m/s for mammalian muscle. 22
cally positioned next to where the muscle fiber's mechanism The unique properties of T tubules result in a number of note-
produces the necessary forces for contraction. The T tubule is worthy effects. Because of the rather small confines of the T
believed to conduct an action potential into the depths of the tubule system, effluxing K accumulates in the region just out-
+
muscle and into the terminal cisternae. side of the sarcolemma during repolarization. The increased
concentration of K just outside the membrane during the latter
+
ELECTRICAL ACTIVITY portion of repolarization tends to depolarize the cell again, i.e.,
the localized extracellular accumulation of K alters the concen-
+
Resting Membrane Potential tration gradient across this portion of the membrane, altering
Inserting a microelectrode into a mammalian muscle cell re- the transmembrane potential. Recall from the Goldman-
veals a resting membrane potential of approximately 80 to 90 Hodgkin-Katz equation that increasing the extracellular concen-
mV negative compared to the extracellular space. By measuring tration of K tends to make the transmembrane voltage more
+
the concentrations of various ions inside and outside of the positive, i.e., a bias toward depolarization. Accumulated K can +
muscle cell (Table 1-2) and using the Goldman-Hodgkin-Katz diffuse back into the cell following repolarization and again de-
equation, it is possible to calculate the resting membrane poten- polarize this aspect of the membrane. This effect can actually be
tial and arrive at - 90 mV. The basis of the resting membrane observed as an afterdepolarization on the falling phase of a
potential in muscle is very similar to that for nerve. muscle's action potential. This afterdepolarization can last for
48
The sarcolemma is a semipermeable membrane that allows several hundred milliseconds and may summate with repetitive
chloride and potassium ions to pass but restricts the movement activity. Fibers may become self-activating and continue to fire
of extracellular sodium cations and intracellular anions. in a self-sustaining manner. This may be the explanation of
Remember that the sarcolemma also has a rather large portion some myotonic conditions in which chloride conductance is re-
extending into the muscle fiber itself. Investigations have duced (see Chapter 27). Nature has obviated repetitive muscle
demonstrated that the T tubules possess permeability to K ap- +
firing from T tubule K accumulation by providing the sar-
+
proximately twice that for the surface portion of the sarcolemma with a 10 times greater Cl~ than K conductance. The
+
colemma. - There is, however, a rather slow effect on the

23 47
increased Cl" conductance allows for a relatively large influx of
membrane potential when the extracellular potassium concen- negative chloride ions to assist in repolarization. An inward neg-
tration is reduced. The narrow T tubules restrict the rapid diffu- ative current essentially means that only a small amount of K +
sion of K through the ion channels and into the extracellular

+
ions need efflux from the cell to repolarize the transmembrane
space. The restriction is a result of a build-up of positive charges potential back to the resting level following depolarization. The
that hinder continued efflux. Just as for nerve, the resting mem- increased CL conductance tends to reduce the total amount of
brane potential for muscle approaches the equilibrium potential K accumulation and minimize any chance of repetitive activity
+
for potassium, but is a bit off when the extracellular concentra- due to potassium ion build-up.
tion of potassium is reduced. The implication is that a second
ion is responsible for just a portion of the resting membrane po- Excitation-Contraction Coupling
tential. As for nerve, the second ion turns out to be sodium. Excitation
Voltage clamp experiments and ion channel poisons have docu- An action potential is initially produced at the end-plate
mented voltage-gated sodium and potassium gates in the sar- region, which then propagates along the muscle membrane and
colemma similar to those in unmyelinated nerve. 21
extends into the T tubule system. Once in the T tubule, the
action potential causes calcium release from the sarcoplasmic
Action Potential Propagation reticulum's terminal cisternae (Fig. 1-27). Although the
2223
As stated previously, an action potential traveling down a exact mechanism of calcium release is unclear, junctional feet
nerve can enter the end-plate region and result in a depolarization projecting from the cisternae to the T tubules presumably assists
Action potential
Figure 1-27. Sarcoplasmic calcium release.Action poten-

tial is shown propagating into the T tubule system carrying t h e
impulse into t h e depths of the muscle and producing C a r e - + +
lease from the sarcoplasmic reticulum.The cycling of calcium ions

is demonstrated following an action potential. (From Guyton AC:
Textbook of Medical Physiology, 9th ed. Philadelphia, W.B.
Saunders, 1996, with permission.)
in calcium release. The sarcoplasmic reticulum contains large

48
in sarcomere shortening: excitation/contraction coupling. In the
quantities of C a ^ that are released to bathe the myofibrils. The absence of ATP, the actin and myosin molecules remain in the
muscle continues to contract as long as C a is present. The C a
++ ++
joined position resulting in rigor mortis.
ions surround the myofibrils in sufficient quantities to facilitate
contraction for approximately 1/30 second. At the end of this
period, a C a pump rapidly sequesters the calcium ions back
++
CONCLUSION
into the sarcoplasmic reticulum reducing the concentration of
C a below that necessary for continued contraction (Fig. 1-27).
++
All practitioners wishing to practice electrodiagnostic medi-
cine must understand the fundamentals of nerve and muscle
Contraction physiology. During the course of an electrodiagnostic medicine
Once the calcium ions have been released from the sarcoplas- examination, the practitioner investigates either directly or indi-
mic reticulum, up to four C a ions are proposed to bind
+ +
rectly nerve and muscle resting membrane potentials and the
strongly to troponin C. Calcium ion binding apparently results
48
generation of action potentials. The manner in which nerve and
in a conformational change of the troponin complex that alters muscle respond to the investigative procedures of the practitioner
the relationship of tropomyosin to actin (Fig. 1-26). In the in both health and disease can only be fully appreciated through
process of changing position, the tropomyosin is thought to an understanding of nerve and muscle physiology. The electrodi-
expose "active" sites on the G-actin molecule. These active sites agnostic medicine principles discussed in the remainder of this
on the actin molecule are believed to be adenosine diphosphate text are based on the concepts delineated in this chapter.
(ADP) molecules. Although the exact mechanism of contraction
is unknown, the "walk-along" hypothesis proposes that as soon
as the active sites (ADP) on actin are uncovered, a linkage REFERENCES
forms between these active sites and the myosin cross-bridges
1. Ackerman M, Clapham DE: Ion channels:Basic science and clinical disease. N
because of a very high affinity of the myosin for the ADP. 22
Eng! J Med 1997;336:1575-1586.

When the linkage is complete, the myosin head tilts toward its 2. Agnew WS: Voltage-regulated sodium channel molecules. Annu Rev Physiol
arm and this motion drags the actin filament in a process re- 1984;46:517-530.
3. Aimers W, Stirling CE: The distribution of transport proteins over animal cell
ferred to as the power stroke. The energy driving the power membranes. J Membr Biol 1984;77:169-186.
stroke is derived from the previous joining of ATP that is 4. Alnaes E, Rahamimoff R: On the role of mitochondria in transmitter release from
cleaved into ADP and Pi owing to the ATPase activity of the motor nerve terminals. J Physiol 1975;248:285-306.
myosin head, thereby "cocking" this portion of the myosin mol- 5. Armstrong CM: Sodium channels and gating currents. Physiol Rev
1981;61:644-683.
ecule (refer to previous section entitled Extrafusal Fibers). 6. Barchi RL: Excitation and conduction in nerve. In Sumner AJ (ed): The Physiology
Following the power stroke, the previously cleaved ADP and Pi of Peripheral Nerve Disease. Philadelphia, W.B. Saunders, 1980, pp 1-40.
are now released, which allows another molecule of ATP to bind 7. Barnard EA, Dolly JO, Porter CW, Albuquerque EX: The acetylcholine receptor
and the ionic conductance modulator system of skeletal muscle. Exp Neurol
to the newly exposed site on the myosin head. Again, this ATP
1975;48:1-28.
is cleaved into ADP and Pi thus re-cocking the myosin and re- 8. Bernstein J: Investigation on the thermodynamics of bioelectric currents.
leasing it from the active site. Once in the ready position, it Pflugers Arch 1902;92:521-560. Translated in Kepner GR (ed): Cell Membrane
quickly binds to the next available actin site to again produce Permeability and Transport. Stroudsburg, PA, Dwoden, Hutchinson and Ross,
1979.
the power stroke. Continued iterations of this process bring the 9. Catterall WA: Structure and function of voltage-gated ion channels. Annu Rev
myosin filaments up to the Z line region, shortening the muscle. Biochem 1995;64:493-531.
It is important to recognize that the filaments themselves do not 10. Chiu SY, Ritchie JM: Potassium channels in nodal and internodal axon mem-
shorten but slide past each other thus resulting in contraction, branes of myelinated fibers. Nature 1980;284:170-171.
11. Chiu SY, Ritchie JM: Evidence for the presence of potassium channels in the in-
i.e., the sliding filament mode of muscle contraction. Muscle ternodal region of acutely demyelinated nerve fibers. J Physiol
contraction can continue only if repetitive action potentials 1981;313:415-437.
invade the T tubules, releasing C a from the sarcoplasmic retic-
++
12. Cole KS, Curtis HJ: Electric impedance of squid giant axon during activity. J
Gen Physiol 1939;22:649-670.
ulum. This in turn alters the troponin complex, thereby expos- 13. Darnell J, Lodish H, Baltimore D: Molecular Cell Biology. New York, Scientific
ing the actin's active site to myosin cross-bridges and resulting American Book, 1986, pp 275-284.
14. DelCastillo J, Engback L: The nature of the neuromuscular block produced by 41. Kuffer SW, Yoshikami D: The number of transmitter molecules in a quantum: An
magnesium. J Physiol (London) 1954;124:370-384. estimate from iontophoretic applications of acetylcholine at the neuromuscular
15. Desmedt JE: The neuromuscular disorder in myasthenia gravis. In Desmedt JE junction synapse. J Physiol 1975;251:465-482.
(ed): New Developments in Electromyography and Clinical Neurophysiology. 42. Latorre R, Alvarez O: Voltage-dependent channels in planar lipid bilayer mem-
Basel, Karger,1973, pp 241-304. branes. Physiol Rev 1981;61:77-150.
16. Doyle DA, Cabral JM, Pfuetzner RA, et al: The structure of the potassium 43. Lee AG: Functional properties of biologic membranes: A physical-chemical ap-
channel: Molecular basis of K conduction and selectivity. Science 1998;280:
+
proach. Prog Biophys Mol Biol 1975;29:5-56.
69-77. 44. Lloyd DPC: Neuron patterns controlling transmission of ipsilateral hindlimb re-
17. Elmqvist D: Neuromuscular transmission defects. In Desmedt JE (ed): New flexes in cat. J Neurophysiol 1943;6:293-315.
Developments in Electromyography and Clinical Neurophysiology. Basel, 45. "Macintosh FC: Formation, storage, and release of acetylcholine at nerve endings.
Karger, 1973, pp 229-240. Can J Biochem Physiol 1959;37:343-356.
18. Erlanger J, Gasser HS: Electrical Signs of Nervous Activity. Philadelphia, 46. Marban E, Yamagishi T, Tomaselli G: Structure and function of the voltage-gated
Philadelphia University Press, 1937. sodium channel. J Physiol 1998;508:647-657.
19. Fawcett DW: Bloom and Fawcett: A Textbook of Histology. Philadelphia, W.B. 47. Maxwell MH, Kleenan CR: Clinical Disorders of Electrolyte Metabolism. New
Saunders, 1986. York, McGraw-Hill, 1952.
20. Fettiplace R, Haydon DA: Water permeaBility of lipid membranes. Physiol Rev 48. McComas AJ: Neuromuscular Function and Disorders. Boston, Butterworth, 1977.
1980;60:510-550. 49. Miller C: Integral membrane channels: Studies in model membranes. Physiol
21. Goldman DE: Potential impedance and rectification in membranes. J Gen Rev 1983;63:1209-1242.
Physiol 1943;27:37-60. 50. Paintal AS: The influence of diameter of medullated nerve fibers of cats on the
22. Guyton AC: Textbook of Medical Physiology, 8th edition. Philadelphia, W.B. rising and falling phases of the spike and its recovery. J Physiol
Saunders, 1991, pp 38-50. 1966;184:791-811.
23. Hille B: Introduction to physiology of excitable cells. In Patton HD, Fuchs AF, 51. Paintal AS: Conduction in mammalian nerve fibers. In Desmedt JE (ed): New
Hille B, Scher AM, Steiner R (eds): Textbook of Physiology, 21st ed. Developments in Electromyography and Clinical Neurophysiology. Basel,
Philadelphia, W.B. Saunders, 1989, pp 1-80. Karger Press, 1973, pp 19-41.
24. Hodgkin AL, Rushton WAH: The electrical constants of a crustacean nerve fiber. 52. Paintal AS: Conduction properties of normal peripheral mammalian axons. In
Proc Soc Lond (Biol) 1946; 133:444-479. Waxman SG (ed): Physiology and Pathobiology of Axons. New York, Raven
25. Hodgkin AL, Katz B: The effect of sodium ions on the electrical activity of the Press, 1978, pp 131-144.
giant axon of the squid. J Physiol (Lond) 1949;108:37-77. 53. Pallotta BS, Wagoner PK: Voltage-dependent potassium channels since Hodgkin
26. Hodgkin AL, Huxley AF, Katz B: Measurement of current-voltage relations in and Huxley. Physiol Rev 1992;72:S49-67.
the membrane of the giant axon of Loligo. J Physiol 1952; 116:424—448. 54. Porter CW, Barnard EA: The density of cholinergic receptors and the end-plate
27. Hodgkin AL, Huxley AF: Currents carried by sodium and potassium ions postsynaptic membrane: Ultrastructural studies in two mammalian species. J
through the membrane of the giant axon of Loligo. J Physiol 1952; Membr Biol 1975;20:31-49.
116:449^72. 55. Ragsdale DS, Avoli M: Sodium channels as molecular targets for antiepileptic
28. Hodgkin AL, Huxley AF: The dual effect of membrane potentials on sodium drugs. Brain Res Rev 1998;26:16-28.
conductance in the giant axon of Loligo. J Physiol 1952; 116:497-546. 56. Rail W: Core conductors theory and cable properties of neurons. In Kandel ER
29. Hodgkin AL, Huxley AF: A quantitative description of membrane current and (ed): Handbook of Physiology, Vol. 1. Bethesda, MD, American Physiological
its application to conductance and excitation in nerve. J Physiol 1952; 117: Society, 1977, pp 39-97.
500-544. 57. Rasminski M, Sears TA: Internodal conduction in undissected demyelinated
30. Hodgkin AL, Keynes R: Active transport of cations in giant axons from Sepia nerve fibers. J Physiol 1971;22:323-350.
and Loligo. J Physiol 1955;128:28-60. 58. Ritchie JM, Rogart RB: The binding of saxitoxin and tetrodotoxin to excitable
31. Hodgkin AL, Horowicz P: Movement of sodium and potassium in single muscle tissue. Rev Physiol Biochem Pharmacol 1977;79:1-50.
fibers. J Physiol 1959;145:405-432. 59. Rogart RB, Stampfli R: Voltage-clamp studies of mammalian myelinated nerve.
32. Hubbard J I, Wilson DF: Neuromuscular transmission in a mammalian receptor In Culp WJ, Ochoa H (eds): Abnormal Nerves and Muscles as Impulse
in the absence of blocking drugs and the effect of D-tubocurarine. J Physiol Generators. Oxford, Oxford University Press, 1982, pp 193-210.
1973;228:307-325. 60. Sakmann B, Neher E: Single Channel Recording. New York, Plenum Press,
33. Hucho F, Tsetlin V: Structural biology of key nervous system proteins. J 1983.
Neurochem 1996;66:1781-1792. 61. Schnepp P, Schnepp G: [Analysis of peripheral nerve fibres in animals of differ-
34. Jewett DL, Rayner MD: Basic Concepts of Neuronal Function. Boston, Little, ent body size. II. Ratio of axon diameter to fibre diameter and internodal length.]
Brown and Co., 1984. Z Zellforsch Mikrosk Anat 1971 ;119:99-114.
35. Katz B, Miledi R: Propagation of electrical activity in motor nerve terminals. 62. Singer SJ: The molecular organization of membranes. Annu Rev Biochem
Proc Royal Soc (London) 1965; 1 6 1 4 5 3 ^ 8 2 . 1974;43:905-835.
36. Katz B, Miledi R: The measurement of synaptic delay, and the time course of 63. Stamplfli R: Overview of studies on the physiology of conduction in myelinated
acetylcholine release at the neuromuscular junction. Proc Royal Soc (London) nerve fibers. Adv Neurol 1981;31:11-23.
1965;161:483-495. 64. Unwin N: Acetylcholine receptor channel imaged in the open state. Nature
37. Katz B: Nerve, Muscle, and Synapse. New York, McGraw-Hill, 1966. 1995;373:37-43.
38. Keynes RD: Ion channels in nerve-cell membrane. Sci Am 1979;240:126-135. 65. Waxham MN: Neurotransmitter receptors. In Zigmond MJ, Bloom FE, Landis
39. Koester J: Resting membrane potential and action potential. In Kandel ER, SC, Roberts JL, Squire LR (eds): Fundamental Neuroscience. San Diego,
Schwartz JH (eds): Principles of Neural Science, 2nd edition. New York, Academic Press, 1999, pp 235-67.
Elsevier, 1985 pp 49-57. 66. Waxman SG, Foster RE: Ionic channel distribution and heterogeneity of the axon [
40. Koester J: Voltage-gated channels and the generation of the action potential. In membrane in myelinated fibers. Brain Res Rev 1980:2:205-234.
Kandel ER, Schwartz JH (eds): Principles of Neural Science, 2nd edition. New 67. Waxman SG: Action potential propagation and conduction velocity—new per-
York, Elsevier, 1985, pp 75-86. spectives and questions. Trends Neurosc 1983;6:157-161.
Chapter 2
E l e c t r i c a l Sources a n d
Volume Conduction
Dick F. Stegeman, Ph.D.
Machiel J. Zwarts, M.D., Ph.D.
CHAPTER OUTLINE
Volume Conduction Theory • Peripheral Nerve Observations • The Leading/Trailing Dipole

Volume Conductors and Electrophysiologic Recording Model • Differences in Volume Conductor Size and Tissue
Principles • Action Potential Source Generation • Local Resistances • The Termination of Excitable Tissue • A Change
Circuit Currents • Circuit Model of Waveform Morphology in Direction of Neural Propagation
• Local Circuit Currents in Volume Conductors • Waveform
Morphology in a Restricted-Volume Conductor • Waveform Waveform Morphology in Muscle
Morphologies in Large-Volume Conductors • Waveform Endplate Potentials • Miniature Endplate Potentials • Endplate
Asymmetry Spikes • Single Muscle Fiber • Motor Unit Potential
Morphology • Compound Muscle Action Potential • Far-Field
Waveform Morphology in Nerve Potentials * Fibrillation Potentials • Positive Sharp Waves
Sensory Nerve Action Potentials (SNAPs) • Clinical Recordings
• SNAP Morphology • Distance Effects on Dispersion and Pitfalls and Possible Sources of Error
Phase Cancellation • Far-Field Potentials • Evoked Potentials Stimulation * Recording
Nerves and muscles are bioelectric source generators. They

VOLUME C O N D U C T I O N THEORY
consist of multiple subcomponent fibers generating transmem-
brane action potentials through which ionic currents spread VOLUME CONDUCTORS AND ELECTROPHYSIOLOGIC
into the extracellular tissue. An extracellular recording from
RECORDING PRINCIPLES
these currents is typically obtained by an electrode placed on
the skin surface, or a needle electrode inserted through the Volume conduction theory describes the three-dimensional
skin and placed in close proximity to the nerve or muscle spread of electrical current (current = charge movement) in any
fibers. The ensuing waveform depends on the summation of conducting m e d i u m . It is governed by the laws of electro-
1419 3 3
individual electric current fields generated by the transmem- statics. Lorente de No made an extensive study of this subject in
brane action potentials of the excited neural or muscular tis- 1947. Due to the various physical shapes and the heteroge-
82
sues. These waveforms have unique shapes and sizes in both neous and anisotropic properties of the human body as a volume
normal and pathologic conditions. Any clinician practicing conductor, the solutions to this problem in terms of physics and
electrodiagnostic medicine should possess a thorough under- intuition are not trivial. The fundamental principles of volume
standing of action potential generation and why these poten- conductor theory may be understood by considering a simple
tials appear as they do to properly formulate a diagnosis. A example of a beaker containing a dissolved table salt solution
flawed understanding of waveform morphology may result in (NaCl). Sodium chloride dissolves and forms individual posi-
an erroneous medical impression. tive sodium ions (Na ) and negative chloride ions (Cl~). These
+
The subtleties of near-field and far-field waveforms as they ions are electrically charged and mobile in solution, and hence
relate to both innervated and denervated tissues are explored in capable of mediating an electric current. If we now connect two
more detail in the appendix to this chapter. wires to the positive and negative terminals of a battery and
27
, Current source mobile in the solution, they form current pathways in three di-
mensions between the two battery poles. In Figure 2-2 current
lines are shown schematically with an increasing and decreas-
ing amount of space separating them. The amount of separation
from one current line to the next depicts the current density
which is greater near the battery poles than further away. - 1419 82
We can use the beaker example to explore a number of con-

cepts important to the understanding of volume conduction. The
two battery poles essentially act as a dipole (+,-) where current
Electrode flows from the positive to the negative pole. Let us focus on
38
just the dipole with its accompanying current (Fig. 2-2).

- Extending out from the dipole at right angles to the current lines
is a set of lines referred to as isopotential lines. One isopoten-
38
Line of highest
tial line represents a constant potential or voltage magnitude
current density
along its entire length. The potential's magnitude associated
with a specific line decreases as one proceeds toward the zero
Current paths isopotential line midway between the two poles (Fig. 2-2).
The spatial gradient is defined as the rate of change of a po-
tential in a volume conductor as a function of distance. The 38
spatial gradient is high when differences in potential are large

Figure 2-1. Volume Conductor. A current source is immersed in between adjacent points. The concept of a spatial gradient can be
a beaker containing an electrolytic solution that serves as a good used to define different aspects of the potential distribution asso-
volume conductor. N o t e that the current lines fan out into t h e volume ciated with a current source in a volume conductor. The designa-
c o n d u c t o r as well as align directly b e t w e e n t h e t w o battery poles. tion near field defines a region of the volume conductor
(From De Lisa JA, Brozovich FV:Volume conduction in electromyogra- surrounding a current source where the spatial gradient from the
phy: Experimental and theoretical review. Electromyogr Clin N e u r o - current source is relatively high. - Potentials recorded in this
38 101
physiol 1983;23:651 -673, with permission.) portion of the volume conductor are called near-field potentials.
Signals detected at some distance from the current source where
place the bared tips in the beaker containing the salt solution, the potential gradient is low are known as far-field potentials. 38
the positive ions (cations) and negative ions (anions) are at- A number of electrodes can be placed at various locations in
tracted to the negative pole (cathode) and positive pole (anode) the volume conductor around a dipole current source (Fig. 2-2).
of the battery, respectively (Fig. 2 - 1 ) . Because the ions are
19
It can be inferred from the above that if two electrodes are lo-
cated on two different isopotential lines, a potential difference
will be detected. No potential difference will be detected be-
tween two electrodes on the same isopotential line, no matter
how far apart they are or how large the voltage magnitudes. In
Figure 2-2, none of the electrodes are on exactly the same
isopotential line. Therefore, a potential difference exists be-
tween any two electrode pairs. This observation suggests that a
potential difference, albeit small, in some electrode pairings
(montages), will be measured between any electrode combina-
tion shown. The potential difference between each electrode
pair, however, is not the same, and different amplitudes can be
anticipated with each different electrode montage.
Consider two electrodes positioned close together on differ-
ent isopotential lines, for example C and D in Figure 2-2, but lo-
cated relatively far from the dipole current source. These two
electrodes constitute a bipolar recording montage, i.e., two
electrodes in close proximity to each other with respect to an
electrical generator. - - - If the potential difference between
58 59 71 72
the two electrodes is negligible, as is the case far from the elec-
trical generator, the detected waveform will also be of small
size. In other words, if the spatial gradient of the voltage distrib-
ution is low, there is a very small difference in potential detected
Figure 2-2. Dipole current/voltage distribution. Distribution of by two electrodes relatively close together.
current lines about a dipole generator in a good volume conductor. A bipolar electrode montage located somewhat closer to the
T h e c u r r e n t lines extend between t h e positive and negative poles, dipole generator can detect a larger potential difference between
while t h e isopotential lines are perpendicular t o the current lines. Five the two electrodes (Fig. 2-2, electrodes A and B). In this in-
recording electrodes a r e positioned a t various locations within t h e stance, the potential difference is relatively large because the
volume conductor. Electrodes A and B are in a region of a high current spatial potential gradient is high near the impulse generator. As
spatial gradient but C, D, and E are in a relatively lower spatial gradi- a result, the voltage difference between the two electrodes (A
ent. (From Dumitru D.Jewett DL: Far-field potentials. Muscle N e r v e minus B) is relatively large and yields a larger potential than C
1993; 16:237-254, with permission.) minus D. Bipolar recordings, therefore, are useful in documenting
Chapter 2 ELECTRICAL SOURCES A N D VOLUME C O N D U C T I O N — 29
the presence or absence of electrical activity in the localized with reversed polarity (swapping electrodes at the amplifier).
region of the electrodes when they are in the near-field, but on So, the distance between the electrodes and the exact positions
different isopotential lines. This fact can be taken advantage of have a pronounced influence on the measured signals.
when attempting to minimize stimulus artifact by rotating the Furthermore, the physical dimensions of the electrodes are im-
stimulus anode about the cathode, thereby attempting to alter portant. It can be generally assumed that the electrode records
the stimulator's isopotential lines so that both recording elec- the average of the potential field under its surface. The mea-
trodes are located on the same isopotential line of the stimulus sured signal can be influenced considerably when large-size
artifact (see Chapter 3). electrodes are in a region with high spatial gradients.
If one of the electrodes close to the generator source (A or B) In electrodiagnostic medicine, an electrical signal is usually
is used with an electrode further away (C, D, or E), then A or B displayed on the instrument's screen. If the net voltage differ-
are said to be "referenced to" C, D, or E. This type of recording ence between the two recording electrodes is negative, the in-
is referred to as a referential recording montage or a monopo- strument's trace moves upward (negative up) or downward
lar recording montage. - The electrode A or B, positioned in
70 72
(positive up). The actual direction of travel is determined by the
a high spatial gradient region, is "referenced" to an electrode lo- manner in which the investigator connects the electrodes to the
cated in an area with a low spatial gradient. A larger potential instrument and is merely a matter of convention. Practitioners
difference exists between A or B connected to C, D, or E com- of electrodiagnostic medicine use the convention of "negative
pared to a bipolar recording between either A and B, or C and up" and "positive down." Unless otherwise stated, this conven-
D. This is because a referential electrode montage usually com- tion is used throughout this book. The just discussed volume
pares a region of high to one of low voltage, i.e., after having conduction as explained by the beaker salt solution can also be
passed many isopotential lines, the result is a large measured viewed as a filtering process. The advantage of this approach is
potential difference. So, referential montages may record poten- that filtering is usually rather familiar to the electrodiagnostic
tials with larger amplitudes than bipolar montages if one of the physician. The conducting medium acts as a spatial low-pass
electrodes is positioned in the same location for both types of (high-frequency) filter resulting in smoothing of the signals.
recordings, e.g., montage A - E (e.g., a monopolar needle elec- This implies that the further away the recording electrodes are
trode) compared to montage A - B (e.g., a concentric needle from the source, the lower the amplitudes are with a more pro-
electrode).8
nounced effect on the high-frequency parts of the original
Electrode terminology can be rather confusing. One electrode signal. The electrode size and interelectrode positions also act
of a pair is typically referred to as an "active electrode" and the as spatial filters. The spatial filter characteristics turn into an
other as a "reference electrode." The term "active" is used by equivalent temporal filtering as soon as, and merely when, one
convention when this electrode is located in the neighborhood deals with the travelling aspect of an action potential. It is im-
of the electrical generator where the spatial gradient is high. The portant to realize, however, that although increasing the distance
reference electrode is usually placed where the spatial gradient between source and recording electrode deeply influences the
is lower than that for the active electrode. The active electrode is shape of the waveform, time relations are not influenced. Thus,
plugged into the (+) amplifier port that magnifies the signal un- measuring a travelling nerve action potential from nearby with
changed, while the reference electrode is connected to the (-) needle electrodes or at large distances with a surface electrode
amplifier port that magnifies and inverts the signal (see Chapter (at a level perpendicular to the direction of the action potential)
3). The two magnified signals are summated (active minus ref-
30
gives different shapes and amplitudes, but nearly identical la-
erence) and displayed or used for further analysis. This proce- tencies for the different waveform subcomponents.
dure is called differential amplification (see Chapter 3). All
recording montages can be said to consist of an active and refer- Action Potential S o u r c e Generation
ence electrode, although in bipolar recordings the active elec- All living cells possess the unique characteristic of a resting
trode is compared to a reference electrode, which is also membrane potential with the intracellular region negative (ap-
relatively close to the impulse generator. proximately - 90 mV) compared to the extracellular space. - 2 67
One may also consider a referential far-field montage (Fig. 2¬ The cell membrane is responsible for developing and sustaining
2). Electrode C or D may be connected to electrode E to com- this transmembrane voltage difference because of its semiperme-
prise a far-field referential montage. Since both electrodes able nature. Most cells contain a high intracellular concentration
(either C or D connected to E) are in a region of low spatial gra- of positive potassium ions (K ) and negatively charged proteins.
+
dient, it is hard to define which of the two is the "reference" In the resting state, the semipermeable membrane allows intra-
electrode. Finally, the distinction between the "active" and "ref- cellular potassium ions to exit the cell "down" a high intracellu-
erence" electrodes also becomes blurred in bipolar montages lo- lar to low extracellular concentration gradient. - This process
2 67
cated in a region where the spatial gradient does not change continues until it is balanced by an inward electrical attraction
significantly (Fig. 2-2, C and D). Therefore, the distinction is from the intracellular protein anions that cannot exit the cell. A
inadequate in many occasions. The terms E-l (electrode 1) and net negative charge results when a dynamic equilibrium is estab-
E-2 (electrode 2), or Gl (grid 1) and G2 (grid 2) may be better lished between the forces driving K out of the cell (concentra-
+
descriptors for recording electrodes where the observed wave- tion gradient) and an inwardly directed force (electrical gradient)
form is a result of E-l minus E-2, formerly "active" minus "ref- from the anions attracting the K into the cell.
+
erence." The terminology E-l and E-2 is preferred over Gl and The resting membrane potential is influenced to a small
G2 since the latter terms refer to vacuum tube amplifiers, which degree by the relatively impermeable positive sodium ions
are no longer used as electrodiagnostic equipment. (Na ). There are two strong forces driving sodium into the cell.
+
As is obvious from the above explanation, the same active The first is the concentration gradient, high extracellular and
bioelectric source can be recorded as a flat line (both electrodes low intracellular N a concentration, and the second is the nega-
+
on an isopotential line), with high amplitude (one electrode in a tive intracellular potential, i.e., electrical gradient. Although rel-
high spatial gradient and the other in a low spatial gradient), or atively impermeable, small quantities of N a do leak into the
+
cell tending to alter the resting membrane potential. A steady- back to - 90 mV (Fig. 2-3A). - The cell, therefore, produces a
48 89
state situation with a constant resting membrane potential is positive monophasic intracellular action potential beginning at
maintained by actively pumping K into the cell and pumping
+
- 90 mV, increasing to about + 40 mV, and then returns to - 90
out N a from the cell (the sodium-potassium ATPase pump). -
+ 2 62
m y 99 j h e action potential's appearance as characterized by its
This steady-state voltage approaches the equilibrium potential amplitude, duration, rapid rise, and relatively slow baseline
of K with a minor modification due to the influence of N a (see
+ +
return is directly dependent upon the temporal aspects of N a +
Chapter 1). and K activation/inactivation (Fig. 2-3A). For a more detailed

+ 87
Nerve and muscle cells are known as excitable cells because description of the resting membrane potential and the action po-
transient action potentials can be induced via an active process tential generation in both myelinated and unmyelinated nerve,
that is based on the behavior of the voltage-gated ion channels see Chapter 1.
in the membrane. If the resting membrane potential is depolar-
ized to the threshold level (approximately - 65 mV), following Local Circuit Currents
appropriate chemical, mechanical, or electrical stimuli, such an The impermeable protein anions located within excitable tis-
action potential is produced. - - T h e action potential arises
2 19 67
sues attract extracellular positive ions, primarily N a , because
+
from an alteration in voltage-gated channels controlling the per- of its abundance, and cause them to align along the extracellular
meability of N a and K . At the threshold voltage, the voltage-
+ +
surface of the cell's membrane. In effect, there is a series of neg-
gated sodium channels open allowing N a to enter and further
+
ative and positive charges aligned along the cell's length sepa-
depolarize the cell, thereby inducing more sodium gates to open rated by the membrane (Fig. 2-3B). During an action potential,
in a positive feedback manner driving the intracellular voltage inwardly flowing positive N a ions proceed longitudinally
+
toward + 55 mV (sodium equilibrium potential; Fig. 2-3A). This along the interior of the axon. These additional intracellular
process of depolarization lasts about 0.1 ms in mammalian positive charges balance a portion of the intracellular anions,
nerves and travels along the nerve fiber. - " The cell's resting
89 90 3
which attracted the extracellular N a . The transmembrane elec-
+
membrane potential depolarizes from - 90 mV to approxi- trical attraction for N a is subsequently reduced allowing those
+
mately + 40 mV. A delayed increase in K permeability lasting

66 +
extracellular N a ions increased freedom to move away from the
+
about 0.4 ms in mammalian unmyelinated nerve in conjunction immediate vicinity. The increased permeability of the depolar-
with inactivation of the sodium gates restores the membrane ized membrane permits those now even more mobile N a ions +
Monophasic Potential
0
+
Figure 2-3. Action potential generation in an unmyeli-
nated nerve. A,Alterations of N a conductance above thresh-
+
old results in depolarization, while increased K conductance

+
helps establish t h e resting membrane potential.The net result of

Oil t h e s e t w o processes is a monophasic positive intracellular
action potential. B,The transmembrane dipolar arrangement of
intracellular and extracellular ions is depicted. C, Local circuit
Saline currents distributed within a thin film of saline surrounded by a
H31SB j-
) p o o r volume conductor (mineral oil). N o t e how t h e local cir-
cuit currents are confined t o the saline and "hug" the membrane
surface. Four sequential electrode locations sample t h e extra-
Oil cellular potential within t h e saline film. In a good volume con-
ductor, such as t h e body, the local circuit current expands away
from t h e nerve surface. D , A monophasic extracellular wave-
form resulting from the local circuits that corresponds t o t h e
• H.^.^......^ . intracellular action potential. (From Dumitru D:Volume conduc-
( ~ — f + ++H tion: T h e o r y and application. In Dumitru D (ed): Clinical
Electrophysiology. Philadelphia, Hanley & Belfus, 1989, pp
665-681, with permission.)
Chapter 2 ELECTRICAL SOURCES A N D VOLUME C O N D U C T I O N — 3I
to enter the cell intensified through the already open N a chan- +

current, results in a potential rise (positive potential). So, a
nels. This sequence of extracellular N a entering the cell and
+
movement against current flow (opposite the local circuit cur-
neutralizing the transmembrane attraction toward extracellular rent's direction) produces a positive trace deflection on the in-
N a ions thereby allowing them to move av/ay from the mem-
+
strument, while travel in the direction of current flow (in the
brane's surface and also to enter the cell is referred to as a local same direction as the local circuit current) results in a negative
circuit (Fig. 2-3C; also see Chapter l ) . - This local circuit
2 67
screen trace deflection.
leads to relatively more positive charges intracellularly and The concept of moving with or against the direction of the
fewer positive charges extracellularly adjacent to the depolar- current may be considered from two perspectives. First, let us
ized portion of membrane. That portion of membrane where ex- imagine that an action potential is propagating along a nerve or
tracellular N a ions are no longer held to the surface of the
+
muscle toward a non-moving electrode. In this instance, the
membrane, but are free to enter the cell through the local circuit local circuit currents associated with the action potential will
current constitutes the current source. The region of membrane pass by the electrode. The direction of current flow with respect
through which the sodium ions enter the cell (open voltage-de- to the electrode, in or out of the cell, will determine the direc-
pendent gates) is called a current sink. - The source current
2 67
tion of the upward or downward deflection. Second, we may
tends to depolarize the cell's adjacent membranous regions sur- consider an action potential with its local circuit currents direc-
rounding the current sink by reducing the local surface positive tion of travel suspended or "frozen" for an instant in time. T h e
charge and reversing the intracellular to extracellular voltage re- electrode can then be moved past the local circuit currents to de-
lationship. The self-sustaining nature of these processes results scribe what it measures at that moment for which the action po-
in action potential propagation. tential is suspended. The results from both scenarios are
The current sink from a propagating action potential permits equivalent with respect to the action potential's configuration.
positive sodium ions to extend not only into that portion of Whether the electrode moves by a non-moving action potential,
nerve or muscle previously depolarized, but also into the region or the action potential propagates past a non-moving electrode,
of the tissue about to be depolarized i.e., a bidirectional intra- the net waveform result is the same. For discussion purposes, it
cellular current flow. Extracellularly, the sodium ions are is often easier to consider the local circuit currents of an action
moving into the sink from portions of the volume conductor potential at one moment in time and move an electrode through
both preceding and following the traveling action potential. In the local circuit currents. The assumption is that the local circuit
effect, there are two intra-/extracellular local circuit currents* currents do not change with action potential propagation along
associated with a propagating action potential (Fig. 2-3C). - 2 67
the course of the excitable tissue examined.
This observation leads to the common source description of a
travelling action potential consisting of a "source-sink-source"
arrangement. 94
LOCAL CIRCUIT CURRENTS IN VOLUME
CONDUCTORS
Circuit Model of W a v e f o r m Morphology
Consider a simple battery circuit with one resistor (Fig. 2¬ W a v e f o r m Morphology in a Restricted-Volume
4 ) . The direction of the current in the circuit is from the posi-
39
Conductor
tive to the negative battery pole across the resistor. The current An excitable cell surrounded by a thin film of saline can be
across the resistor results in a voltage drop. Point A (Fig. 2-4A) placed in a liquid with a high electric resistance, such as mineral
is at a higher voltage than point B. When proceeding in the di- oil (Fig. 2-3C). Mineral oil is a poor conductor because it con-
rection of current across the resistor, the recording instrument tains few ions and, therefore, has a negligible electrical conduc-
records a negative deflection. Moving from the negative to positivity (i.e., a high resistance). If an action potential is induced in
tive battery terminal across the resistor, opposite to the flow of a muscle or nerve fiber suspended in mineral oil, the local cir-
cuit currents are restricted to the thin film of saline solution. 21
All current is then flowing along the immediate surface of the

cell and does not expand into the oil suspension. A referential
electrode montage with the E-l electrode located in close prox-
©-A/VV-G imity to the tissue's surface records a monophasic negative
A B
waveform (Fig. 2-3, C and D).
A *-B
The resistor circuit model can provide an understanding of
0" the extracellularly recorded waveform's appearance arising
from the intracellular monophasic positive potential (- 90 m V
CRT CRT
to + 30 mV, and back to - 90 mV) in poor volume conductors
(Fig. 2-3, C and D). Remember that the local circuit currents
are restricted to the thin film of saline surrounding the nerve
Figure 2-4. Circuit current flow. A, Simple circuit diagram with a and turn in toward the negative sink as soon as their mobility is
c u r r e n t (I) flowing from t h e positive t o negative battery terminal increased. In moving the electrode from left to right, it is ini-
through a resistor (R) from point A t o B. B, Moving in the direction of tially traveling in the direction of the leading source current,
c u r r e n t flow across t h e resistor (points A t o B) results in a voltage i.e., toward the current sink. The battery analogy tells us that
drop with a corresponding trace deflection in the negative o r upward when the electrode movement and the current flow are in the
direction. Proceeding in the direction opposite t o current flow (points same direction, a negative trace deflection can be anticipated
B t o A) produces a positive o r downward deflection of t h e t r a c e . (Fig. 2-4B). The waveform associated with the action potential
(From Dumitru D: Volume conduction: T h e o r y and application. In begins with an initial negative deflection that grows in magni-
Dumitru D (ed): Clinical Electrophysiology. Philadelphia, Hanley & tude as the electrode gets closer to the negative sink. The a m -
Belfus, I989,pp 665-681,with permission.) plitude of the negative deflection is maximized when the
electrode is over the proximal extent of the negative sink (Fig. W a v e f o r m Morphologies in Large-Volume Conductors
2-3, C and D). We will now surround the nerve or muscle cell with a sub-
Continued electrode movement results in meeting a current in stantially larger amount of saline and investigate the action po-
the opposite direction coming from the trailing current source tential's extracellular morphology. To simplify the discussion,
heading into the current sink. It is important to note the relation- again consider the action potential to be suspended at an instant
ship between the direction of electrode movement and current di- in time, and move a referential recording electrode through the
rection, i.e., now traveling in opposite directions. According to action potential's local circuit currents. In a good volume con-
the resistor circuit model, when the current and electrode direc- ductor, the local circuit current lines no longer immediately turn
tion of travel are opposite, a positive screen deflection ensues. In in toward the current sink, but first travel away from the current
our example, the trace deflection is indeed now heading in the sink and fan out an appreciable distance from the cell's surface
less negative direction (Fig. 2-3, C and D). The current density of into the volume conductor, and then head back into the current
the trailing local circuit current is less dense and has a greater spa- sink (Figs. 2-2 and 2-5).
5433
Additionally, as one approaches the
tial extent accounting for the trace's comparatively slower return current sink from any direction, the current density increases
to baseline. As the trailing source current diminishes, the instru- because the source currents are "focused" into the sink region.
ment's trace settles back to baseline. The net result in a poor The extracellularly recorded waveform morphology in a good
volume conductor is a monophasic negative extracellular wave- volume conductor also may be explained with the assistance of
form produced by a monophasic positive intracellular potential. 21
the resistor circuit model (Fig. 2-5, A - D ) . The electrode
32
Nerve
A B C D
Level I
Level 2
Level 3
Figure 2-5. Action potential field distribution. Computer-simulated current field distribution surrounding a suspended action potential
propagating from right t o left. Levels 1—3 represent sequential electrode positions and their accompanying waveforms (A through C) at progres-
sively greater distances from the nerve's surface.The t h r e e levels of recording w e r e simulated at 0.5, 5.0, and 30 mm from the nerve. (From
Dumitru DrVolume conductionrTheory and application. In Dumitru D (ed): Clinical Electrophysiology. Philadelphia, Hanley & Belfus, 1989, pp
moving from left to right first detects local circuit current lines To the right of the second zero isopotential line, the current
moving out and away from the membrane's surface opposite the and electrode are still traveling in opposite directions and the
direction of electrode passage. A positive deflection results. The trace continues below the baseline. As the electrode continues
magnitude of the positive deflection grows as the electrode gets to proceed to the right, it is again perpendicular to the current
closer to the current sink. The peak of the positive deflection is flowing away from the membrane's surface and a third and final
achieved when the electrode is situated just perpendicular to the inflection point is noted. For the remainder of the action poten-
local circuit current lines, i.e., when the current is neither travel- tial, the local circuit current flows out from the membrane away
ing against nor with the direction of electrode movement. The from the current sink prior to curving back toward the mem-
ions are not moving away from or toward the sink, but straight brane. In this region, the electrode and current flow move in the
up relative to the membrane's surface. When the electrode is same direction and the trace proceeds in the negative (less posi-
over this perpendicular aspect of the current flow, the trace stops tive) direction toward the baseline. The current density then di-
moving in the positive direction. Just to the right of this perpen- minishes and the trace settles back to baseline. The end result is
dicular zone, the electrode is still in a region of high current a triphasic waveform: positive-negative-positive (Fig. 2-5 D).
density, but progressing in the direction of current flow. The A crucial notion is that the same intracellular positive
electrode in our example now moves into a region being less monophasic action potential produces different extracellular
positive than previously; this is reflected in the instrument's waveform morphologies depending upon the surrounding
trace now moving in the less positive direction and beginning to medium (Figs. 2-3D and 2-5D, Levels 1-3). In a restricted con-
approach the baseline (Fig. 2-5A). ducting medium, a monophasic negative extracellular potential
With continued movement, the electrode encounters a zero is recorded, whereas in a large volume conductor, a triphasic
isopotential line, and the trace crosses the baseline. As the elec- initially positive extracellular waveform is detected.
trode travels to the right of the zero isopotential line, it continues This statement may be further conceptualized by considering
to detect a current moving in the same direction. The trace rises our previous example of a nerve coated with a thin salt solution
above the baseline. A negative potential is recorded (Fig. 2-5B). (Fig. 2-6). Initially, the local circuit currents are confined to the
The second inflection point is reached when the electrode is again immediate region surrounding the nerve's extracellular surface.
perpendicular to the direction of current flow. This time, however, The isopotential lines demarcating the current sink and its two
the current is flowing into the membrane as opposed to that of the source current regions, two zero isopotential lines, essentially
first inflection point when it was flowing away from the mem- hug the membrane's surface (Fig. 2-6A). If the amount of
brane's surface. Immediately to the right of this perpendicular volume conductor surrounding the nerve is moderately in-
electrode/current flow relationship, the electrode again encoun- creased, the local circuit current expands into the added region
ters current flowing toward it, signifying a region of diminishing and both zero isopotential lines move away from the mem-
negative potential, thereby describing a less negative deflection branes surface with slightly more elevation for the leading
(Fig. 2-5C). With electrode advancement, the instrument's trace zero isopotential line because of the increased current density
continues in this direction until the second zero isopotential line (Fig. 2-6B). The electrode now detects a portion of the initial
is encountered, at which time the baseline is again reached. positive current flow resulting in a positive deflection. Next, a
Direction of action potential propagation
Figure 2-6. Volume conductor influence on waveform morphology. A, Nerve surrounded by a thin film of conducting medium similar
to that described in Figure 2-3. N o t e how the zero isopotential lines are markedly curved so that only t h e negative sink portion of the action p o -
tential is capable of being detected by the recording electrode.The end result is a monophasic negative potential. B, Adding m o r e conducting
medium allows t h e local circuit currents t o expand further from the membrane's surface resulting in a concomitant elevation of both, but prefer-
entially the leading zero isopotential lines.A biphasic positive/negative potential ensues. C, Significantly increasing the surrounding volume conduc-
t o r permits the local circuit currents t o move comparatively further from the nerve's surface fully exposing the leading and trailing positive source
currents t o the recording electrode.The end result is t h e expected triphasic waveform. (Modified from d e W e e r d JPQVolume conduction and
electromyography. In N o t e r m a n s SLH (ed): Current Practice of Clinical Electromyography.Amsterdam, Elsevier, 1984, pp 9-28.)
negative deflection is observed. The low density of the trailing action potential's characteristics for the first several action po-
local circuit current with only a mild elevation of the zero isopotentials generated. - 109 110
tential line allows a minimal, if any, trailing positive current to

be detected. The end result in this case is an initially positive
biphasic potential. Finally, if a sufficient amount of volume con- WAVEFORM MORPHOLOGY IN NERVE
ductor surrounds the nerve, the zero isopotential lines extend
away from the membrane's surface to allow complete recording SENSORY NERVE ACTION POTENTIALS (SNAPS)
of both the initial and trailing positive source currents, produc-
ing a triphasic (positive-negative-positive) potential (Figs. 2-5 Clinical Recordings
and 2-6C). The waveform arising from a referentially recorded digital
As discussed above, the morphology of the extracellularly SNAP is triphasic (Fig. 2-7B). The observed SNAP potential is
8
recorded waveforms depends upon the direction of travel for the the summation of triphasic waveforms from thousands of individ-
electrode with respect to the current as well as the characteris- ual nerve fibers each contributing a triphasic waveform. A nerve
tics of volume conductor present. T h e amplitude of the ob- fiber's action potential approximates 0.5 ms in duration. This 0.5
served potential is directly related to the current density. - The 19 20
ms intracellular monophasic action potential may result in an ex-
current density decreases with an increase in distance from the tracellular triphasic waveform of about 2.0 ms in duration when
generator. Sequential electrode placement at distances further recorded in a good volume conductor because of the longitudinal
from the current generator results in potentials with an earlier local circuit current's spatial expanse preceding and following the
and smaller positive deflection, and a reduced negative peak negative sink. If this potential propagates with a conduction ve-
8
amplitude displaced toward the geometric center of current sink locity of 50 m/s, it has a spatial extent within the volume conduc-
(Fig. 2-5, levels 1-3). The terminal positive phase is also re-
54
tor extending along the nerve for about 100 mm {NCV = D/T;
duced in magnitude. These observations are a result of the 50,000 mm/1000 ms = D/2.0 ms; D = 100 mm}. As the sensory
wider, less dense, and more symmetric current field formed in nerve is composed of thousands of fibers with different conduc-
the volume conductor further from the current generator. tion velocities, the result is an overlapping of multiple waveforms,
each with slightly different spatial distributions. The composite
73
Waveform Asymmetry SNAP can have a duration of 3.0 ms and more (Fig. 2-7).
The triphasic appearance of extracellularly recorded poten-
tials with a referential electrode montage near the generator's S N A P Morphology
source demonstrate a rather distinctive asymmetry (Fig. 2-5D). The morphology of antidromic and orthodromic bipolar
The triphasic waveform's first positive phase is larger in magni- SNAP waveform recordings is typically biphasic and does not
tude and shorter in duration than the terminal positive phase. have the aforementioned triphasic waveform (Fig. 2-7A). The 8
Also, the time from the first positive peak to the subsequent neg-
ative peak, rise time, is shorter than that from this negative peak
to the second positive peak. An intracellular nerve action poten-
tial has a duration (T) approximating 0.5 ms - - and can prop- 89 90 113
agate at a nerve conduction velocity (NCV) of 50 meters/second

(m/s), which produces a negative sink with a spatial extent (D) of
about 25 millimeters (mm) {NCV = D/T; 50 m/s = D/0.5 ms; D
= 25 m m } . Sodium activation (open N a channels) has a dura-+
tion of about 0.1 ms, which represents a spatial expanse along

the nerve of approximately 5 mm. Therefore, when the action
potential is considered from a spatial perspective, most of the
N a channels are open for the first 5 mm of the current sink
+
while the remaining 20 mm (0.4 ms) of the current sink encom-

passes repolarization. This observation suggests that the local
circuit current lines are spatially directed and compressed about
the initial portion of the propagating current sink; this is where
the N a gates are opened. The initial region of current compres-
+ 54
sion causes the local circuit current lines to fan out and flow
away from the beginning portion of the current sink compared to
the terminal area. This difference in current density is directly
21
reflected in the slope of the rise and fall of the intracellular action Figure 2-7. SNAP morphology. Median nerve stimulation at t h e
potential (Fig. 2-3A). Once the gates have closed, the potential is wrist with an antidromic sensory recording montage. Active (A) and
reduced by an effluxing current - (e.g., potassium) or a
211 67
reference (R) recording electrodes are located on the third and fifth
sodium back l e a k . Therefore, the asymmetric extracellular
312
digits. A, Bipolar recording results in a biphasic SNAP. B, Active elec-
triphasic waveform results from the slope characteristics of the trode on the proximal third digit referenced t o the fifth digit results in
intracellular action potential, which depends on the spatial distri- a triphasic SNAP. C . W h e n the active electrode is located on t h e fifth
bution of open sodium channels. digit and referenced t o the distal third digit, an inverted and slightly de-
Of interest, in a number of myelinated nerves, repolarization layed (compared t o waveform in B) triphasic potential is observed. D,
does not occur through a potassium ion efflux, but through a Electronic summation of traces B and C yields the potential recorded
sodium ion back leak. - - - The potassium current is replaced
3 12 84 93
in trace A. (From Dumitru DrVolume conduction:Theory and applica-
by an effluxing sodium current. The blocking of potassium volt- tion. In Dumitru D (ed): Clinical Electrophysiology. Philadelphia, Hanley
age-gated channels produces no significant alteration in the & Belfus, 1989, pp 665-681, with permission.)
biphasic, negative-positive potential is a result of the bipolar It is also possible to predict the optimum intereiectrode sepa-
recording technique and not a violation of the previously pre- ration to maximize the biphasic potential's amplitude in a bipo-
sented extracellular action potential properties. Biphasic SNAP lar recording montage. The critical factor in this instance is the
waveforms can be easily understood by investigating bipolar rise time, baseline to negative peak, of the biphasic potential.
and referential recording montages for median nerve stimula- The recording electrodes must be located at a distance greater
tion at the wrist (Fig. 2-7). Initially, two ring recording elec-
8
than the spatial extent represented by the rise time's duration.
trodes can be placed on the third digit (bipolar montage) with The rise time of most SNAPs approach 0.8 m s , which repre- 74
the E-l electrode (active electrode) proximal. When the median sents a longitudinal extent of 40 mm for a conduction velocity
nerve is excited at the wrist, a clearly biphasic response is ob- of 50 m/s {50,000 mm/1000 ms = D/0.8 ms; D = 40 mm}. If the
tained (Fig. 2-7A). Next, the distal E-2 recording electrode (ref- two recording electrodes are separated by a distance less than
erence electrode) is relocated to the fifth digit that is innervated 40 mm, components recorded by one electrode will overlap
by the ulnar nerve, thereby precluding median nerve activity with those recorded by another electrode resulting in mutual
from influencing the potential at this electrode. This allows one cancellation (differential amplification) of those components
to record a median nerve response from the E-1 electrode with- and producing a potential with a smaller amplitude. Continued
out any contribution from E-2. When the median nerve is again intereiectrode separation toward 100 mm now approaches a ref-
stimulated, the observed waveform is triphasic, positive-nega- erential montage and the two triphasic potentials in Figure 2-7
tive-positive (Fig. 2-7B). One can conclude, therefore, that the B and C become detectable. The potential from E-l precedes
bipolar recording in some way produces the biphasic and not that from E-2, as the two electrodes no longer interact.
the expected triphasic waveform.
Continuing with the median nerve example, we can now Distance Effects on Dispersion and Phase Cancellation
place the E-l electrode on the fifth digit and keep the E-2 elec- We can return to the bipolar antidromic recording example to
trode on the third digit's distal portion. This referential montage investigate the effect different stimulus sites have on SNAP con-
should record electrical activity only from the E-2 electrode. figuration. Activating the median nerve at the wrist yields the
Following median nerve stimulation, an inverted (E-2 elec- expected biphasic SNAP. Relocating the electrical stimulator
trode), somewhat smaller and slightly delayed triphasic poten- sequentially at more proximal activation sites results in a series
tial compared to the previous triphasic waveform is observed of SNAPs with a progressively declining amplitude, and area to
(Fig. 2-7C). In essence, this is the same potential as that the negative phase (Fig. 2-8). This observation may be under-
recorded in the referential montage of Figure 2-7B with several stood if one recalls that the duration of an intracellular action
exceptions. Specifically, the potential is (1) smaller because potential of a single mammalian nerve approximates 0.5 ms and
fewer nerve fibers are present more distally near the fingertip, results in an extracellular triphasic waveform of about 2.0 ms in
(2) a little more delayed temporally as more distance is tra- a volume conductor. A bipolar recording of this extracellular
versed in reaching the finger tip, and (3) inverted because it is waveform results in a biphasic potential with a negative dura-
fed to the amplifier through the inverting port (E-2). tion close to that of 0.5 ms. The negative spike of the individual
In the bipolar recording example above (Fig. 2-7A), the wave- axon (about 0.5 ms) is approximately one-third of the whole-
form from the fingertip is electronically added to the waveform nerve SNAP (roughly 1.5 ms). 88
recorded at the digit's proximal portion. The waveform at the fin- Another important factor is the range of conduction velocities
gertip approximates that from the more proximal location, but it between the fastest- and slowest-conducting fibers, which can
is inverted and slightly delayed in time. Adding one to the other be more than 25 m/s. - This range in velocities results in tem-
27 28
produces a biphasic potential. It is the slight temporal delay and poral dispersion and consequently in asynchronous arrival of
smaller amplitude of the distal potential that prevents mutual single-fiber action potentials at the recording electrode." In
cancellation. The initial positive phase is eliminated and the neg- combining the observation of relatively short action potential
ative peak amplitude is somewhat smaller and occurs slightly durations with the relatively large amount of temporal disper-
earlier in time than for the proximal triphasic potential. This sion, a significant amount of overlap between the negative and
process can be simulated by electronically adding what the E-2 positive phases of individual axons occurs (Fig. 2-9). The over-
electrode alone recorded (Fig. 2-7C) to the waveform obtained lap of positive and negative peaks produces phase cancellation
solely by the E-l electrode (Fig. 2-7B), resulting in what was resulting in a lower amplitude and area than might be expected
originally detected by the bipolar recording (Figs. 2-7D and 2- from the thousands of single nerve fibers excited. It can be seen
7A, respectively). Therefore, bipolarly recorded waveforms are
8
that as the distance between the stimulus and recording sites in-
usually biphasic because of the cancellation effect of some simi- creases, more temporal dispersion and phase cancellation would
lar components being eliminated in differential amplification. be anticipated, thus producing progressively longer SNAP dura-
In a bipolar montage, the recorded waveform only achieves tions and declining amplitudes and areas for SNAP potentials.
its full triphasic morphology and maximum amplitude when the This is indeed the finding in clinical studies. ~ -
73 75 88
E-2 electrode contributes minimally to the response, i.e., the

electrode montage approaches a referential recording. The pre- FAR-FIELD POTENTIALS
ceding example of an action potential with an extracellular
waveform 2.0 ms in duration propagating at 50 m/s demon- Evoked Potentials
strates that the minimum intereiectrode separation required to A far-field potential is simply the waveform detected when
optimally resolve the triphasic waveform is just over 100 mm. both electrodes are located in the far field of the current's spatial
In this scenario, where the E-2 electrode is located 100 mm gradient arising from an action potential. Far-field potentials
from E-l along the nerve's length, it does not record any portion were first described in brainstem auditory evoked responses
of the neural impulse's waveform until it has completely passed (BAERs), which are waveforms generated from the auditory
the E-l electrode, thereby acting as a referential montage. pathway secondary to acoustic stimulation. - - Actually, not
60 61 86
However, few individuals possess a digit longer than 100 mm. all BAER potentials are far-field potentials because wave I is
Median nerve stimulation
Figure 2-8. Distance effects on CMAP/SNAP waveforms. Simultaneous recordings of t h e CMAP from t h e thenar eminence and SNAP
from t h e second and third digits following stimulation of t h e median nerve at the palm, wrist, elbow, and axilla. N o t e that t h e CMAP declines min-
imally, but there is a rather noticeable reduction in the SNAP'S amplitude and area. (From Kimura J, Machida M, IshidaT, e t al: Relation between size
of compound sensory o r muscle action potentials, and length of nerve segment. Neurology 1986;36:647-652, with permission.)
generated in the auditory nerve, and as a near-field, it is absent system. Interestingly, for wave II it has been elegantly shown
from the contralateral ear. For the other waves, it is true that the that it arises at the junction between cerebrospinal fluid and
neural generators are deep within the skull, and the recording brainstem due to the sudden change in conductivity of the
electrodes are located far from the current source on the skull's volume conductor. For the later components, the origin is
85
vertex and earlobes. Seven potentials are usually produced in

36
probably a dipolar field due to the dipolar source activity of
response to auditory clicks given to one of the ears. Lesions large clusters of neurons.
compromising the neural components of the auditory pathway, Far-field potentials also were described in somatosensory
e.g., multiple sclerosis or an acoustic neuroma, result in either evoked potential (SEP) recordings. - - - Following exci-
16 1718 23 24
delayed or absent waveforms. BAERs are, therefore, useful

98103
tation of a peripheral nerve, electrical responses can be
in the diagnosis of pathology involving the central nervous recorded over not just the distal limb, but also proximally in
Sensory A c t i o n Potentials
Individual Summated
responses response
Figure 2-9. Phase cancellation model. A model

depicting phase cancellation effects resulting from in-
creased distance and t h e results of differences b e -
tween t h e fast and slow conducting nerve fibers.With
distal stimulation, t h e phases from t h e individual
SNAPs summate, w h e r e a s with increased distance
t h e phases separate enough by t h e time they reach
t h e recording e l e c t r o d e s t o summate less o r even
cancel. (From Kimura J, Machida M, Ishida T, e t al:
Relation between size of compound s e n s o r y o r
muscle action potentials, and length of nerve seg-
ment. Neurology 1986;36:647-652, with permission.)
/ft*
Chapter 2 ELECTRICAL SOURCES AND VOLUME C O N D U C T I O N — 37
the supraclavicular area, the cervical spine, and from the scalp A series of electrodes can be located along the course of this
overlying the somatosensory cortex of the brain. In short, an nerve and into the distal extent of the second digit beyond the
SEP may be thought of as simply conduction along a long anatomic distribution of the nerve. An additional electrode may
nerve pathway for both the peripheral and central nervous be located on the fifth digit and should show no response when
system. Stimulation of the median nerve produces four posi- the radial nerve is electrically activated. First, the series of adja-
tive potentials preceding the arrival of the impulse at the so- cent recording electrodes can be arranged in a sequential bipo-
matosensory cortex. The interesting finding regarding the
23 2 4
lar montage. The radial nerve is excited and a biphasic radial
far-field potentials is that they were all positive in polarity and nerve SNAP is obtained from each pair of electrodes (Fig. 2-10;
had the same respective latencies despite different electrode left panel). The SNAP'S peak latency increases with each elec-
locations. This observation suggested that these far-field po- trode pair as the induced impulse propagates distally. These are
tentials were not localized phenomena, but extended an appre- completely anticipated results. Secondly, each electrode along
ciable distance and appeared instantaneously throughout the the course of the nerve may be connected in turn to the fifth
body. Because the far-field potentials might facilitate diagno- digit comprising a referential montage. In this instance, radial
sis of lesions affecting the somatosensory pathway, attempts nerve stimulation produces the anticipated triphasic waveform
were made to define their site of origin. Using SEP far-field with a progressively increasing peak latency. However, two
potentials to localize neural pathway pathology is a difficult non-moving positive and negative potentials with the same re-
task because far-field generator sites do not always have a spective latency were noted despite different electrode locations
direct neurophysiologic basis, and variable waveform mor- (Fig. 2-10; right panel). These referentially observed potentials
phology and latency can occur with alterations in body posi- are by definition far-field potentials because they had a distribu-
tion. It is possible to gain an appreciation of far-field
4 2 7 1 1 , 4
tion outside of the nerve's anatomic course, and did not change
potential properties from a number of interesting peripheral latency despite different electrode locations. Comparing when
70
nerve experiments. the far-field potentials first appeared with the anatomic location
of the electrode determined the anatomic site for some of these
Peripheral N e r v e Observations far-field generators. The forearm/hand junction (wrist) and
Let us consider an experiment involving the superficial radial metacarpophalangeal joint were suggested as sites of generation
nerve in the upper limb. This nerve is subcutaneous and sub-
70
for at least two of the observed far-field potentials. Additional
ject to both stimulation and recording from the mid-forearm peripheral nerve recordings with referential montages of both
into the proximal interphalangeal region of the first three digits. the hand and remainder of the upper limb produced a theory of
- v ^
Referential Recording
Figure 2-/0. Peripheral nerve far-field potentials. Sensory nerve potentials across t h e hand along the second digit in a normal subject
recorded antidromically after stimulation of t h e superficial radial nerve 10 cm proximal t o t h e radial styloid. In a bipolar recording (left), the initial
negative peaks, N I (arrow pointing up), showed a progressive increase in latency and reduction in amplitude distally and no response w a s
recorded beyond - I. In a referential recording (right), biphasic peaks, PI -NI and PII-NII (arrows pointing down), showed greater amplitude distally,
with a constant latency irrespective of the recording sites along the digit.The onset of PI extended proximally t o t h e recording electrodes n e a r
the wrist (small arrows pointing down), whereas Pll first appeared at t h e base of t h e digit. (From Kimura J, Mitsudome A,YamadaT, Dickins Q S :
Stationary peaks from a moving source in far-field recordings. Electroencephalogr Clin Neurophysiol 1984;58:351-361, with permission.)
far-field potential production that should apply to both the cen- long as one of the electrodes is on the opposite side of the bound-
tral and peripheral nervous systems as well as to muscle tissue. ary or measuring some portion of the voltage difference's near-
It has been postulated that as a traveling wave of depolariza- field, a far-field potential can be detected. A successful model to
tion reaches (1) an heterogeneity of a volume conductor, e.g., qualitatively explain far-field potentials has been developed and
wrist or shoulder, (2) transitions in extracellular conductivities of is known as the leading/trailing dipole m o d e l . 2663 6 3
the volume conductor, (3) a change in the anatomic orientation

of a traveling action potential, and/or (4) the origination and/or T h e Leading/Trailing Dipole Model
termination of excitable tissue, a voltage difference arises at An action potential can schematically be described as consist-
these "boundary zones." - - - - The voltage difference, in the
26 63 68 69 71
ing of a centrally located current sink flanked by two current
case of property changes in the volume conductor as an underly- sources, i.e., source-sink-source (+ - +: a tripole). The intracellu-
ing cause (see below) also referred to as boundary or junctional larly directed current entering through the sink region extends
potentials, temporarily exist and extend instantaneously both proximally and distally. There are two complete local circuit
throughout the volume conductor for the time an action potential currents in effect splitting the sink into two negative regions (Fig
passes through the transition region arising from one of the 2-3C). - - - The traditional source-sink-source, therefore, can
2 14 26 66
above four conditions. As long as a referential montage is placed also be thought of as a source-sink/sink-source (+ - , - +: a linear
so that the two electrodes are encompassing the relevant bound- quadrupole or double dipole). - This double dipole is referred
25 92
ary, a far-field potential will be recorded. The electrode loca- to as a leading and trailing dipole depending on the direction of
tion's importance cannot be underestimated. If both of the propagation. Each dipole produces a dipole moment, which is
electrodes are in the far-field, but on the same side of the bound- the product of the individual charge magnitude and the charge
ary zone and record the same potential, a far-field waveform is separation. The two dipole moments of an action potential have
not observed. It is eliminated in differential amplification. As opposite directions and balance each other. When an action po-
64
tential encounters one of the four boundary conditions previously

described, the leading/trailing dipole moments no longer balance.
It is the potential difference associated with a dipole moment im-
balance that is recorded as a far-field potential.
Differences in V o l u m e Conductor Size and Tissue

Resistances
Let us consider an induced action potential as it traverses two
cylindrical volumes (Fig. 2-11). Initially, the propagating action
potential is completely contained within the smaller cylinder
and its associated current field is also confined in this space.
The leading and trailing dipole moments exactly balance each
other with a net zero potential in the far field. No voltage is
measured between the two recording electrodes E-l and E-2.
The instrument's trace remains flat (Fig. 2-11 A). As the action
potential continues to propagate, it will reach the transition zone
between the two volumes. For this discussion, we will consider
that point in time when the entire leading dipole is in the large
cylinder while the total trailing dipole remains in the smaller
cylinder (Fig. 2-1 IB). The leading dipole's current field is also
preferentially located in the large cylinder, whereas the trailing
dipole's current field is confined within the smaller cylinder.
Realize that the two dipoles should be considered as current
sources. The current strength is not influenced by the extracel-
Figure 2-/ / . Size changes producing far-field potentials. An lular situation. That means that the potential they set up is en-
action potential represented by a leading/trailing dipole propagates tirely dependent on the resistance in the extracellular medium
along a nerve from the small to the large cylinder. A,The nerve is con- (Ohm's law: voltage = current x resistance). The resistance in
62
tained within the small cylinder and the differential amplifier does not the smaller cylinder is higher than that in the large cylinder. In
record a potential with the trace remaining flat. B,An Instant in time the situation of Fig. 2-1 IB the two dipoles are not completely
when the leading dipole is in the large cylinder while the trailing dipole balancing each other in the extracellular potential. In Figure 2¬
is still within the small cylinder.The amplifier records a potential differ- 1 IB an arbitrary magnitude is recorded between the E-l and the
ence from an imbalance in the dipole m o m e n t s and t h e trace de- E-2 electrodes in the far field. Continued impulse propagation
scribes a positive deflection. C, Continued neural propagation again results in both the leading and trailing dipoles now completely
allows the leading and trailing dipole m o m e n t s t o balance when the entering the large cylinder (Fig. 2-11C). At some distance
nerve is completely within the large cylinder and the amplifier records beyond the boundary zone (several radii of the small cylinder),
no potential with a flat trace. D , T h e leading dipole has been extin- the leading and the trailing dipole set up the same (lower than in
guished at t h e nerve's end while the trailing dipole is still present. An the small cylinder) potential difference between E-l and E-2.
imbalance in dipole moments is again r e c o r d e d by the amplifier and The potentials set up by the two dipoles are again balanced
the trace moves in t h e negative direction. E . W h e n the impulse is no since the extracellular resistances with respect to the leading
longer present, the trace returns t o baseline.Two far-field monophasic and trailing dipoles are equal again.
potentials of opposite polarity is the net result in the above combina- Instead of changing the size of the volume conductor, one can
tion of boundary conditions. also consider a situation in which the specific resistance of the
extracellular medium is changing, e.g., the transition of an action

potential propagation through muscle tissue, which suddenly en-
counters connective tissue with a much higher resistance. Then
also an effective dipole can emerge as a consequence of applying
Ohm's law for the extracellular field. In such situations of chang-
ing extracellular conditions (size changes, resistance changes), it
would be wrong to state that the dipole moments of the propa-
gating source become unbalanced, such as the situation depicted
in Fig. 2-11, leading to an effective dipole moment. It is not the
sources that become unbalanced, rather an imbalance exists in
the extracellular domain encountered during passage of the
dipoles. This is the reason the term "virtual" effective dipole
source has been proposed for situations in which the extracellu-
lar medium produces a far-field potential.' 01
T h e Termination of Excitable Tissue

As previously stated, a far-field potential can be expected to Figure 2-12. Nerve bending producing a far-field potential. A
arise whenever an action potential encounters a boundary. It is 26
nerve located within a cylindrical volume conductor is bent 180° back
important to remember that an action potential has a longitudi- on itself.Two electrodes are placed in t h e far field. A,The action p o -
nal expanse along the nerve. This means that the leading portion tential is completely contained on the straight segment of nerve with a
of the action potential will encounter the termination of ex- balance of dipole m o m e n t s . T h e differential amplifier records no n e t
citable tissue before the trailing aspect. One may anticipate that potential with an associated flat trace. B,At s o m e instant, the leading
when such a situation arises, the leading dipole will be absent dipole will have turned t h e nerve's corner while the trailing dipole r e -
and no longer capable of balancing the trailing dipole. 64
mains on t h e previous segment of nerve. T h e dipole moments will
We can return to the above cylinder example and follow the summate t o produce a negative potential that is reflected in an upward
action potential as it reaches the nerve's termination. Again, for deflection of the trace. C, T h e trailing dipole then propagates around
discussion purposes, we will consider only that instant in time the 180° turn and again balances the leading dipole with the trace r e -
when the leading dipole is no longer present on the nerve and turning t o baseline.A monophasic negative far-field potential is the o b -
all that remains of the action potential is the trailing dipole (Fig. served waveform.
2 1 ID). As it is no longer present, the leading dipole now gives
:
a zero potential between the E-l and E-2 electrodes. A compar-

atively large negative potential is described by the trace because nerve's reversal. One of the nerve's ends is electrically activated
the potential recorded between the El and E-2 electrodes is now and an action potential propagates toward the bent region. The
i completely caused by the trailing dipole, without any balance nerve's length is assumed to be sufficient to completely contain
j from the leading dipole (Fig. 2-1 ID). Continued action poten- the longitudinal extent of the action potential. Again, the action
! tial propagation also results in dissipation of the trailing dipole, potential is represented by a leading and trailing dipole. On the
returning the trace to baseline (Fig. 2-1 IE). Note that reversing straight segment of a nerve, the leading and trailing dipole mo-
the E-l and E-2 electrode locations would result in monophasic ments cancel each other (Fig. 2-12A). A net zero potential is
potentials of opposite polarity to the ones described. Further, if measured between the E-1 and the E-2 electrodes.
the action potential began in the large cylinder and crossed into As the neural impulse continues to propagate, there comes a
the smaller one, a far-field potential with the same morphology point when the leading dipole has turned the corner of the bent
and polarity would be produced provided the recording elec- nerve while the trailing dipole is just about to reach this angle.
trodes remained unchanged. However, the large negative wave- Again, for the sake of simplicity, the point of maximum dipole
form would obviously occur earlier in time compared to the imbalance is chosen to clarify far-field potential production (Fig.
relatively smaller positive waveform. Note that in the above sit- 2-12B). In this case, the two current sinks are oriented in the same
uation of action potential termination, the respective dipole mo- direction toward E-l and the two current sources are pointing at
ments of the source did change because the leading dipole E-2. A negative potential is the result of recording between the E-
extinguished prior to the trailing dipole. Therefore, this is an ex- 1 and the E-2 electrode (Fig. 2-12B). As a result, a relatively large
ample where a "real" (and not a "virtual") effective dipole negative far-field deflection is described by the trace. When the
moment temporally generated a far-field potential. trailing dipole turns the corner, the net dipole moments again bal-
ance and the trace declines to baseline again (Fig. 2-12C). A
A Change in Direction of Neural Propagation monophasic negative potential is the end result from a bent nerve
Clinical observations of SEP and peripheral nerve studies with the above-noted electrode montage. A change in direction of
have shown that abducting the arm or flexing/extending the fin- the propagating action potential can, therefore, result in a far-field
gers produces changes in the latency and morphology of the potential. The case of bending the nerve 180° was used because it
recorded far-field potentials. - - Computer simulations and
25 71 114
represents the maximum dipolar imbalance. In reality, if the nerve
animal experimentation demonstrate that altering the direction is bent to any angle, a far-field potential will be recorded, but of a
of neural propagation can either produce new or alter existing smaller amplitude than that obtained for a nerve bent back on
far-field potentials. -
26 63100
itself. Smaller angles allow some portion of the two dipole mo-
We can place a nerve in a cylindrical volume conductor and ments to still cancel each other resulting in far-field potentials
angle it back on itself 180° at the nerve's midportion (Fig. 2¬ with less magnitude than that for an angle of 180°. 37
12). The E-1 and E-2 recording electrodes are located at some
100
It is important to recognize the relationship between the ori-
distance in the volume conductor on opposite sides of the entation of the recording electrodes to the boundary condition
(bent nerve) producing the far-field potential. If the E-l and E-2 capable of reaching the muscle membrane's threshold level. As
electrode locations are reversed, they will record potentials with a result, the depolarization is localized to the endplate region
the same magnitude but of opposite polarity to the above exam- and represents a miniature endplate potential (MEPP).
ple. This observation implies that the polarity of the far-field po- Clinically, multiple MEPPs are usually observed with an intra-
tentials depends on the orientation of the electrodes and not the muscular recording electrode and the sound is referred to as
direction of impulse propagation. This principle should apply
100
endplate noise or a "sea shell murmur." Numerous MEPPs are
to all far-field-causing phenomena. As in the case of the action detected despite a firing frequency of 0.2 Hz per endplate be-
potential termination, we here again have a "real" temporally ef- cause the recording electrode is relatively large with respect to
fective dipole in the source characteristics causing the far field. the endplate region and detects multiple endplates firing simul-
taneously (Fig. 2-13B).
WAVEFORM MORPHOLOGY IN MUSCLE Endplate Spikes

A second waveform that can be detected with an E-l elec-
ENDPLATE POTENTIALS trode placed in the endplate region is relatively short in duration
( 3 - 4 ms), of moderate amplitude (100-200 jiV), irregularly
Miniature Endplate Potentials firing, and usually but not always biphasic with an initial nega-
An E-l electrode located in the endplate region can record tive deflection, i.e., an endplate spike. A potential that is
9
two distinct waveforms. One waveform that can be observed is biphasic with an initial negative phase is produced when a cur-
a short (1-2 ms), small (10-50 pV), irregularly occurring (about rent sink originates in the vicinity of the E-l electrode and then
once every 5 seconds per axon terminal), monophasic negative propagates away (Fig. 2-14). The terminal source current results
waveform. Considering source and volume conduction charac-
9
in the observed positive phase. A propagating potential implies
teristics, one may suggest that for a potential to be monophasic that the local circuit currents are of a sufficient magnitude to
and negative, the current sink would have to start and finish reach threshold and become a self-sustaining propagating single
within the E-l electrode's recording area. The potential origi- muscle fiber potential.
nates at the electrode's location, but does not propagate away Endplate spikes are believed to be a single-muscle-fiber dis-
(Fig. 2-13A). Local currents generated in excitable tissues that charge induced by the E-l electrode irritating the terminal axon
do not propagate are considered subthreshold. innervating that muscle fiber. It is hypothesized that a portion
9
Subthreshold potentials have indeed been documented in a of the recording electrode abuts against the terminal nerve twig
muscle fiber's endplate region and are believed to arise form the as it approaches the endplate region. The electrode's mechani-
spontaneous release of one or only a few presynaptic vesicles cal force deforms the terminal axon and causes a localized de-
containing acetylcholine (see Chapter I). - The acetylcholine
44 46
polarization that propagates to the endplate, releasing sufficient
that diffuses across the synaptic cleft causes the acetylcholine acetylcholine to produce a suprathreshold endplate potential.
receptor to open, thereby permitting sodium ions to enter the Because the electrode's tip is in close proximity to the endplate
muscle fiber. The resultant depolarization from this small quantity
of acetylcholine is insufficient to produce an endplate potential
A
A
o
+
( 0
( o
( 0
Figure 2-/3. Miniature endplate potentials.Volume conduction
theory explanation of t h e miniature endplate potential's origin. A, A Figure 2-14. Endplate spike potential.The recording electrode is
recording electrode is located over t h e single muscle fiber's endplate placed in the endplate region while the shaft of the electrode depolar-
and records the spontaneous release of neurotransmitter.The depo- izes the terminal axon. A, A suprathreshold depolarization is induced
larization is insufficient t o reach threshold. As a result, the current sink in the endplate.The initial current sink associated with the endplates
does not leave the electrode's location. A monophasic negative trace depolarization is detected, which produces an upward trace deflec-
deflection is the end result. B.The relatively large recording electrode tion. B, As the action potential travels away from the endplate, the ter-
detects MEPPs from multiple endplates.Two endplate spikes are also minal source currents are recorded with a downward trace deflection.
shown as these t w o potentials often occur together. A biphasic negative-positive potential is the net result.
potential (IAP). In muscle, an action potential is approximately

4 to 20 times longer than in nerve; the repolarization process is
particularly prolonged. - - -45
Aside from this longer dura-
52 53 55,83
tion, the concept of a current sink surrounded by two source

currents (source-sink-source) remains the same. Bioelectric
source and volume conductor descriptions therefore also can be
used to predict muscle action potential waveforms that might be
observed under various conditions. Because of the fact that the
local current during the repolarization phase lasts longer and is
more spread out than during the depolarization phase, one may
anticipate that the terminal source current will be expanded,
thereby diminishing its density producing a potentially small
third phase. A triphasic waveform with a small terminal phase
should then be recorded from an extracellular E-l electrode
placed adjacent to a propagating single muscle fiber action po-
tential at some distance from the endplate region (Fig. 2-16).
Triphasic potentials with a somewhat reduced terminal positive
phase are recorded clinically. - Single-muscle-fiber record-
43 96
ings also demonstrate that in addition to triphasic potentials

(large positive/large negative/small positive phases), one also
may record biphasic initially positive or biphasic initially nega-
tive potentials.
It also has been suggested that triphasic single-muscle-fiber
waveforms may appear with a comparatively prominent third
phase the closer an electrode is placed to the endplate region. 94
Figure 2-/5. Triphasic endplate spike. If the recording electrode

is positioned such that its shaft irritates t h e terminal axon but t h e
recording surface is within t h e endplate region, an action potential
originates at t h e neuromuscular junction (upper trace) and travels
toward and past t h e recording electrode, producing a triphasic end-
plate spike.The initial and terminal source currents as well as the sink
current are sequentially detected, producing a triphasic potential.
region, a negative current sink is the initial portion of the extra-

cellular action potential recorded and produces an associated
negative waveform deflection. The single-muscle-fiber action
potential then propagates away from the endplate, and the elec-
trode detects the two trailing source currents resulting in a net
biphasic, initially negative potential (Fig. 2-14). Triphasic end-
plate spikes also may occur if the E-l electrode induces an
action potential in the terminal axon but the electrode's record-
ing surface is a short distance from the endplate (Fig. 2-15). The
action potential's source-sink-source currents are then observed.
Endplate spikes and MEPPs are frequently observed together as
they arise from the same region (Fig. 2-13B). An alternative
theory to explain these potentials state that they arise from the
muscle spindles. These contain intrafusal muscle fibers (nuclear
chain fibers) capable of conducting propagated action poten-
tials. Modulation of the firing rate of these potentials by various
maneuvers seems to support this hypothesis. More study is nec- Figure 2-/6. Single muscle fiber discharge morphology.The
essary to resolve these contradictory explanations. 91
single muscle fiber can have a triphasic appearance if it is recorded a t
s o m e distance from t h e endplate zone but away from the muscle's
SINGLE MUSCLE FIBER tendinous insertions. An action potential originates at the neuromus-
cular junction and propagates toward and past t h e recording elec-
The single-muscle-fiber potential, like a nerve action poten- t r o d e . T h e two source and single sink currents are recorded similarly
tial, depends upon the characteristics of the intracellular action to the triphasic endplate spike.
Because the terminal source current is effectively compressed negative trace deflection suggests that the current sink is pass-
between the current sink and the endplate zone, it must "fit" into ing by the electrode. A complete absence of a terminal positive
a smaller area that may tend to increase the trailing source's cur- phase implies that the final source current never passed by the
rent density. An increased current density usually results in a recording electrode. This would suggest that a biphasic initially
larger potential. The decreased density of the trailing source positive potential could occur if a muscle action potential pos-
current further along the fiber results in a comparatively smaller sessed leading source and sink currents but not a terminal
final positive phase. The combination of a small terminal posi- source current with respect to the recording electrode. This pos-
tive phase and baseline noise may at times prevent recording the sibility might occur if an action potential propagated up to, but
third phase allowing the waveform to appear biphasic. - - 30 43 96
not past, the E-1 electrode because the electrode damaged the
Considering bioelectric source and volume conduction con- fiber rendering it incapable of sustaining an action potential past
cepts, a biphasic, initially negative potential could be produced if the injury point. A waveform recorded from an injured muscle
the E-l recording electrode were located in the area of muscle fiber may produce a biphasic initially positive potential. A
where the current sink is generated. In this situation, the elec- second situation with similar results might occur if the record-
trode will first detect the intracellularly directed N a current pro-
+
ing electrode is placed in proximity to the musculotendinous
ducing a negative trace deflection. As the current sink progresses region. - Again, the source and sink currents would approach
1 41
bidirectionally out into the muscle fiber, the recording electrode and reach the recording surface of the electrode. The terminal
observes the two terminal source currents and records a positive source current, however, would dissipate as soon as the current
deflection. The final potential recorded is a biphasic potential sink was extinguished as it encountered the termination of ex-
with an initial negative phase. We have already described poten- citable tissue (muscle's tendon).
tials with a similar appearance arising because of terminal nerve
irritation from the electrode as opposed to voluntary contraction, MOTOR UNIT POTENTIAL MORPHOLOGY
i.e., endplate spikes. Placing a recording electrode in the end-
plate zone, therefore, and eliciting a voluntary contraction results Anatomy. One anterior horn cell gives rise to a peripheral
in the expected biphasic single-muscle-fiber potentials. - 43 96
nerve fiber that splits into multiple terminal sprouts, each of
A positive deflection of the trace suggests that the source cur- which innervate one muscle fiber (Fig. 2-17). The anterior horn
rent is approaching the recording electrode. The subsequent cell, its peripheral nerve, and all the single muscle fibers sup-
plied by that nerve is referred to as a motor unit. When the
21
anterior horn cell fires, or the axon that arises from it is stimu-
lated, all of the muscle fibers belonging to that motor unit depo-
larize. The electrical activity from all these muscle fibers
summate to produce a motor unit action potential (MUAP).
The anatomic distribution of the terminal axon sprouts with re-
spect to the muscle fibers they innervate is quite interesting and
particularly relevant to the morphologic characteristics of the
MUAP.
The length of an individual terminal axon sprout from the
point it divides to the endplate is quite variable for each muscle
fiber innervated. As a result, the spatial extent of the endplate
region from one motor unit approximates 5 mm longitudinally
along the muscle. - Additionally, the muscle fibers of a single
4-6 50
motor unit are randomly distributed in an oval territory approxi-

mately 4 - 6 mm in diameter (Fig. 2-17). The random distribu-
tion implies that the single muscle fibers may be in groups of
different numbers or singly arranged within the 4-6 mm. Five to
ten different motor units may share this area. The muscle fibers
constituting a single motor unit are randomly interspersed with
the muscle fibers of the other motor units sharing its oval terri-
tory. When considering the morphology of a motor unit poten-
tial with respect to volume conduction, one must consider the
(1) geometric arrangement of individual fibers not only to each
other but neighboring fibers as well, (2) maximum spatial dis-
placement between the two ends of a motor unit's endplate
region, (3) length of the terminal axon sprouts, (4) relationship
of the recording electrode to the various groups of motor unit
fibers (Fig. 2-17 A - D ) , (5) conduction velocities of terminal
axons and muscle fibers, and (6) neuromuscular junction trans-
mission times.
Amplitude/Rise Time. The morphology of a MUAP can be
Figure 2-17. Motor unit territory. Muscle fibers belonging t o o n e described in terms of its amplitude (maximum peak-to-peak
m o t o r unit are randomly distributed within a 6-mm oval territory. trace displacement), rise time (temporal aspect of a potential's
Four separate MUAP waveforms (all representing the same m o t o r peak: duration of initial positive to subsequent negative peak or
unit) are recorded from four (A-D) different locations close t o (A) o r baseline to negative peak), duration (departure from and return
within (B-D) the m o t o r unit. to baseline), and number of phases (baseline crossing plus one)
(Fig. 2-18). In volume-conducting tissue, the potential's am-

84
plitude declines with increasing distance from the current gen- \ /

erator because the surrounding muscle and its supportive \ *
tissues impedes those aspects of the potential that change 1 1 Phases
rapidly. - In other words, the tissue acts as a (spatial) low-

95 96
sf: Turns
Amplitude
pass frequency filter (see above). As a result, the peak-to-peak
50
MUAP's amplitude is believed to arise from fewer than 1 2 4 1 0 7

Baseline / Satellite ^ .
\
and possibly just from 1 or 2 single muscle fibers within 0.5
mm or less of the electrode's recording s u r f a c e . 437796
The
MUAP's rise time is generated from the closest fibers to the
recording electrode and typically occurs over a relatively short r
>$/ w
time span. In recording MUAPs, the usual practice is to locate
the E-l electrode as close as possible to a group of muscle Initial
Portion
Rise
Time
Terminal
Portion
fibers belonging to one motor unit; this is achieved by mini- IS, iv. K,
^
|P
mizing the rise time to less than 0.5-1.0 ms.
2 Duration „
Duration. The MUAP's duration depends upon the (1)
shortest and longest length of terminal axon from the point it
Spikt
^
Total Duration
separates from the parent nerve to the endplate, (2) conduction 1
velocities of the terminal axons and muscle fibers with respect
Figure 2-18. MUAP morphology. Schematic representation of a
to the recording electrode, and (3) most, if not all, of the muscle
single MUAP waveform with the various subcomponents delineated.
fibers contributing their individual waveforms. " For simplicity,
4 7
let the E-l electrode be located some distance away from both
the endplate zone and the muscle's tendinous termination. An the electrode, the positive deflection grows in magnitude until
action potential that propagates down a nerve enters the termi- its negative sink arrives, at which point a negative deflection is
nal axon sprouts that depolarize the single muscle fibers. The observed. Continued action potential propagation results in the
first muscle fiber to depolarize is the one with the shortest dis- electrode observing the action potential terminal positive phase,
tance between the point of separation from the parent nerve and which terminates when the action potential extinguishes at the
its endplate. When this muscle fiber depolarizes, a triphasic musculotendinous junction. This entire process takes the above-
waveform begins traveling along the muscle within the volume noted 13.5 ms. A slight degree of temporal dispersion for the
conductor. The longest distance between the division of the different action potentials comprising the MUAP increases the
nerve into terminal axons and an endplate may signify the last duration slightly, e.g., to 15 ms. If we introduce various noise
waveform to be activated. The remaining muscle fibers with components (environmental/instrument) and use a gain of 100
various intermediate lengths of terminal axons represent activa- pV/div, the MUAP duration will decline to about 10 ms. This is
tion of muscle fibers between the above two extremes. The roughly equivalent to the MUAP durations observed clinically
action potential of each single muscle fiber then propagates in the biceps brachii muscle. From this explanation it can be in-
with a specific conduction velocity between 2 and 4 m/s in ferred that muscles with shorter fiber lengths should have
normal conditions. The multiple action potentials along the
5-7
shorter durations than muscles with longer muscle fibers. T h e
activated muscle fibers belonging to the motor unit activated actual situation is somewhat more complicated than just d e -
travel along the muscle with the individual triphasic extracellu- scribed. For a more complete explanation of MUAP duration,
lar waveforms interacting with each other. At any time, there is consult the appendix to this chapter.
a different phase addition and cancellation effect from the over- Phases. As previously stated, the single muscle fiber usually
lapping of multiple positive and negative phases. The summated has a triphasic appearance when recorded outside the endplate
waveform produced at any instant is slightly different because zone, and away from the musculotendinous junction. The volt-
each terminal axon and each single muscle fiber has a different ages from all of the single muscle fibers belonging to one motor
conduction velocity. The relationship between the muscle unit summate to yield a MUAP that is also usually triphasic:
fiber's waveforms changes constantly because of the spatial and positive-negative-positive. This voltage summation does not
5 6
temporal aspects of both the terminal nerve and muscle fiber always produce a smooth result and small serrations or turns
conduction velocities, and length of the terminal axon to the can occasionally be seen as part of a MUAP's major phase (Fig.
various endplates. 2-18). A phase is defined as the part of a MUAP between two
Perhaps the best way to first conceptualize a MUAP's dura- baseline crossings. Normal MUAPs are considered to have
83
tion is to consider a 100-mm muscle fiber. Our recording elec- four phases at the most. MUAPs with five or more phases are
trode will be located at some location between the endplate and called polyphasic potentials. Recordings of multiple MUAPs
musculotendinous junction. Therefore, the action potential will from normal muscle can have 12-35% polyphasic potentials de-
propagate over a hemi-fiber length of 50 mm and this is tjie pending on the type of recording electrode used (see Chapter
length of muscle of relevance to the MUAP. If we assume a con- 3 ) . ° If one moves the electrode in small increments, passing
duction velocity of 3.7 m/s for the single muscle fiber's action through the motor unit territory while observing a continuously
potential, it will take about 13.5 ms for this action potential to firing MUAP, the morphology of the MUAP may change con-
traverse the 50-mm hemi-fiber length. Let us also initially siderably. The MUAP may increase or decrease in the total
5
assume a recording instrument without noise. Under these con- number of phases or the rise time may increase or decrease.
ditions, as soon as the endplate depolarizes, the positive initial These observations can be explained by considering the geo-
source current is measured as a smaller or larger initial positive metric arrangement of the motor unit. Recall that the single
potential shift over the entire muscle fiber length. An initially pos- muscle fibers of a particular motor unit are randomly located
itive deflection is recorded. As the action potential approaches within an oval territory (Fig. 2-17A). Small groups of muscle
4 4 — PART I FUNDAMENTAL PRINCIPLES
200 / i V above factors with respect to MUAP morphology, a single

normal MUAP can have multiple appearances depending on the
location of the recording electrode within the motor unit's terri-
tory. A nice illustration of the complex arrangement of the
motor unit and the resulting variability of the recorded wave-
forms can be seen from scanning EMG results. With a stepping
motor connected to a EMG needle the territory of a triggered
motor unit can be explored (Fig. 2-19). - 51 97
Pathology can alter the number of phases a MUAP can have.

The motor unit can be affected in two general ways following
5 ms 5 point median filtering injury or disease. A pathologic condition may affect the (1) an-
terior horn cell or a peripheral nerve, or (2) the muscle fibers
Figure 2-19. Example of scanning EMG.With a stepping m o t o r comprising a motor unit. If the neural component of a motor
connected t o an EMG needle the territory of a triggered m o t o r unit unit is compromised severely enough to experience degenera-
can be explored.Time is indicated along t h e horizontal axis.The position, all of the muscle fibers innervated by the parent nerve will
tion of t h e needle tip has changed by 50 micrometers for each trace. become denervated. A denervated muscle fiber, through the
This scan shows 200 different MUAPs of o n e m o t o r unit. (From poorly understood process of peripheral sprouting, induces
Gootzen THJM,Vingerhoets HM, Stegeman DF:A study of motor unit nearby terminal axons to send out neural projections to reinner-
structure by means of scanning EMG. Muscle Nerve 1992;15:349-357, vate the orphaned muscle fiber. 15112
Through peripheral sprout-
with permission.) ing, the total number of muscle fibers belonging to a specific
motor unit increases (Fig. 2-20). Initially, the newly formed
axonal sprouts are poorly myelinated and conduct impulses rel-
fibers from one motor unit may cluster, with each group con- atively slowly. Denervated muscle fibers may also conduct im-
95
taining a different number of muscle fibers, and single muscle pulses more slowly because their diameters may decrease.
fibers from the same motor unit may be separated from the Addition of the newly acquired muscle fibers to the motor unit
small groups with variable distances between them. As a record- may produce altered MUAP waveforms. If the conduction ve-
ing electrode approaches a group of muscle fibers belonging to locities of the new axon terminals and additional muscle fibers
that specific motor unit, these fibers primarily influence the are distributed over a greater range than that of the original, the
major negative spike and rise time. When the electrode passes MUAP duration will increase. More excited muscle fibers in-
by these muscle fibers, it may encounter a localized region crease the amount of simultaneously depolarized muscle fiber
where no muscle fibers of that motor unit are firing. The major membrane, thereby producing a larger summated potential in
negative spike and rise time of the MUAP obviously decline the extracellular tissue. This increase in the number of fibers
dramatically. Continued electrode movement allows the elec- tends to increase the magnitude of the initial and terminal as-
trode to record from a different set of muscle fibers resulting in pects of the MUAP above the baseline noise, permitting an ear-
a major negative spike with an altered appearance from that pre- lier and later detection of the MUAP's initial departure from and
viously recorded (Fig. 2-17, B - D ) . By combining all of the
1
subsequent return to the baseline, thereby increasing the
MUAP's measured duration. As the new fibers may have a
larger spatial distribution and could be located outside of the
original motor unit territory, there is less synchronous summa-
tion of the waveforms from outlying fibers than with those of
the original motor unit. The result would be a MUAP with an
abnormal number of phases (Fig. 2-20). As the terminal sprouts
mature and their conduction velocities improve, the temporal
dispersion of individual fiber action potentials could decrease,
resulting in a further MUAP amplitude increase. The total
number of phases may then decrease; however, they often
remain abnormal. Neurogenic diseases, therefore, may lead to
highly polyphasic MUAPs with longer duration and larger
amplitudes.
If the muscle fibers comprising a motor unit undergo a
random degeneration as occurs in some myopathies, the total
number of fibers belonging to that motor unit will decrease (Fig.
2-21 ). - A decrease in number of muscle fibers is likely to result
2
in a reduction of the MUAP's amplitude. This means that the

initial and terminal portions of the MUAP waveform are re-
Figure 2-20. Neurogenic MUAP. In various neurogenic diseases duced and no longer capable of being detected at the same time
the muscle fibers belonging to a particular m o t o r unit may be deprived interval preceding and following the MUAP's negative spike.
of their innervation. Surrounding terminal axons from intact m o t o r The result of a reduced initial and terminal MUAP amplitude
units form collateral sprouts (dotted lines) t o innervate the orphaned causes a shortening of the MUAP duration. Finally, the dropout
muscle fibers.The increased distance of t h e newly formed endplates of single muscle fiber waveforms combined with different fiber
plus m o r e muscle fibers per m o t o r unit plus slow conduction in t h e velocities (fiber size variation) also produces less voltage with
initially unmyelinated nerve sprouts may result in longer-duration, respect to the spatial summation of single fiber potentials.
larger-amplitude, and polyphasic MUAPs. Fewer muscle fibers may lead to "gaps" in the MUAP waveform,
COMPOUND MUSCLE ACTION POTENTIAL

Consideration of compound muscle action potentials
(CMAPs) with respect to volume conduction is discussed in the
Waveform Morphology in Muscle section because the C M A P
waveform arises from muscle tissue. In attempting to record a
CMAP from a particular muscle, the E-l electrode is located on
the skin's surface directly over the muscle's motor point, i.e.,
endplate region. The endplate region typically lies midway be-
tween the muscle's origin and insertion. The E-2 electrode is
usually placed distal to the muscle's tendinous insertion so as to
not record electrical activity from the activated muscle with that
electrode. Stimulating the peripheral nerve innervating the
muscle under investigation, therefore, results in a relatively
large (several millivolts) biphasic waveform with an initial neg-
ative deflection.
The first approximation of the waveform obtained in the
above example can be explained with the L/TD action potential
model. All muscle fibers innervated by the stimulated nerve
Figure 2-21. Myopathic MUAP. Loss of multiple single muscle fibers are depolarized at the endplate (Fig. 2-22A). T h e
fibers may decrease t h e endplate region's spatial expanse thereby di- muscle's endplates are grouped together at about the middle of
minishing the MUAP's duration and amplitude. A less smooth summa- the muscle. As the muscle fibers depolarize rather synchro-
tion of voltages will occur as t h e r e are "missing" muscle fibers creating nously, the endplate zone of the muscle effectively becomes a
"gaps" in the waveform resulting in more phases and turns. large current sink. When the E-l electrode is located directly
over a current sink, the instrument's trace moves in the negative
direction. The action potential induced in the muscle tissue then
causing an increase in the number of phases. It may be possi-
75
propagates in opposite directions toward the muscle's origin and
ble for enough muscle fibers to degenerate to cause a single insertion (Fig. 2-22B). Two negative sinks, therefore, flanked
MUAP to appear as two distinct MUAPs of very short duration. by source currents travel away from each other. The E-l elec-
This may occur because of insufficient voltage from the remain- trode records the two trailing source currents for each action po-
ing fibers to constitute that portion of the waveform between the tential as they travel toward the muscle's tendinous region. The
extremes of the boundary endplate potentials. A primary myo- trace moves accordingly in the positive direction below the
pathic process could yield a shorter-duration, highly polyphasic, baseline to record the full amplitude of the two summated trail-
low-amplitude MUAP. Considering the above, one should be ing source currents. As the muscle impulse gets further from the
aware of the compensatory mechanism of fiber hypertrophy. E-1 electrode, the strength of the source currents declines and
Hypertrophic fibers may produce single-fiber action potentials the trace returns to baseline. The final recorded CMAP wave-
with large amplitudes, so that the MUAP amplitude does not de- form is a biphasic potential with an initial negative deflection.
crease as far as expected; however, the duration may still be ab- As for MUAPs, the situation is more complex than described
normally short. above, and more details of this process can be found in the ap-
It is also possible to observe small-amplitude, short-duration, pendix. Effectively, from a theoretical point of view the situa-
highly polyphasic MUAPs in disorders characterized by signifi- tion is much like that described above for endplate spike
cant dropout of axons, e.g., a major laceration of a peripheral potentials (Fig. 2-14).
nerve. When the newly arrived axons have regrown and reached An initial sharp negative deflection, however, is not always
muscle fibers, at first only a few muscle fibers are reinnervated. observed. Occasionally, a positive deflection can precede the
These newly formed axonal sprouts conduct with a relatively C M A P ' s negative phase. This observation also can be ex-
wide range of conduction velocities. As a result, the observed plained. The E-l electrode may not be located directly over the
MUAP is small in amplitude because there are only a few motor point. The region of muscle surrounding the endplate
muscle fibers comprising the MUAP. Secondly, the duration is zone serves as that portion of the volume conductor from which
short since only a few fibers are depolarizing within a relatively the source currents initially arise to complete the local circuit
short time of each other. The asynchronous arrival of impulses currents into the current sink. An E-l electrode located off the
at the recording electrode secondary to the difference in termi- motor point, therefore, first records some portion of one of the
nal nerve and muscle fiber conduction velocities leads to a source currents "feeding" the current sink (Fig. 2-23). Recall
highly polyphasic potential. As more muscle fibers are added to that the leading portion of the source current results in a positive
the motor unit, the MUAP begins to appear more normal. deflection. As propagation ensues, the current sink eventually
However, there always may be some abnormal parameters still reaches the E-l electrode producing a negative deflection.
detectable depending upon the severity of the lesion and degree Finally, the terminal source current is detected, generating a
of recovery. positive deflection. Instead of the anticipated biphasic potential,
As noted above, MUAPs configurations are more complex a triphasic positive-negative-positive waveform is recorded.
than presented above. The leading/trailing dipole model is used Effectively, the situation is much like that described above for a
to provide a more complete understanding of the MUAP's mor- triphasic endplate spike potentials (Fig. 2-15). When this situa-
phology, particularly relating to different recording techniques. tion occurs, one should relocate the E-l electrode over the an-
The interested reader is encouraged to read the appendix ac- ticipated motor point region. In some instances, it is not
companying this chapter. possible to find a muscle region capable of generating an initial
Figure 2-22. Biphasic CMAP. A,A surface electrode is located over t h e endplate region of a skeletal muscle.Activation of t h e peripheral
nerve supplying that muscle results in a c u r r e n t sink being generated in the endplate zone.The trace describes an initial upward o r negative de-
flection. B, Propagation of the muscle action potentials allows the recording electrode t o d e t e c t their terminal source currents and produce a
positive deflection which eventually settles t o baseline for a MUAP. A summation of those MUAPs from all of the more o r less synchronously ex-
cited nerve fibers leads t o a biphasic negative-positive CMAP
negative deflection. This situation may arise from trauma to the the abductor digiti minimi and stimulating the ulnar nerve. A 40
area distorting the muscle tissue, profound muscle atrophy, or similar appearing potential, intramuscular nerve action po-
attempted recording from a muscle with a poorly defined motor tential, may be observed with a recording electrode located a
point, such as some of the circular facial muscles. bit more distally than that over the abductor pollicis brevis'
If the CMAP is recorded with a high amplifier sensitivity motor point. This muscle can have a rather diffuse region
8
(500-1,000 pV/division) from the thenar or hypothenar mus- which can generate a CMAP with an initial negative deflection,
cles, a small (10-50 pV) negative waveform preceding the thus accounting for a similarly appearing potential preceding
CMAP can be seen (Fig. 2-24). This potential can create a sit-
31
the CMAP with a distinctly different origin. The intramuscular
uation in which a small positive deflection appears to precede nerve action potential arises from the recurrent branch of the
the CMAP, causing the investigator to incorrectly conclude that median nerve in the midpalm region. This response is a near-
40
E-l is not on the motor point. This negative waveform is the so- field potential as opposed to the premotor potential which is a
called premotor potential. The premotor potential originat-
831
far-field potential. The best course of action is to recognize the
ing from the thenar muscle group to median nerve stimulation premotor potential for what it is, and then measure the latency
has been extensively investigated. Investigators have demon- to the take-off of the large negative potential initiating the
strated that the potential is derived from sensory fibers. The 79
CMAP. Of course, one must ensure that the E-l electrode is op-
premotor potential to the thenar eminence has also been claimed timally placed over the motor point prior to concluding that all
to be the palmar cutaneous branch of the median nerve, but this possible positive deflections are the premotor potential.
has been refuted. Although the origin of this potential is some-
31
Just like SNAPs, distally recorded CMAPs may also demon-
what in dispute, it arises from the sensory branches innervating strate a decrement in amplitude and area with more proximal
various portions of the palmar aspects of the hand. The premo-
47
stimulation sites, but the magnitude of this drop does not ap-
tor potential is in fact a far-field potential arising from the sen- proach that seen in SNAPs. The duration of a single-muscle-
sory branches innervating the thumb as they pass from the fiber action potential is substantially longer than that of a single
relatively flat volume of the palm into the cylindrical volume of nerve fiber. As previously stated, the individual muscle fibers
the first digit. This situation also occurs when recording over
40
belonging to one motor unit are separated by terminal nerves of
Chapter 2 ELECTRICAL SOURCES AND VOLUME C O N D U C T I O N 47
"A 500uV/cm
200uV/cm
50uV/cm
Figure 2-24. Premotor potential.The active electrode is placed

on t h e thenar muscle's m o t o r point and a supramaximal stimulus d e -
livered t o the median nerve at the wrist. N o t e t h e small negative p o -
tential preceding t h e main negative spike of t h e CMAP. This small
waveform is t h e p r e m o t o r potential and can give t h e appearance of
t h e CMAP having an initial positive phase. Increasing t h e amplifier's
sensitivity clearly delineates the premotor potential.
topographical studies of the CMAP from various muscle

groups. 108
FAR-FIELD POTENTIALS
One may assume that obvious volume conduction phenom-
ena should pertain to not only neural, but muscular tissue as
well. It may be reasonable to conclude that if far-field potentials
Figure 2-23. Triphasic CMAP. A,The surface electrode is located are produced in nerves, they also may be generated in muscles.
to o n e side of t h e muscle's m o t o r point. Upon depolarization of t h e Computer simulations have predicted that muscle far-field po-
nerve innervating the muscle, some portion of the source current may tentials should occur. The leading/trailing dipole model pre-
50
be detected by t h e electrode. An initial downward deflection of t h e dicts that whenever the leading and tailing dipole moments do
trace results. B, Subsequent propagation of the muscle's action poten- not balance, a far-field potential should be observed. This situa-
tials then produces the expected local circuit sink and terminal source tion could be predicted to occur when a muscle action potential
currents. C, The final result is a waveform that is triphasic with an ini- extinguishes at the musculotendinous junction. Indeed, far-field
tial positive deflection of the trace. muscle potentials have been demonstrated in the human biceps
muscle (Fig. 2-26). -' The explanation is essentially the same
34 01
as previously elucidated in the examples on nerve conduction.

different lengths and comprise an endplate with a certain
amount of spatial expanse. This separation of endplates and dif- FIBRILLATION POTENTIALS
ferent lengths of nerve fibers results in some desynchrony at the
recording site even when the different muscle fibers are simulta- When a muscle fiber is no longer in contact with its nerve
neously activated. After excitation of the motor axons, all of the supply, the resting membrane potential begins to oscillate
muscle fibers innervated by those nerve fibers depolarize and toward the threshold level. As threshold is reached, the single
ithe sum total of voltages add both spatially and temporally to muscle fiber spontaneously fires at regular and sometimes irreg-
jform the CMAP. The net result is a temporal and spatial disper- ular rates, - and an action potential propagates from the end-
9 10
jsion of the waveforms arising from the depolarized muscle plate region or possibly other sensitized areas along the length
fibers. The dispersive effects of time and distance produce a of the fiber (fibrillation). - ° Fibrillation potentials are
,a,05 , 6
phase cancellation effect at the motor unit level and result in simply spontaneous depolarizations of a single muscle fiber and
MUAPs with a negative spike duration approaching 5-6 m s . 88
will demonstrate waveform morphologies similar to those ex-
Motor nerve conduction velocities display about half the tempo- pected of single muscle fibers voluntarily activated. Fibrillation
ral distribution (12-13 m/s) of sensory fibers. - Because of the
27 28
potentials may not only occur spontaneously but also after elec-
longer duration of MUAPs, compared to single-nerve-fiber po- trode movement in pathologic tissue. Characteristically, the fib-
tentials and this relatively little temporal dispersion of the indi- rillation potential is of short duration (less than 5 ms), less than
vidual motor axons (Fig. 2-25), a CMAP with essentially the 1 mV in amplitude, and can fire at rates between 1-15
same duration and only little decrement following more proxi- time/second (Hertz, Hz). They have a typical high-pitched
mal activation is seen. That CMAP generation can be quite "tick" sound like "rain on a tin r o o f if amplified through a
complicated in a realistic situation can be nicely illustrated in loudspeaker. A recording electrode located in the endplate
10
Individual Summated
responses response
Figure 2-25. CMAP phase cancellation. Stimulating t h e nerve distally (open arrows upper traces) results in t h e discharge of t w o MUAPs
with a slightly different latency after t h e stimulus that summate t o produce a CMAP with approximately double amplitude. Proximal stimulation
(closed arrows lower traces) again results in t w o CMAPs that still summate in phase because of the long duration of the MUAP's negative phases.
The temporal dispersion of individual action potentials only minimally alters t h e CMAP. (From Kimura J, Machida M, IshidaT, et al: Relation be-
tween size of compound sensory o r muscle action potentials, and length of nerve segment. Neurology 19S6;36:647-652, with permission.)
zone of a denervated muscle records biphasic fibrillation poten- recording electrode outside of the endplate zone, but far from
tials similar to the normal endplate spikes already discussed. A the tendinous region detects either biphasic (positive-negative)
or triphasic (positive-negative-positive) potentials. Approxi-
mately 30% of fibrillation potentials are triphasic. - The fibril-
9 10
lation potential may appear biphasic and initially positive if

recorded from the musculotendinous junction. The same
volume conductor explanations applied to single muscle fibers
can again be used for the fibrillation potential. A more detailed
explanation of fibrillation potentials is provided in the appendix
to this chapter.
POSITIVE SHARP WAVES

Jasper and Ballem first described a potential recorded from
denervated muscle with a large primary sharp positive deflec-
tion followed by a small or absent negative potential and called
it the positive sharp wave (PSW). This waveform is believed i
57
to have the same clinical significance as the fibrillation poten-

tial _9.79.8i.84 true nature of the positive sharp wave is not yet
fully explained, but bioelectric source and volume conductor
theory can be used to speculate on the possible nature of this in-
teresting and clinically useful potential. Previous discussion
Figure 2-26. Muscle far-field potential. Measurement of the spa- suggests that a positive trace deflection implies that a source
tiotemporal profile along the skin surface of a motor unit action p o - current is approaching the recording electrode. A subsequent
tential of the biceps brachii muscle. An array of 16 surface electrodes small negative potential indicates that the initial aspect of the
is located in parallel with the main direction of the fibers.The E-2 elec- current sink is detected by the recording electrode. Simply, the
trode for all signals was placed on the ipsilateral elbow. N o t e the prop- action potential approaches, but does not or just barely reaches
agating character of the main negative (upward) peak in t w o directions the electrode's recording surface. It has been suggested that the
starting at t h e end plate region (7th electrode) and t h e non-moving pathologically involved muscle fiber is adversely affected by
positive far-filed peak at the end of t h e waveform at all locations indi- the deforming forces of an intramuscularly placed electrode
cating t h e extinction of the tripolar c u r r e n t source profile. (From such that a region of perielectrode conduction failure is in-
Stegeman DF, Dumitru D, King JC, Roeleveld K: Near- and far-fields: duced. The positive sharp wave, therefore, may be a "blocked"
9
Source characteristics and the conducting medium in neurophysiology. fibrillation potential that cannot conduct past the recording elec-
J Clin Neurophysiol 1997; 14:429-442, with permission.) trode (Fig. 2-27). This explanation, however, suggests that the
Figure 2-27. Positive sharp wave.The generation of a positive sharp wave using volume conduction theory. A,The local source currents of a
fibrillation (single muscle fiber) approach a monopolar recording electrode (M) placed against a single muscle fiber.The initial source current is d e -
tected resulting in a positive trace deflection.The needle has caused a portion of the single muscle fiber t o be compressed and no longer capable
of conducting action potentials (hashed region). B, Only a portion of the current sink can reach t h e recording electrode, resulting in a small nega-
tive trace deflection. C, Because the current sink dissipates, the trace settles back t o baseline describing a primarily positive potential or positive
sharp wave. (From Dumitru D:Volume conduction:Theory and application. In Dumitru D (ed): Clinical Electrophysiology. Philadelphia, Hanley &
Belfus, 1989, pp 665-681, with permission.)
PSW would have a duration no longer than that of a fibrillation and invert the negative spike (cannula = E-2), making it appear
potential instead of > 20 ms. Occasionally short-duration PSWs positive and thereby potentially simulating a positive sharp
can be observed and this may be the likely explanation, i.e., a wave. The above three potentials can be distinguished from a
76
blocked fibrillation potential. A more detailed explanation of positive sharp wave in that they are MUAPs and subject to vol-
the PSW is provided in this chapter's appendix. untary control, whereas a positive sharp wave is not. Asking the
Amplified through a loudspeaker, positive sharp waves have patient to contract and relax the muscle under investigation
a regular firing rate (1-20 Hz) that sounds like a dull thud and should result in a positive potential with variable firing rate,
can have durations of between several milliseconds to 100 ms. while a positive sharp wave typically fires regularly. If any
Although observed to fire spontaneously, PSWs also can be pro- doubt remains, the electrode should be repositioned.
voked by electrode movement. Continued research into this fun-
damental and interesting potential is required to more fully
elucidate its true pathophysiologic basis. 41
PITFALLS A N D POSSIBLE SOURCES OF
A number of other potentials can be observed that have a ERROR
morphology similar to a positive sharp wave but may have dis-
tinctly different origins. As we have already described, a MUAP
STIMULATION
recorded from the tendinous region also can have an initial pos-
itive deflection followed by a negative potential because the A nerve is usually stimulated by electrodes located on the
current sink cannot pass beyond the recording electrode. It is 66
skin surface. The cathode produces a depolarizing current,
highly possible for the recording electrode to damage a number which is passively conducted through the skin into the human
of muscle fibers in close proximity to the recording surface pro- body. A portion of this current excites neural tissue in its vicin-
ducing action potential blockade. An additional possibility is for ity provided the current strength is sufficient to depolarize the
the cannula of a concentric needle recording electrode (see nerve fibers to threshold. The induced current exits the body at
Chapter 3) to be preferentially located in the motor unit territory the anode. Although it is theoretically possible to prevent neural
conduction at the anode because of a hyperpolarizing current In addition to a longitudinal spread of the current field, a
about the anode, anodal block, the nerve is seldom prevented radial expansion of the depolarizing current also may be antici-
from conducting an impulse past the anode. Anodal block has pated, particularly when using high stimulus intensities. An in-
been shown to not occur during routine electrodiagnostic medi- creased radial distribution of current can excite neighboring
cine studies. In fact, the anode can actually depolarize the
29
nerves, especially in small anatomic regions containing numer-
nerve and possibly excite it at an unanticipated location leading ous nerves such as the wrist. Coactivation of multiple nerves
to confusion with respect to latency measurements and conduc- leads to the depolarization of numerous muscle groups, includ-
tion velocities (see Chapter 3). It is important to recognize that ing ones that might obliterate the desired response. This possi-
the site of neural excitation and location of the superficially ap- bility is demonstrated with an actual case in which the
plied cathode only correspond with weak stimulus intensities individual sustained significant trauma to the ulnar nerve in the
approximating 4 milliamperes (mA). If we assume a standard
8
forearm. An absent ulnar nerve SNAP and very small CMAP
distance of nerve conduction, elevating the current's intensity to was obtained with activation of the ulnar nerve. These findings
a supramaximal stimulus level (15-20 mA), 20% greater than suggest profound axonal degeneration that was confirmed by
that required to excite all of the nerve's largest fibers, the site of profuse fibrillation potentials and positive sharp waves with
activation is displaced approximately 3 mm from the cathode. 8
only a few MUAP detected with a needle electrode in the first
The nerve is excited away from the cathode resulting in a reduc- dorsal interosseous and ulnar innervated hypothenar muscles.
tion in conduction time (the "virtual cathode" effect). Fre- Stimulating the median and ulnar nerves in the mid-palm region
quently, pathologic nerves require higher stimulus intensities while recording over the same nerves at the wrist revealed in-
102
for activation occasionally approaching 60 mA, which can po- teresting results (Fig. 2-28, A-D). Excitation of the ulnar nerve
tentially displace the site of neural excitation 12 mm. If undue
8
over the fourth palmar interspace while recording over the ulnar
current spread along the nerve is suspected, this potential source nerve proximally, resulted in an apparent normal response. This
of error regarding the exact location of axonal excitation can be result is inconsistent with profound axonal loss in the ulnar
reduced if a needle cathode is used. The near-nerve cathode en- nerve and requires an explanation. Stimulating the median
sures a lower stimulus intensity as current is no longer re- nerve in the second palmar interspace but recording from the
quired to pass a long way through the volume-conducting tissue. ulnar nerve at the wrist also produced a normal waveform.
It also diminishes artifact from the stimulating pulse itself, and Again, activating the fourth interspace but recording from the
better localizes the desired site of neural activation. A monopo- median nerve yielded a normal appearing potential. Finally,
lar needle (see Chapter 3) is recommended as the cathode. stimulating the second interspace and recording over the median
nerve also produced a normal waveform. Given that the ulnar
nerve is pathologically involved, it is reasonable to conclude
that stimulating the fourth interspace with sufficient current did
not activate the ulnar nerve, but actually depolarized the nearby
median nerve in the palm. Once the median nerve was excited,
it followed its normal course and produced the expected re-
sponse at the wrist. Additionally, the current field in the fourth
Ims palmar interspace activated the median nerve. The resulting
median nerve waveform at the wrist was of a sufficient magni-
tude to radially spread to the electrode located over the ulnar
nerve producing a potential at this site following a fourth palmar
interspace stimulus. In pathologic and occasionally normal oc-
casions, it is important to be aware of the stimulating current
strength to avoid a volume-conducted extension of the depolar-
izing current and its resultant neural or muscular response that
extends to unintended regions of the body.
RECORDING
It is important to understand that placing an electrode on the
skin over a specific muscle does not ensure that the only re-
sponse detected will arise form the intended tissue. Recall from
the preceding section that if the depolarizing stimulus is strong
enough, multiple nerves can be activated. Also, it has already
Figure 2-28. Volume conducted stimulation and recording. been pointed out that a positive deflection may precede the an-
Median and ulnar nerve palmar stimulation with wrist recording over ticipated biphasic negative-positive CMAP. Relocating the
the median and ulnar nerves. A, Stimulation of the fourth palmar inter- active electrode over the muscle's motor point eliminates the
space and recording over the ulnar nerve at the wrist. B, Excitation of initial positive phase as the recording electrode is now directly
the second palmar interspace with the same recording montage as in over the current sink.
A. C , T h e fourth palmar interspace is stimulated and the recording The possibility of multiple CMAP waveforms summating in
electrodes are placed over the median nerve. D, Stimulating the the volume conductor of the body can be reduced by placing a
second palmar interspace with the same recording position as C.A recording electrode intramuscularly. This ensures that the ma-
48
large c u r r e n t intensity was used t o obtain all responses. (From jority of the investigated waveform is arising from excitable
Dumitru D, DeLisa JA: Volume conduction. Muscle N e r v e tissue immediately surrounding the electrode. The observed
1991; 14:605-624, with permission.) CMAP amplitude, however, is no longer an accurate reflection
110 mV
5 ms
Amplitude
Latency Peak
Trace Baseline
onset to
to p e a k
peak
ms mV
Figure 2-29. Volume conducted recording. A, Recording of a A 3.5 11 21
CMAP with the active electrode on the midpoint of the first dorsal in- B 3.5 5 9
t e r o s s e o u s muscle and t h e reference electrode on t h e second
metacarpophalangeal joint. B, Relocating the reference electrode t o
Figure 2-30. Effect of reference electrode. A, Active electrode
the first metacarpophalangeal joint.
located on the m o t o r point of the abductor pollicis brevis and refer-
ence electrode placed on t h e distal aspect of t h e first digit. B,
of the total number of muscle fibers activated within the Relocating the reference electrode o n t o t h e muscle tissue. (From
muscle. This is because the intramuscular recording electrode
56
Dumitru D,Walsh NE: Practical instrumentation and common sources
has a more limited recording area and is preferentially influ- of error. Am J Phys Med Rehabil 1988;67:55-65, with permission.)
enced by the tissue in the immediate vicinity of the small
recording surface.
In addition to accurate placement of the E-1 recording elec- that there is a possibility of axonal loss when indeed the small
trode, equal attention must be given to the location of the E-2 or amplitude is an artifact from poor recording techniques in a
reference electrode. Remember that the E-2 electrode should be volume conductor. The primary factors in electrode placement
of a sufficient distance from the E-l electrode so as to not for CMAP observations is to place E-l on the motor point, and
record similar data because this impairs the differential amplifi- E-2 away from the tissue to be excited; the 4 - c m separation
30
cation (see Chapter 3). In the volume conductor of the body, the used in SNAP recordings becomes irrelevant for motor studies.
E-2 electrode is just as capable of recording source and sink The most important factor is to ensure a reference location away
currents as the E-l electrode. This possibility can be easily from active tissue to reduce volume-conducted potentials, but
demonstrated by a simple example. Let us suppose that a close enough to take advantage of common mode rejection of
recording of the first dorsal interosseous (FDI) CMAP is de- environmental noise. In the hand, for example, placing the
sired. The E-l electrode is located over the midpoint of this active recording electrode on the thenar eminence and the refer-
muscle in an attempt to record from the motor point. It is now ence electrode on the metacarpophalangeal joint satisfies both
necessary to place the E-2 electrode in a convenient location, requirements.
such as the second metacarpophalangeal joint region. In volume conduction, recording from and stimulating ex-
Stimulating the ulnar nerve at the wrist results in an initial posi- citable tissues is a trade-off. Enough stimulation must be used
tive deflection (Fig. 2-29A). The possibility arises that E-l is to excite all of the fibers within a nerve, but not enough current
not on the motor point and so it is relocated to several locations to coactivate multiple nerves resulting in too many tissues depo-
to no avail; the initial positive phase persists. Upon moving the larizing. Interpreting waveform recordings can also be haz-
reference electrode to the first metacarpophalangeal joint ardous as many tissues overlap in the volume conductor of the
region, the desired biphasic negative-positive CMAP with a body and cannot simply be ignored. Localized intramuscular
slightly reduced amplitude compared to those obtained previ- recordings can usually assure one whether a specific muscle has
ously is observed (Fig. 2-29B). It should now be obvious that depolarized, but the amplitude recorded less well reflects the
the E-2 electrode records a CMAP from an unintended muscle muscle as a whole.
that is activated prior to the FDI. Because the first activated
muscles were recorded from the reference electrode, they pro-
duced a positive deflection. The source and sink currents from CONCLUSION
the multiple muscles activated to summate somewhat, causing
the CMAP with an initial positive phase to be larger than the An appreciation of the principles discussed in this chapter
biphasic response CMAP. This example suggests that the refer- permits the practitioner the ability to better comprehend the
ence electrode contributed to the CMAP. generation of the morphology of any waveform observed. This
It is necessary to realize that the majority of electrodes are knowledge is of use in not only identifying specific potentials,
relatively small compared to the muscles from which they are but gaining insight into their generation in both normal and
recording. Locating both electrodes over the muscle to be acti- pathologic situations. It is possible to use volume conduction to
vated can produce CMAPs with amplitudes smaller than antici- one's advantage by systematically applying the knowledge
pated (Fig. 2-30). A small amplitude may cause one to conclude gained in this chapter. On the other hand, the unsuspecting or
uninformed clinician is subject to multiple potential errors with 28. Dorfman LJ: The distribution of conduction velocities (DCV) in peripheral
nerves: A review. Muscle Nerve 1984;7:2-11.
respect to the consequences of volume conduction and source 29. Dreyer S, Dumitru D, King JC: Anodal block versus anodal stimulation: Fact or
characteristics and may make erroneous diagnostic conclusions. fiction. Am J Phys Med Rehabil 1993;72:10-18.
30. Dumitru D, Walsh NE: Practical instrumentation and common sources of error.
Am J Phys Med 1988;67:55-65.
REFERENCES 31. Dumitru D, Walsh NE, Ramamurthy S: The pre-motor potential. Arch Phys
Med Rehabil 1989;10:537-540.
32., Dumitru D: Volume conduction: Theory and application. In Dumitru D (ed):
1. Andreassen S, Rosenfalek A: Relationship of intracellular and extracellular Clinical Electrophysiology. Philadelphia, Hanley & Belfus, 1989, pp 665-681.
action potentials of skeletal muscle fibers. CRC Crit Rev Bioeng 33. Dumitru D, DeLisa JA: Volume conduction. Muscle Nerve 1991;14:605-624.
1981;6:267-306. 34. Dumitru D, King JC: Far-field potentials in muscle. Muscle Nerve 1991:
2. Brazier MAB: Electrical Activity of the Nervous System. London, Pitman 14:981-989.
Medical Publishing Co.. 1977, pp 51-74. 35. Dumitru D, King JC: Far-field potentials in circular volumes: Evidence to sup-
3. BrismarT: Potential clamp analysis of membrane currents in rat myelinated port the leading/trailing dipole model. Muscle Nerve 1992:15:101-105.
nerve fibres. J Physiol 1980;298:171-184. 36. Dumitru D, King JC: Far-field potentials in cylindrical volumes: The effect of
4. Buchthal F, Guld C, Rosenfalek P: Action potential parameters in normal different volume sizes and intercompartmental openings. Muscle Nerve
human muscle and their dependence on physical variables. Acta Physiol Scand 1992;15:949-959.
1954;32:200-218. 37. Dumitru D, King JC: Far-field potential production by quadrupole generators in
5. Buchthal F, Guld C, Rosenfalek P: Multielectrode study of the territory of a cylindrical volume conductors. Electroencephalogr Clin Neurophysiol 1993;88:
motor unit. Acta Physiol Scand 1957;39:83-104. 421-431.
6. Buchthal F, Guld C, Rosenfalek P: Volume conduction of the spike of the motor 38. Dumitru D, Jewett DL: Far-field potentials. Muscle Nerve 1993;16:237-254.
unit potential investigated with a new type of multielectrode. Acta Physiol 39. Dumitru D, King JC: Far-field potentials in cylindrical and rectangular volume
Scand 1957;38:331-354. conductors. Muscle Nerve 1993;16:727-736.
7. Buchthal F, Sten-Knudsen O: Impulse propagation in striated muscle fibers and the 40. Dumitru D, King JC: Median/ulnar premotor potential identification and local-
role of the internal currents in activation. Ann N Y Acad Sci 1959;81:422-445. ization. Muscle Nerve 1995;18:518-525.
8. Buchthal F, Rosenfalek A: Evoked action potentials and conduction velocity in 41. Dumitru D: Electrophysiologic basis for single muscle fiber discharge morphol-
human sensory nerves. Brain Res 1966;3:1-122. ogy. Ph.D. Thesis. University of Nijmegen, Benda BV, Nijmegen, 1999.
9. Buchthal F, Rosenfalek P: Spontaneous electrical activity of human muscle. 42. Eisen A, Odusote K, Bozek C, Hoirch M: Far-field potentials from peripheral
Electroencephalogr Clin Neurophysiol 1966;20:321-336. nerve generated at sites of muscle mass change. Neurology 1986;36:815-818.
10. Buchthal F: Fibrillations: Clinical electrophysiology. In Culp WJ. Ochoa J 43. Ekstedt J: Human single muscle fiber action potentials. Acta Physiol Scand
(eds): Abnormal Nerves and Muscle Generators. New York, Oxford University 1964;52(Suppl 226): 1-96.
Press, 1982, pp 632-662. 44. Elmqvist D, Johns TR, Thesleff S: A study of some electrophysiological prop-
11. Carpenter RHS: Neurophysiology. Baltimore, University Park Press, 1984, pp erties of human intercostal muscle. J Physiol 1960;154:602-607.
30-34. 45. Elmqvist D, Hofmarm WW, Kugelberg J, Quastel DMJ: An electrophysiologi-
12. Chiu SY, Ritchie JM, Rogart RB, Stagg D: A quantitative description of mem- cal investigation of neuromuscular transmission in myasthenia gravis. J Physiol
brane currents in rabbit myelinated nerve. J Physiol 1979;292:149-166. 1964; 174:417^34.
13. Chu-Andrews J, Johnson RJ: Electrodiagnosis: An Anatomical and Clinical 46. Elmqvist D, Quastel DMJ: A quantitative study of endplate potentials in iso-
Approach. Philadelphia, J.B. Lippincott Co., 1986. lated human muscle. J Physiol 1965;178:505-529.
14. Clark J, Plonsey R: The extracellular potential field of the single active nerve 47. Fiaschi A: Observations on sensory nature of intramuscular nerve action poten-
fiber in a volume conductor. Biophys J 1968;8:842-864. tial. J Neurol Neurosurg Psychiatry 1973;36:509-513.
15. Coers C, Telerman-Toppet N, Gerard JM: Terminal innervation ratio in neuro- 48. Gassel MM: Sources of error in motor nerve conduction studies. Neurology
muscular disease. Arch Neurol 1973;29:215-222. 1964;14:825-835.
16. Cracco RQ: The initial positive potential of the human scalp-recorded so- 49. Gasser HS, Grundfest H: Axon diameters in relation to the spike dimensions and
matosensory evoked response. Electroencephalogr Clin Neurophysiol the conduction velocity in mammalian A fibers. Am J Physiol 1939;127:393-414.
1972;32:623-629. 50. Gootzen TH, Stegeman DF, Van Oosterom A: Finite limb dimensions and finite
17. Cracco RQ: Traveling waves of the human scalp-recorded somatosensory muscle length in a model for the generation of electromyographic signals.
evoked response: Effects of differences in recording technique and sleep on so- Electroencephalogr Clin Neurophysiol 1991;81:152-162.
matosensory and somatomotor responses. Electroencephalogr Clin 51. Gootzen THJM, Vingerhoets HM, Stegeman DF: A study of motor unit struc-
Neurophysiol. 1972:33:557-566. ture by means of scanning EMG. Muscle Nerve 1992;15:349-357.
18. Cracco RQ, Cracco JB: Somatosensory evoked potential in man: Far-field po- 52. Hakansson CH: Action potential and mechanical response of isolated cross stri-
tentials. Electroencephalogr Clin Neurophysiol 1976;41:460-466. ated frog muscle fibres at different degrees of stretch. Acta Physiol Scand
19. De Lisa JA, Brozovich FV: Volume conduction in electromyography. 1957;41:199-216.
Minimonograph No. 10. Rochester, MN, American Association of Electro- 53. Hakansson CH: Action potentials recorded intra- and extra-cellularly from the
myography and Electrodiagnosis. 1979. isolated frog muscle fibre in Ringer's solution and in air. Acta Physiol Scand
20. De Lisa JA, Brozovich FV: Volume conduction in electromyography: 1957;39:291-312.
Experimental and theoretical review. Electromyogr Clin Neurophysiol 54. Heringa A, Stegeman DF, de Weerd JPC: Calculated potential and electric field
1983;23:651-673. distributions around an active nerve fiber. J Appl Physics 1989;66:2724-2731.
21. de Weerd JPC: Volume conduction and electromyography. In Notermans SLH 55. Hofmann WW, Alston W, Rowe G: A study of individual neuromuscular junc- i
(ed): Current Practice of Clinical Electromyography. Amsterdam, Elsevier, lions in myotonia. Electroencephalogr Clin Neurophysiol 1966;215:521-537.
1984, pp 9-28. 56. Horning MR, Kraft GH, Guy A: Latencies recorded by intramuscular needle
22. Desmedt JE, Bornstein S: Relationship of spontaneous fibrillation potentials to electrodes in different portions of a muscle: Variation and comparison with sur-
muscle fiber segmentation in human muscular dystrophy. Nature 1975; face electrodes. Arch Phys Med Rehabil 1972;53:206-210.
258:531-534. 57. Jasper H, Ballem G: Unipolar electromyograms of normal and denervated
23. Desmedt JE, Cheron G: Central somatosensory conduction in man: Neural gen- human muscle. J Neurophysiol 1949;12:231-244.
erators and interpeak latencies of the far-field components recorded from neck 58. Jewett DL: Volume-conducted potentials in response to auditor)' stimuli as de-
and right or left scalp and earlobes. Electroencephalogr Clin Neurophysiol tected by averaging in the cat. Electroencephalogr Clin Neurophysiol
1980;50:382-403. 1970;28:609-618.
24. Desmedt JE, Cheron G: Prevertebral (oesophageal) recording of subcortical so- 59. Jewett DL: An averaging response technique for recording potentials relative to
matosensory evoked potentials in man: The spinal PI 3 component and the dual a distant point without EKG interference. Electroencephalogr Clin Neuro-
nature of the spinal generators. Electroencephalogr Clin Neurophysiol physiol 1970;28:414-416.
1981;52:257-275. 60. Jewett DL, Romano MN, Williston JS: Human auditory evoked potentials:
25. Desmedt JE, Nguyen TH, Carmeliet J: Unexpected latency shifts of the station- Possible brain stem components detected on the scalp. Science 1970; 167:
ary P9 somatosensory evoked potential far field with changes in shoulder posi- 1517-1518.
tion. Electroencephalogr Clin Neurophysiol 1983;56:623-627. 61. Jewett DL, Williston JS: Auditory-evoked far fields averaged from the scalp of
26. Deupree DL, Jewett DL: Far-field potentials due to action potentials traversing humans. Brain 1971;94:681-696.
curved nerves, reaching cut nerve ends, and crossing boundaries between cylin- 62. Jewett DL, Rayner MD: Basic Concepts of Neuronal Function. Boston, Little
drical volumes. Electroencephalogr Clin Neurophysiol 1988;70:355-363. Brown & Co., 1984.
27. Dorfman LJ, Cummins KL, Abraham GS: Conduction velocity distributions of 63. Jewett DL, Deupree DL: Far-field potentials recorded from action potentials
the human median nerve: Comparison of methods. Muscle Nerve 1982;5: and from a tripole in a hemicylindrical volume. Electroencephalogr Clin
S148-S153. Neurophysiol I989;72:439^t49.
64. Jewett DL: The leading/trailing dipole model as a means of understanding gen- 91. Partanen J: End plate spikes in the human electromyogram: Revision of the
erators of far-field potentials. Clin Evoked Potentials 1990;7:9-13. fusimotor theory. J Physiol (Paris) 1999;93:155-166.
65. Jewett DL, Deupree DL, Bommmannan D: Far-field potentials generated by 92. Plonsey R: Bioelectric Phenomena. New York, McGraw-Hill, 1969.
action potentials of isolated frog sciatic nerves in a spherical volume. 93. Rogart RB, Stampfli R: Voltage-clamp studies of mammalian myelinated nerve.
Electroencephalogr Clin Neurophysiol 1990;75:105-117. In Culp WJ, Ochoa (eds): Abnormal Nerves and Muscles as Impulse
66. Katz B, Miledi R: Propagation of electric activity in motor nerve terminals. Generators. New York, Oxford University Press, 1982, pp 193-210.
Proc R Soc Lond (Biol) 1965;161:453-482. 94. Rosenfalek P: Intra- and extracellular potential fields of active nerve and muscle
67. Katz B: Nerve, Muscle, Synapse. New York, McGraw-Hill, 1966, pp 11-72. fibers. Acta Physiol Scand 1969;(Suppl 321): 1-168.
68. Kimura J, Yamada T: Short-latency somatosensory evoked potentials following 95. Schwartz MS, Stalberg E, Schiller HH, Thiele B: The reinnervated motor unit
median nerve stimulation. Ann N Y Acad Sci 1982;388:689-694. in man: A single-fiber EMG multielectrode investigation. J Neurol Sci
69. Kimura J, Yamada T. Shivapour E, Dickins S: Neural pathways of somatosen- 1976;27:303-312.
sory evoked potentials: Clinical implications. Electroencephalogr Clin Neuro- 96. Stalberg EV, Trontelj J: Single-Fiber Electromyography: Studies in Healthy and
physiol 1982;36(Suppl):328-335. Diseased Muscle. New York, Raven Press, 1994.
70. Kimura J, Mitsudome A, Yamada T, Dickins QS: Stationary peaks from a 97. Stalberg EV, Antoni L: Electrophysiological cross section of the motor unit. J
moving source in far-field recording. Electroencephalogr Clin Neurophysiol Neurol Neurosurg Psychiatry 1980;43:469-474.
1984;58:351-361. 98. Starr A, Achor JL: Auditory brain stem responses in neurological diseases. Arch
71. Kimura J, Kimura A, Ishida T, et al: What determines the latency and amplitude Neurol 1975;32:761-768.
of stationary peaks in far-field recordings? Ann Neurol 1986;19:479-486. 99. Stegeman DF, de Weerd JPC: Modelling compound action potentials of periph-
72. Kimura J, Ishida T, Suzuki S, et al: Far-field recording of the junctional potential eral nerves in situ I: Model description: Evidence for nonlinear relation between
generated by median nerve volleys at the wrist. Neurology 1986;36:1451-1457. fibre diameter and velocity. Electroencephalogr Clin Neurophysiol
73. Kimura J, Machida M, Ishida T, et al: Relation between size of compound sen- 1982;54:436-448.
sory or muscle action potentials, and length of nerve segment. Neurology 100. Stegeman DF, Van Oosterom A, Colon EJ: Far-field evoked potential compo-
1986;36:647-652. nents induced by a propagating generator: Computational evidence. Electro-
74. Kimura J, Sakimura Y, Machida M, et al: Effect of desynchronized inputs on encephalogr Clin Neurophysiol 1987;67:176-187.
compound sensory and muscle potentials. Muscle Nerve 1988;11:694-702. 101. Stegeman DF, Dumitru D, King JC, Roeleveld K: Near- and far-fields: Source
75. Kincaid JC, Minnick KA, Pappas S: A model of the differing change in motor characteristics and the conducting medium in neurophysiology. J Clin
and sensory action potentials over distance. Muscle Nerve 1988; 11:318-323. Neurophysiol 1997; 14:429-442.
76. Kosarov D; Gydikov A: The influence of the volume conduction on the shape of 102. Stevens JC: AAEE Minimonograph #26: The electrodiagnosis of carpal tunnel
the action potentials recorded by various types of needle electrodes in normal syndrome. Muscle Nerve 1987;10:99-113.
human muscle. Electromyogr Clin Neurophysiol 1975;15:319-335. 103. Stockard JJ, Stockard JE, Sharbrough FW: Detection and localization of occult
77. Krnjevic K, Miledi R: Motor units in rat diaphragm. J Physiol 1958; 140:427-439. lesions with brainstem auditory responses. Mayo Clin Proc 1977;52:761-769.
78. Kugelberg E: Electromyography in muscular dystrophy: Differentiation be- 104. Stolov WC: Instrumentation and Measurement in Electrodiagnosis. Mini-
tween dystrophies and chronic lower motor neurone lesions. J Neurol monograph No. 16. Rochester, MN, American Association of Electrodiagnostic
Neurosurg Psychiatry 1949;12:129-136. Medicine, 1981.
79. Lambert EH: Studies of the origin of the positive sharp wave in electromyogra- 105. Thesleff S, Ward MR: Studies on the mechanism of fibrillation potentials in
phy. [Newsletter]. Rochester, MN, American Association of Electromyography denervated muscle. J Physiol 1975;244:313-323.
and Electrodiagnosis, 1957;3:3. 106. Thesleff S: Fibrillation in denervated mammalian muscle. In Culp WJ, Ochoa J
80. Lambert EH: Electromyography in amyotrophic lateral sclerosis. In Norris FH (eds): Abnormal Nerve and Muscle as Impulse Generators. New York, Oxford
Jr, Kurland LT (eds): Motor Neuron Diseases. New York, Grune & Stratton, University Press, 1982, pp 678-694.
1969, pp 135-153. 107. Thiele B, Bohle A: Number of spike components contributing to the motor unit
81. Li C, Shy G, Wells J: Some properties of mammalian skeletal muscle fibres potential. Z EEG-EMG 1978;9:125-130.
with particular reference to fibrillation potentials. J Physiol 1957;135:522-535. 108. Van Dijk JG, van Benten 1, Kramer CG, Stegeman DF: CMAP amplitude car-
82. Lorente de No R: Analysis of the distribution of action currents of nerve in tography of muscles innervated by the median, ulnar, peroneal, and tibial
volume conductors. Stud Rockefeller Inst Med Res 1947;132:384-477. nerves. Muscle Nerve 1999;22:378-389.
83. Ludin HP: Microelectrode study of normal human skeletal muscle. Eur Neurol 109. Waxman SG, Foster RE: Ionic channel distribution and heterogeneity of the
1969;2:340-347. axon membrane in myelinated fibers. Brain Res 1980;203:205-234.
84. Ludin HP: Electromyography in Practice. Stuttgart, Georg Thieme Verlag, 1980. 110. Waxman SG, Kocsis JD, Eng DL: Ligature-induced injury in peripheral nerve:
85. Martin WH, Pratt H, Schwegler JW: The origin of the human auditory brain- Electrophysiological observations on changes in action potential characteristics
stem response wave II. Electroencephalogr Clin Neurophysiol 1995;96: following blockade of potassium conductance. Muscle Nerve 1985;8:85-92.
357-370. 111. Weber RJ: Motor and sensory conduction and entrapment syndromes. In
86. Moore JK: The human auditory brain stem as a generator of auditory evoked Johnson EW (ed): Practical Electromyography, 2nd ed. Baltimore, Williams &
potentials. Hear Res 1987;29:33-43. Wilkins, 1988, pp 92-186.
87. Noble D: Applications of Hodgkin-Huxley equations to excitable tissues. 112. Wohlfart G: Collateral regeneration from residual motor nerve fibers in amy-
Physiol Rev 1966;46:1-50. otrophic lateral sclerosis. Neurology 1957;7:124-134.
j 88. Olney RK, Budingen HJ, Miller RG: The effect of temporal dispersion on com- 113. Woodbury JW: Potentials in a volume conductor. In Ruch TC, Patton HD,
j pound action potential area in human peripheral nerve. Muscle Nerve Woodbury JW, Towe AL (eds): Neurophysiology, 2nd ed. Philadelphia, W.B.
1
1987;10:728-733. Saunders, 1965, pp 85-91.
89. Paintal AS: The influence of diameter of medullated nerve fibers of cats on the 114. Yasuhara A, Yamada T. Seki Y, et al: Presence of two subcomponents in P9 far-
rising and falling phases of the spike and its recovery. J Physiol field potential following stimulation of the median nerve. Electroencephalogr
1966;184:791-811. Clin Neurophysiol 1990;77:93-100.
90. Paintal AS: Conduction in mammalian nerve fibers. In Desmedt JE (ed): New 115. Zwarts MJ: Evaluation of the estimation of muscle fiber conduction velocity.
Developments in Electromyography and Clinical Neurophysiology. Basel. Surface versus needle method. Electroencephalogr Clin Neurophysiol 1989;73:
Karger. 1973, pp 19-41. 544-548.
Appendix
The Leading/Trailing Dipole

Model and Near-Field/Far-Field
Waveforms
Dick F. Stegeman, Ph.D.
APPENDIX OUTLINE
Near-Field Waveforms Near-Field and Far-Field Waveforms

Innervated Muscle Fibers Far-Field Theory • Motor Unit Action Potential • Compound
• Denervated Muscle Fibers Muscle Action Potential
The leading/trailing dipole (L/TD) model gives a deceptively

1
relatively close to an active fiber. When an electrode records
simple yet powerful explanation for many electrophysiologic phe- electrical activity close to the tissue of interest, the waveform
nomena. We may characterize nerve or muscle fiber electrical ac- can be referred to as a near-field waveform. We will also
21
tivity as consisting of a central negative current sink surrounded discuss far-field waveforms and sources after the groundwork
and supplied by positive current sources. Since the majority of for near-field sources and their associated waveforms have
action potentials propagate for at least a short time during their ex- been described.
istence, a direction of propagation is implied. This directional For the purposes of this discussion, the L/TD model will be
quality stems from the fact that action potentials begin at some de- described with respect to muscle fibers only. However, the same
fined location and extinguish at another. In addition to propagation principles apply to nerve fibers as well. We may begin by con-
direction, action potentials also have spatial characteristics, i.e., sidering an innervated single muscle fiber with a neuromuscular
they span a defined section of the sustaining nerve or muscle fiber. junction located midway between the fiber's two musculotendi-
These two action potential qualities permit one to use a two-di- nous junctions (Fig. 1). Initially, the opening of acetylcholine j
mensional action potential description whereby its central (nega- receptors produces an inwardly directed sodium ion current !
tive) sink region is flanked by a leading (direction in which the forming a negative voltage current sink (Fig. 1A). The sur-
action potential is heading) and trailing (direction from which the rounding region of muscle membrane acts as the positive volt-
action potential originated) positive source, creating the traditional age source "feeding" the negative sink, thereby generating the
tripole (+ - +) model. For conceptual convenience, the tripole
16
growing quadrupole (+ — +). This quadrupole grows in magni-
(+ - +) is converted into a quadrupole (+ — +) with a designated tude as the sodium current maximizes (Fig. IB). A recording
leading dipole (LD: + - ) and a trailing dipole (TD: - +). Finally, electrode is positioned directly over the neuromuscular junction
this quadrupole is spread out over a region of membrane creating region with a distant reference electrode forming a so-called
a spatially distributed compound source (the L/TD model). "referential" or monopolar recording montage. This electrode
records the initial negative voltage, which results in a negative
or upwardly directed deflection on the recording oscilloscope
NEAR-FIELD WAVEFORMS (Fig. 1A and B; E-1 ). The negative deflection continues to
A
grow in magnitude proportional to that of our developing nega-

INNERVATED MUSCLE FIBERS tive sink. Action potential propagation from the neuromuscular
junction out onto the muscle membrane results in the action po-
The L/TD model will first be used to describe a number of tential's trailing dipole (TD) becoming manifest about the neu-
anticipated waveforms recorded with an electrode positioned romuscular junction coincident with membrane repolarization
54
Appendix THE LEADING/TRAILING DIPOLE MODEL A N D NEAR-FIELD/FAR-FIELD WAVEFORMS — 55
deflection grows in magnitude the closer the LD approaches the

electrode. As the LD passes by the electrode, the quadrupole's
negative portion is then recorded, generating an upward oscillo-
scope deflection (Fig. 1C; E-1 ). Continued action potential
B
propagation permits the electrode to detect the TD's positive

pole, generating a terminal positive deflection on the oscillo-
scope (Fig. 1C-H; E - 1 ) . The positive deflection eventually set-
B
tles back to baseline the further the TD gets from the electrode.
The net result is an initially positive triphasic potential. It is this
triphasic potential that is to be expected for all waveforms that
propagate toward, reach, and then travel past a recording elec-
trode provided the reference electrode is far from the active
electrode. Of note, it is possible to record triphasic endplate
spikes with an initial positive deflection using the above L/TD
explanation (Fig. 2B). This is an important issue since these
4
waveforms may be mistaken for fibrillation potentials if careful

consideration is not given to their irregular firing rate that is
identical to that of the initially negative endplate spikes. 4
Next, our recording electrode may be located at the junction

between the muscle and tendon tissues (Fig. 1A; E - l ) . In this
c
instance, the electrode records an initial positive deflection as-

sociated with the LD's approaching positive pole (Fig. 1 B - D ;
E - l ) . The magnitude of the initial positive deflection continues
c
to increase the closer the LD gets to the electrode. When the L D

reaches the electrode, an interesting sequence of events occurs.
Recall, the electrode is positioned at the termination of the ex-
citable muscle tissue and the initiation of connective tissue that
Figure I. Single-muscle-fiber simulation. A single muscle fiber is is incapable of sustaining an action potential. Therefore, when
depicted with a centrally located neuromuscular junction.Three point the LD encounters the muscle tissue terminus, it begins to col-
electrodes (E-1 , E-1 , and E-1 ) are located at t h e endplate, between
A B c
lapse (Fig. IE). The LD continues to dissipate until it is com-
the endplate and musculotendinous region, and at t h e musculotendi- pletely gone and only the TD remains. For this instant in time,
nous junction, respectively.The waveforms recorded by these t h r e e the electrode records a reduction in the waveform's initial posi-
electrodes also are shown. An action potential is generated at the neu- tive deflection and a transition to a negative deflection associ-
romuscular junction (A) with subsequent propagation and ensuing ated with the T D ' s negativity that now "faces" the electrode
action potential dissipation at t h e musculotendinous junctions (B-H). (Fig. IE and F; E - l ) . As with the LD, the TD in turn encoun-
c
(From Dumitru D, King JC, Stegeman DF: Endplate spike morphology. ters the termination of muscle tissue and also begins to dissi-
Arch Phys Med Rehabil 1998;79:634-640, with permission.) pate. Since the T D continues to "face" the electrode until it
completely disappears, the electrode records a declining nega-
(Fig. 1C). The electrode now records the TD's positive voltage tive phase. The final potential recorded from the junction be-
(Fig. 1C and D; E - 1 ) . Continued TD propagation away from
A
tween excitable and inexcitable tissue is an initially positive
the recording electrode results in a declining positive voltage biphasic waveform (Fig. 1; E - l ) . This description clearly
c
source (Fig. 1E-H; E - 1 ) . This sequence of events describes

A
shows that in the endplate region, not only can potentials with
the biphasic initially negative prototypical endplate spike (Fig. morphologies identical to fibrillation potentials be observed, but
2A). The preceding explanation also applies to compound waveforms that to some degree appear similar to positive sharp
muscle action potentials evoked during nerve conduction stud- waves also can be detected (Fig. 2C and D ) . Specifically, if a
4
ies where the electrode is purposely positioned over the needle electrode is angled so as to irritate a terminal nerve and
muscle's motor point region. Failure to obtain an initial positive simultaneously compress the muscle fiber, a biphasic, initially
deflection may indicate that the electrode is not located directly positive potential will result. Again, it is the irregular firing rate
over the motor point zone (see below). In general, the trailing that identifies these potentials as "atypicaf'-appearing endplate
positivity for the above-described endplate spike is smaller in spikes and not pathologic positive sharp waves.
magnitude than that of the preceding negative spike since it has The three above-described waveforms derive directly from a
a longer spatial distribution, which means that although the cur- source description of excitable tissue in a volume conductor. If
rent is equivalent to that of the negative spike, it is spread over a a waveform with an initial negative deflection is observed, one
larger region of membrane and hence its current density and as- may cautiously assume that the electrode is located over an
sociated voltage profile are comparatively smaller. action potential's point of origin. A waveform with an initial
We may next consider a slightly different electrode recording positive deflection, however, can cautiously be assumed to orig-
location. Suppose an electrode is positioned not quite half-way inate away from the electrode. A negative phase following an
between the endplate and musculotendinous junction (Fig. 1A; initial positive deflection suggests that the negative sink arrives
E-1 ). Action potential generation at the endplate again pro-
B
at the active recording electrode's location. If a terminal posi-
duces a quadrupole. This time, however, because of its location, tive phase is detected, then the action potential likely propa-
the electrode detects the positivity associated with the L D ' s gated away from the electrode since the TD's terminal positive
positive pole. The electrode records a positive or downward de- voltage was observed. On the other hand, if a terminal negative
flection on the oscilloscope (Fig. 1A and B; E-1 ). This positive
B
instead of positive phase is observed, then the action potential
\J
200 10 pV
V yV
Figure 3. Orthodromic SNAP. A, O r t h o d r o m i c median nerve

sensory action potential recorded from t h e wrist following third digit
B stimulation. B, Relocating t h e active electrode slightly off t h e nerve
generates t h e same potential in A with t h e exception of an initial posi-
tive deflection.
example, a waveform with an initial positive deflection would

be observed (Fig. 3B). In this case, the two electrodes do not
detect the action potential's initial positivity to the same degree
and the waveform from the same nerve now appears differently,
i.e., it becomes triphasic.
In innervated muscle tissue, the above three potential mor-
phologies effectively describe all waveforms likely to be encoun-
tered depending on the electrode's recording location (Fig. 4).
Any additionally observed waveform morphologies are likely a
result of two or more single muscle fiber potentials summating
to yield a more complex appearing potential (Fig. 5A and B ) . 4
The same three fundamental waveforms are also recorded in

neural tissue, provided comparable recordings are performed. 12
Figure 2. Endplate spike. A, Protypical endplate spikes are shown

with a small (A-1) and relatively m o r e prominent (A-2) so-called "pre-
potential" representing t h e m e m b r a n e approaching threshold. B,A
series of endplate spikes detected from an endplate region with slight
needle movement showing a biphasic endplate spike transition t o a
triphasic endplate spike. C, Endplate spikes also may have an initial pos-
itive deflection appearing as a short-duration "positive sharp wave." D,
It is also possible t o record a series of endplate spikes appearing as a
run of "positive sharp waves." (From Dumitru D, King JC, Stegeman
DF: Endplate spike morphology: A clinical and simulation study. Arch
Phys Med Rehabil 1998;79:634-640, with permission.)
arrived at, but did not propagate past the electrode. The quali-
fiers noted in this description relate to referential recordings
with a reference (E-2) electrode far from the active source.
Exceptions to the above may result from differential amplifica-
tion of bipolarly derived action potentials (see Chapter 3 ) . For
example, recording an orthodromic sensory nerve action poten-
tial (SNAP) from the median nerve at the wrist with both the
active and reference electrodes spaced close together and
pressed firmly over the nerve, typically results in a biphasic ini- Figure 4. Single muscle fiber recording. A muscle fiber is de-
tially negative waveform (Fig. 3 A ) . This initial negativity arises picted with a series of recording electrodes located at t h e endplate
from the fact that the approaching positivity of the L D is de- region and musculotendinous junction as well as several positions be-
tected with relatively the same magnitude by both recording tween. N o t e that there are three primary morphologies (biphasic ini-
electrodes, and is eliminated as a common mode signal. tially negative, triphasic initially positive, and biphasic initially positive)
Therefore, observing a biphasic initially negative SNAP with an that depend on the electrode's recording location with respect t o the
obvious action potential propagating toward the recording elec- action potential. (From Dumitru D, King JC, Stegeman DF: Endplate
trode is not a violation of the above theory. Also, if one were to spike morphology: A clinical and simulation study. Arch Phys Med
displace the active electrode slightly off the nerve in our preceding Rehabil 1998;79:634-640, with permission.)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 with respective durations of approximately 3 ms and 180 ms

(Fig. 6A and B) by the conduction velocity of a single muscle
fiber action potential 3.7 mm/ms (same as 3.7 m/s) to derive the
action potential's spatial extent, or just what portion of the
membrane the action potential physically occupies (Fig. 6C and
D). Once this is accomplished, we can display the respective
IAPs for innervated and denervated muscle tissue with an
equivalent L/TD representation. As discussed above, the inner-
vated IAP is characterized by a quadrupole (+ — +; Fig. 6E).
However, recall the denervated IAP has an additional phase that
is of opposite polarity to that of depolarization, i.e., a hyperpo-
larization phase. Therefore, when converting this IAP to a
spatial L/TD representation, the initial depolarization/repolar-
ization phase is again configured as a quadrupole (+ — + ) ;
however, the hyperpolarization phase also must be represented
Figure 5. M o n o p o l a r n e e d l e r e c o r d i n g . A, Series of potentials by a quadrupole but of opposite polarity ( - + + - ) . The final
recorded with a monopolar needle in a biceps brachii's endplate zone. L/TD model for a denervated muscle fiber is a double quadru-
(From Wiechers D O : Electromyographic insertional activity in normal pole, or an octopole (+ — + , _ + + _ Fig. 6F).
muscle tissue. Arch Phys Med Rehabil 1999; 60:359-363, with permis-
Prior to attempting a complete explanation of the morpholo-
sion). B.The above clinically recorded waveforms were successfully re-
gies single muscle fibers may display in denervated muscle
produced by summating t h e biphasic initially negative and positive
tissue, it is necessary to briefly describe the consequences of a
endplate spikes with each o t h e r in varying magnitudes and time delays.
spatially distributed IAP source, i.e., the amount of membrane
(From Dumitru D, King JC, Stegeman DF: Endplate spike morphology:
occupied by the IAP's dipoles. Additionally, slightly more detail
A clinical and simulation study. Arch Phys Med Rehabil I998;79:
is required with respect to the manner in which action potentials
extinguish at the termination of excitable tissue.
The issue of an IAP's spatial distribution and the ensuing
DENERVATED MUSCLE FIBERS consequences with respect to the extracellularly recorded poten-
tial is described first. An analogy to a moving car is applied to
In denervated muscle, the same L/TD principles apply with the IAP. Let us suppose a car is traveling a distance of 60 miles
the exception of an altered intracellular action potential between points A and B. If it takes one hour for the car to travel
(IAP). It is important to recognize that muscle fibers undergo between points A and B, then on average the car was traveling
dramatic changes following denervation.' - ~ Within several
8 22 24
at a velocity of 60 miles/hour. The rate at which the car changes
days, a separate class of voltage-gated sodium and potassium
channels are newly synthesized. The newly formed sodium
channels are different from the wild type in that they no
longer can be blocked by the poison tetrodotoxin. The im- 24
portant issue regarding the altered sodium channels is that

they display somewhat different kinetics and result in a less
negative resting membrane potential (closer to membrane
threshold) and may have less effective sodium inactivation
characteristics compared to the wild-type sodium channels.
These properties alter the initial portion of the IAP slightly in
that the rise time to peak depolarization requires just a bit
more time than found in innervated tissue. This alteration in
time to peak depolarization, however, does not result in a dra-
matic IAP alteration or its corresponding portion of the extra-
cellularly recorded waveform.
On the contrary, the newly synthesized voltage-gated potas-
sium channels dramatically alter the IAP's configuration, par-
ticularly with respect to its terminal portion. The potassium
channels synthesized following denervation are not only volt- (+ -X- +)
age-gated, but are calcium-activated and identified because of
their binding affinity to the bee venom apamin, which is not the Figure 6. I n t r a c e l l u l a r a c t i o n p o t e n t i a l . A , An IAP recorded
property of the wild types of potassium channels. - - These 8 17 20
from an innervated rat skeletal muscle is shown with a clear monopha-
channels are referred to as SK3 (small-conductance (S) potas- sic morphology. B.The IAP from a denervated rat skeletal muscle
sium (K) channels) which result in a prolonged hyperpolariza- demonstrates a readily apparent hyperpolarization phase of long dura-
tion of the IAP (Fig. 6A and B). It is the hyperpolarization tion resulting in a biphasic action potential. C and D, Multiplying both
phase that may contribute to the repetitive nature of sponta- action potentials by a conduction velocity of 3.7 m/s generates t h e
neous depolarizations of the denervated muscle membrane. 24
spatial representation of t h e t w o action potentials in A and B, respec-
The above description of the altered IAP can be used to better tively. E and F.The leading/trailing dipole representation of the respec-
understand why fibrillation potentials and positive sharp waves tive action potentials in C and D is depicted. (From Dumitru D, King
display their respective morphologies. We may first multiply the JC, Rogers WE, Stegeman DF: Positive sharp wave and fibrillation p o -
IAP profile for both the innervated and denervated muscle fibers tential modeling. Muscle Nerve 1999;22:242-251, with permission.)
distance per time, or its velocity, is the first derivative of its a passive current within and across its membrane over the spec-
motion. If we want to know the car's acceleration between ified crushed region. This type of membrane termination effect
points A and B, then we speak of how the velocity is changing for an IAP is called a "crushed end" or "compressed end." 5
over time, i.e., the car's rate of rate of change. Let us assume the We may now consider the effect a cut versus a crushed end
car traveled at a constant velocity of 60 miles/hour and, there- has on an IAP when it encounters either of these termination ef-
fore, its acceleration was 0 miles/hour/hour. In short, when the fects. At this point, we will consider an innervated IAP's inter-
velocity is not changing, irrespective of its magnitude, or chang- action with a cut and crush end to simplify the explanation and
ing only very little, the acceleration (rate of rate of change) is later consider a denervated IAP terminating at a crush region.
zero or very small, respectively. We may now apply this knowl- As noted briefly above, suppose an action potential is propagat-
edge to the IAP. ing toward the musculotendinous junction or, when considering
Recall that the IAP is represented in our L/TD model as a nerve tissue, the terminal aspect of the nerve. A consequence of
change of voltage (increasing and decreasing) over some speci- the L/TD model for the transmembrane current is a triangular
fied membrane length. The rate at which this voltage changes IAP since the second derivative of a triangle just is the schema-
over distance within the muscle or nerve fiber is proportional to tized double dipole used in this model. The IAP's voltage pro-
the intra-axial current and may be thought of as the IAP's first file goes linearly from resting membrane potential ( - 90 mV) to
derivative. Second, we know that this current must exit the
15
maximal depolarization (~ + 30 to + 40 mV), and is followed by
fiber to complete the local circuit currents. The current exiting a linear repolarization back to the resting membrane level ( - 90
the fiber is known as the transmembrane current or the rate at mV) (Fig. 7A). The location of the poles of the double dipole is
which the intra-axial current is changing from inside to outside indicated along the x-axis in Fig. 7A. The IAP in Fig. 7 A may
the fiber, i.e., the IAP's second derivative. Therefore, how the be conceptualized as existing just prior to encountering the ab-
IAP appears extracellularly depends on its transmembrane cur- sence of excitable tissue at the musculotendinous junction.
rent profile, otherwise represented as the IAP's second deriva- Continued IAP propagation results in the IAP's depolarization
tive. If the transmembrane current changes significantly, a phase, and therefore the LD shortening, as there is no longer
relatively large extracellular waveform is detected. Similarly, if any tissue to sustain continued action potential propagation
the transmembrane current changes little, a negligible extracel- (Fig. 7B). This process continues until the LD is completely
lular waveform is observed. gone (Fig. 7C). The TD now dissipates as its sustaining mem-
It is next necessary to consider the types of action potential brane undergoes repolarization and its corresponding length de-
termination that may exist. We have already addressed the most creases until it is completely gone (Fig. 7 C - D ) . An electrode
common type of termination an action potential can encounter, located at this transition zone records a biphasic initially posi-
namely the musculotendinous junction. Tendon does not pos- tive potential with a terminal negative phase (Fig. 1; E-1 ) .c
sess voltage-gated ion channels and, therefore, cannot sustain The effect a crush end has on IAP termination is quite differ-
an action potential. The abrupt termination of excitable tissue in ent from that of the cut end. For our purposes, we are defining a
a complete absence of ion channels is referred to as a "cut end" crushed membrane as a region of nerve or muscle tissue with
or "sealed end." On the other hand, various experiments can
5
an intracellular space but with a complete absence of functional
use a hemostat to crush previously normal tissue, resulting in membrane voltage-gated ion channels. This means that that
the absence of voltage-gated channels but the continued pres- region of crush membrane can sustain passive currents flowing
ence of a fiber with an intact interior. The crushed region of
9
across it (intracellular to extracellular) only to complete a local
fiber cannot mediate an action potential; however, it can sustain circuit current to some other region of membrane with voltage-
gated ion channels. In this case, the quadrupole (+ — +) ap-
proaches a small region of crush membrane (Fig. 7E). The
initial portion of the LD enters the crush zone because it con-
sists of the initial passive current flow (Fig. 7F). Recall that only
the negative sink depends on the opening of voltage-gated
sodium channels, while the source currents feeding the sink are
primarily passive in nature and simply pass through the mem-
brane to complete the local circuit current to the negative sink.
This process continues as the IAP propagates, with the LD
shrinking in size but not dissipating. The effect of a shrinking
LD with continued current flow sustains a complete triangular
IAP and a balance between the LD and TD. The IAP propagates
up to the crush zone until the portion where the open sodium
channels begin (Fig. 7G). At this point, the action potential
comes to a standstill. Because there is still an LD to balance the
TD, the entire quadrupole simply dissipates with repolarization
and a proportional simultaneous shrinking of the LD and TD
Figure 7. Intracellular action potential. A-D, An IAP is repre- until they are both gone (Fig. 7H). An electrode positioned at
sented by the L/TD model as a triangular intracellular voltage profile the crush zone's termination will detect the IAP's LD and
(cut end, see text). E - H , T h e region of m e m b r a n e that is crushed is record an initial positive deflection (Fig. 8C; E - l ) . The magni-
c
that portion between the vertical dashed lines. Note that this region is tude of this initially recorded positivity continues to increase
smaller than the length of the LD in this case but it may be larger as until the region of open sodium channels reaches the crush zone
well. (From Dumitru D, King JC, Rogers W E , Stegeman DF: Positive (Fig. 8C-E; E - l ) . At this point, the action potential dissipates.
c
sharp wave and fibrillation potential modeling. Muscle N e r v e Since the electrode still "sees" the LD's positivity, it records a
1999;22:242-251, with permission.) decline in the magnitude of the positive voltage (Fig. 8E; E - l ) .c
for the denervated IAP, there also is a terminal quadrupole ( - +

+ - ) representing the action potential's hyperpolarization phase.
Therefore, continued action potential propagation permits the
initial portion of the denervated IAP's terminal quadrupole to
manifest. The electrode at the fiber's origin now detects this
quadrupole's positivity (Fig. 9D; E-1 ). However, as previously
A
noted, the terminal quadrupole has a very long spatial distribu-

tion and a correspondingly low current density. The second de-
rivative, therefore, of this very low current density region
effectively results in a negligible potential, and for all practical
purposes the electrode barely if at all registers a positive deflec-
tion. The net result is that the waveform's terminal positivity re-
turns to baseline, resulting in a net waveform recorded at the
fiber's origin that is biphasic and initially negative of relatively
short duration approximating only 5-7 ms (Fig. 9A-L; E-l ) . A
The result described by the L/TD model is consistent with fib-

rillation potentials recorded in the former endplate zone, i.e., an
initially negative biphasic potential about 5 ms in duration.
An electrode located in the fiber's middle (25 mm) detects
the IAP's initial positivity associated with the LD's positive
pole and records a waveform that begins with a positive deflec-
tion (Fig. 9C; E-1 ). The waveform's positive deflection contin-
B
ues to grow the closer the LD gets to its location. As the IAP's
leading quadrupole's negative sink reaches the electrode, the
waveform correspondingly demonstrates a negative spike (Fig.
9D; E-1 ). Propagation of the IAP permits the electrode to now
B
detect the leading quadrupole's terminal positive polarity with

an ensuing positive waveform deflection (Fig. 9E; E - 1 ) . B
Figure 8. Single muscle fiber simulation. A situation is depicted Further IAP propagation results in the trailing quadrupole and
identical to that described in Figure I with the exception of a crushed its associated low current density presenting at the electrode. As
termination as opposed to a sealed end termination. C A s the quadru- with the fiber's origin, this current density and its second deriv-
pole (action potential) approached the electrode positioned at the ative are far too small to yield a detectable potential. The end
crushed termination, an initial positive deflection is detected (E-I ). D,
c
result is a return of the waveform to baseline, thereby describing
Just before the quadrupole's negative sink encounters the crush zone, an initially positive triphasic waveform (Fig. 9F-L; E-1 ). This
B
the detected potential's positive deflection maximizes ( E - | ) . E-H,

c
triphasic waveform is exactly that anticipated for a fibrillation
Dissipation of the quadrupole upon encountering the crush zone re- potential originating away from, propagating toward, as well as
sults in a monophasic positive potential (E-1 ) .
c
traveling away from the recording electrode.
The electrode located just proximal to the fiber's crush termi-
nation demonstrates an interesting waveform. The oncoming
This process continues until the action potential is completely IAP's LD generates a waveform with the expected initial posi-
dissipated, resulting in a monophasic positive potential (Fig. tive deflection (Fig. 9E; E - l ) . However, as the IAP's LD enters
c
8F-H; E - l ) . A negative deflection is not observed because the

c the crush zone and its negative region arrives at the electrode,
region of open sodium channels (negative sink) never reaches the LD length is shortened and its current density dramatically
the electrode location, as the LD's positivity is always facing increased with a resulting large negative spike recorded (Fig.
the electrode. Of course, if the electrode were positioned over 9F; E - l ) . Recall that a crush zone has the unique ability to alter
c
the region of normal membrane prior to the crush zone, then it yet preserve the LD (Fig. 7E-H). As the membrane repolarizes,
would detect the negative sink and record a corresponding nega- the negative spike settles back to baseline and the waveform's
tive deflection, i.e., a biphasic initially positive potential. negative spike diminishes (Fig. 9G; E - l ) . Dissipation of the
c
We may now address the consequences of the above informa- leading quadrupole permits its trailing quadrupole to enter the
tion on the denervated as opposed to innervated IAP's corre- crush zone with a corresponding increase in its current density.
sponding extracellularly recorded waveform. Let us assume that The second derivative of the trailing quadrupole's LD is no
an action potential is spontaneously generated at the fiber's longer negligible since the crush zone has shortened the LD's
origin (Fig. 9: 0 mm). The electrode (E-1 ) positioned at this lo-
A
length and proportionally increased its current density. The
cation initially detects the LD's negative sink region and records IAP's trailing quadrupole as a whole then dissipates as the hy-
an initial negative deflection (Fig. 9A; E-1 ). The LD continues
A
perpolarization phase returns to the resting membrane potential;
to grow as the membrane fully depolarizes with a corresponding all the while the trailing quadrupole's LD remains in place be-
increase in the generated waveform's initial negative deflection neath the electrode. The electrode records a long but detectable
(Fig. 9B; E - 1 ) . The LD continues to advance onto the fiber
A
terminal positive phase (Fig. 9H-L; E - l ) . This waveform has
c
with development of the TD. The E - 1 recording electrode

A
not been documented clinically and it remains to be verified.
records a decline in the negative spike with a manifestation of The 50-mm electrode located at the crush zone termination
the ensuing positive phase corresponding to the TD's positivity detects the approaching IAP's LD and its associated positivity
(Fig. 9D; E - 1 ) . At this point the denervated IAP's leading
A (Fig. 9E; E-1 ). As the IAP's LD enters the crush zone, it
D
quadrupole (+ — +) is fully appreciated. However, recall that shrinks into it with a corresponding elevation in its current density
60 — P A R T 1 FUNDAMENTAL PRINCIPLES
E-1A E-1 B
9
A (-+)
B
J: ±L
(+ ~K" +)
(+ -x+
(+ -X+
(+ -x+
(+
Figure 9. Single muscle fiber simulation A schematic is shown with a recording electrode in close proximity t o t h e fiber at the following lo-
cations: fiber's origin (0 mm: E-1 A), fiber's midpoint (25 mm: E-1 ), just proximal t o t h e fiber's termination (49.75 mm: E-1 ) , fiber termination (50
B c
mm: E-1 ), and just beyond the fiber's termination (50.35 mm: E-1 ).The potentials recorded from these locations are depicted t o the right for the
D E
corresponding electrode locations. A crush-type termination is modeled. (From Dumitru D, King JC, Rogers WE, Stegeman DF: Positive sharp wave
and fibrillation potential modeling. Muscle N e r v e 1999;22:242—251, with permission.)
(Figs. 7 E - H and 9F; E-1 ). The electrode records a relatively

D (Fig. 9E; E-1 ). In this model, an electrode positioned directly
E
large positive deflection. Membrane repolarization with a corre- at the crush zone termination records a rather large amplitude
sponding dissipation of the IAP's L D and TD results in a de- positive sharp wave, while an electrode a few hundred microns
cline of the waveform's initial positive deflection (Fig. 9G; beyond the crush zone detects a potential with a smaller ampli-
E-1 ). When the IAP's leading quadrupole has completely dis-
D tude. Therefore, the model assumes that the electrode crushes
sipated, the trailing quadrupole approaches the crush zone (Fig. the membrane a few hundred microns prior to the electrode in
9H; E - 1 ) . The IAP's hyperpolarization phase represented by
D order to generate waveforms with amplitudes correlated to those
the trailing quadrupole has its LD enter the crush zone, magni- clinically observed.5
fying the LD's current density. As a result, the electrode now

detects an appreciable negative voltage associated with the trail-
ing quadrupole's LD negativity, thereby generating a negative NEAR-FIELD A N D FAR-FIELD WAVEFORMS
spike (Fig. 9I-L; E-1 ). This negative phase is of long duration
D
since the hyperpolarization phase lasts for such a long time. The FAR-FIELD THEORY
total waveform is identical in appearance to that of the prototyp-
ical positive sharp wave. It is the combination of the IAP as con- In addition to explaining near-field waveforms, the L/TD
sisting of a depolarization and hyperpolarization phase as well model also can be used to conceptualize the generation of far-
as the crush zone mechanism that is proposed to account for the field potentials. In this section, the L/TD model is used to better
formation of the positive sharp wave. The crush zone is hypoth- appreciate the motor unit action potential's (MUAP) near-field
esized to occur from an adverse interaction between the muscle and far-field component waveforms. For this discussion, the
membrane and comparatively large exploring monopolar or MUAP is defined as the summated electrical activity recorded
concentric needle electrode. If there is little membrane affected, by an intramuscular electrode arising from most of the muscle
a fibrillation potential is recorded. However, if a crush type of fibers comprising a single motor unit. A motor unit consists of a
insult is produced by the electrode, the same IAP generates a single motor neuron and all of the muscle fibers it supplies.
positive sharp wave. This model successfully explains both Prior to progressing any further, it is necessary to briefly
types of waveforms (fibrillation potentials and positive sharp review a few aspects of far-field potential production. Un-
waves) originating from denervated muscle tissue as derived fortunately, there is no universally accepted definition of a far-
from the same IAP. An electrode located at any location distal to field potential. Perhaps the most appropriate explanation at this
the crush zone records a waveform with the same morphology time is to simply state that a far-field potential is the waveform
as an electrode at the crush zone itself, but of a smaller magnitude recorded at some distance from the generator source when the
Appendix THE LEADING/TRAILING DIPOLE MODEL A N D NEAR-FIELD/FAR-FIELD WAVEFORMS — 6 1
potential changes little in magnitude and shape over relatively

large distances. By this purposefully vague definition, a near-
field potential changes significantly in amplitude and possibly
shape with small changes in distance from the generator source.
A good place to begin a qualitative understanding of far-field
potentials is by considering the generator source and its sur-
rounding volume conductor. For our purposes a generator
source is either a non-moving or propagating current source
within a medium capable of conducting ions well, i.e., a good
volume conductor. Recall that it is not possible to have a
monopolar current source in isolation. Therefore, the simplest
generator we will consider is a dipole (+ - ) with current flowing
between the two poles and an associated voltage profile of
isopotential lines extending out into the enveloping volume con-
ductor (Fig. 2 - 2 ) .
If a recording electrode located close to a non-moving dipole
source positioned within an infinite volume conductor were
moved toward and then away from the source, a characteristic
voltage profile would be mapped. By definition, at an infinite
distance from the dipole generator, zero voltage would be
recorded (Fig. 10A). If the electrode first approaches the posi-
tive dipole, an initial positive deflection would be recorded that
would increase in magnitude until the positive pole were
reached. Electrode movement toward the negative pole would
i result in a decline in the waveform's positive phase. A zero po-
tential would again be recorded halfway between the positive
and negative pole. Continued electrode movement would result
in the detection of a large negative potential maximizing at the
negative pole. Finally, the potential would again reach zero volt- Figure 10. Dipole source. A , Spatial voltage profile for a dipole
age when the electrode was moved away from the negative pole source in an infinite volume conductor. B,The spatial voltage profile
toward infinity. for t h e same dipole depicted in A as measured along t h e central axis
Regarding far-field potentials in biologic sources as they of a centrally located dipole within an infinitely long cylinder. C, Spatial
relate to most of the electrodiagnostic medicine evaluation, a voltage profile along t h e cylinder's wall as opposed t o its central axis.
specific volume conductor configuration is of particular inter- D,The same dipole source in A, but with t h e reference electrode l o -
est. Specifically, the majority of waveforms recorded with either cated in t h e dipole's far-field region t o t h e left. (From Stegeman DF,
surface or needle electrodes arise from generator sources within Dumitru D, King JC, Roeleveld K: Near- and far-fields: Source charac-
long cylindrically shaped volume conductors such as limbs and teristics and t h e conducting medium in neurophysiology. J Clin
digits. The "long-and-thin" character of such volume conductor Neurophysiol 1997; 14:429—442, with permission.)
shapes has a specific characteristic with respect to potential
fields. If a dipolar current source is positioned within the center
of a cylinder, a unique voltage profile is generated (Fig. 10B). IOC). In practical terms, if the reference electrode were located
At a longitudinal distance of about 1.5 times the cylinder's in one of the far-field regions (beyond 1.5 times the cylinder's
radius from a centrally located dipole source, a region of mini- radius), then a zero potential would be documented when the
mal changing voltage is detected (Fig. 10B). Irrespective of how active recording electrode were positioned in this part of the
far one travels from the dipole source, once within this region, cylinder (Fig. 10D). Of course, the above discussion has cen-
the voltage no longer declines noticeably. This property of tered on dipole current sources. Recall that we defined our typi-
dipole voltage distributions in cylinders accounts for the clinical cal propagating action potential as represented by a quadrupole
observation of the same far-field potential being recorded over (+ — +) consisting of a leading (+ - ) and trailing ( - +) dipole,
long limb distances. Therefore, in both finite and infinite cylin- i.e., two back-to-back dipole sources. We may again conceptu-
ders, unlike truly infinite volume conductors in all directions, alize our action potential to consist of these two dipole sources,
there exists a region where the far-field potential voltage de- but this time acting independently (Fig. 11A-B). In Figure 11 A,
clines minimally, if at all! an action potential's leading dipole for example, is oriented with
It is this non-zero voltage at large distances from the source its anticipated far-field (constant voltage) and near-field (chang-
that is clinically recorded as a far-field potential during periph- ing voltage) voltage profiles displayed. In Figure 11B, the
eral nerve and somatosensory evoked potential far-field experi- action potential's trailing dipole is shown with its accompany-
ments. It should now be clear why these clinically observed ing far-field and near-field voltage profiles. Let us assume that
potentials extend over such large distances without a drop in the action potential is extended over a long piece of active tissue
voltage, i.e., beyond about 1.5 times the limb radius, there is no and the LD and TD far-fields can fully manifest. In this case, it
longer an appreciable voltage drop from the dipole source. can be seen that the opposite polarity far-field voltage distribu-
Similarly, recording along the cylinder's wall as opposed to its tions summate to cancel each other (Fig. 11C). The net result is
center records a different near-field voltage profile but a far- a quadrupolar source (+ — +) displaying the anticipated tripha-
field potential identical to that recorded for the central location sic near-field waveform morphology with cancellation of both
is again detected beyond 1.5 times the cylinder's radius (Fig. dipoles' far-field potentials. This is why balanced quadrupole
A . A .
B . B .
C.
^1V
++
+
Figure 11. Dipole source. A,A dipole source is depicted generat- Figure 12. Dipole source. A, A dipole source is shown with its
ing both a near-field and far-field voltage distribution in an infinite cylin- constituent near-field and far-field voltage distributions. B.This o p p o -
drical volume conductor. B,An equivalent dipolar source oriented sitely oriented dipole source is approximately half t h e magnitude of
opposite t o that in A with its accompanying voltage profile is shown. that shown in A and may have resulted from it crossing into a larger
C, Summating the t w o voltage distribution of t h e oppositely oriented volume conductor and hence lower voltages. C, Summation of t h e un-
dipoles generates a triphasic waveform. (From Stegeman DF, Dumitru equal voltages constituting the quadrupole action potential results in
D, King JC, Roeleveld K: Near- and far-fields: Source characteristics and the net generation of a far-field potential because of a superimposed
the conducting medium in neurophysiology. J Clin Neurophysiol net dipole source. (From Stegeman DF, Dumitru D, King JC, Roeleveld
1997; 14:429-442, with permission.) K: Near- and far-fields: Source characteristics and t h e conducting
medium in neurophysiology. J Clin Neurophysiol 1997; 14:429^442,
sources generate no far-field potentials, i.e., their identical far- with permission.)
field subcomponents exactly cancel each other.
The next logical question is what happens when the LD and be considered as a balanced quadrupole before and after en-
TD far-field regions do not exactly cancel each other. It is possi- countering the transition region. Over this region, however, the
ble for an action potential to encounter several situations or action potential becomes a quadrupole with an imbalance be-
transitions that lead to an imbalance between its LD and TD. tween its LD and TD. This implies that there is a superimposed
For example, an action potential may propagate across two non-propagating dipole source at the junction or boundary be-
compartments of differing sizes (e.g., hand to finger), enter a tween the two differing volume conductor regions. In situations
region of high or low resistance within the same compartment of changing extracellular conditions (size changes, resistance
(e.g., a nerve passing through a muscle), change direction (e.g., changes), it would be wrong to state that the dipole moments of
ulnar nerve across the flexed elbow), or reach the end of ex- the propagating source itself are unbalanced; rather, there is an
citable tissue (e.g., musculotendinous junction). It is the transi- imbalance in the extracellular domain encountered during pas-
tion between these differing circumstances that produces an sage of the dipoles. The term "virtual source" was previously
imbalance between the leading and trailing dipoles at the transi- used in the main text of the chapter. It is this superimposed non-
tion zone. One of the dipole's (e.g., trailing dipole) may be un- propagating dipole source that generates the detected far-field
changed from our previous example and display its potential and is recorded along the limb both proximal to and
subcomponent near-field and far-field voltage distributions (Fig. distal to the transition zone with an appropriate recording mon-
12A). However, the leading dipole may have entered an adjoin- tage. Of note, the far-field polarity will be opposite on one side
ing cylindrical compartment of larger volume with an associ- of the transition region compared to the other. The best elec-
ated smaller voltage for both its near-field and far-field trode montage to detect a far-field potential is to place the two
components (Fig. 12B). Summating both of these dipoles leads recording electrodes on opposite sides of the dipole source;
to a net far-field potential since the L D ' s far-field voltage fails however, other montages also can show far-field potentials. 9
to completely balance or cancel the T D ' s far-field voltage (Fig. The far-field potential will be generated for as long as the dipole
12C). There is a commensurate change in the near-field wave- source is present, which in turn depends on how long it takes for
form morphology as well (Figs. 11C and 12C). The source can the action potential to completely cross the transition zone. In
Table I. "TEN COMMANDMENTS" OF FAR-FIELD

POTENTIAL GENERATION
I. In an infinite volume conductor, a dipolar source cannot pro-
duce a far-field potential, nor can any other type of current
source.
II. A dipolar source located in a finite volume conductor can
produce a non-zero relatively constant potential difference
over distance, i.e., the far-field potential.
III. A far-field potential can best be observed when recordings are
made in a finite volume conductor that is stretched in one di-
rection, i.e. long and thin.
IV. Far-field potentials can be generated by the peripheral neuro-
muscular system.
V. A far-field potential can be best detected when the recording
electrodes are far from and on opposite sides of the dipolar
source.
VI. An impulse propagating along excitable tissue of sufficient
length t o contain t h e entire impulse lacks a dipole component,
and does n o t produce a far-field distribution by itself.
VII. Any disturbance in t h e constant propagation of an impulse
along excitable tissue, including directional changes (real dipo- 1 1
lar sources) and any anatomic alterations in the fiber's sur- ! 1

roundings (virtual dipolar sources), generates a dipolar 1 1
potential field and lead t o far-field potentials. +
3.7 mm 111 m m
VIII. Propagating impulses can induce transient dipolar sources at
sites of fixed anatomic transitions implying the dipolar fields
are not propagating, i.e., dipolar fields are non-moving. Figure 13. Intracellular action potential. A, The traditionally
modeled spatial IAP representation is shown. B.The actual human
IX. Balanced quadrupolar sources only have near fields, but dipo-
skeletal muscle fiber action potential has a much longer repolarization
lar sources have both near fields and far fields.
phase than is usually modeled (A). C,A modeled skeletal muscle fiber
X. A far-field potential can never be recorded with a bipolar intracellular action potential with its associated L/TD spatial represen-
montage using a short intereiectrode distance. tation (D).The horizontal line in A - C r e p r e s e n t s t h e resting m e m -
( F r o m Stegeman D F , D u m i t r u D , K i n g J C , R o e l e v e l d K: N e a r - a n d far-fields: brane potential. (From Dumitru D, King JC, Rogers WE: Motor unit
S o u r c e characteristics and t h e c o n d u c t i n g m e d i u m in neurophysiology. J C l i n action potential components and physiologic duration. Muscle N e r v e
N e u r o p h y s i o l 1997; 1 4 : 4 2 9 - 4 4 2 , w i t h permission.) 1999,22:733-741, with permission.)
our example, as the action potential exits the finger, the LD/TD muscle's insertion reveal the anticipated triphasic waveform
balance is again established and the far-field potential dissipates appearance with a small positive potential incorporated into the
commensurate with the decline of the superimposed dipole MUAP's terminal positive phase (Fig. 14C and D). If the same
source. The above explanation applies to all muscle or nerve MUAP is recorded from different longitudinal locations along
near-field and far-field sources with an associated "ten com¬ the muscle fiber, each further displacement from the endplate,
mandments of far-field potential generation (Table 1).
,,
the MUAP's main negative spike becomes more delayed. This
is to be expected since it takes longer for the action potential's
MOTOR UNIT ACTION POTENTIAL negative sink to reach the electrode. However, the small positive
potential and terminal aspect of the MUAP displays a constant
We can now apply the above information regarding near-field latency (Fig. 14C and D). Similarly, the MUAP's onset also re-
and far-field sources to a more accurate description of MUAPs. mains constant. It is also important to note that these MUAPs
Our beginning point is always the IAP source for a single fiber, demonstrate a total potential duration approaching 30 ms or
which in this case is a single muscle fiber. The traditionally used more as opposed to the anticipated 10-ms MUAP duration. - - 2 3 6 14
IAP to model single muscle fiber potentials is that of Rosen- Volume conduction models can be used to successfully simulate
falek - and is about 4 - 5 ms long (Fig. 13A). However, actual
19 26
arid explain these clinical findings (Fig. 14A and B ) . 6J1
intracellular recordings of human skeletal muscle reveal an Let us suppose an action potential is initiated at the neuro-
IAP with a very prolonged repolarization phase (greater than 20 muscular junction of a 50-mm simulated half muscle fiber (Fig.
ms) derived from the combined repolarization of the surface 15A). As previously discussed, a quadrupole develops about the
membrane and the transverse tubule system (Fig. 13B). It is neuromuscular junction with a leading dipole located to the left
possible to model this action potential (Fig. 13C) and then at- and right of the neuromuscular junction. An electrode posi-
tempt to explain clinical MUAP observations based on the tioned above the neuromuscular junction detects the quadru-
L/TD model representation of the human IAP (Fig. 13D). pole's negative voltage sink, thereby recording an initial
Monopolar needle electrode recordings of MUAPs obtained negative deflection (Fig. 15A; E-1 ). The quadrupole's LD then
A
from the biceps brachii muscle with a reference electrode on the propagates onto the muscle membrane and subsequently a TD is
134.6 \N
-13.2 \N
Figure 14. Clinically obtained and simulated MUAPs. A and B, Simulated MUAPs recorded 25 mm and 30 mm distal t o a muscle fiber's
endplate region. C and D, Clinically recorded MUAPs from t h e biceps brachii muscle.Vertical dashed lines I and 2 represent the onset of t h e near-
field and far-field potentials, respectively.The last vertical dashed line (3) signifies t h e potential's termination.
E-1 A E-1 B E-1 C E-1 A E-1 B E-1 C
Figure IS. Motor unit action potentials.The L/TD model is used t o explain t h e clinically recorded MUAPs.A 50 mm hemi-fiber muscle
length is depicted with action potential initiation at the neuromuscular junction (0 mm).Three point electrodes (E-I , E-I , and E - I ) are posi-
A B c
tioned at 0 mm, 25 mm, and 30 mm, respectively, with the recorded waveforms displayed t o t h e right. In E-H the superimposed dipole source (+
- ) is shown which generates the small-amplitude far-field potential. (From Dumitru D, King JC, Rogers WE: Motor unit action potential compo-
nents and physiologic duration. Muscle Nerve 1999;22:733-741, with permission.)
Appendix THE LEADING/TRAILING DIPOLE MODEL A N D NEAR-FIELD/FAR-FIELD WAVEFORMS — 6 5
generated. The recording electrode now detects the TD's posi- reflects the timeframe of IAP depolarization and repolarization.
tivity and documents a positive phase (Fig. 15B; E-l ) . It is im-
A If the above MUAPs were recorded with a concentric needle
portant to note that the IAP has a total duration approximating electrode, the terminal far-field potential component would be
30 ms because of the above-noted transverse tubular repolariza- very small or simply not appreciated because both the active
tion contributing to the overall IAP duration. If we assume the (core) and reference (cannula) of the concentric electrode are
IAP is propagating at 3.7 m/s, it has an overall spatial expanse located in the far-field region. 6
of about 115 mm. This means that if the muscle fiber were infi-
nitely long, the entire action potential would occupy 115 mm of COMPOUND MUSCLE ACTION POTENTIAL
the muscle's length. Since the muscle fiber in our simulation is
only 50 mm long, by the time the IAP's LD reaches the fiber's The logical sequence of beginning with the IAP, describing
termination, the entire membrane is supporting a single action its associated extracellular waveform, and summating all the
potential (Fig. 15C-H). Therefore, the recording electrode E - 1 A single muscle fibers comprising a single motor unit to generate
in our example (Fig. 15) describes a potential with an initial a MUAP ends with describing the summation of all the MUAPs
negative deflection followed by a terminal positive phase. constituting a compound muscle action potential (CMAP). The
There are two additional electrodes positioned about the L/TD model can be used to describe both the near-field and far-
fiber's middle in our example (Fig. 15; E - 1 and E - l ) . As the
B c field CMAP subcomponents. We may use the prototypical
IAP's LD approaches the electrode at 25 mm (E-1 ), it recordsB CMAP recording from the abductor pollicis brevis (APB) fol-
the initial positive voltage (Fig. 15B). Continued IAP propaga- lowing median nerve wrist activation to begin.
tion permits the same electrode to then detect the quadrupole's In order to record the APB's CMAP, an active recording elec-
negative poles while the E - l electrode documents the leading
c trode is positioned over the muscle's motor point (muscle mid
positive voltage (Fig. 15C; E - 1 and E - l ) . Passage of the IAP
B c belly) with a reference electrode located on the first digit distal
then results in the E - 1 electrode observing the TD's declining
B to the metacarpophalangeal joint. The anticipated CMAP shape
negative voltage, while the E - l electrode detects the quadru-
c
pole's negative voltage (Fig. 15C; E - l ) . Just before the IAP's

c
i LD encounters the fiber's termination, these two electrodes

record a terminal positive phase of slightly different magnitude
(Fig. 15D). So far, the electrodes record the anticipated triphasic
potential. However, when the quadrupole's LD encounters the
' musculotendinous junction, an interesting series of events
occurs.
Since there is no longer any active tissue over which the IAP
can propagate, its LD begins to collapse (Fig. 7 A - D ) .
Dissipation of the LD creates a situation where the LD no
longer balances the TD, resulting in the emergence of a super-
imposed dipole source at the musculotendinous junction (Figs.
12C and 15E). This dipole source grows in magnitude as the im-
balance between the LD and TD increases with a maximum
dipole source magnitude coincident when the LD has just com-
pletely disappeared (Fig. 15F). Because the dipole source is lo-
cated in the cylindrically shaped limb, it generates a far-field
potential. All three recording electrodes ( E - 1 , E - 1 , and E - l )
A B c
are located in the positive region of the superimposed dipole

source and record a positive waveform deflection (Fig. 15E-F).
The maximum positive peak of the far-field potential occurs
when the LD has just completely dissipated. Since the IAP's TD
can no longer propagate, it simply declines, which requires its
complete duration of 30 ms to accomplish. Recall that the LD
has a duration of about 1 ms while the TD has a duration of ap-
proximately 30 ms. Therefore, the dipole source gradually dissi-
pates with its positive pole facing the three electrodes also
slowly diminishing in magnitude (Fig. 15F-H; E - 1 , E - 1 , E- A B
cl ) . The net result is a waveform with an initial negative deflec- Figure 16. C o m p o u n d m u s c l e a c t i o n p o t e n t i a l s . A, Biphasic
tion when recorded at the endplate but an initial positive initially negative CMAP obtained from t h e APB with an active elec-
deflection outside of the endplate region. The portion of the t r o d e on the muscle's m o t o r point and a reference on the distal
waveform from initial baseline onset to initiation of the positive thumb. B, Similar recording montage as in A; however, 30 stimuli w e r e
far-field potential represents the near-field component of the averaged and the sweep speed was slowed t o 10 ms/div. C, Biphasic
single muscle fiber discharge (Fig. 14A-B). The portion of the initially negative CMAP from the ADM with a bilobed negative spike.
waveform from far-field potential onset to waveform termina- D, Relocating the active electrode from t h e ADM t o the lateral epi-
tion signifies the waveform's far-field potential component (Fig. condyle reveals a quadraphasic initially negative far-field potential. E,
14A-B). Summating the electrical potentials of all the single Same recording as that in C, but with a sweep speed of 10 ms/div.
muscle fibers comprising a motor unit generates the MUAP N o t e : NI and N 2 refer t o t h e first and second negative peaks while
with its near-field and far-field subcomponents (Fig. 14C and PI-P2 refer to the first through second positive peaks.The designation
D). Since the total IAP is about 30 ms, the total MUAP duration "F" demarcates the CMAP's associated F-wave.
junction (Fig. 17M; #1 and #2). This results in a large separa-

tion between the terminal positive repolarization phase for the
waveform and the initiation of the far-field potential. With a
short fiber length (Fig. 17A; #1 and #2), the action potential en-
counters the fiber's termination shortly after forming the wave-
form's negative spike with a subsequent fusion of the fiber's
repolarization phase for the recovering membrane as well as the
positive far-field potential derived from action potential dissipa-
tion. The net result is a biphasic initially negative potential
(Figs. 16A and 17A). Therefore, the terminal positive phase
(Fig. 16A; P I ) for the APB CMAP represents some portion of
the anticipated membrane repolarization as well as the sum-
mated far-field potentials of all the APB's fibers encountering
the musculotendinous junction because of the short fiber length.
Altering the sweep speed for the APB recording from 3
ms/div to 10 ms/div reveals that the CMAP's termination re-
quires approximately 62 ms with respect to the waveform's
onset (Fig. 16B). Again, this finding can be explained by con-
sidering the single muscle fiber L/TD explanation (Fig. 15).
Recall that the waveform's long terminal phase arises from the
action potential's long repolarization phase that only becomes
manifest during the formation of a quadrupole imbalance. This
same situation applies to the CMAP's summated action poten-
tials. However, instead of the 30-ms duration, the CMAP has a
terminal phase approximately twice that long. This can be ex-
plained by considering that there is in effect a double discharge
occurring in the muscle following median nerve wrist activa-
tion. The fibers are all excited initially by the neural stimulus,
Figure 17. Single muscle fiber s i m u l a t i o n s . A series of 13 single and then a small percentage of fibers are secondarily activated
muscle fiber lengths (10-130 mm) are shown in A-M with a recording (F-wave, Fig. 16B; "F"). Combining the far-field potentials
electrode positioned at the muscle's endplate (column I) and midway from both activations and averaging a number of responses per-
between the endplate and musculotendinous junction (column 2) with mits detection of the CMAP's far-field potential and its associ-
a reference at infinity. N o t e how the far-field potential (small positive ated long return to baseline.
peak) is merged with t h e short fiber length (10 mm) and is progres- The CMAP derived from the abductor digiti minimi (ADM)
sively displaced from it as the muscle fiber length increases from 10 and other hypothenar muscles following ulnar nerve stimulation
mm to 130 mm for both electrode locations. (From Dumitru D, King usually demonstrates an initially negative biphasic CMAP;
JC: M o t o r unit potential duration and muscle length. Muscle N e r v e however, the negative spike is typically bi-lobed or has a notice-
1999;22:1188—1 195, with permission.) able "hump" in its rising phase (Fig. 16C). This waveform's
morphology has been attributed to the active recording elec-
conforms to that of a biphasic initially negative potential (Fig. trode positioned over several muscles in the hypothenar emi-
16A). This waveform morphology is essentially the same as that nence yielding two negative peaks. However, a similar shape is
previously described for an electrode located over the muscle's not seen for the APB's CMAP, which also arises from several
endplate region (Fig. 15; E-1 ). It is important at this juncture to
A
muscles. Relocating the active electrode from the abductor
point out the differences between endplate recordings per- ADM motor point to the lateral epicondyle reveals primarily a
formed for the single muscle fiber in Figure 15 and the CMAP triphasic initially positive waveform (Fig. 16D). Since the active
from the APB (Fig. 16A). The single muscle fiber waveform re- electrode is on the lateral epicondyle and the reference electrode \
veals a distinct positive far-field potential with respect to the inion the distal fifth digit, the observed waveform must be a far-
tially negative near-field waveform originating from the field potential. Investigations have concluded that the CMAP's
endplate. However, the APB waveform similarly has the ini- second negative peak is derived from the medial interos-
tially negative onset, but a negative spike that seamlessly blends seous/lumbrical muscle's action potentials dissipating at their
into a terminal positive phase (Fig. 16A; N l - P l ) . This discrep- respective musculotendinous junctions forming far-field poten-
ancy can be understood if one considers the effect of different tials that coincide with the ADM's near-field w a v e f o r m s .
5013
fiber lengths on the ensuing waveform morphology. Let us con- The net result is a CMAP with a bi-lobed negative spike par-
struct a series of single muscle fibers ranging in length from 10 tially derived from the hypothenar muscles' near field (Fig.
mm to 130 mm (Fig. 17), We can then place a recording elec- 16C; N l ) and interosseous/lumbrical muscles' far-field compo-
trode at the muscle fiber's endplate (Fig. 17A-M; #1) and nents (Fig. 16C; N2). This waveform's terminal positive phase
halfway between the endplate and the fiber's musculotendinous is likely derived from the same mechanism generating the
junction (Fig. 17A-M; #2) for each fiber length. The far-field APB's terminal positive phase, i.e., dissipation of the ADM's
potential demonstrates the same duration for each fiber length action potentials at its musculotendinous junctions. Finally, al-
since it depends on the IAP shape beginning when the action tering the sweep speed of the ADM recording also reveals a
potential encounters the musculotendinous junction. If the fiber long terminal positive phase duration approaching 60 ms (Fig.
is long, it will take a relatively long time before the action po- 16E; "F") with an explanation similar to that previously given
tential initiated at the endplate terminates at the musculotendinous for the APB.
The above discussion clearly shows that the intuitive notion 7. Dumitru D, King JC: Motor unit potential duration and muscle length. Muscle
Nerve 1999;22:1188-1195.
of recording the CMAP from the ADM when the active ( E l )
8. Hugues M, Schmid H, Romey G, et al: The Ca2 -dependent slow K conduc-
+ +
electrode is placed over the hypothenar eminence is a miscon- tance in cultured rat muscle cells: Characterization with apamin. EMBO J
ception. The recorded C M A P is a complex mixture of both 1982;1:1039-1042.
near- and far-field contributions from all ulnar-innervated hand 9. Jewett DL, Deupree DL: Far-field potentials recorded from action potentials and
from a tripole in a hemicylindrical volume. Electroencephalogr Clin Neuro-
muscles. Depending on the number and configuration of mus- physiol 1989;72:439-449.
cles innervated by a particular nerve the CMAP can have differ- 10. Kincaid JC, Brasher A, Markand ON: The influence of the reference electrode on
ent forms and distribution. This was shown by recording CMAP configuration. Muscle Nerve 1993;16:392-396.
CMAP's at multiple electrode positions over the hand and foot 11. Lateva ZC, McGill KC: The physiological origin of the slow afterwave in
muscle action potentials. Electroencephalogr Clin Neurophysiol 1998; 109:462¬
area and stimulating the ulnar, median, peroneal, and tibial 469.
nerve resulting in C M A P cartography. The distribution of an 12. Lorente de No' R: A study of nerve physiology. Studies of the Rockefeller
area of a high and well-defined negative peak was remarkably Institute for Medical Research 1947; 132:384-^82.
13. McGill KC, Lateva ZC: The contribution of the interosseous muscles to the hy-
different for the studied nerves, being small for the median and
pothenar compound muscle action potential. Muscle Nerve 1999;22:6-15.
peroneal and large for the tibial and ulnar nerves. 25
14. Nandedkar SD, Dumitru D, King JC: Concentric needle electrode duration mea-
surements and uptake area. Muscle Nerve 1997;20:1225-1228.
15. Noble D: Application of Hodgkin-Huxley equations to excitable tissues. Physiol
Rev 1966;46:1-50.
CONCLUSION 16. Nunez PL: Electric Fields of the Brain: The Neurophysiology of EEG. New
York, Oxford University Press, 1981;82-83.
The leading/trailing dipole model can be used to described a 17. Prinbow D, Johnson-Pais T, Bond CT, et al: Skeletal muscle and small-conduc-
tance calcium-activated potassium channels. Muscle Nerve 1999;22:742-750.
number of near-field and far-field waveforms observed during 18. Purves D, Sakman D: Membrane properties underlying spontaneous activity of
the routine electrodiagnostic medicine examination. The above- denervated muscle fibers. J Physiol 1974;239:125-153.
defined principles can be applied to situations not detailed in 19. Rosenfalek P: Intra- and extracellular potential fields of active nerve and muscle
fibers. A physico-mathematical analysis of different models. Acta Physiol Scand
this discussion to explore new aspects of near-field and far-field
1969;(Suppl 321): 1-168.
sources in biologic systems. The interested reader is encouraged 20. Schmid-Antomarchi H, Renaud J, Romey G, et al: The all-or-none role of inner-
to apply the simple quadrupole model when familiar or even vation in expression of apamin receptor and of apamin-sensitive Ca -activated
2+
new waveforms are observed. K channels in mammalian skeletal muscle. Proc Natl Acad Sci USA 1985;
+
82:2188-2191.
21. Stegeman DF, Dumitru D, King JC, Roeleveld K: Near and far fields: Source
characteristics and the conducting medium in neurophysiology. J Clin
REFERENCES Neurophysiol 1997; 14:429-442.
22. Thesleff S: Physiologic effects of denervation on muscle. Ann N Y Acad Sci
1. Dumitru D, Jewett DL: Far-field potentials. Muscle Nerve 1993;16:237-254. 1974;228:89-103.
2. Dumitru D, King JC, Nandedkar SD: Motor unit action potential duration 23. Thesleff S, Ward MR: Studies on the mechanism of fibrillation potentials in den-
recorded by monopolar and concentric needle electrodes: Physiologic implica- ervated muscle. J Physiol 1975;244:313-323.
tions. Am J Phys Med Rehabil 1997;76:488^93. 24. Thesleff S: Fibrillation in denervated mammalian muscle. In Culp WJ, Ochoa J
3. Dumitru D, King JC, Nandedkar SD: Motor unit action potentials recorded with (eds): Abnormal Nerves and Muscle Generators. New York, Oxford University
concentric electrodes: physiologic implications. Electroencephalogr Clin Neuro- Press, 1982, pp 678-694.
physiol 1997;105:333-339. 25. Van Dijk JG, van Benten 1, Kramer CG, Stegeman DF: CMAP amplitude cartog-
4. Dumitru D, King JC, Stegeman DF: Endplate spike morphology: A clinical and raphy of muscles innervated by the median, ulnar, peroneal, and tibial nerves.
simulation study. Arch Phys Med Rehabil 1998;79:634-640. Muscle Nerve 1999;22:378-389.
5. Dumitru D, King JC, Rogers WE, Stegeman DF: Positive sharp wave and fibril- 26. Van Veen BK, Wolters H, Wallinga W, et al: The bioelectric source in computing
lation potential modeling. Muscle Nerve 1999;22:242-251. single muscle fiber action potentials. Biophys J 1993;64:1492-1498.
6. Dumitru D, King JC, Rogers WE: Motor unit action potential components and 27. Wiechers DO: Electromyographic insertional activity in normal limb muscles.
physiologic duration. Muscle Nerve 1999;22:733-741. Arch Phys Med Rehabil 1979;60:359-363.
Chapter 3
Instrumentation
CHAPTER OUTLINE
Basic Electronic Circuit Principles Analog to Digital (A/D) Conversion

Electronic Circuits (Ohm's Law) • Alternating Current (AC) Vertical Resolution • Horizontal Resolution • Latency
Circuits Measurement * Averaging • Averaging Principles • Signal-to-
Noise (S/N) Ratio
Electrodes
Surface Electrodes • Electrode Types • Electrode Size Cathode Ray Tube (CRT) Displays
• Electrical Continuity • Electrode Separation • Needle Cathode Ray Tube • Analog Signal • Digital Signal • Trigger
Electrodes • Electrode Types/Impedance and Delay Line
Amplifiers
Stimulators
Amplification * Input Impedance Activation of Excitable Tissues • Stimulator Types • Surface vs.
Needle Excitation • Stimulator Placement Errors • Stimulus
High- and Low-Frequency Filters Artifact
Biologic Signals • Filters • Low-Frequency (High-Pass) Filter
• High-Frequency (Low-Pass) Filter • Filter Effects on MUAPs Interference
• Notch Filter • Recommended Filter Settings
Sound Electrical Safety
Performing an electrodiagnostic medicine consultation re- number of signal acquisitions to make it more distinct from
quires a thorough understanding of how the electrophysiologic random background noise, i.e., average a number of trials and
instrument processes and displays biologic signals. This knowl- improve the signal-to-noise ratio. Visual and acoustic informa-
edge will help the practitioner avoid misinterpreting erro- tion required to diagnose potential pathology can be gathered as
neously derived data due to instrumentation errors as possible it occurs spontaneously from voluntary activity or evoked
pathology. The basic principles covered in this chapter are through a time-locked depolarizing stimulus delivered from the
shared by the majority of instruments likely to be used. instrument's stimulator. The most important aspect to remember
Recording electrodes located in or on the patient detect the rela- regarding instrumentation is that the skills necessary to practice
tively small voltage changes associated with a nerve or muscle electrodiagnostic medicine ultimately depend on the practi-
action potential and convey them to an amplifier (Fig. 3-1). tioner and not the instrument irrespective of its sophistication.
Following amplification, the signal is filtered to remove any ex-
traneous electrical activity that may distort the waveform under
investigation. The real-time (analog) information can then t>e BASIC ELECTRONIC CIRCUIT PRINCIPLES
presented acoustically and visually. The signal is usually con-
nected to an audio speaker so that the sound associated with an ELECTRONIC CIRCUITS (OHM'S LAW)
action potential can be appreciated. The analog data can also be
presented visually in one or both of the following formats. An Only a basic understanding of electronic circuitry is required
analog waveform can be displayed on a cathode ray tube (CRT) to adequately appreciate the electrophysiologic instrument's
for immediate visual inspection, or with a minimal delay trans- function and effect on biologic signals. We can begin with a
formed into a binary (digital) signal through an analog-to-dig- simple circuit containing a battery (E) (voltage/current source)
ital (A/D) converter and then displayed (Fig. 3-1). Once the connected to a resistor through wires (Fig. 3-2A). The battery is
waveform is digitized, it is possible to store and summate a symbolized by a series of short and long horizontal lines. The
69
circuit or to each of its parts. A unit of resistance is the ohm and

Amplifier (B) is defined as: 1 ohm = 1 volt/1 ampere. A symbol for ohm is
15
o-vw-j—o
TC the Greek letter omega (H). If two resistors are placed sequen-
A/D
o——*—o Converter tially in the above circuit, they are said to be in series. The cur-
rent passing through each resistor is the same, but the battery's
o » —o
4X1
Speaker (F) source voltage is equal to the sum of the voltage drops across
Active Ref Ground each resistor (Fig. 3-2B). The total voltage of the circuit is equal
Patient (A) to the sum of the voltages across each resistor in the circuit:
Etotai = Ej + E and substituting the current and resistance for
2
Isolated each resistor component in the circuit results in E = IR, +

total
Stimulator
IR . By factoring out the constant current term, resistances in
2
series can be summated: E totaI= I(R, + R ) . A circuit with sev-

2
Figure 3-1. Schematic representation of the electrophysio- eral resistors in series, therefore, may act as a voltage divider
logic instrument's component parts. Electrodes in o r on t h e pain that the total voltage drop across all of the resistors consists
tient (A) detect action potentials and transmit them t o a differential of the sum of the voltage dropping across each resistor ( E = lotal
amplifier (B).The differential signal is filtered (C) and then fed t o both E, + E +...+ E ) .
2 n
the loudspeaker (F) for aural analysis as well as t o the analog-to-digital In addition to a resistor, one can also place a capacitor in a
c o n v e r t e r (D) for a digital representation of t h e analog signal. The circuit (Fig. 3-2C). A capacitor is a device that stores charge.
signal is then ready for visual display on t h e cathode ray t u b e (E). Typically, a capacitor consists of a pair of parallel metal plates
Excitation of the peripheral nervous system is possible with the stimu- separated by an insulator, i.e., a material that conducts current
lator (G). (From Dumitru D, Walsh NE: Electrophysiologic instrumen- poorly. The insulator, also called a dielectric, may be air, oil,
15
tation. In Dumitru D (ed): Clinical Neurophysiology. Philadelphia, glass, wax paper, or some other poorly conducting material. The
Hanley & Belfus, 1989, with permission.) capacitor's capacitance (C) is the property that determines its
ability to accumulate charge and is defined as the ratio of the
charge on either plate to the potential difference across both
short line is the negative terminal while the long line is the pos- plates. If an open switch is placed in the circuit, an open circuit
itive terminal. A resistor (R) is a relatively poor conductor that exists and current will not flow. When the switch is first closed,
offers a quantifiable hindrance to the flow of charge and is sym- the circuit is completed and current flows from the positive bat-
bolized by a jagged line in circuit diagrams (Fig. 3-2). The 15
tery terminal to the capacitor plate to which it is connected. This
flow of unit charge per unit time is called current (I). It is as- plate then accumulates a positive charge until the voltage across
sumed that the wires connecting the resistor and battery offer the capacitor plate is equal to the potential of the battery. A sim-
minimal opposition to current flow. By convention, current ilar process occurs at the capacitor's plate connected to the neg-
flows from the positive to negative battery terminal across the ative battery terminal except that the charges are negative.
resistor, whereas in metallic circuits electrons travel in the op- During the charging process, although current did not actually
posite direction. If one measures the current passing through the cross the capacitor, a potential across the capacitor developed in
resistor when the voltage applied across the resistor is changed, that positive charges accumulated on one plate and were re-
the current is found to be directly proportional to the applied moved (accumulation of negative charges) from the other plate.
voltage. This qualitative statement may be expressed quantita- As it takes a finite time to charge the capacitor, current only
tively by a mathematical expression. The current (I) is equal to flows during this charging process. If the switch is again
the voltage (E) times a constant (g) called conductance that is a opened, the capacitor will hold its charge until it is discharged
property of the resistor: I = g x E. It is more common, however, by some external process where the plates are connected by
to use the inverse of the conductance (1/g) i.e., the resistance some form of electrical conductor, e.g., a wire.
(R) of the resistor (R = 1/g). The mathematical expression I = g
x E then becomes: I = E/R or E = I x R (Ohm's law). A restate- ALTERNATING CURRENT (AC) CIRCUITS
ment of the mathematics is that the current in a circuit is directly
proportional to the voltage applied to the circuit and inversely In the above discussion, Ohm's law is described for circuits
proportional to the circuit's resistance (I = E/R). This relation- in which the applied voltage remains constant. In biologic sys-
ship is named for Georg Ohm and may be applied to an entire tems, however, the signals of interest may have a fluctuating or
alternating voltage or current source. Consideration of alternat-
ing current circuits gives one a more realistic appreciation of
B events associated with biologic current flows, i.e., action poten-
Ri
tials. If a circuit only contains a resistive component, Ohm's law
can be used with minimal modification. In an AC circuit, the
hindrance a circuit element provides to AC current is no longer
called resistance (R) but impedance (Z). - A capacitor, how-
15 53
ever, behaves somewhat differently than a resistor in an alternat-

ing circuit. The alternating current generator (counterpart to the
Figure 3-2. Simple circuit diagrams. A,The current source (bat- DC circuit's battery) causes electrons to surge back and forth
tery) generates a voltage difference t h a t d r o p s across t h e resistor. through the generator to one capacitor plate and then the other,
Recall Ohm's law for a direct (E = I x R) and alternating (E = I x Z) i.e., charge/discharge. During the charging phase, like charges
current. B,An additional resistor has been placed in the circuit and a accumulate on the capacitors plates that tend to repel continued
voltage drop occurs across each resistor forming a voltage divider. C, charge addition. As a result, the current in the circuit is some-
O n e of the resistors in B has been replaced with a capacitor. what smaller than it might be if the capacitor were not present.
Chapter 3 INSTRUMENTATION — 71
A B C D
0 ©-> e ® © © ©5*" © 0 © <-© ©

© © «-© © ©->
© ©» © <-© ©
0 ©-*•
© ©*• 0 © ©
0 0 &-> ©
© ©-»• © © 0© © ©
© © 0
e © ©a* © © ©-> 0©©
© © © © ©^ © 0 © <-© ©
0 «r©
0 0©
© ©*»
©
0
© © © © ©
0 0 «e© ©
Figure 3-3. F o r m a t i o n of a n e l e c t r o d e p o t e n t i a l . A, Ions flow from the electrode into the electrolyte solution. B.The ions accumulate in t h e
electrolyte solution and an uneven charge distribution develops. C, Ions flow from the metal surface into solution and from the solution back into
the metal at unequal rates. D, A situation develops at some time where an ionic equilibrium may not be present. Movement of the electrode will
cause this voltage difference t o discharge possibly interfering with recording of the potential under investigation. (From C o o p e r R: Electrodes. Am J
EEGTech 1963;3:91-101, with permission.)
The opposition a capacitor has to an alternating current is called hindrance (impedance) to the energy contained within a wave-
its capacitive reactance (X ) and for our purposes represents
c form. The skin/electrode interface may have an impedance
the capacitor's impedance (Z). The capacitive reactance's units ranging from 500 Q to 9 M Q (M: mega or million) depending
are ohms and its magnitude is expressed mathematically as : X c upon the measurement location. - One way to improve the
39 54
= l/(27ifC) = Z. Although this formula looks imposing, there are signal's amplitude is to apply an electrolyte cream between the
two simple concepts to remember. First, as the capacitance (C) recording electrode and skin's surface stratum corneum layer.
increases, the reactance or impedance to current flow decreases, The ions within the electrolyte improve the transmission of
i.e., mathematically the denominator increases in magnitude, biologic signals through the skin to the electrode's metal sur-
therefore the net result or impedance must decrease. Second, the face. With time, however, a chemical reaction occurs at the elec-
capacitive reactance (X ) is frequency (Independent. In a direct
c trolyte/electrode interface and a voltage difference or electrode
current (DC) circuit, the frequency of the current is zero and the bias potential can be produced. The electrode potential is be-
14
reactance becomes infinite, implying that a constant (DC) cur- lieved to arise from two opposing processes: metallic ions pass-
rent will not flow across the capacitor in the steady state. This ing from the electrode into solution, and the subsequent
suggests that low frequencies have difficulty crossing a capaci- deposition of metallic ions back onto the electrode (Fig. 3-3).
tor. On the other hand, high frequencies reduce the capacitive The rates of these two ionic movements are not necessarily
reactance so that little opposition to current flow is present, al- equal and if there are more ions leaving the metal than return-
lowing these frequencies of current to pass without diffi- ing, an excess of positive ions will be in solution compared to
c u l t y . ' Magnetic fields induced by alternating current flow
553
the electrode's surface. This charge separation generates the
also oppose the current emanating from the generator. For prac- electrode bias potential. Slight movement of the electrode can
tical purposes, this inductive reactance can usually be ignored cause this voltage to suddenly discharge and disturb the wave-
in the human body. The opposition to an alternating current,
47
forms under investigation. A special cup-shaped electrode
therefore, is called impedance (Z) and composed of the resis- design can minimize this possible discharge voltage artifact (see
tive and capacitive elements in a circuit. Ohm's law can be ap- below).
plied to biologic systems using impedance and is usually A second way to reduce the skin's impedance is to gently
written as: E = I x Z. remove several layers of the stratum corneum by mild abrasion
with a pumice solution or fine sandpaper. The smaller the po-
tential to be recorded, the more important it becomes to reduce
ELECTRODES the skin's impedance. When performing routine nerve conduc-
tion studies, it is rarely necessary to aggressively abrade the
SURFACE ELECTRODES skin. The amplitude of somatosensory evoked potentials, how-
ever, may be less than a microvolt and require the skin's imped-
The action potential's magnitude generated by different ex- ance to be reduced below 5,000 Q. 48
citable tissues in the body may vary between several tenths of a

microvolt (uV) to possibly 50 millivolts ( m V ) . In order for 34
Electrode Types
these signals to be analyzed, they must be recorded with an Surface electrodes are used to evaluate sensory nerve and
electrode. This electrode can be located on the skin surface* or motor muscle compound action potentials, somatosensory
inserted through the skin and placed in close proximity to either evoked peripheral and cortical potentials, and occasionally sum-
the nerve or muscle generator. As surface electrodes are used mated motor unit action potential activity. Because different
for many recording purposes, we must consider one of the recordings are made from various regions of the body, a large
major partitions between the action potential and recording variety of electrodes are necessary, e.g., discs, disc electrodes
electrode, i.e., the skin. Although the skin performs vitally im- embedded in a plastic bar, self-retaining clips, adjustable wire
portant biologic functions, it does have a tendency to reduce the nooses, or saline-soaked felt pads. Aside from the saline pads,
amplitude of the potentials under investigation when surface most electrodes are made from a metal that readily conducts
electrodes are used. The potential's magnitude is diminished for current (e.g., silver, gold, tin, stainless steel, platinum, lead, or
surface recordings because the skin offers a certain amount of nichrome).
A commonly used surface electrode is a flat or cup-shaped amplitudes also applies for these self-adhesive electrodes. A
disc with a diameter ranging from 0.5 to 2.5 cm. The electrode single study has found that for a limited number of examined
that is cup-shaped or one with a small depression in its center motor and sensory studies, the potentials latencies were equiva-
may obtain recordings with less artifact than simply a flat sur- lent for the standard stainless steel electrodes and self-adhesive
face. This is because the electrode's central depression can be types. Aside from the above-noted amplitude differences with
3
filled with the electrolyte and any movement between the skin respect to size, all amplitude measurements were still well
and electrode occurs at the skin/electrolyte interface away from within reference values irrespective of the electrode type used.
the recessed electrolyte/electrode interface. - Because move-
14 34
It is still a good practice to establish one's own reference popu-
ment occurs away from the region of the electrolyte/electrode lation when using a new type of electrode.
interface, there is less likelihood of discharging the electrode
bias potential. Electrical Continuity
Aside from the self-retaining electrodes, adhesive tape is After appropriate electrodes have been chosen, a good prac-
commonly used to secure the recording electrodes to the skin. tice is to routinely assess their inherent electrical continuity and
Collodion, a quick-drying liquid adhesive may be used in the the electrode/skin contact. It is necessary to check the electrical
recording of evoked potentials. Although this is a relatively
48
integrity of recording electrodes because the wires underneath
common practice, it is potentially untidy, gives off malodorous the intact plastic coating can unknowingly break. The recording
fumes, requires a drying apparatus, and is somewhat time-con- electrode's resistance is measured with an ohmmeter. The ohm-
suming. Securing electrodes with Collodion offers the advan- meter passes a direct current through the wire and measures the
tage of stable and relatively artifact-free recordings that can be resistance. Intact electrode leads should have a resistance less
15
performed over a prolonged period. An alternative to Collodion than several thousand o h m s . If the resistance is noted to be
48
for the recording of evoked potentials is a thick and sticky elec- greater than this value, a faulty lead is likely present and the
trolyte cream applied to the electrode, which is then covered electrode should be discarded.
with cotton and tape to maintain the electrodes' position. When the electrode is secured to the skin's surface, the elec-
An important aspect of the electrode is also the cable con- trical continuity between the skin and electrode can be assessed
necting the electrodes to the amplifier. This is often a source of by measuring the impedance. Most modern instruments have
artifacts by picking up electromagnetic radiation from the envi- built-in impedance meters that allow the practitioner to easily
ronment. It can be circumvented by not using unnecessarily determine the impedance by passing a weak alternating current
long cables of equal length. The use of twisted cables can of a fixed frequency through the electrodes. The alternating
48
promote the pickup of similar amounts of electromagnetic radi- current is chosen because it is more representative of a time-
ation which by way of the common mode rejection (see varying biologic signal. Additionally, alternating current limits
Amplifiers) will be adequately suppressed. It is also important the polarizing effect a direct current may have on the electrodes.
to arrange the wires in such a way that cables from the stimula- As previously stated, it is common practice to record the imped-
tor and recording part of the equipment are neatly separated and ance when performing evoked potentials. One prefers to keep
do not cross. the impedance between 1,000 O and 5,000 Q for evoked poten-
tial studies. If an impedance less than 1,000 Q. is observed, it
48
Electrode Size is possible that an undesired conduction pathway such as excess

An important but all too infrequently discussed aspect of electrolyte gel on the skin's surface or perspiration linking both
recording surface action potentials, in particular CMAPs, is the electrodes is present. An impedance greater than 5,000 Q. for
effect electrode size has on the potential. There can be consider- surface electrodes can attenuate the signal. Impedances signifi-
able inter-trial variation on CMAP parameters and certainly 27
cantly larger than 5,000 Q. despite adequate skin preparation
one contributing factor is electrode size. - - - In general, larger
2 56 57 59
may suggest a loose electrode/skin contact or broken lead wire.
compared to smaller recording electrodes result in CMAPs with It is important that both electrodes have a similar impedance
slightly smaller amplitudes and area and slower conduction ve- when in contact with the skin otherwise an impedance mis-
locities, but no clinically significant difference regarding onset match can reduce common mode rejection (see Amplifiers) and
latency or negative spike duration. This is understandable if one increase 60-cycle interference, adversely affecting the desired
considers the fact that a larger electrode will obviously record signal.
from a larger region of the volume conductor, thus averaging
out the different contributions from individual motor units. Electrode Separation
Further, the summated electrical activity from a muscle will be Surface electrodes are usually placed in specific locations to
"averaged out" over a larger piece of metal, thereby decreasing optimize waveform recordings. In assessing compound muscle
the overall CMAP's amplitude compared to a smaller electrode action potentials (CMAPs), the active or E-l electrode is
size. It has been shown for CMAP amplitudes that larger elec- placed on the muscle's motor point and the reference or E-2
trode size results in lower CMAP amplitudes and reduced vari- electrode is located in an electrically, inactive region (see
ability; however, the effects on possible diagnostic utility is Chapter 2). This electrode montage will maximize the poten-
unknown. At present, there remains a lack of uniformity of
57
tial's magnitude as it minimizes common mode signals from
opinion and hence practices regarding the preferred size for sur- being recorded (see Amplifiers). The effect various electrode
face electrodes. locations have on CMAPs can be readily demonstrated (Figs.
Manufacturers have introduced self-adhesive silver-silver 3-4A and B). When the E-l electrode is placed over the
chloride (Ag/AgCI) electrodes for nerve conduction studies muscle's motor point and E-2 distally away from activated
similar to those used for years in the electrocardiology (EKG) muscle tissue, a large biphasic CMAP is recorded. Relocating
laboratory. These electrodes are quite flexible and can be easily the E-2 electrode onto the same muscle as E-l results in a
altered in shape by cutting them to fit any desired region. The marked CMAP amplitude reduction, minimal peak latency
above general principles of larger electrodes yielding smaller shortening, but no effect on the onset latency.
j
In antidromic sensory nerve action potential (SNAP) record-
10 mV
ings, a digital intereiectrode separation of 4.0 cm is believed to
be optimal. This particular distance will maximize a sensory
53
action potential's amplitude that has a negative peak rise time of

5 ms approximately 0.8 ms for a maximal fiber conduction velocity
of 50 m/s: 50 m/s = D/0.8 ms; D = 4.0 cm. A distance less than
4.0 cm may result in similar signals being recorded from both
electrodes, resulting in an amplitude reduction. This finding is
easily demonstrated by evoking an antidromic SNAP and se-
quentially increasing the intereiectrode separation in 1.0-cm in-
Amplitude crements (Fig. 3-5A). Changing the intereiectrode distance
18
Latency Peak from 1.0 to 4.0 cm does not alter the onset latency but increases
Trace Baseline
onset to the peak latency by 10% while the amplitude increases 5 3 % .
to peak The SNAP'S onset latency does not change because the active
peak
electrode did not change and, therefore, the distance between
ms mV the site of neural activation and the E-l electrode remained con-
A 3.5 11 21 stant. The SNAP'S amplitude increases because the reference
B 3.5 5 9 electrode records less of what the active electrode detects, i.e.,
the SNAPs rising negative phase. Since less similar information
Figure 3-4. Effect of intereiectrode separation on the ampli- is recorded, less of the same information is eliminated by the
tude of the compound muscle action potential (CMAP). A, differential amplifiers resulting in a larger potential. Similarly,
CMAP detected from the thenar muscles following median nerve acti-
the peak latency increases in latency since the SNAP'S peak is
vation: E-l on the muscle's m o t o r point and E-2 at a distant location
permitted more time to develop since less of it is eliminated as a
away from activated muscle tissue. B, CMAP observed when E-2 is lo-
common mode signal (see Amplifiers). Beyond 4.0 cm of elec-
cated on the thenar muscles near E-l. N o t e t h e reduction in ampli-
trode separation these waveform parameters change minimally.
tude. (From Dumitru D,Walsh NE: Electrophysiologic instrumentation.
Therefore, electrodes that are too close together shorten the
In Dumitru D (ed): Clinical Neurophysiology. Philadelphia, Hanley &
peak latency and reduce the amplitude, but do not affect the
Belfus, 1989, with permission.)
onset latency. A slower maximal fiber conduction velocity
allows one to place the electrodes closer together without com-
promising the amplitude. The same principle applies to ortho-
dromic recording techniques.
It can be appreciated from the above examples that proper
electrode placement is crucial to accurately record waveforms.
Errors usually result when the electrodes are too close together.
A reduction in amplitude may cause one to inappropriately
assume that some form of pathology has resulted in the loss of
excitable tissues when in fact this observation is an artifact re-
sulting from too short an intereiectrode separation. It is also
possible to record too short a peak latency when there may be
actual prolongation. This misreading of latency can result in an
artifactually normal result and thus true pathology may be
missed.
NEEDLE ELECTRODES
Latency Electrode Types/Impedance

Intereiectrode
Trace Amplitude Needle electrodes are usually chosen so as to approach the
distance Onset Peak bioelectric signal's generator source. Near-nerve needle elec-
cm ms »V trodes can be used to record from or stimulate single peripheral
nerves (see Chapter 2). Intramuscular electrodes are used to
A 1.0 2.7 3.0 56
record single-muscle-fiber waveforms or the summated electri-
B 2.0 2.7 3.1 72
cal activity of multiple muscle fibers forming the motor unit
C 3.0 2.7 3.3 77 action potential (MUAP). Subdermal electrodes also are avail-
D 4.0 2.7 3.3 86 able to record somatosensory, auditory, and visual evoked po-
E 5.0 2.7 3.3 86 tentials. The following are commonly used in clinical practice:
monopolar needles, standard concentric needles, single-fiber
Figure 3-5. The effect of varying intereiectrode separation needles, and subdermal electroencephalographic needles.
on an antidromic median nerve SNAP. Sequentially increasing The use of a needle electrode implies that it is placed into the
the e l e c t r o d e s e p a r a t i o n from 1.0 cm t o 5.0 cm p r o d u c e s t h e body's volume conductor and is in contact with the extracellular
above-detailed latency and amplitude changes. (From Dumitru D, fluid, an electrolytic solution. A chemical reaction between the
Walsh NE: Electrophysiologic instrumentation. In Dumitru D (ed): electrode's metal surface and the surrounding electrolytic solu-
Clinical Neurophysiology. Philadelphia, Hanley & Belfus, 1989, with tion can occur for needle electrodes just as it does for the sur-
permission.) face electrode/electrolyte cream interface. The needle electrode,
similar to the surface electrode, has the property of impedance detected in the uptake area. Compared to standard concentric
consisting of resistance, capacitive reactance, and inductance needle electrodes, monopolar needle MUAP recordings have
(usually negligible). As expected, the capacitive reactance is in- been reported to yield somewhat longer durations and larger
versely proportional to the signal's frequency content. amplitudes. Studies demonstrated up to 40% of normal
12
Monopolar Needle Electrode. The monopolar needle was MUAPs (depending upon the muscle) recorded with a monopo-
first used in electromyographic (EMG) analysis by Jasper and lar needle are polyphasic versus 3-20% for standard concentric
Ballem. A monopolar needle is made of solid stainless steel
31
needle electrodes. - - - - Additional investigations have found
6 7 10 29 55
12-75 mm in length and 0.3-0.5 mm in diameter. The entire that although monopolar needle recordings yielded significantly
needle is sheathed in Teflon except of the distal tip, which is higher amplitudes and more polyphasic MUAPs compared to
bare metal for 25-50 pm or more. - The actual recording sur-
10 11
concentric needle recordings, the MUAP's duration for both
face is cone-shaped with a sharp tip and an area approximately needles were not significantly different. - The disparate 30 43
0.17 m m . The monopolar electrode is less expensive and gen-

2
MUAP duration findings of various studies may stem from how
erally better tolerated by the patient but less durable than the the MUAPs were recorded with respect to the rise time, i.e.,
concentric needle electrode (see b e l o w ) . However, today's
5153
recording methodologies. Maximizing amplitude and minimiz-
manufacturing techniques permit disposable monopolar elec- ing rise time (< 500 ps) suggests that one is very close to a
trodes to have good recording characteristics for one-time use. select population of muscle fibers. Ignoring these parameters
Monopolar needle electrodes produce relatively little patient would yield potentially different results with respect to total
discomfort because the sharp pointed tip pierces the skin easily MUAP amplitude and possibly duration. At the present time,
and the shaft's Teflon coating glides readily through muscle commercially available monopolar needles appear to give
tissue. With repeated use of the electrode, the Teflon may peel MUAPs with larger amplitudes, more phases and turns, but sim-
back from the tip or crack and flake along the shaft of the ilar durations compared to MUAPs recorded with standard con-
needle; however, when disposable electrodes are used, this concentric needle electrodes. - - 30 34 43
cern is eliminated. If a monopolar needle electrode is used

53
The monopolar needle's impedance was found to be consis-
several times (following sterilization), the Teflon may peel off tently lower than that of concentric needle electrodes. At a 61
the needle's tip and the recording area subsequently increases. It single frequency of 10 cycles/second (Hertz-Hz), the dry
has been demonstrated that progressive Teflon removal from the monopolar needle's impedance was about 1.4 MO while that of
needle tip to twice the recording surface area of normal results the concentric needle was 4.7 MQ. Placing both electrodes in a
in significant MUAP amplitude reduction. The amplitude is
11
saline bath for 20 minutes significantly reduced both electrode's
believed to be reduced because the removal of Teflon increases impedances. Although previous studies demonstrated a higher
the metal recording area over which an action potential is dis- impedance for monopolar needles than concentric electrodes, - 6 29
tributed. The MUAP's duration does not change until the ex- more recent investigations have not confirmed this finding.
posed length of the needle tip reaches 10 times the original A monopolar needle requires a surface reference and ground
length. At 10 times the exposed needle length, the MUAP area electrode to obtain successful recordings. This may be consid-
32
begins to decline as does the number of turns and phases. ered a minor inconvenience compared to concentric needle elec-
The intramuscular sampling area of a monopolar needle is trodes. The reference electrode can be located near the needle's
spherically shaped and hence nondirectional. The uptake 51
insertion site or over a distant tendinous (electrically silent)
area's radius is approximately 1.5 times that of a standard con- region. It is possible for a reference electrode located close to
centric needle electrode (Fig. 3-6A). The spherical recording
52
the needle site to record voluntary activity and increase the
area can produce MUAPs with larger amplitudes than those recording's noise level. A too distant reference placement may
recorded with the standard concentric needle electrode, possibly also record more distant coactivated MUAPs or diminish com-
because electrical activity from more single muscle fibers is mon mode rejection. It is best to completely relax the limb
Figure 3-6. Pictorial depiction of the different needle recording electrodes available and their uptake areas. A, Monopolar needle
electrodes require a separate reference and ground electrode. B,A standard concentric needle electrode is depicted that uses the cannula as a
reference but does need a separate surface ground. C,Single-fiber electrode with E-l as a side port-pickup and the cannula that serves as the ref-
erence is shown. Again, a separate ground electrode is required. D, Bipolar concentric needle electrode is described with two recording ports
close t o each other.The cannula is the ground while the two central wires are the active and reference electrode. (Modified from Dumitru D,
Walsh NE: Electrophysiologic instrumentation. In Dumitru D (ed): Clinical Neurophysiology. Philadelphia, Hanley & Belfus, 1989, with permission.)
except for the muscle being examined and locate the reference Just as in monopolar recordings, the ground electrode should be
over a nearby bony prominence or tendon. If this fails to reduce placed in a convenient location. The cannula's entire surface is
distant activity, locating the reference or E-2 electrode immedi- capable of detecting electrical activity. Whatever portion of the
ately adjacent to the needle insertion site may reduce the muscle cannula is placed into the body has the potential of detecting
activity. The ground electrode is usually placed in a convenient electrical activity. As a result, the cannula's distal portion and
location. It may be of some assistance to use a second subcuta- active recording surface may record similar data. The informa-
neous monopolar needle instead of a surface disc electrode tion recorded by the cannula, however, is averaged over the
when interference becomes a significant problem. This is be- metal surface exposed within the volume conductor that tends
cause the two different electrodes (monopolar and surface disc) to reduce the signal's amplitude compared to that detected at the
are recording slightly different "noise" potentials which are active electrode's surface. In differential amplification, common
being amplified as opposed to rejected. data from both the cannula and active surface lead to common
The monopolar needle may be used occasionally as a cathode mode rejection that tends to reduce the MUAP's amplitude and
to perform near-nerve excitation. - - Caution has been rec-
544 58 60
background noise compared to monopolar recordings. The end
ommended in stimulating nerves with a needle electrode. The 53
result is a somewhat "quieter" baseline with potentially less
small tip of the needle can concentrate the stimulating current electrical noise and reduced MUAP amplitudes.
over a highly localized region, thereby increasing the current Inserting the cannula deeper into the examined muscle theo-
density to a level that may injure nervous tissue. It is suggested retically results in MUAPs slightly larger in amplitude and
that prior to stimulation the Teflon be removed 3-4 mm from longer in duration than those recorded more superficially. - The
6 8
the tip to reduce the current density. Stimulating with a deeper the cannula is in the volume conductor, the more metal-
monopolar needle is used apparently without detectable neuro- lic surface is available for the detected signal to be averaged
trauma by reducing the excitation pulse's duration to 0.05 over. - This is assumed to reduce the signal's amplitude
6 38
m s 43.60 \i i t necessary to remove the Teflon from a monopo-

s p o recorded by the cannula and fed into the amplifier, which in turn
lar needle used as a cathode provided the stimulus duration is means there is less in common with the active recording elec-
maintained at 0.05 ms or less. trode. As fewer similar signals are subtracted through differen-
32
Commercial needle manufacturers now provide disposable tial amplification, the observed MUAP increases in amplitude.
monopolar needle electrodes at reasonable cost. Disposable This effect does not occur with monopolar needles.
needle electrodes (monopolar and standard concentric) have A study comparing various electrical properties (impedance,
become more popular in light of the prevalence of infectious noise, interference sensitivity, and signal distortion) among five
blood-borne diseases. Although needle quality varies among
33
brands of commercially available concentric needle electrodes
the different companies, those from more reputable firms revealed considerable variability. It was concluded that ob-
16
appear to provide stable noise-free recordings comparable to served differences stemmed from diversity in manufacture, ma-
those obtained from reusable electrodes. If disposable needle terials, construction, and design. Electrolytic treatment appears
electrodes are used, there is no need to worry about Teflon peel- to improve the recording characteristics of all needle brands and
ing with repeated use or infectious disease problems. diminishes the inter-brand electrical variance. The various nee-
Standard Concentric Needle Electrode. The standard dles' recording characteristics may affect the MUAP's parame-
concentric needle electrode was introduced by Adrian and ters and affect comparison of reference data from one laboratory
Bronk. This electrode is essentially a hollow stainless steel hy-
1
to another. Commercially available disposable concentric
podermic needle (cannula) with a centrally located platinum or needle electrodes adequately compete with nondisposable elec-
nichrome-silver wire 0.1 mm in diameter insulated from the trodes with respect to electrical stability and signal fidelity, thus
cannula by a nonconducting epoxy resin. The cannula is bare permitting their use from both a cost-effective and communica-
metal with a diameter of 0.3-1.0 mm and a length similar to that ble disease aspect.
of the monopolar needle. - The tip of the concentric needle
12 41
Monopolar vs. Standard Concentric Needle Electrodes.
electrode has a beveled cutting edge with a 15-20° angle. The41
One may anticipate that the ideal needle recording electrode to
150-pm diameter central wire's recording surface is oval-shaped assess MUAPs does not exist. The monopolar and standard con-
with an exposed surface of 150 pm by 580 pm (0.08 m m ) .2 12
centric needle electrodes each possess inherent advantages and
The three-dimensional recording region of the concentric disadvantages. To some extent, they both demonstrate variation
needle is a hemisphere with a radius range of 1.0-2.5 mm (Fig. in recording characteristics among different manufacturers. The
3-6B). The hemispheric uptake area suggests that the concen-
41
directional and depth-dependent properties of the concentric
tric needle has directional recording properties. The cannula's needle should be kept in mind when quantifying MUAP proper-
beveled tip acts as a shield preventing the single muscle fibers' ties. Standard concentric needle recordings may be somewhat
68
voltages opposite the recording surface from contributing to the more free of electrical interference and background noise than
detected MUAP waveform. This "shield" aspect of the concen- monopolar needles, but possess a cutting edge that may damage
tric electrode applies only to the fibers contributing to the tissue. The standard concentric needle does not require an addi-
MUAP's negative spike, i.e., within 300 pm of the electrode's tional electrode (reference) that needs to be periodically relo-
core. The directional recording characteristics of the standard cated on the patient, but it is more painful than the monopolar
concentric needle electrode are, therefore, quite different from needle. Although the concentric needle may have a slightly
51
that of the monopolar needle as they relate to those few fibers more uniform recording surface, the sensitive diagnostic para-
contributing to the MUAP's negative spike. The recording of meter of MUAP duration appears essentially the same for both
fewer total muscle fibers within the roughly 300-pm hemisphere types of electrodes. In the final analysis, personal preference
30
uptake area may result in smaller MUAP amplitudes compared and training biases seems to be the dominant factors in deter-
to monopolar recordings. mining the type of electrode used in clinical practice. The com-
In concentric needle observations, the cannula serves as the petent practitioner, however, should be thoroughly familiar with
reference electrode but a separate ground electrode is required. using both types of electrodes. A sound practice is to develop
reference data for both types of electrodes in one's practice set- electrode should possess an input impedance of 10 MQ. or
ting and not mix and match MUAP parameters between labora- more. 49
tories and needle types. It is the preference of one of the authors Bipolar Concentric Needle Electrode. A bipolar concen-
(DD) to use monopolar needles when performing routine needle tric needle electrode is similar to a standard concentric needle
electromyographic examinations of multiple muscles, but to use electrode except that a second nichrome-silver wire is located
a concentric electrode when performing quantitative analysis of within the cannula and serves as the electrode's reference (Fig.
individual MUAPs from a limited number of muscles. 3-6D). This reference (E-2) wire is separated from the active
6
Physical modeling of monopolar and concentric needle elec- (E-l) wire and cannula by a nonconducting epoxy resin similar
trodes have verified some but refuted other assumptions made to that of a standard concentric needle. Although the bipolar
in computer models of these two electrodes. ~ - - The physi-
21 24 35 42
concentric needle is rarely used, some of its properties are worth
cal modeling of the uptake area for these two electrodes verifies discussing. Unlike the standard concentric needle, the bipolar
in that the concentric electrode has essentially a hemispherical concentric needle does not require a separate ground electrode
recording area regarding the fibers comprising the main spike because the cannula serves this function. The active ( E - l ) and
for the recorded MUAP. Similarly, the monopolar needle reveals reference (E-2) electrodes are located less than a millimeter
a somewhat spherical recording uptake area for the fibers con- apart, which enables them to record similar data from a very lo-
tributing to the MUAP's negative spike. However, with respect calized or selective region within the volume conductor. As one
to the MUAP's duration, both electrodes' active recording sur- would imagine, there is very little electrical noise in these
faces record from a spherical uptake area. The concept of the recordings. Motor unit action potentials observed with a bipolar
cannula shielding the concentric electrode's core from the fibers concentric needle consist of electrical activity from relatively
"behind" it appears not to be true. This is demonstrated by the few fibers and appear very different than those from monopolar
fact that rotating a concentric needle electrode while recording or standard concentric needle recordings. -
6
The duration and
38,46
from the same MUAP does not lead to any alteration in the amplitude of a bipolarly recorded potential are markedly re-
MUAP's duration. Another interesting aspect of these two elec- duced. Additionally, the bipolar concentric electrode is consid-
trode types is the fact that most studies have revealed that the erably larger than a monopolar or concentric needle and can be
monopolar needle electrode on average records MUAPs with relatively more painful.
larger peak-to-peak negative spike amplitudes. The physical Subdermal Electroencephalographic Electrodes. Sub-
models show that for a single fiber the same distance from dermal electroencephalographic needles are 10-20 mm long
either needle's active recording surface, the concentric needle and 0.8 mm in diameter and made of stainless steel or plat-
records a larger amplitude. However, the rate at which the am- inum/iridium primarily used to record evoked potentials. 48
plitude for this single fiber declines with distance is much These electrodes are usually placed just under the scalp and
greater for the concentric electrode. Therefore, large MUAP held in place with tape. The impedance of these needles ranges
amplitudes recorded by monopolar needle electrodes arise be- between 3,000 and 10,000 Q but yield evoked potentials that are
cause of the larger recording territory and hence more fibers not significantly different than those obtained by surface elec-
within it contributing to the MUAP spike, which more than trodes with impedances less than 5,000 Q . - 4
Comparable
2163
makes up for the relatively larger amplitude per fiber recorded evoked potentials can be obtained with needle electrodes of
by the concentric electrode. Finally, actual measurements of higher impedance than surface electrodes because the needle is
MUAPs recorded with concentric needle electrodes in superfi- placed within the volume conductor and bypasses the skin. The
cial compared to deeper muscle regions failed to support the signal in effect is detected without difficulty by the needle elec-
concept that the more deeply recorded MUAPs should have trodes whereas a surface electrode must have the skin's surface
longer durations and larger amplitudes. The clinical studies
26
barrier considerably reduced to detect the same signal. The in-
showed no difference between these two sets of MUAPs with herent impedance of the needle electrodes' metal (stainless steel
respect to duration or amplitude parameters. This is likely a and platinum/iridium) is larger than that of the surface elec-
result of modern instruments using averagers capable of opti- trodes (gold and silver), and the needle has a considerably
mizing the signal-to-noise ratio sufficiently to detect slight dif- smaller surface area than the surface discs. Surface area is a
ferences between the core and cannula signals irrespective of major determining factor when considering impedance and the
tissue depth. smaller the area of exposed recording surface, the larger the im- i
Single-Fiber Electrode. The single-fiber electrode is basi- pedance as it is more difficult for current to flow. These two fac-
cally a modified standard concentric needle electrode (Fig. 3- tors, surface area and metal type, most likely account for the
6C). The central wire exits approximately 4 mm proximal and
49
different impedances between needle and surface recording
opposite the beveled tip through a side-port in the cannula in- electrodes. However, as long as the amplifier's impedance is
stead of extending along the length of the cannula to the beveled substantially larger than that of the needle electrode, an undis-
tip. While the active recording surface approximates 25 pm in
62
torted signal is obtained. The needle electrode does not require
diameter, the cannula's diameter is about 0.5 mm. A separate the skin to be abraded or an elaborate means of securing it to the
ground electrode is required while the cannula serves as the E-2 skull; application is thus relatively easy. The asserted threat of
electrode. The diameter of the E-1 recording surface approaches increased infection rates with needle electrodes has not been
that of a single muscle fiber and improves the selectivity of this documented and remains only speculation. It should be real-
21
electrode to record primarily from one muscle fiber. The uptake ized that with a larger number of recording channels—for ex-
region of E-l is a hemisphere with a diameter approximating ample in sensory evoked potentials—the use of multiple needle
300 pm. The combination of a small active recording surface
50
insertions may become cumbersome. Of note, both electrodes
and an elevated low-frequency filter of 500-1000 Hz enables should be properly sterilized if reused because the surface elec-
the single-fiber electrode to record from very small areas. A trode requires skin abrasion exposing it to the patient's serum,
small recording surface gives the single-fiber electrode a large and in this respect is just as contaminated as the needle elec-
impedance, which requires that any amplifier connected to this trode piercing the patient's skin. Any device or substance used
for abrading the skin should be discarded. Fortunately, manu-

facturers now offer disposable subdermal electroencephalo-
graphic electrodes.
AMPLIFIERS
AMPLIFICATION
As previously stated, the magnitude of biologic signals can

range from microvolts to millivolts. Following signal detection
Latency
by the electrodes, the biologic signals must be significantly am- Trace Sensitivity Amplitude
plified prior to analysis. One can express amplification in terms Onset Peak
of gain or sensitivity. Gain may be simply stated as the ratio of
the amplifier's output to its input signal. For example, suppose
53 iiV/div ms mV
an input signal of 10 pV emerges from the amplifier's output as A 20 2.8 3.4 78
1 V. The gain in this case is 100,000 (output/input = 1 B 50 2.8 3.4 78
V/0.00001 V = 100,000). It is important to note that there are no C 100 2.8 3.4 78
units to the gain value because it is a ratio. Sensitivity, on the
other hand, is the ratio of the input voltage to the size of a CRT Figure 3-8. Amplifier sensitivity effect on SNAP waveform.
deflection. The CRT's deflection is typically measured in cen- Similar t o Figure 7 except that the amplifier sensitivity is altered for a
timeters (cm). An amplifier with a sensitivity setting of 10 SNAP. N o difference is noted for the various amplifier settings. (From
pV/cm (10 pV/division) means that an input signal of 10 pV Dumitru D,Walsh NE: Electrophysiologic instrumentation. In Dumitru
will result in a 1 cm CRT deflection, or each division on the D (ed): Clinical Neurophysiology. Philadelphia, Hanley & Belfus, 1989,
CRT will represent a display magnitude of 10 pV (10 pV/div). with permission.)
Observing the gain or sensitivity at which a particular re-
sponse is elicited is not a trivial matter. In electrodiagnostic med-
icine one important parameter is the onset latency of the If we begin at an amplifier sensitivity of 500 pV/div, the
compound muscle action potential (CMAP). It is easy to demon- CMAP's onset latency is measured at 3.4 ms. Sequentially de-
strate the variability of the CMAP onset latency when the same creasing the amplifier's sensitivity results in a progressive in-
potential is observed at multiple sensitivity settings (Fig. 3-7). 18
crease in the waveform's onset latency to 3.8 ms at 10,000
pV/div. This example demonstrates the necessity of recording
the amplifier's sensitivity in reporting data from one investiga-
tion to another, and also emphasizes the importance of reproduc-
ing another laboratory's instrument settings when using their
reference data. Significant shifts in onset or peak latencies are
not observed for sensory nerve action potentials, most likely be-
cause these waveforms are rather small to begin with and the in-
cremental change in sensitivity is comparably smaller than
described above for CMAPs (Fig. 3-8).
INPUT IMPEDANCE
Amplifiers, in addition to electrodes, also possess an imped-

ance. Similar to electrodes, the amplifier's resistance and capaci-
tance are considered important parameters, but the inductance is
of negligible magnitude at biologic frequencies of interest. The
electrode and amplifier in combination with the biologic signal
form a relatively simple electronic circuit that obeys Ohm's law:
Trace Sensitivity Latency onset Amplitude
E = I x R. Recall that biologic signals vary over time. Circuits
measuring biologic signals approach AC circuits in which the
\xV/div ms mV
term resistance (R) is replaced by impedance (Z) to take into ac-
A 500 3.4
B 3.5 count both factors of resistance and capacitive reactance so that
1,000
C 3.7 Ohm's law becomes E = I x Z. The action potentials in the body
5,000 23 serve as the battery or voltage source ( E ) , while the imped-
D 10,000 3.8 23
tota!
ance for the recording electrode (Z ) and amplifier (Z^p) can

elec
be simplified for discussion purposes as two separate compo-

Figure 3-7. Amplifier sensitivity effect on CMAP onset la- nents in series (Fig. 3-9). In this simple series circuit, the current
tency. Decreasing t h e amplifier's sensitivity (A—D) d e m o n s t r a t e s a from the biologic source is the same across both Z and Z ^ ;
e l e c
prolongation in t h e CMAP's o n s e t latency (table). (Modified from however, there is a different voltage drop across each of them
Dumitru D, Walsh NE: Electrophysiologic instrumentation. In Dumitru summating to the signal generated in the body, i.e., E . As tolal
5 3
D (ed): Clinical Neurophysiology. Philadelphia, Hanley & Belfus, 1989, stated in the introduction, the total voltage drop in the whole cir-
with permission.) cuit is divided among the individual voltage drops across each of
Table 3-1 . Amplifier Input Im pedance

Etot Zelec Zamp âmp
100 mV 20 MO 10 MO 33.3 mV
100 mV 10 MO 10 MO 50.0 mV
100 mV 5 MO 10 MO 66.7 mV
100 mV 1 MO 10 MO 99.9 mV
amp
Using the voltage divider equation (E = E , x Z^p/Z^ + Z ) one can see
amp tota amp
that most of the biologic signal will be detected by the amplifier when the im-
pedance of the electrode is significantly less than that of the amplifier.
by the manufacturer, dried electrode paste, etc.), too much of

Figure 3-9. An equivalent circuit of one-half of the elec- the voltage drops across the electrode with little left to drop
trode-amplifier system. The impedances of t h e e l e c t r o d e ( Z | ) e ec across the amplifier. The net result is a small and distorted
and amplifier (Za ) form a voltage divider.The voltage drop across t h e
mp signal. This could lead us to believe that the patient has an ab-
amplifier is displayed on the c a t h o d e ray t u b e . (From Stolov W C : normal response secondary to either axonal loss (neuropathy) or
Instrumentation and Measurement in Electrodiagnosis. Minimonograph muscle tissue loss (myopathy).
No. 16, Rochester, MN, American Association of Electrodiagnostic Note that the above discussion concerns only the E-l elec-
Medicine, 1981,with permission.) trode, but the same concepts apply to the E-2 electrode and in-
verting (see below) amplifier to which it is connected.
the circuit's subcomponents, forming a voltage divider. 40
Additionally, the impedance of the amplifier must be signifi-
Applying Ohm's law to this explanation allows one to conceptu- cantly larger than that of the electrode (or electrode plus skin for
alize the interaction between the electrode and amplifier with re- surface recordings) to faithfully reproduce the desired signal on
spect to the magnitude of the recorded signal. We can begin with the CRT screen. The input impedance of most commercially
the previous statement that the biologic signal's voltage is the available amplifiers is in the range of millions of ohms (MO),
sum of the voltage drops across each circuit subcomponent, i.e., whereas the electrode's impedance approximates thousands of
the electrode and amplifier: ohms (kO).
^total =
E lec "F ^
e a m p
DIFFERENTIAL AMPLIFICATION
The current across each circuit's subcomponent is the same and
according to Ohm's law: The amplifiers used in commercially available electrophysio-
logic instruments are differential amplifiers Differential am-
4 7
plifiers are two amplifiers as identical as possible with respect to

P —
âmp - 1 *y^ 7âmp- their electronic amplification characteristics except that one in-
The equivalent terms for E i and E e ec a m p can be substituted into verts the recorded signal compared to the other (Fig. 3-10). The
the summated voltage equation: noninverting (positive) amplifier magnifies the signal received
from the active or E-l electrode without a polarity effect ( E a m p
Etolal I Zelec + I Z
= X X
a m p , OT from the previous discussion). The inverting (negative) amplifier
E-total 1 (Zelec + Z
= X
a m p ). reverses the polarity of the signal's subcomponents through the
The equation for the current across the amplifier (I = E a m p /Z a m p ) reference or E-2 electrode and then magnifies the potential
can be substituted in the above equation: (~E ). These two amplified signals are then electronically sum-
amp
mated, i.e., the signals presenting to the electrodes are effectively

P — âtnp X
(Z lec ~F Z
e a m p ) subtracted ( E + { - E } ) . The net output of this summation
Motal ~* 7 a m p amp
circuit is an amplification of the differences detected from each

We can now solve this equation for that portion of the signal's electrode and the elimination of like signals. The summed ampli-
voltage detected at the amplifier ( E ) because this is what we a m p
fied output of the amplifiers is called the difference-mode
observe on the CRT screen: signal, while the canceled potentials are called the common-
mode signal. A ground electrode is also connected to each of
34
F — ^tolal X
Z a m p the amplifiers that defines the zero potential, i.e., the ground de-
E
amp 7 , 7
-•amp fines the zero potential for the instrument and serves as the refer-
One can clearly see from the above equation that the ob- ence zero potential value for all recorded potentials.
served signal on the CRT screen approaches E | if Z is sig- tota a m p From a practical standpoint, it is impossible for manufactur-
nificantly greater than Z . The impedance effect of the
elec ers to build two identical amplifiers. As a result, there are minor
electrode is diminished when there is a comparatively large am- differences between the two signals as they emerge from differ-
plifier input impedance. If the amplifier's input impedance is ential amplification and are about to be summated. Additionally,
significantly greater than that of the electrode, the above equa- no two E-l and E-2 electrodes are identical with respect to their
tion simplifies to E =E because Z / ( Z
a m p + Z ) ap-
total a m p a m p elec impedances and other recording properties. The end result of
proaches unity ( Z / Z = 1) when Z
a m p is substantially
a m p e l e c these minor differences is an elimination of most but not all of
smaller than Z (Table 3-1). We can now see how the imped-
a m p the common signals presenting to both electrodes. The above
ance of the electrode is an important factor in recording wave- observations explain why it is important to use two similar
forms. If the recording electrode has a very high impedance for recording electrodes if at all possible. Different electrode types
any reason (improperly sterilized, insufficient coating removed can have very dissimilar impedances, producing an impedance
(+) \ ( + ) 200,000/xV ( + )
\^(+)200,000/LtV
10,000 10,000
20u.V 200,000^V
(-)I99,980/IV
„ . Vout 200,000uV ,^
Gain = Vout Gain = • ^ = 10,000
Vin 20/iV ~ Vin 20^ V
10,000
CMRR =
Figure 3-/0. The results of differential amplification are depicted. A, A signal is connected t o both amplifiers in a common mode (same
signal t o each amplifier) manner.The gain is o n e . B, In t h e difference m o d e montage t h e gain in 10,000.The c o m m o n m o d e rejection ration
(CMRR) is 10,000 t o I.(From Dumitru D,Walsh NE:Electrophysiologic instrumentation.In Dumitru D (ed):Clinical Neurophysiology.Philadelphia,
Hanley & Belfus, 1989, with permission.)
mismatch that results in slightly different signals being pre- HIGH- A N D LOW-FREQUENCY FILTERS
sented to the amplifier. These unwanted differences are magni-
fied and not canceled. If one wishes to eliminate our BIOLOGIC SIGNALS
environmentally prevalent 60-Hz interference presenting to both
electrodes, the more alike the electrodes are, the greater the Biologic signals observed in electrodiagnostic medicine can
chance of more perfectly subtracting 60 Hz as a common mode be thought of as the sum of time varying sinusoidal waveforms
signal. Two electrodes constructed of different materials may with different frequencies whereby their individual phases and
amplify undesired environmental noise because of an imped- amplitudes summate or cancel to reproduce the original wave-
ance mismatch. form. An analysis of these potentials can be considered if one
47
The effectiveness of an instrument's differential amplifica- conceptualizes them to consist of an infinite series of sine waves
tion can be quantified by its common-mode rejection ratio with different amplitudes and frequencies. This complex mathe-
(CMRR). The CMRR is the summated output of the amplifiers matical process of harmonic or frequency spectrum analysis
when a signal is amplified differentially compared to when the can be simplified for discussion purposes. Let us consider a
47
same signal is presented in common mode. For example, if a 20 series of five sine waves with particular frequencies, arbitrary
pV signal is presented to both the E-l and E-2 ports (common amplitudes scaled relative to each other, and all in phase with
mode) of an amplifier with respective amplifications of 10,000 respect to each other (Fig. 3-11 A). The individual waveforms in
and 9,999, the amplified summated output signal is 20 pV (Fig. our example have specific frequency characteristics. We define
3-10). In this instance the common mode gain is 1 (20 pV/20 a frequency as the number of times the same event or cycle
; pV = 1). Let us now apply this same 20 pV potential differen- occurs in one second, i.e., cycles/second. As noted above, one
; tially, i.e., 20 pV is fed into the noninverting amplifier port and cycle per second is called a hertz and abbreviated Hz. In elec- 15
0 pV is presented to the inverting amplifier port. The same am- trophysiology, events take place over relatively short time spans
plifiers now yield a summated potential with a magnitude of and can have frequencies of thousands (kilo or simply "k") of
200,000 pV resulting in a differential to common mode gain times per second (kilohertz or kHz). Summating all of the
ratio of 10,000 (200,000 pV/20 pV = 10,000). This ratio is the given waveforms in our example on a point-for-point basis re-
CMRR, or in this instance 10,000 to 1. Another way of thinking sults in a net waveform approximating a square wave. It is im-
about the CMRR is that any difference detected between elec- portant to appreciate how choosing the appropriate number of
trodes is amplified 10,000 times more than when the same sine waves with individual frequencies generates a summated
signal is detected at both electrodes. The majority of commer- waveform whose appearance is distinctly different than each
cially available instruments have CMRRs of 10,000:1 or subcomponent waveform (Fig. 3-11 A).
greater. It is common to express the CMRR as a decibel (dB) From the above discussion it can be seen that our square wave
power function, where a dB = 20 log CMRR. In the above ex- actually consists of a series of subcomponent waveforms of both
ample, 20 log 10,000 = 80 dB (201ogl0,000 = 201ogl0 = 4
relatively high and low frequencies. Adding more appropriate
4(20)logl0 = 80 db). subcomponent higher frequency waveforms results in a final
time spans a time period of approximately 2.0 ms, representing

an occurrence frequency of 500 Hz (1/2.0 ms x 1000 ms/second
= 500/second = 500 Hz). The total negative phase has a duration
of 5.0 ms or an occurrence frequency of 200 Hz. In this exam-
ple, the potential's rise time occurs over a shorter time period
and, therefore, represents a higher frequency (500 Hz) than that
of the entire negative phase (200 Hz). Both portions of this
waveform contain essentially equal numbers of subcomponent
waveforms (see above). The difference is that the lower fre-
quencies' amplitudes do not summate optimally over the region
of the waveform represented by the rise time. The higher sub-
(9.0.O.II) 15.0,0.2)
component frequencies preferentially influence this aspect of
the potential. When considering the total duration of the poten-
Figure 3-11. Subcomponent waveform frequencies. A, A tial's negative phase, the same subcomponent waveforms are
square wave and five subcomponent sine waves that result in the present as for the rise time segment. In this instance, the ampli-
square wave when added together. B,A m o r e "biologic"-appearing po- tude of relatively lower frequencies summate more than the
tential and its subcomponent sine waves.The frequency, phase shift, and higher frequencies, thereby influencing this portion of the
relative amplitude are described below each subcomponent waveform. waveform comparatively more.
In the above example, one can appreciate that rapidly chang-
ing aspects of the biologic signal under investigation (baseline
waveform with less baseline wavering, sharper inflection points, takeoffs, inflection points, rise times, and summits of peaks) are
and a flatter maximum region, in short a true square wave. In this dominated more by the higher frequencies. The subcomponents
example, we can demonstrate that any potential may be repre- of the potential that change over relatively longer time periods
sented by a series of sine waves with appropriate frequency, am- (baseline returns and total potential durations) are influenced
plitude, and phase characteristics. By summating a different more by the relatively lower frequencies. There are also a
family of sine waves, a more "biologic"-appearing potential can number of low-frequency potentials (electrode/electrolyte depo-
be simulated (Fig. 3-1 IB). This same process can be applied to larizations and movement artifact) as well as high-frequency
any waveform. The important concept to remember is that each potentials (radio waves, internal amplifier noise and needle in-
specific point of a biologic waveform may be conceptualized as sertional activity) that are not part of the desired data and should
being the end product of a summated series of sine waves. Each be prevented from contaminating the investigated signal. As a
point of the total recorded waveform contains the same number result, a device with the capability of excluding undesired sig-
of high- and low-frequency subcomponents as every other part nals outside the frequencies of interest (noise) but including
of the potential. The combination of individual subcomponent designated important information is necessary to optimally
waveforms' phase cancellations, additions, and amplitudes, how- assess biologic information. Fortunately, a filter is just such a
ever, determines the net effect or predominant frequency ob- mechanism.
served for any particular point on the observed waveform. An
additional example may help to clarify this concept. FILTERS
We may evoke a compound muscle action potential from the
thenar eminence and investigate only its negative phase for dis- A filter may be defined as a device composed of a variable
cussion purposes (Fig. 3-12). Let us consider two subsegments resistor and capacitor appropriately connected such that they
of this waveform, i.e., (1) the rise time (baseline takeoff to neg- have the capability of excluding certain frequencies from being
ative peak; segment A-B in figure 3-12), and (2) the negative recorded. Frequencies prohibited from being observed depend
5
phase's total duration (A-C in figure 3-12). The potential's rise upon the particular characteristics of the filter used. Two broad
categories of filters are usually employed in electrophysiologic
devices. A high-frequency filter, also called a low-pass filter,
B
limits the higher frequencies from being recorded but allows the
lower frequencies to pass unaffected. The low-frequency
15
filter (high-pass filter) prohibits low frequencies from being

observed, but permits the high frequencies to pass. To ensure
complete passage of a designated frequency, it is good practice
to exceed the desired frequency by at least twice the value or
more. Similar reasoning applies to low frequencies as well. For
example, if a signal contains a subcomponent frequency of
1,000 Hz and requires elimination, it is best to set the high-fre-
quency filter's upper limit to at least 2,000 Hz or higher, be-
5000/A/ cause a typical analog filter set at a particular frequency "filters"
about 70% of the designated frequency's magnitude; therefore,
a somewhat higher cut-off limit is required to eliminate the ma-
Ims
jority of signals containing frequencies with the identified limit.
Figure 3-/2. Compound muscle action potential evoked The concepts of high- and low-pass filters can be illustrated if
from the thenar eminence. Segment A - B represents the poten- one first considers a simple resistor/capacitor (RC) circuit con-
tial's rise time while segment A - C corresponds to the duration of the nected to a variable voltage representing a mixed frequency
potential's negative phase. waveform (Fig. 3-13A). If the resistor is placed first in the circuit
followed by the capacitor, we can measure the voltage drop of the capacitor and resistor, we can allow most of the low fre-
across the capacitor by connecting an amplifier across the two quencies to drop across the capacitor by increasing its imped-
terminals of the capacitor. The voltages contained within the ance greater than that of the resistor described above so that the
mixed-frequency signal representing the waveform must be di- only signals measured by the amplifier are low frequencies. In
vided between two voltage drops: one across the resistor and the this way, the low frequencies can be amplified and detected by
other across the capacitor. Remember that a voltage divider is the instrument. In short, the high frequencies were not allowed
formed by these two "in series" electronic components. The to pass onto the amplifier, but the low frequencies dropped their
higher frequencies in the signal are not impeded by the capaci- voltage across the capacitor thereby being amplified and subse-
tor because the equation Z = l/(27ifC) tells us that as "f' in-
c quently observed, i.e., a low-pass filter. The amplifier measures
creases, the capacitor's impedance diminishes. Therefore, the voltage across the capacitor. If all of the voltage dropped across
capacitor does not impede the higher frequencies contained in the capacitor, then it is high on one side of the capacitor and es-
j the waveform and there is no voltage drop across the capacitor sentially zero on the other side. The amplifier subtracts the zero
for the amplifier to measure. All of the voltage for the high fre- voltage from the high voltage for the passing low frequencies
quencies drops across the resistor and never gets to the ampli- and the voltage amplified is effectively the same as that in the
fier. Another way of thinking of this concept is to realize that body.
the capacitor offers minimal resistance to current flow for the If the order of the resistor and capacitor is reversed, the am-
high-frequency signal. Since the total voltage is divided be- plifier records any voltage drop occurring across the resistor
tween the resistor and capacitor, a minimal resistance to the ca- (Fig. 3-13B ). The voltage divider equation must be rewritten to
pacitor means that most, if not all, of the voltage drop in the reflect our measuring amplifier now placed across the:
circuit must occur across the resistor. The amplifier, however, is
not placed across the resistor and therefore this voltage drop is P _ ^total X
not detected resulting in a "filtering" of these signals, preclud- " Z + Z R c
ing them from being observed on the CRT. This can be ex- For a mixed signal, the subcomponent high frequencies will
pressed mathematically by using the voltage divider equation pass unimpeded across the capacitor (impedance drops to zero
discussed previously where the waveform's voltage detected at by l/(27tfC) to drop their voltages across the resistor to be mea-
, the amplifier ( E ) is expressed in terms of the waveform's
a m p sured by the amplifier. This is because E = E i as Z / ( Z +
a m p tota R R
! voltage content for a particular frequency (E ) and the imped- total 0) = 1. The waveform's subcomponent low frequencies all drop
1
ances of the resistor (Z ) and capacitor ( Z ) :
R c across the capacitor, but since the amplifier is not connected
across it, we are not able to detect this voltage drop. For the low
p _ ^total X
Zc frequencies, the impedance of the capacitor ( Z ) is significantly
c
t a m
p - z + z
R c
greater than Z causing E
R to approach 0, because E
a m p x Z totaI R
For a high frequency, the capacitor's impedance approaches divided by the large quantity Z equates to a very small number.
c
zero (plug zero into the above equation in place of Z in the nu- c
The end result is a high-pass or low-frequency filter.
merator and denominator) and the voltage observed by the am- A completely different approach regarding the filtering of bi-
plifier also approaches zero, thereby eliminating the high ologic signals is the application of digital filtering, which is in-
frequencies, i.e., E = 0 or close to it.
a m p creasingly used in modern apparatuses. Digital filtering can be
Low frequencies may be allowed to pass onto the amplifier if done in both the time and frequency domain. A simple example
the resistor's impedance is substantially lower than that of the is to calculate the average values of the previous and next signal
capacitor. In this case, the voltage for the low frequencies in the value resulting in a smoothing of the signal. The digital repre-
mixed signal pass onto the amplifier without change because sentation of the signal can be transformed to the frequency
the above equation reduces to E = E i when Z » Z be- a m p tola c R domain enabling the selective removal of chosen frequencies.
cause Z / ( Z + Z ) = Z / Z = 1. The high frequencies experi-
C R c c c After this, the signal is shown again as a function of time.
ence a voltage drop across the resistor, but since the amplifier is Digital filtering results in complete and sharp removal of the
not connected across the resistor, they are not observed. The low chosen frequency bands, in contrast with the hardware filtering
frequencies, however, could pass across the resistor but not the just described. Analog filters are subject to drift and variations
capacitor, because the equation l/(27ifC) states that for low fre- in frequency response due to temperature changes. Digital fil-
quencies an impedance is present. Depending on the characteristics ters are very stable and do not change with time.
Figure 3 - / 3 . Two resistor/capacitor ( R C ) circuits are depicted representing analog filters. A, A high-frequency (low-pass) filter is
formed and the high frequencies drop across the resistor while the low frequencies drop across the capacitor to pass o n t o the amplifier. B.The
low frequencies drop across the capacitor (low-frequency filter) while the high frequencies drop across the resistor t o be passed onto the ampli-
fier (high-pass filter).
82 — P A R T 1 FUNDAMENTAL PRINCIPLES
or parameters of the response. A further 10-fold increase in the

low-frequency filter to 100 Hz demonstrates several changes
from the previously obtained waveform. Qualitatively, the re-
sponse seems smaller and appears to be developing a third or
negative phase. Quantitatively, the amplitude has decreased
16.9% while the peak latency has shortened 6.0%, but the onset
latency has not changed. A 28.5% reduction in negative spike
duration has also occurred. An additional 3-fold increase in the
low-frequency filter (300 Hz) produces a 53.8% reduction in
amplitude and a 9.0% shortening of the peak latency compared
to a waveform recorded with a 10 Hz low-frequency filter. The
onset latency remains unchanged, but the negative spike dura-
Frequency tion has decreased 42.8% from the first response. Note that the
morphology of the potential is now clearly triphasic: negative-
Figure 3 - / 4 . Frequency bandwidth. Simple diagram representing positive-negative.
a bandwidth "viewed" by t h e i n s t r u m e n t o n c e the filters have r e - We may return to the concept of segmental subcomponent
stricted t h e frequencies passing o n t o t h e amplifier. (From Dumitru D, frequencies to explain why the above SNAP underwent the ob-
Walsh NE: Practical instrumentation and common sources of error. served changes when elevating the low-frequency filter. The
Am J Phys Med Rehabil 1988,67:55-65, with permission.) waveform's initial departure from the baseline occurs over a rel-
atively short time period. This portion of the waveform has min-
imal amplitude contribution from the low frequencies contained
Most commercially available instruments consist of variable in the potential, i.e., the higher frequencies predominantly influ-
low- and high-frequency filter systems that can be adjusted by ence this aspect of the waveform. Because elevating the low-
the clinician to optimize the frequency content of the signal frequency filter does not affect high frequencies, it should now
under investigation and limit undesired noise. This low- and be clear why the onset latency (high frequency) of the potential
high-frequency filter combination constitutes a "window" to ob- did not change. In short, the waveform's onset goes from a flat
serve a relatively limited frequency domain referred to as a baseline to an abrupt negative deflection over a relatively short
bandwidth (Fig. 3-14). The frequencies above and below the
bandwidth's limitations are severely attenuated by the low- and
high-frequency filters, respectively. It is important to note at this
point that low and high are relative terms that depend on the | 2 0 MV
content of the waveform under investigation. Because filters can 2 ms
potentially eliminate important waveform subcomponent fre-
quencies if set improperly, the waveform's morphology can be
significantly altered. The clinician, therefore, must be thor-
oughly familiar with the effects high- and low-frequency filters
have on normally recorded signals.
LOW-FREQUENCY (HIGH-PASS) FILTER

The majority of instruments have adjustable low-frequency
filters in the range of 0.1-500 Hz. The easiest way to appreciate
how a low-frequency filter affects a biologic waveform is to
Latency Duration
record sequential sensory and motor waveforms while increasing Low
Trace negative Amplitude
the low-frequency filter (e.g., 1, 10, 100, and 300 Hz), but keep- frequency Onset Peak spike
ing the high-frequency filter constant at 10,000 Hz (Figs. 3-15
and 3-16). The waveform parameters investigated are: onset
18
Hz ms ms
and peak latencies, negative spike amplitude, negative spike du-
A 1 2.6 3.3 1.4 65
ration, and number of phases. It is important to keep in mind that
B 10 2.6 3.3 1.4 65
the low frequencies contained in the investigated waveform
below the low-frequency filter's cut-off will be sequentially re- C 100 2.6 3.1 1.0 54
moved. Note both the quantitative (measured) parameters as well D 300 2.6 3.0 0.8 30
as the qualitative (shape) aspects of the waveforms.
Let us first consider the effect a progressive elevation of the Figure 3 - / 5 . Low-frequency filter elevation effect on a
low-frequency filter has on an antidromic SNAP. A bandwidth SNAP. Sequential elevation of t h e low-frequency filter ( A - D ) can
of 1 Hz to 10,000 Hz reveals a biphasic initially negative poten- have rather dramatic effects on t h e quantitative and qualitative aspects
tial with an amplitude of 65 pV, onset and peak latencies of 2.6 of a SNAP. N o t e t h e alterations in the peak latencies as well as t h e po-
ms and 3.3 ms, respectively, and a negative spike duration of 1.4 tential's amplitude reduction. Additionally, a third phase is added as the
ms (Fig. 3-15A-D). Note that the baseline has a very jagged ap- low-frequency filter approaches 100 Hz and higher.The waveform's
pearance that is a composite signal comprised of internal high- slow return t o baseline in A is eliminated thus shortening t h e entire
frequency noise of the instrument, electrical noise from the potential's duration. (From Dumitru D, Walsh NE: Practical instrumen-
patient, and various environmental sources. Elevating the low-fre- tation and c o m m o n sources of error. Am J Phys Med Rehabil
quency filter 10 times does not appreciably alter the morphology 1988;67:55-65, with permission.)
time interval, i.e., this portion of the waveform is predominated

by high frequencies. Since we elevated the low-frequency filter,
there was no effect on the high frequencies with the potential's
onset remaining unchanged. This exemplifies why the low-fre-
quency filter is also called a high-pass filter.
Although the negative peak's inflection point has a short time
span (high frequency), that portion of the waveform preceding
and following the peak covers more time and is influenced by
the summated magnitudes of the relatively lower frequencies
(Fig. 3-11). We might anticipate that elevating the low-fre-
quency filter would prohibit the frequencies that require more
time to be resolved from contributing to the waveform. In one
sense, the low frequencies are being sequentially extracted, re- Low Latency
sulting in a waveform with a progressively greater amount of Trace frequency Amplitude
I high frequencies. Preventing electrical signals from contribut- filter Onset Peak
ing to the potential's total signal content would be expected to
Hz ms mV
result in a smaller potential, i.e., removing frequencies from the
waveform means there is less contained in the observed signal. A 1 3.5 6.7 21
In other words, removing signals from the waveform would be B 10 3.5 6.3 21
expected to reduce and not increase its magnitude. Additionally, C 100 3.5 5.0 11
that portion of the waveform from which the low-frequency sig-
nals were removed could also be expected to occur sooner in Figure 3-J6. Low-frequency filter elevation effect on a
time, because the remaining higher frequencies now have a CMAP. Elevating the low-frequency filter on a CMAP has very clear
greater influence thus biasing the potential to a higher-fre- effects similar t o t h o s e observed for t h e SNAP. (From Dumitru D,
quency content than previously. Higher frequencies occur more Walsh NE: Practical instrumentation and c o m m o n sources of e r r o r .
rapidly per unit of time than lower frequencies, thereby biasing Am J Phys Med Rehabil 1988;67:55-65, with permission.)
| the portion of the waveform with less low frequencies to occur
sooner in time. These two observations are reflected in a SNAP
with less amplitude and a shorter peak latency. Additionally, re- onset latency. The slow return of the potential (low frequen-
moval of low-frequency signals also results in a negative spike cies) is also truncated, thereby reducing the observed poten-
that requires less time to develop. Aside from the onset latency tial's total duration (Figs. 3-15 and 3-16).
(high frequencies), the entire waveform shifts left on the CRT,
i.e., it occurs sooner (Fig. 3-15). HIGH-FREQUENCY (LOW-PASS) FILTER
The observation of a third phase being produced with eleva-
tion of the low-frequency filter also can be explained by con- As previously stated, biologic waveforms have both low-
sidering the concept of subcomponent frequencies. Increasing and high-frequency subcomponents. The high frequencies of
the low-frequency filter essentially removes the low frequen- most waveforms are contained in the portions of the potential
cies contributing to the waveform. The waveform, therefore, that change rapidly, e.g., rise time and inflection points. As it 20
now consists preferentially of higher frequencies. The remain- is reasonably easy to change an instrument's filter settings, the
ing higher frequencies are no longer influenced by the lower clinician should be aware of how a high-frequency filter that is
ones and can be observed. Higher frequencies cycle or recur set too low can distort a waveform. The most practical way to
repeatedly more often over the same time period compared to demonstrate these waveform changes is to evoke an an-
the low frequencies and therefore have more phases. By re- tidromic SNAP or CMAP and sequentially lower the high-fre-
moving the low frequencies from a waveform, one would an- quency filter in the following steps: 10,000, 2,000, 1,000, and
ticipate that the remaining potential would display more 500 Hz. - The low-frequency filter remains at 10 Hz through-
18 45
phases. Another way of thinking about this process is that sig- out the entire example. Similar waveform parameters to those
nals resembling the DC component (relatively low frequency) previously discussed will be investigated: onset latency, peak
of the potential are extracted, leaving the signals that more latency, negative spike duration, and amplitude (Figs. 3-17 and
closely resemble AC (relatively high frequency). In reality we 3-18).
are speaking about voltages versus current, as this is what the Lowering the high-frequency filter incrementally progres-
instrument actually measures. An alternating voltage tends to sively removes the waveform's higher frequencies, leaving a
have the same amount of magnitude above and below the potential with a relatively lower frequency content below the
baseline. A waveform that consists primarily of higher fre- upper cut-off limit. One could anticipate that lower frequencies
quencies or alternating voltages begins to resemble a sinu- occur over a relatively longer period of time compared to
soidal curve. This is exemplified by elevating the low-frequency higher ones. The portions of the waveform that remain for ob-
filter while recording a CMAP (Fig. 3-16A-C). Note how the servation may very well occur comparatively later in time.
CMAP begins to resemble an alternating sinusoidal curve as These predicted findings are indeed found clinically (Fig. 3-
more and more low frequencies are prohibited from contribut- 17). A 20-fold decrease in the high-frequency filter from
18
ing to the potential. The end result of this process is an addi- 10,000 Hz to 500 Hz results in an 1 1 % delay in the onset la-
tional phase. Increasing the low-frequency filter above the tency and a 27% prolongation in the peak latency. The ampli-
frequencies contained in a waveform decreases amplitude, tude decreased about 16% while the negative spike duration
shortens peak latency (phase lead), decreases negative spike increased 30%. Similar changes are demonstrated for a C M A P
duration, increases phases, but does not significantly affect with the exception that the amplitude changes little, if at all, as
(so MV FILTER EFFECTS ON MUAPS

One can anticipate that altering the low- and high-frequency
filters may alter needle electromyographic MUAP parameters
in addition to SNAP and CMAP characteristics. Elevating the
low-frequency filter from 2 Hz to 20 Hz does not affect MUAP
amplitude, area, turns, or negative peak duration for monopolar
ancl standard concentric needle electrode r e c o r d i n g s . The
912
same filter changes, however, significantly reduce the MUAP's

duration for both types of needle electrodes. Increasing the low-
frequency filter to 500 Hz significantly reduces MUAP ampli-
tude, duration, area, and peak duration. The number of phases
High Latency also can increase and is more prominent for monopolar than
Trace frequency Amplitude concentric needle electrodes. The above findings are very simi-
filter Onset Peak lar to those described for SNAPs and CMAPs and similar expla-
Hz ms nV nations can be used. The lower propensity of additional phases
A 10,000 2.7 3.3 76 for a concentric compared to monopolar needle may be under-
B 2,000 2.8 3.4 76 stood if one considers that the low-frequency onset and termina-
C 1,000 2.8 3.8 75 tion of the MUAP is most likely recorded by both the cannula
D 500 3.0 4.2 64 and active site, which accounts for the slightly smaller duration
of concentric needle electrodes. Removal of the lower fre-
51
quencies by common mode rejection prohibits the low-fre-

Figure 3-17. High-frequency filter effect on a SNAP. Reducing
quency filter from artifactually creating additional phases, as is
the high-frequency filter primarily prolongs t h e SNAP and reduces its
the case for SNAPs and CMAPs. Reducing the high-frequency
amplitude somewhat (table). Note how the baseline smooths out as the
filter from 10,000 Hz to 2,000 Hz does not significantly affect
high frequencies (internal noise of the instrument) are removed. (From
any of the MUAP parameters because the rise time of MUAPs
Dumitru D,Walsh NE: Practical instrumentation and common sources
is 0.5-1 ms (2,000-1,000 Hz respectively), which is within the
of error. Am J Phys Med Rehabil 1988;67:55-65, with permission.)
2,000 Hz filter range. However, lowering the high-frequency
filter to 1,000 Hz or less results in MUAPs with a smaller am-
it apparently contains mostly low frequencies (Fig. 3-18). plitude and a longer duration.
Decreasing the high-frequency filter results in onset and peak
latency delays ( p h a s e lag), mild amplitude reductions, and NOTCH FILTER
longer negative spike durations.
The majority of electrophysiologic instruments provide the
practitioner with the option of applying either a 50-Hz or 60¬
Hz n o t c h filter depending on the potential line frequency
used. A notch filter is a filter with the selective ability to
remove only designated frequencies from the waveform. The
most common use of notch filters is during the performance of
the needle electromyographic examination when a large
50/60-Hz signal originating from some power source contami-
nates the potentials under investigation. It is also possible in
some instruments to impose a notch filter during the sensory
and motor conduction portion of the examination, although
little signal distortion usually occurs to either SNAPs or
CMAPs during the nerve conduction examination from notch
filter application. Similarly, little signal distortion occurs to
High Latency fibrillation potentials or MUAPs following implementation of
Trace frequency Amplitude a 60 Hz notch filter. However, it is possible to distort the posi-
filter Onset Peak tive sharp wave somewhat during the needle electromyo-
graphic examination if a 60-Hz notch filter is used (Fig. 3-19).
Hz ms mV The frequency content of this signal certainly contains fre-
A 10,000 3.3 6.9 21 quencies in the 5 0 - 6 0 Hz range with a distinct "notch" noted
B 2,000 3.5 7.1 21 in the terminal negative phase of the positive sharp wave. As
C 1,000 3.8 7.3 21 long as the practitioner is aware of potential signal distortion
D 500 4.2 7.5 21 that can occur with notch filters, they can be used rather effec-
tively. Not uncommonly, immediately following insertion of
Figure 3-18. High-frequency filter effect on a CMAP. Minor the needle electromyographic electrode into muscle tissue, a
changes in the CMAP are described as t h e high-frequency filter is low- considerable 60-Hz signal may be observed which responds to
ered, suggesting that this waveform contains a substantial amount of the notch filter. However, after a few minutes the interference
frequencies below 500 Hz. (From Dumitru D, Walsh NE: Practical in- may subside, most likely a result of a reduction in needle im-
strumentation and common sources of error. Am J Phys Med Rehabil pedance following an interaction with body fluids. Therefore,
1988;67:55-65, with permission.) from time to time during the needle examination it may be
SOUND
After the biologic signal has been amplified, but before it is
digitally processed, the energy from the signal is converted into
sound through a loudspeaker. The sound of MUAPs is particu-
50 pV
larly important to both potential recognition and criteria for
analysis. Normal and pathologic muscle potentials investigated
with needle electrodes produce very characteristic sounds that
help identify them. Biologic signals, therefore, are analyzed
from both a visual and acoustic standpoint. The proximity of a
potential also can be qualitatively judged by its sound proper-
ties. High-pitched sharp sounds signify that the needle electrode
is very near the site of action potential generation. Low-pitched
rumblings suggest that one is relatively far from the site of elec-
50 uV trical activity. A clinician's training is not completed until one
can identify normal and abnormal potentials by both sight and
sound.
Figure 3-/9. Effect of a 60-Hz notch filter on a PSW. A, A pro-

totypical biphasic positive sharp wave (PSW) as recorded from a den- ANALOG T O DIGITAL (A/D) CONVERSION
ervated muscle. B, A similar PSW t o that in A is recorded, but a 60-Hz
notch filter is employed. N o t e t h e "notched" effect in t h e PSWs ter- Once the biologic signal has been detected by the electrode,
minal negative phase. amplified, and then filtered, most commercially available in-
struments convert the real-time or analog waveform into its
digital representation. This conversion process is performed by
worthwhile to check and see if the 60-Hz notch filter can be an A/D converter in which each measured portion of the analog
turned off. It is better to try to eliminate the interfering signal potential is assigned a binary code designation for amplitude
(clean electrodes, electrodes of similar composition, reduce and latency. The total analysis time is manually assigned by the
skin/electrode impedance, place preamplifier close to the pa- clinician and referred to as an epoch or sweep. The sweep ex-
48
tient, etc.) than use a notch filter. tends from one aspect of the cathode ray tube to the other and is
divided into equal time segments known as bins or sample
RECOMMENDED FILTER SETTINGS points. A computer takes a fraction of a second to analyze each
point that is called the dwell time. A reciprocal of the dwell
The practitioner is expected to perform motor, sensory, time is the sampling frequency. Provided the computer h a s
needle electromyographic, and evoked potential studies. Each sufficient memory, it can assign minute portions of the wave-
procedure has particular filter settings based on the optimum form investigated to a corresponding vertical and horizontal il-
frequency content of mean waveforms routinely observed. The luminated point on the CRT, forming an adequate digital
filter settings should be above and below the major subcompo- representation of the original analog signal.
nent frequencies of the waveform to be investigated. These pa-
rameters have been arrived at over the years by simply lowering VERTICAL RESOLUTION
the low- and increasing the high-frequency filters until the
waveform under investigation no longer changed appreciably An A/D converter's resolution is usually discussed in terms
for its respective high- and low-frequency filter settings. of bits of resolution, with a bit representing one binary digit,
Individual clinicians then modified these filter parameters to either 0 or 1. A grouping of bits is referred to as a byte, which
28
their particular needs. There are no universally accepted guide- in turn is used to define some numerical value. If an A/D con-
lines for filter settings but only recommendations (Table 3-2). verter has 12 bits of resolution, it has the ability to resolve a
Following the range of upper and lower frequency filter limits total of 2 (4096) individual amplitude levels. Therefore, let us
1 2
provided yields waveforms with minimal filter distortions and assume we use a sensitivity setting of 20 pV/div for a screen
background noise. with 10 divisions, and a waveform with a maximum potential
amplitude of 200 pV is recorded. This means our A/D converter
samples the potential's amplitude in steps of 0.049 pV (200 pV
Table 3-2. Recommended Filter Settings -r 4096 intervals). If we change the sensitivity to 10 pV/div, w e
Procedure Low Frequency High Frequency can sample the same waveform in 0.024 pV intervals, thereby
resolving smaller waveform changes. Today's possibilities of
NCV (motor) 2 - 1 0 Hz 10,000 Hz
amplifiers and high-resolution A/D converters makes it possible
NCV (sensory) 2 - 1 0 Hz 2,000 Hz to disconnect the described relation between display sensitivity
EMG (routine) 2 0 - 3 0 Hz 10,000 Hz and sampling resolution. The amplification of the signal and
A/D resolution is usually preset in the software as a default
EMG (quantitative) 2 - 5 Hz 10,000 Hz
value depending on the type of study and expected signal ampli-
SFEMG 500-1,000 Hz 10,000-20,000 Hz tude levels. Changing the sensitivity during the measurement
SEP l-IOHz 500-3,000 Hz merely changes the display of the signal on the screen but does
NCV, Nerve conduction velocity; EMG, Needle electromyography; SFEMG, not change the A/D conversion or amplification. This is in effect
single-fiber electromyography; SEP, somatosensory evoked potential. called the "display sensitivity." Unfortunately, the way the system
handles this is usually not transparent to the user. It is the re- CMAPs. For example, a sensitivity of 1,000 pV/div results in
sponsibility of the clinician to understand and use the correct the amplitude marker progressing vertically in 9.7 pV incre-
preset conditions of the apparatus. ments (1,000 pV/div x 10 div/screen x 1 screen/1024 points =
Unfortunately, the majority of instruments available can pro- 9.7 pV/div). Similarly, a sensitivity of 10,000 pV/div causes the
vide approximately 600-1024 discrete vertical resolution or amplitude scale to change in 97 pV steps. As noted above, using
amplitude points representing either millivolts or microvolts too low sensitivity (waveform occupying a small portion of the
with respect to the screens resolution. In this case, the screen
13
screen) does not permit one to assign enough points of resolu-
resolution is obviously considerably lower than that of the A/D tion to differentiate relatively small changes in the waveform
converter. This limitation in screen resolution is often made (Fig. 3-20).
worse by the habit of many manufacturers to display the signals
in windows covering only part of the screen. The clinician has HORIZONTAL RESOLUTION
the responsibility of manipulating the instrument's gain or sen-
sitivity so that the complete waveform occupies approximately Two relatively simple requirements must be fulfilled for ade-
one-half the CRT's vertical height. If the amplifier's sensitivity quate horizontal waveform analysis: (1) total screen analysis
is adjusted so that the waveform completely fills the C R T any time must be sufficient to encompass the entire potential under
variation in the signal or baseline may overwhelm the number investigation; and (2) screen resolution (dwell time or the in-
of vertical resolution points. This can cause the waveform to verse of the sampling frequency) must be sufficient to resolve
extend beyond the upper and lower extremes of the CRT, result- the signal's most rapid changes. The horizontal portion of the
ing in a loss of potentially valuable information (Fig. 3-20). A CRT screen is a time scale usually measured in milliseconds
waveform that occupies significantly less than one-half the CRT and divided into 10 equal divisions. The number of resolution
screen can have significant portions of its signal distributed be- points across the total screen is fixed by the manufacturer and
tween resolution points and again be lost to measurement. typically contains 512 or 1024 data points to resolve the wave-
Fortunately, because the A/D converter's resolution far exceeds form. The clinician has the option of designating the amount of
that of the screen, it is possible for instruments manufactured time allotted for analysis over the screen's width, i.e., the time
presently to sample the waveform at a high resolution and dis- chosen by the instrument's operator determines over how much
play the waveform at many different sensitivities without com- time the limited number of data resolution points are distrib-
promising the waveform's morphology significantly until the uted. Assigning relatively large time intervals to the whole
extremes of the A/D converter's resolution is reached. screen results in less resolution to accurately define minor
The above concepts can be clarified if one considers an in- waveform changes. The resolving power of the instrument is re-
strument to contain 1024 points of vertical resolution for a duced because each individual point on the screen is expected to
screen with 10 vertical divisions. If the sensitivity is initially set cover more time. However, decreasing the total screen time
at 10 pV/div, as used in sensory nerve conduction studies, the allows each point to be responsible for smaller portions of the
instrument's amplitude markers can move in 0.097 pV incre- waveform, thereby increasing the likelihood of detecting minor
ments. This incremental amplitude designation can be calcu- signal variations. This can be understood by providing a few
lated as follows: 10 pV/div x 10 div/screen x 1 screen/1024 simple examples.
points = 0.097 pV/point. If the sensitivity is changed to 1 Let us assume that there are 1000 points of resolution across
pV/div, the amplitude marker now travels vertically in 0.009 pV the screen. If we adjust the instrument so that an entire screen is
jumps (1 pV/div x 10 div/screen x 1 screen/1024 points = 0.009 assigned 1,000 ms (1 second), our sweep speed is 100 ms/div
pV/pt). An amplifier with less sensitivity is used for evaluating (1,000 ms/10 divisions = 100 ms/div). The instrument's resolu-
tion for the present set of parameters can be calculated by con-
sidering the number of points and time across the entire CRT
A ^ B C
screen. The CRT screen's resolution is 1,000 points of resolu-
tion per 1,000 ms or 1 point/1 ms (1,000 points/1,000 ms = 1
point/1 ms). Recall that the dwell time is the time that passes
before the computer analyzes the next point and in this case is 1
ms per 1 point (1 ms/1 point). The sampling rate or sampling
D . E F frequency has already been stated at 1,000 points/1 sec (i.e.,
1,000 Hz), and is the reciprocal of the dwell time or number of
points sampled each second. The sampling frequency is typi-
cally expressed in hertz (cycles/second) and for our example it
becomes: 1,000 points/1,000 ms = 1,000 points/1 sec = 1,000
Figure 3-20. The effect of analog-to-digital conversion on Hz. If for the same 1,000 CRT screen points we can now assign
signal of different amplitude. Sine waves of three amplitudes (A, B, a total of 100 ms across the screen, a new dwell time of 0.1
C) are digitized using vertical intervals of equal size. At each step, t h e ms/point is created (100 ms/1,000 points = 0.1 ms/point). The
amplitude, marked by a dot, is measured.These measures are shown at new sampling frequency is 10,000 Hz (1 point/0.1 ms x 1,000
the bottom (D, E, F) where the dots a r e connected by straight dashed ms/1 sec = 10,000 points/1 sec = 10,000 Hz). It can be seen
lines.The signal in trace "A" exceeds t h e vertical range of the digitizer from the above example that for the same number of points
and is distorted in t h e digital representation in trace "D".The signal in across the screen, the resolution is increased by assigning less
trace "B" fills about one-half of the digitizer range and is fairly well re- time to the same number of points. The resolution increased 10-
solved in trace "E". T h e signal in trace " C " fills only a small portion of fold for one-tenth as much time, i.e., from 1 ms/1 point (1,000
the total vertical range and is poorly resolved in trace "F". (Modified Hz) to 0.1 ms/1 point (10,000 Hz). In other words, for a CRT
from Spehlmann R: Evoked Potential Primer. Stoneham, MA, screen of 100 ms there are 10 resolution points every 1 ms in-
Butterworth-Heinemann, 1985.) stead of 1 resolution point every 1 ms as there are for a total
A B C has an upper limit subcomponent frequency of 5,000 Hz (1 rise

time/0.2 ms x 1,000 ms/1 sec = 5,000 rise times/1 sec = 5,000
Hz). The minimum sampling frequency (Nyqvist frequency) is
twice the highest subcomponent frequency, yielding 10,000 Hz
(2 x 5,000 Hz = 10,000 Hz). To ensure an accurate reproduction
0 2 4 6 8 10 0 2 4 6 8 10 0 2 8 6 8 10 , m s
of the signal, a sampling frequency of 3 times the Nyqvist fre-
(ms) (ms) (ms)
quency (30,000 Hz, or 6 times the rise time frequency) may be
D E F
used. A sampling frequency of 30,000 Hz means that the wave-
form is sampled once every 0.033 ms (30,000 Hz = 30,000
points/1,000 ms = 1 point/0.033 ms). If the screen contains
1,000 resolution points, our sweep speed (number of ms/divi¬
sion) should be set at 3.3 ms/div (0.033 ms/1 point x 1,000
Figure 3-21. The relationship between sampling rate and points/1 screen x 1 screen/10 divisions = 3.3 ms/div) to provide
signal frequency. Effects of different signal frequencies at t h e same a sampling frequency of 30,000 Hz. An instrument with 1,000
sampling rate. Sine waves at t h r e e different frequencies (traces A, B, C) points of resolution across the screen and a sweep speed of 3.3
are sampled at t h e same rate. At each sampling interval, their ampli- ms/div should adequately resolve the given rise time of a single
tude is marked by a dot.These dots, connected by straight dashed lines muscle fiber potential without difficulty. If our instrument had
are shown as a digital display at t h e bottom (traces D, E, F).The sine 500 points of resolution for the same waveform, we would need
wave in trace "A" is sampled a b o u t 14 times p e r cycle and well de- a sweep speed of 1.65 ms/div (0.033 ms/point x 500 points/1
picted in the digital representation in trace "D".The sine wave in trace screen x 1 screen/10 div = 1.65 ms/div). This finding simply im-
"B", being sampled at a rate only 6 time higher than its own frequency, plies that an instrument with half as many screen resolution
is depicted with less but still fair detail after digitization as shown in points requires a sweep speed twice as fast as a comparable in-
trace "E". In contrast, t h e sine wave in trace " C " , being slightly above strument with twice as many resolution points for the same op-
the critical value of one-half t h e sampling rate is n o t resolved ade- timal signal reproduction. In short, increasing the sweep speed
quately in the digital representation in trace "F" showing fewer peaks decreases the total time assigned to the screen, thereby allowing
than t h e analog waveform and is an example of aliasing. (Modified from the same number of resolution points to be applied over a
Spehlmann R: Evoked Potential Primer. Stoneham, MA, Butterworth- shorter interval. For example, changing the sweep speed from 5
Heinemann, 1985.) ms/div to 1 ms/div is what is meant by "increasing" the sweep
speed. Screen resolution and waveform subcomponent fre-
quency resolution are interrelated (Table 3-3). This process re-
CRT screen time of 1,000 ms. The more points per millisecond, sults in assigning more resolution points to better define rapidly
the greater the resolution of the screen, which in turn allows the changing portions of the waveform. Incidentally, a high-fre-
instrument to better define portions of the waveform occurring quency filter setting of at least 10,000 (2 x 5,000-rise time)
very rapidly (high-frequency subcomponents). If the high fre- should be provided to avoid waveform distortions from too low
quencies are adequately resolved, there is no difficulty in accu- a high-frequency filter.
rately defining the low frequencies as there are more than
enough points of resolution for the slowly changing aspects of LATENCY MEASUREMENT
the waveform.
Recall that a biologic waveform may be considered to consist Another important issue to consider with respect to horizontal
of a series of low- and high-frequency subcomponent sine screen resolution is the accurate measurement of waveform la-
waves. To display a sine wave digitally, a minimum of two reso- tency parameters. It can be helpful to know the time of occur-
lution points are required per cycle. As a result, the instrument's rence for particular portions of a waveform, e.g., onset or peak
sampling frequency must be at least twice the highest subcom- latency. Should a clinically relevant aspect of a potential that re-
ponent frequency of that sine wave to minimally resolve the quires measurement lie between two horizontal resolution points,
signal. This critical sampling frequency is known as the Nyqvist one must take the next available higher or lower screen point. The
frequency. The Nyqvist sampling frequency only provides the
48
sweep speed also affects the accuracy of time measurements. For
minimum resolution frequency and does not guarantee that the
waveform will be sufficiently reproduced without distortion. It
Table 3-3. Horizontal Screen Resolution
is a good practice to ensure that the instrument's sampling fre-
quency is several multiples of the Nyqvist frequency. If the sam- Required
pling frequency of the instrument is insufficient to resolve the Rise Waveform High Sampling Sweep
signal's subcomponent frequencies, the recorded waveform will Time Frequency Frequency Resolution Frequency Speed
be distorted (Fig. 3-21). This distortion has the effect of not (ms) (Hz) (Hz) Points (Hz) (ms/div)
recording all of the signal with a potential for missing various 0.2 5,000 10,000 1000 20,000 5.0
components, A reduction in the waveform's components, partic-
0.2 5,000 10,000 500 20,000 2.5
ularly if entire phases are eliminated, has a tendency to make
the waveform appear as if it contains lower frequencies than it 0.2 5,000 10,000 250 20,000 1.25
really does. The effect of reducing a waveform frequency con- 0.5 2,000 4,000 1000 12,000 8.3
tent through inappropriately low sampling frequencies is called
0.5 2,000 4,000 500 12,000 4.2
aliasing. A few simple examples may help explain this impor-
tant concept. The high-frequency filter (HF) is 50% greater than the high-frequency content
of the waveform under discussion to prevent any filter distortion. A screen
Let us begin with a single muscle fiber potential with a rise width of 10 divisions is assumed for these examples.The rise time estimates
time (high-frequency subcomponent) of 0.2 ms. The rise time the highest frequency portion of the waveform.
peak latencies at 5 ms/div than 1 ms/div, resulting in marked

differences between the two sets of values. Also, upon close in-
spection of the potential at 5 ms/div one can see the "ratchety"
appearance of the response indicating that the screen's resolu-
tion is too low to accurately reproduce every point of the analog
signal, thereby losing data through the A/D conversion process.
In less expensive instruments it is also possible for the time cur-
sors to be less accurate than the digital representation of the
waveform, although the waveform is sampled every 0.02 ms,
e.g., the cursor may only be able to record changes in incre-
ments of 0.04 ms. The clinician should understand the limita-
Sweep Latency tions of the instrument and how these shortcomings can affect
Trace Amplitude the accuracy of measurement. This example also demonstrates
speed Onset Peak
how important it is to reproduce exactly all of the recording
ms/div ms conditions used by another laboratory if their reference data are
A 1 2.7 3.2 61 used. Failure to do so may result in artifactually normal or ab-
B 2 61 normal results and diagnostic errors.
2.6 3.3
C 5 2.3 3.4 61
AVERAGING
Figure 3-22. A S N A P r e c o r d e d with different s w e e p speeds. Averaging Principles
N o t e how the potential's onset and peak latencies (table) change with Occasionally, the biologic response under investigation is
progressively slower sweep speeds ( A - C ) . Also, the waveform starts small compared to the surrounding noise of muscle, ECG,
t o break up as fewer points of resolution are assigned t o it. (Modified EEG, and other electrical activity in the body so that the de-
from Dumitru D, Walsh NE: Practical instrumentation and c o m m o n sired potential is not clearly defined but "hidden" within the
sources of error. Am J Phys Med Rehabil 1988;67:55-65.) observed trace. The process of averaging extracts a signal that
is time-locked to a depolarizing stimulus from the surrounding
example, assume a screen has 500 points of resolution, and a background electrical noise by sequentially adding stored
sweep speed of 5 ms/div is used for a screen with 10 horizontal traces and dividing by the total number of responses. For ex-
divisions. The screen's resolution at this sweep speed is 1.0 ample, a peripheral nerve is activated and the ensuing small
point every 0.1 ms (500 pts/50 ms = 1 pt/0.1 ms). A particular electrical response is recorded from another location. The stim-
waveform's peak, however, may lie at 3.16 ms, while the instru- ulating pulse starts the CRT's sweep and the resulting signal
ment can place a time cursor at 3.1 ms and 3.2 ms but not 3.16 plus coincident noise is recorded. Each aspect of the analog
ms. This means that in order to place a time cursor on the wave- signal (noise plus response) is given a digital amplitude and la-
form's peak, it becomes necessary to overestimate or underesti- tency representation within the instrument's computer through
mate the peak latency, thereby affecting the measurement's A/D conversion and stored in its memory (Fig. 3-23). A 48
accuracy. The same rationale applies for onset latency or any second stimulus is delivered by the instrument and the result-
other aspect of the waveform. It is possible to improve the ing analog waveform is again converted into a digital signal
screen's resolution by increasing the sweep speed to 1.0 ms/div. and stored. The second stored response is added point-for-point
In this case, the instrument's resolution is 1 point every 0.02 ms to the first stored trace with respect to latency and amplitude.
(500 pts/10 ms = 1 pt/0.02 ms) or 5 times the previous resolu- The resulting amplitude (voltage) for each digital point is di-
tion. Although it is still possible for a waveform's area of inter- vided in half. If the averaging is stopped at this point, the wave-
est to fall between points occurring every 0.02 ms, one can form is displayed. On the other hand, should a third response
approach the waveform's desired portion with more accuracy. be collected, it undergoes A/D conversion, added to the previ-
Using relatively high (low-resolution) sweep speeds (20-50 ous two waveforms, and then divided by 3, i.e., the total
ms/div) compared to low sweep speeds (1-10 ms/div) tends to number of averages. This process is continued for the number !
"compress" the waveform over a smaller portion of the screen. of averages specified by the investigator. Depending on the
These factors predispose one to greater measurement errors be- amount of background noise present, the signal eventually
cause portions of the waveform can be lost between points of emerges from the noise. Because most electrical noise is
resolution. Note that the designations "high" and "low" sweep random and not time-locked to the delivered stimulus, it occurs
speeds are equivalent to "slow" and "fast" sweep speeds as the in a random fashion with respect to the desired signal that ap-
first set of terms applies to the numbers read on the instrument pears at the same time following each stimulus. The random
that the investigator manipulates and the second set of terms noise is removed,because of phase cancellation while the regu-
refers to how much time is on the screen and thus how rapidly larly appearing signal's phases summate. The final response
the trace "sweeps" across the screen. extracted from the noise can then be analyzed in its digital
As illustrated in Figure 3-22A, the instrument recorded the format or transformed back into an analog signal through digi-
onset and peak latencies for a sensory potential at 2.7 ms and tal-to-analog (D/A) conversion. One only needs to average a
3.2 ms respectively for a sweep of 1.0 ms/div and 500 points of sufficient number of responses to clearly define the signal com-
screen resolution. Decreasing the sweep speed (reducing the pared to the background noise. Sometimes the response in the
resolution) to 5 ms/div (Fig. 3-22C) compressed the potential to signal is time-locked but not in the same phase in every occur-
a smaller portion of the screen and the onset and peak latencies rence. An example is the measurement of reflexes with surface
decreased and increased, respectively. Comparably fewer reso- EMG. A short increased drive of the motor system can result in
lution points were available to accurately resolve the onset and an amplitude rise of the surface EMG, which can be both positive
ANALOG DIGITAL
STIMULUS •
Figure 3-23. Analog-to-digital conversion.Two rather "noisy" analog signals are digitized through A/D conversion.The t w o digital responses
are then summated. Random signals phase cancel while similar signals summate.The summated response is then divided by 2 and t h e process is r e -
peated. For however many averages are performed, the composite signal is divided by the total number of waveforms so that each waveform has
equal weight in the final averaged waveform.The final digital waveform is then converted back into an analog signal.This sequence of events depicts
the manner in which potentials are averaged t o improve the signal-to-noise ratio. (Modified from Spehlmann R: Evoked Potential Primer. Stoneham,
MA, Butterworth-Heinemann, 1985.)
or negative. An example is the measurement of medium- and let us assume a single stimulus is delivered to a nerve and the
long-latency reflexes. In this case it is necessary to rectify the ensuing signal has an amplitude of 2 pV, while the noise ampli-
signal before averaging, i.e., to make all voltage differences be- tude is 4 pV. The S/N ratio is equal to 1/2 (S/N = (2 pV x
tween the two recording electrodes of the same sign (either (Vl)}/4 pV = 1/2), which means that the surrounding noise is
positive or negative). In this way a modulation of the EMG twice as big as the signal and will obviously obscure the signal
output can be measured by adding up all stimulus-related am- from optimal analysis. If we now perform 4 averages, the S/N
plitude changes in the surface EMG. ratio improves to 1 (S/N = (2 pV x (V4)}/4 pV = 1), implying
that the comparative, one trace versus 4, S/N has been en-
Signal-to-Noise (S/N) Ratio hanced by a factor of 2, i.e., the second S/N of 1 is twice that of
The purpose of averaging is to improve the signal's ampli- the previous S/N of 1/2. A S/N of 1 suggests that 4 averages
tude compared to that of the surrounding noise's amplitude, have resulted in a signal that is the same size as the noise.
i.e., the signal-to-noise (S/N) r atio . Once the S/N ratio has Continued averaging to 64 trials improves the S/N by a factor
been optimized, averaging no longer provides additional infor- of 8, and the signal is now 4 times that of the noise. The
mation. Unfortunately, there are a number of regularly occur- number of averages required to best define a response has a
ring biologic and environmental signals (EEG, ECG, 60-Hz somewhat arbitrary end-point. If the investigator suspects that
line interference, and others) that can obscure the desired averaging is required, the number of averages should be set rel-
signal even though they are not time-locked. A portion of the atively high, but stopped when the signal no longer appreciably
60-Hz interference from artificial current/voltage generators changes with additional averages and the baseline noise level
may be reduced if a stimulus rate not evenly divisible into 60 has stabilized. A good practice is to repeat a similar set of aver-
is used (e.g., 2.8 Hz). Obviously, it is impossible to eliminate ages to confirm the waveform's reproducibility. In order to
all extraneous signals from contaminating the desired wave- judge the relative amplitude of remaining noise in relation to
form, but averaging and appropriate stimulation rates can help the response after averaging it is very useful to have a prestim-
considerably. ulus interval. This enables the clinician to estimate the amount
It is known that there is a nonlinear relationship between the of baseline noise and can be helpful in determining if a real
number of responses averaged and the S/N ratio. From a prac-
48
time-locked response is present in the signal. It is possible that
tical standpoint, averaging enhances the S/N ratio by a factor many thousands of averages are required to define a potential,
that is the square root of the number of averages performed rendering its recording impractical, particularly if a poor
(S/N oc V#averages). This statement can be expressed mathe- methodology for recording is used. The clinician must ensure
matically by setting the S/N ratio equal to the signal's ampli- impeccable recording technique with respect to reducing skin
tude (S) times the square root of the number of averages ( \ n ) , impedance, securing properly functioning electrodes with suf-
which is then divided by the random noise's amplitude (A): ficient electrolyte cream to the patient, and providing an ade-
S/N = ( S x (Vn)} -r A. To illustrate the number of averages re- quate stimulus to evoke the desired response. Averaging cannot
quired to make an appreciable difference in the detected waveform, replace good technique.
90 — P A R T I FUNDAMENTAL PRINCIPLES
CATHODE RAY TUBE (CRT) DISPLAYS then became possible for the instrument to repetitively generate
the stored signal without having to continuously shock the pa-
CATHODE RAY TUBE tient. The commercially available instruments presently raster
or continuously regenerate the response multiple times per
A cathode ray tube is an electronic device that uses an elec- second. The raster mechanism works by having magnetic coils
tron beam to produce a visible trace on a coated screen. At the15
instead of deflection plates (for greater control) direct the elec-
narrow end of the CRT is an electron gun capable of generating tron beam to sweep horizontally across the screen for a prede-
a continuous stream of electrons (Fig. 3-24). The electron beam termined number of vertical divisions approaching 256 or more.
passes through a pair of horizontal and vertical deflecting As the beam sweeps from left to right, the computer controls
plates. By changing the potential difference between the hori- whether the beam is off or on. When the beam is on, the screen
zontal plates the electron beam is directed horizontally. Similar lights up. Each point of horizontal resolution may be on or off.
changes in potential on the vertical deflecting plates direct the The horizontal on or off points of resolution are sequentially ac-
electron beam vertically. Simultaneously altering the potential tivated from left to right. The CRT trace is then immediately
on both sets of plates causes the electron beam to move in any brought back to the left-hand side of the screen and displaced
direction to light up the CRT's screen. The screen is coated with somewhat inferior to the next horizontal row of resolution
phosphor, which absorbs the energy from the electrons and points repeating the process. This continuous left/right and
converts it into visible light. The glow of the phosphor screen top/bottom movement traces out the memorized digital wave-
persists for some time after the electrons have passed, allowing form contained in the computer's memory for the 256 or more
one to observe the beam's path. An electronic circuit connected rows. When the bottom right-hand corner of the CRT screen is
to the horizontal deflection plates causes the electron beam to reached, the entire process begins again. In short, rastering re-
repetitively sweep across the screen at rates (sweep speeds) degenerates the digital representation of the analog signal stored
termined by the operator. in the instrument's memory repeatedly over a very short period
of time so that the phosphor that signifies an "on" position is
ANALOG SIGNAL prevented from fading and allows the waveform to appear with
unvarying luminance. The waveform appears "frozen" on the
The original electrodiagnostic instruments used an analog or CRT screen so that the investigator can analyze any portion of
real-time signal to drive the vertical deflection plates, thereby the waveform without causing undue discomfort to the patient.
describing a voltage change over the screen's time input by the
practitioner. For observing MUAPs in needle electromyogra- TRIGGER AND DELAY LINE
phy, a free-running or continuous sweep was used. When per-
forming nerve conduction studies, the current stimulator A method to present a continuously firing MUAP at the same
triggered the sweep to visualize the ensuing SNAP or CMAP. place on the CRT screen was developed to assist in morphologic
Multiple stimuli were necessary because the response would analysis during analog needle electromyographic examinations.
only persist for the duration of the phosphor. The convenience Although this was a real-time signal and required the patient to
of storing or "freezing" a trace on the screen to leisurely exam- continuously activate the same MUAP over time, it essentially
ine its fine details was not available. "froze" the potential on the CRT so that it could be investigated.
Some portion of the real-time MUAP, e.g., initial rise or peak
DIGITAL SIGNAL amplitude, triggered the CRT sweep to begin moving. The
signal was then fed into an electronic circuit to phase delay with
The technologic advance of A/D conversion allowed the de- minimal signal distortion and in a sense "run around" inside the
sired signal to be stored within the instrument's "memory." It instrument until the sweep was several divisions along the
screen. The delayed analog signal was then fed to the electron
Beam-forming mask
beam so that the potential could be traced onto the screen, and
the investigator could determine the amount of delay and the
trigger criteria. The end result of this trigger and delay line was
to delay or place the repetitively triggered waveform at the same
place on the CRT's screen so that it could be studied easily.
Currently available instruments can digitally store a signifi-
cant portion of an analog signal and retrieve selected portions of
this time span. Once the desired MUAP has triggered the CRT's
sweep, the instrument's computer can select a chosen time
period preceding and following the region surrounding the trig-
gered potential. The instrument then presents this information
and continually updates the screen to monitor any changes in
the potential. Because the potential is in the instrument's
memory through analog to digital conversion, it can be "frozen"
on the CRT at any time. When this is accomplished, the instru-
Figure 3-24. A c a t h o d e r a y t u b e ( C R T ) . The electron beam ment continually replenishes the chosen time period so rapidly
sweeps across t h e screen at regular rates by t h e horizontal deflection that to the human eye it appears to be stationary. With respect to
plates while t h e vertical deflection plates describes the waveform de- the trigger and delay line, the patient is still required to continu-
tected by t h e instrument. (Modified from Diamond SR: Fundamental ally activate the muscle under investigation. If voluntary activa-
Concepts of Modern Physics. New York, AMSCO School Publications, tion ceases, the last stored potential usually remains on the
1970.) screen.
STIMULATORS extracellular surface. This local positive charge density "neu-

tralizes" a portion of the intracellular anions' attraction for sur-
ACTIVATION OF EXCITABLE TISSUES rounding intracellular positive ions as well as contributing to a
repulsion force (like charges repel each other). These cations
Electrical activation of excitable tissues can be accomplished now have more intracellular mobility and can migrate toward
by applying a cathode and anode in close proximity to the nerve the region of extracellular negativity surrounding the cathode
or muscle and causing current to flow from the anode to cath- that attracts them (opposite charges attract). An accumulation
ode. A cathode is the negative terminal of a current source and of intracellular positive ions in the vicinity of the cathode d e -
attracts positive ions (cations). The anode is the positive cur- creases the transmembrane attraction of the local intracellular
rent terminal and attracts negative ions (anions). The path of anions for the extracellularly located positive ions. The extra-
current flow is by convention in the direction of positive ions, cellular cations can then proceed toward the cathode, complet-
i.e., from anode to cathode. Action potentials in nerve and ing a current flow. Recall from neural propagation (see Chapter
muscle can be induced by either internally or externally applied 1) that this type of current flow is a capacitive current; it is me-
depolarizing currents. Although a microelectrode can be in- diated through ions that do not physically pass through the
serted into an excitable cell, this technique is rather demanding membrane. Larger current intensities or longer durations of
and selective for that particular nerve or muscle cell. Clinically, current flow will build up enough charge at the cathode's site
one is much more interested in the conduction properties of to eventually decrease the transmembrane potential to thresh-
whole nerves. Extracellular excitation techniques, therefore, are old. Once threshold is achieved, the positive feedback loop of
more commonly performed for routine investigations. opening voltage-dependent sodium ion gates produces an all-
Excitation of peripheral nerves is routinely performed with or-none action potential that results in neural propagation. An
the cathode and anode located on the skin's surface above the alternative explanation is to consider that the - 90 mV resting
nerve's presumed pathway. An electrical pulse is applied across membrane potential of the cell's interior compared to the extra-
the cathode and anode that follows the path of least resistance. cellular space is essentially negligible compared to the several
At low levels of stimulation, the current flows from the anode to hundred volts generated by the stimulator. When the stimulator
cathode in the extracellular fluid surrounding the nerve. At generates a pulse of sufficient strength to create a - 100 V
higher current intensities, some of the current enters the nerve. about the cathode, the inside of the cell is relatively more posi-
If the current is of sufficient strength, the nerve will be depolar- tive than prior to the application of the stimulus, i.e., - 0.090 V
ized and an action potential will be induced to propagate along inside compared to zero outside the cell before, and - 0.090 V
the nerve length. This process of depolarization can be ex- inside compared to - 100.0 V outside the cell during the stimu-
plained if one considers a surface cathode and anode placed lus. In a sense, the inside of the cell is close to zero, which is a
above an unmyelinated nerve located at some depth within an more positive potential with respect to - 100.0 V, and therefore
limb (Fig. 3-25). At a high current intensity, the anode pro- the membrane is effectively depolarized. This large change in
duces a large number of positive charges localized about itself. transmembrane potential effectively depolarizes the cell by
Some of these positive charges accumulate about the nerve's causing the sodium gates to open and beget action potential
propagation.
In myelinated nerve, the depolarization and hyperpolariza-
tion occur primarily at the nodes of Ranvier, which is the path
of least resistance for current flow. Similar principles as those
for unmyelinated nerve apply to muscle with a few modifica-
tions regarding impedance characteristics. Depolarization
occurs under the cathode, while hyperpolarization is produced
under the anode. Anodal hyperpolarization is believed to result
in anodal block. Anodal block does not occur in human beings
under routine clinical conditions using a bipolar stimulator de-
livering currents commonly used in neural activation. T h e17
anode can, however, effectively stimulate or depolarize the

nerve if sufficient current is delivered.
STIMULATOR TYPES
Two basic types of stimulators are commercially available:
constant-voltage and constant-current. The duration over which
a stimulus is delivered can usually be varied between 0.05 ms
and 1.0 ms. The usual standard practice in exciting peripheral
nervous tissue is to use a suprathreshold excitation, i.e., a cur-
Figure 3-25. N e u r a l s t i m u l a t i o n . A stimulator's cathode (-) and rent or voltage intensity 10-20% above that sufficient to depo-
anode (+) are placed on t h e skin overlying an unmyelinated nerve. A , larize the entire nerve. This can be determined by observing the
Small amounts of c u r r e n t a r e delivered insufficient t o activate t h e amount of stimulus required to produce an evoked nerve or
nerve. B,A larger stimulus is presented t o t h e nerve which causes it t o muscle response that no longer increases despite more current,
depolarize. A capacitive current hyperpolarizes t h e neural tissue under and then providing an additional 10-20% more current. As pre-
the anode while depolarization occurs about t h e cathode. (Modified viously stated, an impedance is always present between the
from Brown W F : T h e Physiological and Technical Basis of Electro- stimulator's surface and the tissue to be activated. Additionally,
myography. Boston, Butterworth-Heinemann, 1984.) this impedance can vary during the course of the investigation
or between multiple stimuli at the same site. Ohm's law, E = I x nerve may be impaled by the needle, which apparently does not
Z, can be used to appreciate the characteristics of the two cate- result in long-term sequelae. Signs and symptoms suggesting
gories of stimulators. If the impedance (Z) varies between suc- nerve penetrations include paresthesias in the nerve's sensory
cessive stimuli and a constant current stimulator delivers the area, deep aching pain in muscles innervated by the nerve,
same amount of current (I) during each impulse, then the volt- muscle fasciculations, and less than 1.0 mA required to excite
age (E) output must vary. If the intervening tissue's impedance the nerve. The site of stimulation is well defined by the loca-
5
between the cathode and nerve increases or decreases, the volt- tion of the needle's tip. The anode's placement should be at least
age necessary to drive the same amount of current into the body several centimeters away from the cathode and can be on the
for each impulse must also increase and decrease, respectively. skin's surface or subcutaneous. One should not use a standard
For example, a constant-current stimulator placed on the skin concentric or bipolar needle electrode to deliver a depolarizing
with a total electrode-to-nerve impedance of 5,000 O set to de- current because of the danger of increased current densities
liver 20 milliamperes (0.02 amperes {A}) for every impulse will about the rather small cathode surface. The stimulating surface
require 100 V to drive this current into the body (E = I x Z = area of these two electrodes cannot be increased as they are
0.02 A x 5,000 Q. = 100 V). Should the impedance decrease to fixed by the manufacturer.
2,500 Q after several stimuli, 20 mA will continue to be deliv- When using a surface stimulator, the size of the cathode and
ered, but the voltage required for this amount of current is now anode have been suggested to be at least 3-5 mm in diameter to
50 V. avoid concentrating the current density about the skin and pro-
A constant-voltage stimulator emits the same voltage for each ducing pain in the patient. It is also possible that the site of
5
stimulation, but may vary the amount of current required if the neural excitation does not exactly correspond to the presumed
impedance between the electrode and excitable tissue changes. site of skin cathode placement. The nerve may lie relatively
If 100 V are necessary to excite a nerve with a cathode-to-anode deep to the cathode or take an unexpected course requiring large
impedance of 10,000 Q, then 10 mA is the necessary current for stimulus intensities. An increase in the stimulation's intensity or
nerve excitation. However, if the impedance increases to 10,500 duration may cause the actual site of depolarization to spread
Q, the instrument will deliver 9.5 mA for a constant 100 V stim- some distance from the cathode (see Chapter 2). There are ad-
ulation. Both the constant-current and constant-voltage stimula- vantages and disadvantages to both surface and near-nerve
tor are adequate for routine purposes. If it becomes necessary to needle stimulation techniques (Table 3-4). It is advisable for the
quantify the amount of current delivered for a particular investi- clinician to become familiar with both types of stimulation
gation, e.g., SSEP or facial nerve stimulation, a constant-cur-
48 19
methods and to recognize when they are clinically appropriate.
rent device is preferred. Constant-current stimulators also may
produce less stimulus artifact (see below) than a constant-volt- STIMULATOR PLACEMENT ERRORS
age stimulator. 5
Close inspection of most but not all commercially available

SURFACE VS. NEEDLE EXCITATION surface stimulator devices reveals that the cathode (black) and
anode (red) are color-coded. Additionally, a minus (-) and plus
It is possible to stimulate excitable tissues with any of the fol- (+) sign are located near the cathode and anode, respectively.
lowing cathode/anode combinations: (1) cathode/anode on the During an examination, it is possible to unknowingly reverse
skin's surface, (2) monopolar near-nerve needle cathode placed the location of the cathode and anode at one or possibly more
in close proximity to the nerve and a somewhat distant com- stimulation sites, resulting in potentially erroneous diagnostic
pletely bare (to spread out the current) subcutaneous anode, and conclusions. To illustrate this possible source of error, let us
18
(3) monopolar near-nerve needle cathode and a surface anode. first correctly stimulate the median nerve at the wrist 8 cm prox-
Either a constant-current or constant-voltage stimulator can be imal to the abductor pollicis brevis muscle's motor point, and
used with any of the above three techniques. The first type of again at the antecubital fossa 23 cm proximal to the previous
cathode/anode combination (surface stimulation) has already excitation site. The distal and proximal latencies to the CMAP
been discussed above and is widely used because of conve-
nience. A near-nerve needle cathode also can be used. Most op-
T a b l e 3-4. Surface vs. N e a r - N e r v e N e e d l e S t i m u l a t i o n 4 4 4
erators use needle cathodes because a particular nerve may be

located deep within the body and not be readily accessible (e.g., Advantage Disadvantage
spinal nerve roots) to a surface stimulator or when surface stim- Surface
ulation requires excessive amounts of current to excite a nerve N o need to puncture skin Uncertainty about excitation site
and subsequently causes patient discomfort. The near-nerve Time efficient Excitation requires more current/
cathode requires a low current (voltage) intensity because the voltage t o excite nerve than near-
depolarizing current density associated with the needle's tip is nerve needle and may be relatively
located next to the nerve to be excited. Less current also pro- more painful
duces relatively little discomfort to the patient provided they do
not mind the initial needle insertion. A needle or surface anode Near-Nerve Needle
can be used with similar results. Excitation of deep-lying May take more time
nerves
Caution has been expressed regarding too high a current den-
Less current density/ Can pierce nerve
sity accumulating about the stimulating needle's tip and poten-
less pain
tially injuring the nerve. Peeling the Teflon back from a
monopolar needle decreases the current density and possibly More exact location of May scare patients
protects the nerve from damage. Also, a stimulus duration of
53 nerve excitation (less
0.05 ms without removing the Teflon has been recommended stimulus spread)
when using a near-nerve needle as a cathode. - Occasionally a
44 60 Less stimulus artifact Possibly more expensive
are noted to be 3.1 and 7.1 ms, respectively. The clinically cal-
18
us first consider the upper-limb SEP with a 50 ms analysis
culated nerve conduction velocity (NCV) over the 23-cm seg- time. If the entire waveform requires at least 50 ms (latency of
ment of nerve is 58 m/s (NCV = distance/time = 230 mm/(7.1 action potential to cortex + potential's duration = CRT screen's
ms - 3.1 ms) = 58 m/s). The cathode and anode are separated by analysis time: 50 ms) to be resolved, it can be thought of as re-
a distance of 2.5 cm, which means that it will take the nerve's curring once every 50 ms for a frequency of 20 Hz (1/50 ms x
action potential 0.4 ms to travel this distance (58 m/s = 25 mm/t; 1000 ms/1 sec = 20 Hz). If the entire screen/potential repre-
t = 0.4 ms). sents 50 ms of analysis, a stimulus frequency greater than 20
If the median nerve is inadvertently excited at the wrist with Hz means that two stimuli will be given during one CRT
the stimulator's cathode and anode reversed, the distal motor la- sweep. As a result, 20 Hz is the stimulus frequency's upper
tency is now 3.5 ms. The CMAP's time of onset (3.5 ms) is pro- limit. Two stimuli delivered during one sweep may cause wave-
longed compared to the previous value because the cathode is form distortions because the two waveforms will overlap with
no longer 8.0 cm but 10.5 cm (8.0 cm + 2.5 cm) proximal to the ensuing phase addition or cancellation. In the lower limb, the
muscle because of the added distance of the cathode/anode sep- stimulation frequency's maximum is 10 Hz (1/100 ms x 1000
aration. It will take 0.4 ms longer for the action potential to ms/1 sec = 10 Hz). Although the stimulus rate upper limits
reach the muscle with the cathode and anode reversed distally. have been derived from a purely instrumentatal standpoint, ad-
Stimulating the nerve correctly at the elbow now yields an inter- ditional physiologic factors within the nervous system may re-
stimulus time of 3.6 ms (7.1 ms - 3.5 ms) for an NCV of 64 m/s. quire lower stimulus rates (< 5-7 Hz) to avoid waveform
The distal stimulation error resulted in a 6 m/s elevation above distortion (see Chapter 9 ) .
48
the true conduction velocity. Reversing the cathode/anode ori-

entation proximally but not distally produces a prolonged STIMULUS ARTIFACT
elbow-to-muscle latency of 7.5 ms (7.1 ms + 0.4 ms). The NCV
calculated with a proximal stimulus error is 52 m/s or 6 m/s less When performing peripheral nerve stimulation studies, the
than the true NCV. A consistent error of reversing the beginning of the CRT trace is typically disturbed by a relatively
cathode/anode relationship both distally and proximally cancels large potential decreasing toward the baseline from either the
the error and the correct NCV of 58 m/s is obtained (NCV = positive or negative direction. This potential results from the
230/{7.1 ms + 0.4 ms} - {3.1 ms + 0.4 ms} = 58 m/s). The stray currents of the stimulator, or stimulus artifact, being
distal motor latency, however, remains erroneously prolonged. recorded by the instrument. Occasionally, this potential may in-
This example clearly demonstrates that inadvertent reversal of terfere with the desired response as its decrementing phase co-
the cathode and anode can increase or lower a nerve's NCV. If a incides with the investigated potential's onset or peak latency. It
patient has a normal NCV, it is possible to reduce it into the ab- is important to know how to minimize the occurrence and detri-
normal range. Likewise, an abnormally slow NCV can be elemental effects of stimulus artifacts.
vated into the normal range, causing the investigator to overlook The stimulus artifact is nothing more than the current/volt-
a disease state. The distal motor latency, although prolonged by age delivered from the stimulator's electrodes taking paths of
distal cathode/anode reversal, may lead to the same potential least resistance through the extracellular fluid and being d e -
errors as for NCV. tected by the recording electrodes. It precedes the action poten-
tial because neural conduction is considerably slower than the
STIMULUS FREQUENCY volume-conducted stimulus artifact. Also, the stimulator effec-
tively acts as a dipole current source, thus generating a far-field
The optimal rate at which a stimulus is delivered to evoke a potential (see Chapter 2) which extends instantaneously
particular response may need to be considered, particularly if throughout the body and is detected immediately by the record-
the waveform has a long acquisition time. A potential may re- ing electrodes. Although the stimulator is electrically isolated
quire a relatively long time to be resolved because it may have a from the instrument and its ground, a stimulus artifact is still de-
long duration, arise from a polysynaptic pathway (blink reflex), tected. The phrase isolated from ground means that a stimulat-
traverse the peripheral nervous system twice (F-wave; H- ing pulse is generated in the stimulator through a transformer
reflex), or the site of stimulation may be far from the recording that does not require physical contact through wires but uses an
electrode location. An example of a potential that arises from electromagnetic field. An absence of physical contact within the
some of these conditions is the somatosensory evoked potential instrument reduces a portion of the artifact from being detected
in which a nerve is excited at the wrist or ankle and one set of through the instrument's wiring and limits its observation to the
recording electrodes is located on the scalp. As you might antic- current effects within the body.
i ipate, it will take a relatively long time for the induced action The investigator may have to effect several possible changes
potential at the wrist and ankle to reach the somatosensory at the (1) stimulus site, (2) intervening tissue between the stimu-
cortex. The lower-extremity waveform may take approximately lus location and recording electrodes, and (3) at the recording
40 ms while the upper-extremity potential requires 22 ms to electrodes if the stimulus artifact interferes with the observed
result in waveforms recorded at the brain. In pathologic condi- potential (Table 3-5). The stimulator's anode can be rotated
tions these latencies can be considerably longer. Additionally, about the cathode in small increments in either the clockwise or
evoked potentials contain mostly low frequencies and as a result counter-clockwise direction until the artifact is minimized. 37
are long-duration potentials. The complete waveform may be 20 This can significantly reduce the stimulus artifact because the
ms or greater in duration. A total analysis time of 50 ms and 100 isopotential lines between the cathode and anode are preferen-
ms are usually recommended for upper- and lower-limb SEPs, tially aligned by rotation so that similar lines of potential or
respectively. near-zero potentials originating from the stimulator are placed
Given the extended analysis time required to completely re- on the two recording electrodes (Fig. 3-26). Similar data pre-
solve the entire SEP waveform compared to SNAPs and CMAPs, senting to the differential amplifiers will be eliminated as a
the stimulus frequency becomes an important consideration. Let common mode signal. If the stimulus artifact is large, dissimilar
Table 3-5. Stimulus Artifact Reduction paths of current flow. Wiping the skin thoroughly with alcohol
Remove perspiration circumferentially around limb between often can prevent the stimulus artifact from preferentially trav-
stimulator and recording electrodes eling over the skin to reach the recording site. The impedance
beneath the stimulating and recording electrodes should be re-
Reduce impedance between skin and recording/stimulating
duced (gently abrading the skin, cleaning electrodes) so that
electrodes
both electrodes record as nearly the same signal from the stimu-
Use minimal amounts of electrolyte paste lator as possible to assist in common mode rejection.
Avoid electrolyte paste between cathode and anode Remember that differences between the two recording elec-
Remove all body lotion and makeup between stimulator and trodes are amplified. If there is a marked disparity of impedance
recording electrodes for E-l and E-2, the electrodes record slightly different signals
from the stimulator that will be amplified instead of rejected.
Place ground electrode between stimulator and recording Recording electrode leads should be kept as short as possible. A
electrodes large ground electrode placed adjacent to the E-l recording
Use just supramaximal stimulus electrode between it and the stimulator will help dampen the
Rotate anode about cathode stimulus artifact. Crossing the stimulator's cable with that of the
amplifier or electrode leads must be avoided in order to mini-
Use a needle cathode t o approach stimulus site
mize artifacts. The stimulus artifact is a slowly changing, low-
Try constant-current as opposed t o constant-voltage stimulus frequency potential that can be reduced by elevating the
low-frequency filter slightly. It is important not to increase the
high-pass (low frequency) filter too much because it can distort
lines of potential extending out into the body from the cathode the signal of interest. Any or all of the above techniques can
and anode are recorded and amplified instead of being reduced help alleviate excessive stimulus artifact.
through common mode rejection. As low a stimulus intensity as If the above methods fail to sufficiently reduce the stimulus
possible should be used to evoke a maximal potential. One artifact, one should be aware of potential latency effects result-
should also locate the cathode as close to the nerve as possible ing from superimposition of electrophysiologic waveforms and
and near-nerve needles can help reduce annoying artifacts. the artifact. A negative stimulus artifact (above the baseline)
53
Unnecessarily large stimulus voltages can overwhelm the am- decaying toward the baseline while a potential's negative phase
plifiers and prevent the nerve or muscle response from being is still rising will result in the onset latency shifting artifactually
observed. The distance between the stimulating and recording to the left, i.e., shortened. The rising phase of the waveform can-
electrodes should be as large as possible to still record the de- cels the descending phase of the artifact thereby decreasing the
sired signal. An increased distance allows the stimulus artifact onset latency. A negatively decaying stimulus artifact coincid-
enough time to settle back to baseline before the neural impulse ing with a potential's peak shifts the peak's latency to a shorter
reaches the recording electrodes' location and evokes a re- value. The declining negative aspect of the artifact cancels the
sponse. Perspiration and electrolyte gel must be removed from initial rising negative portion of the waveform, resulting in a po-
the skin separating the various electrodes to eliminate aberrant tential that occurs artifactually earlier in time. On the other
hand, if the negative stimulus artifact is increasing coinciden-
tally with the potential's negative peak, the entire potential
shifts to the right, increasing the onset and peak latency. Of
course, the stimulus artifact also can originate below the base-
line (positive direction) and interfere with waveform recordings.
A stimulus artifact below but decaying toward the baseline can
overlap with a negative spike potential and prolong the peak la-
tency. If the artifact is below the baseline but moving away from
it when a potential is rising, phase cancellation effects will
shorten the onset latency. A waveform's peak occurring with a
positively sloping stimulus artifact shortens the peak latency.
INTERFERENCE
The environment in which most electrodiagnostic examina-
1.0ms tions are performed contains substantial amounts of electrical
interference or noise. For our purposes we may define interfer-
Figure 3-26. Anodal Rotation. An antidromic median SNAP is ence as any signal that can contaminate the desired potential
recorded with the cathode (Ca) located 7 cm from the active record- whether it originates internally from the patient (biologic/artifi-
ing electrode (Ac) with the reference electrode (R) located 4 cm distal cial) or some external source. A number of patient-generated
t o the active e l e c t r o d e . T h e anode (An|) is 2.5 cm proximal t o the noises or artifacts can interfere with the waveform under inves-
cathode.Trace A reveals the SNAP with an ill-defined onset latency, as tigation; it is helpful for the investigator to recognize them so
it is clearly affected by the stimulus artifact. Rotating the anode about that appropriate solutions can be found to reduce them. One of
the cathode in 0.5 cm increments ( A n - A n ) eventually results in sup-
2 4 the more common noises observed is movement artifact, which
pression of that portion of the stimulus artifact that interferes with can occur in two primary forms. If surface electrodes are used,
the SNAP's onset, thus permitting m e a s u r e m e n t of t h e parameter the patient can unintentionally cause the recording electrodes to
(traces B-D). shift positions slightly. This movement can cause the skin to
Chapter 3 INSTRUMENTATION — 9 5
alter its shape creating a voltage difference between the skin and interference problems. When the source of environmental noise
electrode metal (electrode potential) with some frequency com- remains unknown, one can connect the active and reference am-
ponent. If this discharge has a frequency greater than the filters' plifier ports with a jumper cable and use this as an antenna to
bandwidth, a signal may be created and the potential thus inter- search out the directional characteristics of the noise and try to
feres with the desired biologic response. Firmly securing the define its source. Environmental noise often can be eliminated
electrode can help reduce this problem. A second common by surrounding the examination room completely with copper
movement artifact is incomplete relaxation of muscles in close wire, forming the so-called Faraday's box. This is a rather ex-
proximity to the recording electrode. The recorded MUAPs can pensive venture and can be totally defeated by not using an iso-
overlap with the waveform under investigation and obliterate lated earth ground, i.e., grounding the copper shielding to a
the response. It is also possible for regularly firing MUAPs to different ground from that used by other instruments in the
be mistaken for the desired response, particularly when the true building. A simple and effective method to reduce environmen-
response is pathologically absent. Turning the volume up on the tal noise is to ground the electric outlet to an isolated earth
speaker may help identify the contaminating MUAPs and the ground spike. It is also important to plug the instrument into a
patient can be instructed to relax. When performing needle elec- dedicated circuit. This is an electric circuit coming straight
tromyographic examinations on muscles about the thorax, e.g., from the outside power line into your office that is used only to
pectoralis and serratus anterior muscles, a regularly recurring power the instrument. These two relatively simple procedures
waveform of rather large amplitude occasionally can be ob- should take care of most environmental artifacts.
served to impede MUAP observation. This is merely the ECG
and can be ignored as it rarely poses a significant inconve-
nience. Similarly, a regularly recurring spike potential indicates ELECTRICAL SAFETY
that the patient most likely has a cardiac pacemaker.
Interference arising outside of the patient is frequently ob- The clinician must realize that all electrical devices leak some
served when high amplifier gains are used. The most common current to the instrument's chassis and electrodes. Although the
source of environmental interference is the 60-Hz noise. It may manufacturer must ensure that their instruments leak minimal
be recognized by a complete cycle (peak-to-peak time interval) currents, the operator has the responsibility of ensuring periodic
every 16.7 ms (60 cycles/1000 ms = 1 cycle/16.7 ms). The electrical safety checks of both the electrophysiologic device
sudden appearance of 60-Hz interference during the course of and laboratory by qualified personnel. A failure of the instru-
an examination should raise the possibility that an electrode has ment's safeguards with respect to electrical safety can lead to
become disconnected either from the patient or instrument (Fig. large and potentially dangerous current flows into the patient. If
3-27). It is also possible that an electrode lead has failed. An im- a patient comes in contact with a metal bed or other electrical
pedance mismatch between the two electrodes can reduce instrument, a faulty ground plug can allow current to flow into
common mode rejection causing a smooth undulating sine the patient through the amplifier's ground lead. The maximum
wave. Reducing the impedance of the electrode or using the acceptable leakage current from a chassis to ground is 100 pA
same type of electrode can diminish these artifacts. A rather and from the patient's input leads to ground is 50 pA. The fre-
53
characteristic alternating positive/negative spike pattern of fluo- quencies of current to be concerned about range from 0 to 1,000
rescent lighting also is commonly seen (Fig. 3-27). If there are Hz because they are capable of exciting nerve and muscle
no problems with the electrodes, using a simple incandescent tissue. Electrically sensitive patients (having intravenous or
light bulb with a table lamp allows one to turn off the fluores- other lines that may lead to the heart, such as arterial lines or
cent lights. Also, it is possible to purchase fluorescent lights catheters) should not be exposed to more than 10 pA of leakage
with low-noise condensers and this may help eliminate any current. Electrical stimulation of peripheral nerves should be
J
5ms
20 uV J
10ms
5000^V
Figure 3-27. Common forms of interference observed

on a CRT screen. A, Fluorescent light during an electromyo-
graphic examination. B, Pure fluorescent light interference reveal-
J J
ing details of its usual alternating spike appearance. C,
Combination of fluorescent light and 60-Hz interference. D, 25000 uV 5000uV
Almost pure 60-cycle interference. E, Same t r a c e as that in D
except a 60-Hz notch filter has been introduced into the record- 5ms 10ms
ing system.
J5ms
25000uV
performed away from the thorax in patients with cardiac pace- 14. Cooper R: Electrodes. Am J EEGTech 1963;3:91-101.
15. Diamond SR: Fundamental Concepts of Modern Physics. New York, AMSCO
makers. Also, prior to stimulating electrically sensitive patients,
School Publications, 1970.
one should thoroughly dry the skin of any perspiration to avoid 16. Dorfman LJ, McGill KC, Cummins KL: Electrical properties of commercial con-
conducting any current to the electrically sensitive area. centric EMG electrodes. Muscle Nerve 1985;8:1-8.
A few relatively simple safety measures should be employed 17. Dreyer SJ. Dumitru DD, King JC: Anodal block versus anodal stimulation: Fact
or fiction. Am J Phys Med Rehabil 1993;72:10-18.
to ensure optimal patient safety: 18. Dumitru D, Walsh NE: Practical instrumentation and common sources of error.
1. It is necessary to have the device routinely inspected by , Am J Phys Med Rehabil 1988;67:55-65.
qualified personnel familiar with measuring current leakage and 19. Dumitru D, Walsh NE, Porter LD: Electrophysiologic evaluation of the facial
nerve in Bell's palsy. Am J Phys Med Rehabil 1988;67:137-144.
defective electrical instruments.
20. Dumitru D, Walsh NE: Electrophysiologic instrumentation. In Dumitru D (ed):
2. The electrical outlet into which the instrument is con- Clinical Electrophysiology: Physical Medicine and Rehabilitation State of the
nected also should be checked for electrical safety. Art Reviews. Philadelphia, Hanley & Belfus, 1989.
3. In order to avoid power surges, the device should be 21. Dumitru D, Lester J: Comparison of SEP surface and needle electrodes. Arch
Phys Med Rehabil 1991;72:989-992.
turned on or off, respectively, prior to attaching and after remov- 22. Dumitru D, King JC, Nandedkar SD: Motor unit action potentials recorded with
ing electrodes. concentric electrodes: Physiologic implications. Electroencephalogr Clin Neuro-
4. Avoid using extension cords, as they may increase leakage physiol 1997;105:333-339.
23. Dumitru D, King JC, Nandedkar SD: Concentric/monopolar needle electrode
currents. 53
modeling: Spatial recording territory and physiologic implications. Electro-

5. All electrical devices that are to be attached to the patient encephalogr Clin Neurophysiol 1997;105:370-378.
must be plugged into the same outlet to share a common ground. 24. Dumitru D, King JC, Nandedkar SD: Comparison of single-fiber and macro elec-
6. Attach only one ground to a patient during the examination. trode recordings: Relationship to motor unit action potential duration. Muscle
Nerve 1997;20:1382-1388.
Close attention to technical details will ensure the safe acqui- 25. Dumitru D, King JC, Nandedkar SD: Concentric and single fiber electrode spa-
sition of data with minimal risks to the patient. tial recording characteristics. Muscle Nerve 1997;20:1525-1533.
26. Dumitru D, King JC: Concentric needle recording characteristics related to depth
of tissue penetration. Electroencephalogr Clin Neurophysiol 1998;109:124-134.
27. Fleasel AF, Tuck RR: Variability of repeated nerve conduction studies. Electro-
CONCLUSION encephalogr Clin Neurophysiol 1991;81:417-420.
28. Gitter AJ, Stolov WC: AAEM rninimonograph #16: instrumentation and mea-
surement in electrodiagnostic medicine—part II. Muscle Nerve 1995; 18:
An important but frequently overlooked aspect of the electro-
812-824.
diagnostic medicine examination is the practitioner's under- 29. Guld C: On the influence of the measuring electrodes on duration and amplitude
standing of the instrument. This is unfortunate because a faulty of muscle action potentials. Acta Physiol Scand 1951;25(Suppl 89):30-32.
comprehension of how the instrument modifies biologic signals 30. Howard JE, McGill KC, Dorfman LJ: Properties of motor unit action potentials
recorded with concentric and monopolar needle electrodes. Muscle Nerve
certainly can result in significant distortions of the acquired 1988;11:1051-1055.
data. Distorted waveforms can readily lead to a misinterpreta- 31. Jasper H, Ballem G: Unipolar electromyogram of normal and denervated human
tion of the data and lead to a false positive or negative diagnosis. muscle. J Neurophysiol 1949;12:231-244.
A diligent attempt to understand the material presented in this 32. Johnson EW: The EMG examination. In Johnson EW (ed): Practical
Electromyography, 2nd ed. Baltimore, Williams & Wilkins, 1988.
chapter should provide the practitioner with sufficient informa- 33. Karam DB: AIDS and the electromyographer. Arch Phys Med Rehabil
tion to avoid many potential instrumentation errors. 1986;67:491.
34. Kimura J: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and
Practice. Philadelphia, F.A. Davis, 1983.
REFERENCES 35. King JC, Dumitru D, Stegeman D: Monopolar needle electrode spatial recording
characteristics. Muscle Nerve 1996; 19:1310-1319.
1. Adrian ED, Bronk DV: Discharge of impulses in motor nerve fibers. J Physiol 36. Kohara N, Kaji R. Kimura J: Comparison of recording characteristics of mono-
1929;67:119-151. polar and concentric needle electrodes. Electroencephalogr Clin Neurophysiol
2. Barkhaus PE, Nandedkar SD: Recording characteristics of the surface EMG 1993;89:242-246.
electrodes. Muscle Nerve 1994:17:1317 1323. 37. Kornfield MJ, Cerra J. Simons DG: Stimulus artifact reduction in nerve conduc-
3. Barohn RJ, Jackson CE, Adams KK, Gronseth GS: An electrophysiologic com- tion. Arch Phys Med Rehabil 1985;66:232-235.
parison of stainless steel and adhesive recording electrodes for nerve conduction 38. Kosarov D, Gydikov A: The influence of the volume conduction on the shape of
studies. Muscle Nerve 1995; 18:1218-1219. the action potentials recorded by various types of needle electrodes in normal
4. Braddom RL: Somatosensory, brainstem, and visual evoked potentials. In human muscle. Electromyogr Clin Neurophysiol 1975; 15:319—335.
Johnson EW (ed): Practical Electromyography. Baltimore, Williams & Wilkins, 39. Lawler JC, Davis MJ, Everton CG: Electrical characteristics of the skin. J Invest
1988, pp 369-416. Derm 1960;34:301-308.
5. Brown WF: The Physiological and Technical Basis of Electromyography. 40. Misulis KE: Basic electronics for clinical neurophysiology. J Clin Neurophysiol
Boston, Butterworth Publishers, 1984. 1989;6:41-74.
6. Buchthal F. Guld C. Rosenfalek P: Action potential parameters in normal human 41. Nandedkar SD. Sanders DB. Stalberg EV, Andreassen S: Simulation of concen-
muscle and their dependence on physical variables. Acta Physiol Scand 1954:32: tric needle EMG motor unit action potentials. Muscle Nerve 1988; 11:151 -159.
200-218. 42. Nandedkar SD, Dumitru D, King JC: Concentric needle electrode duration mea-
7. Buchthal F, Rosenfalek P: Action potential parameters in different human mus- surement and uptake area. Muscle Nerve 1997;20:1225-1228.
cles. Acta Psychiatr Neurol Scand 1955:30:125 131. 43. Pease WS, Bowyer BL: Motor unit analysis: Comparison between concentric
8. Butler BP. Ball RD, Albers JW: Effect of electrode type and position on motor and monopolar electrodes. Am J Phys Med Rehabil 1988;67:2-6.
unit action potential configuration. Muscle Nerve 1982;5:S95~S97. 44. Pease VVS, Fatehi MT. Johnson EW: Monopolar needle stimulation: Safety con-
9. Chu J, Chan RC: Changes in motor unit potential parameters in monopolar siderations. Arch Phys Med Rehabil 1989;70:412-414.
recordings related to filter settings of the EMG amplifier. Arch Phys Med 45. Pease WS, Pitzer NL: Electronic filter effects on normal motor and sensory nerve
Rehabil 1985;66:601-604. conduction tests. Am J Phys Med Rehabil 1990:69:28-31.
10. Chu J, Johnson RJ: Electrodiagnosis: An Anatomical and Clinical Approach. 46. Pollack V: The waveshape of action potentials recorded with different types of
Philadelphia, J.B. Lippincott. 1986. electromyographic needles. Med Biol Eng 1971;9:657-664.
11. Chu J, Chan RC, Bruyninckx F: Progressive Teflon denudation of the monopolar 47. Reiner S, Rogoff JB: Instrumentation. In Johnson EW (ed): Practical Electro-
needle: Effects on motor unit potential parameters. Arch Phys Med Rehabil myography. Baltimore. Williams & Wilkins, 1980.
1987:68:36^10. 48. Spehlmann R: Evoked Potential Primer. Stoneham, MA, Butterworth Publishers.
12. Chu J, Chan RC. Bruyninckx F: Effects of the EMG amplifier filter settings on 1985.
the motor unit action potential parameters recorded with concentric and monopo- 49. Stalberg EV, Trontelj J: Single Fiber Electromyography. Woking. UK, Mirvalle
lar needles. Electromyogr Clin Neurophysiol 1986;26:627-639. Press, 1979.
13. Cole JL: Equipment parameter determinants in evoked potential studies. Bull 50. Stalberg EV: Macro EMG: A new recording technique. J Neurol Neurosurg
Am Soc Clin Evoked Potentials 1984:3:3-11. Psychiatry 1980:43:475-482.
51. Stalberg EV, Chu J, Bril V, et al: Automatic analysis of the EMG interference 58. Weber RJ, Piero D: Entrapment syndromes. In Johnson EW (ed): Practical
pattern. Electroenceph Clin Neurophysiol 1983;56:672-681. Electromyography. Baltimore, Williams & Wilkins, 1980.
52. Stalberg EV: Personal communication. 59. Wee AS, Ashley RA: Relationship between the size of the recording electrodes
53. Stolov W: Instrumentation and measurement in electrodiagnosis. Mini- and morphology of the compound muscle action potential. Electromyogr Clin
monograph #16, American Association of Electromyography and Electrodiag- Neurophysiol 1990;30:165-168.
nosis, Rochester, MN, 1981. 60. Wongsam PE, Johnson EW, Weinerman JD: Carpal tunnel syndrome: Use of
54. Swain ID, Wilson GR, Crook SC: A simple method of measuring the electrical palmar stimulation of sensory fibers. Arch Phys Med Rehabil 1983;64:I6-
resistance of the skin. J Hand Surg 1985; 10-B:319-323. 19.
55. Thage O, Trojaborg W, Buchthal F: Electromyographic findings in polyneuropa- 61. Wiechers DO, Blood JR, Stow RW: EMG needle electrodes: Electrical imped-
thy. Neurology 1963;13:273-278. ance. Arch Phys Med Rehabil 1979;60:364-369.
56. Tjon-A-Tsien AML, Lemkes HHPJ, van der Kamp-Huyts AJC, van Dijk JG: 62. Wiechers DO: Single fiber electromyography. In Johnson EW (ed): Practical
Large electrodes improve nerve conduction repeatability in controls as well as in Electromyography, 2nd ed. Baltimore, Williams & Wilkins, 1988.
patients with diabetic neuropathy. Muscle Nerve 1996;19:689-695. 63. Zablow L, Goldenshohn ES: A comparison between scalp and needle electrodes
57. Van Dijk JG, Tjon-A-Tsien A, van der Kamp W: CMAP variability as a function for the EEG. Electroenceph Clin Neurophysiol 1969;26:530-533.
of electrode site and size. Muscle Nerve 1995;18:68-73.
Appendix
Basic E l e c t r i c i t y P r i m e r
John C King, B.S.E.E, M.D.
APPENDIX OUTLINE
Elemental Forces Parallel Circuits Bandpass Filters

Resistance Exercise 3 Digital Filtering
Exercise I Capacitance Electronic Devices
Series Circuit Alternating Current Switches • Diodes • Amplifiers • Differential Amplifiers
Exercise 2 Cathode Ray Tube • Digital Processors (Computers)
Inductors
Ground Filters Conclusion
This primer is written to explain as intuitively as possible orbiting negative electrons. Electron is the Greek word for
electricity and the function of the circuit elements: capacitors, amber. The amount of attraction is called voltage or electrical
inductors, resistors, rectifiers, switches, and amplifiers. These potential. Objects that have this attraction possess charge that
circuit elements are concepts that help explain how and why may be positive or negative. Negative charge is the amount of
electronic devices and biologic organisms react as they do to negative electrons present that are not balanced by protons.
electrical influences. All electronic devices can be simplified to Positive charge is the amount of excess protons not balanced by
variations of these six circuit elements. A working knowledge electrons. Objects that have excess positive or negative charge
of these circuit elements can help one understand and anticipate are said to be electrically "charged." Since electrons move more
the electrical behavior of the body, filters, and electrophysio- freely than protons, positive charge is usually due to a deficit of
logic instruments. electrons rather than a surplus of protons (Fig. 2 ) . The concepts
2
Since the circuit elements are really mathematical concepts of charge and electrical potential are analogous to the more
that help us understand how biologic and other real-life systems readily appreciated concepts of mass and gravitational poten-
respond electrically, occasionally mathematical expressions are tial. An object with mass in earth's gravitational field are at-
used. These cannot be read in the usual scanning fashion. tracted from a higher (gravitational potential) to a lower
Mathematical expressions are written in tightly compacted sym- position (Fig. 3 ) . The greater the mass (in kilograms), the
bolism and each symbol must be read as a word with each equa- greater the force of attraction (in Newtons) to the earth.
tion read as a complete sentence. Exercises are provided to help Similarly, an object with a property called positive charge tends
one appreciate the usefulness of the mathematical expressions to move from a higher to lower electrical potential. Oppositely,
1
in solving basic electrical problems. These exercises, once an object with negative charge tends to move from a lower
worked through, will also help one to more fully understand the (more negative) potential to a higher (more positive) potential.
concepts discussed as tools to be applied to electrophysiologic The greater the magnitude of the charge, the greater the magni-
phenomena. tude of force, or voltage, of attraction.
The nylon comb being pulled through dry hair results in re-
moving electrons from the comb's atoms, and the comb be-
ELEMENTAL FORCES comes positively charged (Fig. 1). When small objects, with
their atoms and their mobile electron clouds, are brought near,
Ancient Greeks found that by rubbing amber with a woolen the objects' electrons are attracted to the comb's excess positive
cloth, small objects such as bits of paper would be attracted to charges and migrate to occupy more of that side of those atoms.
the amber. This effect can also be seen in passing a nylon comb This results in the small objects' atoms becoming relatively po-
through dry hair (Fig. 1) that is subsequently attracted to the larized, i.e., with a more negative side near the positively
comb. This phenomenon is called electrostatic attraction. The1
charged comb, leaving the more positive side of the atoms far-
cause of electrostatic attraction relates to the basic structure of ther away from the comb. This electrostatic attraction is in-
atoms and an elemental force of attraction between two sub- versely related to the distance separating the charges (i.e., the
atomic particles. Positive protons in the center of an atom attract amber or comb will pick up paper from nearby but not from far
98
Appendix BASIC ELECTRICITY PRIMER — 9 9
away). The more negative side of the paper is closer and thereby
more attracted to the comb than the more distant positive side is
repulsed, and so the uncharged paper clings. When combing wet
hair, the electrons removed flow back to the comb through the
conductive water and the comb does not become charged.
Likewise, if the charged comb is connected to the earth, e.g.,
dipping it into water in a sink, the excess charges are dispersed
leaving essentially no charge and subsequently the paper bits
will not cling. This charge dissipation is commonly experienced
when one walks on a nylon carpet on a dry day and touches a
light switch plate. The excess electrons, mechanically removed
from the carpet by one's shoes, conduct through the body to the Electrostatic
metal plate that is attached through wiring ("grounded") to the
earth. The crackle and spark are the result of these electrons
"discharging" to the plate. Similar build-up of charges on a
more massive scale occurs in the development of thunder clouds
leading to more dramatic discharges called lightning (Fig. 4). 2
Electricity is the description of the movement of these ele- Figure I. Electrostatic attraction developed by removing electrons
mental charges. Atoms and single charges are so incredibly from t h e nylon comb.The electrons from the bits of paper are then a t -
small that their effects are difficult to measure. Thus a larger tracted t o t h e positively charged c o m b . (From Bergveld P:
number of charges, 6,250,000,000,000,000,000 (6.25 x 10 ) is 18
Electromedical Instrumentation: A Guide for Medical Personnel. N e w
considered the base unit of charge, the Coulomb, whose effects York, Cambridge University Press, 1980, with permission.)
can be more easily measured. When charges move, as during
discharging a comb to earth, a measure of the flow of charges
per unit time is desired and this is called current. One ampere freely mobile in the form of soluble charged ions, such as the
is the base unit of current and equals one coulomb per one sodium (positive) and chloride (negative) ions in normal saline.
second. Ampere is also written 1 A or 1 amp. One volt of attrac- When positive ions move, they do so in the direction defined as
tion of charge is defined as the attraction between 1 coulomb of positive current flow and negative ions move oppositely.
positive charge located 1 meter from 1 coulomb of negative The greater the voltage of attraction for any constant path-
charge in a vacuum. Since most electrical events of biologic in- way, the greater the current, i.e., more charges per unit time will
terest involve very small currents and voltages, the units often flow. The number of charges that can flow, however, also d e -
used are miiliamperes (mA) or millivolts (mV), which equal pends upon the pathway. A pathway made of the same material
1/1000 of an ampere or volt, respectively, or microamps (pA)
and microvolts (pV), which equal 10" ampere or volt, respec-
6
tively. Other prefixes often used in electrical measurements in-

clude M or mega (10 ), k or kilo (10 ), p or pico (10 ), and n or
6 3 -9
nano(10" ).
12
RESISTANCE
The coulomb of positive charge isolated from the coulomb of
negative charge above have an electric field of attraction between
them that totals 1 volt. If a path is provided along which charges
could move, the negative charges, excess electrons, would flow,
or conduct, to the positive charges, which are atoms deficient in
electrons. This charge flow, or current, would continue until both
locations became uncharged, or neutral. At that point, no voltage
of attraction would be present between the two neutrally charged
locations. However, if the one coulomb of charge buildup was
progressively replaced as the charges left and the electrons arriv-
ing at the positively charged area were removed as fast as they
arrived, the attraction, or voltage, between the two locations
would continue unchanged. This is what a battery does. It main-
tains a constant voltage between two locations despite current
flow. This is similar to a water pump that maintains pressure or
push to keep water current flowing (Figs. 5 and 6). A voltage
2
Figure 2. Forming a positive ion. A , Schematic of a balanced
source, such as a battery, has a cathode that replaces the elec- sodium ion.Valence electrons are the o u t e r m o r e easily mobile elec-
trons and an anode that removes the electrons that have flown to trons. B, Sodium atom n o w shown missing an electron. C , A resulting
it. Since electrons have negative charge, positive current is con- positively charged ion results. D.This ion is n o w electrically equivalent
sidered to flow in the opposite direction from the negative elec- t o I positive charge. (From Cromwell L, Arditti M.Weibell FJ, e t al:
tron flow, i.e., from anode to cathode (Fig. 6). In biologic Medical Instrumentation For Health C a r e . Englewood Cliffs, NJ,
systems, both positive and negative charges are present and Prentice-Hall, 1976, with permission.)
High potential (E ) H
Figure 3 . A schematic representation of t h e concept of

electrical potential, voltage, and its tendency t o move
Tendency of Tendency of Voltage (E) charge. (From Bergveld P: Electromedical Instrumentation:
positive charge negative charge = Electrical A Guide for Medical Personnel. N e w York, Cambridge
Potential University Press, 1980, with permission.)
Difference (E - E )
H L
Low potential (E ) L
but having a larger cross-sectional area can more readily accom- (symbolized as "R"). Written mathematically, this is I = E/R, or
modate greater charge flow. Such a pathway is said to have less rewritten: E = IR, which is called Ohm's law. This simple 2
resistance to current flow. equation becomes very useful in understanding how one should
Resistance is the description of how difficult it is for charge to expect certain circuits to behave, including those made by at-
move through a certain path. This is analogous to the resistance taching an electrophysiologic instrument to a living body. The
to water flow caused by the pipes in a water circuit. A narrow units of resistance are ohms, symbolized as "O." Rewriting from
pipe resists the flow of water more than a wider pipe (Fig. 7). A Ohm's law: R = E/I, or 1 Q. = 1 V/l A. Commonly used units of
very tight restriction in the water circuit may almost entirely de- resistance include kO, which is 10 O and MQ, which is 10 O.
3 6
termine the amount of flow for a given pump pressure in a water Charges accumulate on the amber and woolen cloth since
circuit (i.e., the resistance of the rest of the pipe is relatively they do not allow easy movement of charges, and thus can hold
negligible compared to the constriction). Even this flow, how- static charges more easily. They have very high resistance to
ever, can be increased by increasing the pump's pressure. Flow current flow and are called nonconductors or dielectrics.
can also be increased by opening the restriction (decreasing the Electrons are much more free to move in metals than in normal
resistance); thus, water flow is proportional to the pump pres- saline, where sodium and chloride ions transports the charges.
sure, but inversely proportional to resistance. Similarly, electri- Metal has very low resistance, and normal saline has higher re-
cal current flow (symbolized as ' T ' ) is proportional to voltage sistance, although both are considered good conductors.
(symbolized as "E"), but inversely proportional to resistance Conductors can vary greatly in their resistance properties (Table
1). Often the resistance of metals, such as copper or aluminum
Excess
wire, is so much lower than that of the biologic materials that it
is considered negligible, or having essentially zero ohms (indi-
cated by the straight lines in Fig. 6). 2 6
EXERCISE I
In Figure 6, the battery provides a constant voltage of 12 V
A across the resistance which is 2 O. How much current flows?
Table I. Resistivity of Various Biologic Materials

Material Resistivity (ohm-cm)
Copper 2 x lO" 6
Seawater 20
Cerebrospinal flid 64
Blood ISO
Spinal cord (longitudinal) 180
C o r t e x (5 kHz) 230
C o r t e x (5 Hz) 350
White matter (average) 1200
B o n e ( l MHz) 1800
Bone (low frequency) 16,000
W e t skull (low frequency, average) 20,000*
E Pure water 2 x I0 7
Active membrane (squid axon) 2 x I0 7
Figure 4 . Static electrical charge build-up and discharge in

nature. A, Electrically neutral cloud. B, Charge separation by updrafts. Passive membrane (squid axon) I0 9
C, Clouds separating from variable wind effects. D, Lightning, discharging Dry skull (low frequency) I0 13
of excess charges, between clouds. E, Discharging t o earth. (From

*This figure is very rough since skull resistivity is quite variable. From Nunez
Cromwell L, Arditti M,Weibell FJ, e t al: Medical Instrumentation For PL Electric Fields of the Brain. New York, Oxford University Press, 1981, p 80,
Health Care. Englewood Cliffs, NJ, Prentice-Hall, 1976, with permission.) with permission.
Appendix BASIC ELECTRICITY PRIMER — 101
P = Pump pressure
E = Voltage
= Potential Electrical
to move potential
h< water to move
current
+
)P
Water Pump Force and Voltage

Figure 5. Pump analogy applied t o a battery.The pump pushes water and acts as battery (voltage source), which pushes charges.
Answer: (Ohm's law) E = IR pole), the circuit is called a series circuit. Instead of the one re-
12 V = I x 2 O; I = 12 V/2 O = 6 A. Six amperes flow. sistor of Figure 6, two or more may be present (Fig. 8).
Each resistor adds to the circuit's overall resistance. Thus the
If the resistance is increased to 4 k Q , how much current resistance (R) from the voltage source's anode to it's cathode, is
flows? R, + R ; R = Rj + R . Likewise, the current flow will be E = IR,
2 2
E = IR; 12 V = I x 4000 Q.; I = 0.003 A or 3 mA. so E = I (R, + R ) and I = E/R; therefore, I = E/(R, + R ).
2 2
Series circuits are useful because certain predictions can b e

The voltage is now reduced to 4 mV; how much current made about the electrical potential occurring across each re-
flows? sister in the circuit. Since by Ohm's law:
E = IR; 4 mV= 0.004 V = I x 4000 Q. I = E = IR,
0.004 V/4000 Q = 0.000001 A or 1 pA. or E = I ( R , + R ), 2
therefore E = IR,+ IR . 2
This can be rewritten using Ohm's law (E = IR) as:

SERIES CIRCUIT
E = E, + E , 2
When there is only one pathway for current to flow between a where Ej is the voltage "drop" in electrical potential across R {
voltage source from anode (positive pole) to cathode (negative with I flowing, and E is the voltage " d r o p " across R . These
2 2
voltages are proportional to the resistors since the same current,

I, flows through both.
i
Thus, with:
E
f r 1=
R, + R,
and
+ : I R ,
-R,+R, R l
E "±- R
El Ri
Rj + R 2
Thus, E, is proportional to E attenuated by a constant ratio of

R,/(R, + R ).2
This can be used as a voltage divider since Rj and R, + R 2
can be chosen to divide E into any portion desired, noting that

" = actual direction charges Ri^Rj + R ) is always 1 or less. For example, if R = 1H and R?
(electrons) are flowing
2 x
= 9 Q thenR, + R = 10Q and E, = (1/10) E or l/10th of E. This

2
Figure 6. Electrical current flow.The battery creates a "push," of concept is often used to control or attenuate the amount of volt-
magnitude E, t o move positive charges from its cathode (negative port) age applied, for instance across a speaker to control the volume
t o its anode (positive port), which then flow down the electrical poten- of sound produced (Fig. 9).
tial gradient back t o the cathode by means of a resistor. Charge flow, o r
current (I), is considered t h e direction of flow of positive charges. EXERCISE 2 (FIG. 10)
However, the charges that most easily flow through metallic wires are
electrons, which have negative charge.The electrons are actually moving What is E ? 2
in the opposite direction t o what is defined as positive current flow. Answer: = (R /{R, + R })E; E2 = (2Q/{ 1Q + 2Q}) 9 V = 6 V.
2 2
102 — P A R T i FUNDAMENTAL PRINCIPLES
R = Orifice resistance R
- A / W -
W = Water flow
\ P = Pump
pressure
Water C u r r e n t (W) Circuit Electrical C u r r e n t Circuit
E
R
P = WR E = IR, O h m ' s Law
Figure 7. Pump maintains its push even with water current flowing, just as a battery maintains its voltage "push" despite charge flow (called elec-
trical current).The w a t e r pathways have negligible resistance compared t o that of the restricting orifice just as t h e wires (straight lines) of the
electrical circuit have negligible resistance compared to the resistor shown schematically.The amount of current is proportional t o the amount of
push and inversely proportional t o the resistance.This relationship results in Ohm's law.
f r -
fr-
Ri R, = 1 Q E,
+
E = 9V
E ~tr
; R = 2Q
2
Figure \ 0. Series circuit problem is shown with a battery of 9 V and

t w o resistors in series. W h a t is t h e c u r r e n t (I) flowing in t h e circuit
Figure 8. A series circuit is depicted (two resistors connected t o - and how much voltage drops across each resistor?
gether in series).
Potentiometer = Variable Resistor
" 1
-AA/VV
soon
Varying or
Voltage (~) Speaker
Source ^
Figure 11. Schematic for ground, o r ground symbol, is depicted.This

point marks a c o m m o n connection t o the earth, which is usually as-
sumed to be a relative zero potential.
Figure 9. Adjusting t h e voltage applied t o a speaker by means of a
variable resistor, potentiometer, is shown. As t h e potiometer's resis-
tance increases, m o r e of t h e applied voltage is dropped across this
larger resistor, the speaker resistance remaining constant, resulting in
reduced voltage across the speaker and a lower volume of sound pro-
duced.The "volume" control knob on a radio is the potentiometer in
this circuit.
Itotal
fr
f r -
1Q
+
+
E = 9V -±r
E = 9V ~±-
20.
Figure 13. A parallel circuit (resistors have been connected t o -
gether in parallel) is shown.
Figure 12. The circuit of Figure 10 with grounding shown.The point alternating electrical voltage to all appliances, lights, etc. (Fig.
marked with the ground symbol has now been conductively connected 14).
t o the earth.
EXERCISE 3 (FIG. 15)
What is E,?
What is I I , and I
b 2 total ?
Answer: E, = (Ri/{Ri + R }) E = 2
(1Q/{1Q + 2£2})9V = 3 V. What is R (which equals E/ I ai)?

total lol
Answer: I, = E/Rl = 10 V/10 Q = 1 A

I = E/R2 = 10 V/10 O = 1 A
2
GROUND Itotal = Ii + I = 2 A . 2
Rtotai = E / I = 10 V/2 A = 5 a or R,R /(R, + R ) =

total 2 2
Since voltage represents the attraction between charges at dif- (10O)(10Q)/(10Q + 10O) = 100 Q / 2 0 Q = 5 Q . 2
ferent locations, we can only define relative voltages. The earth

is considered to best represent a relative zero voltage and if this A useful circuit that incorporates features of both the parallel
point is present in a circuit it is often represented by a ground and series circuits is the balanced bridge (Fig. 16). If the ad- 2
symbol (Fig. 11). justable resistor, also known as a potentiometer (R ) is manip- pot
In Figure 10, if the battery's cathode was connected also to ulated so that the voltmeter (which measures voltage between
earth it would be drawn as shown in Figure 12. Without such a two points without allowing any current flow between them)
"grounding," a physically wired circuit could pick up stray reads zero, then an identical voltage drop has occurred across
static electricity and develop very high potentials, as occurs in R and the unknown resistor, R (i.e., both points must be at
pot x
thunder clouds, which could make it unsafe to touch. For safety the same potential for the voltmeter to read no voltage differ-
reasons, most circuits are not "isolated" but are usually con- ence between the two points). Current flow in R and R is pot x
nected to earth ground. equal since no current flows through the voltmeter and both
R]S have the same potential across them. This means that R , x
the unknown resistance, can be measured indirectly by reading

PARALLEL CIRCUITS or measuring the resistance needed to "balance" R from the x
potentiometer. The balanced bridge circuit is often used to de-

Resistors also can be connected to the circuit so that current termine an unknown resistance without having to simultane-
branches (Fig. 13). A circuit that applies the same voltage to ously measure the voltage across it and current through it.
each resistor so that the current flow branches is called a paral-
lel circuit. The general equation for the resulting resistance is;
2
E = E
Rtotai = i— 1 r ' ^ w m
*i =
E/R-i and I = E/R ,
total M + h
A 2 2
then: R E
K t o t a l
" (E/R,) + (E/R ) 2
or by eliminating E, which is common in both numerator and

denominator:
P ! RiR? Figure 14. Parallel connections in electrical power circuits.The circu-
*iota|- (1/R ) + (1/R ) ~ R, + R
1 2 2
lar voltage source symbol means a time-varying voltage, in this case 120
Since there are other paths in which the current may flow, V amplitude sinusoidal voltage, which is t h e form of electrical p o w e r
the overall resistance of a parallel circuit is always less than delivered to homes, also called AC (for alternating current). (Modified
that of any single branch's resistance in the circuit. Parallel cir- from Cromwell L,Arditti M.Weibell FJ, et al: Medical Instrumentation
cuits are used in homes in order to apply the same 120 volts of For Health Care. Englewood Cliffs, NJ, Prentice-Hall, 1976.)
Positive charge
Insulator
(air or thin
dielectric)
R S10Q
2
Symbol
Negative charge
Figure 17. Capacitor physical configuration and electrical schematic

symbol is drawn. (From Bergveld P: Electromedical Instrumentation: A
Guide for Medical Personnel. New York, Cambridge University Press,
Figure 15. Parallel circuit problem is depicted. W h a t is the total cur- 1980, with permission.)
rent flowing from t h e voltage source and and how much current flows
through each each resistor? What is the equivalent resistance of these
two parallel resistors?
proportional to Q/E and defined as C = Q/V, thus Q = CV. The
rate of change for both sides of this equation is AQ/At = C
CAPACITANCE AV/At. Substituting I = AQ/At, we have I = C AV/At, since C is
a constant property of the capacitor and it does not vary with
Positive and negative charges attract and do so more strongly time and AV/At = UC. This result means that for a constant cur-
when placed in close proximity to each other. A flat conductive rent flow across the capacitor, the voltage will progressively in-
surface such as a metal foil plate can be placed very close to an- crease. This makes sense if one remembers that no flow is
other foil plate by means of a nonconductive film (dielectric) so occurring across the dielectric, thus all current flow accumu-
that the two metal sheets do not touch each other. This arrange- lates as charge on the two plates. Despite attraction of the
ment allows strong attraction for charges of opposite polarity charges to their opposites across the dielectric, the accumula-
when placed on each sheet (Fig. 17). Since the like charges action of nearby like charges on the same plate causes increased
cumulating on each foil sheet repel each other, the larger the repulsion. This repulsive force lends a voltage push for charges
surface area (A) of the sheets, the more easily charges can be to leave the plates.-For a capacitor to build up charges, an in-
placed. The smaller the distance between the plates (d), the creasing voltage "push" to force the charges to flow onto the
more easily opposite charges can be placed on each sheet due to plates is required.
increased attraction to the opposite foil's charges. The ability When the voltage is first applied, no charge is on the capaci-
(or capacity) to store charge easily (needing lower voltage to tor and thus no voltage is present across it, thus in Fig. 20:1 = 9
push the same amount of charge onto the sheets) is called ca- mA. As the charge accumulates, the voltage across C steadily
pacitance. This is analogous to water stored behind a dam. If a rises until it equals 9 V. Then I equals 0 mA as current ceases to
pump can only push water so high (potential energy or by anal- flow. Thus, the current changes over time (Fig. 20).
ogy voltage level) to fill the reservoir behind the dam, then a With a very small Rj such as l O , the total charge movement
larger reservoir formed behind a dam implies a greater capacity will occur much faster and quit flowing more quickly. For a
of that reservoir (Fig.l 8). This is analogous to the the capacity higher resistance, the charge will take longer to accumulate due
of an electrical capacitor. The more coulombs of charge moved
onto the plates per volt, the greater the capacitance of an electri-
cal capacitor. Thus C is proportional to A/d, which is propor-
tional to Q/E, where " Q " symbolizes the charge in coulombs
and "E" is the voltage between the two conductive sheets. The 1
movement of charge per time (change in charge, AQ, per

change in time, At), AQ/At = I, is the current. Capacitance, C, is n<
Water Storage Dam and Capacitance
AP = (water capacitance) x W AVA C xI

XT At
Figure 18. The height of water behind a dam is proportional t o the

potential gravitational energy of that water. The amount of w a t e r
Figure 16. Balanced bridge circuit. Unknown resistance, R , can be x
stored at any given water height reflects the capacity of the reservoir
determined indirectly by adjusting the potentiometer, R , so that E
pot v
behind the dam. Just as with an electrical capacitor, the g r e a t e r t h e
goes to z e r o voltage. At this point both parallel circuits r e p r e s e n t amount of charge/water stored for a given potential electrical(volt-
equal voltage dividers, thus R = R . T h e potentiometer's resistance is
x pot
age)/gravitational(height) energy, the greater the capacitance/capacity.
often marked on its adjusting screw o r t h e potentiometer can be re- As charge/water is removed from the capacitor/reservoir, the poten-
moved from the circuit and its impedance measured directly. tial electrical(voltage)/gravitational(height) energy decreases.
R = 1 kQ ;
1 R 2 = 2 ka or R 2 = 1 kQ
C , = 0.1pF ; C « 0.1pF 2 or C = 0.2pF
2
Figure 19. Resistor/capacitor circuit is connected to a 9V battery.

T-j = R-JC-J T 2 — R2^2 time
to the lesser rate of charge flow. If the capacitor less readily ac- Figure 21. Current flow decay rates for different resistor/capacitor
cepts charge (small capacitance), then the voltage across it pairs for the Figure 19 circuit. N o t e the product of the resistance and
builds up more quickly and the current stops sooner. A higher capacitance determine how fast the current approaches zero.
capacitance allows more charges to accumulate before its volt-
age equals that of the voltage applied (similar to a large capacity without infinite current flow occurring. Thus capacitors can
reservoir accepting more water inflow before cresting the dam blunt or "filter" rapid voltage changes. For instance, given a
in Figure 18). How fast the capacitor charges is thus inversely square wave voltage source, the blunting of the rapid voltage
proportional to both R and C. The value RC = T, tau, is called changes of the square wave occur as shown in Figure 23.
the time constant of an R and C pair. The larger x is, the longer The voltage change across a large capacitor for a given
time a given resistor and capacitor pair takes to charge or dis- amount of charge shift is relatively small. At high frequencies
charge (Fig. 21). 1
there is less time for charge movement, even with high currents.
Should the voltage source be replaced by a simple wire, the At high frequencies only small voltage shifts occur across a ca-
charges on the capacitor will flow back through the R, and the pacitor despite possibly high alternating currents. This is similar
wire to the opposite plate where they are attracted, eventually to what occurs with a very small resistor. The capacitor's imped-
neutralizing both sides and the voltage across the capacitor will ance (resistance of a capacitor to current flow) at high frequen-
again return to zero resulting in zero current flow (Fig. 22). cies is low. Indeed, at an infinite frequency there would be no
Capacitors are useful because they can store energy. This 2
time to move charge, thus no voltage change and the impedance
energy is actually stored in the electric field of mutual attraction of the capacitor would approach zero. At low frequencies (e.g., 0
between the charges on the separated foil plates. A defibrillator Hz or direct current) such as the battery/resistor/capacitor circuit
is an example of a clinical device that uses capacitor energy. A in Figure 19, once the capacitor is charged sufficiently to equal
selected amount of energy (in joules) is stored on a capacitor the battery's voltage, no current flows. Therefore, the capacitor's
that is connected to the defibrillator paddles through a push impedance approaches infinite resistance to current flow at 0 Hz.
button switch, which can be discharged through a patient At low frequencies, a capacitor has high impedance. Thus, ca-
(equals R in Figure 22) at a relatively known rapid rate. 1
pacitor impedance varies inversely with frequency (Fig. 24). The
Since the capacitor's voltage depends on charge accumulation, frequency equivalent to Ohm's law is V = I Z , where Z , the c C C c
the voltage across a capacitor cannot change instantaneously impedance of the capacitor, is a function of frequency.
I(mA)
9 mA
0 mA 0 t = time
t = time
Figure 22. The c u r r e n t flow upon replacing t h e voltage source, E,
Figure 20. Current flow as charge accumulates (and voltage builds with a straight wire at time t = 0 seconds. Initially the current flows at
up) on the capacitor immediately after applying t h e 9 volts shown in - 9 mA since the current flows in an opposite direction from the pos-
the circuit of Figure 19. With no charge on t h e capacitor initially, t h e itive current flow when the capacitor was charging up t o equal the 9 V
greatest voltage difference is seen across R of Figure 19 and the great-
l voltage source. As c u r r e n t flows, less charges are on t h e capacitor
est current flow occurs. As the voltage increases across the capacitor, plates and the capacitor voltage falls resulting in less current through
less voltage is present across the resistor, R and less current flows,
h the resistor, eventually approaching zero current flow as the capacitor
eventually approaching zero current flow. discharges.
+1V
E=
-1V-
0.5V
•0.5V
Figure 23. The smoothing of t h e leading edges of a square wave

input t o a resistor/capacitor circuit as seen across the capacitor.
ALTERNATING CURRENT
Most bioelectric phenomena of interest vary with time and
their electric potentials are not static as with a battery. Examples Water Momentum and Inductance
include action potentials of nerves or muscles. Even in most
power circuits, such as electric or florescent lights, voltages Figure 25. Analogy of a water fountain t o an inductor.The momen-
vary with time (in the United States at 60 Hz and in Europe at tum of the water in motion will allow it t o continue t o rise and then
50 Hz). When signals vary with time they can do so slowly, in fall back t o t h e ground even if t h e push o r pump is suddenly turned
which case they are said to have low frequency content. An ex- off. An inductor will similarly cause electrical current t o continue t o
ample is the surface recording of the compound muscle action flow momentarily even when t h e voltage originally causing t h e flow is
potential, with typical rise times of 5 ms and 15-20 ms dura- turned off. T h e water's flow continues using up its kinetic energy,
tion. This means frequencies (the inverse of cycle time) are in whereas the electrical current flow uses up its magnetic energy t o mo-
the 5 0 - 2 0 0 Hz range, e.g., (1/5 ms) (1000ms/lsecond) = mentarily continue its flow. Inductance is the property of h o w much
200/second = 200Hz. Signals that change rapidly have higher magnetic energy is stored for a given current flow. This is similar t o
frequency content. An example is the 250 microsecond rise time how powerful is t h e pump creating the fountain's ejection velocity. A
required in a single-fiber electromyographic study, indicating faster pump generates m o r e kinetic energy for the same a m o u n t of
frequencies of 4 kHz or more, e.g., (1/250 ps) (10 ps/1 sec) =
6
water and a larger inductor results in m o r e magnetic energy for the
4,000/sec = 4 kHz. same amount of current flow.
Bioelectric signals usually have many different frequencies
that join to make up the complex waveforms seen. The pres- 1
ence of capacitance in a circuit such as the body can lead to dis- itself can slow the rise time (especially attenuating the high fre-
tortion of the actual signal due to its variable impedance at quencies) of single-fiber action potentials as observed when
different frequencies, thereby acting as a variable voltage di- recording at a distance from the actual fiber. Muscle tissue
vider, altering the relative magnitudes of the frequencies ob- offers various paths for current flow. The bilipid membranes of
served from the biologic signal of interest. Tissue capacitance the many adjacent muscle fibers can act as capacitors, each ac-
cepting a certain amount of current flow before local voltage
changes in response to that current flow due to charge accumu-
lation on these capacitors. This differs from the extracellular
pathways that are essentially resistive. Along these pathways
voltage varies directly and instantaneously with current flow in
Capacitor accord with Ohm's law (E = IR). The electrical model, or equiv-
Impedance alent^ acting electrical circuit, for a needle electrode five fibers
(Z ) from a single-muscle-fiber action potential being recorded
c
looks like Figure 2 6 . The crisp rapid rise of a single-fiber action
potential at the input of Figure 2 6 results in a slower rise at point
A. The current flow must accumulate on capacitor A before
voltage of point A changes. The voltage at point A then causes
f = frequency current to flow through resistor B, but the voltage at point B will
not rise as fast since current must accumulate on capacitor B
Figure 24. Variable capacitor impedance t o current flow at differing before the voltage at point B changes. Each additional fiber
frequencies. The capacitor has infinite resistance at zero h e r t z (no (equivalent R-C circuit) results in further slowing of the rise
direct current, D C , can flow across t h e capacitors insulating dielec- time. This delay or slowing of the rise time is one effect of
tric). At infinite frequency there is no time for charge movement t o tissue capacitance that increases with distance between the bio-
occur, so no voltage changes can occur. W i t h no voltage change re- logic signal source and measuring electrode in muscle tissue.
gardless of current magnitude, zero resistance is implied for the capac- Each resistor and capacitor represents the extracellular pathway
itor at infinite frequency. past the fiber and the capacitance of that fiber's local membrane,
—VvVV- -AAA/V- -A/vW- -AAA/V- -^vVvV-

+ +
-in
E
out
Figure 26. Electrical model of the behavior expected for five muscle fibers from the recorded active fiber. Each fiber's bilipid membrane serves
as a capacitor which can shunt current away from the recording electrode at E . T h e extracellular fluid has a discrete resistance for the path
out
past each fiber.This results in a series of resistor/capacitor circuits, one for each fiber.
respectively. Each resistor and capacitor couple is simply a du- will not allow instantaneous current change without an infinite
plication of the simplest form of a low-pass filter arrangement. voltage being applied. An example is unplugging a home
Having several low-pass filters in a row causes the higher fre- vacuum cleaner while it is running. One is attempting to instan-
quencies to be progressively attenuated. Through this model one taneously change the current flow to zero in a device that due to
can better appreciate why muscle tissue acts as a low-pass filter, its motor has high inductance. The result is that vacuum cleaner
especially with increasing distances, and why slower rise times inductor's magnetic field energy creates a high voltage to resist
result.3
the instantaneous decrease in current, resulting in a spark to the
wall socket (i.e., resulting in a slower decrement than instanta-
neous). A similar principle is used with an automobile engine's
INDUCTORS coil to cause a spark across the spark plugs.
In biologic tissue, three-dimensional current flow occurs. In a
Just as objects with mass tend to remain in motion (Newton's metallic wire current is very concentrated and relatively large
first law), electrical current flow tends to resist change. A water magnetic fields circle it that decline rapidly with distance (Fig.
pump creating a fountain cannot instantaneously stop the flow of 27). The current flow in a volume conductor such as biologic
current (the water already in motion will continue to rise and then tissue can be considered a collection of many single lines of
fall over several seconds after the pump has been turned off) (Fig. current flow. Any two adjacent lines of current flow have mag-
25). Similarly, the flow of electrical current resists instantaneous netic fields that cancel each other and add together outside of
change. This resistance to current change is called inductance both lines. With many adjacent lines of current in a three-di-
and is usually such a small effect in biologic systems that it is mensional volume conductor the cancellation becomes defuse
negligible. Inductances are, however, important for understanding so that a significant magnetic field exists only outside of all the
the principles of isolated preamplifiers and transformers. lines of current. This places the magnetic field very far from the
Electrical inductance provides a voltage kick analogous to the center of all of these contributing lines of current, which due to
kick of a fireman's hose with an on/off nozzle attached. When distance makes the resulting magnetic field very small. The
the nozzle is opened it kicks back as the water current resist its magnetic field strength is also small because current flow in bi-
initial flow. Rapidly closing the nozzle results in a kick in the ologic tissues is very small (microamperes as opposed to tens of
opposite direction. The resistance to water current flow changes amperes in the engine coil). These factors make the inductance
is due to the momentum of the water. The kinetic energy stored effects of normal biologic substrates at electrophysiologic fre-
in its flow provides the kick to the nozzle. In electrical circuits, quencies negligible. At high frequencies such as used in mag-
current flow builds up energy in a magnetic field that surrounds netically coupled short wave or microwave diathermy, the
the path of the current flow (Fig. 27). 2
body's inductance effect does become significant.
With a coil, greater magnetic field amplitudes can be Since inductors resist changes in current per change in time,
achieved with the same amount of current flow due to summa- inductance (L) is inversely proportional to AI/At. Large induc-
tion of the lhagnetic fields from each turn of the current path. tors allow less AI/At per voltage applied than small inductors.
This results in larger resistance to current change as energy Applying more voltage allows a more rapid current change, or
from the magnetic field must be built up or dissipated over time AI/At is proportional to E. Thus, AI/At = E/L.
before the current change can occur. Similar to a capacitor not At high frequencies of applied voltage across an inductance
allowing instantaneous changes in voltage across it, an inductor there is little time for the current to change and thus the alternating
Figure 27. A , T h e magnetic field lines encircle a current carrying

wire, progressively diminishing with distance, as shown by the iron fil-
ings density decreasing with distance from the wire. B,Two parallel
current paths result in opposite magnetic field (curved arrows) direc-
tions in the space between t h e m and thus cancellation of the mag-
netic field intensity in this region. T h e magnetic fields do align in
direction over a larger circle (thus weaker field due t o distance)
around both current lines. C,A coil allows multiple turns of a current
path t o summate t h e magnetic fields down t h e central axis of the
coil.This allows buildup of a larger magnetic field per amount of cur-
rent flow (i.e., more energy stored). Since this buildup is what is re-
sisted, the coil has a high inductance. A. B. c .
•AA/W
R
Varying
Voltage E Inductor, L
Source Coil rotates
Magnet
Figure 28. A simple resistor/inductor circuit.
Voltage generated
as long as coil
current magnitude remains small. At infinite frequencies, the rotates
impedance of an inductor approaches infinity. The body's small
inductance becomes significant at the high frequencies of short Figure 30. T h e dynamo, o r generator, is shown. By forcing t h e coil
wave diathermy, but not at the lower frequencies analyzed in t o r o t a t e in a constant magnetic field, t h e a m o u n t of magnetic field
electrodiagnostic evaluations. Applying 0 Hz voltage, no through t h e center of the coils will continually change.This induces a
change in current is being attempted after the constant DC cur- sinusoidal c u r r e n t that increases and decreases in an a t t e m p t t o
rents begin and thus the inductor's impedance is zero at 0 Hz. negate t h e flux changes through t h e coil.This is similar t o Newton's
Thus, inductor impedance varies directly with frequency. second law.The attempt t o place a magnetic energy field through t h e
Recognizing capacitor and inductor responses at zero and infi- coil is resisted by t h e build up of sufficient current t o create an equal
nite frequency helps one more easily understand their effects in and opposite magnetic field resulting in no net magnetic field build up
any particular circuit. in t h e coil.This requires energy that is provided by the force applied t o
An inductor/resistor series circuit also acts as a variable volt- spin t h e coil. (From Bergveld P: Electromedical Instrumentation: A
age divider over frequency due to the inductor's variable imped- Guide for Medical Personnel. N e w York, Cambridge University Press,
ance over frequency, similar to the variable voltage dividers 1980, with permission.)
over frequency one can produce with resistor/capacitor circuits
(Figs. 23 and 28). also serve to amplify or attenuate original signal according to
Since change in the stored magnetic field energy is what is the ratio of turns. This is how neighborhood transformers
resisted, one circuit can be connected to another by coupling reduce (or transform) high-voltage power lines of 10,000 volts
only this magnetic field. Figure 29 shows such a coupler, called to 120 volts AC for in home use.
a transformer. - Isolation transformers allow one circuit to
1 2
Electric generators, or dynamos, create current flow by me-
follow changes in another without being directly wired together. chanically forcing a coil past a stationary magnet (Fig. 30). As
This helps to reduce some sources of electrical noise. the magnetic field is encountered, the generator's coil has cur-
In Figure 29, the current I is greater than I, proportionally to
2 rents induced that try to oppose the changing magnetic field
the number of turns since two full turns of I magnetic field2 strength. The current will flow in one direction as the magnetic
must cancel three full turns of Ij magnetic field to maintain no field increase is being resisted, and then as the magnetic field
magnetic field change. The current required is 3/2 times larger, declines with movement of the coil's center away from the
i.e., I = 1.5 I,. Thus transformers can both allow indirect or
2 magnet, a current will be induced in the other direction trying to
electrically isolated interconnections between circuits, and can maintain the magnetic field. This is why spinning circular gen-
erators (Fig. 30) intrinsically result in alternating current (AC)
as their primary output. 1
FILTERS
Filters are often used to restrict the frequency content of the
signals evaluated. The desire is not to eliminate the biologic
signal's frequency, since this would distort the true potential's
waveform, but instead to minimize the effect of noise that may
t T
distort the true biologic potential's waveform. Noise is any po-
t tential measured that does not represent the actual biologic
Coil 2 event of interest. Unfortunately, noise is universally present but
Coil 1 Iron core
Figure 29. Coupling circuits through an inductor's magnetic fields, can be attenuated by proper use of filtering without distorting
t h e transformer, is shown.The iron c o r e concentrates t h e magnetic the potential of interest.
field so that almost all of the magnetic field generated by t h e coil la- All resistances generate a certain amount of noise due to the
beled L| tries t o p e n e t r a t e the central axis of coil L .This results in
2 thermal agitation of their atoms and electron clouds. This
c u r r e n t building up in coil L that tries t o resist the magnetic field
2 noise power is distributed uniformly over all frequencies and
being generated by c u r r e n t I, flowing t h r o u g h coil L T h e resulting
h is proportional to the absolute temperature of the resistor. The
current l is said t o be induced and is proportional t o l by the ratio of
2 ( voltage amplitude of this noise ( E ) is defined by the equa-
N
the number of turns of the second coil t o t h e first (since the magnetic tion: E = 8KTBR, where K is Boltzmann's constant (1.38 x
N
2
fields of each try t o be equal and are proportional t o the t h e current 10 joules/°Kelvin), T is the absolute temperature in °K, B is
23
flowing and number of turns). the recording bandwidth, and R is the resistance of the biologic
or physical system being measured. Note that the larger the

resistance or bandwidth, the larger this noise voltage. 1.0-
Lowering resistance helps to reduce this thermal electric
noise. Lowering resistance also helps to decrease noise volt-
age from radiated sources due to the limited power transmit-
ted from such sources (machines, radio waves, 60-Hz
transmission lines, etc.). Since power = I x E, and by Ohm's
law E = IR or I = E/R, power = E /R. Thus, for the same noise
2
power at a higher resistance the noise voltage will be larger.

The same noise power applied to a lower resistance results in
lower noise voltages. Noise power considerations are another
reason to try to reduce electrode impedance as much as possi- F H Frequency, H z
ble. The thermal generation of noise offers a means to exam-
ine the lowest possible theoretical amount of noise one can Figure 31. Idealized and real bandpass filter frequency response. N o
expect. At body temperature of 310°K (37°C) with a resis- attenuation of t h e signals t o be analyzed should occur over the fre-
tance of 20 kQ and bandwidth of 20 kHz one expects, by the quencies of interest- However, outside these frequencies, total attenua-
above equation, E = 2.6 pV. However, with 2 k Q and re-
N
tion is desired as t h e s e signals r e p r e s e n t only noise (undesired
stricting the bandwidth to 2 kHz, thermal noise is reduced to contaminating electrical signals that do n o t represent any part of t h e
0.26 pV. Detection of signals smaller than this is not possible actual biologic signal of interest).
without averaging techniques because of this polluting ther-
mal noise. Of course, actual noise will be higher than these higher- or lower-frequency noise contributions. An ideal filter
theoretical minimums. The benefit of reduction in bandwidth allows no attenuation of the signals in the frequency range of
to reducing overall noise is one reason sensory nerve conduc- interest but above or below those frequencies totally attenuates
tion studies (with very small amplitudes) are performed at all signals, which are all presumed to be noise (Fig. 31). In prac-
more restricted bandwidths than motor nerve conduction tice a "roll off' in frequency response occurs so that the cut-off
studies or needle electromyographic examinations (which frequencies, F (low frequency) and F (high frequency), are
L H
have larger amplitudes and thus are less affected by these defined as the frequency point where V / V = 0.707. Thus at
out in
small electrical noise factors). F and F , almost 3 0 % attenuation is occurring. One should
L H
One can take advantage of the limited range of frequency make sure that F and F are set well outside the frequencies of
L H
content of many electrophysiologic phenomena to reduce interest.
High Pass Filter

or
-AAA/V
+ R +
E i n
+
© Eout
Frequency, Hz f, = — — or — —
2itRC 2TCRL
Low Pass Filter

or
"vaafl&r
L I +
Eout Ein R< E
Frequency, Hz
2TCRC 2TCRL
Figure 32. High-pass (allow high frequencies through unattenuated) and low-pass (allow low frequencies through unattenuated) filters. A ,
Resistor/capacitor filter circuits. B, Inductor/resistor filter circuits. C, Frequency response of these equivalent circuits. N o t e t h e opposite locations
for capacitor and inductors relative t o the resistor for each type.This is due t o their opposite high- versus low-frequency impedance characteris-
tics. F and F define the lowest and highest frequencies that pass relatively unattenuated in the high-pass and low-pass filters, respectively.
L H
The concepts of the capacitor and inductor allow us to see DIGITAL FILTERING
how filters can be made. Since capacitors and inductors have
variable impedance over frequencies, they can be used with a The majority of electrophysiologic instruments now use digi-
resistor to make either high-pass or low-pass filters depending tal processing, including sampling the real-time (or analog)
on the circuit arrangement (Fig. 32). 1
signal at set intervals (which are points in time that occur at the
For the high-pass filter (which blocks the lower frequencies inverse of the sampling rate, i.e., 10 kHz sampling rates means
from being transferred through the circuit), the capacitor pro- values are taken each 0.1 ms) and digital, i.e., computerized
vides near-infinite impedance at the low frequencies, resulting mathematical filtering. The analog, or continuous real-time fil-
in a very attenuated output. However, at high frequencies, ters described above suffer from phase distortion effects so that
almost no impedance is seen across the capacitor and negligible the frequencies near the cut-off settings of F or F are both at-
L H
attenuation occurs. The impedance of the capacitor equals the tenuated (typically to 0.707 of their value without the filter) and
impedance of the resistor at F . An attenuation of 0.5 would
L phase-delayed or phase-advanced, which results in distortion of
therefore be expected, since half the voltage drop would occur these signals near the limits of our cut-offs. If the biologic
over the resistor and half across the capacitor. Due to phase ef- signal of interest has frequencies near these cut-offs, the filter
fects (the voltage across the resistor varies directly and instanta- will distort the resulting signal transferred from V to V .in out
neously with current, whereas the voltage across the capacitor is Digital filtering takes advantage of a digital processor's ability
delayed relative to current flow since charge must accumulate to store data, which can mean that prediction about future
on the capacitor before its voltage changes), the attenuation at values can be made (since time zero can be set arbitrarily at any
frequency F is actually 0.707. Considerations of capacitor and
L stored data value). A more perfect decreased phase distortion
inductor impedances near zero and near infinite frequencies filter with sharper roll-off can be constructed using powerful
help one to deduce the resulting attenuation expected over fre- mathematical digital processing with the advantages of such
quency (called the transfer function or frequency response of predictive properties.
V /V- ^ 1
One major pitfall to digital filtering is that only frequencies
v
out' v
mJ- below 1/2 the sampling rate, which is called the Nyqvist fre-
quency, can be theoretically reconstructed. Any frequency con-
BANDPASS FILTERS tent above 1/2 the sampling rate gets processed as though it
occurs at a frequency equal to the difference between true fre-
The filter with the response comparable to that displayed quency coming into the device and some multiple of the sam-
Figure 33 is called a band-pass filter. It is constructed by seri- pling frequency, always resulting in an interpreted frequency
ally connecting a low-pass filter with its highest-frequency cut between zero and the Nyqvist frequency. This is similar to the
off, F , to a high-pass filter with lowest cut off frequency, F ;
H L strobe effect where an accelerating rotating wheel appears to
where F is greater than F and the range from F to F is the
H L L H speed up correctly, but then falsely appears to slow down and
band pass or frequency range allowed to pass through (Fig. 33). then reverse directions as the rotating frequency approaches and
Most electrophysiologic studies are performed with band-pass exceeds 1/2 the strobe light's (sampling) frequency. As the
filters in which both F and F are selectable.
L H speed exceeds the strobe's sampling rate, it again appears to
speed up, but again slows and reverses as the speed exceeds one
and one half the strobe rate. This process repeats with increas-
ing rotational speed or cycles per second. Thus one can only in-
terpret the true speed for zero to 1/2 sampling rate range of
HPF
frequencies. Since low frequencies are of interest in electro-
physiologic studies, only the range from zero to the Nyqvist fre-
quency (1/2 the sampling rate) is resolvable and higher
frequencies must be first analog filtered out before the digital
Frequency, Hz
filtering computer processing occurs. The processing error that
occurs for frequencies presented above the Nyqvist frequency is
1.0 called aliasing and is a problem inherent to sampling. For-
.707-
tunately, modern digital equipment allows very fast sampling so
LPF the range of biologic frequencies of interest is usually much
Eout/Ein
below the Nyquist frequency for most devices. However, analog
low-pass prefiltering with cut-off below the Nyqvist frequency
to eliminate all frequency content at or above 1/2 the sampling
F Frequency. Hz
rate before sampling occurs is still necessary. If this limitation is
H
not observed, severe distortion of the actual signal can occur as

higher frequencies become inteipreted as lower-frequency con-
tent of the signal.
6
Another limitation to digital filtering has to do with the fre-

quency domain mathematics inherent to the process. To com-5
pletely transform the incoming real-time analog signal into the

Frequency. Hz full frequency domain it must be observed for all time, past and
future. Since we do not have that much time or knowledge, we
Figure 33. Band-pass filters. A low-pass filter (LPF) in series with a compromise by using a segment of time called a window, which
high-pass filter (HPF) result in the band-pass filter (if F < F ; other-
L H is the time required for one trace to sweep across the electrodi-
wise all frequencies would be attenuated). agnostic instrument's video screen. It is then assumed that this
Appendix BASIC ELECTRICITY PRIMER —
trace repeats and has repeated forever. This means the most fun- ORIGINAL RECONSTRUCTED
SQUARE WAVE SQUARE WAVE
damental frequency within this analog signal segment is 1/T,
where T is the time duration, or time window, we have chosen.
The resulting analysis and computations result in creating J } 8 . 9 5 % of A
values only at frequencies that are multiples of this fundamental
frequency. This limits how accurately we can reproduce the
analog signal after applying the desired digital filtering.
Additionally, due to discontinuities at the edges (where the be-
ginning of the trace fails to exactly match the ending position
and rate of change of the end of the trace; remember we have
assumed it is repeating forever), high frequencies are artifactu-
ally added. The sharp transition of the discontinuity at the
Stim. Lt. Median Nerve
edges, with its zero rise time, creates the theoretical need for in-
ANALOG DIGITAL
finite frequency in the frequency domain, which computation-
ally cannot be achieved, leading to an artifact called the Gibbs 1503000Hz 150 3 0 0 0 H z
phenomenon (see Fig. 34). To minimize this phenomen occa-

5
sionally the windows are amplitude-tapered at the edge, i.e., the

actual analog signal is distorted by making it approach zero at 1>
the beginning and end of the trace epoch to better match at the
edges, but this adds its own distortion to the signal. Although 5
digital filtering can eliminate the phase and amplitude distor- E.P.-Sh 2
tions of analog filters, it can create other unexpected errors for

the unwary practitioner. Appreciating these factors can help one
to recognize possible sources of unexpected artifacts. i i i i i i i i 1
» • * *
Figure 3 4 . The original input signal of a square wave is transformed

ELECTRONIC DEVICES t o t h e frequency domain by computer mathematical processing and
then directly transformed back without any o t h e r processing t o t h e
Additional circuit elements exist in electronic devices be-
real-time domain.The limitations of digital filter processing (in this
sides capacitors, resistors, and inductors. All electronic de-
case, a filter that passes all frequencies w i t h o u t any attenuation) r e -
vices, however, can be simplified to these, plus switches,
sults in an overshoot (Gibbs phenomenon) of about 9% of the ampli-
amplifiers, and diodes. Understanding the basics of these ele-
tude of t h e discontinuities at t h e edges of t h e square wave.The lower
ments can help one appreciate how larger electronic devices
traces show an analog and digitally filtered time domain output of a
perform their tasks.
compound action potential recorded at Erb's point. N o t e the artifac-
tually added higher-frequency oscillations found in the digitally filtered
SWITCHES response. (From Maccabee PJ, Hassan NF: AAEM Minimonograph # 3 9 :
Digital Filtering: Basic Concepts and Application t o Evoked Potentials.
Electronics switches are two-state devices. These can be as
Muscle Nerve 1992; 15:865-875, with permission.)
simple as a light switch (one state is closed, i.e., two wires
touch each other mechanically and have essentially zero im-
pedance; the other state is open, i.e., two wires are separated so sensitivity is used on an electrophysiologic instrument, one can
that there is infinite impedance). Multiple setting rotary often see this imperfection as two states. The amplifier satu-
switches are simply a cluster of the simple two-state switches rates up or down with only occasional rapid transition between
described above. Two states, however, can also be produced by these two states, resulting in a random square wave configura-
using the imperfections of practical amplifiers. If excessive tion seen on the screen. This property of "saturating" between
Figure 3 5 . The diode (D), o r rectifier is dia- Time

grammed.This unique device has essentially no resis- Diode, D
tance for o n e direction of current flow but
+
near-infinite resistance in t h e opposite direction.
C u r r e n t flow is allowed in t h e direction of t h e trian-
E
out
gular arrow head but blocked (represented by t h e
A A A
vertical line) from flowing into t h e device from t h e 1.0-
opposite direction.
E ut 0¬
0
Time
-1.0-
two states is purposefully used in digital computers where one AMPLIFIERS

state represents a " 0 " and the other a " 1 " for the binary mathe-
matical processing used in computers. The rapid transitions Amplifiers ideally reproduce the input signal multiplied by a
possible with electronic switches allow very fast processing of factor, called the gain. The sensitivity usually describes the
2
this binary information. size of signals placed at the input, which after amplification
result in a deviation of one division on the viewing screen. The
DIODES ideal amplifier has infinite input impedance, so that no current
is clrawn from the input voltage source (which might distort the
Diodes are specialized resistors that have the unique property signal as compared to what exists without the amplifier at-
of allowing relatively free electrical charge flow in one direc- tached). It produces no distortion, thus it must be constant over
tion (i.e., near zero resistance) but blocking charge flow in the all input voltage ranges and frequencies, and must not add any
opposite direction (i.e., infinite resistance). These devices are noise. However, no amplifiers are ideal. All have input imped-
also called rectifiers since they can help present only one polar- ance and frequency ranges of less than infinite hertz with lim-
ity of the information present (Fig. 35). 2
ited voltage input range, and all add some noise to the signal.
An interesting variant of the diode has a third port that allows With such limitations, one may ask: Why use amplifiers? The
the infinite resistance to be reduced by variable degrees propor- signals pertinent to electrodiagnostic testing are often very
tional to a very small current applied into this port. This device, small, in the pV and mV range. The voltage necessary to oper-
called a transistor, can be placed in circuits that use this prop- ate the instrument's speakers is on the order of volts and for a
erty to amplify the third port's current. The transistor is one of CRT display screen, thousands of volts. Thus, the signal's volt-
the simplest types of amplifiers and has revolutionized both dig- age must be amplified. Amplification, especially by means of a
ital and analog signal processing. 1
preamplifier placed close to the patient, can also help reduce
overall noise (Fig. 36), even though it adds at least some noise
to the amplified signal that is not included in the illustration
below. The benefits of preamplification can be illustrated as fol-
Amplifier with gain of 100
lows: V = biologic signal of interest, V
B N C= noise added
E N C = 10pV through the cabling from the subject to the instrument. If con-
nected directly to the instrument without preamplification the
0 |HK>-^ voltage output = gain x ( V + V ) = gain x V + gain x V .
B N C B N C
With preamplification the voltage output = (gain x V ) + V .

B N C
E
B 0 E I N This means the ratio of the desired signal to the noise (also
= 10|xV called the signal-to-noise ratio) is improved by a factor of the
preamplification gain.
Signal to Noise = DIFFERENTIAL AMPLIFIERS

S/N = V x100 / V x100
B NC 1 mV/1 mV = 1.0
Amplifiers increase the input signal by a selectable gain. The
input signal is often a voltage that can only be defined between
two points. Differential amplifiers allow any two points to be
E = 10LIV
chosen and does not require these to be near the voltage selected
N C
as the relative zero point, or ground, for the circuit used to make
— 1(X> 0 the amplifier. This has distinct advantages in electrophysiologic
studies. Often 60 Hz noise is very prevalent and will be ampli-
fied, possibly swamping the signal of interest. Differential am-
EB <~) E I N plification allows bipolar recordings (where the electrodes are
= 10jiV close together), which cancel out the 60 Hz noise, since it is pre- .
sent relatively equally at both electrodes and, therefore, does |
not represent a voltage difference between these two electrodes.
Signal to Noise = Ideally, any voltage identically present at both electrodes will
S/N = V x100/V = 1mV/1pV= 100 not be amplified, since only the difference between the two
B NC
points is to be multiplied. In practice, differential amplifiers are
Figure 36. The effect of preamplification (using amplifiers before the not perfect and some of this common mode (present at both
signal traverses a long and potentially noisy electrical cable) on im- inputs) voltage appears at the output. A measure of how well
proving t h e signal-to-noise ratio (S/N, a measure of the quality of the such common mode voltages are rejected or not amplified is the
desired signal versus t h e amount of contaminating electrical noise). common mode rejection ratio (CMRR). The CMRR is the gain
W i t h o u t preamplification (top), 1 0 pV of cable noise ( E ) , is added t o NC
of the differential signal divided by the gain of the common
the measured 1 0 uV signal, resulting in just as much noise as signal, o r mode signal and is usually greater than 100,000. This means if
a terrible S/N of 1.0. With preamplification (bottom), t h e biologic the differential amplifier inputs are 100.2 pV and 100.3 pV, re-
signal of interest (E ) is first enlarged by t h e gain of 1 0 0 before t h e
B
spectively, one would ideally expect 20 mV output if the gain is
cable's noise is added, resulting in a much b e t t e r S/N of 1 0 0 . Even if the 200,000 (200,000 x [100.3 pV - 100.2 pV]). If the common
amplifier adds I % noise (which would be very high) the S/N with pre- mode gain is 2, then the CMRR is 100,000, and the actual
amplification would 5 0 [(EB X I 0 0 ) / ( E + l% {E x 100}) = 1 , 0 0 0 / ( 1 0
n c B
output will be 20 mV plus the common mode amplified by its
+ 10) = 5 0 ] , as compared to 0 . 9 8 with later amplification (top) [S/N = gain of 2, which is 2 x 100.2 pV or 0.2 mV, which results in
(E x I 0 0 ) / ( I 0 0 E
B N C + l%{E NC + E } 100) = 1,000/(1,000 + 2 0 ) = 0.98].
B
20.2 mV. This results in only a 1% error despite a 1000-fold
plates. These plates act as capacitors between which large elec-

tric fields are generated. These fields pull the electron beam at a
perpendicular angle to its direction of travel. This causes the
Light spot
beam to strike at different points on the CRT screen. The time
Electron Gun base, lateral plates, has a ramp voltage applied that slowly and
linearly (according to the sweep speed) deflects the beam from
left to right. The lateral plates' voltage then rapidly switches
back to that necessary to again position the spot for a new sweep
Grid beginning from the left. The vertical plates provide the signal's
amplitude information by variably deflecting the beam up or
Anode
down according to the amplified input signal (Fig. 37). 1
Horizontal
deflection Vertical Screen DIGITAL PROCESSORS (COMPUTERS)
deflection
Most modern electrodiagnostic devices have digital (com-
Figure 3 7 . Schematic representation of a cathode ray tube (CRT). puter) processing capabilities. This means a network of rapidly
(From Bergveld P: Electromedical Instrumentation: A Guide for acting electronic switches process information that has been en-
Medical Personnel. N e w York, Cambridge University Press, 1980, with coded into binary numbers, i.e., a series of zeros and ones. This
permission.) processing can be "hard-wired" by built-in circuits, but more
often is flexible and controlled by software that can be varied.
higher common mode input compared to the differential volt- Software authors try to anticipate all possibilities but often fail
age. Therefore, large common mode 60 Hz signals can be sub- to adequately account for all scenarios. Thus, any automated re-
stantially eliminated with an amplifier that has excellent sults should be checked for consistency with the operator's
CMRR. expert opinion. This includes automated placement of latency
and amplitude markers, as well as results based upon these,
CATHODE RAY TUBE such as report findings and nerve conduction velocity calcula-
tions. Due to the speed of information processing available dig-
Even the older analog electrophysiologic instruments re- itally, many tedious or otherwise impossible tasks can now be
quired electronic switching. This is necessary to rapidly return performed. This includes quantitative analysis of multiple
the trace from the right to the left side of the cathode ray tube's waveforms such as quantitative motor unit analysis and presen-
(CRT's) face. The CRT is basically a vacuum tube containing an tation of the frequency content of the waveforms in the semi-
electrophosphorescent coated face (which visually glows where continuous manner called power spectrum analysis.
hit by an electron beam), an electron gun (which is basically a Computers depend on sampling and digitization of real-time
very hot filament that allows electrons to "boil off," which are signals for analysis. Besides the limitations of sampling, i.e.,
focused to impact the CRT face at one spot), and deflection only discreet interval time information, digital processors allow
Sample Sample & Hold • Digitize to - • D/A, Digital to Low Pass
4 Binary Code Analog Conversion Filter Smoothing

of Output
1st digit
0 = positive
1 = negative
Time Amplitude
0.0 10000
'BS VBD
0.1 10010
J
0.2 10010
0.3 10000
0.4 00011
0.5 00001 n
Time Time Time 0.6 00011 I J Time Time
0.7 10001
0.8 10101
0.1ms 0.9 11010
1.0 11100
1.1 11001•
1.2 10011
1.3 10001
1.4 10000
1.5 10000
Figure 38. Digital processing of a real-time (analog) signal is shown.The signal is sampled and held until t h e computers' cycle time allows the dis-
crete value for that particular time t o be recorded. It is then encoded into binary numbers that digital processors can manipulate. If sampled fast
enough, the real-time signal can be reproduced by D/A conversion and smoothing with low-pass filtering, although time-delayed. Since this delay
can be arbitrarily long, digital processors make excellent delay lines for trigger and delay operations o r data storage.
only certain discreet values for amplitudes, usually over a 2 8

predict the electrical behavior of biologic tissues. Electro-
(256) to 2 (64,000) range of values. This is due to the neces-
1 6
physiologic studies inherently require appreciation of the electri-
sary binary encoding. The process of turning a continuous real- cal events under study. Analyzing such events in terms of simple
time signal into a series of binary numbers that a computer can circuit elements can help solidify one's understanding of the
analyze requires an analog-to-digital converter, or A/D (pro- meaning of the recorded bioelectrical potentials and possible
nounced "A to D"). This device samples the real-time signal at electrophysiologic pitfalls involved in interpreting such signals.
discrete intervals and rounds its magnitude off to the nearest
digital value (Fig. 38). An inverse device, the D/A (pronounced
"D to A") or digital-to-analog converter, creates analog stair- REFERENCES
step values that change at each sample interval. This stair-step
output can be smoothed by analog filters, usually low-pass fil- 1. Berg veld P: Electromedical Instrumentation: A Guide for Medical Personnel. New
York, Cambridge University Press, 1980.
ters, to look very similar to the original input if no other pro- 2. Cromwell L, Arditti M, Weibell FJ, et al: Medical Instrumentation For Health
cessing was performed on the signal digitally (Fig. 38). 6
Care. Englewood Cliffs, NJ, Prentice-Hall, 1976.
3. de Weerd JPC: Volume conduction and electromyography. In Notermans SLH:
Current Practice of Clinical Electromyography. New York, Elsevier, 1984, pp
9-28.
CONCLUSION 4. Geddes LA, Baker LE: Principles of Applied Biomedical Instrumentation, 3rd ed.
New York, John Wiley & Sons, 1989.
Basic conceptualizations of electricity and the electrical cir- 5. Maccabee PJ, Hassan NF: AAEM Minimonograph #39: Digital Filtering: Basic
Concepts and Application to Evoked Potentials. Muscle Nerve 1992;15:865-875.
cuit elements like resistors, capacitors, inductors, rectifiers, am- 6. Nunez PL: Electric Fields of the Brain. New York, Oxford University Press, 1981,
plifiers, and switches can be powerful tools to understand and p80.
Chapter 4
Peripheral Nervous System's

Reaction t o Injury
Anthony A. Amato, M.D.
CHAPTER OUTLINE
Severe Neural Injury Model of Action Potential Conduction Slowing

Wallerian Degeneration • Neural Regeneration • Segmental and Blockade
Demyelination • Nerve Injury Classification • Neurophysiology Normal Nerve • Structural Aspects • Electrical Aspects
Correlates of Nerve Injury * Dynamic Electrophysiologic • Anatomic/Electrical Aspects of Demyelination
Observations of Wallerian Degeneration • Clinical Correlation • C o m p u t e r Modeling of Myelin Loss • Clinical
Correlation
Minimal Neural Injury
Temporary Neural Ischemia
Chronic Nerve Compression
Intermediate Neural Injury Toxic Degeneration • Extensive Segmental (Generalized)
Electrophysiologic Findings • Anatomic Findings Demyelination * Motor and Sensory Neuronopathy
Neural insult can be conceptualized as occurring over local- of injury, the axon itself is damaged, and the entire nerve (axon
ized individual nerve segments (focal process), or affecting ex- plus myelin sheath) undergoes disintegration distal to the injury
tensive portions of the peripheral nervous system (generalized and must be completely regenerated prior to the return of func-
process). Localized neural insults can produce three broad cate- tion. A more generalized neural injury can produce either of the
gories of nerve injury (Table 4-1). latter two types of nerve damage but may be more prone to pro-
1. Minimal neural insult: rapidly reversible conduction ducing severe axonal loss or generalized demyelination as op-
block with focal action potential propagation failure secondary posed to a focal loss of conduction because of its generalized
to short periods of ischemia. There is the possibility of nerve nature.
conduction slowing if incomplete neural blockade is present Although there are a multitude of specific entities that can
with sparing of the relatively slower conducting fibers, i.e., pref- affect the anatomic and physiologic status of peripheral nerves,
erential blockade of the fastest-conducting fibers. the nerves' response to an insult is limited (Table 4-1). For ex-
2. Intermediate neural injury: failure or slowing of action ample, a traumatic injury may disrupt the structural integrity of
potential propagation secondary to focal demyelination without an entire peripheral nerve whereby the axon and surrounding
axonal damage producing a prolonged conduction block and re- myelin demonstrate predictable changes (see below). On the
duced nerve conduction velocity (NCV). other hand, a toxic substance may preferentially affect just the
3. Severe neural insult: action potential propagation failure axon with resulting denervation of the muscle fibers and the
with axonal damage, i.e., Wallerian degeneration. supplied sensory end organs. A few hereditary and acquired
In the first two designations of action potential propagation disorders specifically affect the myelin enveloping the periph-
failure, there is a temporary loss of neural conduction across a eral nerves resulting in slowing of neural conduction. Other
focal lesion while the axon's structural integrity is preserved. In hereditary neuropathies and most acquired neuropathies are as-
this situation, it is possible for the enveloping myelin sheath to sociated with a primary axonopathy. Finally, there are meta-
be spared or disrupted, but with restoration of blood flow or bolic diseases that can disrupt both the axon and myelin,
myelin recovery, nerve conduction is restored. In the third type producing a combined demyelinating and axonal injury.
115
Table 4 - 1 . PERIPHERAL NERVOUS SYSTEM RESPONSE Table 4 - 2 . Pathological Basis of Altered Action Potential
TO INSULT Propagation
Insult Terminology Etiology Electrophysiology I. Axonal Injury
A. Wallerian Degeneration
Minimal Rapidly reversible Ischemia, Focal conduction 1. Focal crush or transection
conduction block mild compres- block 2. Disintegration and removal of axon and myelin distal to
injury
sion, antibody- Possible conduc-
3. Alteration of neural properties proximal to injury
mediated tion slowing 4. Reaction of cell body to insult (swelling & chromatolysis)
channelopathy B. Toxic Degeneration
1. Generalized insult to peripheral nervous system
Intermediate Prolonged conduc- Focal demyelin- Focal conduction 2. Similar responses noted for Wallerian degeneration
tion block ation, antibody- block and
II. Neural Repair
mediated slowing
A. Successful Regeneration
channelopathy
1. Regrowth of axon across injury site to appropriate end
Severe Wallerian Loss of axon Loss of conduc- organ
degeneration and myelin tion 2. Remyelination (shorter and thinly myelinated internodes)
3. Action potential propagation returns
sheath
4. Return of normal NCV proximal to injury
5. Improvement of distal NCV to less than 80-90% pre-injury
value
Once the peripheral nerve's anatomic structure has been dis- 6. Return of cell body to normal appearance
rupted, there are a number of processes that ensue depending 7. Clinical function returns depending upon severity of injury
B. Unsuccessful Regeneration
upon the severity of the insult (Table 4-2). A temporary reduc- 1. Failure of axon to reach appropriate end-organ
tion in blood flow to a focal region of nerve typically does not 2. Shrinkage of endoneurial tube distal to injury
result in any alteration of the nerve's structural elements. Nerve 3. Reduced NCV proximal to injury
compression to various degrees or actual severance of a nerve, 4. Possible death of cell body
5. Unsatisfactory clinical function (altered sensation, pain,
however, does generate a response that is directed at anatomic weakness)
repair of the neural damage. This repair process may be simple III. Myelin Insult
when a small segment of myelin must be removed and replaced. A. Focal (segmental, paranodal) demyelination
A significant insult such as a crush or transection of nerve re- 1. Severity of compression insufficient to injure axon
2. Focal segment of myelin injured
quires removal and reconstitution of the nerve distal to the 3. Loss of myelin integrity results in action potential failure
injury. (Conduction Block) and/or NCV slowing
4. NCV proximal to lesion may be slowed if recorded distal
to insult because of fast fiber blockade through site of
lesion
SEVERE NEURAL INJURY 5. Clinically weakness and altered sensation results with
conduction block
For the purposes of discussion, we will first consider a pro- 6. Etiology is usually acute or chronic compression/entrap-
found insult to the nervous system requiring removal and refor- ment
B. Extensive segmental (generalized) demyelination
mation of the nerve's damaged portion. This example allows us 1. Demyelinating neuropathies
to better appreciate the intricate processes developed by the body 2. Yield similar clinical findings with action potential blockade
to reconstitute its injured neural pathways. Additionally, it is im- or slowing
portant for the practitioner to understand the anatomic alterations IV. Myelin Repair
on the cellular level induced in the peripheral nervous system by A. Remyelination
1. Focal segment of injured myelin removed
trauma at various stages of injury and repair to better appreciate 2. Schwann cell proliferation/reformation of new intercalated
the associated electrophysiologic responses (Table 4-2). myelin segment
3. Blocked action potential now conducts across segment but
initially slow
WALLERIAN DEGENERATION 4. Return of clinical function; NCV may remain slow if large
segment affected
The pioneering work of Waller - detailed the physiologic B. Remyelination
298 299
reaction to nerve severance, which has since borne his name and I. In chronic processes cycle of demyelination/remyelination
is known as Wallerian degeneration. Neural reaction to tran- may repeat
section results in alterations in three portions of the nerve seg- V. Neuromuscular Junction/Muscle
A. Degeneration
ment. The first changes to consider are those that occur in the 1. Disintegration of terminal axons and neuromuscular
nerve distal to the lesion site, and it is these reactions that are junction
properly referred to as secondary degeneration or Wallerian 2. Failure of CMAP prior to SNAP because of above
degeneration (Tables 4-1 and 4 - 2 ) . - There are also reac-
298 299 3. Unstable muscle resting membrane potential
4. Extension of ACh receptors to extrajunction regions
tions to the injury just proximal to the lesion for a variable dis- 5. Tetrodotoxin resistant action potentials
tance constituting the second type of injury response. - - 20 87 244
6. Fibrillation potentials/positive sharp waves
Finally, a characteristic alteration becomes manifest in the B. Regeneration
perikaryon or nerve cell body. The latter two reactions are col- 1. Collateral sprouting if incomplete neural injury
2. Repair of above abnormalities
lectively referred to as axonal degeneration or axonal reac-
tion. Essentially any type of injury that produces axonal
214
From Brown WF:The Phsysiological and Technical Basis of Electromyography.
discontinuities whether it is transection, compression, crush, in- Boston, Butterworth Publishers, 1984, p 52; and Gilliatt RW: Recent advances in
the pathophysiology of nerve conduction. In Desmedt JE (ed): New Develop-
jection, ischemia, cold, or other types of inherited or acquired ments in Electromyography and Clinical Neurophysiology. Basel, Karger, 1973,
diseases will result in Wallerian degeneration with the above pp 2-18.
Chapter 4 PERIPHERAL NERVOUS SYSTEM'S REACTION T O INJURY — 117
neural regions affected. - - - The majority of work investigat-

22 26 65 74
distally. As time progresses within the first several days, the
47
ing the various aspects of Wallerian degeneration has been per- myelin continues to retract producing further node of Ranvier
formed in animals and not in humans. Despite this limitation, widening, and there is an accompanying increase in the
the majority of animal findings are believed to be directly ap- Schmidt-Lantermann incisures. It appears that the larger
312
plicable to humans. nerve fibers demonstrate this myelin alteration prior to it being
observed in the comparatively smaller nerves. During this time,
Sequential Changes the Schwann cell's nucleus enlarges and there is an increased
Axonal Component. A nerve fiber consists of a single axon amount of visible chromatin. Of note, the endoneurial fibrob-
with its investing myelin sheath. A peripheral nerve trunk is lasts begin to proliferate (Fig. 4-1). If there has been sufficient
comprised of multiple nerve fibers, both myelinated and un- trauma to surrounding tissues, the vascular structures demon-
myelinated. If the nerve trunk is compressed to the extent that strate endothelial swelling and multiplication of adventitial
the supporting connective tissue structures are not disrupted, but cells. By 72 hours, the myelin sheath formerly comprising the
32
the axon is injured, then axonal degeneration ensues. This type internode region now forms segmented ovoids engulfing the
of injury can be reproduced experimentally by crushing the previously noted axonal segments (Fig. 4 - 1 ) . These myelin 243
nerve with a hemostat. At the site of crush, endoneurial edema ovoids containing fragments of axon are noted to be present
can be observed within 1-2 hours as well as a surrounding zone within 72 hours in all transected nerves. There are also noted to
of hyperemia. - The endoneurial edema occurs because of
204 225
be myelin ovoids without axonal fragments. Also, multiple and
the capillaries' increased permeability secondary to the induced distinct nucleoli are observed in the Schwann cell nuclei.
trauma. - By several hours, the axon within and surrounding
133 227
Beginning approximately 4 days after insult, the Schwann cell
the crushed region is beginning to break apart. At about 72 nuclei start mitotic activity and produce an increase in the
hours following the crush, Schwann cells are in the process of amount of c y t o p l a s m .
2041
The Schwann cells then continue to
digesting myelin and axonal subcomponents, and preparing the undergo significant mitosis and proliferate. This is particularly
endoneurial sheath for the regeneration of axonal sprouts from noticeable at the severed nerve ends where these cells are at-
above the lesion site. - 46 132
tempting to bridge any gap that may be present if the nerve is
Immediately following transection, as opposed to the less severed. At about 10 days, the entire axon length distal to the
1192
severe crush injury, a small amount of intra-axonal fluid leaks lesion has been converted into a series of myelin ovoids. The
from both the proximal and distal severed ends of the nerve segments of the axon contained within the myelin ovoids even-
(Fig. 4-1). - The severed nerve ends retract secondary to the
195 328
tually disappear, prompting some investigators to refer to the
elastic properties of the endoneurium. - The inciting trauma242 281
ovoids as "digestive chambers." The Schwann cells contain-
243
also disrupts capillaries, which results in a localized hemor- ing the remnants of axon and myelin are referred to as
rhage and increased permeability of the surrounding vascula- "myelophages" and the peak of myelin chemical breakdown
ture. Macrophages and mast cells then invade the injury site. occurs about 14 days after the inciting incident.
Within a few hours of injury, the portion of nerve no longer in The inciting traumatic event has caused mast cells to enter
contact with the cell body begins to demonstrate a diffuse the tissue and release histamine and serotonin, thus generating
swelling. Approximately 7 - 1 0 hours after the insult, an eleva- further edema, capillary seepage, and the entrance of macro-
tion in the amount of intra-axonal osmophilic particles and vac- phages into the traumatized region. - Mast cell numbers con-
8 226
uoles are noted. It is believed that this accumulation of particles tinue to increase after the insult and maximize by day 4, and
represents endoplasmic reticulum dilatation and mitochondrial maintain this level until approximately day 15 with a continued
swelling and has been definitively observed by 19 hours. - 114 189
reduction thereafter, reaching normal numbers by about 4
There is also noted to be an aggregation of the neurofilamentous weeks. 282
The above-described processes occur within the
material. At this point, there are no gross light microscopic nerve's endoneurial tube. During this degeneration process, the
structural abnormalities beyond the region of damage. By 48 Schwann cells and invading macrophages from the site of ac-
hours, however, a number of additional changes are noted. companying vascular injury ingest the myelin and disintegrat-
Specifically, the axon appears rather pale and somewhat more ing axonal f r a g m e n t s . -
21
These fragments have usually
,02I34J 84,327
swollen and is more difficult to stain with routine tech- been completely removed by 35 days. After about 4-5 days, the
niques. 20194
Within the nodal and paranodal regions mitochondr- endoneurial tube contains a mass of proliferating Schwann cells
ial aggregations appear. The neurofibrils begin to disintegrate ingesting axonal components and myelin, macrophages, myelin
and decompose into their subcomponents. At about 4 8 - 7 2 ellipsoids, and lipid droplets. At 2 - 4 weeks this mass of
hours, the axons begin to fragment and form spiral or hook-like Schwann cells and minor amounts of cellular debris forms the
segments (Fig. 4-1). Axonal discontinuities originate initially at so-called bands of Biingner (Fig. 4-1). It has been estimated 41
the nodes of Ranvier and later further subdivide the axonal seg- that the number of nuclei approximately 25 days after injury is
ments by forming breaks at the internodal regions. Within 96 8 times greater than is normally present within the endoneurial
hours, the entire axon is completely disintegrated and confined tube. The proliferation of nuclei persists at 225 days but is re-
to small myelin ovoid fragments. At 7 days, there is noted to be duced to approximately 5 times normal. The number of en-
a complete absence of axon organelles. Although there is con-188
doneurial tube nuclei is directly dependent upon the amount of
siderable overlap, the breakdown of the axon may precede that debris required to be ingested. As the axonal material and
of the myelin by a short period. myelin are reduced, there is a concomitant reduction in the cel-
Myelin C o m p o n e n t . Within the first 2 4 - 3 6 hours, and re- lular mitotic activity. The endoneurial tube shrinks to a diam-
123
ported to have begun by 6-16 hours of neural insult, the myelin eter of about 10 pm and can remain in this condition and size
begins to retract from the axons at the nodes of Ranvier. - 46 243
for many months, awaiting the return of an axon from proximal
One group of investigators has observed myelin degeneration to the site of injury. After about a year, however, the endoneurial
within 2 minutes following a crush injury. - This myelin re- 322 323
tube is encompassed by surrounding connective tissue and
traction is initiated at the site of the lesion and then progresses begins to be further reduced to approximately half its original
118— PART I FUNDAMENTAL PRINCIPLES
Figure 4-1. Wallerian degeneration following neural section. A,

Region of nerve under consideration including that portion of nerve tran-
sected and viewed in m o r e detail below (region between horizontal lines).
B, Magnified portion of nerve between horizontal lines t o be included in the
injury zone. C.Appearance of nerve within first day following injury. Leakage
of axoplasm from proximal stump and separation of proximal and distal as-
pects of nerve. D, O n day 2 following trauma, the neurofibrils in t h e distal
axon and for a small distance proximally have disappeared. T h e r e is also
some shrinkage of the axon that appears rather irregular. Myelin is also be-
ginning to retract from the nodes of Ranvier. E, By day 3, there is fragmenta-
tion of the axon and myelin. Schwann cells undergo mitosis and proliferation
and begin t o digest the previous myelin components. F, On about day 8 the
axon fragments have been digested and Schwann cells are attempting t o
bridge the gap between t h e 2 neural portions. Regeneration of proximal
axon portion is noted with several neurites branching and evincing distally
along the outer margins of the myelin ellipsoids. G, At approximately day 12
there is a lessening of t h e gap separating the two nerve portions and there
is continued advancement of the growth cone. H, In this particular instance
the gap separating the t w o neural aspects is closed by the proliferation of
Schwann cells and fibroblasts, which occurs roughly by day 20.The mass of
Schwann cells (and in neural section fibroblasts) forms the band of Bungner
into which the growing axon penetrates. In this particular diagram a neurite
is seen t o n o t only e n t e r t h e band of Bungner but also form an a b e r r a n t
route external t o the nerve proper. I, O n day 100 the neural continuity is
established and all debris has been removed. Neural diameter remains less
than the original size and myelination is not yet present. J, Certainly by day
200 the myelin is established but at a significantly reduced thickness when
compared t o preinjury values.There is also noted t o be a decrease in the in-
ternodal distance, i.e., an increase in the internodes over the same length of
nerve. Modified from Bots G T h : Pathology of nerves. InVinken PJ, Bruyn
G W (eds): Handbook of Clinical Neurology,Vol. 7, Diseases of Nerves (Part
I). Amsterdam, North-Holland Pub. Co., I970,pp 197-243.
cross-sectional a r e a . - ' 2
With respect to the axonal fragmen-
49279
portions of the nerve is examined to gain a better appreciation
tation, it is believed that this process occurs simultaneously over of the histologic basis for clinical and electrophysiologic find-
the entire length of the distal aspect of the nerve. The myelin ings noted in peripheral nerve injuries.
changes, however, may proceed in a more proximal to distal se- Nerve F i b e r Degeneration. In addition to the neural
quence somewhat lagging behind the axonal segmentation changes occurring distal to the injury site, there are degenera-
process, - or occurring concomitantly with axonal disrup-
,27 257
tive alterations of the nerve also transpiring for a variable dis-
tion. - The actual sequence remains debatable.
20 243
tance proximal to the injury site. Depending upon the degree of
damage inflicted on the nerve, there is a similar pattern of
Retrograde Neural Changes Wallerian degeneration progressing along the nerve proximally.
An insult to the peripheral nervous system typically results in Minor injuries that do not disrupt the endoneurium, such as a
a reaction to the inciting trauma not only distal to the injury but crush, lead to proximal Wallerian degenerative changes extend-
proximal as well. This proximal reaction can be localized to ing for several millimeters. Lesion producing severe disruption
three regions: (1) peripherally; at and about the injury (within or transection of the entire nerve can generate retrograde degen-
several centimeters) as well as major portions of the nerve prox- erative changes extending for several c e n t i m e t e r s . - 67193 281
imal to the injury zone (nerve fiber degeneration), (2) cell body Proximal to the injury the axons become swollen, which may be
no longer connected to its axon (axonal reaction), and (3) neural a result of both edema formation and blocked axoplasmic flow
synapses (trans-synaptic neuronal reaction). Each of these
281
(Fig. 4-1). The support for axonal transport disruption is based
Chapter 4 PERIPHERAL NERVOUS SYSTEM'S REACTION TO INJURY — I 19
on the observation that there is an accumulation of intraneural

transport products such as mitochondria, storage granules, and
oxidative enzymes in the perikaryon. - - * If the nerve is
61 95 96 310
transected, the portion of the nerve immediately proximal to the

injury demonstrates a reduction in its diameter. This reduction
is believed to occur as a result of both neural components, axon
i and myelin sheath, decreasing in size. The axon diameter may
! show a reduction in size because the intra-axonal contents are
1
simply escaping into the surrounding tissues (i.e., the trans-
ported substances are accumulating in the soft tissue spaces sur-
rounding the i n j u r y ) . -
6126
A reduction of 4 0 % in axonal
263
diameter has been quantified in the facial nerve between 10-21

days and eventually approached the original diameter to within
13% beyond 3 weeks. Of course, the ultimate fate of the prox-
181
imal portion of the injured nerve is completely dependent upon

the effect of the injury on the axon's cell body. Survival of the
cell body results in continued neural functioning with a chance
of regeneration. Death of the parent cell body, however, results
in Wallerian degeneration of the proximal axonal segment (see
below).
Large portions of the affected nerves proximal to the injury
also demonstrate histologic alterations. The myelin sheath thick-
ness decreases as well as the axonal diameter. If continuity with
the original end organs is established, the size of these two neural
components is re-established. Failure to contact the end organ re-
sults in a permanent reduction in proximal axonal/myelin size.
Axonal Reaction. There are changes occurring in the axon's
cell body simultaneously with the above-noted alterations of the
nerve cell's peripheral extension. The changes demonstrated by
the cell body are referred to as axonal reaction, but also have
been called axonal degeneration, retrograde degeneration,
and central chromatolysis (Table 4-2). A nerve cell may
32 Figure 4-2. Reaction of a motor ventral horn cell to injury of
demonstrate similar changes in not only physical disruption of its axonal extension. I, Normal nerve cell prior t o trauma with
the axon, but also following extended direct neural stimulation, even distribution of Nissl substance. 2, By 4 8 hours following nerve
strychnine poisoning, and poliomyelitis. Not all nerve
16,83 M 8 section, t h e Nissl substance is disappearing and changing in t h e s o -
cells display changes following peripheral nerve insult. called "dust particles." 3 , A t 2 o r 3 weeks t h e cell is swollen with an
Depending upon the severity of the injury, 10-90% of affected eccentrically located nucleus and only a marginal appearance of t h e
neurons can undergo retrograde changes. °- - After a periph-
1 88 329 Nissl substance.The nucleolus is also eccentrically placed in the nu-
eral nerve injury, the parent cell may not be affected and subse- cleus. O n c e the cell has reached this stage, it can proceed in one of
quently demonstrate no changes. The severity of the retrograde t w o directions. 4 a , T h e cell can die and a t first appear as a so-called
reaction is essentially dependent upon the amount of trauma in- ghost cell and then completely disappear. 4 b , It is also possible for t h e
curred by the peripheral nerve. Mild injury generates only mild cell t o recover and again form discrete Nissl bodies. Also, the cell as a
cell body changes while complete transection of a nerve is more whole is no longer swollen and the nucleus is again centrally located.
likely to produce a more intense reaction (see below). Also, 166 (Modified from Bots G Th: Pathology of nerves. In Vinken PJ, Bruyn
proximal compared to distal neural insult results in a greater G W (eds): Handbook of Clinical Neurology,Vol 7. Diseases of Nerves
axonal reaction. This is believed to occur because of the per-
118 (Part I). Amsterdam, North-Holland Pub. Co., I970,pp 197-243.)
centage of total axoplasm removed from the cell. There also 27
appears to be a greater sensitivity for sensory compared to several days to 3 weeks there are additional changes noted in
motor cells. Sensory nerves demonstrate a more rapid and in- the cell. The nucleus also becomes swollen and moves toward
tense axonal reaction to trauma. When the cell body is affected
88
the cell's outer margins. There is an increase in the mitochon-
by peripheral nerve damage, it first displays a characteristic se- dria and only a thin region of fine Nissl granules is noted along
quence of changes called the reactive phase or chromatolytic the outer margins of the cell. The cell body is essentially prepar-
phase (Fig. 4-2). The cell can then proceed in one of two ways. ing to produce the substances required to supply the necessary
It can recover from the insult and undergo chromosynthesis or proteins and cell organelles as well as the axoplasm to regener-
not recover and enter a degenerative phase. 281
ate the axon. It has been estimated that the volume of the axo-
Chromatolytic Phase. After a peripheral nerve insult severe plasm is 200-1000 times that of the parent cell body's contents.
enough to result in Wallerian degeneration, the cell body first Once in the chromatolytic phase, the cell may either undergo
demonstrates an alteration in the Nissl bodies. The Nissl 214
degeneration or recover (Fig. 4-2). The above-noted changes are
bodies or substance (cell body's endoplasmic reticulum) breaks most prominent in the spinal motor neurons, dorsal root gan-
apart into fine dust-like particles within the first 6 - 4 8 hours glion cells, brainstem motor neurons, pyramidal cells, and
(Fig. 4-2). Additionally, the cell body swells and appears more Clark's column neurons. This is because these particular cells
rounded. During this time, the Golgi system is repositioned have distinct aggregates of Nissl particles. Cell bodies without
away from the nucleus to the cell's periphery. Between the first the Nissl substance do not display the above-noted chromatolysis,
but do demonstrate the other characteristics of the reaction to one axonal sprout present for any period of time. Within a short
nerve injury. time there is resorption of the multiple axonal sprouts and the
Chromosynthesis Phase. If a cell and its extensions are formation of one dominant axon. After 24 hours, the advancing
completely located within the central nervous system, degener- terminal axons appear as small terminal buds with multiple
ation is the likely result of neural injury. The majority of cells smaller club-like terminal axonal branches extending into the
projecting peripheral nervous system axons, however, experi- damaged area by 3-8 days (Fig. 4-1). The distal expansion of
ence a recovery phase. Complete recovery is dependent on the the axon tip or growth cone as directed by the cell body serves
axon proximal to the site of injury bridging the injured site and to" increase the longitudinal extent of the axoplasm. The 281
establishing physiologic continuity with its appropriate end growth cone extends several extensions referred to as filopodia
organ. The beginning of chromosynthesis is usually demarcated or neurites. Elongation of the regenerating axon is accom-
by the earliest change noted in the neuron, which is a reforma- plished by an ameboid type of motion of the filopodia. In minor
tion of the granular Nissl substance. Nissl substance formation traumatic injuries with no disruption of the endoneurial tube
is first observed at about 2-3 weeks following injury if there has structure, axon tips have been detected below the injury site
been little disruption of the endoneurial tube. Recovery then within 4 - 1 0 days subsequent to the incident. - - - - It is
79 125 133 216 273
proceeds and requires approximately 10 weeks for the cell to possible for the advancing axonal tip to arrive at the distal por-
again appear as it did prior to the injury (Fig. 4-2). In severe tion of the endoneurial tube when the Schwann cells still con-
nerve trauma with disruption of supporting connective tissue tain debris from the previous axon. In this case, there is little, if
structures, it may take more than 4 months before the neuron any, hindrance to axonal penetration offered by the partially di-
demonstrates signs of recovery. An additional early sign of neu- gested axonal fragments and myelin. - 60 281
ronal recovery is that the nucleus returns to the central portion If there has been sufficient trauma to the nerve resulting in
of the cell and the swelling resolves. - - Also, neurofibrils are
68 69 281
scar formation with a gap between the two nerve ends, success-
reformed and the Golgi system, again, becomes perinuclear. ful regeneration depends upon the length of time since the injury,
The extent to which a cell experiences the axonal reaction de- age of the subject, and content of the scar. Axonal sprouts may
pends upon the severity of injury, the subject's age, and the
41 122
appear within 6 hours of injury. One axon can give rise to mul-
233
distance between the cell body and level of injury. 294

tiple collateral sprouts thus forming an intra-endoneurial tube
Degenerative Phase. Should the injury be severe enough plexus by 48 hours with only a few collaterals eventually pene-
that continuity is not established peripherally, then the cell body trating the zone of injury to enter the distal band of Bungner. Up
is likely to proceed from the chromatolytic to the degenerative to 50 axonal sprouts have be observed to arise from 1 axon and
phase (Fig. 4-2). Degeneration is highly variable and may occur more than one of these sprouts may enter the same endoneurial
in weeks or require several months. Experiments attempting
281
tube. - It is then possible for up to four of these axonal sprouts
62 314
to define the extent of cell body loss have demonstrated a range to acquire a myelin sheath, but one eventually dominates, pre-
of 6 - 5 0 % . The extent of degeneration is dependent upon
13 2 5 6
serving the ratio of one axon per endoneurial tube. Occasionally,
lesion severity. it may be possible for two or more axonal sprouts to remain
Trans-synaptic Neuronal Reaction. Perineuronal glial within one endoneurial tube and successfully reinnervate an end
cells demonstrate an increase in number and size within 24^4-8 organ (e.g., muscle). Although not proven, this may be the expla-
hours of injury. - Both microgliacytes and astrocytes partici-
167 229
nation of the axon reflex. The terminal buds are then converted
pate in this reaction. In rather severe lesions, the microglial cells into terminal repeatedly branching forks by 30 hours, trying to
appear to cause a separation of the affected neuron's synaptic penetrate the wound region containing necrotic tissue by passing
connections with other cells. This action may help in isolating
199
directly through or around it in an attempt to re-enter the en-
the reactive neuron from receiving input by way of neurotrans- doneurial tube. Within the first 3 days, the wound region is filled
mitters, thus allowing it a quiescent period to primarily perform with exudate (blood and plasma) and a fibroblastic network. If 32
the necessary repairs of its peripheral extension. The actual the axon sprout encounters a substance inhibitory to further
process and substances responsible for this reaction are not en- progress (e.g., blood clot, fibrous tissue, fat, scar, etc.), it may bi-
tirely known. Recovery results in a return of the preinjury state furcate, double back, go around the obstruction, or form a termi-
with synaptic terminals being established requiring approxi- nal bulb or spiral of Perroncito (Fig. 4-3). Following neural
mately 4 months after anatomic recovery. These same cells 245
transection or neural repair, it has been estimated that about 1/6
may function in a phagocytic role should the neuron degenerate. to 1/7 of nerve fibers may eventually terminate in the desired end
organ. By day 5, Schwann cells have penetrated the wound
243
NEURAL REGENERATION region and allowed axonal branch passage to be much easier. It
is unclear exactly what factor(s) guide the neurites across the in-
The origin of neural regeneration is that portion of the axon jured site to reach the distal site of the axon. For narrow nerve
proximal to the lesion location contained within the undamaged gaps, fibroblastic tissue rapidly unites the two nerve ends and
endoneurial tube (Table 4-2). Within 6 hours following an injury neurites may follow the fibroblastic bridges. Rather wide gaps
severe enough to result in Wallerian degeneration, the terminal lead to neuroma formation and thus permit only a few axons to
portion of the intact proximal nerve stump becomes differenti- bridge the gap. In optimal situations with only minimal gaps, it
ated from that portion of the nerve about to begin the previously may take up to 8-15 days for the growing axons to reach the en-
described degeneration process. The distal portions of the sur- doneurial tubes. The regenerating axons usually have a diameter
viving viable axons begin to swell and each axon may give rise of roughly 0.5-3.0 pm and penetrate the scar region at 0.15-0.24
to two or more axon branches or sprouts, which have been re- mm/day, but can reach speeds of 2.5-4.0 mm/day once in the
ferred to as neurofibrillar brushes. - In simple lesions with
32 272
distal portions of the endoneurial tubes. Individual human
243
little edema formation and preservation of the endoneurial tube, nerves have various growth rates: median (2.0-4.5 mm/day),
terminal portions of the regenerating axons penetrate the dam- ulnar (1.5 mm/day) and radial (4.0-5.0 mm/day). The rate of 77
aged area and beyond by 24 hours, and there is rarely more than axonal growth is proportional to the distance of the growth cone
from the cell body. The closer the injury to the cell body, the b
faster the rate of growth with a proportional slowing the further
the insult is from the cell body. Also, at any given distance from
the cell body, the rate of axonal growth is the same irrespective
of the original lesion site for a particular nerve. Therefore, for
proximal lesions, the rate of axonal growth is faster than for the
more distal locations, i.e., the growth rate slows as the growth
cone advances. '28
When the axon tip finally reaches the distal aspect of the en-
doneurial tube containing the Schwann cells, there is an align-
ment of Schwann cells about the advancing axon. Once aligned,
the Schwann cell begins to rotate about the axon to form a mul-
tilayered structure of the myelin sheath. The formation of the
myelin sheath about the axon is approximately 9-20 days
behind the advancing font of the axon. < Myelination,
139,240 264
therefore, follows the progression of the axon distally at a rate

of approximately 4 mm/day. The natural separation between
239
the Schwann cells forms the node of Ranvier, while the seg-
ment of myelin containing the Schwann cell nucleus is called
the internode region. Unlike the internode segment prior and
proximal to the injury, there is a shorter distance from one node
of Ranvier to the n e x t . - In other words, there are more
144157 293
nodes of Ranvier over the regenerated nerve compared to the

region prior to the injury (Fig. 4-1). It may take approximately
one year or more for the myelination to fully mature. 250
Within about 3 months following the injury, the denervated

endoneurial tubes are 3 pm in diameter or smaller. This repre- 279
sents the maximum shrinkage of the endoneurial tubes and they

stabilize at this diameter. Upon neural re-entry into the en-
doneurial tube, the bands of Bungner or Schwann cell mass
helps guide the returning axon. It is unlikely, however, for the
endoneurial tube to ever regain its diameter prior to neural Figure 4-3. Endoneural tube disorganization. Following neural
injury. If the endoneurial tubes are not reinnervated by approx-
5 transection t h e r e is often major disorganization of the endoneurial
imately 1-1.5 years after denervation, they are less susceptible tubes as well as interposed connective tissue between the 2 ends of
to receiving an axon because of Schwann cell replacement by the severed nerve.The axon attempts t o grow from the central p o r -
connective tissue. 281 tion of the injured nerve (A) across the scar region (C) to reach t h e
distal aspect of the nerve (B).There is a significant amount of disorga-
In lesions primarily involving crush, cold, concussion, or
nization with respect t o the regrowth of the neurites. Some do indeed
compression, the endoneurial tubes remain intact and there is
cross t h e scar with s o m e bifurcation prior t o entering distal e n -
relatively little difficulty in the axon regrowing across the dam-
doneurial tubes (c) but multiple others are misdirected (a) or simply
aged section. Therefore, the nerve re-establishes contact with
proliferate in connective tissue forming t h e so-called spiral of
the original end organ to again reinnervate the intended struc-
Perroncito (d). (From Bots G Th: Pathology of nerves. InVinken PJ,
ture. When there is physical separation between the cell body
Bruyn G W (eds): Handbook of Clinical Neurology,Vol 7. Diseases of
and end organ, reinnervation is less assured. There is a distinct
N e r v e s (Part I). A m s t e r d a m , North-Holland Pub. Co., 1970, pp
possibility for the newly formed neurites to enter inappropriate
197-243.)
endoneurial tubes (Fig. 4-4). For example, sensory neurites may
enter an endoneurial tube destined to end in a muscle fiber
while a motor neurite can be directed to a sensory end organ. In If the axon eventually reaches the end organ, the process of
both cases, the regeneration is useless—or even counterproduc- axonal maturation is capable of proceeding. The axon begins to
tive—because the inappropriate end organ is reached. The fac- increase in diameter in a proximal-to-distal direction. The 149
tors redirecting neurites toward appropriate endoneurial tubes is Schwann cells forming the bands of Bungner begin to increase
less than 100% effective and poorly understood. There is a com- in size and become longitudinally oriented about the newly
petitive process in muscle tissue following a partial injury to a formed axon. There may be more than one neurite within a
motor nerve. While the axonal sprouts are regrowing down the single endoneurial tube. The eventual fate of these neurites may
endoneurial tube, collateral sprouting is also occurring within be that two individual myelinated nerves are formed or they
the muscle tissue. Intact nerves send out collateral sprouts to may combine into a single axon. As previously noted, the
reinnervate neighboring muscle fibers. In this case, it is en-
82147
Schwann cells begin to rotate about the axon in the process of
tirely possible for there to be little muscle tissue left to reinner- myelin formation. Schwann cell nuclei take on a more elon-
vate by the original nerve as the process of collateral sprouting gated appearance in this alignment process. As myelination pro-
has already accomplished the required reinnervation. It is also gresses, segmentation of the myelin begins with a single
possible for remaining sensory nerves to increase their area of Schwann cell comprising an internode region and nodes of
distribution somewhat to provide cutaneous sensibility to a pre- Ranvier accounting for the production of the internodes. The
viously denervated region of skin prior to neural regrowth from process of myelination may begin as early as 3-4 weeks follow-
the injured site. -
276 313
ing a transection starting proximally and progressing distally. 209
Figure 4-4. N e r v e t r a n s e c t i o n . T h e r e is sprouting of multiple neurites from a single axon (spr) surrounded by a common basal lamina. At the
end of each sprout o r neurite there is a growth cone (gc).The sprouts cross the injured zone associated with Schwann cells (Schw) attempting t o
reach the distal nerve stump. Found within t h e injury zone are macrophages (m), fibroblasts (fb), mast cells (mc), and blood elements. Upon reach-
ing the distal nerve stump, the sprouts attach and enter the band of Bungner. It is rather obvious how misdirection of axonal sprouts is a common
occurrence. (From Lundborg G: Nerve Injury and Repair. Edingurgh, Churchill Livingstone, 1988, with permission.)
If there has been enough neural trauma to result in a separa- the nerve, which requires the Schwann cell to ingest and then
tion of the nerve, the chances of successful reinnervation are de- resynthesize a new myelin layer. Usually, only one or a few in-
creased. As time passes, there continues to be collagen ternodal segments are affected in segmental demyelination,
deposition within and about the endoneurial tube, resulting in a which is quite different from the degree of Schwann cell alter-
progressive thickening of this structure. Should reinnerva-
159 2 7 9
ation noted in Wallerian degeneration. 235
tion occur within the first year, the final axonal diameter may be The process of segmental demyelination is typically re-
reduced to 7 5 % of the original diameter, but this does not stricted to a focal segment limited to several internodes of
appear to impede neurite penetration or the nerve's eventual nerve, although multiple proximal and distal foci can be pre-
physiologic properties. The final result of regeneration is
75,279
sent. The nerve's axon and myelin not immediately adjacent to
usually a nerve fiber that is smaller in diameter, thinly myeli- the damaged region are relatively unaffected. Through an un-
nated, and associated with shorter internodes. known process, the majority of the above nerve diseases exert a
direct toxic effect on the Schwann cell itself or the myelin
SEGMENTAL DEMYELINATION sheath while sparing the axon. Of course, if the demyelinating
disease is particularly profound, it is possible for the axon to be
As noted above, if the axon is injured, one can anticipate the secondarily injured, resulting in Wallerian degeneration.
characteristic pattern of Wallerian degeneration to ensue. It is Unlike Wallerian degeneration, it is much more difficult to
rather common, however, for a nerve to be affected by a disease accurately document a characteristic pattern of changes with
process or other insult that does not damage the axon, but pref- segmental demyelination because of the variable nature of the
erentially injures the nerve's myelin (Table 4-2). In this in- types of diseases capable of generating this lesion. Initially, the
stance, the damaged myelin is removed and replaced in the myelin sheath begins to appear somewhat granular with moder-
process known as segmental demyelination/remyelination (Fig. ately sized vacuoles and lipid droplets compared to the rather
4-5). Throughout the demyelination/remyelination process, the larger ovoids seen in Wallerian degeneration (Fig. 4-5). In ex-
120
axon and surrounding endoneurial connective tissue remain perimental segmental demyelinating investigations (e.g., diph-
undisturbed and intact. A few of the diseases known to produce theric polyneuropathy), 5-8 days following injection of a toxin,
varying degrees of demyelination are: Guillain-Barre syndrome, initial separation pf the myelin lamellae were noted at the nodes
chronic inflammatory demyelinating polyradiculoneuropathy, of Ranvier and Schmidt-Lantermann incisures, and these
diabetes mellitus, leukodystrophies, and several forms of changes correlated with the subject's clinical symptoms. - In
309 316
Charcot-Marie-Tooth disease (Table 4 - 3 ) . - - - - ' - - In

66 9, ,,5 ,22 287 296 3n 325
experimental allergic neuritis, the first changes noted were sep-
short, if a nerve experiences an insult that does not produce aration of the terminal myelin loops in the paranodal region. 14
Wallerian degeneration, chances are that at least segmental de- Within 7 days of these observations, the myelin sheath began to
myelination may appear. This insult results in a malfunction of disintegrate. The Schwann cells increased in number beginning
the Schwann cell to either maintain or continue to produce for a about the nodes of Ranvier. By 8 days following the onset of
period of time, the protein and lipoprotein synthesis required to myelin breakdown, macrophages and Schwann cells began in-
sustain a myelin sheath. It is also possible for compressive or
49
gesting the myelin and other breakdown products. The axon and
ischemic episodes to structurally derange the myelin surrounding immediately surrounding myelin layer were essentially spared
Figure 4-5. Segmental demyelination. I, Overall view of the nerve prior t o injury with t h e section contained within t h e box magnified in t h e
subsequent sections. II, Region of nerve (one internode) affected by a disease process resulting in myelin breakdown of this segment only. Ill, Myelin
is being digested and there is proliferation of Schwann cells. IV,The myelin has been completely removed. V, Remyelination is complete. Note t h a t
over this segment previously containing one internode, there are now three internodes. (Modified from Bots GTh: Pathology of nerves. InVinken PJ,
Bruyn G W (eds): Handbook of Clinical Neurology,Vol. 7, Diseases of Nerves (Part I). Amsterdam, North-Holland Pub. Co., 1970, pp 197-243.)
in most segmental demyelinating processes. The general ap- neural insult. The degree to which a nerve is damaged has impli-
pearance of the affected portion of nerve was that of a relatively cations with respect to its present function and potential for re-
thinly myelinated nerve, often surrounded by Schwann cells and covery. There are two general classification systems (Table 4-4).
macrophages ingesting the affected myelin. Once the affected One is that of Seddon, which considers neural injury from the
myelin was significantly removed, Schwann cells proliferated perspective of a combination of functional status and histologic
and began to remyelinate the affected segment of nerve, usually appearance. Although there are a number of factors that may
by 14 days. Just as in Wallerian degeneration, there were signif- affect the nerve, the devised terminology primarily, though not
icantly more internodes and accompanying nodes of Ranvier exclusively, considers mechanical trauma (compression, stretch,
than prior to the injury (Fig. 4-5). crush, concussion, or various degrees of transection) to be the
inciting incident. In Seddon's scheme, there are three degrees or
NERVE INJURY CLASSIFICATION stages of injury to consider: neurapraxia, axonotmesis, and
neurotmesis (Table 4-4). - 269 270
Seddon's Classification Neurapraxia. The term neurapraxia is used to designate a

In order to intelligently communicate about nerve injuries, it mild degree of neural insult that results in impulse conduction
is appropriate to define a classification based on the amount of failure across the affected segment. It is also acceptable to simply
designate this type of neural insult as conduction block. T h e
Table 4-3. Human Polyneuropathies with Segmental most important aspect of conduction block is its reversibility.
Demyelination When conduction block affects a neural segment, the conducting
properties of the nerve above and below the lesion site are
Minimal Demyelination Significant Demyelination normal. Additionally, the continuity between the cell body and
Alcohol with related thiamin Diphtheria end organ is maintained. Wallerian degeneration does not result
deficiency Some forms of diabetic neuropathy from a conduction block, which implies axonal continuity.
Majority of toxic substances Some forms of Charcot-Marie- Investigators have mimicked this type of neural lesion by care-
Acute porphyria Tooth disease (CMT I, CMT 4) fully compressing nerves to various degrees such that only a focal
Most paraneoplastic neuropathies Leukodystrophies demyelination occurs with little, if any, axonal injury. - T h e
65 218
CMT2 Guillain-Barre syndrome end result of a local demyelinating lesion is action potential slow-
Amyloid neuropathy Chronic inflammatory demyelinating and failure across the compressed aspect of nerve. Nerve con-
Most diabetic neuropathies ing polyradiculoneuropathy duction is preserved proximally and distally to the lesion.
Uremic neuropathy Multifocal motor neuropathy In the above-noted compression injury, conduction again re-
Ischemic neuropathies Neuropathies associated with turns within several weeks or months, i.e., the time required to
IgM (anti-MAG) monoclonal remyelinate the damaged neural segment. The large myelinated
gammopathies fibers are more susceptible to compression and ischemia than
Modified from Gilliatt RW: Recent advances in the pathophysiology of nerve are the thinly myelinated and unmyelinated fibers. Conduction
conduction. In.Desmedt JE (ed): New Developments in Electromyography and block usually affects motor fibers rather profoundly, with rela-
Clinical Neurophysiology. Basel, Karger, 1973, pp 2-18. tive degrees of sensory and sympathetic fiber sparing depending
Table 4-4. Nerve Injury Classification

Type Function Pathological Basis Prognosis
Lundborg
Physiologic Focal conduction block Intra-neural ischemia Excellent; Immediately reversible
Conduction Block: Metabolic (ionic) block. N o
Type a nerve fiber pathology
Physiologic Focal conduction block Intra-neural edema; Recovery in days o r weeks
Conduction Block: Increased endoneurial fluid
Type b pressure;
Metabolic block;
Antibody-mediated channelopathy;
Little o r no fiber pathology
Seddon Sunderland
Neurapraxia Type I Focal conduction block. Local myelin injury, primarily Recovery in weeks to months
Primarily m o t o r functon and larger fibers.Axonal continuity.
proprioception affected. Some N o Wallerian degeneration
sensation and sympathetic
function can be present.
Axonotmesis Type 2 Loss of nerve conduction at Disruption of axonal continuity Axonal regeneration required for
injury site and distally. with Wallerian degeneration, recovery. Good prognosis since
Endoneurial tubes, perineurium original end organs reached.
and epineurium intact.
Type 3 Loss of nerve conduction at Loss of axonal continuity and Disruption of endoneurial tubes,
injury site and distally. endoneurial tubes; Perineurium hemorrhage and edema produce
and epineurium preserved scarring. Axonal misdirection. Poor
prognosis and surgery may be
required
Type 4 Loss of nerve conduction Loss of axonal continuity, endo- Total disorganization of guiding
at injury site and distally. neurial tubes, and perineurium. elements. Intra-neural scarring
Epineurium remains intact. and axonal misdirection. Poor
prognosis and surgery necessary.
Neurotmesis Type 5 Loss of nerve conduction Severance of entire nerve. Surgical modification of nerve ends
at injury site and distally. required. Prognosis guarded and
dependent upon nature of injury
and local factors
Modifified from Lundborg G: Nerve Injury and Repair. Edinburgh, Churchill Livingstone, 1988.
upon the degree of insult. Motor paralysis in conduction block block because muscle innervation is maintained, and secondly,
lesions typically lasts from 1 to 6 months, although most lesions recovery is typically rapid enough to avoid disuse atrophy.
usually resolve by 3 m o n t h s . If sensation is affected, it
2 5 9 m 2 7 8
Fibrillation potentials should not be observed in conduction
appears that touch perception is more profoundly altered than block as the axon is not disrupted. It is important to keep in
pain sensation. These two sensory modalities are less affected mind that a mixed lesion can exist where there is a combination
than motor control and proprioception, and return to a func- of conduction block and axonal loss. In this case, it is certainly
tional status more quickly. Sympathetic fibers are the least af- possible to observe fibrillation potentials.
fected by conduction block. Cutaneous sensibility is usually Conduction block is familiar to most individuals. For exam-
only mildly affected and returns to normal relatively ple, sitting with one's legs crossed, such that the knee of one leg
rapidly. - The severity of conduction block has been graded
259 290
compresses the peroneal nerve for several minutes, results in the
according to the duration of the block: ( 1 ) brief; lasting minutes foot "falling asleep." This sensation corresponds to an ischemic
to hours, ( 2 ) moderate; blockage for up to 4 weeks, and ( 3 ) insult of minor degree to the peroneal nerve with a resultant
severe; several months duration (Table 4 - 4 ) . Unfortunately, it
282
conduction block. Depriving the nerve of its needed blood
is not possible to accurately grade the severity of conduction supply impedes action potential propagation across the is-
block prior to it manifesting resolution. chemic segment. If the compression is of sufficient degree or
Clinically, the onset of motor control and sensory functional length of time, then the individual not only can no longer acti-
loss can be either abrupt or gradual. Depending upon the degree vate the foot and toe dorsiflexors/extensors, but a lack of sensa-
of injury, it is possible to have only partial or complete disruption is present as well. The conduction block prevents both
tion of either motor or sensory modalities. This variability of motor impulses from reaching the peroneal-innervated skeletal
motor/sensory loss is believed to reside in the fact that the dif- muscles as well as sensory impulses from the superficial sen-
ferent fibers are more or less susceptible to injury depending sory peroneal nerve reaching the cerebral cortex. Uncrossing
upon their location within the nerve as well as focal nature of the legs allows blood flow to once again return to that segment
the lesion. Muscle wasting usually does not occur in conduction of compressed nerve with eventual return of both motor and
sensory function. Prolonged compression can eventually lead to tissue is minimized to afford easy axonal growth across the
focal demyelination and, if severe enough, axonal disruption injury site.
with secondary Wallerian degeneration. One can easily see that
the manner in which nerves respond to trauma is dependent Sunderland's Classification
upon the degree of the initial or continuing insult. A second popular and somewhat more detailed classification
Axonotmesis. The second degree of neural insult in Sed- was initially proposed and subsequently modified by
don's classification is axonotmesis (Table 4 - 4 ) . Axonot-
269 2 7 0
Sunderland (Table 4 - 4 ) . - This classification of nerve injury
280 281
mesis is a specific type of nerve injury in which only the axon is is based upon the results of trauma with respect to the axon and
physically disrupted, while the enveloping perineurium and its supporting connective tissue structures. Sunderland's classi-
epineurium are preserved. Compression of a profound nature or fication is divided into five types of injury and depends exclu-
traction on the nerve are the typical etiologies of such a lesion. sively upon which connective tissue components are disrupted
Once the axon has been disrupted, the characteristic changes (Fig. 4 - 6 ) .
previously described for Wallerian degeneration occur. Type 1 Injury. Type 1 injury corresponds to Seddon's des-
Recovery of function is directly dependent upon the time re- ignation of neurapraxia.
quired for the process of Wallerian degeneration and neural re- Type 2 Injury. Seddon's axonotmesis is subdivided into
generation to eventually reach the previously denervated motor three forms of neural insult (types 2-4). A type 2 injury involves
or sensory end organ. Of course, autonomic function is also lost loss of axonal continuity with preservation of all supporting
and must be reestablished. The fact that the endoneurium re- neural structures including the endoneurium and corresponds to
mains intact is a very important aspect of this type of injury. A Seddon's axonotmesis (Table 4-4). Types 3 and 4 injuries result
preserved endoneurium means that once the remnants of the de- in progressively more neural disruption. Sunderland's type 5
generated nerve have been removed, the regenerating axon injury corresponds to Seddon's neurotmesis, i.e., complete
simply has to follow its original course directly back to the ap- neural disruption.
propriate end organ. With this type of injury, there is no misdi- Type 3 Injury. In type 3 injury there is loss of axonal conti-
rection of regenerating axons. A very good prognosis, therefore, nuity as well as the endoneurium, i.e., the endoneurial tube and
is implied when neural damage results only in axonotmesis, contents (axon) are disrupted (Fig. 4 - 6 , Table 4 - 4 ) . With this
I provided the distance between the lesion site and end organ is
1
not too long.
Clinically, because the axon is physically disrupted, one can
anticipate denervation of the corresponding musculature and
complete absence of all sensory modalities. Additionally, there
is an absence of autonomic control to the affected area. Due to
the process of Wallerian degeneration, all tissues become inex-
citable distal to the site of injury and the proximal neural
changes previously described ensue. The length of recovery is
entirely dependent upon the distance between the level of lesion
and the end organs. Obviously, this time interval is longer than
that previously described for conduction block. With respect to
motor recovery, the previously denervated muscles demonstrate
voluntary activity in the sequential order in which they are in-
nervated, i.e., return is proximal to distal. It is often possible to
clinically trace the nerve's recovery based on the advancing
Tinel's sign, as the sensory fibers are sensitive to percussion at
their growing tips. Although the prognosis for recovery is very
good, there are occasions when the effects of retrograde neu-
ronal degeneration has resulted in the loss of some cell bodies.
In this instance, less than complete recovery can be expected.
Neurotmesis. The greatest degree of disruption a nerve can
incur is designated neurotmesis and implies complete disrup-
tion of not only the axon, but all supporting connective tissue
structures. - In this instance not only is the axon disrupted,
269 270
but the endoneurium, perineurium, and epineurium are no

longer in continuity. Neurotmesis implies the nerve is com-
pletely severed even if the outward gross appearance suggests
otherwise. As a result, a neurotmetic lesion has a very poor
prognosis for complete functional recovery. In other words, for
the individual to recover useful function, surgical repair is most Figure 4-6. Sunderland classification of neural injury. I,
likely required. The reason surgical repair is necessary is be- Conduction block. 2, Wallerian degeneration occurring secondary t o a
cause the gap separating the two ends of the nerve becomes lesion confined t o t h e axon with preservation of t h e endoneurial
filled with connective tissue and a serum clot as well as mis- sheath. 3,There is disruption of the axon and endoneurial tube within
alignment of the appropriate endoneurial tubes. It becomes dif- an intact perineurium. 4, Disruption of all neural elements except t h e
ficult for the neurites to work their way through the clot and find epineurium. 5, C o m p l e t e discontinuity of t h e entire nerve t r u n k .
the appropriate endoneurial tubes. Although surgery does not (From Sunderland S: N e r v e Injuries and Their Repair: A Critical
guarantee proper endoneurial tube alignment, at least the scar Appraisal. Edinburgh, Churchill Livingstone, 1991, with permission.)
type of injury, Wallerian degeneration is to be expected as well face widespread loss of continuity and misalignment of en-
as a loss of organization within the funiculus. The perineurium doneurial tubes. Additionally, because of the internal architec-
and epineurium remain intact. Unfortunately, the intrafunicular tural disorganization, regenerating axons now not only enter
disorganization and hemorrhage, edema, and fibrosis tend to inappropriate endoneurial tubes, but also penetrate between fu-
retard optimal neural regeneration to the original end organs. niculi to end blindly in supporting connective tissues. It is likely
Traction and compression are common etiologies for this type that many fourth-degree injuries lead to neuroma formation be-
of neural insult. cause of the above-noted mass of misdirected axons and con-
Neural recovery in third-degree injuries is usually less than nective tissue proliferation. The clinical recovery of these types
optimal. Because of this injury's severity, it is to be anticipated of injuries is rather poor and often requires surgical intervention
that a number of sensory and motor neurons are lost. This loss to reduce the internal disorganization and scar tissue.
of proximal nerve cells results in a reduced number of axons ca- Type 5 Injury. In fifth-degree injuries, the entire nerve is
pable of participating in the regenerative process. Additionally, disrupted such that there is a loss of continuity, i.e., the nerve is
the entire neural recovery process is hindered and delayed be- severed through all of its supporting connective tissue structures
cause of the necessary recovery time for the nerve cell bodies (Fig. 4-6, Table 4-4). The epineurium has been completely tran-
that have undergone the axonal reaction. Some of these cell sected. Profound Wallerian degeneration as well as the proximal
bodies may not recover fully following the traumatic incident. A axonal reaction develop. Any functional recovery is rather rare
second impediment to maximal recovery is the intrafunicular in this type of injury. The severe axonal reaction suggests that
disorganization and scarring. Fibrosis and edema may reduce the number of surviving proximal cell bodies is significantly re-
the neurites' ability to traverse the injury site. The third major duced. Also, very few of the axons capable of regeneration can
factor limiting recovery is the misalignment of endoneurial find appropriate endoneurial tubes to enter because of the phys-
tubes. Axons that do recover sufficiently to penetrate the intra- ical separation between the cut nerve ends and profound en-
funicular disorganization may not be able to relocate their origi- doneural tube misalignment. It is highly likely that the trauma
nal or similar end organ endoneurial tubes. Successful producing the injury will lead to significant scarring, thereby
regeneration across the lesion, therefore, is functionally useless further impeding any axonal regrowth through the injured area.
because of end organ misdirection. Even in the face of excellent surgical repair, the prognosis for
The importance of endoneurial tubes misalignment has optimal functional recovery is limited.
varied consequences depending upon whether the funiculus in-
volved is comprised of different types of fibers or all of the NEUROPHYSIOLOGY CORRELATES OF NERVE INJURY
same fiber type traveling to the same end organ. For example, if
a funiculus contains only motor fibers traveling to foot plantar Following either nerve crush or section, it is important for the
flexors, misdirection of different axons within this funiculus has practitioner to be aware of what can be anticipated with respect
minimal clinical consequences. The foot will still respond to the to neural conduction. The electrophysiologic correlates to
command of plantar flexion. On the other hand, if a funiculus Wallerian degeneration are primarily investigated in animal
contained motor fibers traveling to muscles with different func- models, but nevertheless have direct bearing on what is occur-
tions as well as cutaneous sensory fibers, misdirection of axons ring in humans following nerve injury. Additionally, pertinent
down inappropriate endoneurial tubes can have disastrous clini- findings regarding terminal axon and neuromuscular junction
cal implications. alterations are also noted.
Recovery time in third-degree injuries is longer than that de-
scribed for second-degree injuries primarily resulting from the Neural Crush
axonal reaction of the cell bodies combined with the internal fu- As previously noted, the importance of crushing a nerve is
nicular disruption. Unlike second-degree injuries, the clinical that the endoneurial tubes remain intact even though significant
recovery of motor and sensory function is expectedly less. Also, Wallerian degeneration can occur distal to the injury site. This
muscles that have been denervated longer require more time to affords the nerve the possibility of a complete return to their re-
respond and subsequently recover from denervation. It may be spective end organs. Given that a patient may achieve satisfac-
difficult for reinnervation to occur in the muscle secondary to tory function following reinnervation, it is of interest to examine
intervening connective tissue or fibrous muscle tissue impeding the physiologic status of neural conduction both above and
collateral sprouting. Of importance in third-degree injuries is below the crush site.
the reduced value of the Tinel's sign to demarcate the advancing Once a nerve has been sufficiently crushed to induce
progression of recovery. This is because sensory axons may Wallerian degeneration, one can first explore the effects of the
have been misdirected into inappropriate endoneurial tubes. injury on proximal nerve conduction velocity and the histologic
Thus, there may be a steady advance of the Tinel's sign but this correlates of the nerve damage. That portion of nerve proximal
is to no avail as it may be heading distally in an endoneurial to the lesion demonstrates a reduction in conduction velocity to
tube destined to end in muscle. approximately 90% of normal within 30 days of the injury. - 3 57 8 0
Type 4 Injury. The significance of the fourth-degree By 100 days, the conduction velocity has been further reduced
injury is disruption of the perineurium (Fig. 4-6. Table 4-4). to 80% of its preinjury value. This 80% value is maintained
Wallerian degeneration and loss of neuronal cell bodies occur until about 150 days after neural trauma and then progressively
Disruption of the perineurium results in massive disorganiza- increases to normal within the next 50 days, i.e., 200 days fol-
tion of the internal structure of the peripheral nerve trunk with lowing the original insult. These findings occur only if the re-
significant hemorrhage, edema, and reactive connective tissue generating nerve below the injury site rejoins its appropriate end
proliferation involving multiple nerve fibers. This disorganiza- organ. Should there be failure to re-establish peripheral continu-
tion combined with the results of trauma lead to significant scar ity, the proximal nerve conduction velocity falls to about
formation. Those axons, which eventually recover sufficiently 6 0 - 7 0 % of the original velocity and does not improve above
to attempt regrowth, must penetrate substantial scar tissue and this level even when followed for a period of 400 d a y s . It
1 2 9 1 7 0
appears that even though a significant portion of the injured can be exemplified by a simple example. Let us assume that a
nerve may be spared proximal to the injury, it nevertheless ex- 10-cm segment of nerve has been affected by some disease
periences a reduction in conduction velocity. The slowing of process that has resulted in thinning of the myelin sheath pro-
proximal nerve conduction is explained if one considers the his- ducing an internodal conduction time of 100 ps. We know that
tologic consequences of distal nerve injuries on the more proxi- the normal internodal conduction time is roughly 20 ps, yield-
mal portions of the same nerve. ing a conduction over the 10 cm segment of 50 m/s (20 ps/1
Within the first 150 days after nerve injury, the axonal diame- node x 100 mm x 1 node/1 mm = 2 ms; NCV = 100 mm/2 ms =
ter of the proximal nerves demonstrates an 8.9% reduction in 50 m/s). A conduction time of 100 ps/node results in a con-
246
size while the total fiber diameter diminishes by 5 . 3 % . The 57

duction velocity of 10 m/s (100 ps/1 node x 100 mm x 1 node/1
myelin sheath, however, appears to increase in thickness by ap- mm = 10 ms; NCV = 100 mm/10 ms = 10 m/s). On the other
proximately 5.9%. After 150 days, the axon diameter and total hand, if the damaged myelin is removed and matures with twice
fiber diameter begin to increase so that by 225 days the injured as many nodes of Ranvier and there is restitution of the nodal
and noninjured side are of comparable diameters. There is cer- and paranodal regions, a less than normal neural conduction is
tainly a correlation between fiber diameter and conduction ve- also noted. The conduction velocity over the repaired segment
locity in both normal and the injured nerves. Within the first 150 is expected to approach 25 m/s (20 ps/1 node x 100 mm x 2
days, both conduction velocity and axon diameter decrease. nodes/1 mm = 4 ms; NCV = 100 mm/4 ms = 25 m/s). Note that
After 150 days, axon diameter returns to normal as does nerve the thinning of myelin produces an 80% reduction in conduc-
conduction velocity. tion velocity compared to a 50% reduction resulting from an in-
Interestingly, following both crush injuries and nerve section, creased number of internodes. The alteration in myelin
the internodal length of the regenerated portion of nerve distal thickness can have a much more profound effect on neural con-
to the injury is shortened compared to normal. The normal induction than increasing the number of nodes. This is observed
ternodal distance in large myelinated nerve fibers is roughly be- experimentally in regenerated nerves where the conduction ve-
tween 0.83 mm and 1.3 m m . - Following reconstitution of
154 295
locity approaches the normal conduction velocity. This return
262
the peripheral nerve distal to the injury, the average internodal of almost normal conduction may be the result of an optimiza-
length is approximately 300 pm, i.e., roughly a 2:1 or 3:1 ratio tion of the internodal length (reduced) and the regenerating
compared to normal nerve. The internodal distance of 300 pm is nerve's diameter (also reduced). Of course, during the re-
33
similar to that noted in developing animals when the Schwann myelination process, there is likely to be a combination of both
cells first appear. The Schwann cells are believed to elongate an increased number of internodes and a rather thin myelin
with limb growth to achieve their adult length. In regenera-
172 2 9 5
sheath producing rather profound reductions in conduction ve-
tion of the injured adult nerve, limb growth has ceased and sub- locity. In our example, the NCV would be reduced to 5 m/s
sequently the normal separation of Schwann cells of 300 pm is during myelin regeneration if at some point there was a dou-
re-established, thus accounting for the reduced internodal dis- bling in the number of nodes with an internodal conduction time
tance. It appears, therefore, that in regenerating nerves, the in- of 100 ps. These are of course only theoretical considerations as
ternodal distance is less important than axonal diameter in little controlled long-term information is available regarding
determining nerve conduction velocity. - - 33 262 305
human neural conduction during and after regeneration.
Distal to the site of injury, there is a significant difference in
the eventual conduction velocity compared to the proximal neural Neural Section
segment. The distal conduction velocity eventually reaches ap- Sectioning a nerve leads to somewhat different results than
proximately 6 0 - 9 0 % of the preinjury value. 17.59.90.145,155.171.172 those noted above, particularly when the proximal portion is ex-
Studies in excess of one year fail to demonstrate complete re- amined. Maximum conduction velocities reached in the proxi-
covery of the nerve conduction velocity to normal values. The mal portion of nerves following sectioning and subsequent
primary explanation of this finding is most likely found in the regeneration approached 60-70% of normal, which is consider-
histologic observations of these nerves following regenera- ably less than the 90% or more previously noted. The reduc-
57
tion. Comparison of injured neural diameters with the control

144
tion in velocity is most likely the result of the affected nerves
or noninjured side reveals that there is a distal tapering of the proximal to the transection failing to regain their preinjury di-
nerve. Although the diameter of the nerve within a few centime- ameter. 81126
The profound nature of the insult also may lead to
ters of the injury site is similar to the size of the contralateral significant losses of motor and sensory neurons reducing the
control nerves for that level, the regenerated nerve gradually population of fastest-conducting fibers.
loses size distal to the injury. This axonal atrophy is most likely Assessing nerve conduction velocity distal to the site of nerve
a result of Wallerian degeneration, because as previously de- severance reveals that less than 80% of the normal velocity is
scribed, the endoneurial tube undergoes a considerable reduc- attained even at follow-up times greater than 3 y e a r s . 18
tion in size. There is an obvious failure of the regenerated nerve Histologic examination of the nerve fiber diameters reveals an
to completely re-expand the endoneurial tube. As nerve conduc- interesting finding. Crushing the nerves permits an almost
tion velocity is directly proportional to neural diameter, it is not normal fiber size distribution with a bimodal peak of large and
surprising that the smaller regenerated nerves conduct at slower small fibers. The fiber size distribution after cutting a nerve pri-
velocities than noninjured nerves. This reduction in velocity, marily reveals a large number of small fibers with a signifi-
however, should have little effect on the observed clinical out- cantly reduced population of larger-size nerves. The reduction
come. Complete clinical recovery is expected given that the en- in size is particularly important because the myelin thickness
doneurial tubes maintained their continuity and there is no remains thinner than normal, which may account for the re-
misdirection or failure of nerve regrowth. Also, strength is not duced conduction velocity. The reduction in fiber diameter,
267
dependent on conduction velocity. therefore, most likely results in the failure of a sectioned nerve
The effects on conduction velocity of an increased number of to regain normal conduction velocities. As previously stated, an
nodes following myelin replacement compared to demyelination increased number of internodes also contributes to the amount
of time necessary to convey an action potential over the same rate. It appears that the small myelinated fibers experience the
distance of nerve. above noted sequence of breakdown slightly ahead of the larger
myelinated fibers. Once failure of action potential propagation
DYNAMIC ELECTROPHYSIOLOGIC OBSERVATIONS is noted, there is clearly observed to be axonal fragmentation of
OF WALLERIAN DEGENERATION all fiber sizes. In this sense, failure of neural conduction is cor-
related with loss of axonal continuity, i.e., once the axon is frag-
The previous sections have considered the somewhat static mented, action potential propagation ceases. Also, the failure of
consequences of Wallerian degeneration and regeneration with conduction occurs along the entire nerve length simultaneously.
respect to the eventual clinical correlation to nerve conduction Of scientific interest is the issue of whether or not Wallerian
velocity. Of equal interest is the dynamic failure of both neural degeneration and consequently loss of neural excitability
action potentials and neuromuscular transmission immediately progress in a proximal to distal direction from the site of injury.
following nerve injury. To simplify these observations, the ma- Sectioning a rabbit sciatic nerve and stimulating the nerve at
jority of investigators have used complete nerve section. The two sites distal to the lesion, near and far from the section, while
technique used to assess the evolution of conduction failure in recording the ensuing nerve action potential clearly reveals that
the animal model is to completely severe a selected nerve and the action potential persists longer from the distal stimulus
then activate the distal portion of this cut nerve recording both site. Specifically, by 40-48 hours the proximal nerve action
48
nerve action potentials and compound muscle action potentials. potential is markedly decreased by about 70% while that arising
At selected times, the nerves are removed from a few of the ani- from distal stimulation is significantly larger. The same rela-
mals so that correlative histologic sections of the nerve can be tionship is noted 60 hours following neural section. By 72
made with the electrophysiologic findings. For obvious reasons, hours, both potentials have become difficult to evaluate. This
our knowledge regarding this electrophysiologic and histologic suggests that degeneration is proceeding toward the periphery
relationship is lacking in humans. The information gained from from the lesion s i t e . Histologic evaluation of these nerves
230,257
mammalian species is nevertheless of clinical relevance and will reveals that coincident with failure of conduction along the
be related to what limited information is available in humans. nerve is a progression of myelin retraction from the nodes of
Ranvier. - Early conduction failure (40 hours) as evidenced by
46 47
Nerve Action Potentials nerve action potential amplitude reduction (compared to the
A detailed investigation performed in rabbits allows us to ap- previously noted loss of conduction at 70 hours) is proposed to
preciate the clinical findings following a nerve injury based be the result of the paranodal retraction of myelin leading to
upon the interdependence between the morphologic results of action potential failure at the widened nodes. Subsequent histo-
Wallerian degeneration and the electrophysiologic conse- logic investigations have failed to substantiate the centrifugal
quences of nerve breakdown. Within the first 24 hours, the direction of Wallerian degeneration and this subject remains un-
nerve's conduction velocity as measured directly from nerve settled. It is certainly conceivable that there may be a combi-
71
action potentials, and thus ignoring the muscle's response, is es- nation of failure at the nodes of Ranvier prior to axonal
sentially unchanged. The nerve fibers demonstrate little, if
128
fragmentation depending upon the progress of degeneration in
any, alteration in structure. At 40 hours, the nerve conduction different size fibers of a given nerve and possibly different
velocity continues to be greater than 9 5 % of normal, but the nerves in the same and other animal species.
nerve fibers are beginning to demonstrate irregular outlines and
diffuse swellings. Of importance, however, is that the axon is Nerve-to-Muscle Transmission Failure
still intact as a thin and continuous band of axoplasm can be ob- In addition to action potential propagation cessation, it is also
served. By 48 hours, neural conduction velocity is approxi- necessary to consider neural transmission in the terminal axons
mately 9 5 % of normal and neural continuity continues to be and electrochemical conduction across the neuromuscular junc-
preserved. Observation 60 hours after nerve section reveals the tion following Wallerian degeneration. These processes can be
conduction velocity to be slightly less than 95% of normal, easily examined if similar methodologies to those noted above
while the nerve demonstrates some myelin retraction at the are repeated with the addition of simultaneously recording both
nodes of Ranvier. Axonal continuity is still maintained in all but nerve action potentials and compound muscle action potentials
a few of the smaller-diameter axons that are undergoing frag- from nerve and muscle innervated by the sectioned nerve, re-
mentation. At 70 hours, the conduction velocity is approaching spectively. Again, it is impossible to perform these investiga-
80% of normal and the amplitudes of nerve action potentials are tions in humans; however, subhuman primates (e.g., baboons)
decreasing. The myelin sheath is rather swollen, but ovoids are have been studied and the results most likely parallel what is oc-
not yet observed and there is an increase in the amount of curring in humans following nerve section. When recording
107
axonal fragmentation. Seventy-two hours after neural section, from the muscle innervated by a sectioned nerve, there is a re-
the nerve is no longer excitable (i.e., there is an absent response duction in the compound muscle action potential's amplitude by
to electrical stimulation). Histologically, the nerve demonstrates about 2 1 % at 48 hours after nerve section when stimulating
axonal swellings with significant fragmentation of axons, al- close to the recording site. Four days after the insult, the ampli-
though a few are still observed to be intact. Most of the frag- tude drops to only 3 % of the control value. By 6 days, a muscle
mentation is noted to first occur at the nodes of Ranvier. The response could no longer be observed in any of the experimen-
myelin, however, appears intact at this time despite significant tal animals. When the same nerve is excited at a more proximal
paranodal retraction. Digestion of myelin along with axonal level, the same results are observed. The significance of this
fragments is well underway by 86 hours and the classic descrip- finding is that failure of the motor response occurs below
tion of Wallerian degeneration is noted (see above). (closer to the muscle) the distal point of stimulation, i.e., in the
Several additional observations are of importance during collateral axons within the muscle or at the neuromuscular junc-
Wallerian degeneration. During this process, it is clear that not tion. When the nerve conduction velocities and distal motor la-
all of the myelinated fibers undergo degeneration at the same tencies are examined, it is noted that both parameters essentially
maintain their original values until the response is almost com- Table 4-5. Conduction Failure Distal to Nerve Section
pletely absent. In other words, there is preservation of the Failure Time
fastest-conducting nerve fibers until the time of complete re- Species Nerve (hours)
sponse failure.
Neural Action Potential Failure
In the above study, the nerve action potentials persist approx-
Rabbit 128
Peroneal 71-78
imately 3 days longer than the muscle responses. All electrical
Rat 128
Peroneal 79-81
activity had ceased from muscle tissue by day 6, while ascend-
Guinea pig 128
Peroneal 72-82
ing nerve action potentials could be recorded until day 9. Both
Cat 257
Sciatic 72-101
latency and conduction velocity of the nerve potentials re-
Dog 90
Phrenic 96
mained unchanged from the control data until complete failure
Baboon 107
Lateral popliteal 120-216
of action potential propagation occurred. These findings sub-
Human - 51 234317
- 319
Median, radial 168-264
stantiate the previous preservation of motor conduction until the
Sensory
response has completely disappeared. Histologic observation of
the same nerves at 3 days revealed that the main nerve trunk Muscle Activation Failure
only demonstrated an irregular myelin contour, especially about Rabbit 129
Peroneal 30-32
the paranodal regions. However, there was no widening about Rat 205
Sciatic 24-36
the nodes of Ranvier. Importantly, the intramuscular terminal Guinea pig Sciatic 40^t5
160
axons demonstrated profound myelin fragmentation, particu- Cat Sciatic 69-79

191
larly near the neuromuscular junction. In a number of speci- Baboon 107

Lateral popliteal 96-144
mens, there was complete absence of the axon's terminal 5uo4.i23.234.3i7-3!9
Human d j n , ulnar
F a d a ( m e a 120-216
portion. Within 6 days, the more proximal portions of the nerves Motor
in the leg revealed axonal fragmentation while there was no
trace of the terminal axons within the muscle. There is clear ev-
idence in this preparation to substantiate the previous finding d a y s53.io4j6o.234.3i7.3i8.3i9 T _ 3 days for the
h e r e i s a J a g o f a b o u t 2
that neuromuscular transmission fails prior to action potential sensory response. There is noticeable amplitude loss between
propagation in the proximal portions of the injured nerve days 5 and 7 with disappearance of the sensory response by day
trunk. - -
23 121
• - - -
189,191 197 205 215 261
10 or 11. A complicating factor regarding the interpretation of
There are several additional points of interest regarding mus- side-to-side amplitude comparisons is the inherent variability of
cular activation during Wallerian degeneration. During investi- normal physiologic differences. This parameter has been poorly
gations of the time delay between loss of neural propagation documented and is assumed to be minimal, but in the author's
and muscle activation, there was noted to be a length-dependent experience may reach 30% or more. Further normal side-to-side
relationship between the distance separating the nerve's transec- amplitude ranges need to be recorded prior to fully using the
tion site and the muscle and how long it took for loss of muscle above information.
excitation from neural stimulation. The longer the section of Consideration of the above information explains the typical
nerve between neural section and muscle, the greater amount of observation of obtaining evoked sensory responses for several
time was noted for inducing an action potential in the muscle. more days than the compound muscle action potential (CMAP).
For any given length of transected nerve, there is an additional It is noted that the motor nerve continues to conduct an impulse
45-minute delay between cutting the nerve and that nerve losing even though the neuromuscular junction disintegrates. A poten-
the ability to generate a muscle action potential for each addi- tial can be recorded directly from the motor nerve even though
tional centimeter of length. In other words, terminal intramus-
205
the CMAP may be absent on day 7. This direct motor nerve po-
cular axonal destruction is delayed by 45 minutes per every tential persists until about day 10, which is similar to that for the
additional centimeter of axon length added proximal to the sensory nerve. A sensory potential, therefore, can be recorded
lesion site. The exact reason for this length-dependent sparing is for several more days after the CMAP response is absent b e -
unknown but is postulated to be secondary to a "trophic" influ- cause the technique for recording sensory potentials is one of
ence of some material present in the axon that is transported to essentially a direct nerve response, whereas the motor technique
the periphery. This additional time appears to be related to the relies upon an intact neuromuscular junction.
velocity of fast axonal transport of some unidentified substance. From the above discussion it is evident that a proper classifi-
There is an obvious species difference regarding failure of cation of nerve injuries is very important. The distinction be-
action potential propagation in a nerve undergoing Wallerian tween conduction block and Wallerian degeneration may well
degeneration compared to failure of muscle activation (Table 4¬ be made using electrodiagnostic studies. Classifying a nerve
5). Non-primate species appear to demonstrate preservation of lesion in minimal or intermediate versus severe (Table 4-1), or
nerve action potential propagation between 7 1 - 9 6 hours, in Sunderland's classification between type 1 or type 2 and
whereas primates maintain neural propagation for a time period higher is typically done with the aid of neurophysiologic tech-
approximating 120-264 hours, with humans at the longer aspect niques. However, the distinction between axonotmesis and neu-
of the time spectrum. Loss of muscle excitability due to disinte- rotmesis is an anatomic definition and can never be made on
gration of the intramuscular terminal axons and neuromuscular neurophysiologic grounds. A well-performed classification of
junctions occurs between 30-79 hours in lower mammals. The nerve injuries is thus the result of combining electrophysiologic
same time frame in primates extends to 96-216 hours again data with the clinical context.
with humans taking the most time to demonstrate absence of
muscle activation. Following complete section of a nerve in CLINICAL CORRELATION
man, therefore, one can anticipate loss of the motor response
amplitude to begin by the third to fifth day with absence of a The above findings primarily noted in animal studies are of
compound muscle action potential between the seventh to ninth particular relevance to the investigation of human nerve lesions.
It is important to keep in mind the sequence of events with respect Rest Activity
to loss of motor and sensory evoked potentials to avoid an erro- Muscle PSW/Fibrillation Recruitment
neous diagnosis. An illustrative case may help one to conceptualize (R) APB 0 Normal
the various aspects of a nerve transection regarding electrophysio- (R) Pronator teres 0 Normal
logic findings and the appropriate conclusions to be drawn. (R) Flexor carpi radialis 0 Normal
(R) Extensor digitorum 0 Normal
Case Example (R) First dorsal interosseous
History. A 26-year-old male construction worker sustained (day 2) 0 Absent
a severe blow and laceration to the medial aspect of the right (day 12) 0 Absent
forearm just distal to the medial epicondyle. He complained of (day 15) 2+ Absent
sensation loss along the medial aspect of the affected hand to in- (R) Abductor digiti minimi
clude the fourth and fifth digits immediately following the (day 2) 0 Absent
injury. The patient also complained of difficulty forming a (day 12) 0 Absent
strong grip in the right hand. He denied any previous medical (day 15) 2+ Absent
problems or medication consumption, and had been in a state of (R) Flexor carpi ulnaris
good health prior to the accident. (day 2) 0 Absent
Physical Examination. The patient was examined within 18 (day 12) 1+ Absent
hours of the injury. There is noted to be a complete absence of (day 15) 2+ Absent
sensation to all modalities (touch, pin prick, vibration, and pro-
prioception) in the distribution of the ulnar nerve in the right Surgical Exploration. Surgical exploration 4 weeks follow-
hand to include the volar and dorsal aspects of the fourth and ing injury revealed a complete laceration of the ulnar nerve in
fifth digits as well as the dorsum of the hand. Manual muscle the postcondylar groove.
testing of the hand intrinsic muscles innervated by the ulnar
nerve found them to be 0/5 (adductor pollicis, first dorsal in- Comment
terosseous, abductor digiti minimi, opponens digiti) while the There are a number of clinically relevant aspects to the
median-innervated muscles were 4/5. The flexor digitorum pro- above-noted electrophysiologic portion of the electrodiagnostic
fundus to the fourth and fifth digits 0/5. All remaining muscles medicine consultation. If the patient had been examined within
of the right upper limb were 5/5 and the sensation to the remain- the first 4 days following injury, the ulnar CMAP amplitude
der of the hand was normal. Deep tendon reflexes to the biceps would have been normal following stimulation at the wrist, but
brachii, triceps, pronator teres, and brachioradialis was 2+/2+. CMAPs would not have been obtainable following stimulation
There was noted to be clawing of the fourth and fifth fingers. proximal to the lesion site. In addition, electromyography
Nerve Conduction Studies. Nerve conduction studies were would reveal no recruitable motor unit action potentials in the
performed in the upper limbss bilaterally. The mid-palm tem- ulnar-innervated hand intrinsic muscles. It is impossible at this
perature was 32.5°C on the right and 33.0°C on the left. point to assess whether there is a partial or complete transection
of the ulnar nerve with a component of conduction block or if
DSL S Amp DML M Amp NCV
there is a profound conduction block with minimal axonal
Nerve (ms) (MV) (ms) (mV) (m/s)
injury. If there had been an inability for neural conduction to
Right median 3.1 50.0 3.5 7.0 60.0 propagate across the injury site because of a conduction block
Right ulnar (below injury) (neurapraxia), the findings on days 2-4 would be anticipated.
(day 2) 3.0 20.0 2.9 6.0 62.0 The decreased ulnar motor amplitude to less than 50% of the
(day 4) 3.0 20.0 3.0 2.9 61.0 contralateral side on day 4, however, is suggestive of an axonal
(day 6) 3.2 18.0 3.2 1.2 58.0 injury. There is no confirmatory evidence of a specific
84,92158
type of injury on needle electromyography at this point. The

(day 8) 3.4 10.0 Absent — —
sensory studies are essentially normal and complimentary to the
(day 10) 3.5 6.0 Absent — —
unaffected side. Note that the distal motor latency and nerve
(day 12) Absent — Absent — —
conduction velocity are well within acceptable normal values.
Right ulnar nerve (above injury)
By day 6, the amplitude of the ulnar motor response to stimu-
(day 2-12) Absent — Absent — — lation below the injury is now clearly abnormal and certainly
Right radial 2.9 25.0 2.1 5.0 72.0 implies that there has been a profound injury to the motor fibers.
Left ulnar 2.9 22.0 3.0 8.0 62.0 Note that the distal motor latency and nerve conduction velocity
are still well preserved. Again, the sensory response appears
Left median 3.2 45.0 3.3 8.5 58.0
quite normal and one may believe that the patient has only sus-
DSL, distal sensory latency; S Amp, sensory amplitude; tained a partial nerve injury with sparing of the sensory fibers.
DML, distal motor latency; M Amp, motor amplitude; NCV, On day 8, there is no longer an obtainable motor response for
nerve conduction velocity; ms, milliseconds; pV, microvolts; the affected ulnar nerve. The distal motor latency and conduc-
mV, millivolts; m/s, meter/second. Motor and sensory ampli- tion velocity were quite acceptable until the response was no
tudes are measured baseline-to-peak. Sensory latencies are mea- longer detectable. As noted previously, this implies that there
sured to peak while motor latencies are measured to initial was sparing of the large and fast-conducting fibers until the
negative onset. compound muscle action potential completely disappeared.
Recall from the previous section that the reason for this obser-
Needle Electromyography. A needle electromyographic in- vation is most likely a result of disintegration and fragmentation
vestigation was performed on the left upper limb using a dispos- of the terminal intramuscular axon branches and neuromuscular
able monopolar needle. junctions. It appears that humans undergo phases of neural
response to injury similar to the lower primates. Remember that illustrates what can be expected once a patient has sustained a
the main trunk of the nerve is still excitable. This can be demon- peripheral nerve injury. Examining the patient too early can lead
strated by performing a mixed nerve response of the ulnar to a misdiagnosis. As noted above, there is usually useful infor-
nerve. Activating the ulnar nerve at the wrist and recording from mation to be gained by day 10, but the most complete data can
just below the injury would have yielded a small but detectable be gathered by about day 15. Of course, the surgeon may wish
mixed nerve response. If the ulnar nerve had been injured just to acquire neurophysiologic data prior to that time and this is
above the wrist instead of the elbow, it is highly likely that the certainly acceptable provided the practitioner is aware of what
drop in motor amplitude would have been noticed several days the gathered data means with respect to the natural progression
earlier. Recall that the preservation of the evoked motor re- of Wallerian degeneration and the electrophysiologic manifesta-
sponse is dependent upon the length of nerve between the site tions at each stage of neural disintegration.
of severance and the muscle.
Beginning on day 8 and beyond, the ulnar sensory response
now displays a drop in amplitude compared to the unaffected MINIMAL NEURAL INJURY
side. The sensory latency remains normal, however, and sug-
gests that there is preservation of the large and fast-conducting The term minimal injury is used in the context of an insult to
fibers. This response persists beyond that of the motor because the peripheral nervous system that produces a temporary and
of a lack of a neuromuscular junction. We are recording a nerve completely reversible failure of action potential propagation
action potential, which is known to last several days longer than over a well localized neural segment (Table 4-1). It is to be un-
the motor response because of the previously noted destruction derstood that there are no structural alterations in the axon, en-
in the terminal muscular axons. veloping myelin sheath, or supporting connective tissue
Within the first 12-15 days, the needle electromyographic ex- structures. Once the insult is removed, there is a relatively rapid
amination only reveals abnormally decreased recruitment. This return of neural function to the preinjury condition with normal
finding is certainly consistent with either a complete conduction clinical function.
block, partial axonal loss and profound conduction block, or The term conduction block refers to the inability of an action
complete axonal transection. The presence of membrane insta- potential to propagate beyond a specific region of nerve. Neural
bility is only manifest quite some time after the nerve conduc- conduction can fail for a number of reasons. Following nerve
tion studies have clearly demonstrated a profound abnormality. transection, an action potential is incapable of conducting
It is certainly important to find signs of muscular denervation. If across the transected nerve (see Wallerian Degeneration). If a
the motor and sensory nerve conduction studies had remained traumatic insult to a nerve over a particular location is of suffi-
normal through day 15 and beyond with respect to amplitude cient force to result in disruption of the axon and investing
and the needle examination demonstrated continued absence of myelin sheath with distal disintegration of the axon, the nerve
membrane instability and voluntary motor units, conduction can no longer sustain action potential propagation distal to the
block is the likely conclusion. With the same nerve conduction injured site. Loss of myelin over a localized segment with com-
studies but membrane instability, there is now the distinct possi- plete preservation of axonal continuity can also result in action
bility of a mixed lesion with both axonal loss (probably mild potential blockade. Compression or its induced ischemia can
because of motor amplitude preservation) and conduction result in loss of action potential propagation. This is perhaps the
block. The observation of an absent or markedly reduced most familiar type of conduction block experienced by most in-
CMAP amplitude to stimulation of a peripheral nerve distal to a dividuals at some point following the crossing of one's legs. It is
lesion strongly implies that there has been significant Wallerian rather easy to compress the common peroneal nerve about the
degeneration. When describing the impression of Wallerian de- fibular head by crossing one leg over the other. The induced
generation, it is inappropriate to use the terms axonotmesis or compressive ischemia causes the peroneal nerve to experience
neurotmesis as these are histologic terms and not neurophysio- conduction failure secondary to anoxic disruption of the meta-
logic descriptions. The above electrophysiologic tests cannot bolic processes responsible for action potential propagation
distinguish between disruptions of specific supporting connec- over the affected segment. Relief of the compression within sev-
tive tissue structures. Both axonotmesis and neurotmesis appear eral minutes quickly restores the ability of the nerve to again
the same with respect to nerve conduction studies or needle conduct impulses. The clinical description of an inability to
electromyography. contract the muscles innervated distal to the site of compression
It can be seen from the above example that following an or sense cutaneous stimuli is a result of this temporary conduc-
injury, the earliest sign of Wallerian degeneration is reduction in tion block. Temporary compression of a limb induced by a
the compound muscle action potential. One must be careful, sphygmomanometer cuff serves as the model to explore mini-
however, as there is a wide range of normal with respect to mal neural injury and the transient effects of conduction block
motor amplitude and it is a good idea to compare the affected resulting from ischemia.
with the unaffected side. If a side-to-side amplitude difference
approaches 30^4-0%, it is safe to assume that the patient has sus- TEMPORARY NEURAL ISCHEMIA
tained an axonal lesion provided excitation occurs below the
lesion site. A repeat study within 10-15 days is indicated to con- In an attempt to quantify and simulate the clinical effects of
firm the original suspicion with an absent or reduced sensory re- compressing one's peripheral nerve, a blood pressure cuff can
sponse and membrane instability on intramuscular needle be placed about an elevated a r m . -
103
The pressure in the cuff
190315
examination. It is important not to be confused by the progres- is then increased well above systolic blood pressure to approxi-
sion of Wallerian degeneration. There is a natural sequence of mately 180 mmHg or more. The arm is then comfortably rested
events to be expected following neural transection with respect in a warm water bath and the subject is requested to describe the
to the nerve conduction and needle electromyographic portion ensuing results. Within 13-15 minutes there is noted to be a re-
of the electrodiagnostic medicine consultation. The above case duction in sensation involving the tips of the fingers with the
second and third digits affected first. The first and fourth digits while older subjects (61-82 years) reveal smaller increases in
then experience decreased ability to perceive superficial tactile duration. In addition to the above-noted parameters, sensory
stimulation. Finally, the fifth digit succumbs to the above-noted nerves also demonstrate prolongation of the refractory period
inability to detect sensation at the precompression level. By 17 and elevation of the stimulus threshold required to obtain a re-
minutes, the palm of the hand no longer sustains its previous sponse. Upon release of the compressive force, the amplitude
105
level of sensibility. Within 1-2 minutes of the hand's altered regained 7 5 - 8 5 % of its original value within the first 2 minutes.
ability to detect tactile stimulation, the fingers are beginning to The increase subsequently diminishes such that the response re-
undergo complete anesthesia with respect to touch sensation. gains its preischemic value only after 30 minutes following re-
The sequence of decreased perception followed within several lease of the blood pressure cuff. The remaining parameters
minutes by total absence of tactile sensibility progresses proxi- noted above were not examined in detail.
mally at a rate of 3 - 4 cm/minute. Approximately 30 minutes The above-noted changes in neural amplitude, velocity, and
after the initiation of arm compression, the entire limb distal to duration have been observed in single nerve fibers subjected to
the compression is anesthetic to touch. The loss of immediate anoxia. The amplitude of the single-fiber spike progressively
326
recognition of a painful stimulus occurs within several minutes diminishes until it is no longer detectable following oxygen de-
to that of tactile sensation and lags behind the advancing front privation. This continued decrement is believed to be related to
of decreased touch perception by 15-20 cm. Even after 40 min- the progressive decline in the axonal resting membrane poten-
utes of compression and complete absence of touch sensation, tial. A commensurate decline in the single fiber conduction ve-
pain can still be perceived, but the time of cognitive recognition locity is noted with the diminishing amplitude of the evoked
is delayed. Approximately 25 minutes of compression are re- potential. This conduction velocity decrease results in an in-
quired for motor power of the hand intrinsics to become absent creased temporal dispersion of the nerve's composite fiber ve-
or profoundly decreased. This absence of motor function corre- locities, thus contributing to the decline in the summated
sponds to the time of complete disappearance of touch percep- potential observed. Action potential propagation can even cease
tion in the hand. By 30 minutes, the hand extrinsics are during anoxia despite the amplitude being 7 5 % of the preanoxic
completely paralyzed. Following release of the cuff, the order value. The sensory nerve action potential observed with elec-
326
of functional recovery is the reverse. Altered touch sensation re- trical excitation of the nerve during an ischemic episode is,
covers first in the arm followed by the forearm, hand, and fi- therefore, the summation of single nerve fibers demonstrating
nally fingers. This process of recovery requires approximately individual responses to the ischemic conditions.
2 5 - 5 0 seconds for a compression time of 30-35 minutes. Histologic examination of animal nerves subjected to is-
Unlike sensation, however, motor power requires about 17 min- chemic conditions for periods of up to 4 - 6 hours at 2 5 0 - 3 0 0
utes to regain complete precompression strength. mmHg did not result in detectable structural d a m a g e . - 112 218
It is also possible to repeat the above experiment and record Neither demyelination nor axonal loss with Wallerian degenera-
the sensory nerve action potentials to observe the electrophysio- tion could be observed. The production of ischemia for this time
logic effects of limb ischemia. - - - ' Specifically, ortho-
43 45 213 27
period, however, may result in muscular and subcutaneous
dromic sensory nerve action potentials from the third digit to tissue edema secondary to alterations in capillary permeabil-
wrist can be serially elicited while a cuff is inflated, maintained ity. - Ischemia approaching 8 hours can result in significant
320 321
above systolic pressure, and subsequently deflated. For individ- neural injury. 195
uals 30-59 years of age, there is a 3.2% decrease in nerve con-

duction and a 7.3% amplitude decrement within the first 5
minutes of suprasystolic compression. At 10 minutes, a 5.9% INTERMEDIATE NEURAL INJURY
decrease in maximum conduction velocity and a 9.6% decrease
in the preischemic amplitude are observed. By 15 minutes, a An intermediate type of insult to the peripheral nervous
9.5% and 14.6% decrease in velocity and amplitude, respec- system may be defined as one in which an insult, commonly
tively, are apparent while at 20 minutes the decrease in NCV compression, is of a degree sufficient to result in failure of
and amplitude reaches 15.5% and 27.9%. Twenty-five minutes action potential propagation (conduction block) but not
of arterial occlusion generates a reduction in NCV and ampli- Wallerian degeneration (Table 4-1). The patient typically com-
tude of 22.2% and 50.4%, respectively, and at 30 minutes complains of weakness and sensory loss. Recovery is relatively |
parable NCV and amplitude values of 2 7 . 1 % and 6 6 . 1 % rapid compared to Wallerian degeneration but somewhat longer
reduction are obtained. For the same time periods, individuals than an acute conduction block arising from a brief ischemic
10-28 years of age demonstrate reductions that are generally a episode.
few percent larger, while subjects 61-82 years of age reveal re- This type of intermediate neural lesion was initially described
ductions several percent less. There is an apparent resistance of in 1876 and later referred to as neurapraxia by Seddon. The
89 269
older persons' nerves to ischemia. Compared to digit-to-wrist, pathologic basis of this neural insult became apparent when his-
wrist-to-elbow sensory conduction velocities demonstrate larger tologic investigations of nerves subjected to various degrees of
percent NCV reductions for comparable time periods up to 15 compressive forces were performed in the cat. Following 2
65
minutes for respective age groups. Beyond 15 minutes, re- hours of hind limb compression with pressures between 800 and
sponses could generally no longer be observed across this seg- 1200 mmHg, weakness of the musculature distal to the com-
ment because of profound amplitude reduction. As noted above, pression was noted to persist for several weeks prior to com-
however, wrist stimulation continued to reveal a response de- plete recovery. This time period was far too short for Wallerian
spite the lack of demonstrable potentials from elbow stimula- degeneration to have been the cause of paralysis. Longitudinal
tion. Over the same time frame, the duration of the sensory histologic preparations of the compressed regions demonstrated
nerve action potential incrementally increases for persons that there had been focal or segmental demyelination/remyeli-
30-59 years of age from 3.7% to 101.8% of the preischemic nation located beneath the compressive device only with axonal
value. Younger individuals (10-28 years) demonstrate a larger, sparing, i.e., an absence of Wallerian degeneration. This type of
Chapter 4 PERIPHERAL NERVOUS SYSTEM'S REACTION TO INJURY — 133
segmental demyelination had been described earlier in periph- SI S2 S3

eral nerve lesions caused by lead intoxication. The documen-
115
tation of segmental demyelination was extremely important

because it related a specific type of histologic nerve injury (seg-
mental demyelination) to physiologic conduction block result-
ing in clinical weakness in the absence of Wallerian
degeneration. It also focused attention on a relatively common
type of reversible weakness: acute demyelinating block.
Continued work in compressive segmental demyelinating le-
sions has shed considerable light on the underlying anatomic
changes arising from localized pressure that produce action po-
tential blockade with the associated characteristic clinical find-
ings of weakness and numbness.
ELECTROPHYSIOLOGIC FINDINGS
Reports of patients developing profound weakness and
numbness following the application of a compressive bandage
or tourniquet generated a number of interesting investigations in
which compressive lesions were induced in subhuman primates
like the baboon. ' - It was hoped that subhuman primate
93 109 260
nerves would respond to experimental compression in a similar

manner to pathologically compressed human nerves. We may
begin by first considering the electrophysiologic effects of
neural compression and then explore the anatomic basis of con- Figure 4-7. Neural compressive effects. Evoked muscle action
duction failure. potential from a b d u c t o r hallucis muscle at different intervals after a
Nerve conduction velocity was first investigated in the com- tourniquet was inflated t o 1,000 mmHg a b o u t t h e knee for 95 min-
pressed nerves of cats that had undergone the segmental de- utes. T h e corresponding sites of stimulation and recording are d e -
myelination. The conduction velocity above and below the picted. (From Fowler TJ, Danta G, Gilliatt RW: Recovery of n e r v e
region of nerve with segmental demyelination was normal. 201
conduction after a pneumatic tourniquet: Observations on the hind-

Across the segmentally demyelinated region of nerve, however, limb of t h e baboon. J Neurol Neurosurg Psychiatry 1972;35:638-647,
the velocity of action potential propagation was significantly with permission.)
slowed. This type of finding was explored in greater detail with
particular emphasis on recovery in the baboon. In these ani-
93
mals, the medial popliteal nerve of the sciatic nerve was com- potential's amplitude proximal to the lesion and expressing it as
pressed with a tourniquet applying 1000 mmHg of pressure for a percentage of the distal amplitude allows one to follow the in-
time periods ranging between 1 and 3 hours. All nerves com- jured nerves' rates and durations of recovery (Fig. 4 - 8 ) .
pressed for 1 hour or longer demonstrated evidence of action Conduction block was found to last up to 4 - 6 months prior to
potential blockade. Conduction block was defined as a normal achieving recovery. The recovery rate and extent was quite vari-
motor response induced by stimulating the affected nerve distal able for individual animals but a general trend became apparent.
to the lesion, but there was either a reduced amplitude or com- Longer times of compression resulted in more severe lesions as
plete response absence when exciting the nerve proximal to the judged by longer recovery times and evidence of some
tourniquet (Fig. 4-7). Measuring the compound muscle action Wallerian degeneration. Compressive forces for 3 hours generally
B27*(©C0
Figure 4-8. Recovery of various animals
after conduction block generated by a tourni- 28
quet. The muscle action potentials are recorded 180)
from the abductor hallucis muscle with neural stim-

ulation at the thigh and at t h e ankle. The vertical
scale shows the amplitude of t h e muscle response
to proximal stimulation as a percentage of t h e re- 833
sponse t o distal stimulation.The horizontal scale is (180)
in days after tourniquet application. The animal

number and duration of tourniquet application in
minutes is shown for each nerve. (From Fowler TJ,
Danta G, Gilliatt RW: Recovery of nerve conduction
after a pneumatic tourniquet: Observations on t h e
hind-limb of the baboon. J Neurol Neurosurg
Psychiatry 1972:35:638-647, with permission).
TIME IN D AYS
resulted in 6 months of recovery time and more Wallerian de- of the faster-conducting fibers may have been preferentially
generation than nerves exposed to 1 and 2 hours of compres- slowed with conduction block of other fibers. The continued
sion. It can be seen that mild lesions are primarily composed of demonstration of nerve conduction slowing despite resolution
fibers demonstrating conduction block while severe lesions are of conduction blockade strongly suggests that the fastest-con-
mixed. A mixed lesion is one in which there is a combination of ducting fibers had regained the ability to sustain an action po-
both action potential blockade and Wallerian degeneration. In tential across the damaged segment but at a much reduced
other words, longer times of neural compression can be ex- conduction velocity. The process mediating action potential
pected to generate lesions with a spectrum of nerve injury from propagation had obviously been restored but at a less than
simple action potential blockade with full recovery to Wallerian normal level. Unfortunately, the time required to achieve a max-
degeneration requiring quite prolonged recovery times. imal conduction velocity was not explored. The maximum ve-
Wallerian degeneration was confirmed by the demonstration of locity following compression was not achieved at 6 months and
fibrillation potentials with standard concentric needle explo- may or may not have required more time. Because the ampli-
ration in the affected muscles. tude of the response immediately following compression was
In addition to the complete action potential blockade across smaller and conducted appreciably slower, it is likely that a sig-
the compressed zone, a number of additional electrophysiologic nificant number of the fast-conducting fibers were blocked. One
findings were noted. The nerve conduction velocity range be-
93
can only conclude that a different fiber population may have
neath the tourniquet prior to pressure elevation was 60.4-84.5 been examined with proximal and distal stimulation. As a result,
m/s (mean of 70.3 m/s) while that distal to the tourniquet was the practice of calculating conduction velocities across regions
58.8-80.0 m/s (mean of 68.7 m/s). Following inflation of the of conduction block is of questionable merit as one is compar-
tourniquet for the above-noted times and subsequent deflation, ing the fastest-conducting fibers below the lesion with slower-
the conduction velocity under the tourniquet was initially conducting fibers proximally. The amplitude, however, can be
13.8-44.0 m/s (mean of 29.0 m/s) while that in the noncom- used to gain a rough approximation of the amount of blocked
pressed segment remained essentially unchanged (56.7-79.0 fibers keeping in mind that severe lesions most likely have re-
m/s; mean of 65.8 m/s). Nerve conduction velocity assessment sulted in some axonal loss. After about 7-10 days, expressing
when the nerve had been determined to be recovered (absence the proximal amplitude as a percentage of the distal amplitude
of conduction block: 28-180 days) reveals the proximal con- should eliminate Wallerian degeneration as a contributing factor
duction velocity range had increased but was still below pre- and primarily yield information regarding conduction block.
compression values at 44.3-62.5 m/s (mean of 55.1 m/s), while The reason Wallerian degeneration no longer contributes to the
the distal velocity was within acceptable variation of the pre- distal potential is because the neuromuscular junctions have dis-
compression values, i.e., no change. integrated and these fibers are incapable of contributing to the
Importantly, the conduction velocity through the compressed evoked muscular response to either proximal or distal neural ex-
segment of nerve was dramatically slowed and showed an in- citation. The same theoretical principles apply to sensory nerves
complete recovery, even though conduction block had been de- with the exception that approximately 10 days are required for
termined to be resolved. This suggests that initially there may sensory nerves undergoing Wallerian degeneration to no longer
have been blockade of the fastest-conducting fibers with the be excitable.
slower fibers conducting across the lesion. Additionally, some The morphology of the compound muscle action potential
following compression revealed an interesting change com-
pared to the precompression shape (Fig. 4-9). Prior to compres-
sion, the evoked compound muscle action potential was the
expected biphasic, initially negative waveform. As the potential
recovered from compression, it was no longer a simple biphasic
waveform but appeared with multiple phases. As time pro-
gressed and recovery proceeded, the morphology of the wave-
form again assumed the anticipated biphasic response. This
finding was explained on the basis of altered conduction veloci-
ties of the recovering fibers leading to an initial temporal dis-
persion. As the fiber population assumed a more normal
distribution, the arrival of action potentials at the muscle
became more synchronous, thus generating a biphasic-appear-
ing potential. The onset latency of the recovering waveform also
demonstrated a progressive shortening with recovery.
When nerve action potentials were recorded and compared to
the pure motor response, it was observed that the afferent and ef-
msec ferent nerves wefe affected equally by the compressive forces. 93
Thus, there appears to be no preferential injury to sensory or

Figure 4 - 9 . C M A P c o n d u c t i o n d e l a y / t e m p o r a l d i s p e r s i o n . motor nerves to compression. The demonstration that there is
Evoked muscle action potentials shown from the abductor hallucis mus- electrophysiologic blockade of the action potential propagation
cles before (A) and 73 days after (B) t o u r n i q u e t application at 1,000 across the site of neural compression and slowing of nerve con-
mmHg for 180 minutes. N o t e the prolonged time of conduction and duction velocity within 24 hours of injury is incompatible with
temporal dispersion for the action potential following trauma. (From demyelination being responsible for these findings because the
Fowler TJ, Danta G, Gilliatt RW: Recovery of nerve conduction after a demyelinating process requires several days. The explanation for
pneumatic tourniquet: Observations on the hind-limb of the baboon. J these observations is found in the histologic appearance of
Neurol Neurosurg Psychiatry 1972;35:638-647, with permission.) nerves subjected to the above-noted compressive forces.
ANATOMIC FINDINGS Changes in nerves compressed by tourniquet
In the animal model of neural compression in baboons, longi-

tudinal histologic sections were performed on the injured nerves Proximal
with both light and electron microscopy. - - The blood
217 2 1 8 219 221
pressure cuff covered an expanse of nerve approximating 5.5

cm and histologic examination was carried out just distal,
across, and immediately proximal to this segment of nerve.
Careful examination of single teased nerve fibers were per-
formed within the first several days to several weeks following
compression. Only large myelinated fibers were affected and
demonstrated a unique lesion not previously observed.
Specifically, a mild form of anatomic distortion occurred where Figure 4-11. Region of neural intussusception occurring only
there was invagination of one paranodal region into the adjacent at the edges of the pneumatic tourniquet. Arrows indicate di-
paranodal segment (Fig. 4-10). The extent of this invagination rection of invagination of adjacent paranodes. (From Ochoa J, Fowler
TJ, Gilliatt RW: Anatomical changes in peripheral nerves compressed
by a pneumatic tourniquet.J Anat 1972; I 13:433—455, with permission.)
or intussusception varied from 10 pm in mild lesions up to 300

pm in more profound compressive injuries. These anatomic dis-
tortions were noted to occur in only two regions, i.e., immedi-
ately beneath the proximal and distal edges of the pneumatic
tourniquet (Fig. 4-11). The segment of nerve beneath the mid-
portion of the tourniquet appeared completely normal. This
finding correlated with the observation that nerve conduction
was preferentially slowed beneath the edges of the tourniquet
but was relatively spared in the central portion of compressed
nerve. The significance of these invaginating lesions is that
112
they are the likely etiology of nerve conduction blockade and/or

slowing noted immediately following compression when d e -
myelination has not yet occurred in compressive but not is-
chemic insults. It is postulated that greater than 20 pm of
invagination is required before nerve conduction is affected. 1,2
Once this value is reached or exceeded the anatomic distortion

results in slowing of action potential propagation with mild in-
tussusception and actual blockade following profound intern-
odal disruption. The physiologic mechanism responsible for
slowing and blockade is unknown but may be related to the
physical covering of the invaginated nodal region by the overly-
ing paranodal myelin of the adjacent node. Since this can occur
over several sequential nodes of Ranvier, the action potential
may simply be prevented from exciting the axon as there is no
longer exposed nodal membrane. There is obvious disruption of
the paranodal myelin and this may result in enough current
leakage to delay the spike of the action potential, thus prolong-
T
ing conduction (see below). The proposed mechanism of paran-
nodal end loops of myelin odal invagination is of interest to the physiologic and clinical
consequences of this type of intermediate neural injury.
S c h w a n n cell processes
Increasing the pneumatic tourniquet to 1000 mmHg for one
Figure 4-10. Neural intussusception. A, Microdissection of a or more hours generates a pressure gradient at the proximal and
baboon's normal myelinated tibial nerve fiber demonstrating a node of distal edges of the cuff. The pressure in the cuff's mid-region is
Ranvier. B,An abnormal baboon tibial nerve following acute compres- very high, but uniformly distributed over the 5 cm or so of the
sion with clear evidence of intussusception from right t o left occluding tourniquet's length. At the two edges, however, there is a pres-
the nodal gap region.The indentation of the nerve just t o the right of sure differential, i.e., high intra-axonal pressure beneath the cuff
the intussusception is the pseudonode and marks the previous node of and the relatively low intra-axonal pressure in the nerves both
Ranvier.The horizontal bar represents 10 pm for both nerves. C,A low- proximal and distal to the cuff. At this pressure transition zone
power longitudinal electron micrograph showing an intussusception at the axoplasm is "squeezed" from the high to low pressure re-
t h e Schwann cell junction (J) with the new location for the node of gions. Because of the natural narrowing in axonal caliber for the
Ranvier (N) beneath the myelin folds. D, Diagrammatic representation large myelinated axons at the node of Ranvier, a barrier or dam
of t h e invaginating paranode into an adjacent o n e . (From O c h o a J: is formed offering resistance to the rapid flow of axoplasm. This
Nerve fiber pathology in acute and chronic compression. In O m e r GE, "back pressure" generated at the paranodal region tends to force
Spinner M (eds): Management of Peripheral N e r v e Problems. the paranodal region closest to the tourniquet's edge into the ad-
Philadelphia,W.B. Saunders, 1980, pp 487-501, with permission). jacent paranodal region, thus producing an invagination of one
paranode into the next. The invagination process is noted at both entire internodal segments are demyelinated but more com-
edges of the tourniquet with the invagination pointing away monly it is only the affected portion of the internodal myelin
from both edges (Fig. 4-11). The actual node of Ranvier is thus that is removed. In the lesion produced by compression main-
buried beneath the paranodal myelin of the furthest of the two tained for periods of 3 hours, Wallerian degeneration between
paranodes from the pressure gradient. In order for the paranodal 0-30% was noted in some of the fibers that corresponded to the
myelin to invaginate into the adjacent node, it must be elongated electrophysiologic findings of reduced compound muscle action
as if one were to roll up a single sheet of newspaper and pull on potentials below the lesion site compared with the precompres-
the inner most portion to extend it from the inside outward. This sion amplitude. 93
stretching of myelin disrupts its attachment to the axon at the The process of remyelination repairs the altered portions of
paranodal loop region. The paranodal myelin being invaginated the internodal regions. The Schwann cells apparently undergo
buckles and envelops the invaginating segment. Interestingly, mitosis and remyelinate the newly demyelinated segments, gen-
the Schwann cells do not dislocate but maintain their original erating intercalated thinly myelinated segments (Fig. 4-12). It is
positions. It appears that the Schwann cells are more firmly an- anticipated that these segments eventually regain the preinjury
chored to the surrounding basal lamina and investing connect- myelin thickness although extended observations were not per-
ing tissue. This combination of inner myelin sheath slippage formed. Of interest is the observation in some of the compressed
and stable Schwann cell positioning creates the appearance of nerves that remyelination had not occurred 3 months after injury.
nodes of Ranvier at the sites of the compressive insults. These There was a subpopulation of fibers following compression with
areas are referred to as "pseudonodes," because there is an en- evidence of intramyelin and periaxonal edema, giving the ap-
compassing myelin sheath as opposed to bare axons at true pearance of a grossly swollen myelin sheath separated from but
nodes of Ranvier (Fig. 4-10). surrounding a shrunken axon (Fig. 4-13). It appears that these
Within 7-14 days of the compression, there is paranodal de- swollen nerve segments are eventually invaded by macrophages,
myelination of both the stretched and invaginated myelin seg- which dispose of the edematous material and restore the nerve
ments (Fig. 4-12). By 15 days, there is usually a complete absence segment to a functional level. This prolonged neural dysfunction
of invagination with only demyelination noted. Occasionally, the is proposed to be the etiology of the conduction block, which
lasts for 3-4 months when one would have anticipated neural
repair and functional conduction as opposed to continued action
A potential blockade. ' - The lesion produced beneath the edges
2 7 218
of the pneumatic tourniquet should be viewed as a series of indi-

vidual paranodal regions, each with its own potential for a vari-
able recovery time. As a result, the affected nerve cannot become
physiologically functional until all of the paranodal regions have
been repaired. One or two "delayed recovery" nodes are all that
is necessary to render this particular nerve nonfunctional. In
short, with respect to remyelination a weak link analogy can be
postulated in that " . . . a remyelinated fiber is only as functional
as its most severely affected internode."" 2
An important finding of the myelin intussusception phenom-

enon is that it occurs only in relatively large myelinated fibers,
diameters greater than 5 pm, with sparing of the smaller myeli-
nated fibers. This is significant because it explains the finding
of absent motor function and some sensation, but the preserva-
tion of pain and temperature modalities, i.e., those sensibilities
mediated by the small myelinated fibers. It is postulated that
considerably more force is required to displace the contents of
small axons compared to larger ones. Additionally, there is min-
imal, if any, narrowing in the smaller axons in the paranodal
region. The smaller axons would most likely become crushed
prior to demonstrating the above invaginations so characteristi-
cally noted in the larger myelinated fibers.
Using a small nylon cord to apply a compressive force about
a restricted portion of nerve in subhuman primates revealed
findings similar to those of the tourniquet but that were much
Figure 4 - / 2 . A s e r i e s of single t e a s e d n e r v e fibers. A, Normal more localized. Additionally, there was a direct correlation to
260
nodal and paranodal region of a single nerve fiber. B, Mild degree of in- anatomic injury, clinical findings, and electrophysiologic results
vagination of one paranodal segment into an adjacent one. C - D , to the severity of pressure per unit area. This model of anatomic
Progressively more severe intussusception compared to B.There is thin- disruption of the paranodal myelin is postulated to be the mech-
ning of the myelin sheath in D. E, Demyelination of the affected paran- anism responsible for the paralysis and sensory findings noted
odal region. F.An example of early remyelination with the formation of a in human pressure palsies, such as "Saturday night palsy." It is
so-called intercalated segment involving only a portion of the internodal rather difficult to prove this point in humans as one would have
region that was invaginated. (From O c h o a J, Fowler TJ, Gilliat RW: to expeditiously investigate histologically known individuals
Changes produced by a pneumatic tourniquet In Desmedt JE (ed): New with acute pressure palsies who had coincidentally expired.
Developments in Electromyography and Clinical Neurophysiology,Vol. 2. One may well ask if nerve fibers subjected to segmental
Basel, Karger, 1973, pp 174-180, with permission.) demyelination without loss of axons demonstrate fibrillation
potentials on needle electromyography, i.e., can one see fibril-

lation potentials during prolonged conduction block? In ba-
boons demonstrating prolonged conduction there were
fibrillation potentials only in those muscles where Wallerian
degeneration had occurred." The only abnormalities noted in
1
muscles with prolonged conduction block were: (1) an increase

in the extrajunctional acetylcholine sensitivity that was less
than in denervated muscles, and (2) an increase in insertional
activity after 1-2 weeks to concentric needle electromyo-
graphic exploration of affected muscles (Table 4-2). Contrary
to this observation are the documentation of fibrillation poten-
tials in rat muscle fibers innervated by nerves experiencing
conduction block but to a lesser degree than denervated mus-
cles (7.7% vs 17.6% of fibers examined). These 7.7% of fib-
42
rillating yet blocked fibers were confirmed to be blocked and

not denervated because they responded to neural electrical
stimulation, whereas the 17.6% fiber population did not. Of
note was the observation that the resting membrane potential in
normal muscle fibers was 85.5 mV, while the blocked and den-
ervated muscle resting membrane potential was 74.3 mV and
67.9 mV, respectively.
The above-noted contradiction may be accounted for by
species differences and recording techniques. The absence of
fibrillation potentials in baboons compared to rats may be ex-
plained by lower mammalian species' muscular or neural tis-
sues responding differently than human or subhuman primates.
Additionally, recordings in the baboon were made extracellu-
larly with concentric needle electrodes as opposed to the intra-
cellular microelectrode recordings in the rat. In humans,
fibrillation potentials have been claimed to occur following le-
Figure 4-13. Anatomic basis of prolonged conduction block.
sions producing conduction block. - It is impossible in the
290 292
A, A low-power electron micrograph of a swollen nerve fiber 6 weeks

human preparation to directly document the degree of Wallerian
following an acute compressive episode.The swollen inner aspect of
degeneration and pure conduction block. As previously noted, it
the Schwann cell cytoplasm (v) separates m o s t of the surface of t h e
is difficult to produce profound conduction block without also
axon (ax) from t h e myelin sheath. A s e c t o r remains attached t o t h e
injuring some of the axons, thus resulting in Wallerian degener-
sheath (x 3,000). B, Enlargement of the region demarcated by t h e
ation. It is entirely conceivable that the human subjects with
arrow in A.The mesaxon (m) is observed with swollen Schwann cell
conduction block also may have had some axonal damage to ac-
cytoplasm (v) on either side of it (x 44,000). C , A myelinated fiber
count for the observed fibrillation potentials. Pure conduction
demonstrating an axon (ax) and a region of edema (v).A macrophage
block lesion, segmental demyelination without axonal loss,
(ma) has penetrated into t h e basement m e m b r a n e and entered into
most likely does not produce fibrillation potentials as recorded
the myelin sheath. (From O c h o a J, Fowler TJ, Gilliatt RW: Anatomical
extracellularly with concentric or monopolar needle electrodes.
changes in peripheral nerves compressed by a pneumatic tourniquet. J
However, more detailed investigations are required to defini-
Anat 1972;! 13:433-455, with permission.)
tively answer this question.
Schwann cell, and (3) an axon. The myelin sheath consists of

MODEL OF A C T I O N POTENTIAL multiple layers of the Schwann cell membrane spiraled around
CONDUCTION SLOWING A N D BLOCKADE the axon. A 10 A space separates adjacent lamellae, which is an
extension of the extracellular space sequestered during the orig-
In order to better appreciate the clinical findings associated inal formation of the myelin sheath. This interlamellar por-
141
with demyelination, it is appropriate to formulate a model of tion of the extracellular space is at least partially and most likely
normal and abnormal action potential propagation. This model completely isolated from the extracellular space surrounding
can conceptualize the physiologic occurrences on the cellular the nerve by a series of tight junctions. - The sequential lay-
141 266
level that ultimately lead to nerve conduction slowing or con- ering of the Schwann cell's membrane results in a particularly
duction block with accompanying altered sensation and lossôf important electrical property conveyed to the myelinated nerve
motor function. We may first begin with what is known about by the Schwann cell. This relatively thick myelin sheath
the normal axon/myelin constituents and their electrical proper- wrapped around approximately a 1-mm segment of axon results
ties with respect to action potential propagation. in a rather low capacitance and large transverse resistance with
respect to electrical current attempting to flow from the interior
NORMAL NERVE of the axon to the extracellular space over this 1-mm segment.
In other words, the myelin sheath constitutes an excellent elec-
Structural A s p e c t s trical insulator assuring that significant portions of the current
A normal myelinated mammalian nerve fiber is known to confined within the axon beneath the myelin do not escape
consist of three major components: (1) a myelin sheath, (2) across the myelinated segment.
The Schwann cell extends for approximately 1.0 mm over the channel distribution along the various portions of the axolemma
longitudinal extent of the axon and reaches this length because is found in immunocytochemical, saxitoxin-binding, and volt-
of peripheral nervous system growth elongating its original 300 age clamp m e t h o d s . - ' - ' ' All of these studies strongly
25 55 85 212 252 253
pm length in the immature a n i m a l . By wrapping itself

172295
suggest that sodium channels are primarily localized in high
around the axon multiple times, a substantial thickness of density at the nodal membrane and at significantly lower den-
myelin layering results. Adjacent Schwann cells with their ac- sity in the internodal membrane (Fig. 4-14). Therefore, voltage-
companying myelin sheath approach each other within about sensitive sodium channels are concentrated at the nodes of
1.0 pm or less. As noted previously, this 1.0 pm region of
19
Ranvier and in very low concentrations along the internode.
axonal membrane devoid of myelin is referred to as the node of The issue of potassium channel localization is both interest-
Ranvier. ing and important to understand, particularly with respect to
The axonal membrane, therefore, is enveloped by a discon- mammalian as opposed to squid nerves. The use of voltage
tinuous myelin covering except at approximately 1-mm inter- clamps at the nodes of Ranvier demonstrates that there are very
vals where it is exposed to the extracellular environment. few if any voltage-gated potassium channels. - Two pharma- 34 54
Multiple different experimental techniques have detailed the cologic agents, 4-aminopyridine (4-AP) and tetraethylammo-
molecular structure of the axolemma with respect to ion gates nium (TEA), are known to specifically block rapidly activating
present along various aspects of the axon, both beneath the or "fast" potassium channels and delayed rectifying or "slow"
myelin sheaths as well as the exposed nodes of Ranvier. Various potassium channels, respectively. A combination of these 306
cytochemical investigations have demonstrated that there is a blocking agents and intra-axonal recordings of CNS myelinated
distinct difference between the axolemma of the internode (por- nerve reveal that repolarization of the action potential in these
tion surrounded by myelin) compared to that of the node of fibers is not dependent upon TEA-sensitive potassium channels
Ranvier. The nodal membrane is found to have very similar
238
nor is the action potential morphology altered by blocking the
characteristics to that of the initial segment of the axon (axon 4-AP-sensitive potassium channels. " - ° Unlike the famil- U7173 ,74 3 4
hillock) just distal to the cell body, i.e., high sodium channel iar squid nerve, mammalian nerve does not depend upon potas-
density. Also, freeze-fracture studies of the axonal membrane
70
sium channels for repolarization but instead re-establishes the
demonstrate that the external surface of the nodal membrane resting membrane potential through sodium inactivation and
contains membranous proteins with a density approximating sodium "back-leak" currents. - 34 54
1300/pm . - The same technique, however, reveals that the

2 183 258
The fast 4-AP-sensitive potassium channels appear to be
internodal/paranodal membrane contains a significantly re- preferentially located in the paranodal and internodal axon
duced particle density of 100-200/pm . There is a suggestion
2
membrane, i.e., beneath the myelin sheath (Fig. 4-14). Although
that the physically observed particles in freeze-fracture tech- the exact function performed by these fast potassium channels
niques correspond to the previously noted collections of sodium is unknown, it is postulated that they serve to stabilize the
channels in this same region. Of course, it is possible that the axonal membrane following an action potential to prevent repet-
observed particles may represent a subset of sodium channels or itive firing to a single stimulus. - It is also possible that there
176 254
multiple sodium channels per detected particle. Of interest is is a role for these potassium channels to help generate the rest-
that nonmyelinated fibers demonstrate the above-noted particles ing membrane potential. 56
contained in the external membrane surface at a density of The slow TEA-sensitive potassium channels are present in
150-300/pm . Further support for the heterogeneity of sodium
2 24
the axonal membrane in some mammalian myelinated fibers
(Fig. 4 - 1 4 ) . It appears that the slow potassium channels
1 3 1 7 8 1 7 9
are preferentially activated in prolonged depolarizations such as

those arising from high-frequency stimulation or discharge.
There is a suggestion that in repetitive firing of the axon, the
slow potassium channels help to modulate the ability of the
axonal membrane to participate in successful action potential
°Na Kt
9
regeneration. This is accomplished by the generation of an after-

hyperpolarization that decreases the generation of an undesired ,
burst of action potentials resulting from the initially generated j
potential, thereby resulting in a coordinated as opposed to unco-
ordinated firing pattern. The actual location of these slow chan-
nels is in question, but they appear to be located in both the
Figure 4-14. Model of proposed ion channel organization in nodal and internodal axon membrane.
myelinated mammalian n e r v e f i b e r s . T h e sodium channels are A fourth intramembranous ion channel is believed to be pre-
present in high concentration in t h e axonal membrane at t h e node of sent in mammalian myelinated nerves and referred to as an
Ranvier but are in very low density (less than necessary for action po- inward rectifier channel (Fig. 4 - 1 4 ) . Hyperpolarization of 1386
tential propagation) in the internodal region. Potassium channels (4- the axon membrane is the stimulus that activates this channel to
AP-sensitive) are distributed in a complementary fashion t o t h e allow an inward current into the axon. This inward current
sodium channels, i.e., abundant in t h e internodal region (covered by serves to prevent excessive hyperpolarization, thus regulating
myelin) but less so at t h e node of Ranvier. gNa, sodium channels; gKf, the excitability of the membrane, i.e., helps keep the membrane
fast (4-AP-sensitive) potassium channels; gKs, slow (TEA-sensitive) from deviating too far from the resting membrane potential and
potassium channels; gIR, inward rectifier channels. (From Waxman SG: threshold level. Excessive hyperpolarization tends to prevent re-
Molecular organization and pathophysiology of axons. In Asbury AK, current action potential propagation, particularly at intervals
McKhann GM, McDonald Wl (eds): Diseases of the Nervous System: overlapping the hyperpolarized time frame. There is a sugges-
Clinical Neurobiology. Philadelphia, W.B. Saunders, 1992, pp 2 5 - 4 6 , tion that both sodium and potassium ions may be permeable to
with permission.) these channels.
Electrical A s p e c t s
Extracellular recordings along the surface of single motor
nerve fibers during action potential propagation reveal a number
of interesting findings (Fig. 4-15). Moving a recording electrode
pair across the myelinated internodal region demonstrates an
external longitudinal current with essentially a constant latency
but slightly declining amplitude. At discrete intervals there is a
noticeable shift in latency for the externally recorded current,
which repeats at regular intervals. The regions of latency shift
correspond to the nodes of Ranvier where an inwardly directed
current enters the axon. The direction of current flow is found to
be outward along the internode but inward at the node of
Ranvier. The concentration of current flow about the node with
a progressively diminishing density away from the activated Distance (mm)
node produces the decreasing amplitude noted for the extracel-
lularly recorded potential. A latency shift along the external sur- Figure 4-15. Normal rat single nerve fiber is depicted from a
face of the internode is not observed because the current is ventral root. A, External longitudinal c u r r e n t measurements m a d e
established and manifests essentially instantaneously along the successively along t h e a x o n . T h e records a r e made in 0.2-mm incre-
internode. When the next node is activated, the density of cur- ments.The time scale is 100 us/div and t h e vertical bar is 100 uV. B,
rent for that node is rather large and diminishes toward the next Depiction of electrode recording position along the nerve fiber. C.The
node to be activated, thus repeating the sequence of a series of peak external current latency is shown as a function of distance along
diminishing amplitude potentials shifting as a block. the fiber. Note the discrete jumps of time necessary for action poten-
The shift of latency at the node of Ranvier reflects the time tial propagation, thus yielding t h e concept of saltatory o r "jumping"
required to raise this region of axolemma to threshold and re- conduction in myelinated fibers. (Modified from Rasminsky M, Sears
generate the action potential. A finite amount of time is neces- TA: Internodal conduction in undissected demyelinated nerve fibres. J
sary for the transmembrane voltage to reach threshold and begin Physiol 1972;227:323-350, with permission.)
the self-regenerating process of sodium activation through the
opening of voltage-gated sodium channels. Approximately 19.7
ps is the time for action potential generation at the nodes of next node to be depolarized for three reasons. First, although
Ranvier, i.e., the internodal conduction time. It is this progres-
246
the axolemma and myelin sheath form a very good barrier, there
sive integral shift in latency that gives the appearance of depo- is still some minimal current .capable of passing through the
larization "jumping" from one node to the next, i.e., saltatory membrane by way of its resistive current loss component (Fig.
conduction (Fig. 4-15). Suppose the median nerve conduction 4-16). One can also expect some current to be lost through a ca-
velocity for the forearm segment is found to be 60.0 m/s and we pacitive element as the positive charges neutralize some of the
want to calculate the "average" time required to depolarize each intracellular anions attracting extracellular positive ions. As the
node of Ranvier contained in the measured segment of nerve intracellular anion is now attracting the intracellular sodium it
(20.0 cm). The time required for the nerve to conduct over 20.0 no longer "holds" onto an extracellular positive ion as strongly,
cm is 3.33 ms (60.0 mm/ms = 200.0 mm/t; t = 3.33 ms). The allowing it to escape from the outer surface of the axon. This
time per node is 16.7 ps because 3.33 ms or 3333.3 ps is re- process results in a net transfer of charge, current flow, across
quired to cover about 200 nodes of Ranvier as each node is the myelin sheath thereby further diminishing the magnitude of
roughly 1.0 mm long. Therefore, there is 3333.3 ps per 200 the intracellular current flow. Finally, current is reduced prior to
nodes or 16.7 ps/1 node. It may be of benefit to conceptualize reaching the next node to be activated because of the internal re-
the axon and its enveloping myelin as an electrical circuit, par- sistance of the axoplasm itself (Fig. 4-16).
ticularly when considering the electrical consequences of de- Once the intracellular current from the previous node reaches
myelination. the resting node about to be excited, several processes must
The axon can be thought of as a cylindrical core of axoplasm occur. The bare nodal membrane also has a resistive and capaci-
with a finite resistance to the flow of current (Fig. 4-16). The tive component. The intracellular current must both "charge" up
axolemma and surrounding myelin sheath form an insulating the capacitor of the nodal membrane, thereby displacing charge
covering around the axon's core. These two structures possess extracellularly (current flow across the node), as well as pass di-
both a resistive and capacitive component. A small amount of rectly through some passive ion channels in the membrane (Fig.
ionic current is capable of passing through the axolemma and 4-16). This resistive and capacitive current is concentrated at the
myelin sheath under normal conditions as no substance is a per- nodal region and by flowing across the nodal membrane serves
fect insulator. Additionally, the axolemma and myelin sheath to alter the transmembrane voltage to the required threshold
act as capacitors because they separate charge between the in- level where the voltage-sensitive sodium channels are then
tracellular and extracellular aspects of the axon. Following de- opened. With the opening of the voltage-gated sodium channels,
polarization at a node of Ranvier, an inwardly directed current sodium activation, the process of a relatively large inwardly di-
mediated by sodium ions flows both proximally and distally rected sodium current is generated to repeat the process of ex-
along the intracellular longitudinal core of the axon. Because citing the next node, i.e., action potential propagation.
the node of Ranvier just depolarized prior to the activation of Normally, the amount of current reaching the next node to be
the node under present consideration is refractory, it will be ig- activated is 5-7 times greater than minimally needed to gener-
nored. Instead, the node about to be depolarized is of primary ate the threshold voltage. This difference is referred to as the
285
concern. The inwardly directed current progressing toward the safety factor of neural transmission and can be expressed as:
resting node will decrease in magnitude prior to reaching the current available to reach threshold/current required to reach
Membrane
Figure 4-16. Equivalent circuit diagram for a nerve fiber. Axon with investing myelin sheath shown forming t w o nodes of Ranvier (A and
B) and an internodes segment.The equivalent circuit diagram representing t h e combined resistive and capacitive components of t h e various as-
pects of t h e nerve fiber. Note' that each n o d e of Ranvier can be thought of as consisting of a variable resistor (R ; changes with voltage level, i.e.,
N
sodium activation) and capacitor ( C ) . T h e internodal region of the nerve also consists of a membrane resistance (R ) and membrane capacitance
N M
(C ).The internal resistance of the axoplasm is represented by R .The inward-directed current at node A is shown t o have a rapid rise time and
M A
relatively large amplitude.This same current is reduced in amplitude and rise time by t h e time it reaches node B because of having lost some of its
content through t h e membrane resistance and capacitance as well as axoplasm resistance. If t h e current at node B is of sufficient magnitude t o
reach threshold, an .inwardly directed current generated at node B similar t o that depicted for node A will be produced t o subsequently activate
the next node in line. (Modified from Rasminsky M: Pathophysiology of demyelination. In Didactic Program, AAEM, 1980. Rochester, MN.AAEM,
1980, pp 29-34.)
threshold > 1.0. Should this value fall below unity, one can an- recovers to the predisease state provided the induced lesion was
ticipate failure of action potential propagation. not too severe. One of the earliest electrical abnormalities noted
in the above neural insults was slowing of nerve conduction
Anatomic/Electrical Aspects of Demyelination through the demyelinated segments. - There was an overall
58 2022 0 3
There are a number of ways to produce experimental segmental reduction in the maximum recorded conduction velocity from
demyelination of varying degrees. One method of accomplishing 94.8 m/s to 33.7 m / s . Also, there was a greater proportion of
202
focal or paranodal demyelination has already been described in fibers conducting at the slower conduction velocity than prior to
detail, i.e., compression of varying degrees arising from pneumatic the nerve insult. This suggested that there was significant slow-
tourniquets or nylon bands. Investigators have also used two ad-
218
ing of conduction in the fastest-conducting fibers down to the
ditional methods to induce focal demyelinating lesions with rela- lower value. The slowing of conduction is not believed to result
tive axonal sparing. The first concerns induction of experimental in weakness but may account for loss of deep tendon reflexes
allergic neuritis in animals. Lesions occur in the peripheral nerve, where the synchronous arrival of impulses is important to elicit
spinal ganglia, and nerve roots as focal perivascular accumulations a response. - As previously noted, normal internodal con-
105 280
of inflammatory cells. - - - These regions of inflammatory cel-

58 I35 160 185
duction time is approximately 20 p s . In demyelinated nerves,
278
lular aggregates are associated with the focal breakdown of however, internodal conduction time can reach 500-600 ps. If a
myelin. The other method of generating focal demyelinating le- nerve were previously conducting at 60 m/s over a distance of
sions in the peripheral nervous system is through the application 200 mm, the internodal conduction time would be 16.7 ps. For
of diphtheria toxin-antitoxin m i x t u r e s . This type of inocu-
201202309
this nerve, an increase in the internodal conduction time to 500
lation with 0.5-2.0 ml of the mixture results in weakness onset in ps would result in a dramatic reduction in conduction velocity
5-10 days following injection with a continuation of symptoms to 2.0 m/s. Although most nerves continue to conduct in a salta-
for an additional 5-20 days with subsequent recovery, and occa- tory fashion prior to block, there is evidence that in a least
246
sional death of the animal receiving higher dosages. An additional some nerve fibers conduction may actually become continuous
method used to "dissolve" sequential layers of myelin has also across the demyelinated segment similar to normal unmyeli-
been performed by using saponin, a fat solvent, on amphibian nated neural conduction. In addition to action potential propa-
28
nerves. Although this is obviously not a naturally occurring dis-

284
gation slowing and blockade, there are a number of additional
ease, the method is nevertheless useful for examining conduction electrical abnormalities occurring over demyelinated segments
through controlled demyelinated regions of nerve. The end result of nerve. Close inspection of the waveform morphology of
of the above processes is an initial thinning of the internodal action potentials conducting across a demyelinated region re-
myelin, usually, though not exclusively, beginning about the para- veals that there is usually a reduction in amplitude and an in-
nodal region and extending varying distances along the internode. crease in the potential's duration, thus producing temporal
The myelin sheath becomes significantly thinner and may disap- dispersion. The nonuniform slowing of multiple fibers results in
pear completely. Once the diseased myelin is removed through a less than synchronous arrival of action potentials at the record-
phagocytosis or Schwann cell ingestion, a new myelin layer is ing site compared to the previous unaffected temporal separa-
formed. As noted previously, this newly formed myelin sheath is tion of action potentials. A less synchronous arrival of
typically thinner than the original sheath and may take quite some individual action potentials disturbs the usual summation of
time, if ever, to reach the thickness of the unaffected internodal electrical impulses resulting in a smaller, polyphasic, and in-
myelin sheaths. creased duration waveform.
The clinical consequences of the above-noted lesions are es- Internodal conduction times in excess of 500-600 ps essen-
sentially similar in that they result in weakness, which eventually tially result in failure of conduction, i.e., conduction block. It 246
is the failure of action potential transmission (conduction block) to an elevation in temperature, lowering the temperature by as
that results in the observation of clinical weakness. As one little as 0.5°C decreases the internodal conduction time suffi-
might anticipate, if the action potential does not reach the motor ciently to ensure action potential propagation. It has been deter-
nerve terminal, the muscle membrane cannot be depolarized mined that an increase in temperature results in a decrease in
and there is a lack of muscle activation with ensuing weakness. the action potential rise time, suggesting a more rapid initial in-
Conduction block can be frequency-related, whereby low-fre- crease in the inwardly directed sodium current. - - This 94 228 247
quency stimuli continue to conduct without difficulty, in con- rapid rise time and quicker establishment of optimal sodium
trast to a high-frequency train of stimuli that may result in current flow decreases the internodal conduction time, thereby
action potential blockade. - - This is a direct reflection of an
64 161 307
increasing conduction velocity in normal fibers. There is also
increase in the refractory period of demyelinated nerve. - - 203 246 273
less time for the current to flow as well as sodium channel inac-
Recall that there are few sodium channels beneath the intern- tivation occurring sooner, which shuts down sodium influx and
odal membrane. Removal of the myelin sheath requires the decreases the action potential's total duration. - - This is
146 153 283
action potential to conduct across a segment of nerve with a compatible with action potential propagation in normal nerve
sodium channel density that is most likely too low to sustain because the safety factor is 5-7 times that absolutely needed to
action potential propagation. A small amount of demyelination ensure neural propagation. In demyelinated nerve, however,
may only serve to slow the internodal conduction time by pro- current is lost through the demyelinated internodal region as
longing the time to reach threshold, but if enough of the inex- well as the other factors tending to suppress conduction, result-
citable membrane is exposed, action potential failure most ing in a reduced tolerance for less current flow and the rapid
likely occurs. Additionally, exposure of the potassium channels sodium inactivation (see above). The end result of temperature
in the internodal region will attempt to hold the membrane po- elevation in a demyelinated nerve with a reduced safety factor
tential close to the potassium equilibrium potential, thus resist- and more time required to reach threshold is less current flow-
ing a move of the membrane potential toward the sodium ing for a shorter period of time compared to a lower temperature
equilibrium potential, i.e., suppressing any attempt at depolar- resulting in action potential propagation failure, i.e., conduction
ization. There may also be a "pump-mediated" hyperpolariza-
307
block. This is the most likely cause for the phenomena of in-
tion following an action potential, which also results in a creased weakness of patients with multiple sclerosis when sub-
reduction in the safety factor of transmission. Recall that fol-
31
jected to an elevation in body temperature, i.e., the "hot bath
lowing an action potential, the influx of sodium acts to maxi- test." A reduction in temperature restores action potential
124
mize the exchange of sodium for potassium thus leading to a propagation because sodium inactivation is slowed permitting
slight hyperpolarization in that the sodium-potassium pump current to flow for a longer period of time, thus generating suffi-
overcompensates to a small degree (see Chapter 1). This hyper- cient transmembrane voltage alterations to induce sodium gate
polarization is relatively short-lived, with the resting membrane activation at the next node of Ranvier.
potential eventually settling back to its steady-state level. Additional electrical abnormalities can be detected in de-
There is a further relationship between this "pump" effect and myelinated nerves. Spontaneous generation of impulses have
rate-dependent conduction block of action potential transmis- been observed to arise in chronically demyelinated cat dorsal
sion. Most normal nerves can conduct trains of impulses at
161
column axons. There is also the phenomena of ephaptic con-
274
rates of 50 Hz or more for time periods approaching several duction or "cross-talk" between neighboring nerves with less
hours. Following the generation of multiple action potentials, than an optimal complement of a myelin sheath. Ephaptic con-
the amount of intra-axonal sodium accumulates faster than can duction is a process where impulses interact with adjacent
be removed by the sodium-potasium ATPase pump mechanism. nerves thereby setting up abnormal impulses in them, leading to
It is believed that this sodium accumulation is even greater in various sensory or motor disturbances.' This finding has
75,297
demyelinated nerve still capable of mediating neural conduc- been described in both acutely and chronically damaged nerves.
tion. This is because of a greater surface area of internodal The two most common occurrences of ephaptic conduction
membrane exposure and current flow. The increased intracellu- were noted in nerve fibers ending in neuromas and nerve roots
lar concentration of sodium in turn stimulates the electrogenic of dystrophic m i c e . - There is also some evidence that de-
177 249
sodium-potassium pump, thereby hyperpolarizing the mem- myelination can lead to an increase in mechanosensitivity of
brane and increasing the voltage differential between the new certain nerve fibers and may account for the so-called
resting membrane potential and the threshold level. As a result 31
Lhermitte's sign, paresthesias in radiculopathies, straight leg
of more current required to reach threshold because of the and arm raising signs, and other similar observations. - 152 303
greater voltage difference, the neural safety factor is reduced,

which then results in conduction block, i.e., a rate-dependent C o m p u t e r Nodeling of Myelin Loss
conduction block. Various computer models have been developed to better un-
A number of interesting findings have been noted with re- derstand the consequences of myelin loss on action potential
spect to varying the temperature and observing the conse- propagation. - - A number of known parameters regarding
I80 208 301
quences of this action on neural conduction velocity and myelinated nerves and action potential propagation characteris-
blockade in demyelinated nerves. - - Generally, increasing
63 247 265
tics were programmed to simulate normal action potential prop-
temperature within the normal physiologic range for normal agation. Because normal neural conduction is easy to measure
nerve results in an increase in conduction velocity. In normal and quantify, it is rather simple to verify if the computer model
mammalian nerve fibers the internodal conduction time at 30°C provides accurate results when compared to reality. After
is approximately 30 ps, while at 37°C, it is about 20 ps. In de- having ensured that the program is capable of predicting the an-
myelinated internodes, however, the internodal conduction time ticipated results under normal conditions, the program can be
at 35°C approaches 360 ps. Any further increases in tempera- altered with respect to various anatomic aspects of the nerve to
ture result in action potential conduction block. Once conduc- simulate different disease conditions and assess whether simu-
tion block is produced in some demyelinated fibers secondary lations match clinically observed results. It is then possible to
change one variable at a time to investigate how these alter- sodium ions mediating the current flow to "cross the nodal
ations produce the clinically recorded response. This technique membrane" through a net transfer of charge. The intracellular
has been specifically applied to learn more about action poten- sodium ions "neutralize" some of the intracellular anions' at-
tial failure in conditions resulting in loss of myelin, i.e., seg- traction for the extracellular ions, thus allowing these ions to
mental demyelination from various diseases like compression move away from the outside of the axolemma and generating a
neuropathies and multiple sclerosis. current transfer across the membrane. This current passage
Initially, a myelinated nerve is simulated where an action po- across the node results in altering the transmembrane voltage. If
tential conducts with the expected characteristics of normal the voltage change is sufficient, threshold is achieved and an
neural propagation. For simplicity, an entire internodal portion action potential is generated through sodium activation of the
of myelin is then removed without altering the axon's conduct- voltage-dependent sodium gates in the nodal membrane. This
ing properties. Specifically, the entire aspect of the denuded newly formed inward-directed current then repeats the process
axon is assumed to possess a sodium channel density equal to of nodal depolarization at the next node of Ranvier, producing
that of the node of Ranvier. We now know that this is incorrect the expected saltatory conduction.
but at the time of the study, this information was not available Once the internodal myelin is removed, a separate set of cir-
and the investigators assumed that the internodal axolemma was cumstances now becomes operative to adversely affect action
similar to the nodal region, i.e., excitable. The action potential potential propagation. Thinning or complete removal of the
fails to conduct across the demyelinated segment to excite the myelin component does not alter the intra-axonal resistance to
nodes distal to the region of demyelination, i.e., conduction current flow or the amount of current generated at the node of
block occurs. The finding of action potential failure despite an Ranvier compared to the normal situation. There is, however, a
internodal density of sodium channels equivalent to the nodal profound effect on the internodal resistance and capacitance.
region is rather interesting and implies that even with a rela- Myelin removal now permits the internodal intra-axonal current
tively high sodium channel density, neural propagation could to pass relatively unrestricted through any passive sodium chan-
not cross the demyelinated region. In other words, conduction nels present in the axolemma, i.e., the membrane resistance
did not become continuous as in unmyelinated nerve despite a component is significantly lowered compared to normal. A
high sodium channel availability. The explanation for this ob- greater portion of the intra-axonal current compared to normal,
servation is rather interesting and provides significant insight therefore, is dissipated through the resistive component of the
into the requirements for saltatory action potential propagation membrane. Also, the entire internodal membrane can now act as
in myelinated fibers. In order to appreciate the implications of a large set of capacitor plates and accumulate significant
this finding of action potential blockade, we must return to the amounts of extracellular ions. As the intra-axonal current now
previously discussed "electrical circuit" representation of the travels beneath the exposed internodal axolemma, a large
myelinated axon. amount of current can cross the membrane in the capacitive cur-
As previously noted, depolarization of a node of Ranvier ex- rent transfer. The decreased resistive and increased capacitive
cites the adjacent node by inducing an inward-directed current, aspects of the internodal membrane result in a large current loss
which travels intra-axonally toward the next node. The current across the entire internodal region. As a result, there is little cur-
flow experiences a minimal loss over the internodal region rent available to generate a transmembrane voltage shift of suf-
through both the capacitive and resistive components of the in- ficient magnitude to activate the voltage-dependent sodium
ternodal membrane. Recall that because of the insulation prop- channels at the node of Ranvier because of the diffuse loss of
erties of the internodal myelin, the capacitance is low and the current. Even if the internodal membrane had a high density of
resistance is high. In other words, there is minimal ability of the sodium channels similar to the nodal membrane (see above
intra-axonal current to escape across the axolemma and invest- computer example), conduction continues to fail because there
ing myelin sheath. A finite amount of current is also dissipated is insufficient current to depolarize any aspect of the membrane.
within the axon over the internodal region because of the intra- The decreased resistance and increased capacitance of the in-
axonal resistance of the axoplasm. Despite the three avenues of ternodal membrane depletes the available current to such an
current loss noted, there is still 5-7 times the amount of current extent that the normal 5-7 times current safety factor is reduced
required to effect depolarization at the next available node of below 1, i.e., an insufficient amount of current is available at
Ranvier. any single region of membrane to generate an action potential. |
At the nodal region, the bare axolemma allows the remaining According to computer models, 97.3% of the myelin can be re- !
intra-axonal current to cross the nodal membrane through its removed and action potentials still propagate across the internodal
sistive and capacitive components. Remember that the small region. Once the myelin thickness is reduced to 2.5% of its
region of axolemma at the exposed membrane has some passive normal thickness, there is too much current lost across the in-
sodium channels and therefore a lower resistance than the internodal region and the safety factor becomes unfavorable for
ternodal region, thus allowing current to cross the membrane. action potential propagation resulting in conduction block.
Additionally, the absence of myelin at the node creates a large Introducing a short internodal region of 400 pm just proximal
capacitance because it allows the extracellular sodium ions to to the demyelinated segment does not alter the above situation
accumulate about the node secondary to the intracellular attrac- and there is continued conduction failure. Placing two-200 pm
tion of negative ions. The thin axolemma separating the intra- internodes in the same location noted above, however, results in
cellular and extracellular fluid permits the intracellular anions the action potential propagation into and across the demyeli-
to exert a larger attractive force on the extracellular positive ions nated segment. The two interesting observations in this instance
compared to the increased distance between the intra/extracellu- are (1) action potential propagation across the demyelinated
lar regions across the internode. The nodal membrane behaves segment in a continuous manner similar to that in unmyelinated
as two capacitor plates with significant charges permitted to nerves, and (2) repair of conduction block. The same simulation
build up on either side, i.e., a capacitor with a high capacitance is repeated for an internodal axolemma with a significantly re-
value. As a result, it is relatively easy for the intracellular duced sodium channel density. The same results of action potential
conduction across the unmyelinated segment with repair of con- there is less current available at the node and it must flow for a
duction block are observed. Adding two short internodal regions longer period of time to reach threshold. The increase in time is
just proximal to the demyelinated segment results in action po- reflected as a decrease in the internodal conduction velocity,
tential conduction by increasing the amount of current available thus producing a slowing of neural conduction. This is clearly
for this region, i.e., the safety factor is increased. This increase shown to be the case when conduction times approaching 600
is accomplished by the location of the two internodes. These ps have been documented in demyelinating diseases (see
two internodal regions are simultaneously activated by the prox- above). The same ideas of resistive and capacitive current loss
imal node of Ranvier. As a result of two current generators can be applied to the temperature effects with respect to con-
being coincidentally activated at the same time, a large amount duction block. Increasing neural temperature reduces the safety
of current is injected into the membrane. This large influx of factor by facilitating early sodium inactivation. Less current
current is of a sufficient magnitude to tolerate a significant loss availability may result in conduction block across a partially de-
across the demyelinated membrane through the resistive and ca- myelinated segment because the amount of current at relatively
pacitive current loss mechanisms with enough current left over lower temperatures is just sufficient to exceed capacitive/resis-
to achieve the voltage-dependent sodium channels' threshold tive internodal current loss but is insufficient at the higher tem-
level at the next normal node of Ranvier/internode region. The perature. Reducing the temperature again restores neural
increased amount of current secondary to the two activated conduction as current is permitted to flow for a longer time
nodes, elevated safety factor, is now capable of repairing the period to overcome the capacitive/resistive current loss, thus
previously noted conduction block. The observation of short in- reaching threshold and action potential generation.
ternodes interposed prior to a demyelinated segments has not Of course, the above computer simulations demonstrate
been clinically documented but certainly poses a method to clearly the importance of remyelination. Eistablishing a new in-
repair conduction block. ternodal myelin thickness at least 2.7% of the original sheath
The concept of action potential failure secondary to myelin loss thickness reduces the internodal membrane capacitance and in-
is referred to as impedance mismatch. An imbalance in the opti- creases this region's transverse resistance sufficiently to prevent
mal impedance or various combinations of capacitances and resis- enough current leakage to once again reach a safety factor com-
tances with respect to current flow no longer produces an optimal patible with action potential propagation. Although the conduc-
set of circumstances regarding the electrical properties of the tion velocity may be significantly slowed over this internodal
nerve favoring action potential propagation. Adding short intern- region, at least conduction block has been repaired. The re-
odes to supplement the amount of current available to overcome myelination is also of significance because it again isolates the
the impedance mismatch is one way of repairing action potential paranodal potassium channels from the extracellular compart-
propagation. Removal of myelin from an entire internodal segment, thus decreasing any current loss through this mechanism
ment has been modeled on computers but there is reason to be- to increase the amount of intra-axonal current available for de-
lieve that more limited paranodal demyelination can also result in polarization of the next node of Ranvier. The mechanism of
an insufficient amount of current available for action potential potassium channel blockade with 4-aminopyridine and TEA has
propagation secondary impedance mismatch. There are examples been tried experimentally to restore action potential propagation
of normally occurring regions of impedance mismatch in the pe- with some success in animals. The combination of limited
29 3 0
ripheral nervous system where action potential conduction can be clinical success and potentially serious side effects have d e -
somewhat tenuous, particularly under pathologic condi- creased the initial excitement for the potential of 4-aminopyri-
tions. - - The "initial segment" of the anterior horn cell is a
300 302 305
dine to be clinically useful. - As previously noted, the
162,268 277
region where the nerve becomes somewhat larger and is unmyeli- ability of demyelinated peripheral nerves to mediate impulses is
nated compared to its immediately adjacent distal thinner and in part dependent upon a hyperpolarization resulting from the
myelinated portion. At this transition zone the current is spread sodium-potassium pump mechanism and this has been observed
rather thin over the larger region of unmyelinated membrane, thus in the C N S . Inhibition, at least in part, of this pump by the
1172 2 3
allowing more current to escape and lower the safety factor. cardiac glycoside digitalis may be of benefit in decreasing the
Fortunately, there is believed to be an increased number of volt- voltage differential between the hyperpolarized membrane po-
age-dependent sodium gates most likely capable of generating a tential and the threshold level. Indeed, the administration of this
greater current flow to offset what is lost through the membrane's drug has been shown to be of benefit in both experimental ani-
larger surface area. Branch points of peripheral nerves, which usu- mals with demyelination and a limited number of patients suf-
ally occur at the nodes of Ranvier, are also regions where there is fering from multiple sclerosis. Additional experimental
1 6 3 J6 4 1 6 5
more exposed axolemma permitting a potential for impedance insights from the above computer simulations will hopefully
mismatch. Increased neurogenic jitter and neurogenic blocking continue to bring new and exciting treatment options for pa-
may be a result of action potential failure at these branch points in tients affected by various demyelinating diseases.
the collateral tree of terminal axon branch points in reinnervation
secondary to immature and poorly myelinated nerve. A third po- CLINICAL CORRELATION
tential region of impedance mismatch is the terminal segment-of
the motor nerve where there is a transition between the myelinated The identification of conduction block is an important aspect
and unmyelinated segment about the neuromuscular junction. of the electrodiagnostic medicine evaluation. This is because
The conduction velocity across the demyelinated segment is some inciting incident has produced a temporary and potentially
noted to be markedly reduced compared to the unaffected seg- reversible blockade of neural conduction. Removal of the offend-
ments. As previously noted, when the internodal myelin is re- ing agent and/or protection of the injured nerve should restore
duced there is a loss of current across this region. If there is a complete neurophysiologic function to the affected structures in-
degree of myelin loss capable of reducing the current somewhat nervated. The recognition of conduction block implies the patient
but less than the amount capable of totally eliminating the has a relatively good prognosis for recovery. On the other hand,
safety factor, a delay in conduction can be expected. Essentially, failure to remove an anatomic cause of conduction block such as
a compressive ligamentous structure or repetitive motion can there were no root tension signs in the left or right upper limbs.
result in the transition of conduction block into Wallerian degen- Sensation, reflex, and muscle testing in the lower limbs were
eration associated with a comparatively less optimal functional normal.
outcome. Conduction block has been considered to represent a Nerve Conduction Studies. Nerve conduction studies are
possible transition between a normal nerve and one that has been performed in the upper extremities bilaterally. The mid-palm
injured and may progress onto more profound damage such as temperature is 33.5°C on the right and 32.9°C on the left.
Wallerian degeneration if some type of therapeutic intervention is
DSL S Amp DML MAmp NCV
not performed. Although this may be the case in a number of
112
Nerve (ms) OiV) (ms) (mV) (m/s)

situations, it is also possible for conduction block to persist for
quite some time. As has been previously discussed, animal Right median 3.3 60.0 3.4 10.0 59.0
models of conduction block have persisted for months. 93
Right ulnar 3.2 45.0 3.1 9.0 64.0
Additionally, there are reports in humans of conduction block Right ulnar 8.5 (BE)* 66.0*
persisting from several months to as much as 4.8 years. - -
138 206 291
8.0 (AE)*
One method of recognizing and localizing conduction block Right ulnar CMAP (AE) duration: 6.2 ms
is to observe a marked reduction in compound muscle action Right ulnar CMAP (BE) duration: 6.0 ms
potential amplitude across a particular portion of nerve, particu- Right ulnar CMAP (AE) area: 27.3 mV-ms
larly at known sites of entrapment such as the carpal or cubital Right ulnar CMAP (BE) area: 28.0 mV-ms
tunnels. For example, excitation of the ulnar nerve several cen- Left median 3.1 55.0 3.2 9.5 61.0
timeters proximal and distal to the medial epicondyle may yield Left ulnar 3.4 35.0 3.3 6.5 58.0
a CMAP amplitude of 3.0 mV and 7.0 mV, respectively. This Left ulnar 6.3 (BE)* 40.0*
marked drop in .amplitude suggests that there is conduction 4.0 (AE)*
block of the ulnar nerve's motor fibers in or about the ulnar Left ulnar CMAP (AE) duration: 6.0 ms
groove. Unfortunately, the correct identification of true conduc- Left ulnar CMAP (BE) duration: 5.8 ms
tion block is much more complex than simply comparing Left ulnar CMAP (AE) area: 16.0 mV-ms
CMAP magnitudes. Two hypothetical cases are presented to Left ulnar CMAP (BE) area: 22.3 mV-ms
help clarify the categorization of conduction block and situa-
tions that may lead to an erroneous diagnosis of conduction DSL, distal sensory latency; S Amp, sensory amplitude; DML,
block. An ulnar nerve compromise is illustrated as this nerve is distal motor latency; M Amp, motor amplitude; NCV, nerve con-
frequently involved in neural compression and is a cause of fre- duction velocity; ms, milliseconds; pV, microvolts; mV, milli-
quent referrals for an electrodiagnostic medicine consultation. volts; m/s, meter/second. Motor and sensory amplitudes are
measured baseline-to-peak. Sensory latencies are measured to
ILLUSTRATIVE CASES peak while motor latencies are measured to initial negative onset.
Case # I * Amplitudes and velocities across fully flexed elbow. AE,
above elbow; BE, below elbow.
History. A 22-year-old female competitive ice skater sus-
tained a fall onto the ice while skating secondary to a stress fail- Needle Electromyography. A needle electromyographic in-
ure of one of her skate's blades. Due to the unexpected nature of vestigation was performed on the left upper limb using a dispos-
the equipment failure she was unable to prepare herself for the able monopolar needle.
fall and unfortunately landed on her left elbow region.
Rest Activity
Immediately following the accident she experienced profound
Muscle PSW/Fibrillation Recruitment
pain at the elbow region with inability to feel the fourth and fifth
(L) Abductor pollicis brevis 0 Normal
digits of her left hand as well as difficulty using her left hand.
(L) First dorsal interosseous 2+ Reduced
As she had no difficulty skating and executing her skills she left
(L) Abductor digiti minimi 2+ Reduced
for a competition the next day. Within 2 days her localized
(L) Flexor digitorum profundus 2+ Reduced
elbow pain resolved, and the left hand numbness and weakness
(L) Flexor carpi ulnaris 0 Normal
improved. She was seen by a physician 3 weeks following the
(L) Triceps 0 Normal
initial insult. Radiographic examination of the left upper limb,
(L) Cervical paraspinals 0 Normal
including the elbow, was normal. She was subsequently given a
nonsteroidal anti-inflammatory medication, program of upper Summary of Findings.
limb rest, and referred for an electrodiagnostic medicine evalua- 1. Normal neural conduction parameters of the left and right
tion 4 weeks after the initial elbow trauma. median sensory and motor nerves.
Physical Examination. On physical examination the patient 2. Normal neural conduction parameters of the right ulnar
demonstrated a minimal amount of tenderness about the left nerve.
medial epicondyle with no obvious signs of trauma. There was 3. The left ulnar nerve demonstrated a drop in amplitude
a mild Tinel's sign just distal to the medial epicondyle on the across the elbow region of 37%.
left. There is was also a slight decrease in sensation to touch, 4. The left compared to right ulnar nerve at the wrist reveals
pin prick, and temperature in the cutaneous distribution of the a 28% drop in amplitude.
left ulnar nerve including the dorsal ulnar cutaneous nerve. 5. There is a 2 8 % drop in area across the left elbow region
Additionally, the flexor digitorum profundus and all hand infor the ulnar nerve CMAP.
trinsic muscles except the abductor pollicis brevis and opponens 6. Slowing of conduction for the left ulnar nerve across the
pollicis demonstrated a 3+/5 grade of strength on manual elbow region.
muscle examination. A normal grade of strength was docu- 7. Membrane instability and reduced recruitment are present
mented in all remaining muscles of the left upper limb. Deep in the ulnar innervated muscles of the left upper limb below the
tendon reflexes in the upper limbs are symmetric bilaterally and innervation of the flexor carpi ulnaris.
Impression. The patient demonstrates electrophysiologic DSL, distal sensory latency; S Amp, sensory amplitude; DML,
findings consistent with a combination of Wallerian degenera- distal motor latency; M Amp, motor amplitude; NCV, nerve
tion and conduction block of the left ulnar nerve most likely lo- conduction velocity; ms, milliseconds; pV, microvolts; mV, mil-
cated about the left elbow region. livolts; m/s, meter/second. Motor and sensory amplitudes are
Recommendation. The patient should rest the upper limb for measured baseline-to-peak. Sensory latencies are measured to
3-4 weeks and refrain from left upper limb strength training but peak while motor latencies are measured to initial negative
may continue to perform ice skating routines. The elbow should onset.
be well padded while skating to avoid injury should a fall occur. * Amplitudes and velocities across fully flexed elbow. AE,
A repeat study of the left ulnar nerve should be performed in ap- above elbow; BE, below elbow.
proximately 3-4 weeks to monitor the nerve's electrophysiologic Needle Electromyography. A needle electromyographic in-
status particularly with respect to conduction block. vestigation was performed on the left upper limb using a dispos-
able monopolar needle.
Case # 2
Rest Activity
History. A 46-year-old right-handed male with insulin-depen-
Muscle PSW/Fibrillation Recruitment
dent diabetes complains of numbness involving the fourth and
(L) Abductor pollicis brevis 0 Normal
fifth digits of the left upper limb for approximately 3 months. He
(L) First dorsal interosseous 1+ Mildly Reduced
states that he is not aware of any traumatic incident to the affected
(L) Abductor digiti minimi 1+ Mildly Reduced
limb that may have contributed to the problem, but that he does
(L) Flexor digitorum profundus 1+ Mildly Reduced
have a habit of leaning on his left elbow. The numbness has been
(L) Flexor carpi ulnaris 0 Normal
slowly progressive in nature since starting a new job about 4
(L) Triceps 0 Normal
months ago that requires him to keep detailed ledgers of numbers.
(L) Cervical paraspinals 0 Normal
The patient also notes that he is having difficulty buttoning his
shirts with the left hand. No other medical complaints are noted. Summary of Findings.
Physical Examination. On physical examination the patient 1. Prolonged distal sensory latencis for the left sural, bilat-
demonstrates decreased deep tendon reflexes of the biceps, triceps, eral median, and bilateral ulnar nerves.
and pronator teres of 1+/2+ in both upper limbs. The knee and 2. Borderline nerve conduction slowing of the peroneal,
ankle reflexes are absent in the lower limbs bilaterally. There is a median, and ulnar nerves.
"stocking and glove" type of decreased sensation to touch, temper- 3. Significant slowing of the left ulnar nerve across the
ature, and pin prick in the upper and lower limbs. Of note is that the elbow region.
cutaneous distribution of the left ulnar nerve is significantly more 4. The left ulnar nerve demonstrated a drop in amplitude
reduced compared to the other cutaneous regions of the upper and across the elbow region of 52%.
lower limbs tested. Manual muscle testing of the upper and lower 5. There is a 9.8% drop in area across the left elbow region
limbs reveals a 4/5 grade of strength, which is symmetric in the for the ulnar nerve CMAP.
upper and lower limbs except for the ulnar-innervated muscles of 6. Membrane instability and reduced recruitment are present
the left limb. The ulnar-innervated intrinsic and extrinsic muscles in the ulnar-innervated muscles of the left upper limb below the
demonstrate a 3/5 grade of strength. There is also mild wasting of innervation of the flexor carpi ulnaris.
the left first dorsal interosseous muscle compared to the same Impression. 1. The patient demonstrates electrophysiologic
muscle on the right. There are no upper limb root tension signs or findings consistent with a combination of Wallerian degenera-
limited cervical motion. tion and conduction slowing of the ulnar nerve about the left
Nerve Conduction Studies. Nerve conduction studies are elbow region. The combination of increased temporal disper-
performed in the upper limbs bilaterally. The mid-palm temper- sion and decreased amplitude are suggestive of a demyelinating
ature is 32.0°C on the right and 32.5°C on the left. type of lesion at the elbow with some axonal loss. Additionally,
the drop in amplitude with a normal drop in area suggests that
DSL S Amp DML MAmp NCV
the amplitude reduction across the elbow is secondary to the
Nerve (ms) OiV) (ms) (mV) (m/s)
temporal dispersion and not a conduction block.
Left sural 5.1 5.0 2. There is electrophysiologic evidence to support the diag-
Left peroneal 7.2 3.1 39.0 nosis of a mild generalized sensorimotor peripheral neuropathy.
Right median 4.2 20.0 4.2 8.0 50.0 This finding is consistent with the patient's history of diabetes
Right ulnar 4.1 15.0 4.0 7.0 51.0 mellitus and physical examination.
Right ulnar 7.0 (BE)* 54.0* Recommendation. The patient is instructed to avoid any
6.5 (AE)* type of compression to the left or right elbow region and to
Right ulnar CMAP (AE) duration: 7.5 ms maintain the ledgers in such a manner as to avoid further
Right ulnar CMAP (BE) duration: 7.1 ms injury to the left ulnar nerve. A repeat examination in 4 - 6
Right ulnar CMAP (AE) area: 28.3 mV-ms weeks is suggested to ensure that the lesion does not progress
Right ulnar CMAP (BE) area: 24.8 mV-ms further.
Left median 4.0 18.0 4.4 7.8 51.0
Left ulnar 4.3 8.0 4.5 6.5 48.0 Comment
Left ulnar 6.2 (BE)* 42.0* The above two cases illustrate a number of relevant issues re-
3.0 (AE)* garding the electrophysiologic diagnosis of conduction block.
Left ulnar CMAP (AE) duration: 15.0 ms At first glance one may consider conduction block to be present
Left ulnar CMAP (BE) duration: 8.2 ms if the CMAP amplitude obtained proximal to a suspected site of
Left ulnar CMAP (AE) area: 27.5 mV-ms neural compromise is significantly reduced (greater than
Left ulnar CMAP (BE) area: 29.4 mV-ms 14-20%) compared to the amplitude just below the injury site.
On the other hand, axonal loss results in amplitude reduction at can conclude that the same area, above and below the elbow,
all stimulus locations proximal and distal to the site of does not suggest a dropout of fibers, i.e., pseudo-conduction-
injury. - This is because the recording electrode is typi-
37 2 2 4 317
block. The amplitude is reduced, therefore, secondary to the
222
cally located on an end organ such as skeletal muscle and marked increase in temporal dispersion of the nerve fibers con-
Wallerian degeneration has rendered a number of motor units ducting across the elbow region. Increasing the range of con-
nonfunctional and thus incapable of generating an action poten- duction velocities across a region acts to decrease the
tial. If proximal compared to distal focal amplitude reductions synchronous arrival of the action potentials at the end organ. A
were the only criteria used to diagnose conduction block, a leSs synchronous arrival of action potentials means that there is
number of patients would be misdiagnosed. The reason for this less "in-phase" summation of similar waveform aspects thereby
erroneous conclusion is that a reduction in amplitude can arise leading to phase cancellation and amplitude reduction. The
from another common occurrence besides conduction block, second patient's combination of amplitude, area, and duration
i.e., temporal dispersion. data strongly suggest that the drop in amplitude across the
In Case 1 there is an obvious inciting traumatic incident most elbow is not a result of conduction block, but an increase in
likely producing the patient's complaint. Electrophysiologic eval- temporal dispersion secondary to an elbow lesion, i.e., demyeli-
uation reveals that the CMAP for the left ulnar nerve above com- nation. This is also supported by the slowing in conduction
pared to below the elbow demonstrates a 36.5% amplitude across the elbow. In Case 1, the slowing of elbow conduction is
reduction while the area increases 28.0% (larger distal compared most likely a result of blocking of the fastest fibers with only
to proximal stimulation sites) and the duration decreases 3.3%. the more slowly conducting fibers reaching the end organ, i.e.,
The ulnar CMAP at the wrist on the affected compared to unaf- hypothenar muscles. In this instance, one is essentially compar-
fected side reveals a 27.0% amplitude reduction. Case 2 shows ing two fiber populations to arrive at a conduction velocity. This
that the amplitude reduction across the elbow is 51.6% while area practice is of questionable value as one should not compare two
changes by only 6.4% and duration changes by 45.0%. There is a different fiber populations to arrive at a value supposedly repre-
7.1% difference between the CMAP on the symptomatic com- senting the "fastesf'-conducting fibers. Additionally, in Case 1
pared to asymptomatic hand. These results demonstrate the con- there is both CMAP (left vs. right wrist amplitude difference of
fusion possible when attempting to diagnose conduction block. 27.0%) and needle electromyographic evidence of axonal loss.
The drop in amplitude across the elbow region for the two pa- The concept of phase cancellation is important and should be
tients is certainly significant and more than the 13.6% ampli- understood to better comprehend the underlying mechanism
tude change possible in normal individuals (Table 4 - 6 ) . 2 2 3 2 2 4
producing changes in amplitude so that the distinction between
Specifically, the first patient has a 36.5% reduction and the conduction block and temporal dispersion can be made. A
second individual a 51.6% amplitude decline. These findings simple yet elegant demonstration serves to clarify the interac-
are both suspicious for a lesion producing conduction block. tion of single nerve fibers and motor units with respect to sen-
The important aspect is to confirm this tentative impression. Of sory nerve action potentials (SNAPs) or muscle action
note is the difference in duration and area changes for these two potentials amplitude, duration, and area. If one records an an-
171
individuals. In the first compared to the second patient, there is tidromic SNAP from the digit and stimulates the median or
an abnormal change in area, i.e., 2 8 % vs. 6.4%. Additionally, ulnar nerve at progressively more proximal locations (from
the duration for the across-elbow potential is normal in the first wrist to axilla), a number of characteristic changes in the wave-
(3.3%) but abnormal in the second (45.0%) patient. The combi- form are noted (Fig. 4-17). Specifically, the duration increases
nation of area and duration changes suggest that there is little
temporal dispersion of action potential conduction across the Ulnar Nerve Stimulation
first patient's elbow, while there is significant temporal disper-
sion in the second patient. Using the normal criteria for ampli-
tude, duration, and area, it is possible to distinguish the
individuals with abnormal amplitude reductions over a particu-
lar anatomic region who have sustained conduction block
versus temporal dispersion (Table 4-6). The ulnar nerve CMAP
amplitude change in Case 1 is a result of conduction block be-
cause the duration of the potential is comparable to that antici-
pated in normal individuals. The reduction in area, however,
implies that there is a dropout of fibers that are incapable of
contributing to the CMAP amplitude. This combination of
normal duration with an abnormal area reduction is highly sug-
gestive of conduction block. On the other hand, Case 2 reveals
an ulnar CMAP across the elbow with both an amplitude drop
and duration increases, but little change in area. As a result, one
5ms. 5ms-
Table 4-6. Ulnar Nerve Quantitative CMAP Waveform Figure 4-17. Ulnar nerve stimulation. Stimulation along the
Parameters " 58 0
course of the ulnar nerve with simultaneous recording of the CMAP
from the hypothenar eminence and the fifth-digit SNAP. N o t e that the
Percent Change
CMAP changes little with proximal stimulation while the SNAP demon-
Location Amplitude (mV) Area Duration (msec) strates significant decreases in amplitude and area, but an increase in
Above elbow-wrist 19.2 16.0 15.0 duration. (From Kimura J, Machida M, Ishida T, et al: Relation between
size of compound sensory or muscle action potentials, and length of
Above elbow-below elbow 13.6 6.7 8.3
nerve segment. Neurology 1986;36:647-652, with permission.)
Chapter 4 PERIPHERAL NERVOUS SYSTEM'S REACTION T O INJURY — 1 4 7
dramatically and in some cases doubles while the amplitude and

area are reduced by about 50%. On the other hand, the CMAP
duration may increase only 6.0-8.0% with amplitude and area
declining approximately 10% over comparable distances.
The above findings are understandable if one recognizes that
biphasic initially negative potentials are recorded for both sen-
sory and motor studies in most routine surface techniques. The
negative spike duration of the single-fiber sensory potential is
approximately 0.5 ms, which is about 50% of the corresponding
SNAP. There are thousands of biphasic single-nerve action po-
tentials with a range of conduction velocities between the
fastest and slowest approaching 25 m/s. - Thus, there is signif-
72 73
icant overlap of the positive spike of the first-arriving action po-

tentials with the negative spike of those traveling more slowly.
This positive/negative phase overlap results in cancellation of
-4i A
these aspects of the waveforms, thereby reducing the amplitude
(Fig. 4-18). When the nerve is excited at more proximal loca- V
tions, the disparity of arrival at the recording electrode is magni-
fied because of the 25 m/s conduction differential. The less
synchronous arrival of individual action potentials produces
Figure 4-18. SNAP obtained with distal stimulation.There is a
considerably more phase cancellation as there is increased over-
certain amount of phase cancellation secondary t o t h e disparity b e -
lap of the positive phases from the faster-conducting axons with
tween t h e single-fiber and composite potential's duration. Increasing
the negative phases of the slower-conducting axons. The net
the distance between stimulation and recording sites results in signifi-
result of this increased phase cancellation with progressively
cantly more overlap secondary t o the temporal dispersion of fastest-
more separation between the stimulating and recording elec-
compared t o slowest-conducting fibers.This lead t o phase cancellation
trodes is a reduction in amplitude and area with an increase in
resulting in a potential with smaller amplitude and area but larger du-
duration. Therefore, temporal dispersion has profound effects,
ration. (From Kimura J, Machida M, Ishida T, e t al: Relation between size
even in normal sensory nerve, with respect to the waveform pa-
of compound sensory o r muscle action potentials, and length of nerve
rameters. This finding of dramatic alterations in sensory wave-
segment. Neurology 1986;36:647-652, with permission.)
forms reduces the diagnostic applicability of sensory potentials.
As noted in the above two cases, the sensory potential findings
are essentially ignored when attempting to define conduction (see Chapter 2). The individual motor unit action potentials
block over relatively large segments of the peripheral nervous have a negative spike duration approximating 5-6 m s .
system such as above elbow-to-wrist or digit recordings. Coincidentally, the duration of the C M A P also has a negative
A significantly different situation is found for CMAPs with spike duration of 5-6 ms. As a result, the single motor units are
respect to phase cancellation and hence diagnostic utility re- primarily in phase when they are generated. This is a com-
garding the differentiation of conduction block from temporal pletely different mechanism from the SNAP as there is a tempo-
dispersion. Recall that the CMAP is the three-dimensional sum- ral component of arrival at the recording electrode combined
mation of individual biphasic motor units contained within the with a significant disparity between the SNAP and the individ-
muscle. These motor units are biphasic negative-positive be- ual action potential duration. In muscle, there is little phase can-
cause they are recorded from the end-plate region and hence cellation with distal stimulation because of the similarity
originate beneath the recording electrode and propagate away between the individual and summated negative spike durations.
Figure 4-19. Compound muscle action potentials. Unlike the SNAP, the single-fiber motor unit action potential and CMAP have similar dura-
tions creating a greater tolerance for any phase overlap, thus minimizing phase cancellation. Additionally, there is half the temporal dispersion for t h e
same distance resulting in little alteration in t h e CMAP between proximal and distal stimulation sites. (From Kimura J, Machida M, Ishida T, e t al:
Relation between size of compound sensory or muscle action potentials, and length of nerve segment. Neurology 1986;36:647-652, with permission.)
148 — PARTI FUNDAMENTAL PRINCIPLES
With more proximal stimulation, there is only a small fraction result from compression, but nevertheless profoundly affect
of the phase cancellation compared to the sensory potential be- peripheral nerves. These conditions can be classified as entrap-
cause there is only half the amount of disparity (12-13 m/s) - 72 73
ment neuropathies (Table 4-7). Although a complete discus-
between the fastest- and slowest-conducting motor fibers as that sion of the various types and clinical presentations of
noted for sensory fibers. The combination of similar negative entrapment neuropathies is not presented in this chapter, the
spike durations for single and compound potentials plus half the basic pathophysiology and histologic appearance of compres-
temporal dispersion results in the single motor unit action po- sion is considered. As with acute compression neuropathies, it
tentials maintaining their "in-phase" relationship with little is difficult to perform histologic investigations on humans with
decrement in amplitude or area and minimal increase in dura- known chronic nerve entrapment syndromes. Animal models
tion (Fig. 4-19). Because of small changes in the above three demonstrating similar pathology to that known to exist in
parameters, CMAP quantification is much more reliable diag- humans is the most acceptable alternative.
nostically than the sensory nerve action potential. The concept Fortunately, an animal model of chronic nerve compression
of phase cancellation has been shown to be the most likely ex- naturally occurs in the guinea-pig. In guinea-pigs aged 2 years
planation of the noted findings with respect to negative spike or older, there is known to exist a progressive compression of
amplitude, duration, and area changes by various computer the median and ulnar nerves at the w r i s t . Histologic sections
9100
modeling programs. It should now be clear why only the

253
of the median nerve distal, across, and proximal to the wrist in
CMAP was used in the above two case examples to assess the older animals were compared with younger specimens to assess
presence of conduction block. the amount of demyelination present. Two basic categories of
The quantitative use of CMAP area, duration, and amplitude lesions were noted. The first consisted primarily of demyeli-
appears to be rather important when the question of quantifying nated fibers at the wrist with little or no degenerative changes
conduction block arises. Unfortunately, the above CMAP para- due to Wallerian degeneration. The fiber content of both large
meters have not be collected for all nerves routinely examined. (greater than 6 pm in diameter) and small (2-5 pm diameter)
It is the practice to generalize the findings noted in the upper nerves proximal to the wrist were unchanged. In other words,
limb to lower limb nerves; however, it is most appropriate to
37
demyelination at the wrist without Wallerian degeneration does
first develop normative parameters prior to assuming that upper not lead to fiber loss proximal to the lesion. The second cate-
and lower limbs are identical. gory of nerve lesion was found in the majority of the oldest ani-
mals where the previously noted demyelination at the wrist had
progressed to Wallerian degeneration. In this instance, the fiber
CHRONIC NERVE COMPRESSION distribution of nerves well proximal to the wrist demonstrated
essentially little, if any, loss of the small fibers, but at least a
The previous discussions have primarily considered the ef- 50% reduction in large fiber content. This is consistent with pre-
fects of acute lesions on the peripheral nervous system. There vious work in which sectioning of the nerve with ensuing
are certainly a large number of conditions that are chronic and Wallerian degeneration resulted in reduced diameter and num-
bers of nerve fibers proximal to the lesion site, most likely aris-
ing from the reactive neuronal cell loss. The importance of the
Table 4-7. Chronic Nerve Compression/Entrapment finding in the guinea pig model is that chronic compression that
Syndrome Etiology causes Wallerian degeneration can produce changes in the fiber
content proximal to the site of the lesion. Unlike acute compres-
I. Compression in a Fibro-Osseous Tunnel sive lesions, the histologic appearance of nerves subjected to
Carpal tunnel syndrome Median nerve compression at wrist chronic compression is quite different.
Cubital tunnel syndrome Ulnar nerve compression at elbow
Recall that in acute compressive lesions there is invagination
Canal of Guyon Ulnar nerve compression at wrist
of one paranodal region into the next with a polarity such that
Supinator syndrome Radial nerve compression mid-
the direction of intussusception is directed away from the two
forearm in o r about supinator muscle
edges of the cuff. A quite different type of lesion in the guinea
Meralgia paresthetica Lateral femoral cutaneous nerve com-
pig is noted for chronic lesions. - Specifically, for distances
198 220
pression about anterior superior iliac

approximating 20 mm proximal to the wrist there is a sequential
spine
tapering of myelin for each internode region toward the lesion.
Tarsal tunnel syndrome Medial/lateral plantar nerve compression
The opposite end of the internode reveals a bulbous accumula-
in foot
tion of myelin. This finding also occurs distal to the compres-
II. Angulation and Stretch sion site. The bulbous end consistently pointed away from the
Tardy ulnar palsy Ulnar nerve deformed at elbow wrist region whether proximal or distal to the wrist. This polar-
Thoracic outlet syndrome Angulation of lower trunk/medial ized anatomic distortion is described as a series of tadpoles on
cord of brachial plexus over cervical either side of the lesion aligned head-to-tail, with the tails di-
rib or band rected toward the lesion (Fig. 4-20). In younger animals there
22i
III. Recurrent External Compressive Force is an asymmetry of myelin only. As the animals age, the tadpole
Ulnar nerve at elbow Leaning on exposed ulnar nerve at lesion becomes more apparent, eventually progressing to de-
the elbow with forearm pronated myelination beneath the compressive zone with remyelination.
Deep branch of ulnar nerve Repetitive trauma, e.g., carpet laying In very advanced lesions, Wallerian degeneration is observed to
in hand with implement t o secure carpet occur in the demyelinated zone.
under molding Electron microscopy reveals the fine details of this anatomic
Modified from Gilliat RW: Chronic nerve compression and entrapment. In distortion and suggests a possible mechanism. - The abnor-
198 222
Sumner AJ (ed):The Physiology of Peripheral Nerve Disease. Philadelphia,W.B. mal appearance of the myelin proximal and distal to the com-
Saunders, 1980, pp 316-339. pressed region suggests that there is slippage of the internal
Figure 4-20. C h r o n i c c o m p r e s s i o n . A, Depiction of guinea

pig nerve showing distortion of the myelin segments.The polar-
ity of the distortion is reversed at t h e wrist region. B,
Continued anatomic distortion with a region of demyelina-
tion/remyelination beneath t h e carpal tunnel area as demon-
strated by the small and increased number of internodes. C ,
Advanced lesion with Wallerian degeneration present. (From
Ochoa J: Nerve fiber pathology in acute and chronic compres-
sion. In O m e r GE, Spinner M (eds): Management of Peripheral
Nerve Problems. Philadelphia,W.B. Saunders, 1980, pp 487-501,
with permission.)
myelin lamellae. This slippage is directed away from the lesion, largest myelinated fibers are affected first and to the greatest
thus accounting for the thinning or tadpole tail. As one pro- degree in compressive lesions. The pathologic basis for this
198
gresses toward the lesion site, the amount of myelin thinning in- slowing of latency and conduction velocity is the observed de-
creases to involve more of the internode. This myelin is myelination in the large myelinated and fast-conducting fibers.
subsequently "pushed" toward the opposite paranodal region, Stimulation of the median nerve during operative intervention
thus accumulating in this area and generating the bulbous end. for carpal tunnel syndrome reveals that there is slowing of con-
At the bulbous paranodal site there is buckling and splitting of duction across the carpal tunnel with conduction block in about
the myelin (Fig. 4-21). 25% of fibers. This finding suggests that patients who experi-
150
The constant and repetitive compressive forces are believed to ence significant weakness most likely have had Wallerian d e -
generate pressure waves. The pressure waves are produced in the generation to account for this loss of muscle power. By
central regions of the compression zone and are directed both continually investigating the conduction velocity of both sen-
proximal and distal to this site decreasing with distance (Fig. 4¬ sory and motor fibers during and after operative intervention, it
21). These directed forces then act on the nerve to "push" the is possible to determine the course of recovery for the injured
myelin away from the central regions of the entrapment. Directly nerves. Intraneural microelectrode studies have demonstrated
beneath the constricting zone there is complete demyelination that within a few minutes of surgical release previously blocked
and eventual Wallerian degeneration. On either side of the of- sensory fibers begin to function. Additionally, within three
44187
fending lesion are the progressive distortions of myelin sequen- months of carpal tunnel release many patients demonstrate sig-
tially tapering and buckling less and less as distance from the nificant increases in conduction velocity both distal and across
lesion increases. Unrolling the myelin sheath would reveal a dis- the affected neural segment. This finding suggests that the re-
placement of the inner lamellae (Fig. 4-21). In support of the ap- covered nerves most likely had experienced conduction block
plicability of this model in man is the histologic finding of and/or segmental demyelination, which is within the time frame
similar anatomic distortions involving the median nerve at the of recovery. This rapid rate of recovery rules out regeneration of
wrist and ulnar fibers at the elbow, °- and the lateral femoral
137 2 1 211
previously denervated nerves.
cutaneous nerve about the inguinal ligament. 157
A further interesting observation is the slowing of motor nerve
In humans there is ample electrophysiologic evidence for conduction proximal to the presumed lesion site in compressive
conduction delay or slowing, conduction block, ectopic impulse syndromes, especially the carpal tunnel syndrome. - - One
9 99100 286
production, and Wallerian degeneration in chronic nerve entrap- should not conclude that this finding is exclusively indicative of
ment s y n d r o m e s . - - -
36 39
- The distal motor latency and
40 150168 286
retrograde degeneration proximal to the site of the lesion, e.g., in
conduction velocity are assessed by measuring the fastest-con- carpal tunnel the median nerve for some undefined reason de-
ducting motor fibers. It has been clearly demonstrated that the generates into the mid- or proximal forearm. There is no question
Figure 4-21. Normal and abnormal segment of myeli-

nated nerve. In the upper segment of the diagram is a normal
internodal myelinated nerve segment (left) with its accompa-
nying myelin sheath unfurled yielding the somewhat trape- "* ^^
=
zoidal shape. Hypothesized pressure waves are depicted t o be

directed in the direction of the arrows tending t o distort the
myelin sheath.The lower portion of the figure shows the dis-
tortion of the segment secondary t o the above-noted pres-
sure waves tending t o "push" the myelin sheath t o the left thus
"bunching" up t h e myelin at o n e end of the internode while
thinning it at the opposite end. Unrolling the myelin sheath re-
veals it is altered as shown. (From Ochoa J: Nerve fiber pathol-
ogy in acute and chronic compression. In O m e r GE, Spinner M
(eds): Management of Peripheral Nerve Problems. Philadelphia,
W.B.Saunders, I980,pp 487-501,with permission.)
ISO — P A R T 1 FUNDAMENTAL PRINCIPLES
that retrograde changes do occur in nerve transection in which TOXIC DEGENERATION

Wallerian degeneration is a major component of the lesion.
Retrograde changes in chronic lesions have not been demon- In this chapter toxic substances are only briefly discussed,
strated when only demyelination is present. As a result there
9100
primarily to compare the pathology produced by these sub-
must be some other explanation for this common finding. It is stances with that of trauma-induced Wallerian degeneration
important to consider two factors: (1) pathology of chronic en- (Table 4-8). Superficially, the insult inflicted upon the periph-
trapment neuropathies, and (2) the recording technique used to eral nervous system by various toxic substances resembles the
observe the motor or sensory potentials. Recall that there is doc- neural disintegration following nerve transection. - A toxic236 237
umented preferential demyelination and loss of the largest material that gains access to the peripheral nervous system es-
myelinated (fastest-conducting) nerve fibers only at the site of sentially results in loss of both the axon and its investing myelin
the lesion or at most 1-2 cm proximal. - - This small seg-
9 100198 218
sheath. Although most substances primarily destroy the axon,
ment of demyelination does not account for the reduction in the close relationship between the axon and myelin results in
nerve conduction velocity affecting the entire forearm segment. myelin absorption by the Schwann cells and invading
The forearm nerve conduction velocities for both sensory and macrophages. In this sense, both toxic degeneration and
motor fibers are usually performed by stimulating the median Wallerian degeneration are similar. The possible exception to
nerve at the wrist and elbow while recording from the thenar em- this finding is that in low dosages lead appears to cause signifi-
inence and digits for motor and sensory fibers, respectively. The cant demyelination, but when taken in large quantities can gen-
important aspect of this technique is that the impulse must cross erate preferential axonal loss. On the other hand, Wallerian
98
the affected segment of nerve. It is incorrect to assume that the degeneration is primarily the result of a localized trauma to a
two stimulation sites do not involve this affected neural region peripheral nerve or nerves with focal consequences with respect
because it is supposedly "subtracted out" by being distal to the to clinical function. Toxic degeneration, however, is a more gen-
two stimulation sites. If there is demyelination or conduction eralized process, as one might anticipate given that a toxic sub-
block affecting the fastest-conducting fibers across the wrist stance has gained entry to the human body and attacks multiple
(proximal to both sensory and motor recording electrodes), then nerves resulting in a more generalized "peripheral" neuropathy.
these fibers never reach the recording electrodes. Only the re- Additionally, toxic degeneration is most likely to produce a
maining slower-conducting fibers are capable of mediating characteristic pattern of axonal loss, i.e., a predilection for an
action potential propagation. It is these slower-conducting fibers, initial distal degeneration of the peripheral nerves. This distal
therefore, that reach the recording sites and are subsequently initiation of axon and myelin loss is referred to as a " d y i n g
measured. The reduced conduction velocities in the nerve seg- b a c k " neuropathy.' It is believed that the process of "dying
19
ment well proximal to the presumed lesion is an artifact of the back" is a result of a preferential insult to the nerve's cell body.
recording electrode position and not suggestive of extensive ret- If the cell body is rendered dysfunctional, it is hypothesized that
rograde changes affecting large portions of the nerve proximal to transport of required substances to maintain an optimal meta-
the site of the injury. This simple principle applies to all potential bolic level results first in failure of the most distal expanse of
entrapment locations when the recording electrodes are placed the nerve. - As less material required for maintaining struc-
50 51
distal to the lesion. This is the same reason noted above for the tural integrity is transported to the periphery, these portions of
caution about attempting to document conduction velocities the axon fail first with a progressive loss of the axon proximally.
across regions of conduction block, i.e., comparing different It must be remembered, however, that the entire neuron is af-
fiber populations. fected even though the neural dysfunction and pathology is first
Of note, mixed nerve action potential techniques specifi- manifested distally. This process can affect both motor and sen-
cally investigating the forearm slowing in patients with carpal sory nerves. In short, Wallerian degeneration is a focal process
tunnel syndrome reveal a number of interesting findings. - 232 293
typically following trauma with loss of axonal structural in-
If the median nerve is stimulated at the wrist well proximal to tegrity distal to the site of neural insult, while toxic degenera-
the several centimeters of possible retrograde neural degeneration is a generalized insult to the entire neuron with multiple
tion and the ensuing mixed nerve action potential is recorded nerves displaying a dying-back type of degeneration. In the case
at the elbow, a mild degree of forearm slowing is indeed of the sensory neuron, it should be realized that both the distal
noted. This slowing, however, is not sufficient to account for
52
peripheral and central axons (or in fact dendrites) of the sensory j
the observed forearm motor conduction slowing when mea- nerves can be affected. The last statement means that the central
sured distal from the carpal tunnel and overshadowed by a sig- axons lying in the dorsal column of the spinal cord might also
nificant reduction in the mixed nerve action potential's affected by the disease process. - 231 289
amplitude compared to normal individuals. The conclusion of Interestingly, the nerve conduction velocities in the majority
these findings is that there is some degree of axonal loss in the of toxic degenerative processes investigated are only mildly re-
forearm segment of the median nerve in significant median duced initially. There is usually less than a 10-30% reduction in
nerve wrist compression lesions as demonstrated by a reduced maximal conduction velocity. - - - Of course, the exception
97 101 151 156
mixed nerve amplitude. The reduced forearm conduction ve- to this finding is in low-dose lead intoxication where extensive
locity is most likely a result of the technique artifact of con- demyelination can produce conduction velocities approaching
duction block at the wrist. From the information described in
52
50% of normal. A reduction in conduction velocity irrespective
this chapter, it is likely for patients with profound carpai of etiology usually implies that the large fibers mediating either
tunnel syndrome to have variable degrees of axonal loss sec- the motor or sensory responses are affected. It is also important
ondary to the wrist compression. This may lead to motor and to remember, however, that individual toxic substances may
sensory neuronal cell death with subsequent degeneration and present with unique damage to peripheral nerves with respect to
loss of these fibers from the neuron cell body distally, thereby the amount of axonal loss or demyelination.
accounting for the reduced mixed nerve action potential de- Following repair of the degenerative process, assuming the
tected by some investigators. - - 52 232 293
offending substance has been removed, one can anticipate

Table 4-8. Toxins Producing Peripheral Nerve system can be involved. Large decreases in neural conduction
Degeneration velocity can result from segmental demyelination depending
I. Industrial Chemicals upon the extent of the lesion. - - - In particular, patients
76 78 287 288
A. Affects Peripheral Nervous System Preferentially with chronic familial demyelinating/remyelinating neuropathies
Lead can have conduction velocities approaching 5 m/s or less.
Acrylamide Decreased conduction velocity, therefore, is a hallmark of a pro-
Organophosphates found demyelinating process affecting the peripheral nerve. If
Thallium the particular process results in demyelination affecting some
B. Some Effects on Central Nervous System fibers more than others over a particular portion of the nerve,
Carbon disulfide one can anticipate not only a decrease in the conduction veloc-
Methyl mercury ity but also a reduced amplitude and a dispersed potential. This
Methylbromide is a result of the disparate slowing of conduction for multiple
C. Large Amounts Required fibers, thus leading to a marked increase in temporal dispersion
Arsenic with an associated reduction in potential magnitude.
Trichloroethyiene
Tetrachloroethane MOTOR AND SENSORY NEURONOPATHY
2,4-D (Dichlorophenoxyacetic acid)
Pentachlorophenol A number of disease processes or toxins directly attack the
DDT cell bodies of the peripheral nervous system. These are often re-
D. Some Effects on Other Than Nervous Tissue stricted to either the motor or sensory neurons although an over-
Carbon Tetrachloride lap can occur. Motor neuron diseases are well known examples
Carbon Monoxide and generally result in selective loss of motor neurons. Usually
these are sporadic cases (amyotrophic lateral sclerosis) of un-
II. Pharmaceutical Substances
known cause. Other possible causes are monomelic spinal mus-
Amiodarone Hydralazine
cular atrophy, paraneoplastic syndromes, radiation damage or as
Antinucleosides Isoniazid
part of a more generalized degenerative neurologic disease,
Arsenic Nitrofurantoin
such as different types of dementias. ' - - - Viral invasion
4 36 143 241 255
Chemotherapeutic agents Phenytoin

also can result in motor neuronopathy (herpes zoster or p o -
Chloroquine Sulfonamides
liomyelitis). ' The result is a global Wallerian degeneration of
83 131
Colchicine Thalidomide
the entire axon following the death of the cell body. Elec-
Disulfiram Thallium
trophysiologically, this results in findings compatible with pure
Gold Vitamin B intoxication
6
motor axonal loss without a proximal-distal distribution. In
Modified after Gilliatt RW: Recent advances in the pathophysiology of nerve ALS, for example, the sensory conduction and amplitudes
conduction. In Desmedt JE (ed): New Developments in Electromyography and
remain normal, the motor conduction is only slightly affected,
Clinical Neurophysiology. Basel, Karger, 1973, pp 2-18.
and the CMAPs diminish in amplitude. Needle EMG shows a
combination of de- and reinnervation besides fascicula-
regeneration of at least some of the lost axons. Provided the in- tions.35,255
In addition to the peripheral loss, there is a central
dividual has not ingested too large a quantity of the toxic mater- component due to the involvement of the upper motor neuron.
ial and is capable of reconstituting the injured nerves, one can The sensory neuronopathies are a less well known, heteroge-
anticipate the expected findings based on what is known from neous, and not well understood group of disorders. A selective
nerve repair after degeneration. Because the endoneurial tubes targeting of the primary sensory neurons in the dorsal root gan-
were not destroyed by the toxic substance, aberrant regenera- glia is found. The etiology ranges from paraneoplastic, toxic,
121
tion should be absent. The proliferation of Schwann cells with immune-mediated to viral.' ' '- The onset can be acute with
2 2 324
axonal regrowth is the expected result. Recall that the conse- extensive sensory loss of all kinds of modalities within several
quence of regeneration is smaller internodal distances with thin- days to insidious and slowly progressive over many years (Table
ner myelin sheaths, particularly early after the nerve is 4 - 9 ) . The acute cases usually have prominent dysesthesias. 207
regenerated. As a result, even though the initial conduction ve- Often a special class of neurons is affected, for example the
locities are little affected, once the nerves have degenerated and large-diameter neurons resulting in sensory ataxia, propriocep-
undergone repair, a reduction in conduction velocity over the af- tive disturbances and areflexia. The paraneoplastic sensory neu-
fected segment may be observed with normal conduction proxi- ronopathy associated with small-cell lung cancer is due to a
mally. For nerves experiencing only distal degeneration, a circulating anti-Hu antibody. In rare cases a combination of a
207
prolonged distal motor latency with possible dispersion of the sensory neuronopathy can be seen, and a demyelinating neu-
evoked response may be observed. -' 101 40
ropathy complicating diagnosis further." In an animal model of
toxic sensory neuronopathy, it was found that megadoses of
EXTENSIVE SEGMENTAL (GENERALIZED) pyridoxine resulted in disruption of the cell metabolism fol-
DEMYELINATION lowed by degeneration of the entire axons. Megadosis of pyri- 182
doxin in humans is also known to result in a severe, acute

In addition to the focal segmental demyelination caused by sensory neuronopathy. 7
compression or entrapment, more extensive segmental demyeli- Important findings in electrodiagnostic studies are a decline
nation can result from a number of disease processes. These in sensory amplitudes concomitant with the clinical course and
processes are associated with various peripheral neuropathies loss of H-reflexes. The distribution is not compatible with a
resulting from varied etiologies (Table 4 - 8 ) . Comparatively length-dependent axonopathy. Often generalized absent sensory
more or quite extensive portions of the peripheral nervous responses are found. The motor conduction velocities and
Table 4-9. Rate of Evolution of Sensory Neuropathies and 15. Barr ML, Hamilton JD: A quantitative study of certain morphological changes
Related Puresensory Neuropathies in spinal motor neurons during axon reaction. J Comp Neurol 1948;89:93-121.
16. Barr ML, Bertram EG: A morphological distinction between neurons of the
Type Rate of Evolution Examples male and female, and the behaviour of the nucleolar satellite during accelarated
nucleoprotein synthesis. Nature 1949;163:676-677.
Acute Days/1 - 2 weeks Acute sensory neuronopathy 17. Berry CM, Grundfest H, Hinsey JC: The electrical activity of regenerating
nerves in the cat. J Neurophsyiol 1944;7:103-115.
Subacute Weeks/few months Paraneoplastic sensory neu-
18. Berry CM, Hinsey JC: The recovery of diameter and impulse conduction in re-
ronopathy generating nerve fibers. Ann N Y Acad Sci 1946;47:559-575.
Sjogren's syndrome 19. Berthold CH: Ultrastructural appearance of the node-paranode region of mature
Toxic: cisplatin, pyridoxin abuse, feline ventral lumbar spinal root fibers. Acta Soc Med Uppsala 1968;73(Suppl
9):37-70.
taxol, doxorubicin 20. Bethe A: Allgemeine Anatomie und Pathologie des Nervensystems. Leipzig,
Chronic Years Chronic idiopathic sensory Thieme, 1903.
21. Beuche W, Friede RL: The role of nonresident cells in Wallerian degeneration. J
neuronopathy
Neurocytol 1984;13:767-796.
Chronic sensory neuronopathy 22. Bielschowsky M, Valentin B: Die histologischen Veranderungen in durchfrore-
with dysproteinemia nen Nervenstricken. J Psychol Neurol 1928;22:267-318.
Hereditary sensory neuropathies 23. Birks R, Katz B, Miledi R: Phsyiological and structural changes at the amphib-
ian myoneural junction, in the course of nerve degeneration. J Physiol
Modified from Asbury AK, Brown MJ: Sensory neuronopathy and pure sensory 1960;150:145-168.
neuropathy. Curr Opin Neurol Neurosurg 1990;3:708-711. 24. Black JA, Foster RE, Waxman SG: Freeze-fracture ultrastructure of rat CNS
PNS nonmyelinated axolemma. J Neurocytol 1981;19:981-993.
25. Black JA, Friedman B, Waxman SG, et al: Immuno-ultrastructural localization
of sodium channels at nodes of Ranvier and perinodal astrocytes in rat optic
amplitudes are normal. There are no signs of demyelination, nerve. Proc R Soc Lond (Biol) 1989;238:39-51.
conduction blocks^ or waveform dispersion. - 12 324
26. Blackwood W, Russel H: Experiments in study of immersion foot. Edin Med J
1943;50:385-398.
27. Bodian D: Nucleic acid in nerve cell regeneration. Symp Soc Exp Biol (No. 1,
Nucleic Acids). London, Cambridge University Press, 1947, pp 163-178.
CONCLUSION 28. Bostock H, Sears TA: The internodal axon membrane: Electrical excitability and
continuous conduction in segmental demyelination. J Physiol 1978;280: 273-301.
The information presented in this chapter is vital to appreci- 29. Bostock H, Sherratt RM, Sears TA: Overcoming conduction failure in demyeli-
nated nerve fibers by prolonging action potential. Nature 1978;274:385-387.
ating the manner in which the peripheral nervous system re- 30. Bostock H, Sears TA, Sherrat RM: The effects of 4-aminopyridine and
sponds to the various disease entities discussed throughout the tetraethylammonium ions on normal and demyelinated mammalian nerve
remainder of this book. Irrespective of the disease process pro- fibers. J Physiol 1981;313:301-315.
31. Bostock H, Grafe P: Activity-dependent excitability changes in normal and de-
ducing damage, the manner in which the nervous system is myelinated spinal root axons. J Physiol 1985;365:239-257.
likely to respond is found in this chapter. The practitioner is 32. Bots G Th: Pathology of nerves. In Vinken PJ, Bruyn GW: Handbook of j
urged to master this information prior to reading the succeeding Clinical Neurology, Vol 7, Diseases of Nerves (Part I). Amsterdam, North- '
chapters dealing with specific disease processes, as they all Holland Pub. Co., 1970, pp 197-243.
33. Brill MH, Waxman SG, Moore JW, et al: Conduction velocity and spike config-
build upon the basic principles outlined above. uration in myelinated fibers: computed dependence on internode distance. J
Neurol Neurosurg Psychiatry 1977;40:769-774.
34. BrismarT: Potential clamp analysis of membrane currents in rat myelinated
nerve fibers. J Physiol 1980;298:171-184.
REFERENCES 35. Bromberg MB: Electrodiagnostic studies in clinical trials for motor neuron dis-
ease. J Clin Neurophysiol 1998; 15(2): 117-128.
1. Abercrombie M, Johnson ML: Outwandering of cells in tissue culture of nerves 36. Brown WF, Ferguson GG, Jones MW, et al: The location of conduction abnor-
undergoing Wallerian degeneration. J Exp Biol 1942;19:266-283. malities in human entrapment neuropathies. Can J Neurol Sci 1976;3:111-122.
2. Abercrombie M, Johnson ML: Quantitative histology of Wallerian degenera- 37. Brown WF, Feasby TE: Conduction block and denervation in Guillain-Barre
tion: Nuclear population in rabbit sciatic nerve. J Anal 1946;80:37-50. polyneuropathy. Brain 1984;107:219-239.
3. Acheson GH, Lee ES, Morison RS: A deficiency in the phrenic respiratory dis- 38. Brown WF: The Physiological and Technical Basis of Electromyography.
charges parallel to retrograde degeneration. J Neurophysiol 1942;5:269-273. Boston, Butterworth Publishers, 1984, p 52.
4. Ahlskog JE, Litchy WJ, Peterson RC, et al: Guamanian neurodegenerative dis- 39. Buchthal F, Rosenfalek A: Sensory conduction from digit to palm and from
ease: Electrophysiologic findings. J Neurol Sci 1999;166:28-35. palm to wrist in the carpal tunnel syndrome. J Neurol Neurosurg Psychiatry
5. Aitken JT, Sharman M, Young JZ: Maturation of regenerating fibers with vari- 1971;34:243-252.
ous peripheral connections. J Anat 1947;81:1 - 2 2 . 40. Buchthal F, Rosenfalek A, Trojaborg W: Electrophysiologic findings in entrap- j
6. Aitken JT, Thomas PK: Retrograde changes in fiber size following nerve sec- ment of the median nerve at the wrist and elbow. J Neurol Neurosurg Psychiatry 1
tion. J Anat 1962;96:121-129. 1974,37:340-360.
7. Albin RL. Albers JW, Greenberg HS, et al: Acute sensory neuropathy-neu- 41. Bunger O: Ober die Degenerations und Regenerationsvorgange an Nerven nach
ronopathy from pyridoxine overdose. Neurology 1987;37:1729-1732. Verletzungen. Beitr Pathol Anat 1891; 10:321 -393.
8. Anden NE, Olsson Y: 5-Hydroxytryptamine in normal and sectioned rat sciatic 42. Cangiano A, Lutzemberger L, Nicotra L: Non-equivalence of impulse blockade
nerve. Acta Pathol Microbiol Scand 1967;70:537-540. and denervation in the production of membrane changes in rat skeletal muscle.
9. Anderson MH, Fullerton PM, Gilliatt RW, et al: Changes in the forearm associ- J Physiol 1977;273:691-706.
ated with median nerve compression at the wrist in the guinea-pig. J Neurol 43. Caruso G, Labianca O, Ferrannini E: Effect of ischemia on sensory potentials
Neurosurg Psychiatry 1970;33:70-79. of normal subjects of different ages. J Neurol Neurosurg Psychiatry 1973;36:
10. Andres KH: Untersuchungen Uber morphologische Veranderunkgen in Spinal- 455-466.
ganglien wahrend der retrograden Degeneration. Z Zellforsch Mikrosk Anat 44. Casey EB, Le Quesne PM: Digital nerve action potentials in healthy subjects,
11961:55:49-79. and in carpal tunnel and diabetic patients. J Neurol Neurosurg Psychiatry
11. Antoine JC, Mosnier JF, Honnorat J, et al: Paraneoplastic demyelinating neu- 1972;35:612-623.
ropathy, subacute sensory neuropathy, and anti-Hu antibodies: clinicopalholog- 45. Castaigne P, Cathala HP, Beaussart-Boulenge L, et al: Effect of ischemia on pe-
ical study of an autopsy case. Muscle Nerve 1998;21:850-857. ripheral nerve function in patients with chronic renal failure undergoing dialysis
12. Asbury AK, Brown MJ: Sensory neuronopathy and pure sensory neuropathy. treatment. J Neurol Neurosurg Psychiatry 1972;35:631-637.
Curr Opin Neurol Neurosurg 1990,3:708-711. 46. Causey G, Palmer E: Early changes in degenerating mammalian nerves. Proc R
13. Baker M, Bostock H, Grafe P, et al: Function and distribution of three types of Soc Lond (B) 1952;139:597-609.
rectifying channel in rat spinal root myelinated axons. J Physiol 1987,383: 47. Causey G, Palmer E: The centrifugal spread of structural change at the nodes in
45-67. degenerating mammalian nerves. J Anat 1953;87:185-191.
14. Ballin RHM, Thomas PK: Electron microscope observations on demyelination 48. Causey G, Stratmann CJ: The spread of failure of conduction in degenerating
and remyelination in experimental allergic neuritis. J Neurol Sci 1968;8:1-18. mammalian nerves. J Physiol 1953;121:215-223.
49. Cavanagh JB, Jacobs JM: Some quantitative aspects of cliptheritic neuropathy. 84. Eisen A: Electrodiagnosis of radiculopathies. In Aminoff MJ (ed): Neurologic
Br J Exp Pathol 1964;45:309-322. Clinics: Symposium on Electrodiagnosis, Vol. 3. Philadelphia, W.B. Saunders
50. Cavanagh JB: The significance of the "dying back" process in experimental and 1985, pp 495-510.
human neurological disease. Int Rev Exp Path 1964;3:219-267. 85. Elmer LW, Black JA, Waxman SG, et al: The voltage-dependent sodium chan-
51. Cavanagh JB: Organo-phosphorous neurotoxicity and the "dying back" nel in mammalian PNS and CNS: Antibody characterization and immunocyto-
process. In Norris Kurland (ed): Motor Neuron Diseases. New York, Grune & chemical localization. Brain Res 1990;532:222-231.
Stratum, 1969, pp 292-300. 86. Eng DL, Gordon TR, Kocsis JD, et al: Current clamp analysis of time-depen-
52. Chang M-H, Liao K-K, Chang S-P, et al: Proximal slowing in carpal tunnel syn- dent rectification in rat optic nerve axons. J Physiol 1990;421:185-202.
drome resulting from either conduction block or retrograde degeneration. J 87. Engelmann THW: Ober Degeneration von Nervenfasern. Pflugers Arch Ges
Neurol 1993;240:287-290. Physiol 1876;13:474-491.
53. Chaudhry V, Cornblath DR: Wallerian degeneration in human nerves: Serial 88. Engh CA, Schofield BH, Doty SB, et al: Perikaryal synthetic function follow-
electrophysiologic studies. Muscle Nerve 1992:15:687-693. ing reversible and irreversible peripheral axon injuries as shown by radioautog-
54. Chiu SY, Ritchie JM, Rogart RB, et al: A quantitative description of membrane raphy. J Comp Neurol 1971; 142:465^479.
currents in rabbit myeliated nerve. J Physiol 1979;292:149-166. 89. Erb W: Diseases of the peripheral cerebro-spinal nerves. In Von Ziemssen H
55. Chiu SY, Ritchie JM: Evidence for the presence of potassium channels in the (ed): Cyclopaedia of the Practice of Medicine (Vol XI). London, Samson Low,
internodal region of acutely demyelinated mammalian single nerve fibers. J Marston, Searle and Rivington, 1876.
Physiol 1981;313:415-437. 90. Erlanger J, Schoepfle G: A study of nerve degeneration and regeneration. Am J
56. Chiu SY, Ritchie JM: On the physiological role of internodal potassium chan- Physiol 1946;147:550-581.
nels and the security of conduction in myelinated nerves. Proc R Soc Lond 91. Fisher CM, Adams RD: Diphtheritic polyneuritis: A pathologic study. J
1984;220:415-422. Neuropath Exp Neurol 1962;21:105-115.
57. Cragg BG, Thomas PK: Changes in conduction velocity and fibre size proximal 92. Flesenthal G: An overview of the clinical application of electromyography and
to peripheral nerve lesions. J Physiol 1961;157:315-327. nerve conduction techniques. MD State Med J 1982;31:60-62.
58. Cragg BG, Thomas PK: Changes in nerve conduction in experimental allergic 93. Fowler TJ, Danta G, Gilliatt RW: Recovery of nerve conduction after a pneu-
neuritis. J Neurol Neurosurg Psychiatry 1964;27:106-115. matic tourniquet: observations on the hind-limb of the baboon. J Neurol
59. Cragg BG, Thomas PK: The conduction velocity of regenerated peripheral Neurosurg Psychiatry 1972;35:638-647.
nerve fibres. J Physiol 1964;171:164-175. 94. Frankenhaeuser B, Moore LE: The effect of temperature on sodium and potas-
60. Cravioto H: Wallerian degeneration: Ultrastructural and hisfochemical studies. sium permeability changes in myelinated nerve fibres of xenopus laevis. J
Bull Los Ang Neurol Soc 1969;34:233-253. Physiol 1963;169:431-437.
61. Dahlstrom A, Haggendal J: Studies on the transport and life-span of amine stor- 95. Friede RL: Transport of oxidative enzymes in nerve fibers: A histochemical in-
age granules in a peripheral adrenergic neuron system. Acta Physiol Scand vestigation of the regenerative cycle in neurons. Exp Neurol 1959;1:441-466.
1966;67:278-288. 96. Friede RL, Martinez AJ: Analysis of the process of sheath expansion in swollen
62. Davenport HA, Chor HA, Dolkart RE: The ratio of myelinated to unmyelinated nerve fibers. Brain Res 1970;19:165-182.
fibers in regenerated sciatic nerves of macacus rhesus. J Comp Neurol 97. Fullerton PM, Barnes JM: Peripheral neuropathy in rats produced by acry-
1937;67:483-491. lamide. Br J Industr Med 1966;23:210-221.
63. Davis FA, Jacobson S: Altered thermal sensitivity in injured and demyelinated 98. Fullerton PM: Chronic peripheral neuropathy produced by lead poisoning in
nerve; A possible model for temperature effects in multiple sclerosis. J Neurol guinea-pigs. J Neuropath Exp Neurol 1966;25:214-236.
Neurosurg Psychitry 1971;34:551-561. 99. Fullerton PM, Gilliatt RW: Pressure neuropathy in the hind foot of the guinea-
64. Davis FA: Impairment of repetitive impulse conduction in experimentally de- pig. J Neurol Neurosurg Psychiatry 1967;30:18-25.
myelinated and pressure-injured nerves. J Neurol Neurosurg Psychiatry 100. Fullerton PM, Gilliatt RW: Median and ulnar neuropathy in the guinea-pig. J
1972;35:537-544. Neurol Neurosurg Psychiatry 1967;30:393-402.
65. Denny-Brown D, Brenner C: Paralysis of nerve induced by direct pressure and 101. Fullerton PM: Electrophysiological and histological observations on peripheral
by tournique. Arch Neurol Psychiatry 1944;51:1-26. nerves in aerylamide poisoning in man. J Neurol Neurosurg Psychiatry
66. Denny-Brown D, Sciarra D: Changes in the nervous system in acute porphyria. 1969;32:186-192.
Brain 1945;68:1-16. 102. Gibson JD: The origin of the neural macrophage: A quantitative ultrastructural
67. Denny-Brown D: Importance of neural fibroblasts in the regeneration of nerve. study of cell population changes during Wallerian degeneration. J Anat
Arch Neurol Psychiatry 1946;55:151-215. 1979;129:1-19.
68. Dixon JS: Changes in the fine structure of neuron after axon section. J Anat 103. Gilliatt RW, Wilson TG: Ischaemic sensory loss; in patients with peripheral
1968;103:396. nerve lesions. J Neurol Neurosurg Psychiatry 1954,17:104-114.
69. Dixon JS: Some fine structural changes in sympathetic neurons following axon 104. Gilliat RW, Taylor JC: Electrical changes following section of the facial nerve.
section. Acta Anat 1970;76:473-487. Proc Soc Med 1959;52:1080-1083.
70. Dodge FA, Cooley JW: Action potential of the motor neuron. IBM J Res Dev 105. Gilliatt RW, Willison RG: Peripheral nerve conduction in diabetic neuropathy. J
1973;17:219-229. Neurol Neurosurg Psychiatry 1962;25:11-18.
71. Donat JR, Wisniewski HM: The spatio-temporal pattern of Wallerian degenera- 106. Gilliatt RW, Willison RG: The refractory and supranormal periods of the human
tion in mammalian peripheral nerves. Brain Res 1973;53:41-53. median nerve. J Neurol Neurosurg Psychiatry 1963;26:136-147.
72. Dorfman LJ, Cummins KL, Abraham GS: Conduction velocity distributions of 107. Gilliat RW, Hjorth RJ: Nerve conduction during Wallerian degeneration in the
the human median nerve: Comparison of methods. Muscle Nerve I982;5: baboon. J Neurol Neurosurg Psychiatry 1972;35:335-341.
S148-S153. 108. Gilliatt RW: Recent advances in the pathophysiology of nerve conduction. In
73. Dorfman LJ: The distribution of conduction velocity (DCV) in peripheral Desmedt JE (ed): New Developments in Electromyography and Clinical
nerves: A review. Muscle Nerve 1984;7:2-11. Neurophysiology. Basel, Karger, 1973, pp 2-18.
74. Duncan D, Jarvis WH: Comparison of actions on nerve fiber of certain anes- 109. Gilliatt RW, McDonald WI, Rudge P: The site of conduction block in peripheral
thetic mixtures and substances in oil. Anaesthesiology 1943;4:465-474. nerves compressed by a pneumatic tourniquet. J Physiol 1974;238:31P-32P.
75. Duncan D: Alterations in the structure of nerves caused by restriction of their 110. Gilliat RW: Peripheral nerve compression and entrapment. In Lant AF(ed): The
growth with ligatures. J Neuropath Exp Neurol 1948;7:261-273. Oliver Sharpey lecture, 11th Symposium on Advanced Medicine of the Royal
76. Dunn HG, Lake BD, Dolman CL, et al: The neuropathy of Krabbe's infantile College of Physicians. London, Pitman, 1975, pp 144-163.
cerebral sclerosis. Brain 1969;92:329-344. 111. Gilliatt RW. Westgaard RH, Williams IR: Extrajunctional acetylcholine sensi-
77. Dustin AP: Les lesions posttraumatiques des nerfs. Ambulance de L'ocean tivity of inactive muscle fibres in the baboon during prolonged nerve pressure
1917;1:71-162. block. J Physiol 1978;280:499-514.
78. Dyck PJ, Lambert EH: Lower motor and primary sensor)' neuron disease with 112. Gilliatt RW: Acute compression block. In Sumner AJ (ed): The Physiology of
peroneal muscular atrophy. I. Neurologic, genetic, and electrophysiologic find- Peripheral Nerve Disease. Philadelphia. W.B. Saunders, 1980. pp 287-315.
ings in hereditary polyneuropathies. Arch Neurol 1968;18:603-618. 113. Gilliat RW: Chronic nerve compression and entrapment. In Sumner AJ (ed):
79. Dyck PJ, Hopkins AP: Electron microscopic observations on degeneration and The Physiology of Peripheral Nerve Disease. Philadelphia, W.B. Saunders.
regeneration of unmyelinated fibers. Brain 1972;95:223-234. 1980, pp 316-339.
80. Ebeling P, Gilliat RW, Thomas PK: A clinical and electrical study of ulnar nerve 114. Glimstedt G. Wohlfahrt G: Electron microscopic observations on Wallerian de-
lesions in the hand. J Neurol 1960;23:1-9. generation in peripheral nerves. Acta Morph Neerl-Scand 1960;3:135-146.
81. Eccles JC, Kmjevic K, Miledi R: Delayed effects of peripheral severance of af- 115. Gombault A: Contribution a f etude anatomique de la nevrite parenchymateu.se
ferent nerve fibers on the efficacv of their central synapses. J Physiol subaigue et chronique—nevrite segmentaire peri-axile. Arch Neurol (Paris)
1959;145:204-220. 1880;1:11-38.
82. Edds A: Collateral regeneration of residual motor axons in partially denervated 116. Gordon TR, Kocsis JD, Waxman SG: Evidence for the presence of two types of
muscles. J Exp Zool 1950; 112:517-522. potassium channels in rat optic nerve. Brain Res 1988;447:1-10.
83. Einarson L: On internal structure of motor cells of anterior horn and its changes 117. Gordon TR, Kocsis JD, Waxman SG: Electrogenic pump activity in rat optic
in poliomyelitis. Acta Orthop Scand 1949;19:27-54. nerve. Society Neurosci 1989:15:77.
118. Grant G: Neuronal changes central to the site of axon transection: A method for 152. Howe JF, Loeser JD, Calvin WH: Mechanosensitivity of dorsal root ganglia and
the identification of retrograde changes in perikarya, dendrites and axons by chronically injured axons: A physiological basis for the radicular pain of nerve
silver impregnation. In Nauta WJH, Effesson SEO (eds): Contemporary root compression. Pain 1977;3:25^H.
Research Methods in Neuroanatomy. Berlin, Springer-Verlag, 1970, pp 153. Huxley AF: Ion movements during nerve activity. Ann NY Acad Sci
173-185. 1959;81:221-246.
119. Greenfield JC: The Spinocerebellar Degeneration. Oxford, Blackwell, 1954. 154. Jacobs JM, Cavanagh JB: Species differences in internode formation following
120. Greenfield JG, Meyer A: General pathology of the nerve cell and neuroglia. In two types of peripheral nerve injury. J Anat 1969;105:295-306.
Greenfield JG (ed): Neuropathology. London, Arnold, 1963. 155. Jacobson S, Guth L: An electrophysiological study of early stages of peripheral
121. Griffin JW, Cornblath DR, Alexander E, et al: Ataxic sensory neuropathy and nerve regeneration. Exp Neurol 1965; 11:48-60.
dorsal root ganglionitis associated with Sjogren's syndrome. Ann Neurol 1990; \56. Janz D, Neundorfer B: Klinische und electromygraphische Untersuchungen
27:304-315. nach Triarylphosphat-Polyneuropathie. Dtseh Z Nervenheilk 1962;183;268-304.
122. Gudden H: Klinische und anatomische Beitrage zur Kenntnis der multiplen 157. Jefferson D, Eames RA: Subclinical entrapment of the lateral femoral cuta-
Alkoholneuritis nebst Bemerkungen iiber die Regenerationsvorgange im pe- neous nerve: An autopsy study. Muscle Nerve 1979;2:145-154.
ripheren Nervensystem. Arch Psychiat Nervenkr 1896;28:1-99. 158. Johnson EW, Melvin JL: Value of electromyography in lumbar radiculopathy.
123. Guth L: Regeneration in the mammalian peripheral nervous system. Physiol Arch Phys Med Rehabil 1971;52:239-243.
Rev 1956;36:441-478. 159. Johnson ML, Abercrombie M: Collagen content of rabbit sciatic nerve during
124. Guthrie TC: Visual and motor changes in patients with multiple sclerosis. Arch Wallerian degeneration. J Neurol Neurosurg Psychiatry 1946;9:113-118.
Neurol Psychiatry 1951;65:437-440. 160. Kaeser HE, Lambert EH: Nerve function studies in experimental polyneuritis.
125. Gutmann E, Sanders FK: Functional recovery following nerve grafts and other Electroencephalogr Clin Neurophysiol 1962(Suppl 22):29-35.
types of nerve bridge. Brain 1942;65:373-408. 161. Kaji R, Suzumura A, Sumner AJ: Physiological consequences of antiserum-me-
126. Gutmann E, Sanders FK: Recovery of fiber numbers and diameters in the re- diated experimental demyelination in CNS. Brain 1988; 111:675-694.
generation of peripheral nerves. J Physiol 1943;101:489-518. 162. Kaji R, Sumner AJ: Effects of 4-aminopyridine in experimental CNS demyeli-
127. Gutmann E, Holubar J: Failure of transmission of motor and sensory nerve im- nation. Neurology 1988;38:1884-1887.
pulses after nerve section. Nature 1949;163:328-329. 163. Kaji R, Sumner AJ: Effects of digitalis on central demyelinative conduction
128. Gutmann E, Holubar J: Degeneration of peripheral nerve fibers. J Neurol block in vivo. Ann Neurol 1989;25:159-165.
Neurosurg Psychiatry 1950;13:89-105. 164. Kaji R, Sumner AJ: Oubain reverses conduction disturbances in single demyeli-
129. Gutmann E, Holubar J: Atrophy of nerve fibres in the central stump following nated fibers. Neurology 1989;39:1364-1368.
nerve section and the possibilities of its prevention. Arch Int Stud Neurol 165. Kaji R, Happel L, Sumner AJ: Effects of digitalis on clinical symptoms and
1951;1:1-11. conduction variables in patients with multiple sclerosis. Ann Neurol
130. Gutmann E, Holubar J: Degenerace terminalnich organu v pricne pruhovanem a 1990;28:582-584.
hladkem svalstvu. Ceskoslovenska Fysiologie 1952;1:168-175. 166. Karlstrom L, Dahlstrom A: The effect of different types of axonal trauma on the
131. Haanpaa M, Hakkinen V, Nurmikko T: Motor involvement in acute herpes synthesis and transport of amine storage granules in rat sciatic nerves. J
zoster. Muscle Nerve 1997; 20:1433-1438. Neurobiol 1973;4:191-200.
131. Haftek J, Thomals PK: An electron-microscope study of crush injuries to pe- 167. Kerns JM, Hinsman EJ: Neuroglial response to sciatic neurectomy. I. Light mi-
ripheral nerve. J Anat 1967;102:154-156. croscopy and autoradiography. J Comp Neurol 1973;151:237-254.
133. Haftek J, Thomas PK: Electron-microscope observations on the effects of local- 168. Kimura J: The carpal tunnel syndrome: Localization of conduction abnormali-
ized crush injuries on the connective tissues of peripheral nerve. J Anat ties within the distal segment of the median nerve. Brain 1979;102:619-635.
1968;103:233-243. 169. Kimura J, Machida M, Ishida T, et al: Relation between size of compound sen-
134. Hager H: Die Feinere Cytologic und Cytopathologie des Nervensystems. sory or muscle action potentials, and length of nerve segment. Neurology
Stuttgart, Fischer, 1964. 1986;36:647-652.
135. Hall JI: Studies on demyelinated peripheral nerves in guinea-pigs with experi- 170. Kiraly JK, Krnjevic K: Some retrograde changes in function of nerves after pe-
mental allergic neuritis: A histologic and electrophysiological study. Part II. ripheral section. Quart J Exp Physiol 1959;64:244-257.
Electrophysiological observations. Brain 1967;90:313-332. 171. Kline DG, DeJonge BD: Evoked potentials to evaluate peripheral nerve in-
136. Harding AE, Bradbury PG, Murray NMF: Chronic asymmetrical spinal muscu- juries. Surg Gynecol Obstet 1962;127:1239-1248.
lar atrophy. J Neurol Sci 1983; 59:69-83. 172. Kline DG, Hackett ER, Happel LR: Function of the regenerating distal segment
137. Harriman DGF: Ischemia of peripheral nerve and muscle. J Clin Pathol (Suppl) of nerve. In Gorio A, Millesi H, Miringo S (eds): Post Traumatic peripheral
1977;11:94-104. nerve regeneration. Experimental Basis and Clinical Implications. New York,
138. Harrison MJG: Pressure palsy of the ulnar nerve with prolonged conduction Raven Press, 1981, pp 445-447.
block. J Neurol Neurosurg Psychiatry 1976;36:96-99. 173. Kocsis JD, Waxman SG: Absence of potassium conductance in central myeli-
139. Hentowa F: Regeneration der peripherischen Nerven. Z Ges Neurol Psychiat nated axons. Nature 1980;287:348-349.
1933;147:791-806. 174. Kocsis JD, Waxman SG: Action potential electrogenesis in mammalian central
140. Hern JEC: Some effects of experimental organo-phosphorus intoxication in pri- axons. In Waxman SG, Ritchie JM (eds): Demyelinating Diseases: Basis and
mates. DM Thesis, University of Oxford, 1971. Clinical Electrophysiology. New York, Raven Press, 1981, pp 299-312.
141. Hirano A, Becker MH, Zimmerman HM: Isolation of the periaxonal space of 175. Kocsis JD, Ruiz JA, Cummins K: Modulation of axonal excitation mediated by
central myelinated fibers with regard to diffusion of peroxidase. J Histochem surround electrical activity: An intra-axonal study. Exp Brain Res
Cytochem 1969,17:512-516. 1982;47:151-153.
142. Hirano A, Dembitzer HM: Morphology of normal central myelinated axons. In 176. Kocsis JD, Ruiz JA, Waxman SG: Maturation of mammalian myelinated fibers:
Waxman SG (ed): Physiology and Pathobiology of Axons. New York, Raven Changes in action potential characteristics following 4-aminopyridine applica-
Press, 1978, pp 65-82. tion. J Neurophysiol 1983;50:449-463.
143. Hirayama K, Tomonaga M, Kitano K, et al: Focal cervical poliopathy causing 177. Kocsis JD, Preston RJ, Targ EF: Retrograde impulse activity and horseradish
juvenile muscular atrophy of distal upper extremity: A pathological study. J peroxidase tracing of nerve fibers entering neuroma studied in vitro. Exp
Neurol Neurosurg Psychiatry 1987; 50:285-290. Neurol 1984;85:400-412.
144. Hiscoe HB: Distribution of node and incisures in normal and degenerated nerve 178. Kocsis JD, Gordon TR, Waxman SG: Mammalian optic nerve fibers display two
fibers. Anat Rec 1947;99:447-475. pharmacologically distinct potassium channels. Brain Res 1986;393:357-361.
145. Hodes R, Larrabee MG, German W: The human electromyogram in response to 179. Kocsis JD, Eng DL, Gordon TR. et al: Functional differences between 4-
nerve stimulation and the conduction velocity of motor axons. Arch Neurol aminopyridine and tetraethylammonium-sensitive potassium channels in myeli-
Psychiatry 1948;60:340-365. nated axons. Neurosci Lett 1987;75:193-198.
146. Hodgkin AL, Katz B: The effect of temperature on the electrical activity of the 180. Koles ZJ, Rasminsky M: A computer simulation of conduction in demyelinated
giant axon of the squid. J Physiol 1949;109:240-249. nerve. J Physiol 1972;227:351-364.
147. Hoffman HJ: Local reinnervation in partially denervated muscle. 181. Kreutzberg GW, Schubert P: Volume changes in the axon during regeneration.
Histophysiological sudy. Aust J Exp Biol Med Sci 1950;28:383-397. Acta Neuropath 1971; 17:220-226.
148. Holmes G: On morphological changes in exhausted ganglion cells. Z Allg 182. Krinke G, Schaumberg HH, Spencer PS: Pyridoxine megaviiaminosis produces
Physiol 1903;2:502. degeneration of peripheral sensory neurons (sensory neuronopathy) in the dog.
149. Holmes W, Young JZ: Nerye regeneration after immediate and delayed suture. J Neurotoxic 1980;2:13.
Anal 1942;77:63-96. 183. Kristol C, Sandri C, Akert K: Intramembranous particles at nodes of Ranvier of
150. Hongell A, Mattsson HS: Neurographic studies before, after, and during opera- the cat spinal cord: A morphometric study. Brain Res 1978;142:391-400.
tion for median nerve compression in the carpal tunnel. Scand J Plast Reconstr 184. Krucke W: Erkrankungen des peripheren nervensystems. In Henke F, Lubarsch
Surg 1971;5:103-109. O (eds): Handbuch der speziellen Pathologischen Anatomie und Histologic
151. Hopkins AP, Gilliatt RW: Motor and sensory nerve conduction velocity in the (Vol. XIII). Berlin, Springer, 1955, pp 1-248.
baboon. Normal values and changes during acrylamide neuropathy. J Neurol 185. Lampert P: Electron microscopic studies on ordinary and hyperacute experi-
Neurosurg Psychiatry 1971;34:415-426. mental allergic encephalomyelitis. Acta Neuropath 1967;9:99-126.
186. Landau WM: The duration of neuromuscular function after nerve section in 222. Olney RK, Miller RG: Pseudo-conduction block in normal nerves. Muscle
man. J Nerurosurg 1953;10:64-68. Nerve 1983;6:530.
187. Le Quesne PM, Casey EB: Recovery of conduction velocity distal to a com- 223. Olney RK, Miller RG: Conduction block in compression neuropathy: Recogni-
pressive lesion. J Neurol Neurosurg Psychiatry 1974;37:1346-1351. tion and quantification. Muscle Nerve 1984;7:662-667.
188. Lee JC: Electron microscopy of Wallerian degeneration. J Comp Neurol 224. Olney RK, Budingen HJ, Miller RG: The effect of temporal dispersion on com-
1963;120:65-79. pound action potential area in human peripheral nerve. Muscle Nerve
189. Levinson D, Rosenbluth J: Electrophysiologic changes accompanying 1987;10:728-733.
Wallerian degeneration in frog sciatic nerve. Brain Res 1990;523:230-236. 225. Olsson Y: Studies on vascular permeability in peripheral nerves. I. Distribution
190. Lewis T, Pickering GW, Rothschild P: Centripetal paralysis arising out of arrested of circulating fluorescent serum albumin in normal, crushed, and sectioned rat
bloodflow to the limb, including notes on a form of tingling. Heart 1931; 1:1 - 3 1 . sciatic nerves. Acta Neuropath 1966;7:1-15.
191. Lissak K, Dempsey EW, Rosenblueth A: The failure of transmission of motor 226. Olsson Y, Sjostrand J: Origin of macrophages in Wallerian degeneration of periph-
nerve impulses in the course of Wallerian degeneration. Am J Physiol eral nerves demonstrated autoradiographically. Exp Neurol 1969;23:102-112.
1939;128:45-56. 227. Olsson Y, Kristensson K: The perineurium as a diffusion barrier to protein trac-
192. Logan JE: Protein-bound phosphorus compounds in proximal stumps of periph- ers following trauma to nerves. Acta Neuropath 1973;23:105-111.
eral nerve after nerve section or after nerve crush. Can J Med Sci 1952,30: 228. Paintal AS: Effects of temperature on conduction in single vagal and saphenous
457^62. myelinated nerve fibres of the cat. J Physiol 1965;180:20-49.
193. Lubinska L: Demyelination and remyelination in the proximal parts of regener- 229. Pannese E: Number and structure of perisomatic satellite cells of spinal ganglia
ated nerve fibers. J Comp Neurol 1961 ;117:275-289. under normal conditions or during axon regeneration and neuronal hypertrophy.
194. Lubinska L: Axoplasmic streaming in regeneration and in normal nerve fibres. Z Zellforsch Mikrosk Anat 1964;63:568-592.
In Singer M, Schade JP (eds): Mechanisms of Neural Regeneration. Progress in 230. Parker GH: The progressive degeneration of frog nerve. Am J Physiol
Brain Research (Vol. 13). Amsterdam, Elsevier, 1964, pp 1-71. 1933;106:398-403.
195. Lundborg G: Ischaemic nerve injury. Scand J Plast Reconstr Surg 1970;(Suppl 231. Parry GJ, Bredesen DE: Sensory neuropathy with low-dose pyridoxine.
6): 1-113. Neurololgy 1985:35:1466-1468
196. Lundborg G: Nerve Injury and Repair. Edingurgh, Churchill Livingstone, 1988. 232. Pease WS, Lee HH, Johnson EW: Forearm median nerve conduction velocity in
197. Lunn ER, Brown MC, Perry VH: The pattern of axonal degeneration in the pe- carpal tunnel syndrome. Electromyogr Clin Neurophysiol 1990;30:299-302.
ripheral nervous system varies with different types of lesion. Neuroscience 233. Perroncito A: Die Regeneration der Nerven. Beitr Path Anat 1907;42:354-446.
1990;35:157-165. 234. Pilling JB: Nerve conduction during Wallerian degeneration in man. Muscle
198. Marotte LR: An electron microscope study of chronic median nerve compres- Nerve 1978:1:81.
sion in the guinea-pig. Acta Neuropath 1974;27:69-82. 235. Pleasure DE, Feldman B, Prockop DJ: Diptheria toxin inhibits the synthesis of
199. Matthews MR, Nelson VH: Detachment of structurally intact nerve endings myelin proteolipid and basic proteins by peripheral nerve in vitro. J Neurochem
from chromatolytic neurones of rat superior cervical ganglion during the de- 1973;20:81-90.
pression of synaptic transmission induced by post-ganglionic axotomy. J 236. Prineas J: The pathogenesis of dying-back polyneuropathies. I. An ultrastruc-
Physiol 1975;245:91-135. tural study of experimental triortho-crysyl phosphate intoxication in the cat. J
200. Mayer RF, Denny-Brown D: Conduction velocity in peripheral nerve during ex- Neuropath Exp Neurol 1969;28:571-597.
perimental demyelination in the cat. Neurology 1964;14:714-726. 237. Prineas J: The pathogenesis of dying-back polyneuropathies. II. An ultrastruc-
201. McDonald WI: Conduction in muscle afferent fibres during experimental de- tural study of experimental acrylamide intoxication in the cat. J Neuropath Exp
myelination in cat nerve. Acta Neuropath 1962;1:425-432. Neurol 1969;28:598-621.
202. McDonald WI: The effects of experimental demyelination on conduction in pe- 238. Quick DC, Waxman SG: Specific staining of the axon membrane at nodes of
ripheral nerve: A histological and electrophysiological study. II. Electro- Ranvier with ferric ion and ferrocyanide. J Neurol Sci 1977;31:1-11.
physiological observation. Brain 1963;86:501-524. 239. Quilliam TA: A quantitative study of remyelination. Fed Proc 1955; 14:117.
203. McDonald WI, Sears TA: The effects of experimental demyelination on con- 240. Quilliam TA: Growth changes in sensory nerve fiber aggregates undergoing re-
duction in the central nervous system. Brain 1970;93:583-598. myelination. J Anat 1958;92:383-398.
204. Mellick RS, Cavanagh JB: Changes in blood vessel permeability during degen- 241. Qureshi Al, Wilmot G, Dihenia B, et al: Motor neuron disease with parkinson-
eration and regeneration in peripheral nerves. Brain 1968;91:141-160. ism. Arch Neurol 1996;53:987-91.
205. Miledi R, Slater CR: On the degeneration of rat neuromuscular junctions after 242. Ramon Y, Cajal S: Degeneration and Regeneration of the Nervous System.
nerve section. J Physiol 1970;207:507-528. London, Oxford University Press, 1928.
206. Miller RG, Olney RK: Persistent conduction block in compression neuropathy. 243. Ramon Y, Cajal S: Degeneration and Regeneration of the Nervous System. New
Muscle Nerve 1982;5:S154-S156. York, Hafner, 1959.
207. Molinuevo JL, Graus F, Serrano C, et al: Utility of anti-Hu antibodies in the di- 244. Ranvier L: Lecons sur L'histologic du systeme nerveux. Paris, 1878.
agnosis of paraneoplastic sensory neuropathy. Ann Neurol 1998;44:976-80. 245. Rapos M, Bakos J: Veranderungen der Neuroglia im Nucleus originis Nervi hy-
208. Moore JW, Joyner RW, Brill MH, et al: Simulations of conduction in myeli- poglossi nach mechanischer Beschadigung des Nervus hypoglossus. Z Mikrosk
nated fibers: Relative sensitivity to changes in nodal and internodal parameters. Anat Forsch 1959,65:211-218.
Biophys J 1978;21:147-160. 246. Rasminsky M, Sears TA: Internodal conduction in undissected demyelinated
209. Nathaniel EJH, Pease DC: Collagen and basement membrane formation by nerve fibres. J Physiol 1972;227:323-350.
Schwann cells during nerve regeneration. J Ultrastruct Res 1963;9:550-560. 247. Rasminsky M: The effects of temperature on conduction in demyelinated single
210. Neary D, Ochoa J, Gilliatt RW: Subclinical entrapment neuropathy in man. J nerve fibers. Arch Neurol 1973;28:287-292.
Neurol Sci 1975;24:283-298. 248. Rasminsky M: Pathophysiology of demyelination. In Didactic Program.
211. Neary D, Eames RA: The pathology of ulnar nerve compression in man. AAEM, 1980, Rochester, MN, 1980, pp 29-34.
Neuropath Appl Neurobiol 1975; I ;69-88. 249. Rasminsky M: Ephaptic transmission between single nerve fibers in spinal
212. Neumcke B, Stampfli R: Sodium currents and sodium-current fluctuations in nerve roots of dystrophic mice. J Physiol 1980;305:151-169.
rat myelinated nerve fibers. J Physiol 1982;329:163-184. 250. Rexed B, Swensson A: Uber die Regeneration der Markscheide bei peripheren
213. Nielsen VK, Kardel T: Decremental conduction in normal human nerves sub- Nerven nach Kontinuitatstrennungen. Z Mikrosk Anat Forsch 1941;49:359-387.
jected to ischemia. Acta Physiol Scand 1974;92:249-262. 251. Rhee ED, England JD, Sumner AJ: A computer simulation of conduction block:
214. Nissl F: Uber die Veranderungen der Ganglionzellen am Facialiskern des Kan- Effects produced by actual block versus interphase cancellation. Ann Neurol
inchens nach Ausreizung der Nerven. Z Psychiat 1892;48:197-198. 1990;28:146-156.
215. Novak J, Salafsky B: Early electrophysiological changes after denervation of 252. Ritchie JM, Rogart RB, Strichartz G: A new method for labelling saxitoxin and
slow skeletal muscle. Exp Neurol 1967;19:388-400. its binding to nonmyelinated fibers of the rabbit vagus, lobster walking leg, and
216. O'Daly JA, Imaeda T: Electron microscopic study of Wallerian degeneration in garfish olfactory nerve. J Physiol 1976;261:477-494.
cutaneous nerves caused by mechanical injury. Lab Invest 1967;17:744-766. 253. Ritchie JM, Rogart RB: The density of sodium channels in mammalian myeli-
217. Ochoa J, Danta G, Fowler TJ, et al: Nature of the nerve lesion caused by a pneu- nated nerve fibers and the nature of the axonal membrane under the myelin
matic tourniquet. Nature 1971;233:265-266. sheath. Proc Natl Acad Sci USA 1977;74:211-215.
218. Ochoa J, Fowler TJ, Gilliatt RW: Anatomical changes in peripheral nerves com- 254. Ritchie JM, Chie SY: Distribution of sodium and potassium channels in mam-
pressed by a pneumatic tourniquet. J Anat 1972; 113:433-455. malian myelinated nerve. In Waxman SG, Ritchie JM (eds): Demyelinating
219. Ochoa J, Fowler TJ, Gilliat RW: Changes produced by a pneumatic tourniquet. Diseases: Basic and Clinical Electrophysiology. New York, Raven Press, 1981,
In Desmedt JE (ed): New Developments in Electromyography and Clinical pp 329-342.
Neurophysiology. Vol 2. Basel, Karger, 1973, pp 174-180. 255. Rowland LP: Diagnosis of amyotrophic lateral sclerosis. J Neurol Sci
220. Ochoa J, Marotte LR: The nature of the nerve* lesion caused by chronic entrap- 1998;160 SuppI 1 ):S6-24.
a
ment in the guinea-pig. J Neurol Sci 1973;19:491-495. 256. Romanes GJ: Motor localization and the effects of nerve injury on the ventral
221. Ochoa J: Nerve fiber pathology in acute and chronic compression. In Omer GE, horn of the spinal cord. J Anat 1946;80:117-131.
Spinner M (ed): Management of Peripheral Nerve Problems. Philadelphia, W.B. 257. Rosenbluth A, Dempsey EW: A study of Wallerian degeneration. Am J Physiol
Saunders, 1980, pp 487-501. 1939;128:19-30.
258. Rosenbluth JD: Intramembranous particle distribution at the node of Ranvier 295. Vizoso AD, Young JZ: Internode length and fibre diameter in developing and
and adjacent axolemma in myelinated axons of the frog brain. J Neurocytol regenerating nerves. J Anat 1948;82:110-134.
1976;5:731-745. 296. Waksman BH, Adams RD: A comparative study of experimental allergic neuri-
259. Rudge P: Tourniquet paralysis with prolonged conduction block. An electro- tis in rabbit, guinea pig, and mouse. J Neuropath Exp Neurol 1956;15:293-333.
physiological study. J Bone Joint Surg 1974;56B:716-720. 297. Wall PD, Waxman SG, Basbaum Al: Ongoing activity in peripheral nerve:
260. Rudge P, Ochoa J, Gilliatt RW: Acute peripheral nerve compression in baboon. Injury discharge. Exp Neurol 1974;45:576-589.
J Neurol Sci 1974;23:403-420. 298. Waller A: Experiments on the section of the glossopharyngeal and hypoglossal
261. Salafsky B, Jasinsky D: Early electrophysiological changes after denervation of nerves of the frog, and observations of the alterations produced thereby in the
fast skeletal muscle. Exp Neurol 1967;19:375-387. structure of their primitive fibres. Phil Trans B 1850; 140:423^29.
262. Sanders FK, Whitteridge D: Conduction velocity and myelin thickness in re- 299. Waller A: Sur la reproduction des nerf et sur la structure et les fonctions des
generating nerve fibres. J Physiol 1946;105:152-174. ganglions spineaux. Arch Anat Physiol 1852:392-401.
263. Sanders FK: The thickness of the myelin sheaths of normal and regenerating 300. Waxman SG: Conduction in myelinated, unmyelinated, and demyelinated
peripheral nerve fibers. Proc R Soc (Biol) 1948;135:323-357. fibers. Arch Neurol 1977;34:585-589.
264. Sasybin N: Uber die Regeneration der Nervenfasern im mehrschichtigen Plat- 301. Waxman SG, Brill MH: Conduction through demyelinated plaques in multiple
tenepithel. Z Mikrosk Anat Forsch 1930;22:1-72. sclerosis: Computer simulations of facilitation by short internodes. J Neurol
265. Schauf CL, Davis FA: Impulse conduction in multiple sclerosis: A theoretical Neurosur Psychiatry 1978;41:408^ 16.
basis for modification by temperature and pharmacologic agents. J Neurol 302. Waxman SG: Prerequisites for conduction in demyelinated fibers. Neurology
Neurosurg Psychiatry 1974;37:152-161. 1978,28:27-33.
266. Schnapp B, Mugnaini E: Membrane architecture of myelinated fibers as seen 303. Waxman SG: The flexion-adduction sign in neuralgic amyotrophy. Neurology
by freeze-fracture. In Waxman SG (ed): Physiology and Pathobiology of Axons. 1979;29:1301-1304.
New York, Raven Press, 1978, pp 83-123. 304. Waxman SG, Foster RE: Ionic channel distribution and heterogeneity of the
267. Schroder JM: Altered ratio between axon diameter and myelin sheath thickness axon membrane in myelinated fibers. Brain Res Rev 1980;2:205-240.
in regenerated nerve fibers. Brain Res 1972;45:49-65. 305. Waxman SG: Determinants of conduction velocity in myelinated nerve fibers.
268. Sears TA, Bostock H: Conduction failure in demyelination: Is it inevitable? In Muscle Nerve 1980;3:141-150.
Waxman SG, Ritchie JM (eds): Demyelinating Diseases: Basic and Clinical 306. Waxman SG: Membranes, myelin, and the pathophysiology of multiple sclero-
Physiology. Raven Press, New York, 1981, pp 357-375. sis. New Engl J Med 1982;306:1529-1533.
269. Seddon H: Three types of nerve injury. Brain 1943;66:237-288. 307. Waxman SG: Clinical course and electrophysiology of multiple sclerosis, in
270. Seddon H: Surgical disorders of the peripheral nerves. London, Churchill Waxman SG (ed): Functional Recovery in Neurological Diseases. New York,
Livingstone, 1972. Raven Press, 1988, pp 175-184.
271. Seneviratne KN, Peiris OA: The effect of ischemia on the excitability of human 308. Waxman SG: Molecular organization and pathophysiology of axons. In Asbury
sensory nerve. J Neurol Neurosurg Psychiatry 1968;31:338-347. AK, McKhann GM, McDonald WI: Diseases of the Nervous System: Clinical
272. Shawe GDH: On the number of branches formed by regenerating nerve fibers. Neurobiology. Philadelphia, W.B. Saunders, 1992, pp 25-46.
Br J Surg 1955;42:474-488. 309. Webster H, De F, Spiro D, et al: Phase and electron microscopic studies of exper-
273. Smith KJ, Hall SM: Nerve conduction during peripheral demyelination and re- imental demyelination. II. Schwann cell changes in guinea pig sciatic nerves
myelination. J Neurol Sci 1980;48:201-219. during experimental diphtheritic neuritis. J Neuropath Exp Neurol 1961;20:5-34.
274. Smith KJ, McDonald W I : Spontaneous and mechanically evoked activity due to 310. Webster H, De F: Transient focal accumulation of axonal mitochondria during
a central demyelinating lesion. Nature 1980;280:154-155. the early stages of Wallerian degeneration. J Cell Biol 1962;12:361-383.
275. Soskind L, Anderson FD, Aleu F, et al: Peripheral nerve regeneration: 311. Webster H, De F: Some ultrastructural features of segmental demyelination and
Accelation of rate through administration of sterol substances. Anat Rec myelin regeneration in peripheral nerve. In Singer M, Schade JP (eds):
1966; 154:426. Mechanisms of Neural Regeneration. Progress in Brain Research, Vol. 13.
276. Speidel CC: Studies of living nerves: Growth adjustments of cutaneous termi- Amsterdam, Elsevier, 1964, pp 151-174.
nal arborization. J Comp Neurol 1942;76:57-73. 312. Webster H, De F: The relationship between Schmidt-Lantermann incisures and
277. Stefoski D, Davis FA, Faut M, et al: 4-aminopyridine improves clinical signs in myelin segmentation during Wallerian degeneration. Ann N Y Acad Sci
multiple sclerosis. Ann Neurol 1987;21:71-77. 1965;122:29-38.
278. Sunderland S: Traumatic injuries of peripheral nerves. I. Simple compression 313. Weddell G, Glees P: The early stages in the degeneration of cutaneous nerve
injuries of the radial nerve. Brain 1945;68:56-72. fibers. J Anat 1941;76:65-93.
279. Sunderland S, Bradley KC: Endoneural tube shrinkage in distal segment of sev- 314. Weddel G: Axonal regeneration in cutaneous nerve plexuses. J Anat
ered nerve. J Comp Neurol 1950;93:411-420. 1942;77:49-62.
280. Sunderland S: A classification of peripheral nerve injuries producing loss of 315. Weddell G, Sinclair DC: "Pins and needles": Observations on some of the sen-
function. Brain 1951;74:491-516. sations aroused in a limb by the application of pressure. J Neurol Neurosurg
281. Sunderland S: Nerves and Nerve Injuries, 2nd ed. Edinburgh, Churchill Psychiatry 1947;10:26-48.
Livingstone, 1978. 316. Weller RO: Diphtheritic neuropathy in the chicken: An electron-microscope
282. Sunderland S: Nerve Injuries and Their Repair: A Critical Appraisal. study. J Path Bact 1965;89:591-598.
Edinburgh, Churchill Livingstone, 1991. 317. Wilbourn AJ: Serial conduction studies in human nerve during Wallerian de-
283. Tasaki I, Fujita M: Action currents of single nerve fibers as modified by temper- generation. Electroencephalogr Clin Neurophysiol 1977,43:616.
ature changes. J Neurophysiol 1948;11:311-315. 318. Wilbourn AJ: Nerve conduction study changes in human nerves undergoing
284. Tasaki I: New measurement of the capacitance and resistance of the myelin Wallerian degeneration. Neurology 1981;31:96-97.
sheath and the nodal membrane of the isolated frog nerve. Am J Physiol 319. Wilbourn AJ: AAEM Case Report #12: Common peroneal mononeuropathy at
1955;181:639-650. the fibular head. American Association for Electrodiagnostic Medicine.
285. Tasaki I: Conduction of the nerve impulse. In Field J, Magoun H, Hall VE (ed): Rochester, MN, 1986.
Handbook of Physiology, Section 1, Neurophysiology, Vol. 1. Washington, DC, 320. Williams IR, Gilliatt RW, Jefferson D: Limb ischemia and acute nerve compres-
American Physiological Society, 1959, pp 75-122. sion. Electroencephalogr Clin Neurophysiol 1977;43:592.
286. Thomas PK: Motor nerve conduction in the carpal tunnel syndrome. Neurology 321. Williams IR, Jefferson D, Gilliatt RW: Acute nerve compression during limb is-
1960;10:1045-1050. chemia: An experimental study. J Neurol Sci 1980;46:199-207,
287. Thomas PK, Lascelles RG: Hypertrophic neuropathy. Quart J Med 322. Williams PL, Hall SM: Chronic Wallerian degeneration—an in vivo and ultra-
1967;36:223-238. structural study. J Anat 1971;109:487-503.
288. Thomas PK, Lascelles RG, Hallpike JF, et al: Recurrent and chronic relapsing 323. Williams PL, Hall SM: Prolonged in vivo observations of normal peripheral
Guillian-Barre polyneuritis. Brain 1969;92:589-606. nerve fibers and their acute reactions to crush and deliberate trauma. J Anat
289. Thomas PK: Selective vulnerability of the centrifugal and centripetal axons of 1971; 108:397^08.
primary sensory neurons. Muscle Nerve 1982; 5:S117—S121. 324. Windebank AJ, Blexrud MD. Dyck PJ, et al: The syndrome of acute sensory
290. Trojaborg W: Rate of recovery in motor and sensory fibers of the radial nerve: neuropathy: clinical features and electrophysiologic and pathologic changes.
Clinical and electrophysiological aspects. J Neurol Neurosurg Psychiatry Neurology 1990;40:584-591.
1970;33:625-638. 325. Woltman HW, Wilder RM: Diabetus mellitus: Pathologic changes in the spinal
291. Trojaborg W: Prolonged conduction block with axonal degeneration. J Neurol cord and the peripheral nerves. Arch Intern Med 1929;44:576-603.
Neurosurg Psychiatry 1977;40:50-57. 326. Wright EB: The effects of asphyxiation and narcosis on peripheral nerve polar-
292. Trojaborg W: Early electrophysiological changes in conduction block. Muscle ization and conduction. Am J Physiol 1947;148:174-184.
Nerve 1978;1:400-403. 327. Young JZ: The functional repair of nervous tissue. Physiol Rev 1942;22:318-374.
293. Uchida Y, Sugioka Y: Electrodiagnosis of retrograde changes in carpal tunnel 328. Young JZ: Contraction, turgor and cytoskeleton of nerve fibres. Nature
syndrome. Electromyogr Clin Neurophysiol 1993;33:55-58. 1944;153:333-335.
294. Van Gehuchten A: Le phenomene de chromatolyse consecutif a lesion pathologique 329. Zelena J: Neurofilaments and microtubules in sensory neurons after peripheral
ou experimentale de 1'axone. Bull Acad Roy Med Belg 1897; 11:805-821. nerve section. Z Zellforsch Mikrosk Anat 1971; 117:191 -211.
PART
II
BASIC A N D
ADVANCED
TECHNIQUES
Chapter 5
N e r v e C o n d u c t i o n Studies
Anthony A. Amato, M.D.
CHAPTER OUTLINE
Historical Aspects of Nerve Conduction Common Upper Limb Nerve Conduction Studies
Upper Limb Motor Nerve Conduction Studies
Structure of the Peripheral Nervous System • Median Nerve • Ulnar Nerve • Radial Nerve • Phrenic
Unmyelinated Nerve Fibers • Myelinated Nerve Fibers Nerve
• Axonal Transport Mechanism • Periaxonal Region Upper Limb Sensory Nerve Conduction Studies
• Internodal Schwann Cell/Myelin Sheath • Nodes of Median Nerve • Ulnar Nerve • Dorsal Ulnar Cutaneous
Ranvier • Peripheral Nerve Structure • Action Nerve • Superficial Radial Nerve • Lateral Antebrachial
Potential Generation * Compound Nerve Action Cutaneous Nerve • Medial Antebrachial Cutaneous Nerve
Potential • Comparative NCS of Dual-Innervated Digits
Basic Nerve Conduction Study Principles Mixed Nerve Studies

Patient Consideration • Stimulation • Recording
• Motor Nerve Conduction Studies • Sensory Nerve Common Lower Limb Nerve Conduction Studies
Conduction Studies • Compound Nerve Action Lower Limb Motor Nerve Conduction Studies
Potentials • Femoral Nerve • Sciatic Nerve • Bulbocavernosus
Reflex/Pudendal Nerve • Peroneal Nerve • Tibial Nerve
Factors Affecting Nerve Conduction Studies Lower Limb Sensory Nerve Conduction Studies
Instrumentation Factors • Physiologic Factors • Lateral Femoral Cutaneous Nerve • Posterior Femoral
• Anomalous Innervation Cutaneous Nerve • Saphenous Nerve • Superficial Peroneal
Sensory Nerve • Sural Nerve • Medial/Lateral Plantar
Nerve Conduction Techniques Compound Nerve Action Potentials
One of the fundamental components of the electrodiagnos- HISTORICAL ASPECTS OF NERVE

tic medicine consultation is the assessment of the peripheral
nervous system's ability to conduct an electrical impulse.
CONDUCTION
Measuring the velocity of impulse propagation (nerve con- Although scientists have attempted to measure the velocity of
duction velocity, N C V ) and the ensuing evoked response's neural conduction for quite some time, an initial and novel ap-
magnitude allows one to make inferences about the peripheral proach was attempted in 1762 by Von Haller. His "classic"
343
nervous system's relative health. This information provides approach was to read the Aeneid aloud and tabulate the number
the practitioner with valuable data to assist in the diagnosis of letters pronounced within one minute. Of the 1500 letters per
and prognosis of specific diseases. Prior to performing a minute, the letter " R " supposedly took 10 contractions of the
nerve conduction study (NCS), however, it is important to styloglossus muscle, which was believed to contract and relax
learn the various anatomic and physiologic factors that influ- 15,000 times per minute. Assuming each muscle contraction
ence the accurate collection and interpretation of collected lasted 2 ms combined with a thought impulse requiring 2 ms to
data. It is inappropriate to perform an NCS without first be- cover the 10 cm from brain to muscle, a conduction velocity of
coming thoroughly familiar with peripheral neuromuscular 50 meters/second (m/s) was estimated. Despite multiple un-
anatomy. founded assumptions, this velocity was a rather good guess.
159
160 — P A R T I ! BASIC AND ADVANCED TECHNIQUES
Between 1850 and 1870, Von Helmholtz and associates deter- size of the unmyelinated fiber, which necessitates electron as
mined the median nerve to have a conduction velocity of 61.0 ± opposed to light microscopy and demanding techniques to in-
5.1 m/s by using neural stimulation combined with measuring vestigate various electrophysiologic properties. Electron mi-
the onset of the ensuing muscular contractions. - - Using 345 346 347
croscopy, however, reveals a number of interesting structural
different methods, they also estimated sensory nerve conduction properties of unmyelinated nerve fibers.
to approximate 60 m/s. Major contributions to the determina-
344
The cell body of an unmyelinated nerve fiber is usually lo-
tion of neural impulse propagation were achieved in the 1920s cated at some proximal location within the central nervous
by Erlanger and Gasser - when they stratified conduction ve-
89 90
system or paravertebral ganglia. The peripheral extensions of
locities according to nerve fiber diameter. Further refinements these cell bodies extend to their end organ ensheathed by con-
in both technique and instrumentation allowed motor nerve consecutively aligned satellite cells, i.e., Schwann cells. Unmy-
duction velocities to be applied diagnostically by Hodes et al. in elinated nerve fibers usually consist of collections of multiple
194g 138.139 with the development of instruments by Dawson 60
axons, each surrounded by a single column of Schwann cells. A
that could average multiple responses, sensory nerve conduc- Schwann cell has numerous depressions within which a single
tion studies became practical clinically. Further refinements in peripheral axon is contained (Fig. 5-1). The size of unmyeli-
the stimulating and recording apparatus presently permit exami- nated nerve fibers ranges from 0.5 to about 2.0 jum. Although
201
nation of the peripheral nervous system with good diagnostic a negative descriptor, unmyelinated nerves are axons whose
success. associated satellite Schwann cells do not normally produce an
enveloping multilayered proteophospholipid substance called
STRUCTURE OF THE PERIPHERAL NERVOUS SYSTEM myelin (see below). Developmentally, however, all axons are
initially unmyelinated. Also, there are distinct regions of myeli-
The peripheral nervous system can be divided into two major nated nerves that are always unmyelinated, i.e., nodes of
classes of nerves: myelinated and unmyelinated. In this discus- Ranvier, adjacent to the cell body (axon hillock), and near a
sion, unmyelinated nerves are only considered from the per- motor nerve's termination at the neuromuscular junction.
spective of better understanding of myelinated nerves. The The actual structural relationship between the above noted
rationale for concentrating primarily on myelinated nerves is small axons and the Schwann cell may be imagined if a simple
that the more routine nerve conduction techniques exclusively model is considered. Let us first imagine the Schwann cell as a
assess conduction in the fastest myelinated nerve fibers. Minor two-dimensional object similar to a circle with a malleable
consideration is given to the more slowly conducting autonomic outer membrane boundary. The axon is then pressed against this
nerve fibers and techniques to examine their conduction proper- boundary forming a cleft in the circle surrounding the axon.
ties at the end of this chapter. Continued pressure of the axon results in the intact outer mem-
brane now completely surrounding the axon. At the circle's
Unmyelinated N e r v e Fibers original point of invagination, a double-layered membrane is
The designation ''nerve fiber" is understood to apply to the now formed that gives the appearance of the axon being sus-
axon and its enveloping satellite cells. Information pertaining
329
pended from the edge of the cell's outer boundary. This "sus-
to unmyelinated nerve fibers, particularly regarding biologic pensory" region of cell membrane is known as the mesaxon and
processes, is less available than that for myelinated nerves. The is derived from the term designating the gut's mesentary (Figs.
reason for this relative paucity of information is the rather small 5-1 and 5-2). Additional axons share the same Schwann ceil. In
the living organism, the Schwann cell most likely develops with
the growing axon and envelops the multiple axons it is destined
to support.
Myelinated N e r v e Fibers
We shall consider the mammalian somatic mye ited pe-
ripheral nerve to consist of two major subgroups: (1) d periph-
eral extension of the proximally located cell body, i.e., axon,
and (2) the satellite or Schwann cells. A myelinated nerve may j
be described generally as an axon with a surrounding sheath of
myelin. This cylindrical myelin tube is interrupted at regular in-
tervals exposing the axon at regions designated nodes of
Ranvier (Fig. 5-2A). The surrounding myelin first appears
within a short distance from the cell body and is no longer pre-
sent at 1 or 2 Jim from the nerve's termination. That region of
myelin between two adjacent nodes of Ranvier is called an in-
ternode and arises from a single Schwann cell.
Figure 5-1. U n m y e l i n a t e d n e r v e . Transverse section through a Axon

primate peripheral nerve demonstrating a number of unmyelinated The axon is the portion of the nerve fiber that is the direct pe-
axons (a) contained within the cytoplasm of a Schwann Cell(s).The ripheral extension of the proximally located cell body. The
mesaxons are denoted by arrow heads. Scale = I urn. (From Landon defining boundary of the axon is its surface membrane that is
DN: Structure of normal perpheral myelinated nerve fibres. In referred to as the axolemma. The axolemma is a trilaminar
Waxman SG, Ritchie JM (eds): Demyelinating Diseases: Basic and membrane approximately 8 nanometers (nm) thick. Contained
Clinical Electrophysiology. New York, Raven Press, 1981, pp 25—49, within the axolemma is the axon's cytoplasm known as axo-
with permission.) plasm. Axoplasm is an amorphous and electron-translucent
Chapter 5 NERVE C O N D U C T I O N STUDIES — 161
(does not impede an electron microscope's beam of electrons)

substance containing a number of organelles and filamentous
structures. Electron spin resonance studies have demonstrated
that the viscosity of mammalian peripheral nerve axoplasm is
approximately five times that of water. 124
Filamentous Structures. There are three types of filamen-

tous material identified in mammalian myelinated nerves: mi-
crofilaments, neurofilaments, and microtubules (neurotubules).
Microfilaments are both the smallest type of filament as well
as the smallest individually distinguishable axoplasmic struc-
ture. These microfilaments measure 5-7 nm in diameter, are
irregularly distributed, consist of actin, and comprise approxi-
mately 10% of the axon's total protein content.' The miero- 52 2 0 1
i filaments are relatively sparse in the internodal regions and are

primarily located beneath the axolemma in newly formed
growth cones. Microfilaments are believed to be difficult to
195 3 6 6
observe because of the fragility of F-actin and chemical fixa-

tives used in cell preparation techniques. - 152 227
A second category of filaments located within myelinated

nerve axons are neurofilaments. They are longitudinally ori-
ented along the axon, are of indefinite length, possess a diame-
ter of approximately 10 nm, and have a packing density ranging
from 100 to 300/|im . - ' - - Neurofilaments are thought to
2 3 200 300 365 368
be interconnected by a reticular framework extending at 50-nm

intervals in a radial spoke pattern; however, this retinaculum has
been suggested to be a fixation artifact. - Unlike micro- 1 0 1 1 m 3 2 9
filaments, the neurofilaments are not composed of actin but are

comprised of parallel rows of 4 - 5 chains of a globular protein
called neurofilamin. The functional contribution neurofila-
199
ments make to the axon is unclear, but axonal transport proper-

ties have been implicated. 329
Microtubules (neurotubules) are the third type of filaments

found in myelinated nerves. These structures are hollow cylin-
ders approximately 25 nm in diameter with a 5-nm thick wall
occasionally containing 5-nm granules extending for an indeter-
minate length.2oo.329.37o Subcomponent rings of 4-nm monomer Figure 5-2. M y e l i n a t e d n e r v e . A, Longitudinal phase contrast mi-
molecules forming a triple helix of 13 globular subunits consti- crograph of a mouse sciatic nerve. Nodes of Ranvier are designated by
tute the microtubules' structure. Radiating extensions approxi- the arrows. B, Large and small myelinated nerves are noted in this phase
mately 8 nm in diameter and 20-100 nm in length project from contrast micrograph of a transverse section through a mouse sciatic
the microtubules. The packing density of microtubules varies
365
nerve. C, Electron micrograph of a transverse section through a mouse

between 10 and 20 per p m . Microtubules appear to be pref- 2 101
sciatic nerve. An o u t e r (om) and inner (im) mesaxon is noted. Major

erentially located where the density of neurofilaments is re- dense lines (ml) of the myelin arise from the fusion of the apposed inner
duced. Additionally, microtubules are observed to be closely surfaces of the Schwann cell surface membrane (Sm).The axon (ax) is
associated with axonal organelles such as smooth endoplasmic surrounded by a layer of adaxonal Schwann cell cytoplasm (aSc). (From
reticulum and mitrochondria with some evidence for actual con- Thomas PK, Ochoa J: Microscopic anatomy of peripheral nerve fibers. In
nections through the above noted radiating structures. - 266 300
Dyck PJ,Thomas PK, Lambert EH (eds): Peripheral Neuropathies, 2nd ed.

Alterations in the surrounding environment such as a decrease Philadelphia.W.B. Saunders, 1985, pp 39-96, with permission.)
in temperature to less than 4°C, an increase in the hydrostatic
pressure, exposing the axon to colchicine or vinblastine causes
the microtubules to break down into their subcomponent globu- 0. 1/p.m are seen in relatively larger axons. Time lapse photog-
2 10
lar proteins. The exact role of microtubules is not fully eluci-

311
raphy reveals that mitochondria are rather mobile and travel
dated, but they are believed to play some role in axonal proximally and distally within the axon. - - - Investigators
50 98 99 368
transport. have suggested that the outer wall of the mitochondria may be
Mitochondria. Mitochondria located within the axoplasm continuous with that of the smooth endoplasmic reticulum that
are mophologically different than those in other cells and are transports the inner constituents of the mitochondria (matrix,
believed to be formed in the cell body and then transported to cristae, and inner membrane) within the axon. 305
the axon's periphery. They are relatively long (10 pm) and
286
Smooth Endoplasmic Reticulum/Vesicles. A continuous
thin (0.1-0.3 pm) and contain "concentric funnel-shaped network of tubular structures extending from the cell body to
cristae." The largest aggregation of mitochondria appear to be in the terminal portions of the nerve constitutes the axon's
the rather narrowed nodal and paranodal regions of the smooth endoplasmic reticulum. Primary and secondary com-
76
axon. - Additionally, the abundance of mitochondria is in-

11 199
ponents of the sarcoplasmic reticulum have been identified.
versely proportional to the axon's diameter. Mitochondrial in- The primary portions are relatively thicker tubular structures
ternodal densities of 2-5/p.m are noted in small axons, while 2
aggregated near the axolemma, while the secondary system
162 — PART II BASIC AND ADVANCED TECHNIQUES
consists of smaller and more centrally located tubes. The sec- extend along the entire internode. - - Schmidt-Lanterman
237 283 288
ondary aspects of the sarcoplasmic reticulum have been pro- incisures (see below) are also isolated from the periaxonal
posed to be the source of synaptic vesicles. Ribosomes can be region.
detected in the more distal regions of the axoplasm and at
times associated with the endoplasmic reticular structures. 329
Internodal Schwann Cell/Myelin Sheath
The sarcoplasmic reticulum is believed to be one of the axon's It is suggested that there are a fixed number of Schwann cells
transport mechanisms. - 76 77
associated with each myelinated nerve fiber at the beginning of
A number of vesicular structures are located within the axo- development such that axonal growth is accompanied by elon-
plasm. Chain-like composites of clear vesicles are often ob- gation of the Schwann cell and its cytoplasmic territory. - 287 342
served near the nodes of Ranvier. Lysosomes are typically 11199

The territory of individual Schwann cells defines the internode
identified along the length of most axons. - Near the nodes 145 300
length and ranges from 200 pm to 2,000 pm with a direct corre-
of Ranvier, endocytotic "coated" vesicles are found and have lation between axon diameter and an internode's spatial ex-
been postulated to constitute a mechanism whereby extracellu- panse. - -
399 327
Although the internodal length can vary
359
lar substances are incorporated into the axon and subsequently substantially from one axon to the next, the internodal distance
distributed through the smooth endoplasmic reticulum's trans- for any one nerve is relatively constant.
port system. - 249 295
Schwann Cell. The Schwann cell may be best considered if
it is thought of as unfurled from its axon similar to the unrolling
Axonal Transport Mechanism of a flag (Fig. 5-3). The flat trapezoidal Schwann cell can now
An obvious observation regarding axons is that their cell be appreciated with respect to its constituent parts. That region
bodies give rise to a peripheral extension capable of reaching of the cell's cytoplasm comprising the first wrapping around the
more than one meter in length with an accompanying axoplas- axon (adaxonal) is referred to as the inner belt, while the su-
mic mass significantly exceeding that of the cell body. How perficial aspect of the cell (abaxonal) containing the nucleus
does the cell body sustain the metabolic requirements of its pe- and nodal villi is known as the outer belt. A less dense or semi-
ripheral process? This relatively simple question has resulted in compact myelin is found in both of these areas. The compact
a number of transport mechanisms being postulated that carry myelin is located between the adaxonal and abaxonal regions. 238
various cellular materials in both directions. - The or- 208 249

Connecting the inner and outer belts is a cytoplasmic pathway
thograde or somatofugal (away from the cell body) transport known as the Schmidt-Lanterman incisure.
systems may function at two and possibly five different rates In the adaxonal region one can observe that the Schwann cell
(0.25-3 to 400 mm/day) with one axonal protein being transmembrane and cytoplasm form a spiral of about one turn with
ported per system. - - - Of note, the distally directed slow
119 356 357 360
opposing Schwann cell plasma membrane fusing with the aid of
transport of axoplasmic mass movement approximating 0.25-3 tight junctions to form the inner mesaxon (Fig. 5-2C). The
mm/day is about the rate of axonal regrowth following nerve Schwann cell cytoplasm in the adaxonal area contains a number
transection. Generally, the retrograde or somatopetal (toward
359
of membrane-associated particles of indeterminate function and
the cell body) axonal transport flow rates are roughly one-half no major cellular organelles. A similar formation of Schwann
of the orthograde flow rates. About 7 5 % of the dry weight of the cell cytoplasm is formed at the abaxonal region and contains a
transported material is accounted for in the form of tubulin and limited quantity of cytoplasm except for the perinuclear space.
neurofilamentous proteins traveling at the slowest rate of 0.25 As with the adaxonal spiral, an abaxonal cytoplasmic spiral
mm/day. - - Materials consisting of actin microfilaments
16 143 208
exists that yields an outer mesaxon. Except for the perinuclear
with associated proteins and soluble enzymes traveling at 2-3 region, the outer plasmalemma of the Schwann cell is in close
mm/day make up the remainder of the transported substances. 16
approximation to the myelin sheath. Most of the Schwann cell's
Faster axonal transport rates require the expenditure of energy organelles, rough endoplasmic reticulum, Golgi membranes,
and are temperature-sensitive. The substances transported
99100
and mitochondria are gathered about its nucleus. Finger-like ex-
at the relatively faster rates are glycoproteins, membranous sub- tensions of the Schwann cell's membrane project into the node
components, and enzymes associated with the manufacture of of Ranvier region and are referred to as microvilli. Exterior to
transmitter substances. The actual mechanisms responsible for the Schwann cell's plasmalemma is a basal lamina investing
the faster rates of transport are unknown, but two postulates the entire nerve including the nodes of Ranvier. 289
have been described. One of the transport systems is referred to Myelin Sheath. Close inspection through electron mi-
as axoplasm streaming, which is thought to be a result of microscopy of the axon's myelin sheath reveals a continuous spiral
crotubular actin-myosin interacting with the transported sub- of Schwann cell plasma membrane lamellae (Fig. 5-2C). These
stance through some type of interposed carrier system. - 249 276
lamellae are alternating layers of lipid and protein arising from

Support for this manner of transport is that multiple axoplasmic the Schwann cell "wrapping" its plasma membrane multiple
organelles are frequently noted to be in close association with times around the axon during the process of myelination. 111276
microtubules and at times there appears to be a linkage between Myelin functions as an insulator for the axon allowing depolar-
mitochondria and microtubules. The second type of fast axonal ization only at the regions devoid of myelin, i.e., nodes of
transport may occur within the tubular confines of the smooth Ranvier (see Chapter 1). The presence of myelin, therefore,
endoplasmic reticulum. 77
causes the nerve to become depolarized at nodes of Ranvier, re-

sulting in the action potential "jumping" from one node to the
Periaxonal Region next in the process of saltatory conduction. The velocity of
The periaxonal region is the space between the axolemma nerve conduction is directly related to the axon's diameter, in-
and extracellular aspect of the Schwann cell's plasmalemma ap- ternode length, and efficiency of myelin insulation. An optimal
proximating 20 n m . This portion of the nerve is a self-con-
329
ratio of axon diameter to total fiber diameter has been postu-
tained space and completely sealed from the Schwann cell lated to be 0.6. - This theoretical ratio is often, but not
142 281
through mesaxon tight junctions (zona occludens) which always, found in nature.
Figure 5-3. S c h w a n n cell. An unfurled Schwann cell reveal-

ing its internal structures.The cytoplasm is represented by the
white areas whereas t h e stippled areas designate semi-com-
pact myelin along the inner and o u t e r portions of t h e cell. A
complete and incomplete Schmidt-Lanterman (SL) incisure in
addition t o a longitudinal incisure is included. N o t e t h e mi-
crovilli at either end of t h e Schwann cell's o u t e r belt which
project into the node of Ranvier. (From Mugnaini E, Osen KK,
Schnapp B, et al: Distribution of Schwann cell cytoplasm and
plasmalemmal vesicles (caveolae) in peripheral myelin sheaths:
An electron microscopic study with thin sections and freeze-
fracturing.J Neurocytol 1977;6:647-668, with permission.)
Peripheral nervous system (PNS) myelin is composed of of ions, nutrients, and other small molecules through the com-
lipids (e.g., sphingomyelin) and glycoproteins. The predomi- pact myelin to the innermost layers of the myelin sheath and the
nant protein is myelin protein zero, P , which accounts for ap-
0 axon itself. Connexin-32 also appears to be necessary for
proximately 50% of the total myelin protein. P is a 28-kD 0 normal axonal structure and function. Mutations in the gene en-
transmembrane protein that has an extracellular immunoglobu- coding for connexin-32 located on chromosome X p l 3 cause X-
lin-like struture, a single transmembrane domain, and an intra- linked CMT (see Hereditary Neuropathy chapter).
cytoplasmic tail. It is believed that two P molecules on
302
0 Another major component of PNS myelin is myelin-associ-
opposing surfaces of the myelin membrane interact at the inter- ated glycoprotein ( M A G ) . MAG localizes to periaxonal
302
period lines. This interaction appears to be important in

83
myelin sheath, noncompact myelin at Schmidt-Lanterman in-
myelin compaction and stabilizing the major dense line. Of cisures, and paranodal terminal myelin loops. The protein is felt
note, mutations in the gene that encodes for P located on chro-
0 to contribute to the stabilization of the myelin sheath. There are
mosome lq22-23 can result in Charcot-Marie-Tooth disease no known mutations involving MAG. However, an acquired
type IB (CMT1B) or CMT3 (Dejerine-Sottas disease or con- immune-mediated disorder associated with antibodies (usually
genital hypomyelinating neuropathy) (see Hereditary Neurop- IgM) directed against epitope(s) on the MAG protein result in a
athy chapter). demyelinating sensorimotor polyneuropathy (see Acquired
Peripheral myelin protein-22 (PMP-22) is a 22-kD protein Neuropathy chapter).
that comprises 2 - 5 % of PNS myelin protein. The protein is
302
A series of myelin basic proteins (MBPs) weighing 12-22
composed of four transmembrane domains, two extracellular kD account for approximately 15-20% of PNS myelin pro-
loops, one intracellular loop, and two intracellular tails. PMP- teins. These proteins are also localized to compact myelin. A
302
22 localizes to the compact myelin, where, like P , it is felt to

0 15-kD protein, named P , comprises 1-14% of PNS myelin pro-
2
have a role in stablizing the myelin sheath. In addition, it ap- tein. Thus far, there are no known genetic or acquired periph-
302
pears that normal expression of PMP-22 is required for in- eral neuropathies related to abberant MBPs or P . 2
tegrity of the axon itself. As with P , mutations involving the

0 Schmidt-Lanterman Incisures. Schmidt-Lanterman in-
gene encoding PMP-22 located on chromosome 17p 11.2 may cisures are direct cytoplasmic connections between the abax-
result in CMT1 A, hereditary neuropathy with liability to pres- onal and adaxonal regions that penetrate the compact myelin
sure palsies (HNNP), or CMT3 (see Hereditary Neuropathy layer providing a cytoplasmic pathway within the Schwann cell
chapter). (Figs. 5-3 and 5-4). They may also extend for a short distance
Connexins are a family of gap junction structural proteins, involving only a few lamellae. These structures were initially
which are important in cell-to-cell communication. Connexin- believed to be a result of artifacts created in the cell fixation
32 is a 32-kD myelin protein composed of four transmembrane process, but are now known to be vital regions of the living cell.
domains. Six connexin-32 molecules oligomerize into a hexa-
302
The number of incisures is directly proportional to the thickness
meric structures on the Schwann cell lamella. Connexin-32 is of myelin and can reach 25 incisures per internode in the
localized to the paranodal region and Schmidt-Lanterman in- R I O I . I 1 2 . 1 3 7 The incisures appear to be more prevalent in devel-
A T
cisures, where it forms intercellular channels that allow diffusion oping and remyelinating nerves following demyelination. 113137
164 — P A R T II BASIC AND ADVANCED TECHNIQUES
Figure 5-4. M y e l i n a t e d n e r v e . A.Transverse section through a node of Ranvier with multiple Schwann cell microvilli projecting toward the ax-
olemma. B, Schematic representation of a myelinated axon's paranodal region. N o t e how the axon is fluted and the surrounding myelin conforms
t o the shape of the axon.The microvilli described above are also depicted projecting into the node of Ranver. C , A close drawing of a Schmidt-
Lantermann incisure. A microtubule is shown spiraling through the incisure. D,The longitudinal section through a mouse myelinated nerve is
demonstrated with a Scmidt-Lantermann ineisure.The drawing of the axon in the center of the diagram reveals how the distal paranodal region is
less bulbous than the proximal paranodal region with the proximal portion of the axon t o the right. (From Williams PL, Warwick R, Dyson M,
Bannister LH (eds): Gray's Anatomy. Edinburgh, Churchill Livingstone, 1989, p 900, with permission.)
In longitudinal section, the Schmidt-Lanterman incisures form Nodes of Ranvier

a series of "pockets" between the plasmalemma of the myelin Ranvier was the first to appreciate that an axon's myelin
sheath (Fig. 5-4). Each of the pockets usually contains a single sheath was demarcated at regular intervals by discontinuities in
microtubule and multiple vesicles. Incisures are not rigidly the myelin. Although unknown to Ranvier, these discontinu-
268
fixed structures, but may change their size in response to ities were actually axonal regions completely devoid of myelin.
trauma or the external tonicity of the surrounding medium. 199
A rather sharp angle, at times extending back under the myelin's
Although a number of theories have been suggested regarding exterior most sheath, may be formed at that aspect of the nerve
the purpose of Schmidt-Lanterman incisures, their true func- where the myelin ceases. Two aspects of the node of Ranvier
tion remains unclear. are often described, i.e., the nodal axon and paranodal region. 360
Figure 5-5. Paranodal region. Electron micrograph

longitudinal section through a mouse sciatic nerve at t h e
node of Ranvier. N o t e the manner in which t h e paranodal
myelin terminates by forming terminal loops (tl) and at-
taching t o the axolemma. T h e axon (ax) contains mito-
chondria (m) and multivesicular bodies (mvb). (From
Thomas PK, Ochoa J: Microscopic anatomy of peripheral
nerve fibers. In Dyck PJ,Thomas PK, Lambert EH (eds):
Peripheral Neuropathies, 2nd ed. Philadelphia, W.B.
Saunders, 1985, pp 39-96, with permission.)
Nodal Axon. The diameter of the axon narrows at the nodal The paranodal region of membrane has been suggested to be
region (nodal axon) and has been found to be reduced to one- associated with action potential generation because a significant
third of the internodal area in the ventral and dorsal root fibers of concentration of sodium ions were found in the terminal loops
cats. The middle of the nodal axon, however, increases in size to of Schwann cell cytoplasm. - It is suggested that the terminal
87 330
result in the entire nodal axon appearing somewhat barrel- loops are the source for the inward-directed sodium ions
shaped (Fig. 5 - 4 ) . " Additionally, the region of the axon
199
through the nodal axolemma (see below). At the termination of
devoid of myelin approximates an area 20-25 p m or a distance 2
an impulse, the sodium is pumped out of the axon through the
approaching 1.5 pm between the two internodes in large nodal axolemma and sequestered into the terminal loops by the
fibers. 11133
It has been demonstrated through freeze-fracture Schwann cell's microvilli, which extend into the nodal region.
techniques that the axolemma contains on its outer surface a This postulate remains speculative and awaits further investiga-
number of particles approximating a density of 1200/pm that 2
tion (see below).
have been suggested to be sodium channels. - One also notes
279 290
an increase in the numbers of axoplasmic organelles such as en- Peripheral Nerve Structure
doplasmic reticulum, lysosomal granules, glycogen, and, in par- Typical peripheral nerves are structures consisting of individ-
ticular, mitochondria in excess of five times that of the internodal ual nerve fibers arranged in a characteristic manner. Specifically,
axoplasm. The nodal axoplasm also has a more gelatinous con- single nerve fibers are collectively arranged in large groups
sistency with comparatively higher amounts of sulfated and non- known as funiculi (Fig. 5-6). Funiculi of different numbers
324
sulfated glycosaminoglycans. - Located within the boundaries

2 204
and sizes may be observed in cross-section at any level along
of the basal lamina (the material covering the Schwann cell and the course of a nerve. Individual nerves and funiculi are invested
nodal region), axolemma, and paranodal myelin extensions is the and supported by connective tissue. Three distinct subcategories
gap substance. Although the composition and function of the of connective tissue have been identified in peripheral nerves:
gap substance is not completely known, it is believed to partici-
pate in action potential generation.
Paranodal Region. An interesting observation may be made if EN
one considers the gross morphology of two adjacent internode re-
gions in close proximity to the node of Ranvier, i.e., the paran-
odal region. There is an asymmetric paranodal enlargement of the
myelin sheath extending approximately 40 pm on either side of
the node (Fig. 5-4). Specifically, the myelin bulge is some-
133 2 1 3
what larger on the proximal (closer to the cell body) than distal
side of the node with a distinct series of 3-6 indentations into the
bulbous myelin ends forming a crenated appearance when viewed
transversely (Fig. 5 -4).269.329.35s j regions between the bulging
n e
aspects of myelin are filled with Schwann cell cytoplasm contain-

ing multiple mitochondria at 10-20 times the concentration in the
internode. - Of note, the axon is fluted and surrounded by the
9 358
myelin, which conforms to its shape (Fig. 5-4). The terminal as-
pects of the myelin lamellae turn acutely in toward the axolerqma
to form out-pocketings or terminal cisternae of Schwann cell cyto-
plasm (Fig. 5-4 and 5-5). These terminal loops of Schwann cell
cytoplasm are firmly adherent to the axolemma and their extent
defines the portion of the nerve referred to as the paranodal Figure 5-6. Nerve subcomponents. Diagrammatic representation
region. 199249
A few of the terminal loops do not reach the ax- of a peripheral nerve with supporting connective tissue forming t h e
olemma. Additional cytoplasmic extension of the paranodal nerve's various subcomponents.Three individual fasciculi are shown.
Schwann cell plasmalemma forms several rows of microvilli that The surrounding epineurium (EP) invests t h e perineurium (PE), which
are directed in a radial fashion toward the nodal axolemma and forms t h e fasciculi of a peripheral nerve.The endoneurium (EN) sur-
terminate in the gap substance of the node (Fig. 5-4 and 5-5). rounds individual axons within the funiculi.
endoneurium, perineurium, and epineurium. Each axon is 324

corrugated shape because the elastin fibers tend to shorten the
surrounded by an endoneurial sheath, where multiple axons nerve slightly. Positioning the limb where the joint angle
form the funiculi that in turn are invested by perineurial con- stretches the nerve results in tension on the nerve. The nerve's
nective tissue. Multiple funiculi are invested by areolar connec- undulations disappear as the slack previously present is re-
tive tissue constituting the epineurium forming the peripheral moved. The epineurium's ability to endure stretching, however,
portion of the nerve proper. All three types of connective tissue is limited. Continued stretch placed on the nerve is further con-
structures contain fibroblasts, collagen fibers, mast cells, and trolled by the funiculi unless their tensile strength is exceeded.
macrophages. Additionally, the perineurium contains mesothe- It is believed that the epineurium with its component vessels
lial cells. and adipose tissue serves the function of cushioning and pro-
Epineurium. A considerable amount of areolar connective tecting the individual nerve fibers. 324
tissue surrounds the individual funiculi defining a peripheral nerve Perineurium. Contained within the epineurium are individ-
(Fig. 5-6). At the surface of the peripheral nerve mis areolar con- ual funiculi defined by an investing connective tissue condensa-
nective tissue condenses to form a well-delineated connective tion composed of three distinguishable layers (Fig. 5-6). The 322
tissue sheath lending a specific limiting structure to the nerve. At outermost or external layer of connective tissue forms a smooth
regions of the nerve where individual branches arise, the transition from the well organized and defined perineurium with
epineurium forms clearly demarcated connective tissue condensa- smaller and tightly arranged collagen fibers to the less dense
tions extending for some distance proximal to the site of the epineurium of larger and less organized collagen fibers. An in-
branch's departure from the main nerve trunk. Within the termediate layer has a recognizable layer of lamellated flat per-
epineurium and surrounding the individual funiculi are adipose ineurial cells. The innermost or internal layer consists of a
cells and elastin fibers. The amount of adipose tissue some-
103J04
single row of rather flat perineurial cells bound together with
what depends on the overall body fat content and is highly vari- tight junctions. A potential subperineurial space separates the
able from one nerve to another. Adipose cells are rarely found in innermost aspect of the perineurium from the endoneurium.
any other portion of a peripheral nerve, i.e., endoneurium or per- This inner layer is believed to be responsible for the "diffusion
ineurium. Also contained within the epineurium are nutrient blood barrier" of peripheral nerves. -
278 293
vessels and lymphatics. The total amount of epineurium at any one The perineurial sheath of motor fibers does not extend all the
location along the peripheral nerve is highly variable but charac- way to the neuromuscular junction but ends approximately 1.5
teristically there is more epineurial tissue where nerves cross a pm from it. This is a possible entry into the perinerium for for-
joint. At any portion of the nerve, the amount of epineurial tissue
319
eign substances such as bacteria or toxins. - Because of the
172 283
is inversely proportional to the number of funiculi. subperineurial space, there is believed to be some movement pos-
Observing a peripheral nerve in its tissue bed reveals that it sible for the contained nerve fibers. Funiculi range from 0.04 to
appears to be somewhat convoluted. This appearance is due to 2.0 mm in diameter and the perineurium can range in thickness
the elastin contained within the epineurial tissue. When the limb from 1.3 to 100 pm. In addition to the nerve as a whole appearing
is positioned such that there is little tension on the nerve with somewhat tortuous, the funiculi also have a similar appearance
respect to joint crossings, the nerve takes on the above noted within the epineurial tissue due to the perineurium's elasticity.
Funicular arrangements in peripheral nerves are particularly
interesting. Individual funiculi repeatedly merge and separate
from their neighbors contained within the epineurium, thus
forming funicular plexi (Fig. 5-7). - This results in consider-
317 321
able variation of funiculi over very short distances. It is esti-

mated that the longest portion of nerve with a constant pattern
of funiculi is about 15 mm, but the more typical pattern is a
change every few millimeters. At joint crossings, funiculi are
particularly small and abundant compared to other regions of
the peripheral nerve.
The perineurium serves the functions of protection and main-
taining intrafunicular pressure. As previously stated, the per- j
ineurium is a rather dense connective tissue layer with some
elasticity as well as tensile strength. At regions of joint cross-
ings the increased number of funiculi combined with an in-
creased amount of epineurial tissue affords some protection
from joint movement and repeated trauma. It is also believed
that the inner mesothelial layer is a good diffusion barrier pre-
venting the entrance of foreign substances. - - - The per-
186 238 301 303
ineurium also forms an efficient barrier to the spread of

infection into the funiculi. The circumferential perineurial
layers maintain a certain level of intrafunicular pressure. This is
confirmed by an immediate herniation of the individual nerve
fibers contained within a funiculus when the perineurium is
Figure 5-7. F u n i c u l a r plexus.A 3-cm segment from the musculocu- breached. - The exact role this pressure plays in the func-
306 318
taneous nerve of the arm demonstrating the funicular plexus formation. tioning of the axon is unclear but is believed to aid in the normal
N o t e how transverse section at s h o r t distances results in a different axoplasmic flow.
axonal arrangement. (From Sunderland S: N e r v e s and Nerve Injuries, Endoneurium. Contained within the funiculus and envest-
2nd ed. Edinburgh, Churchill Livingstone, 1978, with permission.) ing the individual nerve fibers is a connective tissue known as
the endoneurium (Fig. 5-6). The endoneurium forms a fine sup- was a result of the elevated electrical insulation properties of the
porting sheath about each nerve fiber and at times subdivides myelin. - He suggested that the action potential would be
209 309
the funiculi into small groups of nerve fibers by way of connec- generated at the nodes of Ranvier and "skip" over the myeli-
tive tissue septae. In addition to the connective tissue, the en-
324
nated internode regions and introduced the term "saltatory"
doneurial tissue contains only capillaries and some localizations with respect to action potential propagation. Approximately 25
of interstitial fluid. The endoneurium is composed of fibroblasts years elapsed before Huxley and Stampfli provided conclusive
and collagen arranged both longitudinally and obliquely. The electrophysiologic evidence of saltatory conduction in myeli-
axon, Schwann cells, and associated myelin are surrounded by nated peripheral nerve fibers. - Longitudinal current mea-
149 150
the endoneurial tissue in a densely packed arrangement forming surements along the excited axon's surface membrane revealed
a supporting structure referred to as the endoneurial tube. that at each node of Ranvier, the current shifted to a slightly
There is some speculation that the collagen comprising the en- later time while it was constant for the entire extent of the in-
doneurium is secreted not by the fibroblasts but by the Schwann ternode (Fig. 5-8). Transverse current measurements over the
cells because of the relative paucity of fibroblasts, i.e., there are same nerve demonstrated that a current entered the axon only at
9 times as many Schwann cells as fibroblasts. - - A fine retic-
7 38 328
the nodes of Ranvier and minimal, if any, current passed
ulum of material forming a tubular sheath is located between through the heavily myelinated internodes. This electrophysio-
the endoneurium and the basement membrane external to the logic evidence required approximately 30 years for anatomic
Schwann cell cytoplasm known as neurilemma, Schwann evidence to provide insight regarding the actual transaxonal
sheath, or Schwann membrane. - One may also simply think
94 324
pathways for ion movement.
of the endoneurium as consisting of two layers, an inner reticu- One would anticipate that transaxonal current movement
lar sheath and an outer collagenous sheath. preferentially localized to the nodes of Ranvier should reveal
Individual nerve fibers located within the endoneural sheaths high densities of sodium ion channels. Indirect gating current
display undulations as do the funiculi. A small amount of stretch measurements suggest that mammalian (rabbit) myelinated pe-
is therefore possible until the undulations disappear with in- ripheral nerves contain approximately 82,000 sodium channels
creasing tension. Although the endoneurium has a limited per node or about 1500/pm of the nodal region in order to sus-
2
amount of tensile strength, the majority of stretch is resisted by tain the observed action potentials. - In rat sciatic nerve this 45 46
the epineurium and perineurium. This is exemplified when number is on the order of 21,000 sodium channels per node
nerve roots, which lack a perineurium, fail under tensile forces while frog myelinated nodal membranes contain sodium chan-
prior to peripheral nerve rupture. - The endoneurium is not a
322 324
nels approximating 5,000/pm . - Similar electrical studies
2 243 245
rigid structure but conforms to its contents. Recall that there is a also suggest that the internodal density of sodium channels is
certain amount of pressure contained within the endoneurium less than 2 5 / p m . - Morphometric analysis through freeze-
2 273 274
that helps to maintain its tubular appearance. If the axon under- fracture methods demonstrated a maximum of 3,000/pm ap- 2
goes degeneration, the endoneurium shrinks but again expands propriately sized particles in the paranodal region and
upon axonal regrowth. The endoneurial sheath is also be-
320
2,000/pm particles in the nodal region that were assumed to be
2
lieved to serve the function of insulating neighboring nerves sodium channels. - - - Potassium channels are sparse in the
13 14 86 356
from each other with respect to their electrical properties thus nodal region but are present in the internodal and paranodal
avoiding cross-talk or ephaptic conduction. 91110
region covered by the myelin sheath (Fig. 5-9). Electron 352 3 5 3
microscopy combined with cytochemical identification tech-

Action Potential Generation niques reveal preferential concentrations of sodium ions in the
Correlation Between Anatomy and Physiology terminal loops of the myelin lamellae within the paranodal
In 1925, Lillie first proposed that the 10 times greater con- region. These terminal loops appear to serve as a reservoir for
86
duction of myelinated compared to unmyelinated nerve fibers the sodium ions required for nodal depolarization. Additionally,
Figure 5-8. Myelinated a x o n c u r r e n t profile.

A, Longitudinal currents detected along the c o u r s e
of a myelinated axon.The series of three measure-
ments from each internode is essentially t h e same
but d e m o n s t r a t e s a noticeable prolongation of la-
tency following each node of Ranvier. B, Recordings
demonstrating that significant c u r r e n t flow associ-
ated with an action potential occurs at the nodes of
Ranvier while the internode is a passive cable con-
ductor, thus supporting the concept of saltatory con-
duction. (From Stampfli R: Overview of studies on
the physiology of conduction in myelinated nerve
fibers. In Waxman SG, Ritchie JM (eds): Demyelinating
Diseases: Basic and Clinical Electrophysiology. N e w
York, Raven Press, 1981, pp I I - 2 3 , with permission.)
168 — P A R T II BASIC A N D ADVANCED TECHNIQUES
terminal Schwann cell loops. Sodium ions in the terminal loops

are subsequently concentrated into the small triangular spaces
or junctional clefts between the axolemma and the paranodal
loops. When the transmembrane voltage is sufficient to reach
threshold and activate the voltage-dependent sodium gates, the
previously sequestered sodium ions now enter the axon convey-
ing the inwardly directed current flow associated with action
potential passage. Following sodium inactivation, the intra-
axonal sodium ions are pumped into the node of Ranvier's gap
substance by the sodium/potassium ATPase pump. These
sodium ions are then reabsorbed by the Schwann cell microvilli
Figure 5-9. Distribution of sodium and potassium channels
to complete the sodium ion cycle. In this proposal, the sodium
in myelinated nerve. N o t e that t h e sodium channels are primarily
ion cycle is complete and serves the function of action potential
located in t h e node of Ranvier; however, t h e potassium channels are
generation. Of course, this is only speculation and further direct
beneath t h e myelin sheath in the internode region. Na, sodium chan-
evidence is required for its complete acceptance.
g
nels; Kf, fast potassium channels sensitive t o a blocking agent of 4-AP;

g
Ks, slow potassium channels sensitive t o TEA block; and IR, inward Several indirect lines of evidence support the above hypothe-
g g
rectifier channels. (Ffrom: Waxman SG: Molecular organization and sis. Morphometrically, the appropriate number of assumed
pathophysiology of axons. In Asbury AK, McKhann GM, McDonald W I sodium channels are present in the paranodal membrane beneath
(eds): Diseases of t h e Nervous System: Clinical Neurobiology, 2nd ed. the terminal loops to generate an action potential. Laser coagula-
Philadelphia,W.B. Saunders, 1992, with permission.) tion of the paranodal membrane prevented action potential gen-
eration at that node despite a normal resting membrane
potential. Additionally, depleting the surrounding environment
239
immunocytochemical analysis demonstrated that the Na ~K - + +

of sodium ions did not prevent the nerve from continuing to con-
ATPase macromolecule coupling ATP hydrolysis with the active duct impulses. - - The large numbers of mitochondria sur-
86 87 330
transport of K and N a is primarily localized to the axolem-

+ +
rounding the paranodal and nodal region suggest that energy is
mas' nodal region. 363
consumed to support the metabolic functions of extruding, ab-
An interesting hypothesis has been developed in order to ex- sorbing, and concentrating the sodium ions. Finally, heat produc-
plain impulse generation by combining the above-noted electri- tion calculations during ionic movement yield results that are too
cal, chemical, and morphometric data (Fig. 5-10). Initially, the
86
great for the nodal membrane to dissipate, suggesting that more
terminal paranodal loops of myelin sequester and concentrate membrane than that available at just the node is required. 3135
sodium ions for storage, i.e., any sodium ions in the node of Of particular note in the above hypothesis is the absence of a
Ranvier region are absorbed by the microvilli of the Schwann potassium requirement for repolarization. Although unmyeli-
cells extending into the gap substance. The absorbed sodium nated nerves in mammals require a voltage-dependent potas-
ions are then transported to and concentrated in the paranodal sium conductance change, - - this is not the case in many
22 189 263
Figure 5-10. Cyclic flow of sodium ions during an action potential. During action potential production at t h e node of Ranvier, the
sodium ions contained in the junctional cleft between the axolemma and terminal loops (PN: paranodal loops) enter into the axon constituting
the inward current flow when the voltage-dependent sodium gates are opened. Following sodium inactivation, t h e sodium is actively transported
t o the exterior of the axon through the sodium-potassium pump.The extruded sodium is then reabsorbed by the Schwann cell's increased surface
area through t h e microvilli (MV) and again concentrated in t h e junctional clefts and terminal loops that act as sodium ion reservoirs. (From
Ellisman MH: Molecular specializations of t h e axon m e m b r a n e at nodes of Ranvier a r e n o t dependent upon myelination. J Neurocytol
1979;8:719-735, with permission.)
myelinated nerves. - 15
- This is also consistent with what is
118J87 190
velocity is shown to be 6 m/s/pm for relatively large-diameter
presently known about mammalian myelinated nerves. There myelinated mammalian nerve fibers. For example, a nerve
148
are primarily two major groups of potassium channels, one with fiber with a diameter of approximately 8 pm would be predicted
fast gating kinetics (K ) and a second group with slow gating ki-
f to have a conduction velocity approaching 48 m/s. Different
netics (K ; Fig. 5-9). The slow-gated potassium channels are lo-
s conversion factors have been found for nerve fibers with smaller
cated primarily but not exclusively in the node of Ranvier diameters, e.g., 1.7 m/s/pm for unmyelinated nerve fibers. 110119
region, while the fast-gated channels are situated in the paran- The above noted designation of nerve fibers by Greek letters
odal region under the myelin sheath. The slow-gated potassium has now been generally accepted as a means to classify nerves
channels most likely contribute in part to the resting membrane based on their size and associated conduction velocity. The clas-
potential. However, neither channel type participate in the for- sification is as follows: A - a fibers are myelinated fibers with an
mation of the propagated action potential since blocking either approximate external diameter of 6-12 pm, while A-8 fibers
type does not alter the action potential's morphology. The im- have diameters of about 2 - 6 pm. The fibers that are unmyeli-
portance of the fast-gated potassium channels is primarily to nated and have external diameters of 0.5-2 pm are the C fibers.
prevent bursting or repetitive firing of the axon following action A second classification system considering only afferent fibers
potential propagation past a node of R a n v i e r . There may351-353
was introduced by Lloyd, in which Roman numerals are used. 210
be several subtypes of fast-gated potassium channels but their This system's classification is based on fiber diameter and its
explicit roles in action potential burst suppresion remain to be relation to the Greek letter classification is designated: I (21-12
fully elucidated. In myelinated fibers, voltage clamp experi- pm); II (12-6 pm: A - a ) ; III (6-1 pm: A-8); and IV (< 1 pm: C
ments document that repolarization is not conveyed by a potas- fibers).
sium current but occurs through sodium inactivation and a
sodium or potassium back-leak current mediated through non- Fiber Diameter and Velocity Correlates
voltage-gated or passive ion channels. - 25
Intra-axonal inves-
44136
In order to compare the relationship between fiber diameters

tigations have supported these observations. - 188 189
and various aspects of the compound nerve action potential in

living healthy humans, it is possible to remove 2-5 fascicles
C o m p o u n d N e r v e Action Potential
10-15 cm in length from the sural nerve. Microscopic analysis
Fiber Diameter and Conduction Velocity of this small portion of the nerve results in a histogram display-
If a maximal depolarizing stimulus applied to a mixed periph- ing the range of myelinated fiber diameters contained within the
eral nerve and the ensuing resultant action potential is recorded sural nerve (Fig. 5-11). For fiber diameters between 2 and 12
several centimeters distal to the activation site, a characteristic pm, there is a bimodal distribution with respect to size and cor-
spike potential is recorded. When this experiment is performed, responding numbers of nerves. One group of nerve fibers
198
it is noted that as the distance between the stimulus and record- ranges from 4 to 12 pm, while the second major subgroup lies
ing sites increases, the observed nerve action potential changes between 2 and 4 pm. A third large category of nerve fiber diam-
from a single spike into a temporally dispersed potential with eters range from approximately 0.25 to 1.5 pm. Individual nerve
several peaks separated in time (Fig. 5-1 l)ô.io7,i98 These investi- fibers 11-12 pm in diameter comprise the largest fibers exam-
gations clearly demonstrate that the recorded action potential ined in any quantity.
arising from a mixed peripheral nerve is composed of multiple Recording the compound nerve action potential from the
subcomponent action potentials likely correlated to individual above nerves results in the anticipated three-peak potential. The
nerve fibers each with their own and slightly different conduc- three peaks correspond to the A-a, A-8, and C subcomponents.
tion velocities, thus yielding a compound nerve action poten- By knowing the distance between the stimulation and recording
tial. The original studies were performed on frog sciatic nerves points as well as the arrival times for various waveform portions
and produced a prominent waveform designated the "A poten- to the site of recording, it is relatively simple to calculate a con-
tial," which consists of several subcomponent peaks referred to duction velocity for the identified subcomponents. For example,
in increasing times of arrival (decreasing conduction velocity) if the distance between the site of stimulation and recording is
by the Greek letters a, (3, y, and 8. In other words, the fastest- 100 mm and the time it takes the leading portion of the action
conducting fibers ( a fibers) arrive at the recording site first while potential to cover this distance is 10 ms, the nerve conduction
the slowest-conducting nerves (8 fibers) are recorded later. The velocity is 10 m/s (NCV = distance/time; NCV = 100 mm/10
A - a and A-p peaks of the compound nerve action potential ap- ms = 10 mm/ms = 10 m/s). In the teased sural nerve fibers, the
parently are one and the same as different investigators labeled beginning of the A - a peak was determined to be 61 m/s. The A-
the same peaks differently. Additionally, the ypeak was later
198
8 peak had a conduction velocity of 19 m/s, while the fastest C
shown to be a recording artifact. As a result, there are only two fibers propagated at 1.6 m / s . Because we know that the
198
peaks now designated in the A potential, i.e., A - a and A-8. A fastest-conducting nerve fibers are the largest-diameter fibers,
second major peak designated the "B potential" is believed to we can align the histogram from largest to smallest and com-
consist of slower-conducting fibers than those in the A potential pare the fiber diameters with the compound nerve action poten-
and are found only in preganglionic autonomic fibers, which are tial's various peaks. For example, the fibers propagating at 61
small-diameter myelinated nerves. Finally, there is another class m/s are the 11- and 12-pm diameter fibers. This yields a nerve
of even slower-conducting fibers known to comprise a third conduction velocity to diameter ratio of 5.3:1, which is in close
major compound nerve action potential peak and referred to as agreement with the 6.0:1 ratio previously noted. Similar corre-
the "C potential." The C potential is generated from small un- lations can be drawn for other fiber diameters and velocities. In
myelinated nerve fibers that consist of sympathetic and afferent addition to examining teased fascicles in vitro, corresponding
dorsal root fibers.198
clinical studies have also been performed in vivo on healthy
The fibers producing the A potential demonstrate a linear re- sural nerves with near-nerve recording and stimulating tech-
lationship between conduction velocity and external fiber diam- n i q u e s . - Findings for whole nerve are similar to those for
196 198
eter. This relationship between fiber diameter and conduction teased preparations.
01—
12 8 4 2.0 1.5 1.0 0.5
DIAMETER (jjm) DIAMETER (urn)
Figure 5-11. Compound action potential and histograms of fiber diameters of normal nerve. Myelinated fibers are represented on the
left, unmyelinated fibers on the right (From Lambert EH, Dyck PJ: Compound action potentials of sural nerve in vitro in peripheral neuropathy. In
Dyck PJThomas PK, Lambert EH, Bunge R (eds): Peripheral Neuropathy, 2nd ed. Philadelphia,W.B. Saunders, I984,pp 1030-1034, with permission).
Biopsy specimens from normal humans reveal that the sural nerve action potential's major spike recorded at the mid-calf is
nerve contains approximately 6600 fibers. About 2600 fibers
32
about 14 pV and is estimated to arise from 1800 large nerve
are in the 7-13 pm range, while roughly 4000 fibers have diam- fibers. The slower components of the recorded potential (spike
eters ranging from less than 1.0 pm to 7 pm. Stimulating the components following the main spike) and are believed to arise
sural nerve maximally (enough current to activate all of the from fibers 7 pm in diameter or less. At the mid-calf, the slow-
fibers) and recording at two proximal sites (e.g., mid-calf, 15 est velocities averaged 15 m/s, which corresponds to fibers 4
cm and popliteal fossa 50 cm) reveals a number of interesting pm in diameter. Recording in the popliteal fossa for the same
findings (Fig. 5-12). The peak-to-peak amplitude of the compound sural nerve revealed that the amplitude of the major spike had
decreased markedly to about 1 pV and the duration of the po-

tential had increased. This implies that the greater distance al-
lowed a greater separation between the fastest and slowest
fibers, i.e., increased temporal dispersion. The potentials de-
creased in amplitude because the individual action potentials
were less synchronized and, therefore, their voltages summated
less, producing a smaller potential as well as some phase can-
cellation. Additionally, note how much longer (increased dura-
tion) the potential is because of the small-caliber slowly
conducting fibers. This finding has already been discussed in
POPL. 0.5 UV
detail in the chapter addressing volume conduction factors (see
FOSSA
Chapter 2). + FIBRES
600
Compound Nerve Action Potential Modeling
Although the above observations are based on sound electro-
physiologic and histologic principles, credence would be lent to - 400
these assertions if one could model the compound nerve action
potential based upon the above findings. Through painstaking
histologic and electrophysiologic measurements on the myeli- - 200
nated fibers of the cat saphenous nerve, Gasser and Grundfest 108
were able to model the compound nerve action potential. A his-

tologic analysis of 2119 myelinated nerve fibers demonstrated a T l J 0
clear bimodal distribution of fiber diameters. Electro- 6 4 2 0

physiologic analysis of fiber diameter and conduction velocities DIAMETER, pm
allowed these investigators to arrive at several assumptions: (1)
conduction velocity is directly proportional to external nerve Figure 5-/2. Constituents of the sural nerve sensory action
fiber diameter, (2) externally recorded monophasic action po- potential and distribution of myelinated fiber diameters.
tential spikes are proportional to fiber cross-sectional area, and Maximal stimulation at t h e lateral malleolus resulted in potentials as
(3) action potentials from different fiber sizes all have the same recorded from t h e mid-calf and popliteal fossa regions. Dashed lines
morphology and duration that can be approximated by triangles connect neural components conducted at t h e same velocity and point
with a rising phase one-third, and a falling phase two-thirds the t o t h e corresponding fiber diameter in t h e histogram. (From Buchthal
total potential's duration. By aligning the correct number of tri- F, Behse F: Sensory action potentials and biopsy of t h e sural nerve in
angles with appropriate nerve fibers' sizes and numbers, an ex- neuropathy. In Canal L, Pozza G (eds): Peripheral N e u r o p a t h i e s .
cellent approximation of the externally recorded potential can Amsterdam, Elsevier, 1978, p I, with permission.)
be described (Fig. 5-13). This amazing demonstration con-
firmed that the above noted assumptions were correct and one This finding has obvious implications for attempting to under-
could directly correlate fiber diameter with conduction velocity. stand various disease processes that affect specific nerve fiber
12 10 8
DIAMETER ( JJ) TIME ( m s )
Figure 5-13. Compound nerve action potential modeling. A series of triangles were correlated t o the fiber diameter distribution of a cat
saphenous nerve and subsequently summated t o yield a predicted action potential.The correlation between the predicted potential and that actu-
ally recorded agree amazingly well.The only difference is noted by t h e dotted line. (From Gasser HS, Grundfest H:Axon diameters in relation t o t h e
spike dimensions and t h e conduction velocity in mammalian A fibers. Am J Physiol 1939; 127:393-414, with permission.)
populations with respect to anticipating the effect on conduction wise to take a lesson from our surgeon colleagues regarding pa-
velocity. Subsequently, more elegant and sophisticated tech- tient positioning. Three principles should guide the practitioner:
niques have confirmed this deceptively simple method of sum- (1) patient comfort, (2) examiner comfort, and (3) anatomic ex-
mating triangles. Of note, essentially the same method is posure. As noted previously, the patient must be comfortable
employed in present-day highly sophisticated computer model- with the practitioner, surroundings, explanation of procedure,
ing techniques. and physically able to relax. Patient comfort and relaxation
ensure an optimal examination. In order to collect the necessary
data, the practitioner must be comfortable. Many studies can be
BASIC NERVE C O N D U C T I O N STUDY (NCS) quite time-consuming and the uncomfortable practitioner has a
PRINCIPLES tendency to rush the examination thus potentiating improper
data collection or an incomplete consultation. Finally, the
A nerve conduction study may be defined as the induction region of the body to be examined should be readily accessible
of a propagating action potential in the peripheral nervous to the stimulating electrodes and measurements required.
system and the subsequent recording of its neural impulse at Attempting to stimulate a nerve or measure distance in an awk-
some location distant to the impulse's initiation site. One may ward position only predisposes one to experimental error (see
investigate a number of different types of nerve conduction below). Of course, proper exposure should not cause the patient
studies such as: pure motor nerve evaluations by assessing the any undue discomfort.
evoked compound muscle action potentials (CMAPs), pure Before beginning neural excitation, the patient must be
sensory nerve action potentials (SNAPs), and compound warned that an electrical impulse is about to be delivered. Many
nerve action potentials (CNAPs) from mixed (motor and sen- individuals are uncomfortable with electricity and afraid of
sory) nerves. By calculating the ability of peripheral nerves to being shocked. The patient should be told what to expect when
conduct an electrical impulse, it is possible to assess their the current is delivered. Some statement such as "a tingling sen-
functional status in a healthy state and responses to various sation will be felt down your hand and into your fingers (de-
disease processes. Nerve conduction studies allow one to ac- scribe which fingers) and your muscles might jump" should be
curately localize focal lesions or detect generalized disease given to the patient. Also, it is a good idea to begin at a zero cur-
processes along accessible portions of the peripheral nervous rent intensity and slowly approach the desired intensity allow-
system. By first considering the general principles pertinent ing the patient to become familiar with the electrical sensation.
to performing nerve conduction studies, we can better appre- Indirect lighting may also be of help. A patient in the supine po-
ciate the various factors affecting our ability to perform opti- sition stares at the ceiling by default and possibly into bright
mal investigations. overhead fluorescent lights. This can be disconcerting and
simply annoying, especially if the examination begins to ap- !
PATIENT CONSIDERATION proach 1 hour. Considerations given to the patient as those
noted above will ultimately result in a fruitful electrodiagnostic
The practitioner must first direct attention to the patient. medicine consultation for both the patient and physician.
Placing the patient in the supine position on a comfortable
plinth allows one to examine significant portions of the periph- STIMULATION
eral nervous system while at the same time affording the patient
an opportunity to maximally relax. Relaxation is of prime im- For the purposes of NCS, the peripheral nervous system may
portance in optimizing the electrodiagnostic medicine examina- be activated in one of two ways: (1) electrical stimulation or (2)
tion. Because the majority of electrophysiologic responses are magnetic stimulation. In this chapter magnetic stimulation is
small and significant amplification is necessary, any noise aris- not covered as its full diagnostic potential with respect to pe-
ing from anxiety-induced muscle contraction may well obscure ripheral nerve disease remains to be explored in more detail (see
the desired response. Patient cooperation and relaxation can chapter 10 on Magnetic Stimulation). In short, magnetic stimu-
most easily be obtained by completely explaining the procedure lation cannot fully depolarize the full complement of fibers con-
about to be performed in simple and understandable language. tained within a peripheral nerve. This technical shortcoming
92
A number of patients often arrive at the examiner's office misin- leads to rather obvious limitations in its diagnostic usefulness.
formed by well-intentioned acquaintances or physicians with At this time the greatest utility of magnetic stimulation appears
respect to the procedures performed and the amount of discom- to be limited to excitation of the cerebral cortex and possibly the
fort experienced. A full explanation of the procedure, purpose spinal cord. When referring to stimulation in this chapter, there-
and importance of the examination, as well as the sensation fore, it is assumed that electrical and not magnetic excitation is
about to be experienced is important to convey. This can only be employed.
carried out with credibility if the practitioner has experienced Electrical depolarization of the peripheral nervous system
electrical excitation of the peripheral nervous system firsthand. can be accomplished by using either surface or needle elec-
All practitioners, therefore, should have nerve conduction stud- trodes and delivering a square wave pulse. The more commonly
ies performed on themselves prior to subjecting patients to these used type of stimulation uses a surface cathode and anode. The
procedures. After the patient has been informed as to the benign cathode or negative pole of the stimulator results in a localized
nature of the nerve conduction study and why it is important to region of negative potential while the anode produces a sur-
obtain the desired information, they are usually able to cooper- rounding region of excess positive potential. Current flows be-
ate fully with the practitioner. tween the anode and cathode with decreasing density as
Once the electrodes have been attached to the patient (see distance from the poles increases (see Chapter 3). Should these
below) and before a nerve is excited, the limb must be properly two poles be located over excitable tissue such as peripheral
positioned. Positioning is one of the most important aspects of nerve, a region of depolarization is induced about the nerve
the nerve conduction examination. The practitioner would be under the cathode. If the cathode produces a depolarizing impulse
of sufficient magnitude to exceed the transmembrane voltage intense stimulus deliveries (elevated current/voltage and/or
threshold for sodium activation, then a self-propagating action pulse duration) may be necessary in nerves difficult to excite
potential is generated. The action potential generated under the secondary to the previously noted conditions or disease states.
cathode will propagate both proximally and distally from the In these situations, consideration should be given to the amount
cathode along the nerve's longitudinal course. Those action po- of stimulation required. As the current is gradually increased
tentials traveling toward the anode may be partially extin- from a threshold to maximal intensity, the larger myelinated
guished at the hyperpolarized region surrounding the anode axons are activated preferentially because of their lower thresh-
(anodal block) or collide with action potentials produced under old 60.61.354 with supramaximal stimuli, the remainder of the rel-
the anode (anodal break excitation). The occurrence of
140
atively smaller myelinated axons are excited. Also, the larger
anodal block in routine conduction studies is questionable as fibers may be preferentially activated by using submaximal cur-
one can typically detect a similar response by reversing the lo- rent intensities with relatively long pulse durations of 500 or
cation of the anode and cathode with the expected delay from a 1000 p s . -
257 258
more distant excitation site (see Chapter 3). Anodal break exci- When stimulating the peripheral nervous system, the pulse du-
tation, however, is capable of producing an action potential ration should initially be set between 0.1-0.2 ms and a starting
through neural depolarization at the cessation or "break" of the current intensity of zero delivered. It is also important not to
stimulus. This process has been postulated to occur by imped- change the pulse duration between different sites of neural acti-
ing sodium inactivation for the duration of the hyperpolarizing vation. A gradual elevation of the current delivered allows the
p u l s e . As a result there is an inward current flow into the
140
patient to adjust to the stimulating pulse. The evoked response is
membrane attempting to depolarize it. Because of the anodal carefully observed to ensure that it no longer increases in magni-
current, however, the nerve is prevented from generating an tude despite an increase in the current intensity. Should the re-
action potential. Once the stimulus pulse is over, the sodium in- sponse's amplitude continue to increase despite reaching the
activation persists long enough to result in an action potential, highest intensity on the stimulator, a longer pulse duration can
i.e., anodal break excitation. It is this action potential that may be used. The pulse duration is increased to the next highest set-
collide with the one originating under the cathode. No doubt, an ting and the current intensity is again reset to zero. The above
action potential propagates away from both the cathode and process continues until a supramaximal waveform is docu-
anode. mented. The frequency of stimulation is usually set to 1 Hz.
In addition to surface electrodes, needle electrodes may also Some practitioners prefer to stimulate not with a fixed stimulus
be used to evoke an action potential. Needle electrodes are usu- frequency but by applying every stimulus manually using the
ally used when the nerve is hard to stimulate with surface elec- footswitch on the EMG machine. Basically, it is important not to
trodes, although some investigators use needles routinely. If a deliver more stimuli to the patient than necessary. This implies
nerve is surrounded by excess subcutaneous tissue or edema that the investigator should have time between every stimulus to
preventing optimal stimulation with surface electrodes, a needle judge the respons and make changes with regard to stimulus pa-
cathode in combination with a needle or surface anode may be rameters. Stimulus rates higher than 1 Hz are generally unneces-
required. The advantage of needle stimulation is a more precise sary except with the purpose of averaging the response.
localization of the depolarizing pulse with less current/voltage If any difficulty is noted in obtaining the anticipated re-
required. Less current/voltage can also mean less discomfort.
28
sponse, several procedures should be explored prior to conclud-
Of course, there is a small risk of trauma to the nerve with ing that the response is absent. First, as noted above, a
needle stimulation secondary to accidental neural insertion. A supramaximal stimulus must be assured. If the current intensity
pulse duration of 50 ps usually suffices for needle stimulation. is maximal and a pulse duration of 0.5 ms or more is attained,
For a more detailed discussion of needle stimulation refer to one can be assured that a supramaximal stimulus has been de-
Chapters 2 and 3. livered. The question then arises as to whether the stimulus was
The magnitude of the stimulus can be categorized qualita- delivered to the nerve under investigation. Of course, one
tively in number of ways. A stimulus pulse with an intensity just should not attempt to perform an electrodiagnostic examination
capable of evoking an observable response is referred to as a without first being thoroughly familiar with peripheral nerve
threshold stimulus. Stimulus intensities less than the thresh-
1
and muscle anatomy. Moving the cathode several centimeters
old level are called subthreshold stimuli. Maximal stimuli are from side-to-side ensures covering the nerve's anticipated loca-
those that do not produce any further increase in the response. tion. All electrode leads must be checked for continuity as the
Those stimuli at intensities between threshold and maximal stimulus may not be reaching the patient or the response may
levels are known as submaximal stimuli. Finally, a stimulus not be arriving at the instrument. Although obvious, it is always
delivered at approximately 2 0 - 3 3 % above the maximal inten- a good idea to verify that the amplifier/preamplifier is on.
sity is called a supramaximal stimulus. It is the supramaximal Finally, averaging the response several times may improve the
stimulus that is the preferred intensity to be delivered in the ma- signal-to-noise ratio sufficiently to resolve a small response ob-
jority of routine nerve conduction studies. One can also easily scured by baseline noise (see Chapter 3). The especially deter-
quantify the stimulus used in peripheral nerve studies. An in- mined practitioner may consider recording from the nerve or
strumentation parameter defining the duration during which a muscle under investigation with a needle electrode to maximize
stimulus is delivered is known as the pulse width or pulse du- the response's amplitude. Should all of these measures fail to
ration and is measured in milliseconds or microseconds. A typ- produce a response, it is most likely absent secondary to pathol-
ical pulse duration for most nerve conduction studies is 100 ps ogy and not because of a technical deficiency.
or 0.1 ms with a range of 50-1000 ps. Once the practitioner has
chosen the appropriate pulse duration, greater for nerves deep RECORDING
below the skin surface, the intensity or amount of current/volt-
age delivered is adjusted to a supramaximal level. Typical volt- Following action potential generation it is necessary to record
age levels are between 100 and 300 volts or 5-35 mA. More the induced response. As with stimulation, one may use either
surface or needle electrodes. The optimal type of electrode de- about the needle electrode now includes the slow conduction of
pends on the situation (see Chapter 3). Surface electrodes in muscle fibers ( 4 - 6 m/s) and varies from one site to another.
general are somewhat more convenient to use because they can Instead of the one latency reflecting depolarization of the end-
be taped over the appropriate location and easily repositioned to plate zone, a needle located at some location in the muscle de-
another location. Additionally, because the skin's protective bar- pends upon slow conduction along the muscle fibers prior to
rier is not penetrated (except for somatosensory evoked poten- reaching the particular location of the needle. Intramuscular
tials, see Chapter 9), there is less patient discomfort and less of needle placement, particularly a concentric needle, does have
a concern regarding blood-borne infectious agents. As their the'advantage of selectively recording from a small area. This
name implies, surface electrodes are located on the body's sur- can be used to one's advantage in order to avoid volume-con-
face, and as a result are relatively far from the activated neural ducted responses from neighboring muscles (see Chapter 2).
or muscular tissue. This "distant" location has the advantage of When recording a response from nerve or muscle, three elec-
recording the summated response from a large portion of the de- trodes are required to obtain a response: active, reference, and
polarized tissue. This detected response is more representative ground electrode. The active electrode is placed as close to
of all of the tissue's fibers undergoing depolarization. As a the tissue under investigation as possible. This electrode is con-
result, surface electrodes are more appropriate when the total nected to the noninverting amplifier port (see Chapter 3) and
amount of depolarization is important, thereby attempting to may be referred to by several terms in addition to active elec-
quantify what portion of axons or muscle fibers were activated. trode. Initially the active electrode was called G l , which arose
Needle recording electrodes, although used less frequently, from the physiology and electroencephalography literature des-
also have some advantages in particular circumstances despite ignating the active electrode as Grid 1. More recently, the term
the small amount of discomfort and possible patient anxiety as- Gl is also called E - l for electrode 1 because the designation
sociated with their use. When there is excess swelling about a "active" becomes inadequate to accurately describe the situa-
muscle or nerve and inadequate surface recordings are obtained, tion encountered in far-field recordings (see Chapter 2). Also,
a needle electrode may be capable of better delineating the pres- this electrode may occasionally be designated the noninverting
ence or absence of a response. The advantage of a needle elec- lead. The second or reference electrode is plugged into the in-
trode is that it can be located next to, or within the tissue under verting amplifier port and was previously known as G2 repre-
investigation and record action potentials in nerve or muscle senting the Grid 2 electrode. As with E - l , the reference
with only a few active fibers. It is possible, therefore, to docu- electrode is designated E-2 or the inverting lead. This electrode
ment the presence of surviving axons or muscle fibers when sur- is usually placed at some distance from the active electrode so
face recordings may detect no response. This same capability, that the information from E-l is "referenced" or compared to E-
however, may also be a limitation when recording within nerve 2. Through the process of differential amplification, the ob-
or muscle. Placing a needle next to the nerve may produce a rel- served potential on the cathode ray tube (CRT) is really the
atively large amplitude even if surface recordings yield a small difference in voltage between the two electrodes, i.e., ( E - l ) -
response, but slight needle movement can produce large shifts (E-2). The actual locations of these two electrodes designate
in the amplitude. Positioning the needle in the same location be- whether the recording is a bipolar or referential montage (see
tween examinations to give identical amplitude readings is Chapter 2).
somewhat challenging and requires significant skill to ensure The anticipated site of E-l and E-2 placement should be
reproducibility. Surface electrodes yield more reproducible re- properly prepared with fine sand paper or a commercial pumice
sults despite smaller amplitude recordings because the previ- solution in order to reduce the skin's impedance if difficulty
ously noted distance effects on amplitude are comparatively with interference is encountered. Reducing skin impedance
reduced. Serial studies, particularly regarding blocked fibers, ensures that the electrical activity generated within the body is
may be difficult to perform with needle recording techniques. 250
actually recorded by the instrument to be amplified and subse-
When a needle electrode is used, it must be properly sterilized quently displayed. Skin preparation is of particular importance
or discarded. when recording sensory or mixed nerve potentials because of
If a needle is located relatively close to a nerve, the observed their particularly small size. A commercial electrode paste or
response adequately reflects the number of axons excited (given gel is then placed on the anticipated site of electrode placement
the above noted limitations) because it records the summated to assist in the electrode/skin electrochemical coupling, thus op-
response within the volume conductor. This principle is also timizing the response's presentation to the instrument.
true for recording the induced depolarization in a muscle pro-
vided the needle electrode is located subcutaneously superior to MOTOR NERVE CONDUCTION STUDIES
the fascia overlying the muscle belly. Placing a needle electrode
within the activated muscle, however, does not result in re- Using the above principles of peripheral nervous system stim-
sponses that accurately reflect the summated electrical activity ulation and recording, it is possible to assess only the motor
arising from total muscle depolarization. - This is because
147 225
fibers contained within a peripheral nerve. Recall that there are
the depolarized tissue surrounding the needle's small active sur- no pure motor nerves accessible to the nerve conduction tech-
face preferentially contributes to the amplitude of the recorded niques. Every nerve innervating a muscle not only contains
potential while at the same time acting as a low-pass filter to motor efferents to that muscle, but also multiple sensory or affer-
reduce the high-frequency spike contribution from more distant ent nerves conveying data from the muscle spindles and Golgi
muscle fibers. The subcutaneous needle, especially with a large tendon organs as well as autonomic fibers. Given this stipulation,
surface area exposed, however, is different because it is located we can nevertheless evaluate the action potential characteristics
superior to the muscle but within the body's volume conductor of the motor fibers innervating a particular skeletal muscle pro-
and summates the response from the muscle similar to the sur- vided recording electrodes can be placed on or in the muscle.
face electrode. Also, the onset latency can change with different The technique of performing a motor study on any peripheral
locations within the muscle because the time to depolarization nerve first begins with the proper location of recording electrodes.
The E-l electrode is always located on the muscle's motor

point, i.e., the endplate region. The majority of skeletal muscles
possess one motor point region typically located halfway be-
tween the muscle's origin and insertion. A practical way to
45
locate the motor point is to palpate the origin and insertion of

the muscle and bisect this distance. This rule is particularly A
useful in relatively small muscles without long tendinous exten-
sions. In muscles with particularly long tendons, the motor
point is approximately between the proximal one-third and
distal two-thirds of the distance between the origin and inser-
tion, e.g., tibialis anterior, flexor carpi radialis, flexor carpi ul-
naris, etc. Particularly long muscle like the sartorius may have
two distinct motor point regions. The active surface electrode B
should be secured to the presumed motor point region once it is
identified. An E-2 electrode is then placed just distal to the
tendinous insertion of the muscle presumably on an electrically
inactive region. Recall that a relatively silent electrical site is
preferred for the reference electrode to minimize electrical ac-
tivity that might interfere with the potential recorded from the
Figure 5-14. U l n a r n e r v e C M A P t o A D M . A, A CMAP is o b -
endplate region. The concept of 4 cm of separation for the E-2
tained from the abductor digiti minimi following ulnar nerve wrist acti-
compared to E-l electrode is completely irrelevant for CMAP
vation. N o t e t h e initial positive deflection (arrow) preceding t h e
recordings and applies only to sensory studies. The placement
CMAP's main negative spike. B, Repositioning the E-1 electrode slightly
of an E-2 electrode on the muscle's tendinous insertion consti-
results in a CMAP with an initial negative deflection, suggesting t h a t
tutes a referential montage. When the nerve is excited at some
the E-1 electrode was not over die muscle m o t o r point originally (A).
point proximal to the muscle's motor point at an intensity to
produce a muscle twitch, the instrument's CRT records a bipha-
sic negative-positive compound muscle action potential or may not necessarily directly correlate with the muscle true
CMAP. The CMAP, also referred to as an M response, is the anatomic motor point. To be sure, the anatomic spatial extent of
summated electrical activity arising from the synchronous de- a motor point for any given muscle may not be confined to a
polarization of the muscle fibers innervated by the depolarized very small region. Further, multiple muscles at different layers
nerve. The current intensity should be increased to a supramaxi- innervated by the same nerve may create a relatively large nega-
mal level. Once a supramaximal response is obtained, there are tive sink spatially with the ability to generate a CMAP with an
a number of CMAP parameters that can be recorded in the NCS initial negative deflection over a large muscle region, i.e., a
portion of the electrodiagnostic medicine consultation. large "electrical motor point." A further complicating factor is
The above information suggests that if the E-1 electrode is the influence the reference electrode has on the final potential
properly positioned over the muscle's motor point, one may recorded. Although it is presumed that the reference or E-2 elec-
assume that the CMAP's amplitude will be maximized and re- trode is electrically silent, it is all too frequently very electri-
peatable if the same technique is used by another investigator. cally active. The CMAP for the abductor hallucis in the foot
23 1 8 5
These concepts presume that a muscle's motor point is accu- and the abductor digiti minimi in the hand usually have a
rately localized by obtaining a CMAP with an initial negative bilobed negative peak that has been shown to be comprised of at
deflection. Unfortunately, the issue is somewhat more complex least in part very large muscle far-field potentials (Fig. 5-i 4 ) .
than described to this point. Considerable variation in CMAP The initial portions of these far-field potentials may begin with
parameters can be obtained by different assessments of the same a positive deflection that can in part subtract from the initially
muscle by the same or different investigators. A muscle's
26
negative deflection recorded by the E-l electrode from the
motor point, particularly large muscles but even the hand intrin- muscle it is recording from. As a result, there can be partial can-
sic muscles, can have a location that is not confined to a few cellation of the these two portions of the summated waveforms
square millimeters. Further, multiple muscles in close prox-
340
resulting in a slightly delayed but still initially negative CMAP.
imity may extend the three-dimensional spatial expanse of dif- Further, if the far-field potential happens to fire first because of
ferent electrode locations that can continue to yield a CMAP a shorter neural pathway, the CMAP may display an initial pos-
with an initially negative onset. Additionally, different electrode itive deflection even though the E-l electrode is over the in-
sizes can also affect the measured parameters of the CMAP. 341
tended muscle's motor point (Fig. 5-15A). Repositioning the
The fact that one can position the E-l electrode at multiple loca- reference electrode only can result in the CMAP now appearing
tions over any muscle's presumed motor point and obtain a with an initial negative deflection (Fig. 5-15B). The only alter-
CMAP with an initial negative deflection has led to the concept ation was the location of the reference electrode, strongly sug-
of "false motor points." - Although a CMAP can have an ini-
66 341
gesting that the initial positive deflection was generated by a
tial negative deflection at these different locations, the latency far-field potential that was at least in part eliminated by relocat-
can vary with each electrode position. Comparatively large ing the reference electrode on the same side of the far-field gen-
recording electrodes can be used to minimize this effect; how-331
erator, thereby eliminating it through differential amplification
ever, this does not offer an explanation as to why this phenome- (Fig. 5-15C). Despite these complexities, it is still convention to
non is observed. attempt to obtain a C M A P with an initial negative deflection.
There are a number of factors to consider in attempting to un- The above information, however, should be kept in mind when a
derstand why a muscle can have a relatively large area capable CMAP initial positive deflection is observed (Fig. 5-14A). The
of generating a C M A P with an initial deflection that may or practitioner should always try to move the E-l electrode first in
the waveform between baseline deflections and crossings.

Another way to define a phase is the number of CRT baseline
crossings plus one. For example, in the above typical CMAP,
A the CRT trace describes one baseline crossing. The initial base-
line deflection and final return do not cross, but only deviate
from or return to the baseline. Because only one baseline cross-
ing is described, the entire waveform consists of two phases.
The first phase between the upward CRT deflection and subse-
quent baseline return is the initial negative phase or spike. It is
B considered negative because it is contained within the negative
region of the CRT, i.e., above the baseline. The second or posi-
tive phase is demarcated by the baseline crossing and the final
return of the potential to the isoelectric line. A CMAP may have
an increase in the number of phases if the synchronicity of
C nerve fibers' arrival times at the recording electrode is de-
creased, i.e., abnormally increased temporal dispersion.
Onset Latency. As described above, the optimal CMAP
demonstrates an initial upward or negative deflection in re-
5mV sponse to neural excitation and subsequent depolarization of the
entire muscle. For discussion purposes, we assume a supramax-
5 ms imal stimulation for all responses thereby assuring complete
neural activation for the muscle under investigation. The initial
Figure 5 - / 5 . U l n a r n e r v e C M A P t o F D I . A.An E-l electrode is
baseline departure for the CMAP, onset latency, is negative be-
located over the first dorsal interosseous' mid-belly with an E-2 elec-
cause we have purposefully positioned the E-l recording elec-
t r o d e secured t o the second metacarpophalangeal joint.The CMAP
trode over the muscle's endplate region (Fig. 5-16). Upon
begins with an initial positive deflection. B, Repositioning the E-2 elec-
muscle depolarization, the current is directed inward at the site
t r o d e t o the first metacarpophalangeal joint results in a CMAP with an
of action potential initiation, the motor point, thus constituting a
initial negative deflection. C,An E-1 is located on the radial styloid with
negative current flow. The voltage associated with the negative
an E-2 electrode positioned on the second metacarpophalangeal joint.
current is what is measured by the E-l electrode and is reflected
N o t e that a large potential is recorded by t h e s e two distant electrodes
as a CRT trace deflection in the negative direction. Because the
suggesting that the waveform is primarily a far-field potential.
instrument initiates the sweep at the start of the cathode's depo-
Therefore, t h e waveform recorded in B derives in part from an FDI
larizing pulse, the CMAP's first baseline departure reflects the
near-field potential and interosseous/lumbrical far-field potentials (C).
arrival time at the muscle of the fastest-conducting nerve fibers.
This time is composed of the following: latency of activation
case the muscle's motor point has been missed (Fig. 5-14B); (utilization time), neural conduction, and neuromuscular junc-
however, repositioning the reference electrode may improve the tion transmission. Recall that the onset latency is dependent
situation in a few select cases (Fig. 5-15). upon the instrument's amplifier sensitivity (see Chapter 3). The
higher an amplifier's gain, the earlier a departure from the base-
Parameters line can be detected. This fact emphasizes the necessity of du-
A number of parameters regarding the motor nerve's CMAP plicating individual laboratories' instrument settings prior to
can be measured in an attempt to quantify the response, thus in- using their reference data.
creasing its diagnostic utility. Given most instruments' capabili- Following neural excitation, a CMAP is detected on the CRT
ties to automatically calculate a number of the waveform's screen. The onset latency is measured by placing the instru-
parameters, a substantial amount of data can be gathered in a ment's latency marker on the region of the CMAP identified as
relatively short time. It is important to keep in mind, however, the initial departure from baseline. From the above discussion, it
that the quantification of a waveform's characteristics does not can be appreciated that there are a number of processes that must
necessarily substantiate its clinical utility. We will define a occur prior to the observation of the CMAP. Once the stimulator
number of CMAP parameters that have been shown to have has been activated, there is a certain amount of time required for
clinical relevance or may be of importance in the future. The the transmembrane voltage difference induced by the cathode to
waveform constituents to be discussed are: phase, onset latency, reach the depolarization threshold. The time to peak cathodal
termination latency, nerve conduction velocity, negative wave- negativity is minimal as all commercially available instruments
form duration, amplitude, rise time, area, and stability. The clin- use a square wave pulse that achieves its peak essentially instan-
ical utility of a number of these parameters remains to be fully taneously. Neural propagation can be considered to have been
appreciated and awaits the quantification in both normal sub- initiated once threshold is reached and sodium activation begins.
jects and patients with specific disease states. The duration of time between cathodal current initiation and that
Phase. The optimal CMAP morphology is an immediate necessary to achieve sodium activation with ensuing propagation
negative (upward rise from the CRT's baseline) deflection that is referred to as the latency of activation (utilization time) and
rises to a peak with a subsequent return to the baseline. The has a duration of approximately 0.1 ms. Once the neural im-
CRT trace then continues below the baseline (positive direction) pulses are initiated, action potentials travel along the peripheral
to form a second peak. A gentle sloping return to the baseline nerve to the muscle's neuromuscular junctions. The time neces-
forms the terminal aspect of the CMAP (Fig. 5-16). During the sary to reach the muscle is variable because it is directly depen-
course of describing the typical CMAP, the CRT trace describes dent upon the length of nerve between the point of stimulation
a number of phases. We may define a phase as that portion of and the muscle. Upon reaching the neuromuscular junction, the
complex process of electrochemical induction of an action po- Termination Latency. The CMAP's termination latency
tential from nerve to muscle must occur. The time required to is defined as that portion of the waveform where the gradual
achieve neuromuscular transmission and muscle action poten- return of the CRT trace intersects the baseline (Fig. 5-16).
tial induction with subsequent CMAP appearance on the CRT is Because the terminal portion of the C M A P ' s positive phase
approximately 1.0 ms. The onset latency, therefore, is the result slowly returns to the baseline, it is occasionally somewhat diffi-
of neural activation at the cathode (latency of activation), action cult to decide this aspect of the waveform. That time period rep-
potential conduction along the nerve, and neuromuscular junc- resented between the onset and termination latency defines the
tion transmission. potential's total waveform duration.
Occasionally the CMAP does not demonstrate an abrupt neg- Nerve Conduction Velocity (NCV). One of the more com-
ative deflection but an initial positive deflection describing a monly calculated C M A P parameters is the rate at which a
positive phase of variable amplitude preceding the major nega- neural impulse propagates along the motor fibers stimulated,
tive CMAP spike. This finding indicates that the active elec- i.e., its nerve conduction velocity. The NCV is calculated by
trode detected an initial positive current source prior to dividing the distance over which an action potential travels by
detecting the negative sink (see Chapter 2). When a positive de¬ the time required to cover this distance, i.e., NCV = D (dis-
] flection precedes the negative spike, one can assume that the tance)/t (time). Once the NCV for a patient is known, it can be
active electrode is not located on the muscle's motor point, or a compared to a reference database to assess whether pathology
complex interaction between the electrodes detecting both near- affects the nerve under study. If the calculated NCV is less than
field and far-field potentials is occurring. Fortunately, this situa- normal, one must conclude that the detected slowing is a result
tion can be easily rectified by slightly repositioning the active of demyelination/remyelination, loss of the fastest-conducting
electrode until a negative deflection is achieved indicating a fibers, or both.
recording location over the motor point, or relocating the refer- The accepted manner of determining motor conduction ve-
ence electrode. Infrequently, a muscle may be encountered in locities is to stimulate the nerve at two different locations and
which the motor point simply cannot be found because of measure the distance between these stimulation sites. This dis-
anatomic distortion due to a normal variation or pathology. In tance is then divided by the interval separating the two stimuli.
this instance, one should measure the onset latency to the initial Convention dictates that it is less desirable to stimulate a nerve
baseline departure in the positive direction. Consistency should in only one location and then calculate the NCV based upon the
be applied by also measuring other responses obtained from this separation between the stimulus and recording site. The ratio-
muscle with different stimulation sites to the same initial posi- nale in this instance is to remove the uncertainty and variability
tive deflection. of neuromuscular transmission time and latency of activation. If
' A second problem may arise from the premotor potential, the NCV is to be calculated with only one stimulation site, the
which is a small sensory potential preceding the CMAP to both latency of activation and neuromuscular transmission time must
median and ulnar nerve stimulation at the w r i s t .8197197
This po- be subtracted from the onset latency, i.e., onset latency minus
tential's amplitude usually does not exceed 5 0 - 7 0 pV and, 1.1 ms. By stimulating the nerve in two locations and determin-
therefore, is not a problem unless high amplifier sensitivities are ing the NCV over this distance, however, the neuromuscular
used. The small negative deflection may result in the appear- junction and theoretically that portion of the nerve between the
ance of a slight positive phase preceding the CMAP, thus sug- distal stimulation site and the motor point is precluded from in-
gesting that the recording electrode is not located on the fluencing the NCV Because the latency of activation is assumed
muscle's motor point. If numerous attempts at repositioning the to be the same for all portions of the same nerve, it is of no con-
E-l electrode fail to result in an initially negative onset, the am- sequence in the final NCV determination. Although the above is
plifier sensitivity should be increased to 50 pV/div in order to generally true, a word of caution must be expressed regarding
more closely inspect the CMAP's onset. If a small negative po- the concept of eliminating the portion of the nerve between the
tential is observed, one may conclude that the premotor poten- distal stimulating site and the motor point.
tial is present and one then measures to the CMAP's onset As an illustrative example, let us assume that we are record-
following the premotor potential. Also, the amplifier sensitivity ing from the motor point of the median innervated thenar mus-
may simply be decreased until the response is no longer visible, cles. Our two stimulation sites are just proximal to the distal
thus making CMAP onset measurement much easier. Of course, wrist crease and the antecubital fossa. Suppose a lesion exists at
altering the amplifier's sensitivity may delay the CMAP's onset the wrist just distal to the wrist stimulation site and has resulted
latency, but hopefully this will not extend into the "abnormal" in the loss of nerve fibers capable of conducting at 50
range. If it does, the increased amplifier sensitivity option may meters/second (m/s) and greater, but sparing those conducting
be required. One should realize that the descending portion of 49 m/s or less. By stimulating the nerve just proximal to the
the premotor potential's negative peak can interfere with the lesion site and elbow, it is clear that the fastest fibers measur-
CMAP's negative onset, thus altering the CMAP's true onset la- able are 49 m/s even though fibers of 50 m/s and faster are com-
tency. In this case, consistency is the key and the practitioner pletely intact between the two stimulus sites. Even though the
should decide the most appropriate course of action (see above) 50 xn/s or faster fibers are stimulated, they do not contribute to
and use this technique on all such responses (see Chapter 2). the CMAP secondary to wrist blockage. One should now recog-
Irrespective of the amplifier gain used, the same setting should nize that a lesion not incorporated into the neural segment under
be used for all stimulation sites in order to calculate a valid con- investigation can continue to have quite a profound impact upon
duction velocity. the NCV results.
The onset latency is frequently used for diagnostic purposes. The units placed into the NCV equation are important to con-
Disease processes starting at the distal expanse of the nervous sider. The conventional manner for reporting an NCV result is in
system can usually be detected because of a prolongation in the the units of meters per second (m/s). The electrodiagnostic instru-
CMAP's onset latency. So-called "dying back" neuropathies ment reports time intervals in milliseconds (ms) when the onset
may manifest initially as onset latency prolongation. latencies are measured on the CRT. If one records the distance
between stimulation sites in millimeters (mm), it is simple to the waveform's total duration may be increased, i.e., temporally
quickly convert to m/s. For example, if the time interval for 200 dispersed.
mm is 5 ms, the NCV is 40 m/s (NCV = distance (d)/time (t) = A m p l i t u d e . In addition to onset latency and NCV, the
200 mm/5 ms = (40 mm/ms) x (1 m/1000 mm) x (1000 ms/ls) CMAP's amplitude is one of the more frequently used parame-
= 40 m/s. As long as the numerator is in millimeters and the de- ters in assessing pathology. There are two ways in which to mea-
nominator is in milliseconds, the two numbers associated with sure the CMAP's amplitude: baseline-to-peak and peak-to-peak
the units can simply be divided and m/s incorporated into the (Fig. 5-16). Baseline-to-peak amplitudes are calculated by first
NCV. On the other hand, if the distance is in centimeters (cm), a extending an imaginary baseline with that preceding the CMAP
conversion factor is required to convert cm to mm, i.e., multiply from the onset latency to an intersection with the falling negative
the result by a factor of 10. portion of the negative spike. The distance between the highest
Negative Waveform/Spike. As previously stated, by con- peak of the negative spike and the imaginary baseline immedi-
vention, phases comprising the C M A P above the baseline are ately beneath it represents the baseline-to-negative peak ampli-
considered to be negative. In CMAP studies there is usually one tude of the CMAP. It is this amplitude that is believed to most
major negative phase, often called the CMAP's negative spike accurately represent the total number of axons and their inner-
(Fig. 5-16). This portion of the C M A P is an important compo- vated muscle fibers depolarized. This is because the negative
nent because it contains several parameters of diagnostic utility. sink contained within the motor point for all muscle fibers depo-
The onset latency defines the initiation of the negative spike. larized generates a summated voltage representative of the
That portion of the negative spike opposite the potential's onset number of fibers activated. If fewer fibers are depolarized, a cor-
intersecting an imaginary line drawn through the previous iso- responding drop in voltage results in a lower CMAP magnitude.
electric line defines the CMAP's negative spike duration. The Peak-to-peak amplitude is calculated by first determining the
duration of the negative spike may become prolonged in dis- greatest magnitudes of the major initial negative and subsequent
eases producing differential slowing or temporal dispersion of positive peaks. A horizontal line parallel to the baseline is drawn
the motor fibers' conduction t i m e s . I 8 U 8 2
just intersecting each of the tips of the above-noted major peaks.
Duration. The CMAP's duration can be conceptualized as The separation between these two lines signifies the CMAP's
consisting of two separate components. The first or negative peak-to-peak amplitude. Because the major positive peak arises
spike duration has previously been defined as that amount of from the two terminal source currents and propagates away from
time between the CMAP's onset latency and intersection with the electrode as well as consisting of the CMAP's far-field com-
the baseline (Fig. 5-16). The second duration of possible inter- ponents and possibly those of other muscles, peak-to-peak am-
est is the time between CMAP onset and eventual return to plitudes do not yield a good correlation with the number of
baseline of the terminal positive phase. This is the total poten- axons/muscle fibers depolarized. Also, the positive portion of the
tial's duration and is rarely considered of diagnostic benefit. In CMAP is where there is comparatively more phase cancellation
some demyelinating diseases with nerve fibers affected differently, of the positive phase of the faster fibers with the negative phase
of slower fibers. Therefore, this region of the CMAP bears less
one-to-one correspondence between amplitude and number of
fibers depolarized. Practically, however, if good reference data
are available, either technique suffices for routine studies.
One may note that the CMAPs generated by stimulation sites
further removed from the recording site are somewhat reduced
compared to sites closer to the recording electrode. This obser-
vation arises because of the temporal dispersion normally pre-
sent in motor fibers. There is approximately a 13 m/s difference
between the fastest- and slowest-conducting fibers in normal
motor nerves. - Over increasing distances, the separation be-
72 73
tween these population of fibers increases. As a result, the ter-

minal positive phase of the fastest fibers will increasingly
overlap the negative spike of the slower fibers. Increasing
amounts of phase cancellation produce a potential with a reduc-
tion in the baseline-to-peak and peak-to-peak amplitudes. The
minimal increase in temporal dispersion is relatively offset by
the naturally long duration of the negative spike of muscle
2000/zV fibers. The net effect is a mild but noticeable amplitude reduc-
2ms tion. In pathologic conditions, it may be possible for the above
noted difference between the fastest and slowest conduction ve-
Figure 5-16. T h e n a r m u s c l e C M A P . Example of a typical CMAP locities to increase, thus resulting in a pathologic reduction in
recorded from the thenar eminence following median nerve excitation CMAP amplitude over a commonly used distance. Also, if there
at the wrist.The time represented by the segment A-B is referred to is greater than approximately 5 0 % reduction in C M A P for a
as the rise time, while A - C is t h e duration of the negative spike. proximal compared to distal stimulation site, some degree of
Segment A-E represents the duration of the total potential.The ampli- conduction block of neural propagation may exist. By incre-
tude of A - B is the baseline-to-peak magnitude of the potential, while mentally moving the stimulator, it is sometimes possible to lo-
B-D is t h e peak-to-peak amplitude.The portions of the CMAP be- calize the site of conduction block within a few centimeters by
tween A - C and C-E each constitute one phase of this biphasic poten- monitoring for the drop in amplitude.
tial. T h e latency of point A is t h e o n s e t latency, while point B Rise Time. As previously noted, if the active electrode is lo-
represents the peak latency. cated directly over a muscle's motor point, the initial departure
from the baseline is in the negative direction. The time required antidromic nerve conduction recording. If the nerve were stimu-
for the CRT trace to describe the peak of the CMAP's negative lated at the wrist and a more proximal forearm recording site
spike, peak latency minus onset latency, is referred to as the rise chosen, induced action potentials would be propagating in the
t i m e (Fig. 5-16). In general, the rise time's slope is proportional physiologic direction with respect to the recording electrode.
to the distance between the action potential's source and the This type of recording is called an o r t h o d r o m i c conduction
recording electrode. If the recording electrode is not located on technique. In the previously described motor nerve conduction
the motor point and an initial positive deflection is observed, the studies, stimulating the nerve and recording from a muscle is an
rise time in this instance is from the first positive deflection's orthodromic technique because the motor impulses are traveling
peak to the negative spike's peak. The reversal of potential from toward the muscle as they would physiologically.
positive to negative signifies that the negative current sink has Let us assume we are next going to stimulate the median nerve
reached the E-l recording electrode. sensory fibers in one of the first three digits and record from the
Area. In addition to amplitude, area is an alternative way to mixed median nerve at the wrist. This type of setup constitutes
estimate the number of axons/muscle fibers depolarized. The an orthodromic conduction study with a SNAP being recorded
term area as it is commonly used in electrodiagnostic medicine from a mixed nerve. Additionally, the median nerve at the wrist
actually refers to the area under the negative spike of the CMAP. can be stimulated while recording from one of the first three
Commercially available instruments can quite readily calculate digits in the hand (Fig. 5-17). We are now stimulating the mixed
the CMAP's negative spike area. Studies comparing amplitude median nerve at the wrist and recording only from an anatomic
and area for diagnostic purposes with respect to axonal loss and region where sensory fibers originating from the excited nerve
prognosis found comparable results. It is important to recog-
361
are distributed. A pure sensory response occurs because only the
nize that the area of CMAPs obtained from both proximal and sensory fibers are located on the digit. In this instance, an an-
distal stimulation sites is not equal. This suggests that area is
182
tidromic sensory nerve conduction study is performed.
subject to temporal dispersion effects producing phase cancella- Antidromic and orthodromic techniques are equivalent with
tion just as previously described for amplitude. As with ampli- respect to onset and peak latency but not amplitude. Although 30
tude, the area under the positive portion of the CMAP is several studies suggested that the latency differed depending
potentially unreliable in estimating axonal loss when phase can- upon the direction of propagation, hand temperature and record-
cellation effects are prominent, as in a demyelinating neuropa- ing electrode separation were not controlled. A subse-
47 2 4 0 3 2 6
thy causing differential slowing (some fibers altered more than quent study documented that when these factors were
others) of affected fibers. 58
controlled, there was no significant difference between these
S t a b i l i t y . In normal nerve and muscle, the recorded CMAP two techniques with respect to latency The amplitude of digi-
4 9
should maintain the same amplitude and morphology with each tal antidromic responses, when recorded with surface elec-
successive stimulation. The constancy of the potential from one trodes, is generally larger than orthodromic potentials because
neural excitation to the next refers to the CMAP's stability. If the recording electrodes are closer to the subcutaneous neural
the nerve is repetitively stimulated at high rates (e.g., 50 Hz), a tissue. In orthodromic studies, the recording electrode is posi-
less than 50% increase in amplitude may be noted and is re- tioned over a proximally located nerve with respect to the stim-
ferred to as p s e u d o f a c i l i t a t i o n . The exact mechanism pro-
254
ulation site and thus the nerve is typically lying somewhat
ducing this phenomenon is unknown but is postulated to arise deeper in the tissue, which reduces the recorded response. 30
secondary to increased synchronization of motor unit firing. If

sequential peripheral nerve stimulation results in a CMAP
decrement or amplitude variability, one should suspect faulty B
neuromuscular junction transmission or inadequately secured
electrodes combined with movement artifact.
SENSORY NERVE C O N D U C T I O N STUDIES
In addition to investigating the conduction properties of

motor nerves, one can also study action potential propagation in
pure sensory nerves. It is possible to record a pure sensory nerve 20/xV
action potential (SNAP) by stimulating (1) a pure sensory nerve
and recording from this nerve, (2) a pure sensory nerve and
recording from a mixed (motor and sensory fibers) nerve, and
(3) a mixed nerve and recording from a pure sensory nerve.
These three possibilities are best conceptualized by discussing
several examples. The superficial radial nerve on the dorsum of
the forearm can be activated proximal to the wrist while.a
recording electrode is located over this nerve as it crosses the
extensor pollicis longus tendon at the wrist. In this instance, we Figure 5 - / 7 . Antidromic median sensory nerve action po-
are activating and recording the same pure sensory nerve. When tential (SNAP) recorded from the third digitThe same descrip-
stimulating a sensory nerve proximal to the recording site and tions noted in figure 5-16 for the potential's various segments apply t o
recording distally, the induced action potential is propagating in the above recorded SNAP. In t h e upper trace a bipolar recording mon-
the opposite direction the nerve conducts impulses physiologi- tage (both recording electrodes on the third digit) is shown.The lower
cally, i.e., peripheral to central. This type of neural conduction, trace depicts the result of relocating the reference (E-2) electrode t o
opposite physiologic propagation, is referred to as a n t i d r o m i c the fifth digit resulting in a referential recording montage and an ini-
p r o p a g a t i o n . Therefore, the above-described technique is an tially positive triphasic SNAP.
When using needle recordings, however, orthodromic potentials 5-17). Because of the smaller magnitude of the SNAP, an ampli-
are comparatively larger than antidromic responses because the fier sensitivity of 10 or 20 pV/div may be required compared to
needle can only be located near a few digital fibers in an- the 500 pV/div or greater used for CMAPs. This results in more
tidromic studies, while it can be placed close to more fibers at a baseline noise, which at times makes it somewhat difficult to
proximal region in orthodromic investigations. Specifically, accurately determine the waveform's departure from baseline.
placing a ring electrode around a digit excites all nerve fibers Averaging 10-20 responses may remove the baseline artifact
near the ring cathode, which can be recorded with a needle situ- and produce an acceptably sharp potential take-off. Historically,
ated next to the nerve at the wrist. Exciting the nerve at the wrist the aforementioned baseline/instrument noise and lack of so-
(antidromic technique), however, allows one to place the needle phisticated averagers necessitated measuring SNAPs to the
next to only a few fibers in the subcutaneous tissue of the digit. peak. The peak latency is much easier to define in noisy record-
As with all techniques, antidromic and orthodromic sensory ings. As instrumentation improved the ability to record optimal
recordings both have advantages and disadvantages. Antidromic SNAPs even without averaging, onset latency measurement in-
sensory responses in the hand are typically larger when using creased in popularity but peak latency remains because of previ-
surface electrodes than orthodromic potentials (see above). The ous convention. Of course, the peak latency does not reflect the
disadvantage of the antidromic recording is that it can occasion- fastest propagating axon velocities. To some extent, the onset
ally record muscle artifacts, which may interfere with the de- also is not a completely accurate descriptor of the fastest veloci-
sired sensory potential. This volume-conducted muscle ties since it is an artifact of the recording electrodes and differ-
response may originate from the lumbrical muscles as the ring ential amplification. The fastest recording fibers are most
electrodes are located near their insertion. The inexperienced accurately measured to the inflection of the initial positive
practitioner may also mistake the above noted muscle response phase in referential recordings (triphasic SNAP), which signi-
for a sensory potential. Muscle artifact is usually not a problem fies the arrival of the negative sink at the E-l recording elec-
when using orthodromic techniques as only a sensory nerve is trode. Negative onset and peak latencies are now routinely
excited, thus obviating a concomitant activation of motor measured with standardized techniques. If a triphasic SNAP is
nerves. In antidromic studies, the motor artifact is usually not a detected, the onset latency is defined as the peak of the initial
problem but can be minimized by moving the E-l electrode positive deflection. As long as these techniques are used in
slightly more distal on the digit or having the patient abduct the defining pathology and the various recording artifacts are rec-
digits. Also, the motor response is usually delayed with respect ognized, it is acceptable to measure onset and peak latency. This
to the sensory response and its negative peak potential is con- is in contrast to a CMAP where one attempts to always record
siderably longer in duration than the sensory response. It is rec- from the muscle's motor point and hence generate an initial
ommended that the practitioner become familiar with both types negative deflection.
of techniques and employ them judiciously when appropriate Termination Latency. The termination latency of the
circumstances arise. SNAP is calculated in the same manner as that for CMAPs. As
The basic principles of nerve conduction stimulation and for CMAPs, this parameter is seldom used at the present time
recording are similar for motor and sensory studies. There are a for defining pathologic responses.
few pertinent differences necessary to point out prior to per- Nerve Conduction Velocity. Calculating nerve conduction
forming either study in patients. We will review the same as- velocities for sensory nerves is slightly different than the tech-
pects for the SNAP components as for the CMAP with nique employed for motor nerves. The sensory nerve does not
emphasis on the aspects unique to the sensory nerve conduction contain a neuromuscular junction but the latency of activation is
study. still present. When performing sensory NCVs, either one or two
stimulation sites can be used. If one stimulus site is used, then
Parameters the distance between the stimulus and active recording electrode
Phase. The typical SNAP morphology with a bipolar record- is divided by the onset latency minus 0.1 ms (latency of activa-
ing montage is biphasic with an initial negative spike followed tion). When two stimulus sites are used, the latency of activa-
by a positive spike (Fig. 5-17). In the case of the SNAP there is tion can be ignored as it is common to both. Because a SNAP is
no motor point from which to record and it is reasonable to often recorded from the digits where the nerve fibers are quite
question why it is biphasic and not triphasic. Indeed, one would small, it is important to realize that the distal NCV can be
anticipate that the initial positive source current should be de- slower than that calculated for a proximal segment of the same
tected prior to the negative sink, thereby producing an initially nerve. This is because nerve fibers taper as they travel distally
positive triphasic SNAP. However, this is not what is observed and NCV is proportional to the nerve's diameter. This problem
in most digital nerve studies when ring and bar electrodes are is also present for motor fibers, but is less prominent as the
used. If you recall from volume conductor theory (see Chapter distal several centimeters of the nerve are common to all stimu-
2), when two electrodes are placed relatively close together on lation sites and thereby not directly measured. When an NCV is
the same excited nerve, the propagating action potential is going measured, one must realize that the end result is a reflection of
to be detected at both electrodes. This is because the duration of the "average" NOV over the segment studied. The proximal
the SNAP is long enough to reach the E-2 electrode before it is portion of the nerve may be conducting slightly faster than the
completely extinguished at the E-l electrode. The result is a distal segment, but the entire nerve segment is what is measured
cancellation of the first positive phase with an apparent initial for the calculation, thus reflecting proximal (faster) and distal
negative deflection thereby generating a biphasic potential. 30
(slower) aspects of the nerve.
Should the distance between the two recording electrodes be in- When calculating sensory nerve NCVs, it is only appropriate
creased, or the E-2 electrode rotated off the nerve fibers, the ex- to use the onset and not peak latencies. In this way, the fastest-
pected triphasic morphology of the SNAP reappears (Fig. 5-17). conducting fibers are detected at both the proximal and distal
Onset Latency. Just as for CMAP studies, the onset latency sites of stimulation. Peak latencies do not reflect the fastest-con-
is measured to the initial negative deflection of the SNAP (Fig. ducting fibers and are subject to uneven changes over distance
because of temporal dispersion effects (see below and Chapter potentials. The end result can be a significant amount of phase
2). Peak latencies, however, can be used in isolation of onset la- cancellation, generating a corresponding reduction in the
tencies and NCVs if reference data are presented for individual SNAP'S amplitude.
nerves and standard distances. Rise Time. In biphasic SNAP responses the rise time is the
Negative Waveform/Spike. Because the SNAP may be time period from the potential's onset latency to negative peak
either biphasic or triphasic, the negative waveform portion of latency (Fig. 5-17). Similarly, in triphasic SNAPs, the rise time
this potential depends on the waveform's morphology. The is from the initial positive to subsequent negative peaks. T h e
SNAP'S negative spike is that portion of the waveform between rise time is particularly important when performing near-nerve
the onset latency (initial negative or positive-to-negative deflec- recordings, i.e., attempting to locate a needle active recording
tion) and where the negative spike intersects the baseline to electrode as close as possible to the nerve. In this technique, the
form the terminal positive phase (Fig. 5-17). needle electrode is placed in proximity to the nerve and moved
Duration. The duration of the SNAP may be either the nega- in small increments while continuously stimulating the nerve.
tive spike's duration or the entire waveform's duration. The neg- The optimal position for the recording electrode is determined
ative spike duration is calculated as noted above, i.e., from the by the rise time. The shortest rise time signifies that the needle
SNAP'S onset to the intersection with the baseline. This para- is next to the nerve. If the rise time increases, then movement is
meter is believed to be rather sensitive in assessing pathology, occurring in the direction away from the nerve instead of toward
but little reference data are available. Further work is needed in it. When using a near-nerve technique, the potential's morphol-
this area to better define negative spike duration with respect to ogy is typically triphasic because the reference electrode is pur-
its diagnostic utility. The total potential's duration is not rou- posefully located at some distance from the nerve forming a
tinely used for diagnostic purposes. referential recording montage. Coincidentally, a maximal p o -
Amplitude. The SNAP'S amplitude is usually measured tential amplitude usually corresponds to the shortest rise time.
from the initial baseline deflection or positive peak to subse- Area. The area of the SNAP is characteristically determined
quent negative peak, i.e., baseline-to-peak or peak-to-peak. This by measuring that portion of the waveform demarcated by the
description may be somewhat confusing depending upon the negative spike. Area, like amplitude, is representative of the
SNAP'S morphology and recording technique. If a bipolar number of axons depolarized. The area is not immune from the
recording technique is used where both electrodes are located effects of temporal dispersion arising from increases in distance
on the nerve and are relatively close together, an initial negative between the site of stimulation and recording. Care must be ex-
1 deflection typically ensues. As already described for CMAPs, a ercised when attempting to define axonal loss based either on
more accurate reflection of the number of axons associated with area or amplitude for proximal sensory studies because of tem-
neural depolarization is the negative spike's magnitude without poral dispersion effects.
considering the terminal positive peak. In motor fibers, collat- Stability. Just as for CMAPs, the SNAP should remain es-
eral sprouting of terminal axons following denervation can rein- sentially unchanged from one stimulus to the next. A slight al-
nervate newly denervated muscle fibers and increase the motor teration in the potential may occur secondary to baseline
units' muscle fiber content, thereby decreasing the accuracy of wavering because of the high amplifier sensitivities required to
motor amplitudes in defining axonal loss. The process of collat- perform sensory studies. Should clearly recognizable changes
eral sprouting diminishes the correlation of the CMAP's ampli- appear between successive stimuli, electrode movement or
tude with the number of axons excited over time. Sensory faulty electrode leads should be suspected, or intrinsic neural
nerves, however, do not sprout collaterals to replace lost fibers pathology causing drop-out of fibers or pathologic degrees of
but must regrow from the lesion site. Thus, a decreased ampli- temporal dispersion.
tude comparatively more accurately reflects the axonal content
with respect to the number of functional nerve fibers. Therefore, COMPOUND NERVE ACTION POTENTIALS
when considering the SNAP'S amplitude, one should really
measure it from initial negative onset to negative spike peak. If In addition to investigating pure motor and sensory nerve
an initial positive potential is present, as would be encountered fibers, it is also possible to study the action potentials from both
in referential recordings, or widely spaced electrode's on the sensory and motor nerve fibers simultaneously, i.e., mixed
same nerve, then the initial positive to subsequent negative peak nerves. - -
27 39
When one attempts to record action potentials
471141,5
most accurately reflects the number of axons excited. arising directly from a mixed nerve, the observed potential is
The magnitude of the SNAP'S amplitude demonstrates a known as a compound nerve action potential, mixed nerve
marked diminution from distal to proximal stimulation sites. action potential, or occasionally compound mixed nerve
This is the same phenomenon previously described for motor action potential. This technique is relatively simple and can be
nerves but is considerably more pronounced in sensory fibers. readily performed on most nerves. For example, suppose we
The explanation of the significant amplitude differences is pri- wish to record the compound nerve action potential arising from
marily twofold. First, the duration of the SNAP'S negative spike the median nerve in the forearm. One method is to record from
is shorter than that for the CMAP. This means there is less toler- the median nerve at the antecubital fossa while stimulating this
ance for asynchronous summation at the recording electrode of nerve at the wrist. In this instance, the detected potential at the
sequentially arriving action potentials before phase cancellation elbow region consists of orthodromically conducting sensory
becomes a prominent factor (see Chapter 2). Secondly, the dis- fibers and antidromically conducting motor fibers. A readily de-
persion between the fastest and slowest sensory nerves is twice tectable potential is observed. The latency is usually measured
that for motor nerves at about 25 m/s. - When a nerve is ex-
72 73
to the potential's initial negative departure from the isoelectric
cited proximally, the increased distance traveled allows the tem- line or first positive deflection if it is triphasic, thereby repre-
poral dispersion between fast and slow fibers to become senting the fastest fibers contained in the nerve. It is also possi-
prominent, which is manifested at the recording electrode as a ble to record from the wrist following stimulation at the elbow.
comparably asynchronous arrival of the multiple SNAP action The obvious difficulty commonly encountered with proximal
stimulation is motor artifact arising from the activated forearm cm is to provide enough electrode separation to maximize the
muscles contaminating the mixed nerve response. Distal stimu- SNAP'S amplitude by ensuring that the potential's peak has
lation results in considerably less muscle artifact. passed by the E-l electrode before the initial aspect has reached
When conduction velocities are attempted for compound the E-2 electrode. If the intereiectrode separation is too small,
nerve action potentials, one stimulus site is required. Recall that then the leading portion of the action potential detected by the
we are recording a response directly from a nerve without an in- E-2 electrode is subtracted from the E-l electrode. Should the
tervening neuromuscular junction. As a result, the distance can peak of the action potential be under the E-l electrode at the
be directly divided by the response's latency after 0.1 ms (la- same time the leading portion is under the E-2 electrode, the
tency of activation) has been subtracted. Compound nerve action amplitude of the negative spike is reduced. The magnitude of
potential data may be of benefit in the early diagnosis of periph- reduction depends upon how close the two electrodes are to
eral entrapment syndromes. Despite a few promising reports,
312
each other. Four centimeters is chosen because the time neces-
there are a few limitations regarding this technique. At the pre- sary for the typical SNAP'S peak to maximize is about 0.8 ms. If
sent time there are small numbers of normal values for only a the action potential's fastest-conducting fibers are propagating
few nerves. Additionally, this technique has not been widely
39
at 50 m/s, then they have traveled 4 cm in the time necessary to
applied to multiple diseases in order to fully characterize its clin- reach the negative spike's peak (NCV = 50 m/s = distance/0.8
ical utility. The sensitivity and specificity of the compound nerve ms = 40 mm or 4.0 cm). Of course, it is not always feasible to
action potential must also be compared to pure sensory and achieve 4 cm of separation, e.g., short fingers in an antidromic
motor studies in pathologic situations. There have been sugges- recording. Also, it is important to remember that most elec-
tions that it is inappropriate to equate compound nerve action po- trodes imbedded in a plastic bar are not separated by 4 cm and
tential velocities with sensory velocities. One cannot assume that can result in artificially low amplitudes. The consequence of
the sensory fibers are preferentially the first and fastest fibers ex- closely spaced electrodes is that the observation of small SNAP
cited simply because of the size principle. This cautionary note responses may cause one to erroneously conclude that axonal
is necessary because a study comparing sensory and motor loss is the reason for the amplitude reduction.
thresholds revealed that median nerve sensory thresholds were There is no optimal separation distance in motor studies. The
lower than motor fibers in only 56% of 55 digital nerves tested major consideration in studying C M A P ' s is to locate the E-l
and 40% of 37 nerves examined in the axilla. 30
electrode on the muscle's motor point while the reference elec-
trode is placed in an electrically "silent" area, i.e., on the
muscle's tendinous insertion. If the reference electrode is situ-
FACTORS AFFECTING NERVE ated on muscle tissue innervated by the excited nerve, the
CONDUCTION STUDIES CMAP's amplitude will be reduced for reasons similar to those
noted above for the SNAP. Again, a low CMAP amplitude may
Prior to discussing the more commonly performed nerve con- give the appearance of axonal or muscle fiber loss when it is
duction techniques, a number of issues regarding technical and merely an artifact of too closely spaced electrodes.
physiologic factors that can affect the NCS should be consid- Filters. When the low-frequency (high-pass) filter is ele-
ered. It is necessary for the practitioner to be cognizant of the vated above the frequencies contained in the waveform, signifi-
various ways in which the measured SNAP and CMAP parame- cant alterations in the potential's frequently measured
ters can be altered by either the instrument, electrodes, or pa- parameters can occur. Specifically, for both SNAPs and CMAPs
tient. The reason these issues are important is because the amplitude decrease, peak latencies shorten, an extra phase
false-positive and occasionally false-negative results can arise appears, negative phase duration shortens, and the total poten-
predisposing the unsuspecting clinician to an erroneous diagno- tial duration decreases, but onset latency is unaffected. - - On
78 79 80
sis. Although the actual technical performance of the NCS is the other hand, lowering the high-frequency (low-pass) filter
relatively easy, the underlying physiologic mechanisms result- below the frequencies contained in the waveform results in a re-
ing in the detection of an impulse are by no means simple. duction in amplitude, increases in both the onset and peak laten-
Careful study, however, allows one to readily comprehend these cies, and an increase in the negative spike's duration. Consult
principles and avoid diagnostic errors. Chapter 3 for a complete description of these changes and why
they occur.
INSTRUMENTATION FACTORS Amplification. Increasing the gain or sensitivity of the in-
strument allows one to detect deflections from the baseline ear-
One of the essential components of the electrodiagnostic lier than at comparatively lower amplifications. This has
medicine consultation is the instrument with which the patient's obvious implications regarding onset latency measurements.
electrophysiologic responses are recorded. As previously stated, Sweep Speed. When recording both SNAPs and CMAPs it
it is imperative for the practitioner to be fully aware of the in- is good practice to use a sweep speed that allows the potential to
strument's capabilities and shortcomings. A poor understanding appear approximately in the middle of the CRT. This permits
of how the instrument processes the recorded signal can lead to the instrument to devote an adequate number of resolution
data misinterpretation. One of the most important concepts to points to avoid morphologic distortions secondary to subopti-
remember when considering instrumentation effects is to use mal sampling frequencies. Additionally, if the sweep speed is
the same parameters at which the reference database was ac- too low (e.g., 20 ms/div vs. 1 ms/div), the waveform is com-
quired. Only a few of the more common instrumentation factors pressed into the stimulus artifact that can produce possible
are noted and the reader is urged to read the chapter addressing waveform distortions.
instrumentation in detail (see Chapter 3). Averaging. All commercially available instruments have the
Electrode Separation. When considering SNAPs, it is rec- capability of averaging multiple stimuli in order to improve the
ommended that the active and reference intereiectrode separation signal-to-noise ratio. When axonal loss results in potentials of
should meet or exceed 4 cm. - This recommendation of 4
78 80 84
relatively small size, it is possible to erroneously consider them
as part of the baseline noise, particularly if high amplifier gains placed 8 cm proximal to and directed toward the active record-
are used. If these small potentials are dismissed as artifact, the ing electrode at the wrist over the median nerve and 20 cm prox-
practitioner may be missing valuable diagnostic information. imal to the wrist site in the antecubital fossa. The separation
Averaging multiple response trials allows one to "extract" the between the anode and cathode is 2.0 cm. A nerve conduction
diminished potential from the surrounding background noise velocity of 50 m/s is calculated with a distal CMAP onset la-
that may be of equal magnitude as the response. As a result, it is tency of 3.0 ms. Suppose we now wish to repeat the study but
good practice when even the slightest doubt exists as to the inadvertently reverse the location of the cathode and anode at
presence of a response to employ the instrument's averager the wrist except with the anode now 8 cm proximal to the active
prior to concluding a potential is absent. recording electrode and hence the cathode 10 cm from E-l. The
Stimulators. Neural action potential generation occurs nerve conduction velocity is an intrinsic property of the nerve
about the region surrounding the cathode. As previously stated, and is not going to change irrespective of what we do with
convention dictates that a supramaximal stimulus be used at all recording or stimulating electrodes, thereby remaining 50 m/s.
times when eliciting an action potential in the peripheral ner- Originally the time required to traverse the 20 cm is 4.0 ms (50
vous system for NCV measurements. A byproduct of the stimu- m/s = 200 mm/time; time = 4.0 ms). By reversing the cathode
lation process may interfere with detection of the CMAP and and anode the interstimulus distance is now only 18 cm be-
particularly the SNAP or compound nerve action potential. cause the cathode is 2.0 cm (cathode-anode distance) closer to
Impulse delivery from the stimulator is associated with an ini- the antecubital fossa. The distal onset latency is subsequently
tial pulse of current/voltage from the cathode and anode, which prolonged by the time required for the action potential to travel
is conveyed instantaneously throughout that portion of the body the extra 2.0 cm between the cathode and anode, i.e., 0.4 ms.
enveloping the nerve, i.e., stimulus artifact. Because the Because we are hypothetically unaware of cathode/anode re-
recording electrodes are located some distance from the source versal, the interstimulus distance is still measured at 20 cm
of the stimulus artifact, it can be regarded as a far-field potential when in reality it is 18 cm. The actual interstimulus interval is
(see Chapter 2). A large stimulus artifact may have a sufficient no longer 4.0 ms but 3.6 ms because of the stimulus reversal
duration so as to coincide with the induced neural action poten- (4.0 ms - 0.4 ms = 3.6 ms). The velocity is now increased to 55
tial. In this instance, the response of interest is either masked m/s because our time has decreased (NCV = 200 mm (erro-
completely or distorted to such an extent that latency and ampli- neously measured)/3.6 ms = 55 m/s). If this mistake is made
tude are of questionable value. There are a number of important initially as the only calculation, our NCV may be erroneously
actions the practitioner should take in order to minimize stimu- high.
lus artifact. Reversing the cathode and anode also raises the possibility of
Simply, one wishes to have the stimulus artifact appear at the anode hyperpolarizing the nerve thus preventing conduction
both recording electrodes simultaneously and with the same through this segment of nerve, i.e., anodal block. The produc-
magnitude. If the stimulus artifact appears equally at the E-l tion of anodal block is theoretically possible and has been suc-
and E-2 electrodes, a substantial portion of the artifact will be cessfully performed in vitro. Anodal block does not occur in
267
eliminated through differential amplification and common routine clinical practice and can be ignored. By performing the
mode rejection. There are a number of ways to ensure this above noted stimulator reversal one can readily demonstrate
occurs. The impedance under both electrodes should be re- that the latency prolongation is exactly what one would predict
duced and of a similar value with respect to the input imped- based on distance effects of cathode/anode reversal without
ance of the amplifier thus ensuring a similar recording. Also, anodal block either blocking or slowing neural conduction.
both electrodes should be of the same type of material so the A potential problem associated with neural excitation is stim-
signal is presented similarly at the amplifiers. Aberrant con- ulus spread either longitudinally along the nerve beyond the site
duction pathways, surface collections of perspiration, or elec- of stimulation or laterally toward other nerves. Recall that a
trode paste can preferentially convey the stimulus artifact to supramaximal excitation as previously defined is considered op-
one of the two electrodes and should be avoided. A ground timal in exciting peripheral nerves. Increasing the intensity of
electrode located between the active recording electrode and stimulation can reduce the localization of neural depolarization.
stimulator also helps suppress some of the artifact. In addition A 4 times supramaximal stimulus intensity produces a decrease
to these precautions, one of the most useful ways to reduce in conduction time by approximately 0.75 m s . Also, increas-
262
stimulus artifact is to rotate the anode about the cathode, i.e., ing the stimulus intensity to just greater than supramaximal can
the cathode remains on the nerve while only the anode is repo- shift the site of stimulus by 5 mm in older persons and 3 mm in
sitioned. This simple procedure helps to align the isopotential younger individuals. This phenomenon is most likely sec-
30
lines of voltage generated between the cathode and anode to ondary to the depolarizing pulse associated with the cathode ex-
present in a similar manner at both recording electrodes assist- tending within the body's volume conductor beyond the region
ing in common mode rejection. where the cathode contacts the patient. Nerves affected by
In addition to stimulus artifact, the practitioner should visu- pathology can require significantly more current to achieve de-
ally inspect the orientation of the cathode with respect to the E- polarization, thereby increasing the possibility of stimulus
1 recording electrode. The cathode is usually positioned such spread and producing erroneously short latencies. Because the
that it is directed toward the E-l recording electrode. When per- body's interstitial milieu is a relatively good conductor of cur-
forming routine nerve conduction studies, there are only two ex- rent flow, excess stimulus intensities can produce coactivation
ceptions to this rule, i.e., F waves and H-reflexes (see below). A of neighboring nerves to those intentionally excited. A typical
simple example will demonstrate the consequences of reversing example is median nerve stimulation at the wrist. If the current
the cathode and anode in relation to the active recording elec- intensity is too strong, it is relatively easy to activate the ulnar
trode. - "
7 8 8 0
Let us suppose that we have located an E-l
6 3 3 3
nerve. Should the median nerve conduction actually be patho-
recording electrode over the motor point of the abductor pollicis logically affected across the wrist requiring atypical current in-
brevis in the hand. For discussion purposes, the cathode is tensities for stimulation, and coincidentally, the electrode is also
conduction velocity are the distance between stimulation sites

and the onset latencies of the proximal and distal response. In 294
this discussion only the motor NCV is explored but similar

comments can be made regarding sensory nerves.
Distance Between Stimulation Sites. In examining NCVs on
the same subject separated by a time interval of 1 week, approx-
imately 4 - 6 m/s difference can be calculated in the NCV. When
calculating the NCV for the same distance and nerve, different
practitioners produce differences between 2.0 to 2.7 m / s . A 146
second set of studies that performed NCVs on two separate days

of variable separation results in a standard deviation 5.9 m / s . 246
In children, similar investigations demonstrate differences be-

tween successive days of up to 4 m/s. It is possible to develop a
simple formula in an attempt to quantify the amount of experi-
mental and biologic error typically present in NCVs. One can
CONDUCTION VELOCITY ( M / S E C j
begin by first considering the basic nerve conduction equation:
C (nerve conduction velocity) = D (distance)/t (time). By 231
Figure 5 - / 8 . Conduction velocity error. A number of error taking the natural logarithm of this equation one can place it in
curves demonstrating t h e conduction velocity error ( A C ) as a func- a format that can be readily differentiated, i.e., In C = In D - In t.
tion of conduction velocity at multiple distances. (From Maynard FM, Differentiating both sides of the equation results in: A (In C) = A
Stolov W C : Experimental error in determination of nerve conduction (In D) - A (In t). Because the derivative of (In x) is dx/x, we then
velocity.Arch Phys Med Rehabil 1972;53:362-372, with permission.) arrive at: AC/C = AD/D - At/t. For convenience, the minus sign
in the above equation is converted to a plus sign as At can be
either positive or negative and the variables are rearranged. 231
spatially over the motor point of the adductor pollicis as well as The result is: AC = (AD/D + At/t) C. In this equation AC repre-
the abductor pollicis brevis, coactivation of the ulnar nerve can sents the experimental error for a single NCV determination and
yield a false-negative onset latency. In this case a misdiagnosis is clearly dependent upon the error in measurement for both dis-
might arise because the ulnar nerve's activation of the adductor tance (AD) and latency (At), and inversely dependent upon the
pollicis was recorded instead of the slowed conduction to the magnitude of the distance and conduction time. The error in
median nerve's abductor pollicis brevis muscle. Should stimu- time measurement can arise from stimulus strength, identifica-
lus spread be suspected, it is best to use a needle cathode and tion of the CMAP's onset, amplifier sensitivity, and subtraction
nearby surface or subcutaneous anode thus necessitating less errors. Distance measurements may result from limb position,
current, and decreasing the possibility of coactivating nerves or anatomic course of the nerve, cathode localization, skin-subcu-
excessive current spread along the nerve. Neural stimulation taneous movement, skin movement during measuring, and tape
with either a needle or surface cathode at the same location pro- measure reader error. The above equation attempts to better
duces no difference in the final NCV calculated. 334
define the time and distance errors in order to avoid some of
Measurement. When performing nerve conduction studies, these errors. Applying this equation in some simple examples
one must measure a number of parameters to arrive at the infor- may help the practitioner conceptualize the diagnostic signifi-
mation necessary for an appropriate diagnosis. Two of the most cance of time and distance errors.
fundamental parameters required in order to calculate a nerve Onset Latencies of Proximal and Distal Responses. If the
ulnar nerve is stimulated at the wrist and elbow, distal and prox-
imal onset latencies can be recorded for the abductor digiti
minimi's CMAP. At an amplifier sensitivity of 1000 pV/div
231
the proximal and distal latencies are 5.58 ± 0.21 ms and 2.42 ±
0.15 ms for a conduction time of 3.17 ± 0.25 ms, respectively,
over a distance of 222.2 ± 1.8 mm. The manner in which the la-
tencies were measured was by 20 skilled electromyographers
reviewing a photograph of the trace and counting an electrical
time marker. Digital latency markers similar to those presently
available on instruments were not used because the technology
did not exist at the time of the study. Given these differences,
the principles of time error measurements continue to be valid.
Because it is common to use two standard deviations, the exper-
imental error for conduction time (At) and distance (AD) is 0.5
ms and 3.6 mm for the nerve under investigation at an amplifier
setting of 1000 pV/div. The experimental error for conduction,
CONDUCTION VELOCITY ( M / S E C )
velocity for various conduction velocities and distances or con-
duction times can be calculated, thus producing a table and
Figure 5 - / 9 . Conduction velocity error. A set of error curves family of curves (Tables 5-1 and 5-2; Figures 5-18 and 5-19).
revealing t h e conduction velocity error ( A C ) as a function of conduc- The tables and experimental curves allow us to graphically visu-
tion velocity at different times of conduction. (From Maynard FM, alize the relationship of experimental error to time and distance.
Stolov W C : Experimental error in determination of nerve conduction For example, if we assume a lower limit of normal conduction
velocity.Arch Phys Med Rehabil 1972;53:362-372, with permission.) for the ulnar nerve to be 50 m/s, the experimental error equation
Chapters NERVE C O N D U C T I O N STUDIES — 185
Table 5-1. Calculated Experimental Error in Conduction Velocity (m/s) for

Different Conduction Distances and Conduction Velocity Magnitudes
Conduction Conduction distance (mm)
velocity
(m/sec) 50 75 100 125 150 175 200 225 250 275 300 350 400 500
15 3.4 2.2 1.7 1.3 I.I 1.0 0.8 0.7 0.7 0.6 0.6 0.5 0.4 0.3
20 5.5 3.7 2.7 2.2 1.8 1.6 1.4 1.2 I.I 1.0 0.9 0.8 0.7 0.5
25 8.1 5.4 4.1 3.2 2.7 2.3 2.0 1.8 1.6 1.5 1.4 1.2 1.0 0.8
30 1 1.3 7.5 5.6 4.5 3.8 3.2 2.8 2.5 2.3 2.0 1.9 1.6 1.4 I.I
35 14.9 9.9 7.4 6.0 5.0 4.3 3.7 3.3 3.0 2.7 2.5 2.1 1.9 1.5
40 19.0 12.7 9.5 7.6 6.3 5.4 4.8 4.2 3.8 3.5 3.2 2.7 2.4 1.9
45 23.7 15.8 11.8 9.5 7.9 6.8 5.9 5.3 4.7 4.3 3.9 3.4 3.0 2.4
50 28.8 19.2 14.4 11.5 9.6 8.2 7.2 6.4 5.8 5.2 4.8 4.1 3.6 2.9
55 34.5 23.0 17.2 13.8 1 1.5 9.9 8.6 7.7 6.9 6.3 5.7 4.9 4.3 3.4
60 40.7 27.1 20.3 16.2 13.6 11.6 10.2 9.0 8.1 7.4 6.8 5.8 5.1 4.1
65 47.3 31.5 23.7 18.9 15.8 13.5 11.8 10.5 9.5 8.6 7.9 6.8 5.9 4.7
70 54.5 36.3 27.2 21.8 18.2 15.6 13.6 12.1 10.9 9.9 9.1 7.8 6.8 5.4
75 62.2 41.4 31.1 24.9 20.7 17.8 15.5 13.8 12.4 11.3 10.4 8.9 7.8 6.2
80 70.3 46.9 35.2 28.1 23.4 20.1 17.6 15.6 14.1 12.8 11.7 10.0 8.8 7.0
From Maynard FM, Stolov WC: Experimental error in determination of nerve conduction velocity. Arch Phys Med Rehabil 1972;53:362-373, with permission.
provides us with some idea of the confidence with which the measured conduction velocity is normal within experimental
nerve can be judged normal based solely on experimental error. error. For a conduction distance of 150 mm, the AC is calculated
Suppose we consider a conduction distance of 250 mm for to be 13.6 m/s. Our calculated conduction velocity of 60 m/s
nerve fibers conducting at 60 m/s with a resulting conduction may actually be as low as 46.4 m/s. Because of the increased
time of 4.16 ms. Substituting the appropriate values into the ex- experimental error secondary to a reduced distance of measure-
perimental error equation yields: AC = 60 m/s (3.6 mm/250 mm ment, we no longer can state with 97% assurance that the NCV
+ 0.5 ms/4.16 ms) = 8.1 m/s. The value of 8.1 m/s represents we measured is normal. The experimental error tables and
two standard deviations about the calculated NCV with respect graphs (Tables 5-1 and 5-2; Figures 5-18 and 5-19) suggests that
to experimental error. Subtracting 8.1 m/s from the measured 60 we should increase the distance over which the nerve's conduc-
m/s produces a normal conduction velocity of 51.9 m/s. We can tion is calculated in order to reduce the experimental error into
state with 97% assurance (two standard deviations) that the an acceptable range, i.e., measured NCV is greater than 50 m/s.
Table 5-2. Calculated Experimental Error (2 Standard Deviations) in Conduction Velocity (m/s) for
Different Conduction Times and Conduction Velocity Magnitudes
Conduction Conduction time (msec)
velocity
(m/sec) 1.5 1.75 2.0 2.5 3.0 3.25 3.5 4.0 4.25 4.5 5.0 5.5 6.0 7.0
15 7.5 6.4 5.6 4.5 3.7 3.4 3.2 2.8 2.6 2.5 2.2 2.0 1.9 1.6
20 9.1 7.8 6.9 5.5 4.6 4.2 3.9 3.4 3.2 3.0 2.7 2.5 2.3 2.0
25 10.8 9.3 8.1 6.5 5.4 5.0 4.6 4.1 3.8 3.6 3.2 3.0 2.7 2.3
30 12.5 10.7 9.4 7.5 6.3 5.8 5.4 4.7 4.4 4.2 3.8 3.4 3.1 2,7
35 14.2 12.2 10.6 8.5 7.1 6.5 6.1 5.3 5.0 4.7 4.3 3.9 3.5 3.0
40 15.9 13.6 11.9 9.5 7.9 7.3 6.8 6.0 5.6 5.3 4.8 4.3 4.0 3.4
45 17.5 15.0 13.2 10.5 8.8 8.1 7.5 6.6 6.2 5.8 5.3 4.8 4.4 3.8
50 19.2 16.5 14.4 11.5 9.6 8.9 . 8.2 7.2 6.8 6.4 5.8 5.2 4.8 4.1
55 20.9 17.9 15.7 12.5 10.5 9.6 9.0 7.8 7.4 7.0 6.3 5.7 5.2 4.5
60 22.6 19.4 16.9 13.6 1 1.3 10.4 9.7 8.5 8.0 7.5 6.8 6.2 5.6 4.8
65 24.3 20.8 18.2 14.6 12.1 11.2 10.4 9.1 8.6 8.1 7.3 6.6 6.1 5.2
70 25.9 22.2 19.5 15.6 13.0 12.0 1 I.I 9.7 9.2 8.6 7.8 7.1 6.5 5.6
75 27.6 23.7 20.7 16.6 13.8 12.8 11.8 10.4 9.8 9.2 8.3 7.5 6.9 5.9
80 29.3 25.1 22.0 17.6 14.7 13.5 12.6 1 1.0 10.3 9.8 8.8 8.0 7.3 6.3
From Maynard FM, Stolov WC: Experimental error in determination of nerve conduction velocity.Arch Phys Med Rehabil 1972;53:362-372, with permission.
Conduction Velocity. The experimental error curves for AC as standardized distances versus anatomic landmarks for recording
a function of distance and time reveal that experimental error is distal motor latencies or sensory latencies. - - For example,
229 233 234
increased as the conduction velocity increases for a given dis- it is intuitively obvious that there is a significant amount of in-
tance or conduction time. Also, at any particular conduction ve- terindividual variation regarding the size of one hand versus an-
locity, the experimental error decreases as the conduction other. Examining distal motor or sensory latencies by placing
distance or time increases. In our above example, it can be seen recording or stimulating electrodes at a specified anatomic loca-
that the final contribution to experimental error at 250 mm from tion or a number of centimeters proximal to it, predisposes one
distance (0.01) is less than that arising from time measurement to more variability and error. A very large number of subjects
194
(0.12). The error in distance, therefore, contributes only 7.7% is necessary to account for the wide normal variation of hand
(0.13/0.01 x 100 = 7.7%) to the total experimental error while sizes. Also, to be completely representative, the effect of gender
time assessment contributes 92.3% to the error. This simple ex- must be separately examined because of an anticipated gender
ample clearly illustrates the practical utility of being aware of ex- hand size difference. Standardizing the distance over which a
perimental error and the necessity of carefully measuring distance nerve is measured, however, eliminates the above noted differ-
and in particular latencies. It is also important to round off the ences and should generate much more uniform data with wider
calculated conduction velocity after the calculations. This implies applicability to the "normal" population. A way to avoid this
194
that distances should be measured in millimeters and not rounded normal variation is to use conduction velocities. Reporting only
off to centimeters before the conduction velocity is calculated* distal motor or sensory latencies without a standard distance is
Amplifier Sensitivity. In the same investigation in which the less than optimal. It is also crucial to use standardized distances
above experimental error for time and distance were calculated, when defining reference data for sensory amplitudes. Small in-
it was noted that the standard deviations at amplifier sensitivi- creases in distance can profoundly affect the SNAP'S amplitude.
ties of 5000, 1000, and 200 pV/div, respectively, for distal la- Recording a SNAP amplitude and comparing it with a "norma-
tencies were greater (6%, 6%, and 9%) than those for proximal tive" data base using anatomic landmarks as opposed to stan-
latencies (2%, 4%, and 4%). Additionally, the standard devia- dardized distance is inappropriate.
tions or how much variation exists about the mean value, for Also, when measuring distances, one should record the dis-
both proximal and distal latencies, are less at 5000 pV/div than tance between the center of two sequential cathode positions or
those at 200 pV/div. One explanation of these findings may be from the center of the cathode to the center of the E-l elec-
that distal latencies are somewhat harder to accurately measure trode. - This is because the stimulus is depolarizing the
250 333
because of the stimulus artifact typically interacting with the nerve about the cathode and is not displaced from it unless cur-
onset latencies, particularly for short conduction times or rents significantly in excess of the supramaximal (as defined
recording distances. Also, at 5000 pV/div the onset of the po- above) level are used. Repositioning the anode about the cath-
116
tential is better defined as it quickly arises from the baseline, ode does not affect the onset latency of a mixed nerve potential
while at higher sensitivities there is a more gradual baseline de- thus demonstrating that neural excitation occurs in a region lo-
parture as smaller changes are detected, thus resulting in more calized about the cathode and not some intermediate distance
uncertainty. Amplifier sensitivities at 1000 pV/div are interme- between the cathode and anode. Stimuli delivered with 6 mm or
diate in terms of the standard deviations and is the reason some 20 mm diameter disc electrodes results in the same onset la-
investigators prefer this setting to optimize onset sensitivity yet tency of the CMAP, thus suggesting that the nerve is activated at
minimize experimental error. Of course, more stimuli are re- the center of each individual electrode supporting the con-
quired as the amplifier must be adjusted to then record the entire tention that measurements should be performed from the center
magnitude of the potential that often exceeds the CRT's vertical of the stimulation site. 36
resolution. The sensitivity of 1000 pV/div is usually too high to Limb Position and Anatomic Nerve Course. A final note re-
view the entire potential on the CRT. garding measurement factors affecting NCS is that of limb posi-
Biologic Variation. Once the experimental error is quanti- tion and anatomic course of the nerve. Studies comparing the
fied, it is then possible to calculate the amount of variation in- presumed course of the forearm nerves by surface estimates
volved in nerve conduction velocities arising solely from versus actual anatomic course and length reveal that surface
"normal" biologic differences. Suppose the median nerve has a measurements are 3-8 mm shorter than the anatomic length. A 36
conduction velocity 56.7 ± 3 . 8 m/s over a distance of 250 mm

64
1% length discrepancy in surface compared to actual length is
for a conduction time of 4.4 ms, and the distance and time errors noted in the peroneal nerve in the leg. In addition to this, poten-
8
noted above are operative. The experimental error contributing tial problem is the difference in surface compared to anatomic
to the conduction velocity is calculated to be 7.2 m/s: AC = 56.7 measurements of nerves coursing across a joint. The most obvi-
(3.6 mm/250 mm + 0.5 ms/4.4 ms) = 7.2 m/s. Because this ous example is that of the ulnar nerve traversing the elbow.
number represents two standard deviations, one standard devia- Calculating ulnar nerve conduction across an extended com-
tion is equal to 3.6 m/s. The total variation is the summation of pared to flexed elbow consistently produces lower conduction
the experimental plus biologic variations, which is equivalent to velocities. This slowing is obviously secondary to a measured
41
the square of the total variation consisting of the square of the distance too short for the actual length of nerve conveying the
standard deviations for the experimental and biologic varia- electrical impulse. In the flexed position the nerve must be long
tions. Applying this relationship to our data reveals that the
231
enough to accommodate elbow flexion; therefore, the nerve be-
normal biologic variation is approximately 1.2 m/s: (3.8 m/s) = 2
comes redundant in the extended position. Anatomic dissections
(biologic variation) + (3.6 m/s) ; biologic variation = 1.2 m/s.
2 2
in cadavers confirm that the ulnar nerve buckles upon itself in
Applying these principles to any experimentally obtained data the extended positioned and unfolds with elbow f l e x i o n . A
4130
can yield the amount of variation resulting from normal in- detailed discussion of the various techniques available for cal-
terindividual variation and that arising from experimental error. culating ulnar nerve velocities across the elbow is presented
Standard Deviation vs. Anatomic Landmarks. An impor- when ulnar nerve NCV techniques are described (see below).
tant issue with respect to measurement and NCV is the use of Obstetric calipers are recommended when attempting to measure
nerve distances across a spiral pathway. For example, appro- 250

myelinated fibers proportional to the decreasing numbers of
priate uses of obstetric calipers may be in determining the posterior spinal nerve root fibers was noted. A loss of fibers 51
length of radial nerve traversing the spiral groove, plantar was also noted in the peripheral and central projections of the
nerves from the sole of the foot to the medial malleolus, and acoustic nerve. Degeneration of the eighth nerve extensions was
transbrachial plexus measurements. also observed in the pathways through the brainstem into the
white matter of the cerebrum. The optic nerve displays simi-
129
PHYSIOLOGIC FACTORS lar changes to those previously noted for both peripheral and
central nerves. 71
There are a number of physiologic factors that have a direct Nerve Conduction Studies and Aging. One may consider
effect on nerve conduction studies. When considering physio- either motor or sensory responses with respect to nerve conduc-
logic factors, it is convenient for discussion purposes to divide tion studies and the effects of aging. Several generalizations can
them into those that can be altered by the practitioner and those be made regarding peripheral evoked sensory nerve actions po-
intrinsic to the subject and beyond control. The most important tentials (SNAPs). The conduction velocity demonstrates a con-
factor readily amenable to change is a limb's surface tempera- sistent decline approximating 1-2 meters/second per decade
ture. A number of physiologic variables beyond the control of (Table 5 - 3 ) . - - - The SNAP'S duration is about 10-15%,
17 47 105 342
the clinician are gender, age, height, digit circumference, and and 20% longer in the 4 0 - 6 0 - and 70-88-year-old individuals
anomalous innervation. than the 18-25-year-old persons, respectively. Compared to the
G e n d e r . Only a few studies have attempted to characterize 18-25-year-old group, the SNAP'S amplitude is one-half and
nerve conduction studies between males and females. There one-third, respectively, for the 4 0 - 6 0 - and 70-88-year-old
was noted to be a slight increase in the antidromic sensory nerve groups. The distal sensory latencies revealed a similar prolon-
30
amplitudes for both the median and ulnar nerves recorded from gation with age.
the digits in women. Also, females demonstrated a greater
17
The results of aging on conduction velocity have been exam-
nerve conduction velocity for upper and lower limb nerves. - 56 296
ined in a number of upper and lower limb nerves (Table 5-3).
Both of these differences, however, are minimized when one Motor nerve conduction velocities reveal similar changes to
considers limb length, height, and digit circumference (see sensory nerves. The newborn's motor nerve conduction veloci-
b e l O W ) . 17.277,304 ties are about half of adult values, which are reached by 3 - 5
Aging. Histologic Alterations with Aging. The density of years of age. After the age of 50 years, there is a progressive
6
large myelinated fibers in the distal portions of the sural nerve decline in the conduction velocity of the fastest motor fibers ap-
increases rapidly from birth to 3 years of age and reaches a proximating 1-2 m/s per decade. - - - There is a concur-
120 230 246 349
stable adult value by 3 years. A maximum fiber density of 6480 rent increase in the distal motor latency and decrease in the
fibers/mm is achieved by the third decade of life. The large
2
motor response's amplitude with advancing age. The decrease
fiber density subsequently progressively declines to 54% (3480 in amplitude is difficult to ascertain clinically as there is such a
fibers/mm ) of the second decade by 90 years of age. After 60
2 332
wide range of normal.
years of age, stenosis of the vasa nervorum is rather pro- The above nerve conduction study findings may be explained
nounced. - One may conclude, therefore, that only about half
52 332
to some degree by considering the histologic changes associated
of the large sensory myelinated fibers innervating the distal por- with the aging of the peripheral nervous system. The maximum
tions of the lower limbs survive the aging process. Of additional
interest is the relationship between the distances separating the
nodes of Ranvier (internodal length) and diameter of myelinated T a b l e 5 - 3 . D e c r e a s e in N C V w i t h A g e
fibers. Below the age of 65 there is a linear correlation of intern- p e r D e c a d e A f t e r 20 Y e a r s of A g e
odal length and diameter. At ages above 65 years there is less of Nerve NCV Range (m/sec)
a linear correlation and increased scatter of values, suggesting a
Motor Nerve Conduction
shortening of the internodal length. - A lessening of the in-
5 207
Median 0.6-2.3
ternodal length can result from demyelination and remyelina-
Ulnar 0.6
tion. This process is ongoing with age and does not affect all
342
Peroneal 0.4-0.8
of the large myelinated fibers equally. The above findings have
Posterior Tibial 1.7
been documented in multiple peripheral nerves of both the
upper and lower limbs. Sensory Nerve Conduction
With surface electrodes
Histochemical studies of limb muscles from elderly individu-
Median, orthodromic 3.0
als (65 years or older) without evidence of neuromuscular dis-
Median, antidromic 2.0
ease revealed fiber size variation, hyaline or granular
Ulnar, orthodromic 4.0
degeneration, loss of striations, clumps of pyknotic nuclei, in-
With near-nerve techique
creased fat and connective tissue, and most significantly neuro-
Median 1.8
genic fiber type grouping. - The fiber type grouping is more
161 247
pronounced in the lower than upper limbs and distally more Ulnar, up t o 54 years 1.2
than proximally. The etiology of fiber type grouping was pre- Ulnar, over 55 years 3.3
sumed secondary to degeneration and regeneration of the pe- Ulnar 0.1
ripheral nervous system with secondary reorganization of the Sural, calf, orthodromic 0.5 I.I
muscle fibers belonging to individual motor units. Mixed Nerve Conduction
In addition to age-related changes of the peripheral nervous Median 4.0
system, the central nervous system also demonstrates alterations Ulnar 3.5
associated with aging. The most conspicuous changes were From Oh SJ: Clinical Electromyography: Nerve Conduction Studies, 2nd ed.
noted in the posterior columns. - A progressive loss of large Baltimore, Williams & Wilkins, 1993, with permission.
236 251
188 — PART II BASIC A N D ADVANCED TECHNIQUES
adult nerve conduction velocities are achieved coincidentally muscles showed a systematic difference: the longer axons were
when myelination of the large fibers is completed, usually by alway slower in both the proximal and distal segment. This 371
the age of 5 years. - The subsequent reduction in sensory and

6 349
was true for both arm and leg motor nerves. The additional
motor conduction velocity and amplitude with increases in the distal slowing due to tapering did not clearly add to this effect.
distal latency and response duration correlates well with ad- These findings agree with anatomic studies indicating greater
vancing age. The aging nervous system demonstrates loss of nerve diameter in roots to proximal portions of the limb than to
large fibers and evidence suggestive of progressive demyelina- more distal muscles in humans. Overall a slowing of 3.8 m/s
270
tion and remyelination, particularly after the sixth decade. - - 5 207 342
per every extra 10 cm of axonal length was found. It can be
371
The combination of large fiber loss and segmental demyelina- understood that this has a considerable impact on normal values
tion/remyelination has been suggested as possible cause leading given a range of body heights between 1.50 to 2 meters. It is be-
to a slowing of conduction velocities. Buchthal has postulated lieved the previously noted gender differences for NCV are in
that the primary cause of the noted electrophysiologic changes reality height effects as women are characteristically shorter
is a direct result of an alteration in the nerve's membranous than men and display faster conduction velocities solely for this
properties required to sustain the appropriate current density reason. These height differences have not been found by all in-
necessary for maximum conduction velocities. 30
vestigators and previous findings of neural conduction slowing
Digit Circumference. Females consistently demonstrate a with height have been ascribed to poor techniques and lack of
significant difference in antidromic SNAP amplitudes for the temperature control. - - Continued investigation is needed
336 337 371
ulnar and median nerves from the second and fifth digits. A 17
to further define the influence and etiology of height on NCV.
negative linear correlation exists between finger circumference Temperature. One of the most profound factors influencing
and amplitude for these two nerves. Additionally, it is known nerve conduction studies is temperature. In order to fully appre-
that as the distance between the recording electrode and neural ciate the effects temperature can have on SNAP and CMAP pa-
generator increases, the amplitude precipitously declines. rameters, it is perhaps best to first consider how a nerve's action
Increasing the circumference of the finger, especially with re- potential morphology changes with different temperatures.
spect to subcutaneous tissue, displaces the electrode further When single myelinated fibers of the frog are cooled, a number
from the nerve. Men have significantly larger finger circumfer- of interesting observations are made with respect to the nerve's
ences than women, thus providing a size difference explanation excitability and its action potential parameters. As the tem-
291
for the amplitude dissimilarities rather than an intrinsic neural perature of the nerve is lowered, the amount of current required
difference between male and female nerves. to generate an action potential increases. In other words, neural
Height. Several investigations have documented slower excitability is lowered with a reduction in temperature. This de-
nerve conduction velocities in taller compared to shorter indi- creased excitability is a direct temperature effect on the nerve's
viduals. - This difference was found to be independent on the
34 202
action potential generating mechanism at the nodes of Ranvier, |
limb's temperature or subject's age. ° - ' 1
Although the etiol-
5 06,202
and not a result of membrane resistance changes, i.e., the trans-
ogy of this finding is unknown, an interesting hypothesis has membrane resistance is not increased by a drop in temperature.
been suggested to explain the observation. It is known that prox- In addition to excitability, the morphology of an action potential
imal compared to distal nerve conduction velocities in different is profoundly affected by a drop in temperature.
nerves and more rostral nerve segments compared to distal por- Effect on Action Potential The action potential's amplitude
tions of the same nerve conduct faster. - - For example, com-
8 228 329
increases as the nerve's temperature declines. In addition to am-
paring proximal median and ulnar motor NCVs to distal ones plitude, the action potential's rise and fall times are also in-
revealed an 18% and 11% greater proximal conduction velocity, creased. Specifically, the time required for the action potential
respectively. A 10-20% difference between proximal and
333
to reach its peak depolarization from the resting membrane level
distal NCVs was found for lower limb sensory nerves. Recall 8
is increased approximately 3 3 % . Also, the time necessary for
that nerve conduction velocity is proportional to axon diameter. the action potential to return to its resting level is also increased
It is suggested that the more proximal nerve fibers conduct com- but much more so than the rise time (69%). Because both the
parably faster than the distal ones because the individual nerve's duration and spike height have increased, the area of the action
diameter tapers in the distal aspects of the limb, although some potential increases dramatically at lower temperatures. Similar
documentation exists to support distal tapering. This is not 96
data regarding action potential parameter changes at lower tem-
universally found for all nerves. Combining the above infor-
270
peratures are found in the giant axon of the squid, i.e., increases
mation with the postulate that there is an abrupt decrease in the in action potential duration (rise time and descent time), spike
axonal diameter of individual nerves at some uniform distance amplitude, and area. Application of the mathematical equa-
140
from the anterior horn cell irrespective of the patient's height tions described by Hodgkin and Huxley for the giant axon of
has been suggested to account for the inverse relationship for the squid also predicts that the action potential should increase
height and NCV. If the axon suddenly decreases in diameter at a in amplitude as well as duration, and fall time greater than rise
set distance from the motor neuron, then tall persons would time, just as observed in animal preparations. Additionally, the
have a greater percentage of smaller axon diameters than short peak sodium conductance is also predicted to increase approxi-
subjects for a given segment, such as the leg. The shorter per- mately 3 8 % .
141
sons would still contain larger-diameter fibers when tall people Alteration of Conduction. A decrease in temperature is also
possess only small-caliber axons. Since NCV displays a linear found to alter conduction differently in nerves of various diame-
relationship to axon diameter, it should not be unexpected that ters. The large-diameter fibers comprising the A group require
tall persons have demonstrably slower conductions than their less of a drop in temperature to produce action potential block-
shorter counterparts. An alternative explanation is that long ade than the C fibers. Within the A fiber group, cessation of
axons are smaller and thus conduct slower from their origin. It action potential propagation following temperature reduction
was found that proximal conduction—measured by root stimu- occurs first in the delta fibers and last in the alpha fibers.
lation—over identical segments to proximally and distally located Comparing the motor and sensory fibers contained in the alpha
size category, action potential propagation fails first in the sodium channels to flow for a longer period, i.e., by as much as
motor fibers. The rate of NCV decline with a drop in tempera-
74
3^4 times (see above). Because we know the resistance of the
ture is proportionately the same irrespective of the fiber's size, membrane changes minimally if at all, an increase in current
i.e., the rate of NCV drop per degree is the same for different should result in elevated action potential magnitude, i.e., Ohm's
nerves based on the percent of their normal NCV (see below). law states E = IR; if R is constant and I increases, then E must
Prior to myelinated nerve conduction failure at about 7-8°C, also increase. Indeed, as already demonstrated in several
NCV may be reduced to 1-2% of normal, i.e., for fibers with species, the nerve's action potential's amplitude and duration
conduction velocities of 6-100 m/s, low temperatures (8-10°C) increase, particularly the action potential's decline associated
can produce conductions of 0.06 to 1.0 m/s. - 255 256
with sodium inactivation. Simply, if the sodium gates remain
Refractory Periods. The mechanism producing the above ef- open longer, then current flows for an increased period of time,
fects may be understood if one considers the nerve membrane's thereby generating a larger and longer action potential. Also,
refractory period. Following neural excitation, there are two failure of gate closure means there are more gates open for a
time periods of interest known as (1) the absolute refractory longer time period as not all the gates respond immediately but
period, and (2) the relative refractory period. For a short time stay open for variable time periods prior to closing. Thus, nor-
period after an action potential, the nerve's membrane cannot be mally the gates that open first also close first. A drop in temper-
excited irrespective of how large the depolarizing stimulus; this ature, however, causes these initially open gates to remain open
time is called the absolute refractory period. After the ab- while their neighbors also begin to open, resulting in more cur-
solute refractory period, a relative refractory period occurs rent flow. Therefore, delayed sodium inactivation appears to ex-
during which an action potential can be induced, but only with a plain the morphologic alterations noted in action potentials with
stimulus intensity greater than that normally required. The time changes in temperature. We can now apply this information to
of the absolute refractory period is essentially the duration of whole nerves and understand the changes noted.
the action potential plus a subsequent short interval. Contained Local Cooling. Regarding sensory nerves in humans, two
within the absolute refractory period are sodium activation and different effects with respect to temperature must be considered:
sodium inactivation. The termination of the absolute refractory (1) local cooling, and (2) regional cooling. Local cooling is de-
period is initiated by neural repolarization to where there are fined as decreasing the temperature of the nerve only about the
sufficient sodium gates available to again generate an action po- E-1 recording electrode site or for a finite distance within the
tential. In normal myelinated nerve fibers, the duration of the region of the recording electrode. If the sural nerve is recorded
action potential is about 0.4 ms with an additional 0.3-0.5 ms of posterior to the lateral malleolus and stimulated proximally,
membrane refractoriness for a total absolute refractory time of local cooling is accomplished by decreasing the temperature
roughly 1.0 ms with individual variations. The limiting factor in over the recording electrode only. The amplitude of the sural
203
how fast a nerve can conduct impulses is thus the absolute re- SNAP increases as the temperature surrounding the active
fractory period resulting in a maximum firing rate of 1,000 Hz. recording electrode is decreased. Also, the rise time of the
The value in studying the refractory periods in nerve is that one SNAP is increased as the temperature is lowered. Upon rewarm-
can gain indirect evidence as to how the sodium gates respond ing, the SNAP returns to its previous amplitude and rise time.
to temperature fluctuations that in turn yield information re- Similar observations have been documented in other sensory
garding the mechanism of temperature effects on nerve mem- nerves. - - - These findings are certainly expected as the re-
,8 ,9 2n 214
branes and NCVs. sponse of the entire nerve is merely the summation of what is
Decreasing the temperature surrounding human nerves alters occurring on a single-fiber bases. As previously described, indi-
the absolute refractory period in a characteristic manner. vidual nerve fibers generate action potentials with increased
Lowering the temperature of median nerve fibers from 34°C to amplitudes, longer peak rise and fall times, as well longer total
20°C results in an absolute refractory period change from 1.8 potential durations. The increased SNAP amplitude, negative
ms to 5.5 m s . The ulnar nerve sensory fibers demonstrated
63
peak duration, and increased peak rise and fall times simply re-
similar changes in the absolute refractory period of 0.54 ms to flect the composite changes occurring at the single fiber level
3.07 ms from 35°C to 20°C, respectively. Using a slightly dif-
212
because of delayed sodium inactivation. The net effect of all
ferent technique, median nerve sensory potentials demonstrated these physiologic changes is to slow the fastest-conducting
an absolute refractory period change from 0.8 ms to 1.8 ms for fibers somewhat more than the slower-conducting fibers.
temperatures of 36°C and 24°C. Relative refractory periods in Therefore, there is an increase in the temporal synchronicity
these same nerves were 2.5 ms and 10.0 ms, respectively. Mixed with which the individual nerve fiber SNAPs summate at the
ulnar nerve fibers demonstrated absolute refractory periods of recording electrode. A more synchronous arrival of individual
0.8-2.1 ms for fast-conducting fibers and 1.0-3.8 ms for fibers results in a significant increase in the composite SNAP
slower-conducting fibers. 12.30.175-178.255.256 j mammalian nerves
n
amplitude measured for all the fibers by the electrode.
the action potential occurs only at the nodes of Ranvier, which The CMAP arising from cooled muscle tissue demonstrates
contain significant numbers of sodium channels and essentially similar changes as those noted for SNAPs, although the increase
no potassium channels. - As a result, the action potential
25 46
of amplitude with cooling is less pronounced than that noted for
arises exclusively from sodium channel activation. The refrac- the SNAP. - CMAP amplitude, duration, rise time, and area
1819 67
tory period, therefore, essentially consists of sodium inactiva- all increase as the muscle's temperature is reduced. Intra-
tion and passive leak currents. Decreasing temperature alters muscular recordings also reveal that those motor units in close
sodium activation slightly but significantly affects sodium inac- proximity to the recording electrode are also increased in the
tivation by slowing it down considerably. - It has been sug-
25 211
same parameters noted a b o v e . However, there is some dis-
272
gested, therefore, that a prolongation of sodium inactivation is agreement in this area because one study documented a reduc-
the major cause for increases in human nerve refractory periods tion in the motor unit action potential amplitude with a cooling
with cooling. The increased time of sodium gates remaining of muscular tissue. It is these authors' opinion that lowering
29
open allows the depolarizing current associated with open muscular tissue temperature should result in an increase in the
amplitude, rise time, and duration of the individual motor unit latency correlation of 0.2 ms/°C. Using this information al-
action potentials as observed with needle electromyography. lowed a correction formula to be devised to correct median and
Unlike neural tissue, cooled muscular tissue demonstrates a ulnar motor or sensory NCV or distal latencies (DL) measured
resting membrane potential that fluctuates directly with alter- at a particular temperature in the above-noted range to an opti-
ations in temperature. This fluctuation, however, is of ques-
242
mal skin temperature of 33°C. This formula is: NCV or D L { l e m p
tionable clinical significance at this point. Refractory periods in corrected) ~~

human muscle demonstrate increases in duration when the tem- CF (33°C - skin temp ° C ) + NCV or ( m e a s u r e d ) ( c a l c u l a t e d )
perature is lowered. - The magnitude of these temperature ef-

63 93
^(calculated)' where CF equals the appropriate correction
fects reveals a 2.3 ms absolute refractory period at 34°C which factor. The same investigators also-determined the correlation
127
is elevated to 9.3 ms at 20°C. The increased refractory period between temperature and NCV in commonly investigated lower
suggests that the same mechanism of prolonging sodium inacti- limb nerves. The peroneal nerve demonstrated a correction
vation applies to muscle membrane as well as nervous tissue. In factor of 2.0 m/s/°C when the skin temperature is measured 15
short, with respect to decreased temperature, CMAP and SNAP cm above the lateral malleolus for a temperature range of
parameters demonstrate similar findings to local cooling of the 26-32°C. Again a formula can be used to correct the observed
active tissue surrounding the recording electrode. NCV to a temperature of 32°C: N C V = 2.0 (32°C - ( l e m p c o r r e c l e d )
Relationship Between NCV and Temperature. In addition to skin temp ° C ) + NCV . Finally, temperature cor-
( m e a s u r e d ) (calculated)
the morphologic characteristics of the SNAP and CMAP, one relations for the tibial and sural nerves were also described for
must also consider the rate at which sensory and motor action the same surface temperature range measured 15 cm proximal
potentials propagate at different temperatures, i.e., the relation- to the medial malleolus. Correction factors of 1.1 and 1.7
ship between NCV and temperature. Based upon the single m/s/°C were used for the tibial and sural nerves, respectively.
nerve fiber's response to a lowering of temperature, prolonga- The tibial nerve correction formula is: N C V =1.1 ( t e m p c o r r e c l e d )
tion of the rise and fall time, we should be able to infer what (32°C - skin temp ° C ) + NCV . For the sural
( m e a s u r e d ) ( c a l c u l a t e d )
might be anticipated by the whole nerve's response to cooling nerve a correction formula is: N C V = 1.7 (32°C - ( l e m p c o r r e c t e d )
with respect to NCV. Because propagation is saltatory in myeli- skin temp ° C ) + NCV . If one is more accus-
( m e a s u r e d ) ( c a l c u l a t e d )
nated nerve, the above effect of decreased temperature results in tomed to calculating sural nerve peak latencies for a 14-cm dis-
an increase in the amount of time necessary to reach the action tance a correction formula may be used as follows: sural
potential's peak at each node of Ranvier. As more time is re- latency =1/{0.012 (32°C - skin temp ° C
( t e m p c o r r e c t e d ) ) + ( m e a s u r e d )
quired at each node, over comparable segments of nerve the 1/ \ 1 2 5
cooler nerve should have a slower conduction velocity. This is

1
' (measured latency) J •
exactly what is observed in cool extremities when calculating Regional Cooling. As previously noted, local cooling pro-
both sensory and motor nerve conduction velocities. duces an increase in the duration, rise time, and amplitude of
The first detailed investigation of temperature effects on both CMAPs and SNAPs. Regional cooling of a nerve segment
NCV in human nerves revealed an NCV-to-temperature correla- over which conduction is measured, however, results in a pro-
tion of 2.4 m/s/°C for median and ulnar motor conduction. 132 longation of the response and possibly a slight decrease or no
With every 1° drop in temperature, one could anticipate a 2.4 change in the amplitude compared to a warm limb. This is a 310
m/s decrease in the conduction velocity. Reductions in conduc- consistent finding provided the E-l recording site is not signifi-
tion velocity for upper limb motor nerve fibers have also been cantly involved in the temperature reduction. The explanation
found to approximate a decrease of 4 - 5 % per degree for the drop in amplitude is most likely the result of a differen-
Celsius. - Multiple investigations have revealed slightly dif-
56 163 tial effect of temperature on NCVs for fast- as opposed to slow-
ferent correlations between NCV and temperature, which is conducting fibers. The anticipated increase in temporal
62
most likely due to different measurement techniques. One of the dispersion would lead to a greater separation of fast and slow
difficulties in attempting to define a relationship between NCV action potentials over the same nerve segment for a comparable
and temperature is deciding if surface, subcutaneous, or intra- cooler segmental temperature. An increase in temporal disper-
muscular temperatures are most appropriate. Because each of sion leads to greater phase cancellation and less phase summa-
these three regions yields different temperatures for the same tion, thus producing a reduction in the evoked response
site, it should come as no surprise that different correlations are amplitude and area as well as a prolongation in the total poten-
obtained for each depth. Fortunately, there is a linear correlation tial's duration. This concept may be better understood with an
between the three tissue depths justifying the use of surface example.
temperature measurements in attempting to calculate correction One way to study the effect temperature has on nerve con-
factors. Of course, correction factors using subcutaneous and
125 duction velocity is to calculate the change over a 10° tempera-
intramuscular readings are also correct. It is significantly more ture difference thereby producing a ratio known as Q , i.e., Q 0 1 0
convenient for most practitioners to use surface measurements. = NCV , o / N C V , . For example, an NCV for the median
( T + 0 C (T C)
The studies that carefully attempted to control for temperature nerve sensory fibers at 37°C conducting over the wrist to elbow
and use standardized surface sites for measuring temperature segment for the fast fibers is found to be 60 m/s with a known
are described. 125 127 Q of 1.5. The slow fibers are found to conduct at 30 m/s for
i 0
62
the same temperature range with a similar Q . The Q , is found

In the upper limb the relationship between temperature and
10 0
to be the same for both slow- and fast-conducting fibers. We 62
NCV was investigated for the surface temperature range of

can then ask, what is the temporal dispersion at 37°C and for a
26-33°C measured at the wrist's midline on the distal crease.
5°C drop in temperature over a 25-cm segment of nerve for the
Calculations revealed that for median motor and sensory nerves,
fast- and slow-conducting fibers? First, we must find the tem-
NCV was altered 1.5 and 1.4 m/s/°C, respectively, while the
perature correction factor or just how much does the nerve con-
distal latency for both changed 0.2 ms/°C. The ulnar nerve
duction velocity change per degree drop. By applying the Q
demonstrated motor and sensory temperature relationships of
1 0
ratio for a temperature range of 37°C to 27°C we discover that

2.1 and 1.6 m/s/°C, respectively, and a distal motor and sensory
the correction factor is 2.0 m/s/°C: (60 m/s) o /(X)27°c = 1-5; 37 C
X 7°c = 40 m/s, i.e., the conduction velocity for the fast fibers at
2
27°C is 40 m/s; (60 m/s - 40 m/s)/10°C = 2.0 m/s/°C. Applying correcting the NCV to a given temperature irrespective of the
the correction factor to our 5°C drop yields a conduction veloc- time involved. It is recommended that the practitioner use at
ity of 50 m/s at a temperature of 32°C (2.0 m/s/°C x 5°C = 10 least a surface temperature between the stimulating and record-
m/s drop in NCV, i.e., 60 m/s - 10 m/s = 50 m/s). At 37°C the ing electrodes of 32°C and 30°C for upper and lower limbs, re-
time required for the fast-conducting nerve fibers to traverse a spectively, to record data.
25-cm neural segment is 4.2 ms (60 m/s = 250 mm/t; t = 4.2
ms). The slow-conducting fibers cover this same segment of A n o m a l o u s Innervation
nerve at 37°C in 8.3 ms. The temporal dispersion, therefore, be- There are a number of variants in both the upper and lower
tween fast and slow fibers at 37°C is 4.1 ms. The slow fibers limb with respect to muscular innervation. It is important for the
conduct at 25 m/s at 32°C (apply above reasoning to arrive at practitioner to be fully aware of these so-called anomalous in-
this solution) and have a correction factor of 1 m/s/°C. Note that nervation patterns as they can often pose diagnostic challenges
for the same Q the fast and slow fibers have different correc-
1 0 in various peripheral nerve injuries. We examine three upper
tion factors. Therefore, the conduction velocity for the slow and one lower limb diagnostically relevant innervation variants
fibers at 32°C is 5 m/s less than at 37°C or 25 m/s. At 32°C, the from an electrophysiologic perspective. The manner in which
temporal dispersion for the 25-cm cooled nerve segment be- these anomalies present is stressed to avoid an erroneous con-
tween the fast and slow fibers is 5.0 ms or a net increase of tem- clusion given a particular set of findings.
poral dispersion of 0.9 ms. This difference increases if slower
fibers are considered, particularly at lower temperatures. It is Upper Limb
important to recognize the differential effect of temperature on M a r t i n - G r u b e r A n a s t o m o s i s . In the upper limb the
slow- and fast-conducting fibers through a region of cooled median and ulnar nerves normally travel as two distinct nerves
nerve producing temporal dispersion that tends to reduce the without an anatomic connection following their departure
amplitude of the recorded potential. Even though the action po- from the brachial plexus. All of the forearm flexors are inner-
tentials are increased for all cooled nerve fibers, the competitive vated by the median nerve except for the flexor carpi ulnaris
process with respect to amplitude of increased temporal disper- (FCU) and the medial two muscle bellies of the flexor digito-
sion and subsequent phase cancellation results in a compara- rum profundus (FDP). It is possible for the median nerve to
tively smaller potential. This is just the opposite effect of local also supply the FDP to the fourth and fifth digits. Rarely, the
cooling that produces an increase in amplitude. If both the ulnar nerve may supply the FDP to the second digit. In the
region of nerve where the response is recorded as well as the hand, however, all of the intrinsic muscles are innervated by
portion conveying action potentials is cooled, the two opposing the ulnar nerve except for the abductor pollicis brevis (APB),
effects yield unpredictable responses; personal observations opponens pollicis (OP), one-half of the flexor pollicis brevis,
suggest that the response is delayed but the amplitude may be (superficial head) and the first two lumbrical muscles. This
larger than at a warmer temperature. classic innervation pattern generally serves the practitioner
Given the previous information regarding temperature and its quite well, but is certainly incomplete. A relatively common
profound effect on NCV, one can readily appreciate the neces- communication between the median and ulnar nerves in the
sity of recording and attempting to control the temperature. A forearm is both clinically and electrophysiologically impor-
cool limb, irrespective of the ambient room temperature, can tant, and should be understood by all individuals practicing
result in abnormal latencies, NCVs, and amplitudes. In short, a electrodiagnostic medicine.
normal limb can certainly yield abnormal results because of a The above-noted neural connection was first described
low temperature. An equally important issue concerns an abnor- anatomically by Martin and Gruber resulting in the eponym
224 121
mal nerve, which is presumably why an individual is seeking associated with this anomaly bearing their names, i.e., Martin-
the expertise of a specialist in electrodiagnostic medicine. Gruber anastomosis. The Martin-Gruber anastomosis is be-
Although significant work has been performed in attempting lieved to have an autosomal dominant pattern of inheritance 55
to define the necessary correction factors relating conduction and an incidence of 7 . 7 - 3 4 % , l23153 221
- - - - with 68% of af-
315 316 355
velocity with skin temperature, there remain serious questions fected persons having this anomaly bilaterally. In 9 1 % of per- 175
as to the relationship between abnormal nerves and how they re- sons with the Martin-Gruber anomaly the anatomic connection
spond to temperature variations. 70364
In a study carefully investi- between the median and ulnar nerve is a communicating branch
gating the response of normal and abnormal nerves to reduced from the anterior interosseous nerve directly to the ulnar
limb temperature, healthy nerves responded differently than ab- nerve. Because the anterior interosseous nerve does not typi-
308
normal nerves. For the sake of brevity, let us consider only two cally convey cutaneous sensory fibers, this may explain why the
parameters,J.e., CMAP distal motor latency and SNAP ampli- Martin-Gruber anastomosis typically does not affect sensation
tude. In normal persons the CMAP distal motor latency changed to the hand. Additional anatomic pathways include a branch
the expected amount of 0.28 m s / ° C . T h e same parameter in
19
from the flexor digitorum superficialis to ulnar nerve (6%) and
patients with compressive neuropathies and uremic peripheral a direct link between the main median and ulnar nerves (3%). 308
neuropathies demonstrated changes of 0.40 and 0.23 ms/°C, re- The nerve fibers involved in this anomaly are derived from the
spectively. SNAP amplitudes revealed an increment in normal C8/T1 nerve roots, travel with the median nerve (anterior in-
subjects of 1.96 pV/°C, while entrapment and uremic neu- terosseous branch), and cross over to the ulnar nerve to primar-
ropathies changed 0.94 and 0.84 pV/°C, respectively. Although ily supply the first dorsal interosseous (FDI) muscle 95-100%
it is recommended that applying correction factors are less time- of the time, hypothenar muscles (41-61%), and adductor polli-
consuming than heating the patient, it is questionable how ac-
126
cis (AP) (14%). - A number of reports state that the thenar
123 355
curately correction factors for temperature in normal persons muscle is innervated in the Martin-Gruber anastomosis. - It 175 355
apply to the individuals with abnormal nerves. Until more data is important to realize that this implies that the AP or flexor pol-
are available regarding the best correction factor for diseased licis brevis are the muscles referred to and not the APB or OP.
nerves, warming of the limb should be considered superior to The Martin-Gruber anomaly does not refer to innervation of the
typically median-innervated intrinsic hand muscles but to the positive deflection; however, with a concomitant Martin-Gruber
manner in which the ulnar-innervated hand intrinsics are sup- anastomis as the motor point of the AP can occasionally align
plied by fibers that travel with the median nerve in the arm and with the surface APB recording electrode. A Martin-Gruber
forearm. anastomosis should be suspected in a median nerve lesion distal
Clinically, the Martin-Gruber anomaly is important because to the median/ulnar nerve anatomic connection when the fol-
it is a way to convey innervation to the ulnar hand intrinsic lowing are observed: (1) initial positive CMAP deflection with
muscles despite an ulnar nerve lesion proximal to the commu- proximal median nerve stimulation, (2) significantly elevated
nication between the median and ulnar nerve. For example, if median nerve NCV, or (3) larger proximal than distal CMAP on
the neural fibers in the anastomosis innervate the majority of median nerve stimulation. One would also anticipate that the
hand intrinsics, it is possible for the patient to sustain a com- CMAP from ulnar-innervated hand intrinsic muscles with ulnar
plete lesion of the ulnar nerve at the elbow and continue to have nerve activation should be larger from the wrist compared to
a functional hand. Although the patient's strength in this case elbow stimulation. This is the expected finding and as a result is
may be somewhat reduced because of some concomitant inner- often overlooked. Additionally, the ulnar nerve fibers traveling
vation from the main ulnar nerve fibers, the typical ulnar claw proximally with the median nerve may not innervate the ADQ
hand may not appear. This can give the impression of an "all and, therefore, not yield a noticeable difference. Unfortunately,
median hand" when indeed the innervation to the hand intrin- the above-noted proximal and distal amplitude differences for
sics merely traveled with the median nerve proximal to the the median and ulnar nerve are not adequately quantified for the
ulnar nerve lesion. Nerve conduction studies in this patient Martin-Gruber anastomosis and are subject to the interpretation
would reveal that CMAPs are obtained from the ulnar-inner- and experience of the practitioner.
vated hand intrinsic muscles to stimulation of the median nerve Ulnar-to-Median Nerve Communication. An electrophys-
proximally, but not at the wrist, and a similar result with distal iologic communication conveying fibers to the median nerve
ulnar nerve excitation but not activation of the ulnar nerve from the ulnar nerve has not been documented in large detailed
above the lesion site. Needle electromyography may reveal ev- series of patient studies. A number of investigations using
178
idence of denervation in the ulnar nerve intrinsic muscles be- surface electrodes or physical examination, however, have sug-
cause of some fibers in the main ulnar nerve innervating the gested this type of c o n n e c t i o n . These results must be
174191222
examined muscles were injured. questioned, however, because of the possibility of a connection
Observation of the Martin-Gruber anastomosis in routine between the median and ulnar nerves in the hand (see below)
nerve conduction studies requires a keen eye and knowledge of and volume conduction from various nearby muscles confusing
how the results differ compared to the classic innervation pat- the electrophysiologic surface-recorded interpretation of vari-
tern. Typically, the amplitude of the C M A P recorded from the ous findings. One well-documented case may represent the
117
APB and abductor digiti minimi (ADM) when stimulating the presence of this anastomosis. The CMAP from the APB with
314
median and ulnar nerves, respectively, is somewhat larger dis- median nerve stimulation was significantly smaller at the elbow
tally than with proximal nerve activation. This is because of less than wrist. In this case, the possibility of conduction block in
temporal dispersion of the individual action potentials with the forearm was not considered and may have explained the
distal stimulation (see Chapter 2). If a Martin-Gruber anastomo- findings. In any event, prior to concluding that an ulnar-to-
sis is present, the proximal CMAP with median nerve stimula- median anomaly exists, other physiologic, pathologic, or tech-
tion while recording from the APB is typically larger than that nical reasons should be considered.
obtained distally. This is because the nerve fibers contained in Riche-Cannieu Anomaly. Riche and Cannieu indepen- 271 35
the median nerve at the elbow but destined to eventually join the dently described an anatomic communication of the median and
ulnar nerve are concomitantly activated at the elbow along with ulnar nerves in the hand between the recurrent branch of the
the median nerve fibers. The resulting volume conducted median nerve and the deep branch of the ulnar nerve. Although
CMAP from the AP, the deep head of the flexor pollicis brevis, largely ignored compared to the Martin-Gruber anastomosis, the
and possibly the FDI activated through the anastomosis sum- Riche-Cannieu anastomosis is anatomically present in approxi-
mates with that of the APB and OP to yield a larger proximal re- mately 77% of hands. The physiologic functional integrity of
131
sponse to antecubital fossa stimulation of the median nerve. this communication, however, remains to be documented.
When stimulating the median nerve at the wrist, the fibers trav- Although there is little doubt as to the presence of this neural
eling with the median nerve destined to join the ulnar nerve connection, the exact percentage of either median or ulnar
have previously departed the median nerve proximally and are muscle fibers innervated through this anastomosis is unknown.
no longer capable of summating with the AP and OP. Clinically, it is possible for the ulnar nerve, through the
Should a lesion of the median nerve exist at the wrist (e.g., Riche-Cannieu anastomosis, to supply in part the thenar mus-
carpal tunnel syndrome) with concomitant slowing of median cles normally supplied exclusively by the median nerve, thus
nerve conduction across this region, a somewhat different set of providing a dual innervation to these muscles. A complete
electrodiagnostic findings is noted. Stimulation of the median median nerve lesion at the wrist with clinical function of thumb
nerve at the wrist may result in a prolonged distal motor latency abduction and opposition has clearly been documented on mul-
to the APB. Proximal median nerve excitation, however, pro- tiple o c c a s i o n s . -
48
It is this connection that most likely ac-
241280
duces a relatively normal proximal motor latency. The resultant counts for the so called "all ulnar hand." In such cases, one may
NCV is significantly increased to that above normal (50 m/s) at anticipate that despite complete median nerve severance at the
times reaching 100 m/s or more. It is also possible for the prox- wrist, needle electromyography would reveal signs of denerva-
imal motor latency to actually be shorter than the distal motor tion in the APB and OP, but a variable number of voluntary
latency. Additionally, one may commonly note that the
153
motor units depending upon the extent of innervation by way
CMAP arising from proximal stimulation is not only larger than of the communicating ulnar-to-median branch. One could con-
that from the wrist, but also begins with a positive deflection. ceivably have an individual with both a Riche-Cannieu and
Proximal median nerve stimulation does not always require a Martin-Gruber anastomosis, thus allowing median nerve elbow
stimulation to produce a response from the APB and OP with nerve at the ankle. Irrespective of the stimulus strength at the
variable responses from wrist stimulation. In this case, a com- ankle, however, a larger response than that from the fibular head
plete severance of the median nerve at the wrist results in func- should not be obtained. Clinically, a complete lesion of the deep
tional preservation of the typically median-innervated thenar peroneal nerve in the leg may spare the EDB leading one to er-
muscles, i.e., an all ulnar hand. Failing to understand the Riche- roneously conclude that there is an incomplete peroneal nerve
Cannieu anomaly can lead to potential confusion. It may also be lesion. Failure to recognize this anatomic variant may also pre-
possible for the ulnar nerve itself to provide direct neural com- dispose one to conclude that some corrective surgical procedure
munication to all of the thenar muscles obviating the necessity may have led to improvement when indeed the accessory per-
of a Riche-Cannieu anastomosis, but this is a rare occurrence. 324
oneal nerve is providing the EDB's innervation. Preservation of
Finding signs of denervation in the ulnar-innervated muscles, the EDB and peroneus brevis and longus muscles despite lack
the APB, and OP would normally lead one to suspect the possi- of voluntary motor unit activity in the tibialis anterior or exten-
bility of a C8/T1 or brachial plexus injury. A complete diagnos- sor digitorum muscles should cause one to suspect the possibil-
tic assessment, however, failed to reveal such a lesion in just ity of an accessory deep peroneal nerve.
such a patient. Careful clinical examination in this person re-
80
It is relatively easy to confirm the presence of an accessory
vealed that the ulnar nerve demonstrated a lesion at the elbow. deep peroneal nerve. When one encounters a larger peroneal
Intramuscular recording techniques combined with selective nerve response from the EDB with fibular head compared to
neural blockade demonstrated a connection between the median ankle stimulation, one should first ensure proper activation of the
and ulnar nerve in the hand thus accounting for the APB and OP peroneal nerve at the ankle. Occasionally, the nerve may lie rather
findings. An initial awareness of this simple anatomic connec- deep to muscular and tendinous tissues. Increasing the current
tion would have resulted in the institution of proper treatment at strength and duration should result in an increase in the magni-
a more appropriate time interval. tude of the EDB's CMAP. If the response does not change, stimu-
lating posterior to the lateral malleolus while recording from the
Lower Limb EDB then results in a response. Summating the EDB's C M A P
Accessory Deep Peroneal Nerve. In the lower limb an ex- from both the anterior and lateral stimulation sites should slightly
tension of the superficial peroneal nerve, the accessory deep exceed that obtained from the fibular head. If signs of complete
peroneal nerve, proceeds posterior to the lateral malleolus to denervation (florid membrane instability and absent voluntary
then innervate the lateral portion of the extensor digitorum motor units) are noted in the muscles innervated by the deep per-
brevis (EDB) muscle. The EDB, therefore, can receive dual
361
oneal nerve except the EDB, one should consider the possibility
innervation from both the deep peroneal nerve and the acces- of an accessory deep peroneal nerve and proceed with the above-
sory deep peroneal nerve. Investigations of normal populations noted stimulation technique. Caution is recommended when
have revealed that the accessory deep peroneal nerve may exist stimulating posterior to the lateral malleolus and recording from
in up to 28% of persons with 57% of those individuals having it the EDB. When an excessively strong current is used, a volume-
bilaterally. ' - - There is also the suggestion that an auto-
123 151 196 244
conducted response from the foot intrinsic muscles other than the
somal dominant inheritance pattern may be present. This 54
EDB can be detected because of tibial nerve depolarization. The
anomaly may be a cause of confusion if the practitioner is un- initial positive deflection should result in the discarding of this
aware of this variant. Specifically, stimulating the peroneal potential and not concluding that an accessory peroneal nerve is
nerve about the fibular head typically results in a CMAP from present. However, a negative deflection may be observed because
the EDB with a smaller magnitude than that evoked from the of a spatial alignment of the foot intrinsic muscles' motor point
ankle. This is a result of less temporal dispersion of action po- with the recording electrode may occur. It is always a good prac-
tentials distally compared with proximal nerve stimulation. A tice to place a concentric needle electrode in the muscle to verify
larger amplitude proximally may lead one to consider the possi- that a volume-conducted potential is not present as the needle
bility of a less than supramaximal stimulation of the deep peroneal electrode is less prone to recording volume-conducted potentials.
T a b l e 5-4. Amplitude Changes along Nerve S e g m e n t

Distal CMAP Greater than Proximal CMAP Amplitude
Nerve Submaximal Costimulation Anatomic Variant Disease
Median At elbow Median and ulnar at wrist Ulnar t o median CB/dispersion

Ulnar At elbow Median and ulnar at wrist Martin-Gruber CB/dispersion
Peroneal At fibular head - CB/dispersion
Tibial At knee - CB/dispersion
Distal CMAP Less than Proximal CMAP Amplitude
Nerve Submaximal Costimulation Anatomic Variant Disease
Median At wrist Median and ulnar at elbow Martin-Gruber -

Ulnar At wrist Ulnar and median at elbow Ulnar t o median -
Peroneal At ankle Peroneal and tibial at knee Accessory peroneal -
Tibial At ankle Tibial and peroneal at knee - -
CB, conduction block.
Different causes of abnormal amplitude change along the major nerves of arm and leg. (Modified from: van Dijk JG: Motorisch geleidingsonderzoek. In van Dijk JG
(ed):EMG voor de algemeen neuroloog. Boerhaave Commissie, 1998, Leiden, with permission).
Table 5-4 summarizes the different possibilities for the major cutaneous fibers of the median nerve extending from the tip of
nerves that should be considered upon encountering an unex- the digits to about the interphalangeal region. However, the
pected higher or lower CMAP at distal or proximal stimula- motor fibers from the C8 and T l levels innervate the hand in-
tion. It is always wise to have an extra channel available to
365
trinsic muscles while C6-T1 cervical levels innervate the fore-
have the possibility of simultaneous recording an extra muscle arm and hand extrinsic muscles. The motor fibers of the median
to exclude or prove costimulation or the existence of an ana- nerve are commonly investigated by evaluating nerve conduc-
tomic variant. tion to the median-innervated thenar muscles (APB, OP, and
one-half of the flexor pollicis brevis). Conduction studies to the
first and second lumbrical muscles are possible, but not rou-
NERVE CONDUCTION TECHNIQUES tinely performed. When performing nerve conduction studies
98
to the thenar muscles, therefore, motor fibers arising primarily

Once the above-noted fundamental aspects of the peripheral from C8 and T l are evaluated. It is possible to evaluate the am-
nervous system and the various factors influencing peripheral plitude change across the transverse carpal ligament with stimu-
nerve action potential propagation have been mastered, it is ap- lation proximal and distal to the ligament. This method of
259
propriate to consider diagnostic nerve conduction studies. The median nerve assessment permits one to determine if there is
technical aspects of stimulating and recording potentials from conduction block of median motor fibers as they traverse the
the peripheral nervous system are relatively straightforward. carpal tunnel region. Caution must be exercised when perform-
The intent of this section is not to present an encyclopedic litany ing this study because it is relatively easy in some individuals to
of every NCS described, but to familiarize the practitioner with also activate the ulnar nerve's deep branch that lies just deep to
the more common diagnostic techniques available. Additionally, the median nerve's recurrent branch in the palm. Therefore, the
proper technical skills are stressed that can be applied to less CMAP morphology above and below the carpal ligament must
commonly used nerve conduction studies. Upper and lower be very similar, otherwise ulnar activation is likely and the re-
limb motor and sensory nerves are described as well as nerve sponse can not be accurately evaluated.
root stimulation and additional techniques. Recording Electrodes. For motor nerve conduction studies,
Although there are multiple techniques reported to evaluate E-l is preferentially located on the muscle's motor point under
motor and sensory fibers, very few control for temperature and investigation (Fig. 5-20). The motor point of most small mus-
distance, or clearly report the amplifier gain and filter settings. cles is approximately one-half the distance between the
The effects of distance and temperature on nerve conduction ve- muscle's origin and insertion. Muscles with long tendons pos-
locity have already been stressed. A lack of control for tempera- sess somewhat more difficult motor points to define, but these
ture and distance predisposes one to a less than standardized are mentioned in detail when discussed. Prior to securing the
methodology inviting inter-individual variation, thus diminish- recording electrodes to the patient, it is recommended that the
ing the sensitivity of the test. An attempt is made at including skin be slightly abraded to reduce the impedance and a mod-
only the investigations using a standardized distance where icum of electrode gel be applied. Enough tape or other securing
some mention of temperature is described. material should fasten the electrodes securely to the patient to
avoid movement artifact. E-2 is best located at or just beyond
the muscle's tendinous insertion away from E-l on the first
COMMON UPPER LIMB NERVE digit. Similar comments noted above are appropriate for all
recording electrodes irrespective of the nerve being studied and
CONDUCTION STUDIES
will not be repeated for each nerve discussed.
E-l. The E-l electrode is positioned on the thenar eminence
UPPER LIMB MOTOR NERVE CONDUCTION STUDIES
halfway between the midpoint of the metacarpophalangeal
A pertinent aspect of performing motor nerve conduction joint's volar aspect for the first digit (thumb), and the midpor-
studies is placing the E-l electrode on the muscle's motor point tion of the distal wrist crease (Fig. 5-20). This site should ap-
that is innervated by the nerve under investigation. The E-2 proximate the most prominent portion of the thenar muscles.
electrode is located on a relatively electrically silent region, i.e., Slight repositioning of E-l may be necessary to obtain an opti- j
the tendinous insertion of the muscle studied. A ground elec- mal CMAP response as defined by an immediate negative de- !
trode is then placed just proximal to E-l between it and the flection following median nerve excitation.
cathode. The stimulator employs a supramaximal pulse to evoke E-2. The actual distance between E-2 and E-l is really not
a CMAP from at least two and possibly more locations along critical provided the distance is not too close. The most conve-
the nerve's peripheral extent. The NCS's parameters of interest nient location is just distal to the insertion of the APB approach-
are: CMAP's amplitude, CMAP's onset latency, also called ing the first digit's distal interphalangeal joint (Fig. 5-20).
distal motor latency (DML), and NCV. Ground. The ground electrode should be secured to the pa-
tient following appropriate skin preparation to ensure a low skin
Median Nerve impedance. The most appropriate location for any ground elec-
The median nerve is a mixed nerve composed of motor fibers trode is adjacent to E-l between the stimulator's cathode and E-
arising from cervical and thoracic root levels C 7 - T 1 , and sen- 1. It is assumed that this is the position for all nerves examined
sory fibers from cervical levels C 6 - C 7 . The median nerve's
144
unless otherwise stated.
cutaneous distribution encompasses the volar aspect of the lat- Stimulation. The cathode is positioned over the median
eral two-thirds of the hand and three and one-half digits begin- nerve such that it is directed toward E-l while the anode is lo-
ning with the thumb. This nerve supplies C6 cutaneous fibers
206
cated away from E - l . This distal cathode and proximal anode
to the thumb and index finger while providing C7 fibers to the relationship is assumed to be the case in all following nerves in-
third and one-half of the fourth digits. Additionally, the dorsal vestigated unless specified. With respect to the median nerve,
aspect of the three and one-half digits are also innervated by the the distal site of stimulation is at the wrist between the tendons
of the flexor carpi radialis (most lateral of the prominent ten-

dons with slight wrist flexion) and palmaris longus (tendon just
medial to the flexor carpi radialis tendon) muscles. In the event
that the palmaris longus muscle is absent (normal anatomic
variant), the cathode is located just medial to the flexor carpi ra-
dialis tendon. The placement of the cathode is critical and
should be located at a standardized distance. A commonly used
technique is to note the E-l site and measure to the mid-wrist
crease. From this point the tape measure is directed proximally
along the course of the median nerve between the above-noted
two tendons until a total distance of 8 cm is reached (Fig. 5-20).
The 8-cm site is the desired location for the placement of the
cathode. The anode is typically aligned with the median nerve
proximal to the cathode. A supramaximal stimulus is applied to
the median nerve and a biphasic initially negative CMAP is the
desired response. The instrument's latency marker is positioned
at the beginning of the CMAP's negative deflection. If a small
positive deflection precedes the main C M A P response, E-l
should be slightly repositioned about the thenar eminence's
midpoint until a negative deflection is obtained. Should it be im-
possible to denote an initial negative deflection, the point of ini-
E-2
tial positive deflection is chosen with the latency marker. Figure 5-20. Median/ulnar nerve assessment. Diagrammatic
The above technique allows one to record a distal motor la- representation of recording and stimulating electrode locations for
tency. A second stimulus is typically applied to the median median and ulnar m o t o r nerve conduction studies. For the median
nerve at the antecubital fossa. The cathode is positioned just nerve the active (E-l) electrode is located mid-way between the origin
medial to the brachial artery pulsations. Again, a supramaximal and insertion of the abductor pollicis brevis (APB) m o t o r point, on t h e
impulse is delivered and the CMAP recorded. Once again the prominence of t h e t h e n a r mass. The reference (E-2) electrode is
instrument's latency marker is positioned at the onset of the en- placed distally on the first digit beyond the tendinous insertion of t h e
suing negative deflection. The distance between the two stimu- APB. For an ulnar nerve study,an E-l electrode is secured to the mid-
lus points is divided by the time difference between the distal portion of the abductor digiti minimi (ADM) while the E-2 electrode is
and proximal CMAPs to yield the NCV. It is also possible to situated distally on t h e fifth digit.The median nerve is stimulated 8 cm
excite the median nerve at more proximal sites to arrive at an proximally on a line attempting to trace the median nerve's anatomic
NCV for the median nerve's arm segment. course (dotted line connecting APB E-l t o cathode (-) between t h e
For both proximal and distal stimulation sites, a pulse dura- t e n d o n s of t h e flexor carpi radialis (FCR) and palmaris longus (PL)
tion of 0.1 or 0.2 ms is typically used. This allows one to easily muscles.The ulnar nerve is excited 8 cm proximal t o the ADM's E-l
apply a supramaximal stimulus at mid-range current intensities. on a straight line medial t o the tendon of the flexor carpi ulnaris
Occasionally, a person with significant muscle or adipose tissue (FCU) muscle. Gnd represents the ground electrode placement on t h e
may require an elevation of the pulse duration, particularly at dorsum of the hand.
the antecubital fossa, in order to supramaximally stimulate the
nerve.
If stimulation of the median nerve in the mid-palm is desired, location is quite close to the recording electrodes, stimulus arti-
it can be accomplished but it may be rather difficult to excite fact may pose a significant problem. This can usually be solved
solely the median nerve because the deep branch of the ulnar by ensuring all palmar perspiration has been removed and by
nerve lies immediately beneath the median nerve. The mid- rotating the anode about the cathode thus eliminating a large
palm motor stimulation site is approximated by asking the pa- portion of the stimulus artifact through differential amplifica-
tient to touch the base of the thenar eminence with the tion. A needle electrode can also serve as the cathode by plac-
192
ipsilateral fourth digit. The site of digital contact denotes the

168
ing it just under the skin thereby reducing the amount of current
anatomic location for the recurrent branch of the median nerve; required as well as the stimulus artifact.
motor nerve innervating the median supplied thenar muscles. I n s t r u m e n t P a r a m e t e r s . Three instrument parameters are
The cathode is positioned at this site with the anode distal. of interest with respect to performing optimal CMAP record-
Slight repositioning of the cathode is then performed until the ings: amplifier sensitivity, CRT sweep speed, and filter settings.
amplitude of the response is maximized. Because the palmar The reference data noted below are obtained at an amplifier sen-
fascia and skin are rather thick, several cautionary notes are nec- sitivity of 1,000 pV/div. It is important to recognize that the ma-
essary when attempting a mid-palm median nerve stimulation. jority of normal individuals possess CMAPs that exceed the
When a surface as opposed to needle cathode is employed, a maximum amplitude capable of being displayed at this sensitiv-
pulse duration of 0.2 ms or slightly more may be required. Care ity. As a result, the entire CMAP is not visualized on the C R T
must be exercised to not concomitantly activate the ulnar nerve. screen. Should one wish to view the entire potential, a less sen-
This is assured by comparing the mid-palm CMAP morphology sitive amplifier setting is required but one can no longer use the
with that obtained by median nerve stimulation at the wrist. same latency data. Two separate amplifier settings are required
Should the appearance of these two potentials differ, less cur- with this technique: (1) 1,000 pV/div to record latency, and (2)
rent or a diminution in the pulse duration should be applied. 2,000 or 5,000 pV/div to measure amplitude.
Repositioning the cathode slightly may also help to localize just An appropriate sweep speed is necessary to assign enough
the recurrent branch of the median nerve. Because the stimulus points of resolution to the screen to optimally resolve the potential.
T a b l e 5-5. Median N e r v e : M o t o r are necessary to assess the electrical integrity of the nerve
DML (ms) Amplitude (mV) NCV (m/s) through the above-noted areas.
One of the major sources of contention regarding ulnar nerve
3.7 ± 0.3 (3.2-4.2) 13.2 ± 5.0 (5-25) 56.7 ± 3.8 (50.0-67.3)
conduction studies is the position of the elbow. Because the
8.8 ± 3 . 1 (3.5-15)
ulnar nerve can be compromised at the level of the elbow, it be-
Mid-palm Amplitude (% change)*
comes necessary to determine the optimal position of the elbow
8 ± 8.5 (-7 to - 4 3 )
when performing nerve conduction studies across this segment.
Temperature was not recorded for these values. Amplitudes are measured There is noted to be a significant difference in conduction ve-
peak-to-peak except for mid-palm study. Amplifier sensitivity is set at 1,000 locity across the elbow depending on whether the elbow is
u.V/div, and filter settings are 8 Hz to 8kHz. - 233 234
* Amplitude for transcarpal vaiues are reported in percent change with any flexed or extended, and to what degree. The nerve conduction
change greater than 25% considered abnormal. DML, distal motor latency;
259 velocity is markedly slower with elbow extension compared to
NCV, nerve conduction velocity. flexion. This finding is believed to occur because of the nerve
length required to extend across the elbow in the flexed posi-
Generally, the recorded potential should appear near the middle tion without rupturing and subsequently becoming slack and
third of the CRT so as to avoid the stimulus artifact contaminat- redundant with elbow extension. Although the length of nerve
41
ing the desired response. A sweep of 2 or 5 ms/div typically does not significantly change between a point above and below
suffices for most routine studies of the forearm. Should exami- the elbow, a discrepancy occurs with respect to the surface dis-
nation of the arm be required, a sweep speed of 5 ms/div is pre- tance measured in flexion versus extension. In extension, the
ferred to visualize the response. measured distance is shorter compared to the same points when
Perhaps the most variable and least universal instrument pa- . the elbow is in flexion. When calculating a nerve conduction
rameter is the high- and low-frequency filter settings. The band- velocity, essentially the same time of conduction for both posi-
width noted for the included reference data is 8 Hz to 8 kHz. tions is divided into a longer distance for flexion compared to
Some variation about these numbers should not dramatically extension thus resulting in a larger value for the NCV. There is
alter the CMAP morphology provided the low- and high-fre- no change in the CMAP's duration or amplitude when recorded
quency responses do not restrict the bandwidth in excess of 10 in flexion compared to extension. The only parameter that
41
Hz to 5 kHz. changes is the NCV because of the difference in distance mea-

Reference Values. The reference data utilized in this text is sured when the elbow is flexed. Normal values are provided for
derived using the above noted instrumentation parameters both extension and flexion for the across elbow ulnar nerve
(Table 5 -5). - Any significant deviation from the instrument
233 234
segment (Table 5 - 6 ) . A comparison of nerve conduction in
4 U 8 4
settings requires the practitioner to develop a new set of refer- the distal segment of the ulnar nerve is also provided. Because
ence data. Mid-palm amplitudes are noted, however, latencies of a possibility of compromise preferentially involving the
were not reported. 259
deep branch of the ulnar nerve distal to the innervation of the
abductor digiti minimi, a technique is described comparing la-
Ulnar N e r v e tency differences between the abductor digiti minimi and first
The ulnar nerve is comprised of motor and sensory fibers dorsal interosseous.
originating primarily from C8 and T l cervical root levels. 144
Cutaneous fibers of the ulnar nerve innervate the volar aspect of Ulnar Nerve Conduction: Forearm, Elbow, and A r m
the palm as well as the fifth and one-half of the fourth digits. 206
Recording Electrodes. The first technique involves ulnar
The cutaneous sensibility also innervates the dorsal aspects of nerve conduction in the forearm, across elbow, and arm seg-
these digits to approximately the proximal interphalangeal joint ments recording from the abductor digiti minimi muscle. A
region. The dorsal aspect of the hand and remaining portion of second technique investigating conduction in the terminal ex-
these digits is innervated by the dorsal ulnar cutaneous nerve. panse of the ulnar nerve within the hand records a CMAP from
There is believed to be a motor contribution from the C7 level, not only the previously noted muscle but also the first dorsal
particularly innervating the flexor carpi ulnaris. The muscles in- interosseous. 252
nervated by the ulnar nerve in the forearm are the flexor carpi E-l. An E-1 electrode is secured to the motor point of the ab-
ulnaris and the flexor digitorum profundus to the fourth and ductor digiti minimi. This is accomplished by locating the
fifth digits. Anatomic variants to the flexor digitorum profundus halfway point between the distal wrist crease or pisiform bone
exist, but are not discussed in this portion of the text. All of the and the metacarpophalangeal joint of the fifth digit (Fig. 5-
hand intrinsics are innervated by the ulnar nerve except for 20). -233.234 y j distance is measured on the most medial aspect
41
n s
those previously noted to be supplied by the median nerve. This of the hand over the muscle's main bulk.
nerve is essentially a direct extension of the brachial plexus' E-2. The E-2 electrode can be positioned over the distal
lower trunk and medial cord. The nerve is secured at the inter- aspect of the digit or just distal to the metacarpophalangeal
muscular septum in the arm (arcade of Struthers) and somewhat joint. A "bar" recording electrode should not be used to ensure
restricted between the two heads of the flexor carpi ulnaris the reference electrode is not located over muscle tissue.
muscle. Between these two anatomic regions the nerve is rela- Stimulation. The ulnar nerve can be stimulated in at least
tively exposed about the ulnar groove at the elbow in that it is four locations to evaluate conduction over various neural seg-
subcutaneous and vulnerable to trauma. There is also a possibil- ments. 41184
Let us first assume that the elbow is flexed. For dis-
ity that the ulnar nerve may become entrapped within the sub- cussion purposes, within this text the elbow is fully flexed to
stance of the flexor carpi ulnaris muscle. A second possible site approximately 135° with the forearm supinated (Fig. 5-21). The
of ulnar nerve compromise is at the wrist within the canal of arm and forearm are positioned for maximum patient comfort.
Guyon as the ulnar nerve traverses the region between the pisi- One way to achieve this goal is to have the patient abduct the
form bone and hook of the hamate. Because of these possible
311
arm 90° and rest it on a pillow such that the hand approximates
anatomic regions of compromise, electrophysiologic techniques the ipsilateral ear. This position exposes the desired stimulation
Table 5-6. Ulnar Nerve: Motor

DML NCV Amplitude Duration
(ms) (m/s) (mV) (ms)
Elbow Flexed 110°
Segment
Wrist 3.2 ± 0.5 6.1 ± 1.9 4.2 ± 0.5
Forearm 61.8 ± 5.0 5.6 ± 1.9 4.2 ± 0.5
Across Elbow 62.7 ± 5.5 5.8 ± 1.8 4.2 ± 0.5
Elbow Flexed 135°
Forearm 63.3 ± 5.2 I 1.2 ± 2 . 1
Across Elbow 62.8 ± 7.1 10.8 ± 2.1
Axilla 61.9 ±6.0 10.4 ± 2 . 0
Above Elbow 63.0 ± 4.7
t o Wrist
Elbow Extended
Wrist 6.1 ± 1.6 4.2 ± 0 . 5
Forearm 62.5 ± 4.5 5.3 ± 1.9 4.2 ± 0.5
Across Elbow 49.9 ± 7.9 5.4 ± 1.5 4.3 ± 0.5
Elbow Extended
Forearm 65.7 ± 6.7
Across Elbow 50.3 ±5.9
Axilla 60.9 ± 7.0
Above Elbow 59.0 ± 4.9
t o Wrist
Segmental Difference: < 11.4 ms
(Elbow-Forearm)
ADM/FDI240
Figure 5-2/. Ulnar nerve stimulation sites. Stimulation sites for ADM 2 FDI (ms) < 2.0 ms
ulnar nerve (arrows) m o t o r and sensory conduction at wrist, across L/R FDI (ms) < 1.3 ms
elbow, and arm. Ring electrodes on t h e fifth digit and surface elec-
Filter settings for A and C are not stated but are noted to be 8 Hz-8 kHz with a
t r o d e s on the abductor digiti minimi can be used for the same three
sensitvity of 1,000 JIV/ cm, while B and D used filters of 1.6 Hz-16 Hz and a sensi-
stimulation sites. N o t e that the arm is abducted and externally rotated tivity of 5 mV/cm. Instrumentation parameters for F were not provided.The
4,J84
with the forearm slightly supinated.This positioning allows adequate amplitude differences between A and C and B and D are because the smaller
exposure for all stimulation sites. values are measured base-to-peak, while the larger values are peak-to-peak. ADM
= abductor digiti minimi; FDI = first dorsal interosseous; L/R = left-right difference.
sites while simultaneously relaxing the patient and allowing the ulnar groove. The two stimulus points should first be delin-
easy measurement of the desired neural segments. eated on the patient. A tape measure is then secured to the
It is desirable to only stimulate the nerve at fixed distances to below-elbow site and carefully placed just posterior to the
minimize anatomic variability and inter-observer variation due medial malleolus and then proximally along the medial aspect
to distance measurement errors and uncertainty. One particular of the arm to the above-elbow site. In the flexed position, one
study is used for uniformity of data; however, one may wish to may first wish to optimize the ulnar nerve CMAP with respect
use slightly different distances for the distal site as this should to amplitude to accurately locate the ulnar nerve as it is often
not significantly affect the nerve conduction velocity. surrounded by subcutaneous tissue impeding ulnar nerve palpa-
Specifically, a distance of 6.5 cm proximal to E-l along the tion. Several measurements may be necessary to ensure the
85
ulnar nerve at the wrist is one recommendation but 8.0 cm may most accurate distance.
also be u s e d . The only caution is to use the appropriate ref-
41184
The ulnar nerve can also be excited at the above-noted four
erence data for distal latencies for each respective distance. A stimulation sites with the elbow in 180° of e x t e n s i o n . 41184
second site of stimulation is 4 cm distal to the medial epi- Unlike elbow flexion, the stimulation sites are less well exposed
condyle. The third stimulation point is at least 10 cm proxi-
184
and the examiner experiences somewhat more difficulty in at-
mal to the below elbow stimulation area. Finally, the fourth tempting to measure and stimulate the appropriate neural seg-
position for the stimulator is 11 cm proximal to the above-elbow ments. For localizing an entrapment of the ulnar nerve at the
site. With this technique one can assess 5 separate neural seg- elbow it may be necessary to use a shorter segment over the
ments: (1) forearm, (2) across elbow, (3) arm, (4) above elbow elbow in order to measure a slowed conduction velocity. It is
to wrist, and (5) wrist to recording electrode. Of course, should also possible to stimulate the ulnar nerve in short segments of 1
the clinical situation arise, it is relatively easy to also stimulate cm over a presumed lesion to help better localize the exact site
the ulnar nerve in the axilla or its constituent fibers from supra- of pathology. If short segmental stimulation is to be used, it is
169
clavicular activation. It is necessary to discuss the measurement essential to ensure accurate nerve activation. Recall that neural
technique across the elbow. excitation of a nerve that lies deep to muscle and subcutaneous
A certain amount of practice is required to accurately mea- tissues (ulnar nerve deep to the flexor carpi ulnar muscle) may
sure the across-elbow segment with the elbow fully flexed. It is require an elevated stimulus intensity/duration, which predis-
best to pursue the anatomic course of the nerve as it traverses poses to errors between the presumed and actual site of nerve
stimulation. In the authors' opinion, it is best to perform short loss as opposed to demyelination. Therefore, the needle elec-
segmental stimulation with a needle cathode to minimize the tromyographic examination usually detects an abnormality
amount of current spread between the site of skin penetration prior to the latencies reaching the above-noted abnormal values.
and neural activation.
Instrument Parameters. The amplifier sensitivity can vary Radial N e r v e
between 1,000 and 5,000 pV/div and is specified in the table de- The radial nerve is a direct continuation of the posterior cord
lineating the reference data. A sweep speed of 5 ms/div is as- of the brachial plexus. Cervical root levels C5-C8 and occasion-
sumed for most motor studies particularly when examining the ally T l comprise the neural fibers forming the radial nerve. 144
more proximal neural segments. Filter settings are important and After traversing the spiral groove about the humerus, the radial
should be reproduced with respect to the reference data used. nerve enters the forearm. The superficial radial nerve, a pure
Reference Values. Two sets of reference data are included sensory nerve, arises from the main trunk of the radial nerve fol-
for both the flexed and extended position (Table 5 - 6 ) . 4,J84
lowing innervation of the extensor carpi radialis longus. This
Despite different filter and gain settings there is good agreement nerve provides cutaneous sensibility to the dorsum of the hand
between the two studies. A value of less than 11.4 m/s is noted not innervated by the median and ulnar sensory nerves noted
when comparing across-elbow conduction velocities with those above. The continuation of the radial nerve innervating the re-
in the forearm. 184
mainder of the hand's extrinsic extensor muscles is referred to
as the posterior interosseous nerve. The terminal muscle inner-
Ulnar N e r v e Conduction: Hand vated by the posterior interosseous nerve (C7 and C8) is the ex-
Recording Electrodes. A lesion affecting the deep palmar tensor indicis proprius.
branch of the ulnar nerve may occur distal to the innervation of Recording Electrodes. The placement of recording elec-
the ADM but prior to the innervation of the AP or FDI. In this trodes for this nerve is somewhat more technically demanding
case, it would be of value to compare the time of conduction than for the median or ulnar nerves. This is because the usual
through the potentially affected deep branch to the motor inner- site of E-l placement, extensor indicis proprius, is not as
vation to the ADM. A comparison technique is described to anatomically prominent as the thenar or hypothenar groups.
assist in the diagnosis of possible injury to this nerve. Bony anatomic landmarks are therefore necessary to ensure
E-l. The E-l electrode is located on the abductor digiti proper electrode placement. The commonly used techniques re-
minimi as noted previously (Fig. 5-20). A second E-l electrode quire that a standard concentric needle electrode be placed
is placed over the first dorsal interosseous over the muscle's within the extensor indicis proprius muscle. A needle electrode
major protuberance. located within a muscle helps to ensure that the observed re-
E-2. An E-2 for the abductor digiti minimi is positioned on sponse arises from only the desired muscle. However, because
the fifth digit distal to the muscle's insertion. The recommended of a needle E-l electrode, amplitudes are of limited value par-
position for the first dorsal interosseous muscle's E-2 is on the ticularly in attempting to define axonal loss or conduction
second metacarpophalangeal joint. One may wish to place E-
252
block. Recall that the muscle as a whole is not evaluated but
2 distal to the first metacarpophalangeal joint; however, in order only the fibers located within a short distance of the E-l record-
to reduce an initial positive deflection occasionally observed ing surface.
with the previously noted E-2 placement. The use of surface electrodes to record the radial nerve re-
Stimulation. The ulnar nerve is stimulated at the distal wrist sponse has been suggested; - however, one must be prepared
215 216
crease in the original study. Although a standard distance is pre- for a positive deflection as it is very difficult to locate the exten-
ferred, it is acceptable to use an anatomic location in this in- sor indicis proprius or any other radial nerve muscle's motor
stance because the parameter of interest is the latency difference point. This positive deflection should then be observed with all
between the two muscles. The same stimulation site is used for CMAP's from proximal stimulation sites. Additionally, the sur-
each recording. A supramaximal stimulus evokes a CMAP from face electrode located in the distal forearm is subject to record-
each muscle and the onset latency from the abductor digiti ing volume-conducted potentials from a different set of
minimi muscle is subtracted from the onset latency to the first radial-innervated muscles as the stimulus site progresses more
dorsal interosseous muscle. A two-channel recording is helpful proximally. This is unlike the median and ulnar nerves where ,
because fewer stimuli are required. the hand intrinsic muscles are relatively isolated from the fore- !
Instrumentation Parameters. These data were not speci- arm muscles and somewhat less prone to detecting volume-con-
fied in the original study. However, as long as the same parame- ducted electrical activity.
ters are used for both latencies the same time difference should E-l. The central core of a standard concentric needle elec-
apply. Similar instrumentation parameters to those used for the trode serves as E-l for recording the radial nerve CMAP. This
median nerve are recommended. needle electrode is inserted into the extensor indicis proprius
Reference Values. The difference between the two onset la- muscle by measuring approximately 4 cm proximal to the ulnar
tencies should not exceed 2.0 ms. Additionally, when this tech- styloid and 1 or 2 cm toward the radius, i.e., just lateral to the
nique is applied bilaterally, the time difference between the tendon of the extensor carpi ulnaris muscle. The patient is re-
159
left/right first dorsal interossei's onset latencies should not quested to repeatedly extend and relax the second digit while
exceed 1.3 ms (Table 5-6). A time difference exceeding those
250
the examiner palpates for muscle contraction. - A monopolar
159 334
noted above suggests that the deep branch of the ulnar nerve needle electrode also may be used. 183
displaying the prolonged conduction time is compromised distal E-2. If a standard concentric needle electrode is used, the
to the innervation of the hypothenar muscles. needle's cannula serves as E-2. Should a monopolar needle
In the authors' opinion, the latency between these two mus- serve as E-1, a surface E-2 may be located on the fifth metacar-
cles is usually normal even in persons with lesions. The major- pophalangeal joint. 183
ity of canal of Guyon lesions examined by the authors have Stimulation. The radial nerve may be excited in several lo-
been a result of a ganglion cyst and produced primarily axonal cations. For the technique described in this text, a needle electrode
is used for stimulation purposes. At each designated stimulation filter approximating 10 Hz and a high-frequency filter in the
site a monopolar needle can be used as the cathode, and a sur- range of 10 kHz should suffice.
face anode located at a convenient site near the cathode. The Reference Values. The reference data provided are applica-
original description of the technique used a monopolar needle ble only when recording with intramuscular needle electrodes
with 3 mm of Teflon removed from the tip. A pulse duration of (Table 5-7). The amplitudes are measured peak-to-peak.
334
0. 2 ms at an intensity of 2-4 mA was used. It is also possible to Onset latencies are to the initial deflection of the recorded re-
j use a standard monopolar needle, but reduce the pulse duration sponse irrespective of its polarity.
to 0.05 m s . Slightly different currents may be necessary as a
260
reduced stimulus duration is delivered to the nerve. Whenever a Radial Nerve: Proximal S e g m e n t
needle electrode is chosen as the cathode, one optimally posi- The radial nerve is the continuation of the posterior cord and
tions the electrode by delivering a stimulus that is supramaxi- is the largest branch of the brachial plexus. Nerve fibers origi-
mal but with the smallest current required to achieved the result. nating from cervical segments C 5 - C 8 and possibly from T l
This is accomplished by moving the needle in small increments comprise this peripheral n e r v e . All extensor muscles of the
144324
with each pulse delivery to ensure near-nerve placement, i.e., a arm and forearm are innervated by the radial nerve. In the axilla
supramaximal response with the minimal amount of current and proximal portion of the arm, several branches are given off
necessary. the main radial nerve trunk to innervate the various heads of the
The first stimulus site is in the forearm 8 cm proximal to the triceps muscle. Examination of conduction in the radial
324
ulnar styloid just lateral to the extensor carpi ulnaris muscle. nerve's distal segment has been described previously; however,
The radial nerve is subcutaneous for approximately 5 cm at this in this section we examine conduction from Erb's point to the
region and should be excited easily. A second stimulus loca-
159
triceps muscle, i.e., exclusively the proximal neural segment of
tion for the radial nerve is at the elbow region in the groove the radial nerve to the triceps muscle.
formed by the lateral margin of the biceps brachii tendon and Recording Electrodes. Because the triceps muscle does not
the medial border of the brachioradialis muscle 6 cm proximal have a well defined motor point as detected by surface record-
to the lateral epicondyle. Radial nerve excitation is also per- ing electrodes, a standard concentric or monopolar needle elec-
formed in the axillary region in the depression formed by the trode is preferred for recording.
coracobrachialis and medial border of the triceps muscles. E-l. The needle electrode is located deep within the sub-
When performing stimulation at these three sites it is impera- stance of the triceps muscle. Three mean distances are examined
tive that the same CMAP be obtained for each neural location from Erb's point as measured with obstetric calipers to assist in
to ensure that the recording electrode did not move signifi- the examination of persons with different arm lengths. These dis-
cantly. The above three stimulus locations allow the practi- tances are: 21.5 cm, 26.5 cm, and 31.5 cm (Table 5-7). 105
tioner to examine two separate segments of the radial nerve, E-2. The standard concentric needle electrode's cannula
1. e., arm and forearm. Surface stimulation may replace needle serves as the E-2 electrode, while a nearby surface electrode
excitation; however, a longer pulse duration approximating 0.5 functions as E-2 in surface or monopolar needle recordings.
ms may be necessary as the radial nerve is commonly sur- Stimulation. A surface cathode and anode are located at
rounded by significant muscle and adipose tissues. A fourth Erb's point. The pulse duration should be between 0.5 and 1.0
stimulus location can be performed at Erb's point. Erb's point ms with an intensity capable of producing a supramaximal re-
is just lateral to the insertion of the sternocleidomastoid muscle sponse. Firm pressure applied at Erb's point is often necessary
and a few centimeters superior to the clavicle, i.e., supraclavic- to evoke an optimal response.
ular fossa. Surface stimulation should only be attempted in this Instrumentation Parameters. A sweep speed of 2 ms/div
region to avoid pneumothorax induction from a needle cath- and sensitivity capable of displaying the entire response on the
ode. A pulse duration of 0.5 or 1.0 ms may be necessary to screen are sufficient to obtain the desired responses. Also, low-
completely activate the posterior cord from which the radial and high-frequency filters of less than 10 Hz and greater than or
nerve arises. Firm pressure on the stimulator is typically neces- equal to 8 kHz, respectively, are employed.
sary to ensure that the cathode and anode are as close to the
brachial plexus as possible.
T a b l e 5-7. Radial N e r v e : M o t o r
Instrumentation Parameters. When measuring the dis-
tance across the brachial plexus (Erb's point to axilla) or arm DML NCV Amplitude Distance
segment (axilla to elbow) the use of obstetric calipers is recom- (ms) (m/s) (mV) (cm)
mended to more accurately reflect the distance traversed by the Segment
nerve. - - Recall that the radial nerve does not pursue a
160 193 334
straight course in the arm but passes posterior to the humerus in Erb's-Axilla 72 ± 6.3
the spiral groove. About 10° of arm abduction may assist in ex- Axilla-Elbow 69 ± 5.6 11 ± 7.0 15.7 ± 3.3
amining the proximal segment of the radial nerve. Optimally,
Elbow-Forearm 62 ±5.1 13 ± 8.2 18.1 ± 1.5
the forearm is fully pronated and the elbow should not be flexed
more than 10° or 15°. Because a needle electrode is located
159
Forearm-ElP 2.4 ± 0.5 14 ± 8 . 8 6.2 ± 0.9
within the muscle tissue, a rather large response can be expected
Erb's-Triceps 4.5 ± 0 . 1 20-30
that requires a sensitivity of about 5 mV/div. It is important to
4.9 ± 0 . 1 25-28
emphasize that the amplitude as recorded with needle electrodes
5.3 ± 0 . 1 2 30-33
is of limited diagnostic value with respect to assessing axonal
loss or conduction block because the entire muscle is not evalu- Filter setting are not specified but those approximating a low filter of 10 Hz
ated. Sweep speeds of either 2 or 5 ms/div are appropriate de- and a high filter of 10 kHz should suffice. Erb's data from Jebsen, remainder
159160
from Trojaborg and Sindrup. Amplitudes are measured peak-to-peak and

1053 3 4
pending upon the length of the subject's arm. Filter settings onset latencies to the recorded potential's initial deflection. Performed with in-
were not provided in the quoted study ; however, a low-frequency tramuscular needle recordings.
2 0 0 — PART II BASIC AND ADVANCED TECHNIQUES
Reference Values. The onset latencies to the initial deflec- medial 1-2 cm of the thenar eminence is supplied by this
tion of the recorded response are documented for the above- nerve. A lateral branch of this nerve also innervates the proxi-
81
noted three distances (Table 5-7). 105

mal and lateral hypothenar eminence. The remainder of the
hand and digits are innervated by the terminal sensory branches
Phrenic Nerve of the median nerve distal to the carpal tunnel. Additionally, the
The phrenic nerve is a branch of the cervical plexus com- distal two-thirds of the dorsal aspect of the first three and one-
posed of fibers originating from the anterior primary rami of hajf digits are also supplied by the cutaneous braches of the
C3/C4/C5 with occasional contribution from C2 or C 6 . In the 144
median nerve.
cervical region, the nerve courses medially, anterior to the ante-
rior scalene muscle to lie posterior to the sternocleidomastoid Antidromic Techniques
muscle. Once it enters the thoracic cavity, the phrenic nerve As previously defined, antidromic refers to the direction of
travels between the pericardium and pleura of the mediastinum. induced impulse propagation with respect to the recording elec-
The nerve then pierces the diaphragm to innervate this muscle. trodes and the direction of naturally elicited action potential
One may wish to examine this nerve in persons suspected of propagation. In this section, antidromic implies that E-l is lo-
having traumatic or acquired damage to the phrenic nerve. cated distal to the cathode. The action potential generated be-
Recording Electrodes. There are several techniques for neath the cathode, therefore, is recorded at a more distal
recording the C M A P from the phrenic nerve. - - One can 59 218 223
location. Physiologically, however, an action potential induced
locate the recording electrodes laterally over the ribs in the mid- by natural stimuli would have propagated from the recording
axillary line or centrally over the xyphoid process. Our pre- electrodes to the cathode. In this sense, the recording is per-
ferred technique is placement of the active electrode proximal formed such that the induced action potential travels in the op-
to the xyphoid process. - The patient should be supine without
20 42
posite direction (anti) to what it would naturally take.
a pillow supporting the head. Both wrist and mid-palm antidromic stimulation techniques
E-l. A surface E-l electrode is secured 5 cm proximal to the are described. It has been suggested that in median nerve en-
xyphoid process following appropriate preparation of the skin. trapment at the wrist, selectively examining median nerve sen-
This location suffices for both left- and right-sided phrenic sory conduction across the transverse carpal ligament increases
nerve excitation. As a surface recording is used, the recorded the diagnostic sensitivity of neural conduction. - Although
155 362
CMAP yields a response appropriate for side-to-side amplitude the original study examined subjects' third digit, it is reasonable
comparisons. to assume that the same technique also can be applied to the
E-2. The surface E-2 recording electrode is best located ipsi- second digit.
lateral to the side of stimulation at the costal margin approxi- Digits II and III. The straight anatomic course of the
mately 16 cm distal to the E-l electrode. median nerve's sensory fibers from the wrist to the second and I
Stimulation. A surface stimulator is located just posterior to third digits allows one to easily study sensory conduction. 164
the posterior border of the sternocleidomastoid muscle at the Either the second or third digit can be used for investigating the
level of the cricoid cartilage or just above the clavicle. - It is 20 59
median nerve's SNAP. As previously noted for motor fibers, we
not uncommon with this placement to concomitantly excite the prefer techniques that use standard distances to those employ-
brachial plexus. If this occurs, the stimulator should be reposi- ing anatomical landmarks. Although there are multiple investi-
tioned until the patient describes a hiccup-like sensation and an gations examining median nerve sensory conduction, very few
abdominal pulsation may be palpated. 216
investigators standardize the distance over which the nerve is
Instrumentation Parameters. See radial nerve stimulation. recorded.
Reference Values. Mean onset latencies for the CAMP are Recording Electrodes. Ring or clip electrodes are secured
6.54 ± 0.77 ms (5.5-8.4 ms) with base-to-peak amplitudes of to the fingers and used for both E-l and E-2. Several technical
660 ± 201 pV (301-1198 pV). 42
comments are relevant to the application of the ring recording
electrodes. It is preferable to rotate the electrodes a small
UPPER LIMB SENSORY NERVE CONDUCTION STUDIES amount while at the same time applying pressure prior to finally
securing the electrodes to the digit to help reduce the skin's im- |
The more common sensory nerve conduction techniques of pedance. A small amount of electrolyte gel is then placed along s
the major sensory nerves in the upper limb are described. Both the ring electrode and it is again slightly rotated to ensure an
orthodromic and antidromic methods of eliciting SNAPs are equal amount of recording gel around the circumference of the
presented. We do not subscribe to the philosophy that an- electrode. The open portion of the ring electrode should be pos-
tidromic studies are somehow inferior to orthodromic studies terior, thus maximizing contact between the active recording
because of an occasional motor artifact possibly interfering with surface and the median nerve. A piece of tissue paper should
the SNAP. Anyone attempting to practice electrodiagnostic separate the ring electrodes from the adjacent digits to minimize
medicine should be familiar with both methods of examining movement artifact secondary to the electrodes contacting other
peripheral sensory nerves to be able to use an alternative tech- fingers following stimulation.
nique when a given one yields less than optimal results. E-l. The E-l ring electrode is secured to either the second or
third digit as noted above (Fig. 5-21). The actual placement
Median N e r v e should be 1-2 cm distal to the metacarpophalangeal joint to
The median nerve provides cutaneous sensibility to the lat- minimize volume-conducted motor responses from the acti-
eral aspects of the palm as well as to the volar areas of the first vated hand intrinsic muscles.
three and one-half digits. - The first cutaneous branch of the
144 206
E-2. An E-2 ring electrode should be positioned 4 cm or
median nerve is the palmar cutaneous branch of the median more distal to E-l to avoid amplitude reduction and latency
nerve. This nerve innervates the skin just distal to the distal shortening secondary to differential amplification effects of
wrist crease for a distance of several centimeters. Only the common mode signals being recorded (Fig. 5-22). - - 78 79 80
Figure 5-22. Diagrammatic representation of electrode placement for antidromic median and ulnar nerve sensory conduction
studies. A, For median nerve examination the E-l recording electrode is located on the second o r third digit just distal t o t h e metacarpopha-
langeal joint region and t h e E-2 electrode is placed at least 4 cm more distal on t h e respective digit.The median nerve is excited 7 cm (S ) and 14 2
cm (S|) proximal t o E-l between t h e tendons of t h e FCR and PL. B.With ulnar nerve studies, t h e E-l is positioned on t h e fifth digit just distal t o
the metacarpophalangeal joint with t h e E-2 electrode as far distal as possible. Stimulation (S) of t h e ulnar nerve is performed 14 cm proximal and
medial t o the tendon of the FCU.
Ground. As for motor studies, the ground electrode should SNAP as it occurs within several milliseconds of the stimulus
always be located between E-1 and the cathode, preferably ad- artifact. The relatively small size of the SNAP requires an am-
Ijacent to E-l. This is the assumed position for all remaining sen- plifier sensitivity of 10-20 pV/div. Occasionally, an exception-
sory studies noted in this text. Alternatively, placing the ground ally large SNAP may require a sensitivity of 50 pV/div. Filter
on the dorsum of the hand is acceptable, provided stimulus arti- settings approximating 20 Hz to 2 kHz should yield clearly re-
fact is minimal. producible SNAPs.
Stimulation. The median nerve is easily excited at the wrist Reference Values. Distal sensory latencies were recorded to
14 cm proximal to E-l measured on a straight line. The cathode the peak of the negative deflection or the onset latency of the
is positioned between the tendons of the flexor carpi radialis lat- initial deflection while amplitudes are calculated from the ini-
erally and the palmaris longus muscles medially. A pulse dura- tial deflection to the major negative spike (Table 5-8). Negative
tion of 0.1-0.2 ms using a supramaximal current intensity is spike durations were also noted. The mid-palm temperature for
necessary. The anode is located proximal to the cathode. In this the data reported exceeded 31 °C.
arrangement, the cathode is directly opposite E-l on a 14 cm
line between E-l on the digit and the cathode proximally. "Inching Technique"
Should a mid-palmar stimulation be necessary, the cathode is Both orthodromic and antidromic sensory studies clearly
located 7 cm proximal to E-l (anode proximal) (Fig. 5-22). This demonstrate that there is often focal slowing across the trans-
stimulation site optimally falls in the mid-palm region. If the verse carpal ligament in the carpal tunnel s y n d r o m e . - A 31 57179362
actual measured distance of 7 cm places the cathode near the technique exciting median nerve sensory fibers in 1-cm incre-
distal aspect of the transverse carpal ligament, the recording ments across the wrist and into the hand while recording from
electrodes should be repositioned to a more distal location thus the second digit is described, i.e., short-segment stimulation. 180
localizing the mid-palm stimulation more toward the middle of The intent of this type of stimulation is to localize an area of
the palm. The antidromic technique quoted used a needle cath- nerve that displays focal slowing thus suggesting a correlation
ode, but the authors have used a surface cathode with little diffi-
culty. When stimulating in the mid-palm with a surface cathode,
the pulse duration may need to be increased slightly above that Table 5-8. Median Nerve: Sensory
used at the wrist because of subcutaneous tissue thickness in OL (ms) ' 49 6 4
PL ( m s ) 49164
Amplitude (pV) -
49 234
Duration (ms) 49
most persons' palmar regions. Care must be exercised to not el-

Antidromic
evate the pulse duration too much, otherwise a more distal site
of excitation than anticipated may occur. The use of a needle 2.4 ± 0 . 3 3.0 ± 0 . 3 10-90 1.2-2.4
monopolar cathode obviates the concern for crossing the palm's Orthodromic
impedance barrier and a pulse duration of 0.05 ms typically suf- 2.4 ± 0 . 3 3.0 ± 0 . 3 15-50
fices for generating a supramaximal response. A surface anode,
ring electrode, located on the thumb or preferably a dispersive Antidromic: mid-palm t o 14 c m 362
electrode on the dorsum of the wrist yields good results with the Mid-palm 1.6 ± 0 . 2 67 ± 20 1.0 ± 0 . 1
above-noted needle cathode. 14 cm 3.1 ± 0 . 2 52 ± 1 3 I.I ± 0 . 1
P/W 52 ± 4% 128 ± 2 9 %
Instrumentation Parameters. A sweep speed of 1 or 2 OL, onset latency; PL, peak latency; P/W, palm to wrist ratio for latency and
ms/div is recommended to optimally resolve the recorded amplitude.
2 0 2 — P A R T II BASIC AND ADVANCED TECHNIQUES
between an anatomic site of neural compromise and an electro- or distal wrist crease for a total of 12 points of neural excita-
physiologic conduction abnormality. Although the method se- tion. At approximately 3 cm distal to the zero reference, the
quentially stimulates the median nerve in 1-cm increments, this distal edge of the transverse carpal ligment is believed to be
study is commonly known as the "inching technique." 235
present. 168180
Recording Electrodes. As this is an antidromic technique, An important cautionary note is necessary with respect to this
recordings are performed on digital sensory nerves distal to the technique. Because the median nerve passes beneath the trans-
site of stimulation. Although the original method described verse carpal ligament and then into the palm under rather thick
recordings from the second digit, the same data most likely tissues, there is a varying depth of this nerve below the skin sur-
apply to the third digit as well. face along its course. It is often necessary to use different cur-
275
E-l. A ring E-l recording electrode is located on the second or rent intensities and pulse widths to obtain supramaximal
third digit just distal to the metacarpophalangeal joint (Fig. 5-23). responses. There is a danger that in some persons, particularly
E-2. The E-2 electrode is positioned distal to E-l at a distance individuals with rather calloused hands, excessive current may
approximating 4 cm if possible. Because this technique is pri- be required to maximally excite the median nerve in or about
marily concerned with comparative latencies and possibly am- the carpal tunnel region or mid-palm. There is a possibility,
plitudes over short segments, absolute times and distances are of therefore, of stimulus spread exciting the nerve slightly more
less importance than for the previously described investigation. distally than anticipated, thus producing a region of "abnormal"
Stimulation. The stimulation sites were determined by first segmental latencies. - Judicious use of current intensity to
180 354
localizing the distal wrist crease, which is designated the zero induce supramaximal responses in this technique, particularly
or reference mark (Fig. 5-23). The distal wrist crease approxi- by the beginning practitioner who has not had much practice, is
mates the proximal edge of the transverse carpal ligament. A 275
mandatory with this method of short-segment stimulation.
line overlying the median nerve, between the tendons of the Alternatively, a needle cathode can be used.
flexor carpi radialis and palmaris longus muscles proximally, Instrumentation Parameters. An amplifier sensitivity of
and the third digit distally, serves as the referential measuring 10-50 pV/div and a sweep speed of 1 or 2 ms/div should result
line. Six proximal and 5 distal stimulation sites are designated in clearly defined SNAPs. Filter settings of 20 Hz to 2 kHz were
with respect to the reference line where it crosses the zero line used in the original study; however, high and low filter settings
Sfte of
Stimulation
-6
-5
-4
-3
~2
-1
0
1
2¬
3
4
5
| 50 uV
B 1 ms
Figure 5-23. " I n c h i n g " t e c h n i q u e . Sensory "inching" stimulation of the median nerve can be performed with recording electrode placement
as previously described for the median nerve.The ensuing antidromic median nerve SNAPs are shown and correspond t o the various stimulation
sites (horizontal lines across the palm of the hand). (From Kimura J:The carpal tunnel syndrome: Localization of conduction abnormalities within
the distal segment of the median nerve. Brain 1979; 102:619-635, with permission.)
Chapter 5 NERVE C O N D U C T I O N STUDIES — 2 0 3
approximating these values are acceptable as comparative laten- used for more proximal segments. The reason the antidromic
cies are the important parameters of interest. SNAP amplitudes are comparatively larger is because the
Reference Values. The SNAP'S onset is recorded for each recording electrodes are located over the digits where the nerve
stimulus site, and the latency difference between adjacent fibers are subcutaneous and relatively close to the electrodes. In
neural excitation points is calculated. Should the onset latency orthodromic techniques, the nerve is usually deeper in the limb
be difficult to detect, one can also use the negative peak latency and subsequently farther from the recording electrodes, thus
to determine segmental conduction times. Normal sensory generating smaller amplitude responses. Orthodromic SNAPs,
axons demonstrate a segmental latency shift between 0.16 and because of their small amplitude, may be difficult to observe at
0.21 m s . Short-segment latency shifts exceeding these
1 8 0 1 8 3
increasing distances from the stimulus site, e.g., when attempt-
limits may be abnormal and indicate focal slowing of neural ing to calculate conduction velocities over several segments. It
conduction. Again, caution must be exercised not to use too has been recommended that needle recordings be used for or-
much current and produce an erroneous segmental latency shift. thodromic studies; however, care must be exercised in near-
! These reference values were obtained in forearms maintained at nerve recordings. The amplitude of the recorded response is
| or above 34°C with infrared heating lamps. very dependent upon the distance between the nerve and needle.
Failure to ensure a near-nerve placement at consecutive sites
Orthodromic Technique may cause one to erroneously conclude that pathology has
When performing orthodromic techniques, the median caused a comparative amplitude reduction when indeed it is a
nerve's sensory fibers in the digits are excited and the ensuing result of poor needle placement. This is particularly a problem
response is recorded at the wrist. This type of stimulation and when inexperienced examiners are following patients serially
recording is referred to as orthodromic because the induced and attempting to compare amplitudes with those previously
neural impulse propagates in the same direction as a sensory obtained, especially by a different examiner.
action potential generated by natural stimuli. Additionally, the Early studies investigating both antidromic and orthodromic
response is recorded proximal to where it is produced. In this SNAPs concluded that there was no significant difference be-
way, the sensory action potential travels centripetally, i.e., phys- tween the two techniques with respect to latency as measured to
iologically toward the central nervous system. the potential's peak or onset. - - Later studies concluded that
30 233 234
Recording Electrodes. The recording electrodes are typi- orthodromic median nerve studies resulted in potential latencies
cally positioned over the median nerve at the wrist. This is the significantly shorter than those obtained with antidromic tech-
same general location as previously described for cathodal niques. - The implication was that there is some type of poorly
47 326
placement in antidromic excitation of the median nerve. understood physiologic explanation for the way in which nerves
E-L A surface electrode is situated over the median nerve conducted impulses in an antidromic compared to orthodromic
between the tendons of the flexor carpi radialis and palmaris manner. In reality, the difference is secondary to the technique
longus muscles. The distance between E-l and the cathode (see used to investigate neural conduction. The inter-electrode sepa-
below) is standardized at 14 cm. - 233 234
ration for recording antidromic responses was 4 cm while that
E-2. Placement of the E-2 recording electrode is crucial (see used for orthrodromic studies was less. Reducing the inter-elec-
below: antidromic vs. orthodromic conduction). It is recom- trode separation tends to reduce the recorded potential's ampli-
mended to use separate surface disc electrodes for an inter-elec- tude and latency. Using similar distances between E-l and
78-80
trode separation between E-l and E-2 of 4 cm. - 233 234

E-2 results in similar latencies for both antidromic and ortho-
Stimulation. A ring electrode secured around the second or dromic responses. Thus, the original investigators were correct
49
third digit serves as the cathode while a second ring electrode in finding that nerves conduct the same whether propagating an
several centimeters more distal on the digit is the anode. The impulse orthodromically or antidromically.
placement is the same as that noted above for E-1 and E-2, re-
spectively, in antidromic sensory nerve recordings. An appro- Ulnar Nerve
priate pulse duration and current intensity is used to produce a The ulnar nerve's cutaneous distribution innervates the
supramaximal SNAP. medial aspect of the volar and dorsal regions of the hand not
Instrumentation Parameters. See antidromic median nerve supplied by the median or radial nerves. Approximately 10
144
instrumentation parameters. cm proximal to the ulnar styloid, the first cutaneous branch of
Reference Values. Peak and onset latencies have been the ulnar nerve, the palmar ulnar cutaneous nerve, becomes
recorded for temperature and distance controlled orthodromic distinct and travels distally to innervate a small patch of skin on
investigations of the median nerve (Table 5-8). The reference the proximal hypothenar eminence. This nerve is somewhat
324
data vary depending on the inter-electrode separation between inconstant and is often replaced by terminal sensory ulnar nerve
E-l and E-2 (see below). fibers. Just distal to the origin of the palmar cutaneous branch of
the ulnar nerve a second cutaneous nerve, the dorsal ulnar cu-
Antidromic vs Orthodromic taneous nerve, arises to supply the dorsum of the hand.
Although one may read that antidromic studies are prone to Specifically, the skin overlying the medial two metacarpal
volume-conducted muscle responses contaminating the SNAP bones is innervated by this nerve. Within the hand, the ulnar
under investigation, - this criticism is infrequently of clinical
58 312
nerve divides into the superficial and deep ulnar nerves. The su-
significance. Locating the recording ring electrodes just distal perficial branch of the ulnar nerve innervates the hypothenar
to the metacarpophalangeal joint significantly reduces concomi- eminence and provides several superficial terminal branches to
tant overlap of motor and sensory responses. Additionally,
217
innervate the fifth and medial one-half of the fourth digits.
asking the patient to abduct the fingers during neural activation These superficial terminal branches also innervate the dorsal
can help minimize any contaminating motor artifact. The ampli- terminal portions of the above-noted digits. The dorsal ulnar cu-
tude of antidromic responses is typically larger than those ob- taneous nerve provides cutaneous sensibility to the proximal re-
tained through orthodromic techniques, and therefore can be gions of the digits' dorsum similar to the radial nerve's function
with respect to the median nerve and the lateral three and one- then flexed as noted above. These two skin designations serve
half digits. as the two stimulation sites with the elbow flexed to 135°. A
As with the median nerve, a number of nerve conduction fourth stimulation location is delineated approximately 11 cm
techniques exist to investigate the integrity of the ulnar nerve's proximal to the above elbow site. The distance between the 3
sensory fibers. An attempt is made to only note the more popu- segments is then measured. It is important to calculate the dis-
lar techniques controlling for distance and temperature. The in- tance across the elbow in the flexed position as previously de-
terested reader is encouraged to consult several nerve fined (see ulnar nerve motor section) by following the course of
conduction manuals for additional techniques. As for the
64 2 , 7
the ulnar nerve. Once can certainly use distances less than those
median nerve, both antidromic and orthodromic methods of in- described; however, recall that as the segmental distance de-
vestigating the superficial terminal branches of the ulnar nerve creases, the percent measurement error increases.
are described. Additionally, nerve conduction velocities across Instrumentation Parameters. Instrument parameters simi-
the elbow segment are also described to allow the practitioner lar to those noted for the median nerve's sensory fibers are used.
the opportunity to investigate ulnar nerve sensory conduction Reference Values. Distal sensory latencies are measured to
across a common site of ulnar nerve entrapment. the peak of the negative deflection (Table 5-9). When comput-
ing sensory nerve conduction velocities, however, the potential
Antidromic Technique is measured to the initial deflection of the negative peak where
Recording Electrodes. The general comments regarding the waveform departs the baseline.
ring electrode application noted for median nerve sensory stud-
ies also apply for ulnar nerve sensory investigations. Orthodromic Technique
E-L The E-l recording electrode is located on the fifth digit The orthodromic technique for evaluating the ulnar nerve is
just distal to the metacarpophalangeal joint or midway along the essentially the same as that used for the median nerve.
proximal phalanx (Fig. 5-22). If motor artifact is encountered, Essentially, the E-l and E-2 ring recording electrodes are now
positioning the electrode slightly more distally may help, and used as the cathode and anode, respectively. The recording elec-
asking the patient to abduct the digits can also be of assistance trodes become separate disc electrodes located on the ulnar
in some persons. nerve just lateral to the tendon of the flexor carpi ulnaris muscle
E-2. One should attempt to locate E-2 as distal on the fifth 14 cm proximal to the cathode. A supramaximal stimulus is ap-
digit as possible to maximize the amplitude of the SNAP. In plied to the fifth digit. The orthodromic SNAP'S latency is
other words, the optimal separation of 4 cm is preferable, but recorded to the peak while the amplitude is calculated from the
the length of the fifth digit may preclude achieving this goal. initial deflection to the negative peak (Table 5-9).
Stimulation. The ulnar nerve is stimulated with a surface
cathode 14 cm proximal to the E-l recording electrode. The Dorsal U l n a r C u t a n e o u s N e r v e
actual location of the cathode is just lateral to the tendon of the Recording Electrodes. Separate disc recording electrodes
flexor carpi ulnaris muscle (forearm supinated). The anode is are located on the hand's dorsum. Because the dorsal ulnar cuta-
always positioned proximal to the cathode. neous nerve and its parent mixed nerve are in close proximity, it
When attempting to compute sensory conduction velocities is difficult to avoid concomitantly exciting both nerves. This
for various segments of the ulnar nerve, the same proximal stim- simultaneous neural excitation results in the SNAP closely
ulation sites for motor studies are used. Specifically, the elbow
is flexed 135° and supinated. Again, the arm is abducted ap-
proximately 90° and is rested on the same pillow supporting the
patient's head. This position clearly exposes all sites to be stim-
ulated. Although the original investigation excited the ulnar
nerve digitally at 11 cm, a 14-cm distal excitation should not
alter the nerve conduction velocities for the arm, across-elbow,
and forearm regions. With the elbow first extended, a mark on
the patient's skin is made 4 cm distal to the medial epicondyle
and approximately 10 cm proximal to this site. The elbow is
Table 5-9. Ulnar Nerve: Sensory

OL (ms) '4
PL (ms) 2;
Amplitude (pV) 49
Antidromic
2.4 ± 0.3 3.0 ± 0.3 34 ± 12.1
Orthodromic
2.4 ± 0.3 2.9 ± 0.3 15-50
NCV (m/s)' 84
FA 64.6 ±5.1 15.8 ±6.8

AE 68.5 ± 7.5 12.4 ±5.2
A 67.9 ± 9.3 9.8 ± 4.2
Figure 5-24. Dorsal ulnar cutaneous nerve. Location for E-1
Dorsal Ulnar Cutaneous N e r v e 156
(R ) and E-2 (R ) recording electrodes and stimulus site (S) for dorsal
a r
2.0 ± 0.3 20 ± 6 ulnar cutaneous nerve evaluation. (From Ma DM, Liveson JA: N e r v e
OL, onset latency; PL, peak latency. NCV calculated using onset latencies. Conduction Handbook. Philadelphia, F.A. Davis, 1983, with permission.)
approximating the appearance of the motor response arising

from the hand intrinsic muscles. The dorsal ulnar cutaneous
SNAP, therefore, is typically found at a slightly shorter latency
than the motor response.
E-L The E-l electrode is positioned over one of the superfi-
cial branches of the dorsal ulnar cutaneous nerve as it cross the
fourth and fifth metacarpal b o n e s . Specifically, E-l is lo-
156173
cated at the V-shaped approximation of the fourth and fifth

metacarpal bones proximal joining (Fig. 5-24). Slight medial or
lateral repositioning of this electrode may be necessary to opti-
mize the nerve's response. The hand is pronated and completely
relaxed.
E-2. An E-2 recording electrode is secured to the base of the
fifth digit.
Stimulation. The cathode is located 8 cm proximal to the
ulnar styloid with the anode proximal. The cathode must be se-
curely held in place between the ulna and flexor carpi ulnaris
muscle. Following supramaximal sensory stimulation, the
SNAP should appear slightly earlier than the relatively larger
motor response. It may be necessary to apply a current intensity
that is submaximal for the motor fibers in order to avoid a large
motor response that can potentially obliterate the SNAP.
Instrumentation Parameters. See median sensory nerve
instrumentation setup. Figure 5-25. Superficial radial nerve.The E-l (R ) electrode is
a
Reference Values. Latency is measured to the negative peak located over the tendon of t h e extensor pollicis longus while the E-2
as there may be an ill-defined initial deflection. Amplitudes of (R ) electrode is positioned distally close t o t h e distal aspect of t h e
r
the main negative deflection are calculated (Table 5-9). second metacarpal b o n e when investigating t h e superficial radial
nerve. (From Ma DM, Liveson JA: N e r v e Conduction Handbook-
Superficial Radial N e r v e
Philadelphia, F.A. Davis, 1983, with permission.)
The superficial radial nerve arises from the main trunk of the
radial nerve at approximately the level of the lateral epicondyle.
This nerve provides cutaneous sensation to the dorsal aspect of effect on the reported data. An amplifier sensitivity of 10 or 20
the lateral two-thirds of the hand as well as the dorsal aspect of pV/div should suffice to resolve most responses.
the first three and one-half digits. Two antidromic stimulation
144
Reference Values. The superficial radial nerve SNAP'S la-
techniques are described for this nerve. These techniques use tencies are reported to the initial deflection as well as to the neg-
surface stimulation and recording as well as attempt to control ative peak. Amplitudes are measured from baseline to the apex
for temperature and distance. Although it is possible to perform of the major negative peak (Table 5-10).
orthodromic techniques for this nerve, it is less painful to evalu- 2. Recording From Thumb. Because the superficial
ate this nerve with antidromic methods. - - 53 75 95
radial nerve innervates the dorsum of the first digit it is also
possible to record a SNAP from the thumb. Since this is the
Antidromic Techniques termination of this nerve, the fibers are quite small and share
1. Recording From Wrist. The first antidromic technique innervation of this digit with the median nerve, thereby pro-
described records the potential from the radial nerve at the ducing a comparatively reduced amplitude to the previously
wrist. Using this method, potentials are relatively easy to obtain. described technique.
E-L An E-l recording electrode is positioned over the super- E-L A ring recording electrode placed at the proximal por-
ficial radial nerve at the wrist (Fig. 5-25). This nerve can be tion of the first digit serves as E-l for this method. The open 215
easily located by asking the patient to extend the thumb while portion of the ring should be positioned to face the volar aspect
the practitioner palpates the extensor pollicis longus of the thumb. This affords maximum contact between the metal
tendon. - Just distal to the wrist the superficial radial nerve
216 217
surface of E-1 and the radial nerve fibers.
can be felt as it crosses the above-noted tendon. The site where
the nerve can be palpated is the optimal location for E-l.
E-2. The E-2 electrode is secured to the distal aspect of the Table 5-10. Superficial Radial Nerve
second metacarpal bone on the dorsum of the hand. Distance (cm) OL (ms) PL (ms) Amplitude (pV)
Stimulation. A surface cathode can be located at 10, 12, x>r
W r i s t Recording 217
14 cm proximal to E - l . The cathode is positioned over the lat-

217
10 1.8 ± 0 . 3 2.3 ± 0.4 31 ± 2 0 (13-60)

eral margin of the radius at one of the above three noted sites.
12 2.1 ± 0 . 3 2.6 ± 0.4
These distances are calculated by keeping the wrist in neutral
14 2.4 ± 0.3 2.9 ± 0.4
with the thumb adducted.
Instrumentation Parameters. A sweep speed of either 2 or Thumb Recording 215
5 ms/div can be used but 2 ms/div may help resolve the poten- 14 2.8 ± 0.5 3.3 ± 0.6 12 ± 8 (7-19)
tial better. Filter settings of 2 Hz to 2 kHz were recommended 16 3.1 ± 0 . 5 3.6 ± 0.6
by the original investigators, but those previously used filters 18 3.5 ± 0.6 3.9 ± 0.6
settings for other sensory studies also can be used with little OL, onset latency; PL, peak latency.Temperature at or exceeded 32°C.
E-2. The E-2 recording electrode is also a ring electrode sim- roots supplying this nerve originate from C5/C6. The nerve
144
ilar to that for E-l and is located distally on the first digit near becomes rather superficial at the lateral margin of the biceps
the base of the thumbnail. It is important to place a tissue or dry brachii tendon and forms two branches. An anterior branch pro-
gauze pad between the thumb and remainder of the hand to min- vides sensibility to the anterolateral aspect of the forearm proxi-
imize movement artifact from the rings brushing against the mal to the wrist. A dorsal branch innervates the skin on the
palm of the hand. dorsolateral forearm. Although it is possible to examine this
Stimulation. A surface cathode can be located on the lateral nerve with an orthodromic needle technique, an antidromic
335
margin of the radius 14, 16, and 18 cm proximal to E - l . The surface recording and stimulation method is described as it is
wrist is in neutral position and the thumb adducted for measur- considerably easier to perform. 307
ing the above-noted distances. Recording Electrodes. Unlike a number of the previously
Instrumentation Parameters. See previously described su- described studies, this method uses a plastic bar with embedded
perficial radial technique. disc electrodes 3 cm apart in order to reproduce the original
Reference Values. The methods used to calculate latencies technique. This allows one to use the reported reference data.
and amplitudes are the same as those previously described for Should one wish to use separate disc electrodes at more than 3
recordings at the wrist (Table 5-10). cm separation, slightly longer peak latencies and larger ampli-
tudes can be anticipated. As always, if the original method is
Lateral Antebrachial Cutaneous N e r v e modified, new reference data must be developed.
The lateral antebrachial cutaneous nerve is the cutaneous sen- E-l. A line is constructed connecting a site just lateral to the
sory termination of the musculocutaneous nerve. The nerve biceps brachii tendon at the antecubital fossa and the radial
pulse at the wrist (Fig. 5-26). At 12 cm distal to the antecubital
location, the E-l electrode is secured to the forearm. 307
E-2. The E-2 electrode embedded in the plastic bar is posi-

tioned on this same line distal to E-l. If a separate surface elec-
trode is used, care should be taken to align both E-l and E-2 on
the above-noted line.
Stimulation. The surface cathode is located at the above-
noted site just lateral to the biceps brachii tendon (Fig. 5-26).
For this technique it is crucial to have firm pressure applied to
the stimulator to position the cathode deep enough into the
arm's soft tissues and ensure close proximity to the nerve.
Should this response not be readily detected, the E-l and E-2 i
electrodes should be repositioned slightly prior to concluding
the response is pathologically absent. A normal contralateral
(unaffected) response helps implicate a pathologic process
rather than poor technique. The current intensity is best deliv-
ered in small increments to preferentially activate the sensory
fibers of the lateral antebrachial cutaneous nerve. Too strong an
excitation pulse may coactivate the nearby median nerve with
volume-conducted potentials, overwhelming the small sensory
response.
Instrumentation Parameters. See median nerve sensory
technique.
Reference Values. Although temperature was not recorded,
the forearm should be subject to less temperature variation than
that encountered in the digits. The SNAP'S onset latency was
1.8 ± 0.1 ms (1.6-2.1 ms) and the peak latency was 2.3 + 0 . 1 ms ;
(2.2-2.6 m s ) . Conduction velocities using onset latencies

307
were 65 ± 3.6 m/s (57-75 m/s) and amplitudes were 24 ± 7.2

pV (12-50 pV). Left/right latency differences for both onset
and peak were 0.1 ± 0.1 ms (0.0-0.3 ms).
MEDIAL ANTEBRACHIAL C U T A N E O U S NERVE

The medial antebrachial cutaneous nerve originates from the
medial cord or rarely the lower trunk, and is derived from seg-
ments C8 and T l . This nerve is subcutaneous just proximal
2 2 6 2 6 5
to the medial epicondyle and course along the medial aspect of

the arm to provide cutaneous sensation to the medial forearm.
Figure S-26. L a t e r a l a n t e b r a c h i a l c u t a n e o u s n e r v e . W h e n ex- Recording Electrodes. As for the lateral antebrachial cuta-
amining t h e lateral antebrachial cutaneous nerve the E-l (R ) elec- a neous nerve, disc electrodes embedded in a plastic bar are used.
t r o d e is located distal t o the biceps t e n d o n (see text) while the E-2 E-l. The medial epicondyle is palpated with a line con-
electrode (R ) is situated distal to E-l.The nerve is excited just lateral
r structed between the medial epicondyle and ulnar styloid. The
t o the biceps brachii tendon. (From Ma DM, Liveson JA: N e r v e E-l electrode is placed on this line 8 cm distal to the medial epi-
Conduction Handbook. Philadelphia, F.A. Davis, 1983, with permission.) condyle (Fig. 5-27).
E-2. The E-2 electrode is located distal to the E-l electrode

on the same line noted between the medial epicondyle and ulnar
styloid.
Stimulation. The cathode is located 4 cm proximal to the
medial epicondyle in the groove overlying the neurovascular
bundle between the arm's anterior and posterior muscle com-
partments. It is important not to apply too much pressure to the
cathode and slowly increase the current intensity. It is quite easy
to apply too much current activating the median and ulnar
nerves. Additionally, the cathode may need to be moved slightly
anterior or posterior to find the best position for activating the
nerve.
Instrumentation Parameters. See median nerve sensory
technique.
Reference Values. The forearm temperature was noted to be
greater than 3 1 ° C . Mean peak SNAP latencies for this an-
265
tidromic technique were 2.1 ms (1.7-2.6 ms) with conduction

velocities of 49.3 m/s (45-55 m/s) and an amplitude of 20 pV
(10-30 pV).
Comparative N C S of Dual-Innervated Digits

In the hand there are two digits with dual cutaneous innerva-
tion, i.e., first digit (radial/median) and fourth digit
(median/ulnar). These digits have been used to evaluate com-
parative latencies of both innervating nerves, particularly when
median nerve entrapment at the wrist is suspected. In theory, if
the median nerve is preferentially affected, then its latency to
the dual innervated digit should be comparatively delayed.
These techniques are presented primarily because median neu-
ropathy at the wrist is rather common and may assist in the di-
agnosis of equivocal cases of entrapment neuropathy. Caution
must always be exercised whenever a digit with dual innerva-
tion is used. A delay in the affected response is a clear method
Figure 5-27. M e d i a l a n t e b r a c h i a l c u t a n e o u s n e r v e . T h e E-l
of detecting an abnormality. If the response is absent, then a
electrode Is located 8 cm distal to the medial epicondyle with the E-2
single peak is detected when both are nerves are simultaneously
electrode positioned approximately 3 cm distal t o E-l.The medial an-
activated through a volume-conducted wave of depolarization.
tebrachial cutaneous nerve is stimulated 12 cm proximal to E-1 along
Only the techniques employing separate nerve stimulation
the medial arm. (From Ma DM, Liveson JA: N e r v e Conduction
should be relied upon when recording from a digit yielding a
Handbook. Philadelphia, F.A. Davis, 1983, with permission.)
single response.
Median/Radial Nerves to Digit I palmaris longus muscles. The location for the cathode is deter-
As previously stated, the median and radial nerves supply the mined by measuring from E-l to the center of the distal wrist
cutaneous innervation to the first digit. By applying electrodes crease and then proximally until 10 cm is reached. By moving
to the thumb, one can perform either orthodromic or antidromic the cathode on a line halfway between the median and radial ex-
evaluations of both cutaneous nerves. Antidromic studies are citation sites, it is possible to coactivate both median and radial
preferred because they yield larger and more easily recorded nerves. This last method of stimulation is not recommended for
SNAPs. - -
37 261 298
attempting to detect pathology.
Recording Electrodes. Ring electrodes are used to record Instrumentation Parameters. See median nerve sensory
the two nerves' SNAPs. The region of noncontact for the ring parameters.
electrodes should be the medial aspect of the thumb (hand Reference Values. When coactivating both median and su-
supinated). This electrode position assists in recording the po- perficial radial nerves, a characteristic pair of potentials is ob-
tential from both nerves. served. Normally, either one fused response is noted or there is
E-l. The E-l ring electrode is secured to the base of the first less than a 0.5 ms difference between the two potentials' peaks
digit approximating the metacarpophalangeal joint (Fig. 5-2&). (Table 5-11). A prolongation of one of the peaks in re-
167 2 6 0
This location is similar to that noted for recording the superfi- sponse to individual or dual nerve excitation more than 0.5 ms
cial radial nerve response from the thumb. 215
suggests entrapment. Stimulating each nerve singly at their op-
E-2. A ring E-2 electrode is positioned distally on the thumb timal location defines which nerve is slowed. As the median
close to the base of the thumbnail. nerve is commonly entrapped at the wrist, it is more likely that a
Stimulation. There are three stimulation sites for this tech- lesion affects this nerve.
nique. First, the superficial radial nerve is excited 10 cm proxi-
mal to E-l on the lateral margin of the radius (Fig. 5-28). The Median/Ulnar Nerves to Digit IV
second site of stimulation is over the median nerve proximal to The fourth digit or "ring finger" is supplied on its volar sur-
the wrist between the tendons of the flexor carpi radialis and face by the median nerve on its lateral aspect and ulnar nerve on
between the metacarpophalangeal and proximal interphalangeal

joints.
E-2. The E-2 ring electrode is situated 4 cm distal to E - l .
Stimulation. A supramaximal stimulus is delivered sequen-
tially to the median and ulnar nerves at the wrist 14 cm proxi-
mal to E - l . The stimulation sites are similar to those described
previously. Although the original study used a needle cathode,
more recent studies found no difference between needle and
surface stimulation in a limited number of subjects.
Instrumentation Parameters. See parameters for the sen-
sory median nerve.
Reference Values. Median nerve SNAP peak latencies were
3.14 ± 0.24 ms while the ulnar nerve revealed peak latencies of
3.03 ± 0.21 m s . The latency difference in 9 3 % of hands was
166
less than 0.3 ms and less than 0.6 ms in all hands examined.
Amplitudes were not noted but temperature was controlled at
greater than 30°C for the hand.
//////>?/
MIXED NERVE STUDIES
Figure S-28. Costimulation of t h e superficial radial and
median sensory nerves. The superficial radial nerve and median A popular technique used to examine both median and ulnar
nerve can be individually stimulated at t h e wrist 10 cm proximal t o the nerve fibers in the hand involves exciting the respective mixed
recording electrodes located on t h e first digit. W h e n exciting t h e nerves in the mid-palm and recording a response proximally.
median nerve, the 10-cm distance is measured along the thenar emi- The term mixed nerve is preferred because motor efferents,
nence t o t h e insertion of the APB and then between the tendons of the motor afferents, and cutaneous sensory fibers are all stimulated.
FCR and PL For the superficial radial nerve, a straight line 10 cm long is The terms antidromic and orthodromic do not adequately char-
constructed from t h e recording electrodes proximally along t h e radius. acterize the stimulating/recording technique. The actual wave-
Both nerves can be simultaneously stimulated through a volume-conform detected is neither a sensory nor a motor potential but a
ducted stimulus midway between t h e t w o aforementioned stimulus compound nerve action potential, i.e., a mixed nerve response.
sites. (From Johnson EW, Sipski M, LammertseT: Median and radial sen- Recording Electrodes. E-l/Median Nerve. A surface disc
sory latencies t o digit I: Normal values and usefulness in carpal tunnel electrode secured to the skin between the tendons of the flexor
syndrome. Arch Phys Med Rehabil 1987;68:140-141, with permission.) carpi radialis and palmaris longus muscles is the E-l location
(Fig. 5-29). The actual location approximates the region be-
312
tween the distal and proximal wrist crease. This site is some-
the medial aspect. By stimulating the median and ulnar nerves
144
what variable as it depends on the size of the subject's hand.
at the wrist and recording from digit IV, it is possible to com- There must be an 8 cm separation between E-1 and the cathode
pare antidromic SNAPs from both nerves. The rationale for this that is placed in the mid-palm area between the flexor tendons
study assumes that compromise of the median nerve at the wrist to the second and third digits.
will result in a comparatively longer median than ulnar nerve la- E-J/Ulnar Nerve. The E-l recording electrode for the ulnar
tency. In short, this technique is suggested as a screening test for nerve is placed adjacent to the recording location for the median
possible median nerve entrapment at the wrist. Of course, nerve over the ulnar nerve at the wrist (Fig. 5-29). This is just
should the ulnar nerve be preferentially affected, one may con- lateral to the flexor carpi ulnaris tendon at the same level as the
sider screening for ulnar neuropathies. Unfortunately, anatomic median nerve's E - l .
variants of an all-ulnar or median-innervated fourth digit render E-2/Median Nerve. An E-2 electrode is situated 3.5 cm prox-
this technique useless for comparison studies in some persons. imal to E-l over the median nerve, i.e., between the two above
Fortunately, these variants are rare. 206
noted tendons.
Recording Electrodes. Because this is an antidromic E-2/Ulnar Nerve. E-2's site is 3.5 cm proximal to E-l just
method, the median and ulnar nerves are excited proximal to the lateral to the flexor carpi ulnaris muscle.
recording electrodes. It is also possible to perform an ortho-
166
Stimulation. Median Nei-ve. The cathode is located over the
dromic technique by stimulating the fourth digit and recording sensory fibers of the median nerve in the mid-palm region 8 cm
from both the median and ulnar nerves at the wrist. distal to E-l (Fig. 5-29). Should the distally placed anode
E-L A ring electrode is located on the fourth digit with the extend beyond the palm of the hand or into the digital area, E-l
opening dorsally. This E-l is positioned approximately midway must be relocated to a slightly more proximal position. Because
there is a possibility of coactivating motor fibers to the lumbri-
cals or the recurrent branch of the median nerve, a mixed nerve
Table 5-11. Median/Radial Sensory Latencies ' 167 261
response is most likely the result. Of course, should only cuta-

Nerve PL (ms) Amplitude (pV) LD (ms) neous sensory fibers be activated, then a pure SNAP is gener-
Radial
ated. The anode may need to be rotated about the cathode to
2.4 ± 0 . 2 (1.9-2.8) 12 ± 1.0
minimize the stimulus artifact. A pulse width of 0.1 ms usually
Median 2.5 ± 0.2 (2.0-2.9) 30 ± 2.0 suffices; however, there may be an occasional need to increase
Radial/Median <0.5 it to 0.2 ms.
PL, peak latency; LD, latency difference between peak median and radial SNAPs. Stimulation. Ulnar Nerve. The ulnar nerve is excited 8 cm
Amplitude measured from baseline to peak negative deflection. distal to E-l in the mid-palm area between the flexor tendons to
Figure 5-29. Median/ulnar mixed nerve response. Location for stimulating and recording electrodes for performing a mixed nerve evalua-
tion of t h e median (left) and ulnar (right) nerves. (From Stevens JC:The electrodiagnosis of carpal tunnel syndrome. Muscle Nerve 1987; 10:
99-1 13, with permission.)
the fourth and fifth digits (Fig. 5-29). Similar statements regard- of most of the reported lower limb motor studies are referred to
ing stimulation of the median nerve are equally pertinent to the standard t e x t s . Five NCS techniques (femoral, peroneal,
642,6
ulnar nerve. tibial, sciatic, and pudendal nerves) are described as these
Instrumentation Parameters. A sweep speed of 2 ms/div assess most of the lesions likely to be encountered even by ex-
with an amplifier sensitivity of approximately 10-20 pV/div aminers involved in a large practice.
should suffice for obtaining clearly defined responses. Filter set-
tings approximating 10 Hz and 2 kHz are recommended. Femoral N e r v e
Reference Values. Negative peak latencies for the median The femoral nerve is comprised of nerve fibers originating
nerve have a mean of 1.8 ± 0.02 ms with a range of 1.5-2.2 from the dorsal aspects of the lumbar ventral primary rami
m s . Peak-to-peak amplitudes are noted at 100 ± 5.1 pV with a
312
L2-L4. 1 4 4
It courses through the psoas muscle and lies b e -
range of 50-180 pV. The ulnar nerve has the same normal la- neath the iliacus muscle. The iliacus muscle is innervated
tency parameters as the median nerve. The ulnar nerve's peak while the nerve is located within the abdominal cavity, and it
latency should not differ from the median response by more then enters the femoral triangle. T h e femoral nerve travels
than 0.2 ms. Amplitudes should exceed 15 pV. Temperatures into the leg under the inguinal ligament approximately
measured over the first dorsal interosseous muscle are recom- halfway between the anterior superior iliac spine and the
mended to be between 33-36°C. pubic tubercle. Once in the leg, the femoral nerve can be lo-
cated proximally by palpating the femoral artery and sweep-
ing 1 or 2 cm laterally. Roughly 4 - 5 cm distal to the inguinal
COMMON LOWER LIMB NERVE ligament, the femoral nerve divides into an anterior and pos-
CONDUCTION STUDIES terior division. The anterior division innervates the skin over
the anterolateral thigh, and also supplies the pectineus and
There are a number of relatively easy-to-perform lower limb sartorius muscles. The posterior division innervates the hip
nerve conduction techniques. Essentially, the basic principles and knee joints and the quadriceps muscle and terminates as
applied to upper limb nerve conduction studies pertain to lower the saphenous nerve.
limb nerve conductions. NCVs for lower limb nerves are gener- Recording Electrodes. Unless otherwise stated, all lower
ally slower than for upper limb nerve segments. It is particularly limb recordings are performed with surface electrodes. The ma-
important to master a number of motor and sensory lower limb jority of motor evoked responses are of sufficient amplitude to
nerve conduction methods because the majority of generalized be easily detected with surface electrodes. Of course, in patho-
processes affecting the peripheral nervous system begin in the logic situations, should a response not be observed with surface
distal aspects of the lower limb and progress proximally. As electrodes, one may wish to consider an intramuscular needle
with upper limb NCS, the techniques attempting to standardize recording to document that a response is still present. An NCV
nerve conduction methods for distance and temperature are can be obtained with this technique.
noted. E-L An E-l recording electrode is located over the mid-por-
tion or most prominent aspect of the vastus medialis muscle. If
LOWER LIMB MOTOR NERVE CONDUCTION STUDIES an initial positive deflection is observed with neural excitation,
E-l may need to be repositioned slightly in order to generate a
There are a number of lower limb motor nerve conduction negative deflection from the baseline.
studies commonly performed. Although it is impractical to at- E-2. E-2 is most conveniently secured to the patella as this is
tempt a litany of every described technique for lower limb a relatively electrically silent region.
motor studies, those believed to be used most often in clinical Stimulation. Either a needle or surface stimulator may be
practice are detailed. Practitioners interested in a compendium used. Although a surface cathode/anode are convenient for the
practitioner, the deep location of the nerve may require high intensity are optimized by adjusting both the needle's depth and
current intensities and pulse durations. This possibility can current delivered.
result in a rather uncomfortable examination for the patient. Three stimulation sites are recommended just as for upper
Should the patient have difficulty tolerating the stimulation, a limb nerve conduction studies: above and below the inguinal
needle cathode should be considered as less current intensity is ligament, and at Hunter's canal along the medial aspect of the
necessary because the cathode is positioned next to the femoral thigh. Again, the femoral artery is located in the inguinal region
nerve. In this instance, a surface or needle anode located several at a site 1-2 cm lateral to the femoral artery. The separation be-
centimeters more proximally is used. As long as one is lateral to tween the two proximal cathodal stimulation points should ap-
the femoral artery, there should be little danger of piercing the proximate 5-6 cm. The original description of this technique
artery provided the needle electrodes are perpendicular to the used a mean distance between stimulation locations of 5.5 ±
skin. Even though the nerve lies rather deep, the needle is ad- 1.6 cm. The total length of the femoral nerve between the prox-
vanced slowly delivering a stimulating pulse of moderate inten- imal stimulation site and E-l was noted to be 35.4 ± 1.9
sity. As the muscle contraction increases, the current intensity is c m 64.165.216
reduced until a small amount of current produces a large Instrumentation Parameters. The instrument's sweep
quadriceps muscle contraction. The needle location and current speed and amplifier settings were not noted in the original study
describing the above noted technique. However, a sweep
165
speed of 2 or 5 ms/div, amplifier gain of 2,000 to 5,000 pV/div

with filter settings approximating 10 Hz to 10 kHz result in
easily obtainable CMAPs. The same parameters can be used for
all lower limb studies.
Reference Values. The latencies from above and below the
inguinal ligament to E-l are 7.1 ± 0.7 ms (6.1-8.4 ms) and 6.0 ±
0.7 ms (5.5-7.5 ms), respectively. Conduction velocities from
165
above and below the inguinal ligament to the mid-thigh location

are 66.7 ± 7.4 m/s (50-96 m/s) and 69.4 ± 9.2 m/s (50-90 m/s),
respectively. The time of conduction across the inguinal liga-
ment region is 1.2 ± 0.4 ms (0.8-1.8 ms).
Sciatic N e r v e
The sciatic nerve is the largest nerve in the human body and
formed by nerve fibers originating from lumbosacral segments
L 4 - S 3 . The muscles innervated are the medial and lateral
144
hamstrings as well as all of the muscles distal to the knee joint.

The above-noted root levels fuse to form the sciatic nerve as it
travels along the posterior aspect of the pelvis to enter the
gluteal area through the greater sciatic foramen. Traveling deep
to the gluteus maximus muscle the sciatic nerve enters the lower
limb beneath the biceps femoris muscle. At about the mid-thigh
to distal one-third of the thigh, two major divisions of the sciatic
nerve, tibial and peroneal divisions, become physically distinct
as separate nerves. The tibial division of the sciatic nerve sup-
plies the long head of the biceps femoris, semitendinosus, and
semimembranosus, as well as a portion of the adductor magnus
muscles. The peroneal division innervates the short head of the
biceps femoris. Distal to the knee all muscles are innervated
either by the tibial or peroneal nerves (see below).
This nerve may be injured by traumatic dislocations of the
hip, femur fractures, pelvic ring fractures, or by local injection
injuries. When called upon to examine the sciatic nerve for the
above or other causes, the described technique may be of some
assistance in determining neural injury.
Recording Electrodes. Surface or needle electrodes may be
used to record from either the tibial or peroneal component of
the sciatic nerve. Surface recordings may offer the advantage of
using the amplitudes for diagnostic purposes.
Figure 5-30. S c i a t i c n e r v e e v a l u a t i o n . T h e sciatic nerve is stimu- E-l Tibial Component. An E-l electrode is located over the
lated at the gluteal fold (S ) using a 75-mm monopolar needle as the
( abductor hallucis muscle 1 cm distal and 1 cm inferior to the
cathode with a surface anode positioned nearby.The tibial and per- navicular bone in the foot (Fig. 5-30).
oneal nerves can then be excited at the popliteal fossa (S ). Recording 2 E-1 Peroneal Component. The E-l recording electrode is sit-
electrodes are positioned on the e x t e n s o r digitorum brevis and ab- uated over the main bulk of the extensor digitorum brevis
ductor hallucis for examining the peroneal and tibial portions of the muscle on the foot's dorsum (Fig. 5-30).
sciatic nerve, respectively. (From Ma DM, Liveson JA: N e r v e E-2 Tibial Component. E-2 is located on the medial aspect of
Conduction Handbook. Philadelphia, F.A. Davis, 1983, with permission.) the foot at the first metatarsophalangeal joint.
E-2 Peroneal Component. E-2 is secured to the lateral aspect

of the foot over the fifth metatarsophalangeal joint.
Stimulation. The sciatic nerve is excited proximally by
using a monopolar needle 5 0 - 7 0 mm in length. The needle is
inserted just below the gluteal fold in line with the popliteal
fossa's apex (Fig. 5 - 3 0 ) . A surface anode can be positioned
2,6367
several centimeters more proximally. Surface stimulation of the

sciatic nerve should not be attempted as the nerve is too deep to
be maximally activated in most persons. Similar localization
techniques are used to those previously described for the
femoral nerve.
The two divisions of the sciatic nerve are stimulated in the
popliteal fossa. Tibial nerve excitation is performed in the mid-
popliteal fossa region while the peroneal nerve is activated lat-
eral to the tibial nerve (Fig. 5-30). Surface stimulation works
quite nicely for neural activation at this location. For both prox-
imal and distal stimulations, proper activation is ensured by ob-
serving the clinical response following stimulus delivery. Figure 5 - 3 / . Bulbocavernosus reflex study. The bulbocaver-
Specifically, the tibial portion of the sciatic nerve produces nosus reflex is examined electrically by stimulating the dorsal nerve of
plantar flexion when activated. On the other hand, foot eversion the penis (clitoris) and recording with a needle electrode located in
and dorsiflexion indicate the peroneal portion of the sciatic the bulbocavernosus muscle. (From Siroky MB, Sax DS, Krane RJ:
nerve has been activated. It is possible to preferentially stimu- Sacral signal tracing: The electrophysiology of t h e bulbocavernosus
late these aspects of the sciatic nerve below the gluteal fold by reflex. J Urol 1979; 122:661 - 6 6 4 , with permission.)
slightly positioning the needle medially (tibial division) or later-
ally (peroneal division). placement. The instrument is then set up for recording nerve
Instrumentation Parameters. See femoral nerve. conduction latencies.
Reference Values. Nerve conduction velocities for the tibial E-2. The cannula of the standard concentric needle electrode
and peroneal divisions of the sciatic nerve are 52.8 ± 4.7 m/s serves as E-2.
(46.7-59.6 m/s) and 54.3 ± 4.4 m/s (48.5-61.5 m/s), respec- Stimulation. Ring electrodes are located on the shaft of the
tively. Because of the large number of muscles innervated in
216
penis with the cathode proximal. A current intensity above sen-
the lower limb, recordings can be made from multiple muscles sory threshold but capable of being tolerated by the patient is
using standard concentric needles placed within these muscles. delivered. The pulse duration has been reported between 0.1 to
Nerve conduction velocities for the leg from the two stimulation 0.5 m s . - In women, the cathode of the stimulator is located
69 296
sites noted above to a number of muscles are: abductor hallucis over the clitoris and stimuli applied.
51 ± 5 . 8 m/s; extensor digitorum brevis 51 ± 7.0 m/s; tibialis Instrumentation Parameters. A sweep speed of 10 ms/div
anterior 55 ± 4.5 m/s; soleus 56 ± 5.5 m/s, and gastrocnemius may be necessary for this response because of its relatively long
56 ± 5.6 m/s. Temperature is most likely of little consequence
106
latency. The amplifier gain must be rather sensitive at about 200
as the nerves are rather deep and, unlike cutaneous nerves, not pV/div as the response is small. Filter settings similar to the
subject to major temperature fluctuations. femoral nerve are used.
Reference Values. A mean onset latency of 35 ± 2.0 ms
Bulbocavernosus Reflex/Pudendal N e r v e with a range of 2%~42 ms was observed. Abnormal responses
296
Although examination of the pudendal nerve is not performed were defined as the waveforms with latencies in excess of 45
directly but depends upon a reflex response, it is nevertheless ms.
described in this section. Clinically one evaluates the integrity
of the pudendal nerve by compressing the glans penis or clitoris Peroneal Nerve
and observing for anal sphincter or bulbocavernosus contrac- The peroneal or common peroneal nerve is one of the two
tion. - Because the bulbocavernosus muscle is somewhat dif-
21 205
major divisions of the sciatic nerve. Lumbosacral segments
ficult to observe clinically, the anal sphincter is more commonly L 4 - S 2 constitute the neural fibers comprising the peroneal
examined. The pudendal nerve can be investigated electrically nerve. As previously stated, the peroneal nerve becomes a dis-
144
by stimulating the dorsal nerve of the penis or clitoris and tinct entity from the sciatic nerve in the popliteal fossa region. It
recording a response from the bulbocavernosus muscle. This is travels medial to the biceps femoris but lateral to the lateral
an important electrical test to master as it can be of use in pa- head of the gastrocnemius muscles toward the fibular head.
tients with voiding or sexual dysfunction secondary to neuro- Upon reaching the head of the fibula, the nerve courses poste-
logic causes. rior to this structure and then deep to the peroneus longus
Recording Electrodes. E-L A standard concentric needle muscle. The peroneal nerve then separates into the superficial
electrode is inserted into the bulbocavernosus muscle (Fig. 5¬ and deep peroneal nerves. Within the leg, the superficial per-
31). The bulbocavernosus muscle is located just lateral to the oneal nerve innervates the peroneus longus and brevis muscles
scrotum on either side within the perineum. Both left and right and continues distally to provide cutaneous sensation to the
responses can be assessed by placing the needle on the appro- dorsum of the foot except the region between the first and
priate side following stimulation. Once the needle electrode has second digits. On the other hand, the deep peroneal nerve tra-
pierced the skin, the instrument should be placed in the needle verses the leg on the interosseous membrane to pass beneath the
electromyographic mode. Needle insertion is continued until extensor retinaculum at the ankle just lateral to the dorsalis
definite insertional activity is detected indicating intramuscular pedis artery. At the ankle, the deep peroneal nerve splits into
two terminal branches, one sensory and the other motor. The with manually pressing the cathode and anode into the depth
sensory nerve provides cutaneous sensibility to the area be- of the tissue medial to the biceps femoris tendon and away
tween the first and second digits while the motor branch inner- from the tibial nerve. Unfortunately this does not always pro-
vates the extensor digitorum brevis. During the deep peroneal duce the desired results and it may be impossible to activate
nerve's course in the leg, the following muscles are innervated: just the peroneal nerve with a surface stimulator. An alterna-
tibialis anterior, extensor digitorum longus, peroneus tertius, tive technique is to locate the stimulator's cathode/anode on
and extensor hallucis longus. As one of the major nerves inner- th,e lateral aspect of the thigh at the popliteal fossa level on the
vating the lower limb, it is important to be able to assess the per- skin overlying the biceps femoris tendon. Some pressure in
oneal nerve electrophysiologically. The peroneal nerve may be combination with an increased pulse width may be required to
compromised in pelvic fractures, traumatic hip fractures and obtain a response. If both surface techniques fail, the practi-
dislocations, femoral fractures, and entrapment or trauma at the tioner should consider a needle cathode. A current capable of
fibular head. exciting just the peroneal nerve is delivered after the needle
Recording Electrodes. There are two common recording has been positioned over the peroneal nerve. One must exer-
techniques typically employed in assessing the integrity of the cise some caution and review the pertinent anatomy to avoid
peroneal nerve. Recordings can either be obtained from the ex- puncturing the popliteal vessels.
tensor digitorum brevis (EDB) or tibialis anterior (TA) muscles.
Both methods are described as they are complementary in at- Tibialis Anterior
tempting to define the level of possible neural compromise as E-L The motor point of the TA is located by placing E-l at
well as the extent of a lesion. the junction of the muscle's proximal one-third and distal two-
thirds between the tibial tuberosity and prominence of the lat-
Extensor Digitorum Brevis eral malleolus over the bulk of the muscle. 68
E-L A surface E-l recording electrode is positioned over the E-2. An E-2 electrode is positioned over the medial aspect of
main bulk of the EDB muscle (Fig. 5 - 3 0 ) . - Occasionally,
,62 216
the tibia about the ankle.
slight repositioning of this electrode is required to obtain a Stimulation. Same as the two proximal stimulation sites
sharp initial negative deflection. noted above for EDB recordings.
E-2. The E-2 surface electrode is located on the fifth Instrumentation Parameters. A sweep speed of 5 ms/div,
metatarsophalangeal joint or on the fifth digit. sensitivity of 1,000 pV/div, and filter settings approximating 10
Stimulation. In evaluating conduction for the deep peroneal Hz to 10 kHz is recommended when using the reference data
nerve with recordings obtained from the EDB, three stimula- noted below. Less amplifier sensitivity than noted above is often
tion sites are suggested to evaluate both proximal and distal necessary to measure the potential's full amplitude.
nerve segments. Excitation of the peroneal nerve distally is car- Reference Values. Distal motor latencies and conduction
ried out by locating the cathode just lateral to the TA tendon. velocities for the EDB and TA are provided (Table 5-12). Leg
Occasionally, the nerve may be rather deep beneath subcuta- temperature at the recording sites should be monitored and
neous and tendonous tissues requiring an elevation in the stim- maintained above 29-30°C.
ulus' pulse width. The second stimulus of the same pulse
duration is delivered to the peroneal nerve a few centimeters Tibial N e r v e
distal to the fibular head. Finally, the peroneal nerve is excited The tibial nerve is the medial and larger portion of the sci-
a third time in the popliteal fossa as the nerve courses medial to atic nerve arising from lumbosacral nerve fibers L 4 - S 3 . At
144
the biceps femoris tendon. This last site of stimulation is some- approximately the level of the popliteal fossa, the tibial por-
what difficult and requires both patience and experience to se- tion of the sciatic nerve becomes a distinct nerve trunk. All
lectively depolarize the peroneal nerve. This difficulty arises posterior compartment leg muscles and foot intrinsic muscles
because of the tibial nerve's close proximity. A stimulus deliv- except the EDB are innervated by the tibial nerve. The nerve is
ered to the popliteal fossa frequently activates both the per- superficial in two regions, at the popliteal fossa and posterior
oneal and tibial nerves. The distally located E-l electrode then to the medial malleolus. Traveling posterior to the flexor reti-
detects the depolarization associated with both the EDB and in- naculum at the ankle, tarsal tunnel, the tibial nerve enters the
trinsic foot muscle innervated by the tibial nerve. This situation foot to terminate as two separate nerves, i.e., the medial and
may be detected if the CMAP from the proximal stimulation lateral plantar nerves.
site does not appear similar to the response obtained at the Recording Electrodes. There are two sites from which to
ankle. It is obviously important to only excite the peroneal record a CMAP following tibial nerve stimulation. One can ex-
nerve. This may be accomplished by using just the amount of amine conduction to not only the tibial nerve segment in the leg,
current necessary to stimulate the peroneal nerve combined but also the terminal portions of the medial and lateral plantar
nerves. Routine tibial nerve conductions are performed with
recordings to the abductor hallucis (AH) muscle, medial plantar
T a b l e 5-12. Peroneal Nerve nerve, to assess conduction in the leg (Fig. 5-30). Should one
216
desire information on the lateral plantar nerve, recordings can

DML (ms) PNCV(m/s) AMP (mV) DNCV (m/s) AMP (mV)
be performed to the abductor digiti quinti pedis.
TA Recording 68
E-l: Medial Plantar Nerve. A surface E-l electrode is lo-

3.0 ± 0.6 66.3 ± 12.9 3.9 ± 1.2 cated 1 cm posterior and inferior to the navicular tubercle on the
EDB R e c o r d i n g 40162 medial aspect of the foot (Fig. 5-30).
4.5 ± 0.8 53.9 ± 4 . 3 44 ±1.4 51.6 ±4.1 5.0 ± 1 . 5 E-2: Medial Plantar Nerve. An E-2 electrode is secured to
DML, distal motor latency; PNCV, proximal NCV from popliteal fossa to fibular the metatarsophalangeal joint or distal aspect of the first digit.
head regions; AMP, amplitude of CMAP measured from baseline to negative E-l: Lateral Plantar Nerve. The E-l recording electrode is
peak; DNCV, distal NCV for fibular head to ankle segment. positioned immediately inferior to the lateral malleolus halfway
Table 5-13. Tibial N e r v e medial and lateral plantar nerves. As with previously described
DML (ms) 102
AMP* (mV) - 102 162
AMR (mV) 162
NCV 102
nerve conduction techniques, an effort is made to present only
the methods using standard distances, surface recordings, and
MPN (8 cm)
temperature controls. It is not always possible to comply with
3.4 ± 0 . 5 11.8 ± 4 . 5 8.8 ± 3.4 54.9 ± 7.6 these criteria, however, and the most reliable methods in the au-
LPN (8 cm) thors' experience are presented.
3.6 ± 0.5
MPN (10 cm) Lateral Femoral Cutaneous Nerve
3.8 ± 0.5 7.54 (3.5-22) The lateral femoral cutaneous nerve is a pure sensory nerve
LPN (10 cm) formed from the ventral primary rami of nerve roots L2 and
3.9 ± 0.5 7.25 (3.0-10.0) L 3 . This nerve is posterior to the psoas muscle and passes
144
from the lateral border of the psoas muscle to traverse the ilia-
DML, distal motor latency from cathode posterior to medial malleolus to E-l; cus muscle. It then enters a tunnel in the lateral aspect of the in-
MPN, medial plantar nerve; LPN, lateral plantar nerve. AMP*, amplitude of
CMAP negative spike from distal stimulation; AMP*, amplitude of CMAP nega- guinal ligament. Distal to emerging from this tunnel the nerve
tive spike from proximal stimulation; NCV, NCV over leg segment from may pass over or through the sartorius muscle. The lateral
popliteal fossa to ankle region. MPN and LPN designate medial and lateral femoral cutaneous nerve divides into anterior and posterior
planter nerves, respectively.Table is incomplete because only data obtained branches to provide cutaneous sensation to the anterolateral
from studies using similar techniques have been included. DML ; AMP ; 102 162250
AMP and NCV. 162

aspect of the thigh. Injury to this nerve typically leads to altered
sensation in its cutaneous distribution and is referred to as mer-
algia paresthetica. 313
between the sole of the foot and the inferior tip of the lateral Recording Electrodes. Because of the rather small ampli-
malleolus approximating the abductor digit quinti pedis' motor tude of most lower limb SNAPs, the skin beneath both E-l and
point. E-2 recording electrodes should be gently abraded to reduce the
E-2: Lateral Plantar Nerve. For the lateral plantar nerve an skin impedance. Also, the patient must be completely relaxed as
E-2 electrode is positioned over the fifth metatarsophalangeal motor artifact may interfere with detection of the desired re-
joint or distal portion of the fifth digit. sponse. It is important to ensure that all electrodes are properly
Stimulation. The tibial nerve is routinely excited in two po- secured to the optimal location as lower limb hair may result in
sitions. A stimulus in the popliteal fossa and a second posterior loosening of the electrode/patient interface.
to the medial malleolus proximal to the flexor retinaculum E-L A tape measure is used to form a line connecting the an-
forms a segment over which the NCV can be calculated. The terior superior iliac spine with the lateral border of the patella
distal stimulus site may be either 8 cm or 10 cm proximal to the (Fig. 5-32). Approximately 16-18 cm distal to the anterior su-
E-l electrode location for the medial plantar nerve measured perior iliac spine on the above noted line is the designated site
over the course of the nerve 1 cm posterior to the medial malle- for E - l . 33
olus with the ankle in a neutral position, i.e., 9 0 ° . Two dis- 102
E-2. E-2 is positioned 3 cm distal to E-l on the same line
tances are provided to accommodate for various foot sizes. connecting the anterior superior iliac spine and the lateral aspect
Instrumentation Parameters. See peroneal nerve. of the patella.
Reference Values. Reference data are provided for both Ground. The ground electrode should be positioned adja-
stimulation distances with respect to cathodal placement (Table cent to E-l between the cathode and E - l . Again, the underlying
5-13). Acceptable temperature ranges are 29-34°C with a mean skin must be gently abraded to improve electrical contact. This
of 32.1 ± 1.4°C inferior to the medial malleolus. 102
is the same general location for all lower limb sensory tech-
niques and is assumed to apply for each method.
LOWER LIMB SENSORY NERVE Stimulation. The cathode is located approximately 1 cm
CONDUCTION STUDIES medial to the anterior superior iliac spine. Although some au-
thors use a needle cathode, one may wish to consider surface
There are a number of lower limb sensory nerve conduction stimulation. When a needle cathode is employed, a monopolar
techniques important for the electrodiagnostic medicine practi- needle is inserted 1 cm medial to the anterior superior iliac
tioner. The majority of lower limb SNAPs are slightly more spine. A stimulating pulse delivered at a rate of 1 Hz with a
33
technically demanding than upper limb sensory methods. The moderately intense current is slowly inserted perpendicular to
primary reason for this increased demand in technical profi- the skin surface until a paresthesia is described by the patient in
ciency is the small size of the responses. This small size is due the distribution of the nerve. Careful manipulation of the needle
in part to the size of the nerves investigated and the fact that and current is then performed until the response is optimized. A
unlike upper limb sensory nerves, in the lower limb the sensory surface anode may be located several centimeters proximal to
nerves are surrounded by more subcutaneous tissue. Also, we the cathode. The use of a surface cathode is also acceptable and
cannot encircle these nerves with a ring electrode and maximize is positioned at the same site noted for needle placement.
the contact between the recording electrode and neural tissue. Rotating the anode about the cathode may be of great assistance
The sensory nerve techniques discussed in this section are used in minimizing shock artifact. The locations of the stimulating
to evaluate the following nerves: lateral femoral cutaneous, pos- and recording electrodes obviously result in an antidromic tech-
terior femoral cutaneous, saphenous, superficial peroneal, and nique. Orthodromic techniques are also described, and are e s -
sural nerves. Although there are methods described to assess the sentially reversing the locations of the stimulator and recording
superficial fibers of the medial and lateral plantar nerves as they electrodes. - 33 285
innervate the skin to the digits, these responses are very small Instrumentation Parameters. A sweep speed of 1 or 2
and in the authors' opinion unreliable. A mixed nerve technique ms/div with an amplifier sensitivity of 10 pV/div should be suf-
is presented in detail to evaluate nerve action potentials in the ficient to optimize detection of the response. Filter settings
nerve trunk. Once below the popliteal fossa, it may extend for a
variable distance sharing innervation of the calf with the sural
nerve. This nerve is occasionally injured through injection in-
juries or prolonged bicycle riding. 313
E-L The E-l electrode is located in the midline of the poste-

rior thigh 6 cm proximal to the mid-popliteal fossa (Fig. 5-33). 82
E-2. The original technique used a bar electrode with a separa-

tion of 3.0 cm. A separate disc electrode may also suffice as E-2.
Stimulation. A tape measure connecting E-l with the ischial
tuberosity forms a reference line with which to locate the cath-
ode (Fig. 5-33). Twelve centimeters proximal to E-l on the
above-noted line is the desired stimulation site. The cathode is
positioned proximally and usually requires rotation to arrive at
an acceptable stimulus artifact.
Instrumentation Parameters. See lateral femoral cuta-
neous nerve.
Reference Values. Negative peak latencies were noted to be
2.8 ± 0.2 ms (2.4-3.2 ms) with amplitudes of 6.5 ± 1.5 pV
(4.4-11.0 p V ) .
82
Saphenous N e r v e
Of all the femoral nerve branches, the saphenous nerve is the
longest and largest. It is a pure sensory nerve and is com-
144
prised of lumbar segments L3/L4. The cutaneous distribution of

this nerve is distributed to two main regions. The most proximal
is the infrapatellar area while the second covers the anterome-
dial and posteromedial aspects of the leg. There is also a vari-
able distribution along the medial aspect of the foot
occasionally extending to the base of the first digit. Approx-
imately 4 - 5 cm distal to the inguinal ligament is where the
saphenous nerve becomes a distinct nerve as it separates from !
the main trunk of the femoral nerve. The nerve then enters the
adductor canal deep to the sartorius muscle in the anteromedial
thigh. Just superior to the medial epicondyle of the femur and *
posterior to the sartorius muscle, the saphenous nerve becomes
Figure 5-32. L a t e r a l f e m o r a l c u t a n e o u s n e r v e . O n e can stimu- subcutaneous traveling between the tendons of the sartorius and
late the lateral femoral cutaneous nerve either superior or inferior to gracilis muscles. In the leg, the saphenous nerve is located pos-
t h e inguinal ligament with a surface o r needle cathode.The E-I (R ) a
terior to the medial edge of the tibia as it courses toward the
and E-2 (R ) recording electrodes are located 16 cm distal t o the ante-
r
foot. About 7 cm proximal to the medial malleolus the nerve
rior superior iliac spine on a line connecting this landmark with the proceeds anterior and distal to it into the foot. The saphenous
lateral margin of t h e patella. (From Ma DM, Liveson JA: N e r v e nerve terminates along the medial border of the forefoot at
Conduction Handbook. Philadelphia, F.A. Davis, 1983, with permission.) times reaching the first metatarsophalangeal joint. This nerve
may be injured in the infrapatellar region either secondary to
approaching 10 Hz and 2 kHz are recommended to resolve the physical trauma or following surgical procedures to the knee. 313
SNAP. Recording Electrodes. There are two popular techniques

Reference Values. Negative peak latencies are 2.6 ± 0.2 ms used in assessing antidromic saphenous nerve conduction, i.e., .
with a range of 2.3-3.1 ms. Mean conduction velocities are re-
33
proximal and distal. Either may be used and serve as an alterna-
ported at 47.9 ± 3 . 7 m/s (43-55 m/s) with negative peak ampli- tive procedure when one fails to produce the expected response
tudes of 10-25 pV. in a particular patient. The proximal method may yield slightly
larger responses, but the stimulation site is harder to locate, es-
Posterior Femoral Cutaneous N e r v e pecially in persons with significant subcutaneous tissue about
The posterior femoral cutaneous nerve is formed by the ante- the knee. Distally, the anatomic landmarks are somewhat easier
rior and posterior divisions of sacral segments S I - S 3 with occa- to locate, but the nerve's diameter is reduced, resulting in rather
sional contributions from L4, L5, and S 4 . Once the nerve 144
small potentials. Patient relaxation is paramount in attempting
exits the pelvis, it passes anterior to the piriformis muscle and to record saphenous SNAPs.
posteromedial to the sciatic nerve. Inferior to the piriformis
muscle, multiple small branches are given off to innervate the Proximal Recording
perineum and inferior gluteal area. The main portion of the pos- E-L The knee is slightly flexed at 10-20°. The E-l recording
terior femoral cutaneous nerve then comes to lie in the inter- electrode position is located by first palpating the tendons of the
muscular groove between the medial and lateral hamstring sartorius and gracilis muscles 1 cm proximal to the inferior
muscles. Traversing the posterior thigh in this location, the border of the patella (Fig. 5-34). On a line 15 cm distal to the
216
nerve provides cutaneous sensation to the posterior aspect of the previously noted site to the medial border of the tibia, an E-l
thigh through numerous small branches arising from the main electrode is secured to the patient. Sufficient tape is required to
gracilis
sartorius
Figure 5-34. Saphenous nerve.The saphenous nerve is activated

along t h e medial aspect of t h e knee (S).The recording electrodes (E-
I {RJ and E-2 {R }) are positioned in the depression posterior t o t h e
r
tibia. (From Ma DM, Liveson JA: Nerve Conduction Handbook.

Philadelphia, F.A. Davis, 1983, with permission.)
Distal Recording
E-l. For recording the saphenous nerve SNAP distally, the
E-1 electrode is positioned in the anatomic depression just pos-
terior to the tibialis anterior tendon 3 cm proximal to the medial
malleolus' greatest prominence at the ankle region. 350
E-2. E-2 is placed posterior to the tibialis anterior tendon and

anterior to the medial malleolus' largest prominence.
Proximal Stimulation. The cathode is located at the pre-
viously designated site between the sartorius and gracilis
tendons 1 cm proximal to the inferior border of the patella
with the anode proximal (Fig. 5-34). Firm pressure combined
Figure 5-33. Posterior femoral cutaneous nerve. Recording with a pulse duration of 0.1 or 0.2 ms may be required to op-
and stimulating electrode location for evaluation of the posterior timally excite the saphenous nerve in this location depending
femoral cutaneous nerve. (From Dumitru D, Nelson MR: Posterior upon the amount of subcutaneous tissue. Stimulus artifact
femoral cutaneous nerve conduction. Arch Phys Med Rehabil may pose a particular problem with this SNAP because of the
1990;71:979-982, with permission.) small size and difficulty in obtaining the response. As a
result, the practitioner is strongly encouraged to rotate the
anode about the cathode prior to concluding that the re-
maintain the electrode in the depression posterior to the tibia sponse is unobtainable.
and the practitioner may wish to apply slight pressure to this Distal Stimulation. Cathodal placement is 14 cm proximal
electrode with a plastic pen during recording. to E-l between the medial border of the tibia and the gastrocne-
E-2. A surface E-2 electrode is placed 3 cm distal to E-l mius muscle in the anatomic depression between these two
along the medial border of the tibia in the anatomic depression structures. Similar comments regarding stimulus artifact noted
between the tibia and the gastrocnemius muscle. above for proximal stimulation apply.
Instrumentation Parameters. See lateral femoral cuta- Recording Electrodes. E-L A bar electrode with an inter-
neous nerve. eiectrode separation of 3 cm is located posterior to the lateral
Reference Values. With proximal stimulation, the latency malleolus with E-l proximal. E-l should be located at approxi-
measured to the SNAP'S onset is noted to be 2.5 ± 0.2 ms mately the level of the major prominence of the lateral malleolus.
(2.2-2.8 ms) with peak-to-peak amplitudes of 10.2 ± 2.1 pV E-2. An E-2 electrode imbedded in the plastic bar is posi-
(7.0-15.0 p V ) . Conduction velocity for this technique is 58.8
216
tioned as close to the lateral malleolus as possible without com-
± 2.3 m/s (53.7-63.7 m/s). For distal stimulation, peak latency promising E-l's location.
is 3.6 ± 0.4 ms with a mean amplitude of 9.0 ± 3.4 pV and con- Stimulation. On a line 14 cm proximal to E-l the sural
duction velocity approximating 41.7 ± 3.4 m/s. 350
nerve is stimulated with the cathode distal just lateral to the
leg's midline. Slight medial/lateral movement may be necessary
Superficial Peroneal Sensory N e r v e to position the cathode directly over the sural nerve. This is
The superficial peroneal sensory nerve is the cutaneous con- achieved when the amplitude of the response is maximal. The
tinuation of the superficial peroneal nerve that arises from the patient should describe a paresthesia associated with the stimu-
common peroneal nerve and is composed preferentially of nerve lus along the lateral aspect of the foot.
fibers originating from the L5 segment. This nerve becomes
144
Instrumentation Parameters. See lateral femoral cuta-
subcutaneous in the anatomic depression between the peroneus neous nerve.
longus and extensor digitorum tendons at the proximal portion Reference Values. At 14 cm with a skin temperature of
of the distal third of the leg. It travels anterior to the extensor about 32.2 ± 1.2°C, negative peak latency of 3.5 ± 0.2 ms with a
retinaculum at the ankle as two distinct sensory branches, i.e., conduction velocity of 39.6 ± 2.3 m/s and amplitude between
the intermediate and medial dorsal cutaneous nerves of the 10-50 pV is described. 292
foot. These nerves provide cutaneous sensation to the antero-

154
lateral aspect of the distal leg area and the majority of the foot's Medial/Lateral Plantar C o m p o u n d
dorsum except for that region between the first and second N e r v e Action Potentials
digits that is innervated by the deep peroneal nerve. A number of techniques exist to evaluate the sensory nerve
Recording Electrodes. A number of orthodromic and an- distribution of the medial and lateral plantar nerves, but the re-
tidromic techniques have been developed to examine this sponses are rather small and inconsistently obtained. Near-nerve
nerve; however, only a reliable and simple antidromic tech- needle recordings have also been described for orthodromic
nique is described. 157
evaluation of the medial/lateral plantar nerves, but most practi-
E-L This method records the response only from the inter- tioners prefer surface stimulation and recording. As a result, a re-
mediate dorsal cutaneous branch of the superficial peroneal sen- liable surface method of evaluating the mixed nerve action
sory nerve as it crosses the ankle. An E-1 electrode is positioned potential of the medial and lateral plantar nerves is described.
1-2 cm medial to the lateral malleolus. It is important to note Beneath the flexor retinaculum, the tibial nerve divides into
that this technique uses a bar electrode with an intereiectrode the calcaneal, medial, and lateral plantar nerves. The medial
65
separation of 3 cm. Slight repositioning may be required to opti- and lateral plantar nerves proceed into the foot to innervate the-
mally locate the nerve branch under investigation secondary intrinsic foot muscles. Specifically, the medial plantar nerve
slight anatomic variations. innervates the following muscles: abductor hallucis, flexor digi-
E-2. The E-2 electrode imbedded in the plastic bar is located torum brevis, flexor hallucis brevis, and first dorsal in-
distally. terosseous. The remaining foot muscle are innervated by the
144
Stimulation. The cathode is located 12 cm proximal to E-l lateral plantar nerve: abductor digiti minimi, flexor digiti
firmly pressed into the soft tissue just anterior to the anterior minimi, adductor hallucis, and flexor digitorum brevis. The
margin of the fibula. A pulse duration of 0.05 ms is recom- medial and lateral plantar aspects of the foot are innervated by
mended but some individuals may require a slightly greater the medial and lateral plantar nerves, respectively. Also, the
pulse duration. medial three and one-half digits are supplied by the medial
Instrumentation Parameters. See lateral femoral cuta- plantar nerve while the remainder receive cutaneous innervation
neous nerve. from the lateral plantar nerve.
Reference Values. The negative peak latency is noted to be Recording Electrodes. E-L An E-l electrode imbedded in
2.9 ± 0.3 ms with a negative peak amplitude of 20.5 ± 6.1 pV a plastic bar with an inter-electrode separation of 3 cm is posi-
and a conduction velocity of 65.7 ± 3.7 m/s. 157
tioned with E-l just proximal to the superior edge of the flexor
retinaculum over the tibial nerve. This site is located by con-
282
Sural N e r v e necting an imaginary line between the posterior tip of the calca-
Two branches, one from the tibial (medial sural nerve) and neus and the medial malleolus' prominence (Fig. 5-35). E-l is
the other from the peroneal (lateral sural nerve) nerve fuse just located just proximal to this line. The bar electrode is firmly
inferior to the popliteal fossa region to form the sural nerve pressed into the space posterior to the medial malleolus and se-
proper. The sural nerve has representation from primarily the S1 cured with as much tape as necessary to maintain close contact
nerve root. Proximally, the sural nerve traverses the groove
144
with the skin.
between the two head of the gastrocnemius muscle. At about the E-2. The E-2 electrode is positioned proximal to E-l on the
proximal lower third of the leg, it becomes subcutaneous and tibial nerve.
continues distally posterior to the lateral malleolus. The sural Stimulation. The site for medial plantar nerve excitation is
nerve then terminates along the lateral border of the foot. The located by first measuring 10 cm from E-l to the interspace be-
cutaneous regions innervated by the sural nerve are the distal tween the first and second metatarsals on the plantar aspect of
dorsal lateral aspect of the leg and lateral portion of the foot. the foot. One then continues an additional 4 cm toward the
Although multiple techniques have been described, only a reli- digits on this inter-metatarsal line to stimulate the medial plan-
able antidromic method is described. 292
tar nerve (Fig. 5-35). The cathode is situated such that it faces
Temperatures on the plantar aspect of the foot should exceed

29°C.
CONCLUSION
After completing this chapter, one can appreciate the intricate
interconnections between neural anatomy and impulse conduc-
tion. The anatomic arrangement of the nerve on a microscopic
level serves to provide a flexible and strong "cable" system to
convey a self-sustaining action potential. This action potential
can be used to determine the peripheral nervous system's physi-
ologic status with respect to health or disease. Multiple tech-
niques are available to assess individual nerve integrity in both
the upper and lower limbs. It is essential for the practitioner to
master most of these techniques without referring to a technical
manual. This is possible only with a great deal of practice under
the supervision of a recognized expert. All of the previous prin-
ciples of volume conduction and instrumentation converge on
the proper performance of technically demanding nerve con-
duction techniques. The information gained from the nerve con-
duction portion of the electrodiagnostic medicine consultation
is combined with the history and physical examination to assist
in the formation of an appropriate diagnosis.
REFERENCES
1. Abbott BL, Hill AV, Howarth JV: The positive and negative heat production
associated with a nerve impulse. Proc R Soc Lond (Series B) 1958; 148:167—
175.
2. Abood LG, Abul-Haj SF: Histochemistry and characterization of hyaluronic
acid in axons of peripheral nerve. J Neurochem 1956;1:119-125.
Figure 5-35. Medial/lateral plantar nerve. Locations for stimulat-
3. Angevine JB: The nervous tissue. In Fawcett DW (ed): Bloom and Fawcett: A
ing and recording electrodes for the medial plantar nerve are depicted in Textbook of Histology. Philadelphia, W.B. Saunders, 1986, pp 311-366.
the upper trace. Lateral plantar nerve excitation 14 cm distal t o record- 4. Apfelberg DB, Larson SJ: Dynamic anatomy of ulnar nerve at elbow. Plast
ing electrodes located proximal t o the medial flexor retinaculum is Reconstr Surg 1973;51:76-81.
5. Arnold N, Harriman GF: The incidence of abnormality in control human pe-
shown in the lower trace. (From Saeed MA, Gatens PF: Compound nerve ripheral nerves studied by single axon dissection. J Neurol Neurosurg
action potentials of t h e medial and lateral plantar nerves through the Psychiatry 1970;33:55-61.
tarsal tunnel.Arch Phys Med Rehabil 1982;63:304-307, with permission.) 6. Baer RD, Johnson EW: Motor nerve conduction velocities in normal children.
Arch Phys Med Rehabil 1965;46:698-704.
7. Barton AA: An electron microscope study of degeneration and regeneration of
nerve. Brain 1962;85:799-806
the calcaneus while the anode is direct toward the digits. Lateral 8. Behse F, Buchthal F: Normal sensory conduction in the nerves of the leg in
plantar nerve stimulation occurs at the intersection of the origi- man. J Neurol Neurosurg Psychiatry 1971;34:404-414.
nal line extending 4 cm beyond the previously noted 10 cm site 9. Berthold CH, Skoglund S: Histochemical and ultrastructural demonstration of
mitochondria in the paranodal region of developing feline spinal roots and
and the interspace between the fourth and fifth metatarsals (Fig. nerves. Acta Soc Med Upsal 1967;72:37-70.
5-35). The cathode and anode are positioned in a similar manner 10. Berthold CH: Ultrastructure of the node-paranode region of mature feline ven-
to that for medial plantar nerve excitation. A pulse duration of tral lumbar spinal root fibres. Acta Soc Med Uppsala 1968;73(Suppl 9):37-70.
0.2 ms or more may be necessary to deliver a supramaximal 1 1. Berthold CH: Morphology of normal peripheral axons. In Waxman SG (ed):
Physiology and Pathobiology of Axons. New York, Raven Press. 1978, pp 3 - 6 3 .
stimulus as the plantar aspect of most persons is rather thick. 12. Betts RP, Johnson DM, Brown BH: Nerve fibre conduction velocity and refrac-
Stimulus artifact poses a rather significant problem in most in- tory period distributions in nerve trunks. J Neurol Neurosurg Psychiatry
dividuals and rotating the anode about the cathode is often of 1976;39:694-700.
i 3. Black JA, Waxman SG, Friedman B, et al: Sodium channels in astrocytes of rat
great help in obtaining the desired responses. Firm pressure into
optic nerve in situ: Immuno-electron microscopic studies. GLIA 1989: 1:353¬
the sole of the foot is often of assistance in eliciting the mixed 369.
nerve action potential. 14. Black JA. Friedman B, Waxman SG. et al: Immuno-ultrastructural localization
- of sodium channels at nodes of Ranvier and perinodal astrocytes in rat optic
Instrumentation Parameters. Because the mixed nerve re-
nerve. Proc R Soc Lond (Series B) 1989;238:39-51.
sponse is similar to a SNAP in amplitude and duration, typical 15. Black JA, Kocsis JD, Waxman SG: Ion channel organization of the myelinated
sensory nerve settings are used. As a result, refer to the lateral fiber. Trends Neurosci 1990;13:48-54.
femoral cutaneous nerve for appropriate parameters. 16. Black MM, Ladek RJ: Slow components of axonal transport: Two cytoskeletai
networks. J Cell Biol 1980;86:616-623.
Reference Values. When the above methods are followed, 17. Bolton CF, Carter KM: Human sensory nerve compound action potential ampli-
one should record medial plantar nerve mixed nerve action po- tude: Variation with sex and finger circumference. J Neurol Neurosurg
tentials with a negative peak latency of 3.2 ± 0.3 ms (2.6-3.7 Psychiatry 1980;43:925-928.
18. Bolton CF, Sawa GM, Carter K: The effects of temperature on human com-
ms) and an amplitude greater than 10 pV (10-30 pV). Lateral 282
pound action potentials. J Neurol Neurosurg Psychiatry 1981;44:407-413.

plantar nerve latencies should approximate 3.1 ± 0.3 ms 19. Bolton CF, Carter K, Koval JJ: Temperature effects on conduction studies of
(2.7-3.7 ms) with amplitudes greater than 8 pV ( 8 - 2 0 pV). normal and abnormal nerve. Muscle Nerve 1982;5:SI45-S147.
218 — PART 11 BASIC AND ADVANCED TECHNIQUES
20. Bolton CF: AAEM Minimonograph #40: Clinical neurophysiology of the respi- 56. Cummins KL, Dorfman LJ: Nerve fiber conduction velocity distributions:
ratory system. Muscle Nerve 1993;16:809-818. Studies of normal and diabetic human nerves. Ann Neurol 1981;9:67-74.
21. Bors E, French JD: Management of paroxysmal hypertension following injuries to 57. Daube JR: Percutaneous palmar median nerve stimulation for carpal tunnel syn-
cervical and upper thoracic segments of spinal cord. Arch Surg 1952;64:803-812. drome. Electroenceph Clin Neurophysiol 1977;43:139-140.
22. Bostock H, Sherrat RM, Sears A: Overcoming conduction failures in demyeli- 58. Daube JR: Nerve conduction studies. In Aminoff MJ (ed): Electrodiagnosis in
nated nerve fibres by prolonging action potentials. Nature 1978;274:385-387. Clinical Neurology. New York, Churchill Livingstone, 1986, pp 265-306.
23. Brasher A, Kincaid JC: The influence of the reference electrode on CMAP con- 59. Davis JN: Phrenic nerve conduction in man. J Neurol Neurosurg Psychiatry
figuration: Leg nerve observations and an alternative reference site. Muscle 1967;30:420-426.
Nerve 1996;19:63-67. 60. Dawson GD, Scott JW: The recording of nerve action potentials through the
24. Braddom RL, Johnson EW: Standardization of H reflex and diagnostic use in skin in man. J Neurol Neurosurg Psychiatry 1949;12:259-267.
SI radiculopathy. Arch Phys Med Rehabil 1974;55:161-166. 61. Dawson GD, Merton PA: "Recurrent" discharges from motoneurones. XX
25. BrismarT: Potential clamp analysis of membrane currents in rat myelinated International Congress of Physiology, Brussels, Abstract of Communications,
nerve fibres. J Physiol 1980;298:171-184. 1956, pp 221-222.
26. Bromberg MB, Spiegelberg T: The influence of active electrode placement on 62. de Jesus PV, Hausmanowa-Petrusewicz I, Barchi RL: The effect of cold on
CMAP amplitude. Electroencephalogr Clin Neurophysiol 1997;105:385-389. nerve conduction of human slow and fast nerve fibers. Neurology
27. Brown BH: Theorectical and experimental waveform analysis of human com- 1973;23:1182-1189.
pound nerve action potentials using surface electrodes. Med Biol Engin 1968;6: 63. Delbeke J, Kopec J, McComas AJ: Effects of age, temperature, and disease on
375-386. the refractoriness of human nerve and muscle. J Neurol Neurosurg Psychiatry
28. Brown WF: The Physiological and Technical Basis of Electromyography. 1978;41:65-71.
Boston, Butterworth Publishers, 1984. 64. DeLisa JA, Mackenzie K, Baran EM: Manual of Nerve Conduction Velocity
29. Buchthal F, Pinelli P, Rosenfalk P: Action potential parameters in normal and Somatosensory Evoked Potentials, 2nd ed. New York, Raven Press,
human muscle and their physiological determinants. Acta Physiol Scand 1987.
1954;32:219-229. 65. Dellon AL, Mackinnon SE: Tibial nerve branching in the tarsal tunnel. Arch
30. Buchthal F, Rosenfalek A: Evoked action potentials and conduction velocity in Neurol 1984;41:645-646.
human sensory nerves. Brain Res 1966;3:1-122. 66. Del Toro DR, Park TA: Abductor hallucis false motor points: Electrophysio-
31. Buchthal F, Trojaborg W: Electrophysiological findings in entrapment of the logic mapping and cadaveric dissection. Muscle Nerve 1996; 19:1138-1143.
median nerve wrist and elbow. J Neurol Neurosurg Psychiatry 1974;37:340-360. 67. Denys EH: M wave changes with temperature in amyotrophic lateral sclerosis
32. Buchthal F, Behse F: Sensory action potentials and biopsy of the sural nerve in and disorders of neuromuscular transmission. Muscle Nerve 1990;13:613-617.
neuropathy. In Canal L, Pozza G (eds): Peripheral Neuropathies. Amsterdam, 68. Devi S, Lovelace RE, Duarte N: Proximal peroneal nerve conduction velocity:
Elsevier, 1978, p 1. Recording from anterior tibial and peroneus brevis muscles. Ann Neurol
33. Butler ET, Johnson EW, Kaye ZA: Normal conduction velocity in the lateral 1977;2:116-119.
femoral cutaneous nerve. Arch Phys Med Rehabil 1974;55:31-32. 69. Dick HC, Bradley WE, Scott FB, et al: Pudendal sexual reflexes. Urology
34. Campbell WW, Ward LC, Swift TR: Nerve conduction velocity varies inversely 1974;3:376-379.
with height. Muscle Nerve 1981;4:520-523. 70. Dioszeghy P, Stalberg E: Changes in motor and sensory nerve conduction para-
35. Cannieu A: Note sur une anastomose entre le branche profunde de cubital et le meters with temperature in normal and diseased nerve. Electroencephalogr Clin
median. Bull Soc Anat Physiol Horm Path Bordeaux 1897;17:339-342. Neurophysiol 1992;85:229-235.
36. Carpendale MTF: Conduction time in the terminal portion of the motor fibers 71. Dolman CL, McCormick AQ, Drance SM: Aging of the optic nerve. Arch
of the ulnar, median, peroneal nerve in healthy subjects and in patients with Opthalmol 1980;98:2053-2058.
neuropathy. Unpublished thesis, Mayo Foundation, University of Minnesota, 72. Dorfman LJ, Cummins KJL, Abraham GS: Conduction velocity distributions of
Rochester, 1956. the human median nerve: Comparison of methods. Muscle Nerve 1982;5:
37. Carroll GJ: Comparison of median and radial nerve sensory latencies in the S148-S153.
electrophysiological diagnosis of carpal tunnel syndrome. Electroenceph Clin 73. Dorfman LJ: The distribution of conduction velocities (DCV) in peripheral
Neurophysiol 1987;68:101-106. nerves: A review. Muscle Nerve 1984;7:2-11.
38. Causey G: The Cell of Schwann. Edinburgh, Livingstone, 1960. 74. Douglas WW, Malcolm JL: The effect of localized cooling on conduction in cat
39. Chassin SL, Little JW, DeLisa JA: Compound nerve action potentials from the nerves. J Physiol 1955;130:53-71.
median and ulnar nerves. Arch Phys Med Rehabil 1987;68:31-35. 75. Downie AW, Scott TR: An improved technique for radial nerve conduction
40. Checkles NS, Bailey JA, Johnson EW: Tape and caliper surface measurements studies. J Neurol Neurosurg Psychiatry 1967;30:332-336.
in determination of peroneal nerve conduction velocity. Arch Phys Med Rehabil 76. Droz B, Rambourg A, Koenig HL: The smooth endoplasmic reticulum:
1969;50:214-218. Structure and role in the renewal of axon membrane and synaptic vesicles by
41. Checkles NS, Russakov AD, Piero DL: Ulnar nerve conduction velocity: fast axonal transport. Brain Res 1975;93:1-13.
Effect of elbow position on measurement. Arch Phys Med Rehabil 1971 ;52: 77. Droz B, Koenig HL, Di Gamberardino JY, et al: The importance of axonal
362-365. transport and endoplasmic reticulum in the function of the cholinergic synapse
42. Chen R, Collins S, Remtulla H, el al: Phrenic nerve conduction study in normal in normal and pathological conditions. In Tucek S (ed): Progress in Brain
subjects. Muscle Nerve 1995;18:330-335. Research: The Cholinergic Synapse, Vol. 49. Amsterdam, Elsevier, 1979, pp
43. Cherington M: Long thoracic nerve: Conduction studies. Dis Nerv Syst 23-44.
1972;33:49-51. 78. Dumitru D, Walsh NE: Practical instrumentation and common sources of error.
44. Chiu SY, Ritchie JM, Rogart RB, et al: A quantitative description of membrane Am J Phys Med Rehabil 1988;67:55-65.
currents in rabbit myelinated nerve. J Physiol 1979;292:149-166. 79. Dumitru D, Walsh NE, Porter LD: Electrophysiologic evaluation of the facial
45. Chiu SY: Asymmetry currents in the mammalian myelinated nerve. J Physiol nerve in Bell's palsy. Am J Phys Med Rehabil 1988:67:137-144.
1980;309:499-519. 80. Dumitru D, Walsh NE, Weber CF: Electrophysiologic study of the Riche-
46. Chiu SY, Ritchie JM: Ionic and gating currents in mammalian myelinated Cannieu anomaly. Electromyogr Clin Neurophysiol 1988;28:27-31.
nerve. Adv Neurol 1981;31:313-328. 81. Dumitru D, Walsh NE, Ramamurthy S: The premotor potential. Arch Phys Med
47. Chodoroff G, Tashjian EA, Ellenberg MR: Orthodromic vs. antidromic sensory Rehabil 1989;70:537-540.
nerve latencies in healthy persons. Arch Phys Med Rehabil 1985;66:589-591. 82. Dumitru D, Nelson MR: Posterior femoral cutaneous nerve conduction. Arch
48. Clifton EE: Unusual innervation of the intrinsic muscles of the hand by median Phys Med Rehabil 1990;71:979-982.
and ulnar nerve. Surgery 1948;23:12-31. 83. D'Urso D, Brophy PJ, Staugaitis SM, et al: Protein zero of peripheral nerve
49. Cohn TG. Wertsch JJ, Pasupuleti DV, et al: Nerve Conduction Studies: myelin: Biosynthesis, membrane insertion and evidence for homotypic interac-
Orthodromic vs. antidromic latencies. Arch Phys Med Rehabil 1990;71:579¬ tion. Neuron 1990;4:449-^60.
582. 84. Eduardo E, Burke D: The optimal recording electrode configuration for com-
50. Cooper PD, Smith RS: The movement of optically detectable organelles in pound sensory action potentials. J Neurol Neurosurg Psychiatry 1988;51:684¬
myelinated axons of Xenopus laevis. J Physiol 1974;242:77-97. 687.
51. Corbin KB, Gardner ED: Decrease in number of myelinated fibers in human 85. Eisen A: Early diagnosis of ulnar nerve palsy: An electrophysiologic study.
spinal roots with age. Anat Rec 1937;68:63-74. Neurology 1974;24:256-262.
52. Cottrell L: Histologic variations with age in apparently normal peripheral nerve 86. Ellisman MH: Molecular specializations of the axon membrane at nodes of
trunks. Arch Neurol Psychiatry 1940;43:1138-1150. Ranvier are not dependent upon myelination. J Neurocytol 1979;8:719-735.
53. Critchlow JF, Seybold ME, Jablecki CJ: The superficial radial nerve: techniques 87. Ellisman MH, Friedman PL, Hamilton WJ: The localization of sodium and cal-
for evaluation. J Neurol Neurosurg Psychiatry 1980:43;929-933. cium to Schwann cell paranodal loops at nodes of Ranvier and of calcium to
54. Crutchfield CA, Gutmann L: Hereditary aspects of accessory deep peroneal compact myelin. J Neurocytol 1980;9:185-205.
nerve. J Neurol Neurosurg Psychiatry 1973;36:989-990. 88. Elmer LW, Black JA, Waxman SG, et al: The voltage-dependent sodium chan-
55. Crutchfield CA, Gutmann L: Hereditary aspects of median-ulnar nerve commu- nel in mammalian CNS and PNS: Antibody characterization and immunocyto-
nications. J Neurol Neurosurg Psychiatry 1980;43:53-55. chemical localization. Brain Res 1990;532:222-231.
Chapters NERVE C O N D U C T I O N STUDIES — 219
89. Erlanger J, Gasser HS: The compound nature of the action current of nerve as 121. Gruber W: Uber die Verbindung des Nervus Medianus mil dem Nervus Ulnaris
disclosed by the cathode ray oscillograph. Am J Physiol 1924;70:624-666. am Unterarm des Menshen und der Saugethiere. Arch Anat Physiol Wissenshaft
90. Erlanger J, Gasser HS, Bishop GH: The absolute refractory phase of the alpha, Med 1870;37:501-522.
beta and gamma fibers in the sciatic nerve of the frog. Am J Physiol 1927;81: 122. Gutmann L: Atypical deep peroneal neuropathy. J Neurol Neurosurg Psychiatry
473-474. 1970;33:453-456.
91. Esplin DW: Independence of conduction velocity among myelinated fibers in 123. Gutmann L: Median-ulnar nerve communications and carpal tunnel syndrome.
cat nerve. J Neurophysiol 1962;25:805-811. J Neurol Neurosurg Psychiatry 1977;40:982-986.
92. Evans BA, Litchy WJ, Daube JR: The utility of magnetic stimulation for routine 124. Haak RA, Kleinhaus FW, Ochs S: The viscosity of mammalian nerve axoplasm
peripheral nerve conduction studies. Muscle Nerve 1988; 11:1074-1078. measured by electron spin resonance. J Physiol 1976;263:115-137.
93. Farmer TW, Buchthal F, Rosenfalek P: Refractory period of human muscle after 125. Halar EM, DeLisa JA, Brozovich FV: Nerve conduction velocity: Relationship
the passage of a propagated action potential. Electroenceph Clin Neurophysiol of skin, subcutaneous and intramuscular temperatures. Arch Phys Med Rehabil
1960; 12:455^66. 1980;61:199-203.
94. Fawcett DW: A Textbook of Histology. Philadelphia, W.B. Saunders, 1986, pp 126. Halar EM, DeLisa JA: Peroneal nerve conduction velocity: The importance of
311-366. temperature control. Arch Phys Med Rehabil 1981;62:439-443.
95. Feibel A, Foca FJ: Sensory conduction of radial nerve. Arch Phys Med Rehabil 127. Halar EM, DeLisa JA, Soine TL: Nerve conduction studies in upper extremi-
1974;55:314-316. ties: Skin temperature corrections. Arch Phys Med Rehabil 1983;64:412-416.
96. Fernand VSV, Young JT: The size of the nerve fibres of muscle nerves. Proc R 128. Hallin RG, Torebjork HE: Single unit activity in human skin nerves during rest
Soc 1951;139:38-58. and various maneuveres. Acta Physiol Scand 1974;92:303-317.
97. Fiaschi A: Observations on sensory nature of intramuscular nerve action poten- 129. Hansen CC, Reske-Nielsen E: Pathological studies in presbyacusis. Arch
tial. J Neurol Neurosurg Psychiatry 1973;36:509-513. Otolaryngol 1965;82:115-132.
98. Fitz WR, Mysiw WJ, Johnson EW: First lumbrical latency and amplitude: 130. Harding C, Halar E: Motor and sensory ulnar nerve conduction velocities:
Control values and findings in carpal tunnel syndrome. Am J Phys Med Rehabil Effect of elbow position. Arch Phys Med Rehabil 1983;64:227-232.
1990;69:198-200. 131. Harness D, Sekeles E, Chaco J: The double anastomotic innervation of thenar
99. Formann DS, Padjen AL, Siggins GR: Axonal transport of organelles visualized muscles. J Anat 1974;117:239-331.
by light microscopy: Cinematographic and computer analysis. Brain Res 132. Henrikson JD: Conduction velocity of motor nerves in normal subjects and pa-
1977;136:197-213. tients with neuromuscular disorders. Thesis. University of Minnesota,
100. Formann DS, Padjen AL, Siggins GR: Effect of temperature on the rapid retro- Minneapolis, 1956.
grade transport of microscopically visible intra-axonal organelles. Brain Res 133. Hess A, Young JZ: The nodes of Ranvier. Proc R Soc Lond 1952;140:301-320.
1977;136:215-226. 134. Hill DK: The volume change resulting from the stimulation of a giant nerve
101. Friede RL, Samorajski T: Axon caliber related to neurofilaments and micro- fibre. J Physiol 1950;111:304-327.
tubules in sciatic nerve fibers of rats and mice. Anat Rec 1970;167:379-387. 135. Hill AV, Howarth JV: The initial heat production of stimulated nerve. Proc R
102. Fu R, DeLisa JA, Kraft GH: Motor nerve latencies through the tarsal tunnel in Soc Lond 1958;148:149-187.
normal adult subjects: Standard determinations corrected for temperature and 136. Hille B: Potassium channels in myelinated nerve: Selective permeability to
distance. Arch Phys Med Rehabil 1980;61:243-248. small cations. J Gen Physiol 1973;61:669-686.
103. Gamble HJ: Comparative electron microscopic observations on the connective 137. Hiscoe HB: Distribution of nodes and incisures in normal and regenerated nerve
tissues of a peripheral nerve and a spinal nerve root in the rat. J Anat 1964;98: fibers. Anat Rec 1947;99:447-475.
17-26. 138. Hodes R, Larrabee MG, German W: The human eleetromyogram in response to
104. Gamble HJ, Eames RA: An electron microscope study of the connective tissues nerve stimulation and conduction velocity of motor axons. Arch Neurol
of human peripheral nerve. J Anat 1964;98:655-663. Psychiatry 1948;60:340-365.
105. Gassel MM: A test of nerve conduction to muscles of the shoulder girdle as an 139. Hodes R: Selective destruction of large motor units by poliomyelitis virus. 1.
aid in the diagnosis of proximal neurogenic and muscular disease. J Neurol Conduction velocity of motor nerve fibers of chronic poliomyelitis patients. J
Neurosurg Psychiatry 1964;27:200-205. Neurophysiol 1948;12:257-266.
106. Gassel MM, Trojaborg W: Clinical and electrophysiological study of the pattern 140. Hodgkin AL, Katz B: The effect of temperature on the electrical activity of the
of conduction times in the distribution of the sciatic nerve. J Neurol Neurosurg giant axon of the squid. J Physiol 1949;109:240-249.
Psychiatry 1964;27:351-357. 141. Hodgkin AL, Huxley AF: A quantitative description of membrane current and
107. Gasser HS, Erlanger J: The role played by the sizes of the constituent fibers of a its application to conduction and excitation in nerve. J Physiol
nerve trunk in determining the form of its action potential wave. Am J Physiol 1952;117:500-544.
1927;80:522-547. 142. Hodgkin AL: The Conduction of the Nervous Impulse. Liverpool, University
108. Gasser HS, Grundfest H: Axon diameters in relation to the spike dimensions Press, 1971.
and the conduction velocity in mammalian A fibers. Am J Physiol 1939; 127: 143. Hoffman PN, Ladek RJ: The slow component of axonal transport: Identification
393^14. of major structural polypeptides of the axon and their generality among mam-
109. Gasser HS: Unmedullaied fibers originating in dorsal root ganglia. J Gen malian neurons. J Cell Biol 1975;66:351-366.
Physiol 1950;33:651-690. 144. Hollinshead WH: Anatomy for Surgeons: The Back and Limbs, 3rd ed.
110. Gasser HS: Properties of dorsal root unmedullated fibers on the two sides of the Philadelphia, Harper & Row, 1982.
ganglion. J Gen Physiol 1955;38:709-728. 145. Holtzman E, Novikoff AB: Lysosomes in the rat sciatic nerve following crush. J
111. Geren B: The formation from the Schwann cell surface of myelin in the periph- Cell Biol 1965;27:651-669.
eral nerves of chick embryos. Exp Cell Res 1954;7:558-562. 146. Honet JC, Hebsen RH, Perrin EB: Variability of nerve conduction velocity de-
112. Ghabriel MN, Allt G: The role of Schmidt-Lanterman incisures. I. A quantita- terminations in normal persons. Arch Phys Med Rehabil 1968;49:650-654.
tive teased fibre study of remyelinating peripheral nerve fibres. Acta 147. Horning MR, Kraft GH, Guy A: Latencies recorded by intramuscular needle
Neuropathol 1979:48:93. electrodes in different portions of a muscle: Variation and comparison with sur-
113. Ghabriel MN, Allt G: Schmidt-Lanterman incisures. I. A quantitative teased face electrodes. Arch Phys Med Rehabif 1972;53:206-211.
fibre study of remyelinating peripheral nerve fibres. Acta Neuropathol 148. Hursh JB: The conduction velocity and diameter of nerve fibers. Am J Physiol
1980;52:85-95. 1939;127:131-139.
114. Gilliatt RW, Sears TA: Sensory nerve action potentials in patients with periph- 149. Huxley AF. Stampfli R: Saltatory transmission of the nervous impulse. Arch Sci
eral nerve lesions. J Neurol Neurosurg Psychiatry 1958;21:109-118. Physiol 1949;3:435-448.
115. Gilliatt RW, Thomas PK: Changes in nerve conduction with ulnar lesions at the 150. Huxley AF, Stampfli R: Evidence for saltatory conduction in peripheral myeli-
elbow. J Neurol Neurosurg Psychiatry 1960:23:312-320. nated nerve fibres. J Physiol 1949;108:315-339.
116. Gilliatt RW, Melville ID, Velate AS et al: A study of normal nerve action poten- 151. Infante E, Kennedy WR: Anomalous branch of the peroneal nerve detected by
tials using an averaging technique (barrier grid storage tube). J Neurol electromyography. Arch Neurol 1970;22:162-165.
Neurosurg Psychiatry 1965;28:191-200. 152. Ishikawa H: Identification and distribution of intracellular filaments. In Hatano
117. Golovchinsky V: Ulnar-to-median anastomosis and its role in the diagnosis of S, Ishikawa H, Sato H (ed): Cell Motility: Molecules and Organization.
lesions of the median nerve at the elbow and the wrist. Electromyogr Clin Baltimore, MD, University Press, 1979. pp 417-444.
Neurophysiol 1990;30:31-34. 153. Iyer V, Fenichel GM: Normal median nerve proximal latency in carpal tunnel
118. Gordon TR, Kocsis JD, Waxman SG: Evidence for the presence of two types of syndrome: A clue to coexisting Martin-Gruber anastomosis. J Neurol
potassium channels in the rat optic nerve. Brain Res 1988;447:1-9. Neurosurg Psychiatry 1976;39:449-452.
119. Grafstein B: Axonal transport: The intracellular traffic of the neuron. In Brookhart 154. Izzo KL, Sridhara CR, Lemont H, et al: Sensory conduction studies of the
JM, Mountcastle VB (ed): Handbook of Physiology: The Nervous System (Vol 1. branches of the superficial peroneal nerve. Arch Phys Med Rehabil 1981 ;62:
Part 1). Belhesda, MD, American Physiological Society, 1977, pp 691-717. 24-27.
120. Gregersen G: Diabetic neuropathy: Influence of age, sex, metabolic control, 155. Jablecki CK. Andary MT, So YT, et al: Literature review of the usefulness of
and duration of diabetes on motor conduction velocity. Neurology 1967; 17: nerve conduction studies and electromyography for the evaluation of patients
972-980. with carpal tunnel syndrome. Muscle Nerve 1993;16:1392-1414.
156. Jabre JF: Ulnar nerve lesions at the wrist: New technique for recording from the 191. Komar J, Szegvari M, Gloviczky Z: Traumatischer Durchschnitt des Nervus
sensory dorsal branch of the ulnar nerve. Neurology 1980;30:873-876. medianus ohne komplette motorische Parese: Martin-Grubersche Anastomose.
157. Jabre JF: The superficial peroneal sensory nerve revisited. Arch Neurol Nervenarzt 1978;49:697-699.
1981;38:666-667. 192. Kornfield MJ, Cerra J, Simons DG: Stimulus artifact reduction in nerve con-
158. Jabre JF: Surface recording of the H-reflex of the flexor carpi radialis. Muscle duction. Arch Phys Med Rehabil 1985;66:232-235.
Nerve 1981;4:435-438. 193. Kraft GH: Axillary, musculocutaneous and suprascapular nerve latency studies.
159. Jebsen RH: Motor conduction velocity of distal radial nerve. Arch Phys Med Arch Phys Med Rehabil 1972;53:383-387.
Rehabil 1966;47:12-16. 194. Kraft GH, Halvorson GA: Median nerve residual latency: Normal value and use
160. Jebsen RH: Motor conduction velocity in proximal and distal segments of the in diagnosis of carpal tunnel syndrome. Arch Phys Med Rehabil 1983;64:
radial nerve. Arch Phys Med Rehabil 1966:47:597-602. v
221-226.
161. Jennekens FGI, Tomlinson BE, Walton JN: Histochemical aspects of five limb 195. Kuczmarski ER, Rosenbaum JL: Studies on the organization and localization of
muscles in old age: An autopsy study. J Neurol Sci 1971;14:259-276. actin and myosin in neurons. J Cell Biol 1979;80:356-371.
162. Jimenez J, Easton JKM, Redford JB: Conduction studies of the anterior and 196. Lambert EH: The accessory deep peroneal nerve: A common variation in inner-
posterior tibial nerves. Arch Phys Med Rehabil 1970;51:164-169. vation of extensor digitorum brevis. Neurology 1969; 19:1169-1176.
163. Johnson EW, Olsen KJ: Clinical value of motor nerve conduction velocity de- 197. Lambert EH: Electromyography in amyotrophic lateral sclerosis. In Norris JR,
termination. JAMA 1960;172:2030-2035. Kurland LT (eds): Motor Neuron Diseases, New York, Grune & Stratton, 1969,
164. Johnson EW, Melvin JL: Sensory conduction studies of median and ulnar pp 135-153.
nerves. Arch Phys Med Rehabil 1967;48:25-30. 198. Lambert EH, Dyck PJ: Compound action potentials of sural nerve in vitro in
165. Johnson EW, Wood PK, Powers JJ: Femoral nerve conduction studies. Arch peripheral neuropathy. In Dyck PJ, Thomas PK, Lambert EH, Bunge R (ed):
Phys Med Rehabil 1968;49:528-532. Peripheral Neuropathy, 2nd ed. Philadelphia, W.B. Saunders, 1984, pp
166. Johnson EW, Kukla RD, Wongsam PE, et al: Sensory latencies to the ring 1030-1044.
finger: Normal values and relation to carpal tunnel syndrome. Arch Phys Med 199. Landon DN, Hall S: The myelinated nerve fibre. In Landon DN (ed): The
Rehabil 1981;62:206-208. Peripheral Nerve. London, Chapman & Hall, 1976, pp 1-105.
167. Johnson EW, Sipski M, Lammertse T: Median and radial sensory latencies to 200. Landon DN: Structure of normal peripheral myelinated nerve fibres. In
digit I: Normal values and usefulness in carpal tunnel syndrome. Arch Phys Waxman SG, Ritchie JM (eds): Demyelinating Diseases: Basic and Clinical
Med Rehabil 1987;68:140-141. Electrophysiology. New York, Raven Press, 1981, pp 25-49.
168. Johnson RK, Shrewsbury MM: Anatomical course of the thenar branch of the 201. Landon DN: The structure of the nerve fiber. In Culp WJ, Ochoa J (eds):
median nerve—usually in a separate tunnel through the transverse carpal liga- Abnormal Nerves and Muscles as Impulse Generators. New York, Oxford
ment. J Bone Joint Surg 1970;52A:269-273. University Press, 1982, pp 27-53.
169. Kanakamedala RV, Simons DG, Porter RW, et al: Ulnar nerve entrapment at the 202. Lang AH, Forsstrom J, Bjorkqvist SE, et al: Statistical variation of nerve con-
elbow localized by short segment stimulation. Arch Phys Med Rehabil 1988; duction velocity: An analysis in normal subjects and uraemic patients. J Neurol
69:959-963. Sci 1977;33:229-241.
170. Kaplan P, Sahgal V: Residual latency: New applications of an old technique. 203. Lang AH, Puusa A: Dual influence of temperature on compound nerve action
Arch Phys Med Rehabil 1978;59:24-27. potential. J Neurol Sci 1981;51:81-88.
171. Kaplan PE: Electrodiagnostic confirmation of long thoracic nerve palsy. J 204. Langley OK: A comparison of the binding of Alcian blue and inorganic cations
Neurol Neurosurg Psychiatry 1980;43:50-52. to polyanions in peripheral nerve. Histochem J 1971;3:251-260.
172. Kerjaschki D, Stockinger L: Zur Struktur und Funktion des Perineuriums. Z 205. Lapides J, Bobbitt JM: Diagnostic value of bulbocavernosus reflex. JAMA
Zellforsch Mikrosk Anat 1970; 110:386-400. 1956;162:971-972.
173. Kim DJ, Kalantri A, Guha S, et al: Dorsal cutaneous ulnar nerve conduction: 206. Laroy V, Spaans F, Reulen J: The sensory innervation pattern of the fingers. J
Diagnostic aid in ulnar neuropathy. Arch Neurol 1981;38:321-322. Neurol 1998;245:294-298.
174. Kimura I, Ayyar DR, Lippmann SM: Electrophysiological verification of the 207. Lascelles RG, Thomas PK: Changes due to age in internodal length in the sural
ulnar to median nerve communications in the hand and forearm. Tohoku J Exp nerve in man. J Neurol Neurosurg Psychiatry 1966;29:40-44.
Med 1983: 141:269-274. 208. Lasek RJ: Protein transport in neurons. Int Rev Neurobiol 1970;13:289-324.
175. Kimura J: Electrophysiological study of anomalous innervation of intrinsic 209. Lillie RS: Factors affecting transmission and recovery in the passive iron wire"
hand muscles. Arch Neurol 1976;33:842-844. model. J Gen Physiol 1925;7:473-507.
176. Kimura J: A method for estimating the refractory period of motor fibers in the 210. Lloyd DPC: Neuron patterns controlling transmission of ipsilateral hind limb
human peripheral nerve. J Neurol Sci 1976;28:485-490. reflexes in cat. J Neurophysiol 1943;6:293-315.
177. Kimura J: Collision technique: Physiologic block of nerve impulses in studies 211. Louis A A, Hotson JR: Regional cooling of human nerve and slowed Na inacti-
+
of motor nerve conduction velocity. Neurology 1976; 26:680-682. vation. Electroenceph Clin Neurophysiol 1986;63:371-375.
178. Kimura J, Murphy MJ, Varda DJ, et al: Electrophysiological study of ana- 212. Lowitzsch K, Hopf HC, Galland J: Changes in sensory conduction velocity and
malous innervation of intrinsic hand muscles. Trans Am Neurol Assoc refractory periods with decreasing tissue temperature in man. J Neurol 1977;
1976:101:257-259. 216:181-188.
179. Kimura J: A method for determining median nerve conduction velocity across 213. Lubinska L, Lukaszewska I: Shape of myelinated nerve fibres and proximo-
the carpal tunnel. J Neurol Neurosurg Psychiatry 1978;41:1-10. distal flow of axoplasm. Acta Biol Exp (Wars) 1956;17:115-131.
180. Kimura J: The carpal tunnel syndrome: Localization of conduction abnormali- 214. Ludin HP, Beyeler F: Temperature dependence of norma! sensory nerve action
ties within the distal segment of the median nerve. Brain 1979;102:619-635. potentials. J Neurol 1977;216:173-180.
181. Kimura J, Yanagisawa H, Yamada T. et al: Is the F wave elicited in a select 215. Ma DM, Kim SH, Spielholz N, et al: Sensory conduction study of distal radial
group of motoneurons? Muscle Nerve 1984;7:392-399. nerve. Arch Phys Med Rehabil 1981;62:562-564.
182. Kimura J, Machida M, Ishida T, et al: Relation between size of compound sen- 216. Ma DM, Liveson JA: Nerve Conduction Handbook. Philadelphia, F.A. Davis.
sory or muscle action potentials, and length of nerve segment. Neurology 1983.
1986;36:647-652. 217. Mackenzie K, DeLisa JA: Distal sensory latency measurement of the super-
183. Kimura J: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and ficial radial nerve in normal adult subjects. Arch Phys Med Rehabil 1981;
Practice. Philadelphia^ FA. Davis, 1989. 62:31-34.
184. Kincaid JC, Phillips LH. Daube JR. The evaluation of suspected ulnar neuropa- 218. MacLean 1C, Mattioni TA: Phrenic nerve conduction studies: A new technique
thy at the elbow. Arch Neurol 1986;43:44-47. and its application in quadriplegic patients. Arch Phys Med Rehabil 1981:62:
185. Kincaid JC, Brashear A, Markand ON: The influence of the reference electrode 70-73.
on CMAP configuration. Muscle Nerve 1993:16:392-396. 219. MacLean IC: Spinal nerve stimulation. In AAEM Course B: Nerve Conduction
186. Kiemm H: Das Perineurium als Diffusionsbarriere gegen-uber peroxydase bei Studies—A Review. AAEM, Rochester, MN, 1988.
epi- und endoneuraler Applikation. Z Zellforsch Mikrosk Anat 1970; 108: 220. Magladery JW, McDougal DB: Electrophysiological studies of nerve and reflex
431-445. activity in normal man. I. Identification of certain reflexes in the electromyo-
187. Kocsis JD, Waxman SG: Absence of potassium conductance in central myeli- gram and the conduction velocity of peripheral nerve fibres. Bull Johns
nated axons. Nature 1980;287:348-349. Hopkins Hosp 1950;86:265-290.
188. Kocsis JD, Malenka C, Waxman SG: Enhanced parallel fiber frequency-follow- 221. Mannerfeil L: Studies on the hand in ulnar nerve paralysis: A clinical experi-
ing after reduction of postsynaptic activity. Brain Res 1981;207:321-331. mental investigation of normal and anomalous innervation. Acta Orthoped
189. Kocsis JD, Waxman SG, Htldebrand, et al: Regenerating mammalian nerve Scand Suppl 1966;87:23-142.
fibres: Changes in action potential waveform and firing characteristics fol- 222. Marinacci AA: The problem of unusual anomalous innervation of hand mus-
lowing blockage of potassium conductance. Proc R Soc Lond 1982;217:77¬ cles: The value of electrodiagnosis in its evaluation. Bull Los Angeles Neurol
86. Soc 1964;29:133-142.
190. Kocsis JD, Waxman SG: Ionic channel organization of nomial and regenerating 223. Markand ON, Kincaid JC, Pourmand RA, et al: Electrophysiologic evaluation
mammalian axons. In Sell FJ, Herbert E, Carlson BM (eds): Progress in Brain of diaphragm by transcutaneous phrenic nerve stimulation. Neurology I984;34:
Research. New York, Elsevier, 1987, pp 89-101. 604-614.
224. Martin F: Tal om nervers allmanna egenskaper i manniskans kropp. Stockholm, 262. Pinelli P: Physical, anatomical and physiological facotrs in the latency mea-
L. Salvius, 1763. surement of the M response. Electroenceph Clin Neurophysiol 1964;17:86.
225. Maryniak O, Yaworski R, Hayes KC: Intramuscular recording in neurodiagnos- 263. Preston RJ, Waxman SG, Kodsis JD: Effects of 4-aAminopyridine on rapidly
tic studies. Am J Phys Med Rehabil 1989;68:186-191. and slowly conducting axons of rat corpus callosum. Exp Neurol 1983;79:
226. Masear VR, Meyer RD, Pichora DR: Surgical anatomy of the medial ante- 808-820.
brachial cutaneous nerve. J Hand Surg 1989;14A:267-271. 264. Pribyl R, You SB, Jantra P: Sensory nerve conduction velocity of the medial an-
227. Maupin-Szamier P, Pollard TD: Actin filament destruction by osmium tetrox- tebrachial cutaneous nerve. Electromyogr Clin Neurophysiol 1979;19:41-46.
ide. J Cell Biol 1978;77:837-852. 265. Race CM, Saldana MJ: Anatomic course of the medial cutaneous nerves of the
228. Mavor H, Libman 1: Motor nerve conduction velocity measurement as a diag- arm. J Hand Surg 1991 ;16A:48-52.
nostic tool. Neurology 1962;12:733-744. 266. Raine CS, Ghetti B, Shelanski ML: On the association between microtubules
229. May M: The Facial Nerve. New York, Thieme Inc., 1986. and mitochondria within axons. Brain Res 1971;34:389-393.
230. Mayer RF, Mawdsley C: Studies in man and cat of the significance of the H- 267. Ranck JB: Which elements are excited in electrical stimulation of mammalian
wave. J Neurol Neurosurg Psychiatry 1965;28:201-211. central nervous system? A review. Brain Res 1975;98:417-440.
231. Maynard FM, Stolov WC: Experimental error in determination of nerve con- 268. Ranvier LA: Recherches de l'histologie et la physiologie des nerfs. Arch
duction velocity. Arch Phys Med Rehabil 1972;53:362-372. Physiol 1871;4:I29.
232. McLeod JG: Electrophysiological studies in Guillain-Barre syndrome. Ann 269. Ranvier LA: Lecons sur l'Histologie du Systeme Nerveux. Paris, Savy, 1878.
Neurol 1981;9(Suppl):20-27. 270. Rexed B: Contributions to the knowledge of the postnatal development of the
233. Melvin JL, Harris DH, Johnson EW: Sensory and motor conduction velocities peripheral nervous system in man. Acta Psychiatr Scand 1944;Suppl 33.
in the ulnar and median nerves. Arch Phys Med Rehabil 1966;47:511-519. 271. Riche P: Le nerf cubital et les muscles del l'eminence thenar. Bull Mem Anat
234. Melvin JL, Schuchmann JA, Lanese RR: Diagnostic specificity of motor and Paris 1897:251-252.
sensory nerve conduction variables in the carpal tunnel syndrome. Arch Phys 272. Ricker K, Hertel G, Stodieck G: Increased voltage of the muscle action poten-
Med Rehabil 1973;54:69-74. tial of normal subjects after local cooling. J Neurol 1977;216:33-38.
235. Miller RG: The cubital tunnel syndrome: Diagnosis and precise localization. 273. Ritchie JM, Rogart RB, Strichartz G: A new method for labelling saxitoxin and
Ann Neurol 1979;6:56-59. its binding to non-myelinated fibres of the rabbit vagus, lobster walking leg,
236. Morrison RL, Cobb S, Bauer W: The effect of advancing age upon the human and garfish olfactory nerves. J Physiol 1976;261:477-494.
spinal cord. Cambridge, MA, Harvard University Press, 1959. 274. Ritchie JM, Rogart RB: Density of sodium channels in mammalian myelinated
237. Mugnaini.E, Schnapp B: The zonula occludens of the myelin sheath: Its possi- nerve fibers and nature of the axonal membrane under the myelin sheath. Proc
ble role in demyelinating conditions. Nature 1974;251:725-727. Natl Acad Sci USA 1977;74:211-215.
238. Mugnaini E, Osen KK, Schnapp B, et al: Distribution of Schwann cell cyto- 275. Robbins H: Anatomical study of the median nerve in the carpal tunnel and eti-
plasm and plasmalemmal vesicles (caveolae) in peripheral myelin sheaths: An ologies of the carpal tunnel sydrome. J Bone Joint Surg 1963:45A:953-966.
electron microscopic study with thin sections and freeze-fracturing. J 276. Robertson JD: The ultrastructure of adult vertebrate peripheral myelinated
Neurocytol 1977;6:647-668. nerve fibers in relation to myelinogenesis. J Biophys Biochem Cytol 1955;I:
239. Mulier-Mohnssen H, Tippe A, Hillenkamp F, et al: Is the rise of the action po- 271-278.
tential at the Ranvier node controlled by a paranodal organ? Naturwissenshaften 277. Robinson LR, Rubner DE, Wahl PW, et al: Influences of height and gender on
1974;61:369-370. normal nerve conduction studies. Arch Phys Med Rehabil 1993;74:1134-1137.
240. Murai Y, Sanderson I: Studies of sensory conduction. Comparison of latencies 278. Rohlich P, Weiss M: Studies on the histology and permeability of the peripheral
of orthodromic and antidromic sensory potentials. J Neurol Neurosurg nervous barrier. Acta Morph Hung 1955;5:335-347.
Psychiatry 1975;38:1187-1189. 279. Rosenbluth J: Intramembranous particle distribution at the node of Ranvier and
241. Murphy F, Kirklin JW, Finlayson Al: Anomalous innervation of the intrinsic adjacent axolemma in myelinated axons of the frog brain. J Neurocytol 1976;
muscles of the hand. Surg Gynecol Obstet 1946;83:15-23. 5:731-745.
242. Nakanishi T, Norris FH: Effect of local temperature on the resting membrane 280. Rowntree T: Anomalous innervation of the hand muscles. J Bone Joint Surg
potential in rat muscle. Electroenceph Clin Neurophysiol 1970;28:633-636. 1949:3 lB:505-510.
243. Neumcke B, Stampfli R: Sodium currents and sodium-current fluctuations in 281. Rushton WHA: A theory of the effects of fibre size in medullated nerve. J
rate myelinated nerve fibers. J Physiol 1982;329:163-184. Physiol 1951;115:101-122.
244. Neundorfer B, Seiberth R: The accessory deep peroneal nerve. J Neurol 282. Saeed MA, Gatens PF: Compound nerve action potentials of the medial and lat-
1975;209:125-129. eral plantar nerves through the tarsal tunnel. Arch Phys Med Rehabil 1982;
245. Nonner W, Rojas E, Stampfli R: Gating currents in the node of Ranvier: Volt- 63:304-307.
age and time dependence. Phil Trans R Soc 1975;270:483-492. 283. Saito A, Zacks SI: Ultrastructure of Schwann and perineural sheaths at the
246. Norris AH, Shock NW, Wagman IH: Age changes in maximum conduction ve- mouse neuromuscular junction. Anat Rec 1969;164:379-390
locity of motor fibers of human ulnar nerves. J Appl Physiol 1953;5:589-593. 284. Sandri C, Van Buren JM, Akert K: Membrane morphology of the vertebrate ner-
247. O'Sullivan DJ, Swallow M: The fibre size and content of the radial and sural vous system: A study with freeze patch technique. Progr Brain Res 1977;
nerves. J Neurol Neurosurg Psychiatry 1968;31:464-470. 46:1-384.
248. Ochoa J: The unmyelinated nerve fibre. In Landon DN (ed): The Peripheral 285. Sarala PK, NishiharaT, Oh SJ: Meralgia paresthetica: Electrophysiologic study.
Nerve. London, Chapman & Hall, 1976, pp 106-158. Arch Phys Med Rehabil 1979;60:30-31.
249. Ochs S: Axoplasmic transport-energy metabolism and mechanism. In Hubbard 286. Scharf JH, Blume R: Uber die Abhangigkeit der axonalen Mitochondrienzahl
JI (ed): The Peripheral Nervous System. New York. Plenum Press, 1974, pp von Kaliber der segmentierten Nervenfaser auf Grund einer Regressions-
47-72. analyse. J Hirnforsch 1964;6:361.
250. Oh SJ: Clinical Electromyography: Nerve Conduction Studies, 2nd ed. 287. Schlaepfer WW, Myers FK: Relationship of myelin internode elongation and
Baltimore, Williams & Wilkins, 1993. growth in the rat sural nerve. J Comp Neurol 1973;147:255-266.
251. Ohnishi A, O'Brien PC, Okazaki H, Dyck PJ: Morphometry of myelinated 288. Schmidt WJ: Doppelbrechung und feinbau der markscheide der nervenfasern. Z
fibers of fasiculus gracilus of man. J Neurol Sci 1976;27:163-172. Zellforsch Mikrosk Anat 1936;23:657.
252. Olney RK, Wilbourn AJ: Ulnar nerve conduction study of the first dorsal in- 289. Schnapp B, Mugnaini E: The myelin sheath: Electron microscopic studies with
terosseous muscle. Arch Phys Med Rehabil 1985;66:16-18. thin sections and freeze-fracture. In Santini M (ed): Golgi Centennial
253. Omer GE, Spinner M: Management of Peripheral Nerve Problems. Symposium Proceedings. New York, Raven Press, 1975, pp 209.
Philadelphia. W.B. Saunders, 198oT 290. Schnapp B, Mugnaini E: Membrane architecture of myelinated fibers as seen
254. Ozdemir C, Young RR: The results to be expected from electrical testing in the by freeze-fracture. In Waxman SG (ed): Physiology and Pathobiology of Axons.
diagnosis of myasthenia gravis. Ann N Y Acad Sci 1976;274:203-222. New York, Raven Press, 1978, p 83.
255. Paintal AS: Block of conduction in mammalian myelinated nerve fibres by low 291. Schoepfle GM, Erlanger J: The action of temperature on the excitability, spike
temperatures. J Physiol 1965:180:1-19. height and configuration, and the refractory period observed in the responses of
256. Paintal AS: Effects of temperature on conduction in single vagal and saphenous single medullated nerve fibers. Am J Physiol 1941;134:694-704.
myelinated nerve fibres of the cat. J Physiol 1965:180:20-49. 292. Schuchmann JA: Sural nerve conduction: A standardized technique. Arch Phys
257. Panizza M, Lelli S, Hallett M: H-reflex in nonhomonymous muscles in the Med Rehabil 1977;58:166-168.
human forearm. Neurology 1989;39:785-788. 293. Shanthaveerappa TR, Bourne GH: The "peripheral epithelium:'" A metaboli-
258. Panizza M, Nilsson J, Hallett M: Optimal stimulus duration for the H reflex. cally active, continuous, protoplasmic cell barrier surrounding peripheral nerve
Muscle Nerve 1989;12:576-579. fasciculi. J Anat 1962;96;527-537.
259. Pease WS, Cunningham ML, Walsh WE, et al: Determining neurapraxia in 294. Simpson JA: Fact and fallacy in measurement of conduction velocity in motor
carpal tunnel syndrome. Am J Phys Med Rehabil I988;67:117-119. nerves. J Neurol Neurosurg Psychiatry 1964;27:381-385.
260. Pease WS, Fatehi MT, Johnson EW: Monopolar needle stimulation: Safety con- 295. Singer M, Krishnan N, Fyfe DA: Penetration of ruthenium red into peripheral
siderations. Arch Phys Med Rehabil 1989;70:412^H4. nerve fibers. Anat Rec 1972;173:375-390.
261. Pease WF, Cannell CD, Johnson EW: Median to radial latency difference test in 296. Siroky MB, Sax DS, Krane RJ: Sacral signal tracing: The electrophysiology of
mild carpal tunnel syndrome. Muscle Nerve 1989;12:905-909. the bulbocavernosus reflex. J Urol 1979:122:661-664.
297. Smith DS, Jarlfors U, Cameron BF: Morphological evidence for the participa- 332. Tohgi H, Tsukagoshi H. Toyokura Y: Quantitative changes with age in normal
tion of microtubules in axonal transport. Ann N Y Acad Sci 1975;253:472-506. sural nerves. Acta Neuropath (Berl) 1977;38:213-220.
298. Smith J: Radial nerve conduction in patients with carpal tunnel sundrome. Appl 333. Trojaborg W: Motor nerve conduction velocities in normal subjects with partic-
Neurophysiol 1981;44:363-367. ular reference to the conduction in proximal and distal segments of median and
299. Smith KJ, Hall SM: Nerve conduction during peripheral demyelination and re- ulnar nerve. Electroenceph Clin Neurophysiol 1964;17:314-321.
myelination. J Neurol Sci 1980;48:201-219. 334. Trojaborg W, Sindrup EH: Motor and sensory conduction in different segments
300. Smith RS: Microtubule and neurofilament densities in amphibian spinal root of the radial nerve in normal subjects. J Neurol Neurosurg Psychiatry 1969;
nerve fibers: Relationship to axoplasmic transport. Can J Physiol Pharmacol 32:354-359.
1973;51:798-806. 335. Trojaborg W: Motor and sensory conduction in the musculocutaneous nerve. J
301. Smith RS: The short-term accumulation of axonally transported organelles in Neurol Neurosurg Psychiatry 1976;39:890-899.
the region of localized lesions of single myelinated axons. J Neurocytol 336. Trojaborg WT, Moon A, Andersen BB, et al: Sural nerve conduction parameters
1980;9:39-65. in normal subjects related to age, gender, temperature, and height: A reap-
302. Snipes J, Suter U: Molecular anatomy and genetics of myelin proteins in the pe- praisal. Muscle Nerve 1992;15:666-671.
ripheral nervous system. J Anat 1995:186:483-494. 337. Trojaborg WT, Moon A, Andersen BB, et al: Height and nerve conduction: A
303. Soderfeldt B, Olsson Y, Krisiensson K: The perineurium as a diffusion barrier reply. Muscle Nerve 1993; 16:563.
to protein tracers in human peripheral nerve. Acta Neuropath 1973;25:120- 338. Van Dijk JG, van der Kamp W. van Hilten BJ, van Someren P: Influence of
126. recording site on CMAP amplitude and on its variation over a length of nerve.
304. Soudmand R, Ward LC, Swift TR: Effect of height on nerve conduction veloc- Muscle Nerve 1994;17:1286-1292.
ity. Neurology 1982;32:407^10. 339. Van Dijk JG, Tjon-A-Tsien A, van der Kamp W: CMAP variability as a func-
305. Spacek J, Leiberman AR: Relationships between mitochondrial outer mem- tion of electrode site and size. Muscle Nerve 1995;18:68-73.
branes and agranular reticulum: Ultrastructural observations and a new inter- 340. Van Dijk JG: Motorisch geleidingsonderzoek. In van Dijk JG (ed): EMG voor
pretation. J Cell Sci 1980;46:129-147. de algemeen neuroloog. Leiden, Boerhaave Commissie, 1998, pp 14-46
306. Spencer PS, Weinberg HJ, Raine CS, et al: The perineurial window—a new 341. Van Dijk JG, van Benten I, Kramer CGS, Stegeman DF: CMAP amplitude car-
model of focal demyelination and remyelination. Brain Res 1975;96:323- tography of muscles innervated by the median, ulnar, peroneal, and tibial
329. nerves. Muscle Nerve 1999;22:378-389.
307. Spindler HA, Felsenthal G: Sensory conduction in the musculocutaneous nerve. 342. Vizoso AD, Young JZ: Internode length and fibre diameter in developing and
Arch Phys Med Rehabil 1978;59:20-23. regenerating nerves. J Anat 1948;82:110-134.
308. Srinivasan R, Rhodes J: The median-ulnar anastomosis (Martin-Gruber) in 343. Von Haller A: Elementa physiologiae corporis humani. Lausanne, Grasset,
normal and congenitally abnormal fetuses. Arch Neurol 1981;38:418-419. 1762.
309. Stampfli R: Overview of studies on the physiology of conduction in myelinated 344. Von Helmholtz H: Vorlaufiger Bericht uber die fortpflanzungsgeschwindigkeit
nerve fibers. In Waxman SG, Ritchie JM (ed): Demyelinating Diseases: Basic der Nervenreizung. Arch Anat Physiol Med I850;71.
and Clinical Electrophysiology. New York, Raven Press, 1981, pp 11-23. 345. Von Helmholtz H: Messungen Uber Fortpflanzungsgeschwindigkeit des Reizes
310. Stegeman DF, De Weerd JPC: Modeling compound action potentials of periph- in den Nerven. Arch Anat Physiol Med 1852; 199.
eral nerves in situ. II. A study of the influence of temperature. Electro- 346. Von Helmholtz H, Baxt N: Versuche uber die Fortpflanzungsgeschwindigkeit
encephalogr Clin Neurophysiol 1982;54:516-529. der reizung in den motorischen Nerven des Menschen. Berlin, Mber Konigl
311. Stephens RE, Edds KT: Microtubules: Structure, chemistry, and function. Preuss Akad Wiss, 1867, pp 228-234.
Physiol Rev 1976;56:709-777. 347. Von Helmholtz H, Baxt N: Neue Versuche uber die Fortpflanzungs-
312. Stevens JC: The electrodiagnosis of carpal tunnel syndrome. Muscle Nerve geschwindigkeit der Reizung in den motorischen Nerven des Menschen. Berlin,
1987; 10:99-113. Mber Konigl Preuss Akad Wiss, 1870, pp 184.
313. Stewert JD: Focal Peripheral Neuropathies. New York, Elsevier, 1987. 348. Wager EW, Buerger AA: A linear relationship between H-reflex latency and
314. Streib EW: Ulnar-to-median nerve anastomosis in the forearm: Electromyo- sensory conduction velocity in diabetic neuropathy. Neurology 1974;24:
graphic studies. Neurology 1979;29:1534-1537. 711-714.
315. Streib EW, Sun SF: Martin-Gruber anastomosis: Electromyographic studies. 349. Wagner AL, Buchthal F: Motor and sensory conduction in infancy and child- •
Electromyogr Clin Neurophysiol 1983;23:261-270. hood: Reappraisal. Devel Med Child Neurol 1972;14:189-216.
316. Sun SF, Streib EW: Martin-Gruber anastomosis: Electromyographic studies. 350. Wainapel SF, Kim DJ, Ebel A: Conduction studies of the saphenous nerve in
Electromyogr Clin Neurophysiol 1983;23:271-285. health subjects. Arch Phys Med Rehabil 1978;59:316-319.
317. Sunderland S: The intraneural topography of the radial, median and ulnar 351. Waxman SG, Foster RE: Ionic channel distribution and heterogeneity of the
nerves. Brain 1945;68:243-299. axon membrane in myelinated fibers. Brain Res Rev 1980;2:205-234.
318. Sunderland S: The effect of rupture of the perineurium on the contained nerve 352. Waxman SG, Kocsis JD, Eng DL: Ligature-induced injury in peripheral nerve:
fibres. Brain 1946;69:149-152. Electrophysiological observations on changes in action potential characteristics
319. Sunderland S, Bradley KC: The cross-sectional area of peripheral nerve trunks following blockade of potassium conductance. Muscle Nerve 1985:8:85-92.
devoted to nerve fibres. Brain 1949;72:428-449. 353. Waxman SG, Ritchie JM: Organization of ion channels in the myelinated nerve
320. Sunderland S, Bradley KC: Endoneurial tube shrinkage in the distal segment of fiber. Science 1985;228:1502-1507.
a severed nerve. J Comp Neurol 1950:93:411-420. 354. Wiederholt WC: Threshold and conduction velocity in isolated mixed mam-
321. Sunderland S, Marshal RD, Swaney WE: The intraneural topography of the cir- malian nerves. Neurology 1970;20:347-352.
cumflex musculocutaneous and obturator nerves. Brain 1959:82:116-132. 355. Wilbourn AJ, Lambert EH: The forearm median-to-ulnar communication: Elec-
322. Sunderland S, Bradley KC: Stress-strain phenomena in denervated peripheral trodiagnostic aspects. Neurology I976;26:368.
nerve trunks. Brain 1961;84:120-124. 356. Wiley CA, Ellisman MH: Rows of dimeric-particles within the axolemma and j
323. Sunderland S, Bradley KC: Stress-strain phenomena in human spinal nerve juxtaposed particles within glia, incorporated into a new model for the paran-
roots. Brain 1961;84:125-127. odal glial-axonal junction at the node of Ranvier. J Cell Biol 1980;84:261-280.
324. Sunderland S: Nerves and Nerve Injuries, 2d ed. Edinburgh, Churchill 357. Wiliard MB, Cowan WM, Vagelos PR: The polypeptide composition of intra-
Livingstone. 1978. axonally transported proteins: Evidence for four transport velocities. Proc Natl
325. Taborikova H, Sax DS: Conditioning of H-reflexes by a preceding subthreshold Acad Sci USA 1974;71:2183-2187.
H-reflex stimulus. Brain 1969:92:203-212. 358. Williams PL, Landon DN: Paranodal apparatus of peripheral nerve fibres of
326. Tashjian EA, Ellenberg MR. Gross N, et al: Temperature effect on antidromic mammals. Nature 1963;198:670-673.
and orthodromic sensory nerve action potential latency and amplitude. Arch 359. Williams PL, Hall SM: Chronic Wallerian degeneration—an in vivo and ultra-
Phys Med Rehabil 1987;68:549-552. structural study. J Anal 1971;109:487-503.
327. Thomas PK, Young JZ: Internodal lengths in the nerves of fishes. J Anat 360. Williams PL. Warwick R, Dyson M, Bannister LH: Gray's Anatomy.
1949;83:336-350. " Edinburgh, 1989, p 900.
328. Thomas PK: The connective tissue of peripheral nerve: An electron microscope 361. Winkler G: Le nerf peronier accessoire pro fond: Etude d'anatomie comparee.
study. J Anat 1963;97:35-44. Arch Anat (Strasbourg) 1934; 18:181.
329. Thomas PK, Ochoa J: Microscopic anatomy of peripheral nerve fibers. In Dyck 362. Wongsam PE, Johnson EW, Weinerman JD: Carpal tunnel syndrome: Use of
PJ, Thomas PK, Lambert EH (eds): Peripheral Neuropathies. 2n ed. Phila- palmar stimulation of sensory fibers. Arch Phys Med Rehabil 1983;64:16-19.
delphia, W.B. Saunders, 1985, pp 39-96. 363. Wood JG, Jean DH, Whitaker JN, et al: Immunocytochemical localization of
330. Tippe A, Muller-Mohnssen H: Further experimental evidence for the synapse the sodium, potassium-activated ATPase in knifefish brain. J Neurocytol
hypothesis of Na current activation and inactivation at the Ranvier node.
+
1977;6:571-581.
Naturwissenschaften 1975:62490-491 364. Wu PBJ, Baig M. Huang S, Kingery WS, Date E: Temperature correction fac-
331. Tjon-A-Tsien AML, Lemkes HHPJ, van der Kamp-Huyts AJC, van Dijk tors derived from normal subjects may be invalid in demyelinating neu-
JG: Large electrodes improve nerve conduction repeatability in controls as ropathies. Am J Phys Med Rehabil 1993;72:369-371.
well as in patients with diabetic neuropathy. Muscle Nerve 1996:19:689¬ 365. Wuerker RB, Kirkpatrick JB: Neuronal microtubules, neurofilaments, and mi-
695. crofilaments. Int RevCytol 1972;33:45-75.
366. Yamada MK, Spooner BS, Wessels NK: Ultrastructure and function of growth 369. Zenker W, Mayr R, Gruber H: Axoplasmic organelles: Quantitative differences
cones and axons of cultured nerve cells. J Cell Biol 1971;49:614-635. between ventral and dorsal root fibers of the rat. Experientia 1972;29:77-78.
367. Yap CB, Hirota T: Sciatic nerve motor conduction velocity study. J Neurol 370. Zenker W, Mayr R, Gruber H: Neurotubules: Different densities in peripheral
Neurosurg Psychiatry 1967;30:233-239. motor and sensory nerve fibers. Experientia 1974;31:318-320.
368. Zelena J: Bidirectional movements of mitochondria along axons of an isolated 371. Zwarts MJ, Guechev, A: The relation between conduction velocity and axonal
nerve segment. Z Zellforsch Mikrosk Anat 1968:92:186-192. length. Muscle Nerve 1995;18:1244-1249.
Chapter 6
Special N e r v e C o n d u c t i o n
Techniques
CHAPTER OUTLINE
Motor Nerve Conduction Studies Late Responses

Nerve Root Stimulation • Erb's Point (Supraclavicular) F-Wave • Physiology of F-Wave Production
Stimulation • Nerve Root Stimulation: Lumbosacral Plexus • Diagnostic F-Wave Techniques • F-Wave Clinical Utility
Conduction Latencies • Cranial Nerve Conduction Studies • H-Reflex • Physiology of t h e H-Reflex • Factors Affecting
the H-Reflex • Diagnostic H-Reflex Techniques • Peripheral
Miscellaneous Techniques Nervous System Applications • Central Nervous System
Residual Latency • Collision Technique • Refractory Period Applications
* Clinical Utility • Sensory and Motor Nerve Refractory
Periods • Refractory Periods in Muscle
Once the practitioner has mastered the basic nerve conduc- proximal nerves and assess neural conduction time across the
tion techniques, it is important to pursue more specialized plexus. Conduction times as opposed to conduction velocities are
methods of evaluating the peripheral nervous system. Ad- preferred as it is difficult to accurately measure the neural seg-
ditionally, nerves requiring needle excitation and less comment's length. Nerve root stimulation is a relatively advanced nerve
monly studied nerves are of importance. From time-to-time conduction technique and should only be attempted once the fun-
patients may present with lesions affecting specific sensory damentals of more routine procedures are mastered. Nerve roots
branches that yield small amplitude responses or require aver- can be stimulated electrically with needle electrodes and magneti-
aging techniques to better define the desired waveform. With cally with a coil over the skin. Because root stimulation with needle
the majority of nerve conduction studies described in this chap- electrodes can be done with standard apparatus, this technique is
ten the difficulty lies not in the inherent technique or nerve, but described in the following sections. It is presumed that with
more so in unfamiliarity. Most, if not all, of the techniques de- monopolar needle stimulation the root is depolarized just proximal
scribed in this chapter can be mastered with simple practice to the intervertabral foramen. 142
and repetition.
N e r v e Roots C 5 - C 6
Because the nerve roots are located under a relatively sig-
MOTOR NERVE CONDUCTION STUDIES nificant amount of muscle tissue, attempts to localize just one
nerve root in a "blind" manner is rather difficult. Addi-
A number of motor nerve conduction studies can be per- tionally, considerable expertise in addition to adjunctive fluo-
formed. By stimulating specific nerve roots or Erb's point and roscopy is required to accurately localize a particular root
recording from particular muscles, it is possible to selectively level. A more simple yet effective approach is to place the
evaluate distinct portions of the brachial or lumbosacral stimulating cathode (needle electrodes are required) just lat-
plexus. Conduction times across the hial plexus also can be eral to the spinous process (see below) so that it overlies the
assessed. posterior arch of the cervical vertebrae, thus preventing the
needle from piercing the nerve root or other vital neurovascu-
lar structures.
NERVE ROOT STIMULATION
Recording Electrodes. When stimulating the C 5 - C 6 nerve
The purpose of attempting to stimulate the nerve roots and roots, recordings are typically obtained from the biceps brachii
record CMAPs is primarily to evaluate conduction in various muscle. This muscle allows the practitioner to assess neural
225
from it. A pulse duration of 0.1 ms or more may be required to

achieve a supramaximal activation of the cervical nerve root
under study. To optimally excite the C 5 - C 6 nerve roots, a
monopolar needle is inserted perpendicular to the skin 1 or 2 cm
lateral and just inferior to the spinous process of C5 until the
posterior spinal arch is encountered (Fig. 6-1). The needle elec-
trode is then withdrawn several millimeters to ensure a volume-
conducted spread of the depolarizing stimulus. A needle
electrode 50 mm in length is recommended because the depth
of needle insertion is usually between 25^40 mm. It is impor-
tant to maintain the needle perpendicular to the skin surface to
avoid directly encountering sensitive neurovascular or lung
structures.
A similar needle electrode has been recommended to be in-
serted contralateral to the side of stimulation and serve as the
anode. Using a rather strong current intensity may activate
144
both left and right nerve roots, simultaneously allowing one

to record from both sides should a two-channel instrument be
available. When the contralateral side is examined, then the
cathode and anode are reversed. A surface anode also can be
positioned several centimeters distal to the needle insertion
site should a recording obtained from one side at a time be
desired.
Instrumentation Parameters. Specific instrumentation set-
tings were not provided; however, similar latencies to those
originally obtained should be approximated when routine set-
Figure 6-1. C5/C6 nerve root stimulation. Needle placement tings are used. A sweep speed of 2 ms/div and sensitivity ca-
144
for excitation of the C5/C6 spinal nerves used in assessing the upper pable of displaying the entire response on the screen are
trunk and lateral cord of the brachial plexus. (From MacLean IC: Spinal sufficient to obtain the desired responses. Also, low- and high-
nerve stimulation. In AAEM Course B: N e r v e Conduction Studies—A frequency filters approximating 10 Hz and greater than or equal
review c o u r s e . Rochester, MN, American Association of Electro- to 8 kHz, respectively, are used. ;
diagnostic Medicine, 1988, with permission.) Reference Values. The anticipated latency to the biceps
brachii muscle from the C5-C6 region is between 4.5 to 6.6 ms
impulses originating in C5-C6 nerve roots traversing the upper with a mean of 5.3 ± 0.4 m s . An expected left/right difference •
144
trunk, lateral cord, and musculocutaneous nerve. 96

less than 0.6 ms is anticipated (Table 6-1).
E-L A surface E-l electrode is positioned over the mid-point
of the biceps brachii in an attempt to record from the muscle's N e r v e Roots C 6 - C 8
motor point, thus resulting in an initial negative deflection. As previously noted, exact localization of specific nerve roots
Standard concentric needle electrodes may also be used; how- is difficult because of the overlying muscular tissue and
ever, a more localized recording ensues, limiting the value of volume-conducted spread of the depolarizing current. The pos-
the CMAP's amplitude. The latency with needle electrodes lo- terior divisions and posterior cord of the brachial plexus can be
cated deep within the muscle are as valid as those recorded with evaluated by recording from the triceps muscle following
surface electrodes. 102
C6-C8 nerve root excitation.
96
E-2. The E-2 recording electrodes is usually a surface elec- Recording Electrodes. Because the motor point of the tri-
trode positioned on the tendinous insertion of the biceps ceps muscle is rather difficult to locate, one may wish to con-
brachii. sider using an intramuscular needle recording electrode. If a 1
Stimulation. Stimulation for all nerve root studies is per- needle electrode is chosen, it should be a standard concentric or
formed with a monopolar needle serving as the cathode posi- monopolar needle and inserted deeply into the main bulk of the
tioned on the posterior arch of the cervical vertebra. In the case triceps muscle. Although surface electrodes are capable of
of cervical nerve root excitation, the needle electrode is not po- recording a response, an initial negative deflection may be diffi-
sitioned directly next to neural tissue but somewhat removed cult to reproduce in all patients.
E-L A standard concentric needle electrode is positioned
within the depth of the main bulk of the triceps muscle on the
Table 6-1. Cervical Nerve Root Stimulation posterior or posterolateral aspect of the arm. This allows one
Stimulation Recording Latency (ms) L/R (ms) to obtain a clearly recognizable deflection from the baseline
whether in the positive or negative direction and should be
C5/C6 Biceps brachii 5.3 ± 0.4 (4.5-6.6) 0.0-0.6
noted for determining the onset latency. A recording from
C6/C7/C8 Triceps brachii 5.4 ± 0 . 4 (4.4-6.1) 0.0-0.6 the triceps muscle permits the practitioner to assess the
C8/TI Abductor digiti 4.7 ± 0.5 (3.7-5.5) 0.0-0.7 C6-C8 neural fibers trasversing the brachial plexus' posterior
minimi divisions and posterior cord. A surface electrode may be
used; however, onset latency determination may be some-
As amplitude is not considered, one may use needle recordings to assess onset
latency.The time to the abductor digiti minimi represents the transbrachial plexus what difficult because of less than distinct deflections from
latency as calculated by subtracting the axillary latency from the C8/TI latency.
144 the baseline.
Chapter 6 SPECIAL NERVE C O N D U C T I O N TECHNIQUES — 2 2 7
E-2. If a standard concentric needle electrode is used, the E-

2 electrode is the surrounding cannula. In monopolar needle
recordings, a surface E-2 should be located on the olecranon.
Stimulation. Again, a monopolar needle cathode electrode is
inserted perpendicular to the skin surface 1-2 cm lateral and just
inferior to the spinous process of C6 and positioned a few millime-
ters superior to the posterior arch of C6 (Fig. 6-1). A supramaxi-
mal stimulation is delivered by optimizing the CMAP recorded
from the triceps muscle. An anode can be placed in a similar posi-
tion contralaterally or ipsilaterally as previously described.
Instrumentation Parameters. See C5-C6 nerve root stimu-
lation.
Reference Values. A triceps brachii latency between 4.4 to
6.1 ms with a mean of 5.4 ± 0.4 ms is expected in normal indi-
viduals (Table 6-1). Additionally, a right-to-left difference of
less than 0.6 ms is expected. 144
N e r v e Roots C 8 - T I
Perhaps one of the more commonly performed nerve root
stimulation procedures involves excitation of the C8-T1 nerve
roots. This may be a result of the regional diagnostic popularity
of C8-T1 root, lower trunk, or medial cord compression sec-
ondary to possible anatomic compromise of these structures,
i.e., the thoracic outlet syndrome. Although not discussed in
detail at this time, evaluation of C8-T1 proximal nerve fiber
Figure 6-2. C8/TI nerve root stimulation. Needle electrode loca-
conduction is one objective electrophysiologic way in which to
tion for C8/TI nerve stimulation for lower trunk and medial cord evalu-
evaluate possible neural compromise in a patient suspected of
ation. (From MacLean IC: Spinal nerve stimulation. In AAEM Course B:
having the thoracic outlet syndrome.
Nerve Conduction Studies—A review course. Rochester, MN.American
Recording Electrodes. Locating the recording electrodes on Association of Electrodiagnostic Medicine, 1988, with permission.)
the hand intrinsic muscles, either median- or ulnar-innervated
muscles, allows one to assess C8-T1 neural fibers traversing the
lower trunk and medial cord of the brachial plexus. Because of 6-1). Left-to-right conduction time differences range from 0.0
the long conduction route, a second proximal stimulation site to 0.7 ms.
(see below) is necessary to preferentially consider this segment
of the C8-T1 fiber course. ERB'S POINT (SUPRACLAVICULAR) STIMULATION
E-l. A surface electrode is recommended to be positioned
over the motor point of the abductor digiti minimi muscle. It is A number of proximal nerves are not amenable to direct
certainly acceptable to use a standard concentric needle electrode neural excitation because of the surrounding musculoskeletal
as long as quantitative amplitude measurements are not desired. structures. An indirect means is required to assess their integrity
E-2. The E-2 electrode in a surface recording is placed just
distal to the insertion of the muscle (see ulnar nerve conduc-
tion). If a standard concentric needle is used, the cannula serves
as the E-2 recording electrode.
Stimulation. A 50-mm monopolar needle electrode is in-
serted perpendicular to the skin surface approximately 1 cm J
distal and lateral to the spinous process of C7 until the posterior
bony arch is contacted (Fig. 6-2). The needle cathode is then
25cm
withdrawn several millimeters. Again, a contralateral needle i
anode is possible or an ipsilateral surface anode located several i
centimeters distal to the needle insertion site. The onset latencies 1
for left and right abductor digiti minimi CMAPs are recorded. Stimulation of Median \
A second stimulus is then applied at the axilla on a line 25 cm and Ulnar Nerves
in length from the mid-sternal notch with the arm abducted 90°
and externally rotated (Fig. 6-3). This procedure is repeatedJor
the contralateral limb. The onset latencies to the left and right Figure 6-3. Axilla stimulation. Location for arm stimulation of
abductor digiti minimi muscles are recorded to the CMAP's ini- the median and ulnar nerves used in conjunction with C8/TI n e r v e
tial departure from baseline. Onset latencies from axillary stim- root stimulation t o assess conduction time across the brachial plexus.
ulation are subtracted from the nerve root excitation latencies to The arm is externally rotated and abducted. Stimulation of the median
arrive at a transbrachial plexus conduction time. and ulnar nerves is performed 25 cm from t h e sternal notch over t h e
Instrumentation Parameters. See C 5 - C 6 nerve root stimu- neurovascular bundle in t h e arm. (From MacLean IC: Spinal nerve stim-
lation. ulation. In AAEM C o u r s e B: N e r v e Conduction Studies—A review
Reference Values. The range of conduction times across the c o u r s e . Rochester, MN, American Association of Electrodiagnostic
brachial plexus is 3.7-5.5 ms with a mean of 4.7 ± 0.5 ms (Table Medicine, 1988, with permission.)
with respect to potential pathology. Specifically, techniques fingers. A standard concentric needle electrode is then carefully
have been developed to examine a number of these proximal inserted between the two fingers until the rib is encountered and
nerves, including long thoracic, suprascapular, axillary, muscu- then withdrawn just a few millimeters. The response is mea-
locutaneous, and proximal radial nerves. Each of these nerves sured to the initial deflection from the baseline.
innervates a muscle that can easily be used to record from; how- E-2. In a surface recording, an E-2 is positioned on the same
ever, the individual nerves are not readily accessible. As a result, rib as E-1 several centimeters anteriorly. Of course, if a standard
it becomes necessary to deliver a rather strong depolarizing concentric needle is used, the cannula serves as E-2.
pulse to the brachial plexus as a whole in an attempt to activate Stimulation. Stimulation is carried out at Erb's point with
in a supramaximal manner all of the above-noted nerves. This an intensity and pulse duration capable of delivering a supra-
can be accomplished if the excitation pulse is delivered at Erb's maximal excitation. A pulse duration between 0.5 and 1.0 ms is
point. usually sufficient to achieve the desired result. Initially, the pa-
Recording Electrodes. Because of the nature of the stimu- tient can be requested to turn the head opposite to the side of
lus producing depolarization of the brachial plexus as a whole, stimulation, making accurate identification of Erb's point rela-
one can record from multiple muscles simultaneously provided tively easy. Once Erb's point is localized, the patient's head can
one's instrument possesses more than one channel. This is ad- be turned slightly toward the stimulus site just past midline.
vantageous because it limits the number of rather intense shocks This is necessary to relax the skin and subcutaneous tissues al-
delivered to the patient. Also, the original studies described lowing the practitioner to position the cathode (distal) and
below used intramuscular recordings with standard concentric anode as deeply into the supraclavicular fossa as possible with-
needle electrodes, thus limiting the utility of amplitudes. out producing undue patient discomfort. It is anticipated that the
entire plexus will be activated, but an attempt should be made to
Long Thoracic N e r v e confirm contraction of the serratus anterior by palpating the
The long thoracic nerve arises directly from cervi'cal nerve muscle over the anterolater aspect of the rib cage.
roots C 5 - C 7 and innervates the serratus anterior muscle. - 95 96
Instrumentation Parameters. See nerve root stimulation.
This muscle is of clinical value because its nerve originates Reference Values. Onset latencies to the initial deflection of
proximal to the formation of the brachial plexus and may help the response are in the range of 2.6-4.0 ms (Table 6-2). - 29 111
to assess the extent of a possible brachial plexus injury. Should

this nerve be spared, it suggests that the injury site is distal to Suprascapular N e r v e
the C5-C7 root level. The suprascapular nerve originates just distal to the formation
Recording Electrodes. Surface recording electrodes are of the brachial plexus' upper trunk and is composed of nerve
typically used for this study; however, there is a report of fibers from C 5 - C 6 . This nerve then proceeds posteriorly
96
monopolar needle recordings from this m u s c l e . The use of 29111

through the suprascapular notch to innervate the supraspinatus
surface recording electrodes is understandable given the prox- muscle. The nerve continues to course around the lateral aspect
imity of this muscle to the chest wall and the potential for inad- of the scapula to innervate the infraspinatus muscle. We describe
vertently piercing the intercostal space and possibly producing a latencies to both the supraspinatus and infraspinatus muscles.
pneumothorax. Recording Electrodes. The majority of recordings from
E-l. An E-l surface recording electrode is positioned over proximal muscles following Erb's point stimulation are per-
the fifth or sixth rib at the midaxillary line. In persons with sig- formed with standard concentric needle electrodes, although it
nificant subcutaneous tissue, an intramuscular E-l may be is possible to also use monopolar needles.
preferable for recording the response's latency. If a standard Supraspinatus Muscle. E-l. A standard concentric needle
concentric needle is chosen, it is imperative to use proper tech- electrode is located in the supraspinatus muscle midway be-
nique to avoid the possibility of inducing a pneumothorax. The tween the medial border of the scapula and the acromion. 72
interspaces flanking the desired rib must be identified by palpat- Using obstetric calipers, two separate distances were measured
ing the interspaces and locating the rib between the palpating from the center of the cathode/anode at Erb's point to the
recording locations for the supraspinatus muscle. The two mean
distances were 8.5 cm and 10.5 cm for respective E-l locations.
T a b l e 6-2. Erb's P o i n t Stimulation
This allows the practitioner to choose one of the distances most j
Nerve Recording Latency (ms) appropriate to the size of the patient. Surface electrodes are of
Long t h o r a c i c 291
" Serratus anterior 2.6-4.0 questionable value for this muscle because of the overlying
trapezius muscle that may diminish the CMAP's magnitude.
Suprascapular 72
Supraspinatus (8—9 cm) 2.6 ± 0.07
E-2. When using standard concentric needle electrodes, the
Supraspinatus (10—11 cm) 2.7 ± 0.07
cannula serves as the E-2 recording electrode. If a monopolar
Suprascapular 72
Supraspinatus (7.4-12 cm) 2.7 ± 0.5 needle is used as E-2. it should be located several centimeters
Suprascapular 72
Infraspinatus (13-15 cm) 3.4 ± 0.09 lateral to E - l .
143
Infraspinatus (16-18 cm) 3.4 ± 0 . 1 3 Stimulation. A surface cathode and anode are located at
Suprascapular 128
Infraspinatus (10.6—15 cm) 3.3 ± 0.5 Erb's point. The pulse duration should be between 0.5 and 1.0
ms with an intensity capable of producing a supramaximal re-
Musculocutaneous - 128 123
Biceps brachii (19-21 cm) 4.6 ± 0.14
sponse. Firm pressure applied at Erb's point is often necessary
Biceps brachii (23—25 cm) 4.7 ± 0 . 1 5
to evoke an optimal response. See long thoracic nerve for de-
Biceps brachii (27-29 cm) 5.0 ± 0 . 1 3
tails of head positioning.
Axillary 72128
Deltoid (15-16 cm) 4.3 ±0.11 Instrumentation Parameters. See nerve root stimulation.
Deltoid (18-19 cm) 4.4 ± 0.08 Reference Values. The onset latencies to the initial deflec-
Recording performed with standard concentric needle electrodes and ampli- tion of the response is recorded for either distance chosen (Table
tudes not recorded. 6-2).
Infraspinatus Muscle. E-l. A standard concentric needle recorded with a concentric needle electrode can range between
electrode is inserted into the mid-portion of the infraspinatus 6 and 32 mV.
muscle 14 cm and 17 cm from the stimulation site at Erb's
point. This distance is measured with obstetric calipers. Axillary Nerve
E-2. See supraspinatus muscle. The axillary nerve, also called the circumflex nerve, is
Stimulation. See supraspinatus muscle. formed by nerve roots C 5 - C 6 . This nerve arises from the pos-
96
Instrumentation Parameters. See nerve root stimulation. terior cord of the brachial plexus. There are two muscles inner-
Reference Values. As for the supraspinatus, the initial onset vated by the axillary nerve: teres minor and deltoid. T h e
latency is of interest (Table 6-2). relevant anatomy of the axillary nerve is that it courses through
the quadrilateral (quadrangular) space, i.e., teres minor supe-
Musculocutaneous N e r v e riorly, teres major inferiorly, surgical neck of the humerus later-
The musculocutaneous nerve is the continuation of the ally, and long head of the triceps muscle medially. This nerve
brachial plexus' lateral cord and consists of fibers from nerve then travels posterolaterally around the humerus to divide into
roots C 5 - C 7 . This nerve innervates the coracobrachialis,
95
anterior and posterior neural branches to innervate the deltoid
biceps brachii, and brachialis muscles. It is possible to injure muscle. Dislocations or fractures of the humerus may injure the
this nerve as a result of shoulder dislocations. axillary nerve.
Recording Electrodes. Because of a relatively well defined Recording Electrodes. The accessibility of the deltoid
motor point along a band in the middle of the biceps brachii muscle permits surface-recording electrodes to be u s e d . 128
muscle, surface electrodes can be used to obtain a CMAP. It 128

Standard concentric needle electrodes have also been used to
is also possible, however, to use standard concentric needle record onset latencies. 72
electrodes (Table 6-2). E-l. If a surface E-l electrode is chosen, it should be secured
E-l. When recording with surface electrodes E-l is located to the most prominent portion of the deltoid muscle in the upper
at the mid-point of the biceps brachii muscle. A standard con- lateral aspect of the arm. The mid-portion of the muscle con-
centric needle electrodes is placed in the muscle at 20 cm, 24 tains the motor point and a recording from this region should
cm, or 28 cm depending upon the patient's stature. As for the result in a well-defined negative onset. The use of a standard
suprascapular nerve, this distance is measured with obstetric concentric needle electrode requires the needle to be placed
calipers from the Erb's point stimulation site. deep in the substance of the muscle. As for needle recordings
E-2. In the surface recording technique, E-2 is located on the from other proximal muscles, there are several distances mea-
biceps brachii tendon in the antecubital fossa. As previously sured with obstetric calipers from the point of stimulation to ac-
noted, the cannula of the standard concentric needle electrode count for different arm lengths. The distances for E-l placement
serves as E-2. are 15.5 cm and 18.5 cm. 72
Stimulation. There are two stimulation techniques for elicit- E-2. For a surface recording, E-2 is located at the tendinous
ing a CMAP from the biceps brachii muscle. It is possible to insertion of the deltoid muscle in the mid-arm area. As previ-
either stimulate Erb's point or directly excite the musculocu-
128
ously noted, the cannula is the E-2 electrode for concentric
| taneous nerve in the anterior aspect of the axilla. The latter
223
needle recordings.
stimulation site more selectively activates the musculocuta- Stimulation. See long thoracic and suprascapular nerves.
neous nerve as opposed to activating the entire brachial plexus Instrumentation Parameters. See suprascapular nerve.
and may be of use during repetitive stimulation studies obviat- Reference Values. Onset latencies are similar for both stan-
ing the need to excite the entire brachial plexus. Stimulation at dard concentric needle and surface recordings (Table 6-2). It is
Erb's point is performed as previously noted for the long tho- important to recall that should a needle recording be used, the
racic and suprascapular nerves. needle electrode is placed deep into the substance of the muscle
When attempting to directly excite the musculocutaneous to avoid erroneously long latencies. Only surface recordings
102
nerve in the arm, one can use either surface or needle stimula- are optimal for comparing side-to-side amplitudes.
tion. For surface stimulation, the cathode is positioned close to
the insertion of the pectoralis major muscle on the humerus N e r v e Root Stimulation: Lumbosacral
distal and somewhat posterior to its inferior margin, whereas the Plexus Conduction Latencies
anode is located proximally. A needle cathode is located be- It is possible to evaluate conduction across the lumbosacral
tween the coracobrachialis tendon laterally and the axillary plexus by stimulating the nerve roots constituting the plexus and
artery medially just proximal to the latissimus dorsi tendon. subtracting the time of conduction from either the femoral or sci-
Similar parameters for needle stimulation previously noted are atic nerves. The lumbar plexus is assessed by simultaneously ex-
again used. The needle anode is located transversely at a dis- citing roots L2-L4 and recording a response from the vastus
tance of 3 cm. medialis muscle. The femoral nerve is depolarized in the in-
96
Instrumentation Parameters. See suprascapular nerve. guinal region. The femoral nerve latency is then subtracted from
Reference Values. When using a surface recording elec- the root latency and a conduction time across the lumbar plexus
trode the latency is measured to the initial deflection of the re- results. For sacral plexus analysis, a CMAP from the AH is ob-
sponse (Table 6-2). Unlike needle electrodes, the amplitude of tained following L5-S1 nerve root activation. The sciatic nerve is
the surface-recorded C M A P best reflects the summated re- then stimulated at the gluteal fold. This latency is subtracted from
sponse of the muscle and may be used for diagnostic purposes. the L5-S1 latency for a trans-sacral plexus conduction time.
Amplitudes obtained with needle recording should be used with Recording Electrodes. Surface recording electrodes for the
caution regarding any attempt to quantify axonal loss. femoral and sciatic nerves are used to calculate lumbosacral
If the musculocutaneous nerve is directly excited in the conduction times.
axilla, one can anticipate onset latencies in the range of 1.3-3.6 Lumbar Plexus (L2-L4). E-l-E-2. See femoral nerve
ms for recording distances of 7-13 cm. Peak-to-peak amplitudes
223
(Chapter 5).
Conduction Across Sacral Plexus (L5-S1). E-l-E-2. See sciatic nerve (Chap-
Lumbar Plexus ter 5).
Stimulation (L2-L4). A monopolar needle electrode 75 mm
in length is used for the cathode. Approximately 2-2.5 cm lat-
eral to the spinous process of the L4 vertebral body, a needle
cathode is inserted perpendicularly to the skin in a sagittal plane
(Fig. 6-4). The needle is positioned on the periosteum of the
142
vertebral arch overlying the L4 nerve root. The anode, a similar

needle electrode to the cathode, is located in the same position
on the contralateral aspect of the body. Stimulation as described
above allows one to activate the lumbar nerve roots bilaterally.
It is important to rest the tip of the cathode and anode on the
posterior bony aspect of the vertebral arch and not the inferior
or superior interspaces. Sufficient current is delivered by adjust-
ing both the intensity and pulse duration to achieve a supramax-
imal response. The patient should be sufficiently warned as this
can be uncomfortable.
Stimulation (L5-S1). The same needle cathode and anode
used for lumbar stimulation are also used for excitation of
L5-S1 nerve roots (Fig. 6-5). In this instance, however, the
144
cathode/anode are inserted perpendicular to the skin surface just

medial and a bit caudal to the posterior superior iliac spine.
Similar comments noted above for L2-L4 nerve root excitation
also apply to activating L5-S1 nerve roots.
Instrumentation Parameters. See femoral and sciatic
nerve conduction study instrumentation recommendations
(Chapter 5).
Reference Values. Calculated means, ranges, and left/right
differences are provided for lumbosacral nerve root stimulation
Figure 6-4. L2/L3/L4 nerve root stimulation. Needle electrode (Table 6-3).
placement for activation of the L2/L3/L4 nerve roots. Additionally,
stimulation of the femoral nerve is depicted for the determination of CRANIAL NERVE CONDUCTION STUDIES
transplexus conduction times. (From MacLean IC: Spinal nerve stimu-
lation. In AAEM C o u r s e B: Nerve C o n d u c t i o n Studies—A review Three of the cranial nerves can be readily studied with rou-
course. Rochester, MN, American Association of Electrodiagnostic tine nerve conduction studies previously described for upper
Medicine, 1988, with permission.) and lower limb peripheral nerves. The cranial nerves discussed
in this text are: cranial nerve VII (facial nerve), cranial nerve V
(trigeminal nerve, afferent component only), and cranial nerve
XI (spinal accessory nerve). The techniques discussed are per-
formed with surface stimulation and recordings and of proven
value in the authors' experience.
Cranial N e r v e VII (Facial N e r v e )

The seventh cranial nerve's nucleus is located within the cen-
tral nervous system in the pons. This nerve provides motor in-
132
nervation to the muscles of facial expression, i.e., all facial j

muscles except those innervated by the trigeminal nerve (mas-
seter, temporalis, and pterygoid muscles). Additional neural
components mediated by the facial nerve include taste sensation
to the anterior two-thirds of the tongue (chorda tympani nerve),
sensation to a portion of the external ear and soft palate, and
Table 6-3. Lumbosacral Nerve Root Stimulation 142
Stimulation Recording Latency (ms) L/R (ms)

L2/L3/L4 Vastus medialis 3.4 ± 0.6 (2.0-4.4) 0.0-0.9
(femoral nerve)
Figure 6-5. L5/SI nerve root stimulation. L5/SI nerve stimula- L5/SI Abductor hallucis 3.9 ± 0.7 (2.5-4.9) 0.0-1.0
tion is shown along with sciatic nerve activation in order to determine (sciatic nerve)
the transplexus conduction times for the L5 and SI nerve root fibers.
T h e a b o v e - n o t e d t i m e s r e p r e s e n t t h e latency across t h e lumbosacral plexus
(From MacLean IC: Spinal nerve stimulation. In AAEM Course B: Nerve
w i t h f e m o r a l and sciatic n e r v e latencies s u b t r a c t e d f r o m t h e a b s o l u t e n e r v e
Conduction Studies—A review c o u r s e . Rochester, MN, American r o o t latencies. As a m p l i t u d e is n o t c o n s i d e r e d , o n e may use needle recordings
Association of Electrodiagnostic Medicine, 1988, with permission.) t o assess onset latency.
Chapter 6 SPECIAL NERVE C O N D U C T I O N TECHNIQUES — 231
finally the parasympathetic supply to the lacrimal and salivary thereby avoiding coexcitation of the masseter muscle. Of 152
glands. The anatomic course of the facial nerve can be separated course, the latency is significantly shortened in this case, but the
into an intracranial and extracranial portion. Intracranially, the response is acceptable for side-to-side amplitude comparisons.
seventh nerve arises from the pons to enter the facial canal via E-l. The E-l surface-recording electrode can essentially be
the internal auditory meatus. The facial canal consists of the placed on any facial muscle desired. Three commonly examined
labyrinthine, tympanic, and mastoid segments of which the muscles are the orbicularis oculi, orbicularis oris, and nasalis
labyrinthine is the smallest. - The termination of the mastoid
55 56
(Fig. 6-6). Should the orbicularis oculi be chosen for recordings,
segment, stylomastoid foramen, is where the facial nerve exits the E-l electrode is usually positioned inferior to the eye's
the skull to begin its extracranial course. After exiting the skull, lower canthus aligned with the pupil or at some point laterally
the nerve enters the substance of the parotid gland and divides to the outer margin of the eye. Some repositioning of the elec-
into a number of divisions to innervate various muscles of facial trode may be required to achieve an initial negative deflection.
expression. These muscles are relatively easy to evaluate with For orbicularis oris recordings, E-l is located at the angle of the
nerve conduction techniques because of their superficial loca- mouth just lateral to where the upper and lower lips join. The
tion. Also, the facial nerve can be readily excited anterior to the nasalis muscle area is perhaps the easiest region to record from
earlobe. when exciting the facial nerve (Fig. 6-6). It is located by having
Recording Electrodes. As previously noted, only surface the patient "crinkle" the nose as if a foul scent has been encoun-
recordings are described as this method provides the best as- tered. The prominent bulge just superior to the lateral nasal ala
sessment of the total number of muscle fibers excited. is the nasalis muscle area. The paretic side should be compared
Essentially any muscle can be used to record a CMAP follow- with the normal side in order to properly position the electrode.
ing facial nerve activation. This gives the opportunity to selec- Recording from the nasalis muscles usually result in the best
tively measure the different branches of the facial nerve (e.g., CMAPs. 186
zygomatic, mandibular etc.). Facial muscles do not necessarily If amplitude is not of interest when performing facial nerve
have well-defined motor points and subsequently may yield recordings, it is acceptable to use standard concentric needles
CMAPs with an initial positive deflection. One can attempt to placed into the muscle under investigation. Relatively sharp
reposition the electrodes, but this may not always result in a onsets of either a positive or negative direction should be ob-
waveform with an initial negative onset. When this occurs, one tained. It is important to remember, however, that the amplitude
is advised to accept the response and measure the onset latency
to the beginning of the initial positive deflection. When calcu-
lating the amplitude of any CMAP, it is better to measure the
potential from the initial negative deflection to the peak of the
negative spike. If it is impossible to obtain an initial negative
deflection, an initial positive to subsequent negative peak suf-
fices. The major value in facial nerve studies with respect to
prognosis is comparing side-to-side amplitudes. - Hopefully,
55 56
both sides of the face have similar-appearing potentials for com-

parison purposes. There may be occasions when one side of the
face has a pronounced positive deflection, whereas the con-
tralateral side begins with the expected negative deflection. This
poses a significant problem for comparative evaluations. If
repositioning the E-1 electrode with the positive deflection does
not resolve the problem, one cannot use two morphologically
different CMAPs for comparative purposes. All factors being
equal (recording electrode position, stimulus location, current
pulse width and intensity, and manual pressure on all elec-
trodes), a marked side-to-side amplitude discrepancy of greater
than 5 0 % is suspicious. This is a conservative estimate as
normal side-to-side variations may reach approximately
o
3-20%.55.56.103.104 ne y
ma w \sh p c e e d to a different muscle
t 0 r0
in the hope of finding relatively symmetric CMAPs for left and

right sides of the face.
A second problem in facial nerve studies is a volume-con-
ducted response from the masseter. When stimulating the facial
nerve anterior to the earlobe it is relatively easy to directly acti-
vate the masseter muscle. In patients with profound facial nerve
loss, a volume-conducted masseter CMAP can coincide with
the expected facial nerve response's position and be mistaken
for a facial CMAP. The practitioner must be aware of this poten- Figure 6-6. Facial n e r v e a c t i v a t i o n . T h e facial nerve is stimulated
tial problem to avoid an erroneous conclusion that there is facial either anterior or posterior t o the ear (S) with subsequent recording
nerve function when indeed this nerve may have undergone from any facial muscle. In the above diagram a recording from the left
complete degeneration. Should this be encountered, it behooves nasalis (E-hRJ is depicted with E-2 (R ) on the superior aspect of t h e
r
the practitioner to palpate the masseter muscle for a contraction nose away from muscle tissue. W e believe the posterior stimulation is
when stimulating the facial nerve. There is also a recommenda- preferable. (From Ma DM, Liveson JA: Nerve Conduction Handbook.
tion to excite the facial nerve as it passes beneath the zygoma, Philadelphia, F. A. Davis, 1983, with permission.)
obtained with needle recordings is not valid to be used for as- side-to-side amplitudes within the first 2 weeks following a
sessing axonal loss with respect to prognosis. lesion such as Bell's palsy, sparing of 10% or more of the re-
E-2. A surface E-2 is usually located in an area devoid of sponse compared to the uninvolved side suggests a good prog-
muscle if at all possible. The most likely location on the face is nosis for recovery. - Normal threshold stimulation currents are
53 56
on the tip or bridge of the nose as it is mostly cartilage or bone between 3.0-8.0 mA with a side-to-side difference less than 2.0
(Fig. 6-6). Although this location is not "electrically silent," it is mA. 125
a convenient location to assist in differential amplification and

common mode rejection. Of course, should a standard concen- Cranial N e r v e V (Trigeminal N e r v e )
tric needle be used, the cannula serves as E-2. That aspect of the trigeminal nerve capable of being exam-
Ground. As with other nerve conduction techniques, the ined with peripheral nerve stimulation involves primarily the
ground electrode should be located close to E-l between it and sensory afferent fibers originating in the supraorbital nerve.
the cathode. This nerve can be located by palpating the supraorbital notch
Stimulation. Surface stimulation can be applied to one of along the medial aspect of the supraorbital ridge. Afferent im-
two convenient locations. A cathode may be placed either ante- pulses arising from the cutaneous distribution of this nerve,
rior or posterior to the earlobe (Fig. 6-6). Anteriorly, the cathode vertex of skull to supraorbital area, travel to their cell body lo-
is pressed into the substance of the parotid gland several cen- cated in the trigeminal ganglion. From this ganglionic region,
90
timeters superior to the angle of the mandible. Slight the action potentials travel into the pontine portion of the central
superior/inferior movement may be required to optimally locate nervous system and apparently diverge into two separate path-
the facial nerve. Postauricular activation of the facial nerve is ways. An oligosynaptic path proceeds superiorly to synapse in
accomplished by positioning the cathode posterior to the neck the principle sensory nucleus (Fig. 6-7). A second-order path-
of the mandible inferior to the mastoid process. Again, it is im- way then travels caudally to synapse with the facial nerve nu-
portant to avoid direct masseter activation . " - Despite recom-
42 44 74
cleus causing depolarization of this structure with an ensuing
mendations in the literature, we strongly believe that all facial contraction of the orbicularis oculi muscle ipsilateral to the side
nerve stimulations should occur in proximity to the stylomas- of excitation. A second pathway from the point of divergence in
toid foramen, i.e., behind the ear. the rostral pons courses caudally in the lateral medullary plate
Because there are several parameters one can measure fol- region a variable distance (Fig. 6-7). At some point in the lower
lowing facial nerve stimulation, the characteristics of the stimu- medulla and several synapses later, two separate tracts head su-
lator must be specified. If facial nerve latency or amplitude is of periorly, both ipsilateral and contralateral to the side of stimula-
primary interest, then either a constant-current or constant-volt- tion. These two pathways eventually synapse with their
age stimulator with sufficient current intensity capable of deliv- respective facial nerve nuclei and induce a contraction of both
ering a supramaximal response is all that is required. On the orbicularis oculi muscles, which is the clinically observed blink.
other hand, should one wish to measure the amount of current The above described and presumed pathway describes the elec-
necessary to evoke just a minimal facial muscle contraction, a trically induced "blink reflex." Apparently, a relatively strong
125
constant-current stimulator with a pulse width of 0.2-0.5 ms is depolarization of this nerve is required to generate a blink reflex •
necessary. It is very easy to stimulate the facial nerve intracra- as cutaneous stimulation to the distribution of the supraorbital
nially with a magnetic stimulator. In contradistinction with nerve does not produce the clinically observed blink response.
magnetic root stimulation, it is possible to obtain maximal The above-described pathway is believed to be slightly different
CMAPs, even with low stimulus strength. The magnetic coil is than that taken by impulses generated with tactile stimulation of
positioned over the parietal region. The facial nerve is depolar- the cornea, i.e., the clinical blink reflex.
ized just at the proximal part of the facial c a n a l . Attempting
186199
The electrical blink reflex examines the afferent trigeminal
to define the minimal amount of current that just produces a tract through the supraorbital nerve and the efferent facial
minimal twitch of a facial muscle is known as the nerve ex- nerve pathway to the orbicularis oculi muscle. It is possible to
citability test (NET). elicit a blink reflex with excitation of the infraorbital and
To perform a NET study, the patient is comfortably posi- mental nerves but with significantly less consistency than the
tioned with a bright light directed across the side of the face so supraorbital nerve. Facial muscles other then the orbicularis ,
that sharp shadows are cast by the facial structures to aid in vi- oculi do not typically yield a consistent blink response. |
sualizing muscle contraction. The current intensity is slowly in- Because of the time resolution of the electrodiagnostic equip-
creased until a minimal twitch of a facial muscle is observed. ment, it is possible to resolve both the early ipsilateral response
The current is recorded and compared with a similar procedure (RI) and the later bilateral response (R2) (Fig. 6-7). By assess-
for the unaffected side. The muscles usually observed for this ing the presence, absence, or delay of various components of
minimal twitch are the orbicularis oris and orbicularis oculi. Of the blink reflex, it is possible with some assurance to localize
course, any other muscle may be used. the lesion's presumed site. Both central nervous system and pe-
Instrumentation Parameters. Facial muscle CMAPs are ripheral nerve lesion affecting the supraorbital and facial
considerably smaller than those obtained in the limbs and thus nerves can be investigated with this technique.
require a sensitivity of about 200-1,000 pV/div. The latency is Recording Electrodes. The most efficient manner to re-
rather short to the CMAP's onset necessitating a sweep speed of cord the blink reflex is using two channels to detect the three
about 1-2 ms/div. Filter settings are the same as those used for responses generated with stimulation of one supraorbital
median nerve motor studies. Stimulator parameters are noted nerve, specifically, the early ipsilateral RI and the bilateral de-
above. layed R 2 . The ipsilateral E-l to stimulation records two ip-
113
Reference Values. Stimulation of the facial nerve anterior silateral facial nerve responses, RI and R2. The contralateral
to the ear lobe yields a mean onset latency of 3.57 ± 0.35 ms E-l only records its orbicularis oculi R2. It is also possible to
(2.8-4.1 ms). Postauricular stimulation generates a mean onset record the blink reflex with only one channel, but more stimuli
latency of 3.88 ± 0.36 ms (3.2-4.4 m s ) . When comparing
141
are required.
Both cathode and anode should be maintained posterior to the of distal segment conduction by providing a smaller standard
sternocleidomastoid muscle as anterior placement may activate deviation and tighter normal range than distal latencies for both
the brachial plexus or phrenic nerve. If the brachial plexus is ac- motor and sensory studies. For example, let us assume that
1,0 , 2 9
tivated, depolarization of the supraspinatus muscle may be mis- two individuals have a DML for their right median nerve of 4.0
taken for the trapezius muscle response because of its close ms. This DML would be considered normal by most practition-
proximity. When the spinal accessory nerve is excited, the prac- ers. There may be diagnostic significance, however, in this
titioner should observe contraction of the trapezius muscle re- DML if one person had a forearm conduction velocity of 65 m/s
sulting in shrugging of the shoulder ipsilateral to the side of compared to the other subject with a forearm NCV of 52 m/s as-
stimulation. suming the DML is measured over an 8-cm segment in both in-
Instrumentation Parameters. The relatively short distance dividuals. The respective RLs would be 2.8 ms and 2.5 ms. The
between the stimulus and recording sites requires a sweep speed implication in these findings is that the comparative difference
between 1 and 2 ms/div. Other than sweep speed, routine motor between the predicted and actual DML is larger for the person
nerve conduction study parameters are used. with a forearm NCV of 65 m/s. This suggests that the distal seg-
Reference Values. The spinal accessory nerve should nor- ment of nerve for the subject with a forearm NCV of 65 m/s is
mally generate an onset latency of 1.8-3.0 m s . This is an im-
28
conducting slower than for the individual with the lower proxi-
portant technique to master because spinal accessory nerve mal NCV. The question then arises as to possible pathology af-
injuries are common and this technique can be quite productive fecting the distal segment of nerve with the larger R L .
when performing repetitive nerve stimulation in neuromuscular Normative data are available for both median and ulnar nerves
junction disorders or following lesions due to surgical proce- for motor and sensory studies (Table 6-5). Unfortunately, the
dures of the neck. clinical utility of the RL has only been examined in a limited
number of patients and needs further study to assess its true
clinical applicability. - no 129
MISCELLANEOUS TECHNIQUES
COLLISION TECHNIQUE
A number of specialized nerve conduction techniques may be
of clinical assistance under certain circumstances. Occasionally, Most routine studies excite the distal portions of peripheral
alternative methods may help to define a particularly challenging nerves where they are separated from neighboring nerves by
diagnosis. The residual latency, collision study, and refractory ' sufficient distances to allow selective neural excitation. Unless
period are electrophysiologic techniques that electrodiagnostic one is using large current intensities and durations, a single
medicine practitioners should be capable of performing. nerve can usually be examined. The selective delivery of a de-
polarizing pulse becomes much more difficult when attempting
RESIDUAL LATENCY to excite nerves in a proximal location such as the axilla. T h e
close proximity of the median and ulnar nerves often precludes
It is well known that nerve conduction velocities in proximal exciting either one individually. The result is a significant depo-
nerve segments are greater than in the distal portion of the larization of multiple upper limb muscles with occasional over-
nerve. Because NCV in general is directly proportional to axon lap of distal electrical responses. For example, suppose a
diameter, slowing is attributed to tapering of the nerve as it selective recording from the median-innervated thenar muscles
reaches the distal regions of the limb. - Consequently, in an
40 41
is the desired goal. This should pose no particular problem
upper limb a nerve cannot be expected to conduct with the same when activating the median nerve at the wrist or elbow provided
velocity within a few centimeters of the nerve's termination excessive current intensities are not used. The difficulty arises if
compared to a region in the forearm. However, if one were to a proximal conduction velocity of the median nerve is desired,
apply the forearm conduction velocity to the distance over i.e., axilla to elbow segment. It is highly probable that axillary
which the distal motor latency were measured, a time difference stimulation will result in coactivation of both the median and
between the predicted and observed distal motor latency would ulnar nerves as well as possibly the radial nerve. The recorded
arise. This difference is referred to as the residual latency CMAP from the thenar muscles may not be a true reflection of
(RL). - The concept of residual latency is perhaps best un-
110 129
the activity arising solely from the median-innervated thenar
derstood by using an example. Let us suppose a median nerve muscles. There is a good chance that the observed CMAP re-
conducts with a velocity of 60 m/s in the forearm and has a flects not only the median-innervated thenar muscle electrical
distal motor latency of 4.0 ms over an 8-cm segment. If one activity, but may also contain volume-conducted potentials from
were to assume that the NCV over the distal 8 cm also was 60
m/s, then the predicted distal motor latency would be 1.3 ms (60
T a b l e 6-5. Residual Latency ( m s ) 1 1 0 1 2 9
m/s = 8 cm/DML; DML = 1.3 ms). The difference between the

predicted and observed DMLs, residual latency, is 2.7 ms. In Control Neuropathy
other words, there is a 2.7-ms discrepancy between the observed Ulnar nerve (S) 1.3 ± 0 . 3 (0.8-1.8) 2.4 ± 1.0 (2.0-3.0)
and calculated DML. This same principle may be applied to
Ulnar nerve (M) 1.4 ± 0 . 8 (1.0-1.9) 3.0 ± 0.8 (2.7-3.3)
sensory as well as motor nerves only using the distal latency (to
initial takeoff of the SNAP) as opposed to the DML. A general Median nerve (S) 1.3 ± 0 . 3 (0.8-1.8) 3.4 ± 1.2(2.0-4.0)
formula may be used to determine the residual latency: RL = Median nerve (M) 1.5 ± 0 . 3 (1.0-2.0) 3.3 ± 1.0 (2.7-3.8)
DL - (cathode to E-l distance in mm/forearm NCV in mm/ms).
Median nerve (M)* 1.9 ± 0 . 2 (1.4-2.5)
The proposed diagnostic utility of residual latencies is to
S, Sensory RL; M, motor RL
compare the distal aspect of the nerve segment to the more f Median nerve RL (From Kraft GH, Halvorson GA: Median nerve residual la-
proximal aspect of the same nerve. Residual latency determina- tency: normal value and use in diagnosis of carpal tunnel syndrome. Arch Phys
tions should theoretically eliminate the intersubject variability Med Rehabil 1983;64:221-226.)
the neighboring ulnar-innervated hand intrinsic muscles such as solely from axillary excitation. It is then possible to calculate
the first dorsal interosseous (FDI) and adductor pollicis (AP). If the conduction velocity from this segment involving only the
the action potentials conducting in the median nerve fibers median nerve fibers. Delaying the axillary stimulation slightly
reach the thenar eminence first, then a correct DML is detected compared to that delivered at the wrist results in slightly more
with an appropriate proximal median nerve conduction velocity. separation between the two recorded CMAPs should this be
The amplitude, however, may be erroneous as it reflects activity necessary in selected cases. The collision of the two induced
from both median- and ulnar-innervated muscles. Depending ulnar nerve impulses is why the method is known as a collision
upon phase interactions of the two potentials, the amplitude technique. Of course, the principle of collision can be used for
may be larger or smaller than anticipated, although it is typi- any nerve and not just the ulnar nerve. Additionally, applying
cally larger. This situation would change if there was preferen- collision principles and appropriately separated stimulus inter-
tial slowing of the median nerve conduction across the wrist vals, one also can examine slower-conducting nerve fibers by
segment; the fastest-conducting fibers would be prevented from selectively blocking the faster-conducting axons. The collision
reaching the thenar muscles either by a conduction block or technique also may be of assistance in selectively blocking con-
axonal loss. duction in anomalous neural conducting pathways. - - 76 85 100 1 8 7
Stimulation of the median nerve at the wrist and elbow in the

above case would accurately reflect this slowing, yielding both REFRACTORY PERIOD
a prolonged DML and lower conduction velocity. Remember,
even though the distal segment is supposedly removed from Immediately following depolarization, that portion of an
nerve conduction velocity determinations for the elbow-to-wrist axon is completely inexcitable and cannot generate an action
segment, if the fastest fibers never reach the muscle, then the potential for a brief time. Within the next several milliseconds,
onset latency of the slower-conducting fibers determines the the axonal membrane becomes relatively excitable and can pro-
CMAP's onset latencies for all stimulus sites and hence the reduce an action potential, eventually returning to its resting
spective conduction velocities. When performing the axillary
118
state. It is possible to investigate the axon's membranous elec-
stimulation, it is highly likely that the coactivated ulnar nerve trical properties by delivering two successive stimuli with vary-
impulses will reach the hand intrinsic muscles prior to the ing interstimulus intervals. By convention, the first excitation
median nerve because of its slowing at the wrist. If the instru- pulse is referred to as the conditioning stimulus. The second
ment's sensitivity is relatively low or the ulnar nerve's nearby or test stimulus is then delivered at a predetermined interval.
muscles happen to coincidentally align their motor point with This terminology is used because the first excitation conditions
E-l, then an initial positive deflection is not observed and one the nerve, whereas the second depolarization tests the effect of
may erroneously conclude that the observed CMAP's negative the first stimulus on the nerve's voltage-dependent ion gates.
onset latency reflects median nerve conduction. The prolonged That time period after the conditioning excitation during which
antecubital median nerve latency combined with the shortened a test stimulus fails to evoke a response is referred to as the ab-
axillary median nerve latency results in a rather fast axilla-to- solute refractory period. A depolarization pulse, irrespective
elbow conduction velocity that is not a true reflection of the of strength, is incapable of inducing an action potential. At
median nerve's proximal neural segment conduction. Should a some point in time a test response can generate an action po-
positive deflection be observed with axillary excitation, it is tential but it is smaller than the conditioning response and de-
clear that one is not observing median nerve fiber excitation and layed in latency compared to the anticipated time of
no conduction velocity should be attempted. Should the positive observation with respect to when the nerve is activated. At
deflection be used to compute the conduction velocity, a similar some longer interval following the conditioning stimulus, the
situation to that described above results. The question remains, test response again resembles the conditioning response re-
is it possible to examine the proximal segment of the median garding appearance latency and amplitude. That segment of
nerve without contamination from the ulnar nerve? time following the absolute refractory period and detection of a
The proximal segment of the median nerve can be investi- test response identical to the conditioning potential is known as
gated by using coactivation of both the median and ulnar nerves the relative refractory time.
at appropriately separated time or distance intervals. If a supra- The proposed physiologic mechanism generating the two as-
maximal stimulus is delivered to the axilla and coincidentally at pects of reduced neural excitability is believed to be sodium in-
the wrist to only the ulnar nerve, an interesting electrical event activation. Recall that immediately following activation of
139
ensues. An early volume-conducted response from the ulnar-in- voltage-dependent sodium gates, action potential generation,
nervated hand intrinsic muscles is recorded from E-l located on the same voltage-dependent gates close, thus significantly re-
the thenar eminence secondary to ulnar nerve stimulation at the ducing sodium conductance. The closure of sodium gates is an
wrist. Because the origin of this CMAP is known to arise from intrinsic property of these proteinaceous channels and they
the ulnar nerve, it is ignored. The impulse induced at the wrist remain closed for a finite period of time irrespective of an addi-
also conducts proximally along the ulnar nerve. Recall that the tional depolarizing stimulus.
axillary impulse is traveling distally in both the ulnar and It is important to remember that sodium channel opening is
median nerves. At approximately the mid-arm level, the proxi- dependent upon a voltage difference and that their opening
mally and distally propagating ulnar impulses collide and anni- spans a finite time period. If the voltage applied to a nerve is
hilate each other. The median nerve impulse, however, slowly and progressively increased, it is possible to exceed the
continues distally to reach the thenar eminence generating a threshold level at which an action potential is generated. This
pure median nerve response. Because the median nerve action occurs because only a few sodium channels are induced to open
potentials originated in the axilla, the CMAP produced is suffi- at a time. As new channels are opened at a slightly greater volt-
ciently delayed in time so as to not overlap with the volume- age difference, the previously opened channels are closed or be-
conducted CMAP generated at the wrist by ulnar nerve ginning to close. The process of exceeding the nerve's threshold
excitation. The end result is a pure median nerve CMAP arising without action potential production is called accommodation.
On the other hand, just after the passage of an action poten- period. - People with motor neuron diseases also display pro-
4 151
tial, sodium gate closure or sodium inactivation renders the longed refractory periods.
membrane incapable of sustaining action potential induction. A limited number of investigations have been performed to
This time of complete inexcitability during which the sodium determine the clinical utility of neural refractory characteristics
gates are closed accounts for the absolute refractory period. in disease states. The relative ease with which refractory periods
Sodium gate closure and subsequent opening occur over a can be applied to the peripheral nervous system with commer-
finite time in that the gates do not all open and close simultane- cially available equipment should allow investigators to pursue
ously, i.e., this process occurs over a little less than 1 ms. As this area in the future. Direct muscle stimulation reveals that in
more and more of these voltage-dependent gates begin to re- muscle suffering from various forms of muscular dystrophy, the
cover from their mandatory inexcitable phase, at some point absolute and relative refractory periods are reduced compared
there is enough potentially excitable gates to again generate an to normal. Denervated muscle, on the other hand, reveals a
161
action potential, but one of less magnitude that takes longer to prolongation in both the absolute and relative refractory times.
generate the amount of current required to excite the next node The pathophysiology underlying these changes remains to be
I of Ranvier, i.e., propagation. A stimulus of sufficient magni- completely elucidated.
tude above the resting state's previous supramaximal level can Refractory period observations have been performed in ani-
induce a synchronous opening of the available sodium gates to mals for quite some time but this requires removal of the nerve.
produce a relatively small and delayed action potential. With As this is unacceptable for human studies, a simple yet elegant
progressively longer interstimulus intervals, more and more methodology has been developed that can be performed rou-
sodium gates capable of being excited become available. tinely by most practitioners with the appropriate equipment.
Correspondingly, less and less current is required to generate The actual methodology requires that one's instrument have the
an action potential. The increasing number of potentially ex- capability of delivering two stimuli with varying interstimulus
citable sodium gates allows threshold to be reached progres- intervals. With this type of stimulus delivery, it is relatively
sively earlier. Also, the larger number of sodium gates allows straightforward to examine either mixed or pure sensory nerves.
more current to flow, which in turn produces a larger action po- Mixed Nerve Studies. To perform mixed nerve refractory
tential until the test and conditioning waveforms are the same. period measurements, the technique of Gilliatt and Willison 75
The time between sufficient sodium gates to just generate an can be used.
action potential and enough to produce similar conditioning Recording Electrodes. E-L The E-l surface recording
I and test responses is the relative refractory period. Following electrode is located over the median nerve just proximal to the
I the relative refractory period is a supernormal period during antecubital fossa.
which the propagating test stimulus conducts at a velocity E-2. A surface E-2 electrode is positioned over the insertion
somewhat greater than normal. 75
of the deltoid on the lateral aspect of the arm.

Although the above description is correct, the actual tech- Ground Electrode. The ground electrode should be secured
nique requires propagated action potentials to be recorded at a to the forearm just distal to E-1.
distance from their production site. In other words, there may Stimulation. The median nerve is excited at.the wrist in a
be a time where an action potential may be produced locally at similar manner to that used for routine median nerve motor
the region of axonal membrane depolarization but it is of insuf- studies except the cathode is located proximal, i.e., pointing
ficient magnitude to result in propagation. Indeed, this is found toward E-l. A pulse duration of 0.2 ms may be used. Initially, a
to be the case and the time period between the absolute refrac- minimum threshold and single supramaximal stimulus is deliv-
tory period and the observation of a small and delayed propa- ered. The supramaximal response is then used to determine the
gating action potential is known as the critical interval of optimal recording electrode position for the mixed median
conduction. Of course, this time interval can best be mea-
213
nerve waveform.
sured with near-nerve microelectrodes. For practical purposes, An instrument with the capability of delivering sequential
however, the absolute and relative refractory periods can be pair of stimuli from the same cathode at predetermined inter-
conceptualized depending upon the detection or lack of a test stimulus intervals is required. Specifically, it is helpful if inter-
stimulus following a conditioning pulse. stimulus intervals between two successive stimuli of 0.1 ms can
be delivered. A stimulus exceeding the suprathreshold magni-
Clinical Utility tude 4—6 times is delivered at 0.1 ms intervals following the
By investigating the refractory periods of peripheral nerves, it conditioning stimulus to determine the absolute refractory
is possible to assess the effects of various disease states. In ex- period.
perimental demyelinating diseases of the peripheral nervous Once the absolute refractory period is determined it is possi-
system, experimental allergic neuritis, and diphtheria-induced ble to determine the relative refractory period. Beginning at the
demyelination, the refractory periods are significantly in- point when the second response was first detected with the max-
creased. - - In demyelination secondary to lysophosphatidyl-
31 32 179
imal stimulus, a response is attempted at the next 0.1-ms inter-
choline, refractory periods demonstrated a better correlation val. In this instance, however, only enough current is used to
with histologic findings than did conduction velocities. Of in-
204
produce a detectable response. This procedure continues at in-
terest is the finding of abnormal refractory periods in patients creasing intervals until the originally determined baseline stim-
with multiple sclerosis, suggesting that peripheral nervous ulus is reached. That stimulus interval between a just visible
system membrane characteristics may be altered in this response at 4-6 times stimulus threshold to the resting value de-
disease. Also, in patients with various peripheral neurop-
101
fines the relative refractory period. Continuing to increase the
athies, the relative refractory period appeared to be a more sen- interstimulus interval and measuring minimum stimulus excita-
sitive indicator of abnormality involving neural structures than tion levels allows one to calculate the supranormal period. The
the absolute refractory period. On the other hand, hypokalemia time when the original threshold value is required to just elicit a
has been found to actually shorten the relative refractory potential defines the cessation of supranormality.
It is important to note that delivery of the high-intensity cur- Refractory Periods in Muscle
rents/voltages required to properly study the refractory periods In addition to measuring the refractory periods in nerve, it is
can be quite uncomfortable and not tolerated by all patients. also possible to determine the absolute and relative refractory
Additionally, proper skin site preparation with commercially periods in muscle fibers. Using the paired stimulation tech-
available abrasives to reduce impedance is recommended. nique, direct muscle fiber stimulation can be performed while
Instrumentation Parameters. A sweep speed of 1 ms/div and recording from single muscle fibers. I6,
The studies reveal that
2(>6
amplifier sensitivity of 20 pV/div should suffice for most persons. the absolute refractory period in muscle with a stimulation in-
Filter settings of 10-20 Hz to 2 kHz will yield detectable responses. tensity 2 5 - 3 5 % above the conditioning stimulus is 4.12 ± 1.73
Reference Values. The absolute refractory period measured ms (2.69-8.13 ms). The relative refractory period for muscle
with above technique was found to be less than 0.6-0.7 ms. In 75
fibers is 5.99 ± 2.7 ms (2.88-12.40 ms). A supranormal period
other words, the second potential was first observed at an inter- also can be observed at 10.19 ± 3.2 ms (4.86-15.7 ms). As for
stimulus interval of 0.6-0.7 ms. The relative refractory period nerve, the waveforms in the relative refractory period are
lasted between 2.5 and 3.5 ms. Following the relative refractory smaller and demonstrate an increase in the rise time and a
period, a supranormal time interval extended for 5-8 ms. longer total duration.
Sensory and Motor Nerve Refractory Periods

In addition to examining mixed nerves, it is also possible to LATE RESPONSES
measure the refractory periods of pure motor and sensory
nerves using paired stimuli techniques similar to those noted Following the CMAP or M response in motor NCS a
above. Sensory nerve refractory periods can be calculated by number of secondary or late responses can be observed on the
placing stimulating ring electrodes, cathode proximal, on the CRT several milliseconds later. Depending upon the particular
second digit. Recordings were initially performed with near-
210
physiologic conditions, there are three late responses of interest
nerve needle E-l /E-2 recording electrodes at the wrist separated that are discussed in this section: F-wave, H-reflex, and axon
by 3 cm. The sural nerve was also examined in a similar reflex. These three individual waveforms are essential to gain
manner. The absolute refractory periods for the median and insight into the physiologic mechanisms underlying the periph-
sural sensory nerve fibers were approximately 0.7 m s . - In 210 211
eral and central nervous systems. Additionally, a number of in-
these studies the relative refractory period was assessed by both vestigators have proposed various techniques whereby the late
amplitude and latency criteria. Amplitude criteria suggested rel- responses may be used for diagnostic purposes with regard to
ative refractory durations of 5 times the absolute refractory pathology involving specific regions of the peripheral nervous
period, whereas latency criteria revealed a length of 3 times the system. Each response is discussed in detail and their clinical
absolute refractory period. relevance to particular disease entities is noted during the re-
Refractory periods in motor nerves also can be studied; how- mainder of this text when appropriate.
ever, the rather long duration of the conditioning CMAP inter-
feres with the necessary latency measurements of the test F-WAVE
response. An alternate method of calculating the refractory
times other than direct paired stimuli is required. A collision In 1950, Magladery and McDougal first detected a small and
technique (see above) was developed to eliminate the interfer- late response occurring after the CMAP elicited from the per-
ing effects of the first stimulus while continuing to investigate oneal innervated foot muscles and designated it the F-wave (F
the interactions of the conditioning and test r e s p o n s e s . 117120122
from foot). The above two investigators noted that the F-wave
145
For example, surface recordings are obtained over the hy- increased in amplitude and reached a maximum at supramaximal
pothenar eminence while CMAPs resulting from paired stimuli stimulation of the peripheral nerve, varied in amplitude from
at the axilla combined with a solitary pulse at the wrist are ex- subject to subject, displayed different morphologies from one
amined. With this technique, the conditioning stimuli is blocked stimulus to the next as well as slightly different latencies, and
when it collides with the action potentials propagating toward it that not all CMAPs were followed by an F-wave (Fig. 6-9). Of
from the wrist. The second stimulus from the axilla is then free interest was the observation that moving the stimulus site from
to propagate to the hypothenar muscle and produce a response the elbow to the distal forearm resulted in a shortening of the
provided the nerve is not in the absolute refractory period in- CMAP but a prolongation of the F-wave from 26 ms to 31 ms.
duced by the axillary conditioning response. The CMAP result- The decrease in the CAMP onset latency was expected because
ing from wrist stimulation is sufficiently displaced from the the excitation site moved closer to the muscle from which the re-
axillary CMAP to offer no interference. By appropriately ad- sponse originated. The increase in F-wave latency, however, sug-
justing paired stimuli at the axilla, one is free to investigate the gested that the neural impulses generating this response had a
membrane properties regarding refractory characteristics of longer pathway to travel prior to reaching the hypothenar mus-
pure motor nerves in a similar manner used for sensory and cles. Additionally, F-waves were noted to be absent when a stim-
mixed nerves. Absolute refractory period for the ulnar motor ulus was delivered to the ulnar nerve distal to a complete
nerve is 0.77 ± 0 . 1 8 ms, and the relative refractory period is procaine block of the nerve. Faced with these observations,
2.03 ± 0.57 ms. It is also possible to investigate the refractory Magledary and McDougal concluded that the F-response could
periods of single motor units by stimulating a mixed nerve but not arise from repetitive firing of the motor nerve, neuromuscu-
recording from just one motor unit with intramuscular record- lar junction, or muscle but must be a delayed potential that first
ing techniques. 15
In the peroneal nerve, the absolute refrac-
1 8 1 2 7
travels centripetally toward the central nervous system and then
tory period is 1.28 ± 0.22 ms. This is most likely the case centrifugally back to the muscle. The F-wave, therefore, some-
because the peroneal nerve has slightly lower conduction veloc- how involved the central nervous system via motor neuron dis-
ities than upper limb nerves and the refractory period is in- charge, either through a backfiring of the anterior horn cells or
versely proportional to conduction velocity. 14168
through a reflex mechanism involving afferent-to-efferent central
connections. Also, the small amplitude of the F-wave implied

that only a select population of motor neurons responded to the
peripheral depolarization pulse. They concluded from ischemic
conditions applied to peripheral nerves that the F-wave was a
result of a reflex mechanism and not backfiring of the anterior
horn cell. The major issues to be resolved were the pathways in-
volved in the production of the F-wave, an explanation of the
small amplitude, variable latency, and changing morphology,
and the diagnostic utility of this response.
Physiology of F-Wave Production

In addition to Magledary and McDougal, a number of other
investigators assumed that the F-wave was a reflex response
mediated through an oligosynaptic or polysynaptic pathway re-
quiring afferent fiber a c t i v a t i o n .
83133134
Shortly after the F-wave
was first described, a group of investigators suggested that in-
stead of a reflex pathway, the F-wave was produced by a motor
neuron activated through an antidromic impulse, i.e., a backfir-
ing mechanism. Sectioning the posterior roots supplying limbs
46
to be examined in both animal and human subjects demon-

strated little change in the production of F - w a v e s .
73154157
-160 218
Further support for the lack of a reflex or synapse involved in F-

wave production occurred when single-fiber electromyography 5ms
demonstrated essentially the same delay or jitter (see Chapter 8)
from one F-wave to the next as observed in the same muscle Figure 6-9. F - w a v e s e r i e s . A series of F-waves resulting from
fiber. 95.228.229 \ other words, only one neuromuscular junction
1
n
median nerve wrist stimulation and recording from the abductor pollis
or synapse was involved in F-wave generation that was present brevis. N o t e the variable latency and morphology of the F-waves. O f
in the muscle. If a reflex were involved in the F-wave, a synapse interest, each CMAP is preceded by a p r e m o t o r potential.
interposed between the afferent and efferent neural pathway
would be necessary. This synapse would significantly add to the following dorsal root section in cats, stimulating a motor nerve
transmission variability from one F-wave firing to the next, thus resulted in the anticipated large antidromic impulse being con-
increasing the jitter. When removal of the anatomic pathways ducted toward the central nervous s y s t e m . Recording di-
181182
conveying the afferent electrical impulses resulted in F-wave rectly from the same motor nerves revealed a second impulse
generation, it had to be concluded that the F-wave did not only 2 - 3 % of the original amplitude that required-a central turn-
depend on a reflex. The only alternative clearly suggested that around time or delay of approximately 1 ms. These neural im-
following activation of a mixed nerve, a small late response was pulses correspond to the F-wave response described by
observed that originated from the antidromic motor impulses Magladary and M c D o u g a l even though Renshaw recorded
145
propagating centripetally and activating a small population of them from the nerve, whereas the F-wave was observed in
motor neurons. The limited number of excited motor neurons muscle. In other words, Renshaw documented the neural re-
then generated an impulse that traveled orthodromically in sev- sponse responsible for the muscular potential produced by the
eral motor nerves to activate the muscle fibers they innervate. backfiring neural impulses. In both animal and human investi-
These reactivated motor units were the potentials designated as gations, the F-wave is between 1 and 3 % of the CMAP, which
the F-wave. corresponds nicely to the percentage of total nerves found to be
Given that F-waves are believed to be generated by an an- activated and represents roughly 1-2% of the available motor
tidromic backfiring of motor neurons, it is reasonable to ask neuron p o o l . - - - When individual F-waves are examined
5, 8U24 155 182
why the F-wave amplitude is significantly less than the previ- with needle recording techniques, each F-wave is found to consist
ously generated CMAP. When considering the amplitude of the of 1-3 motor units, roughly supporting the previously noted
F-wave, it is important to first consider factors that may affect data. The actual explanation for the small number of motor
190
the magnitude of the motor units contributing to the F-wave. neurons activated by an antidromic impulse is poorly understood.
The number of muscle fibers and their cross-sectional diameter In order to consider the relatively few motor neurons acti-
comprising a particular motor unit and how closely these fibers vated following depolarization of an entire mixed nerve, it is
are arranged in space can influence a potential's amplitude. The first necessary to briefly consider the anatomy of the anterior
more fibers per motor unit and a given area, the more voltage horn cell. The anterior horn cells concerned with motor function
produced during depolarization and the bigger the F-wave ob- consist of a relatively large soma or main body with several sub-
served. Also, the total number of motor units activated and their stantial projections emanating from it. One rather large projec-
temporal dispersion with respect to each other directly affect F- tion is the axon destined to innervate all of the muscle fibers
wave amplitude. Several motor units temporally synchronized innervated by that motor neuron. The unmyelinated portion of
(superimposed) yield a larger potential than if they were more the motor neuron forming the junction between the last myeli-
separated in time. The implication in the relatively small F- nated segment of the axon and the main portion of the soma is
wave amplitude compared to the CMAP is that only a small referred to as the axon hillock. The axon hillock's threshold for
subpopulation of available motor neurons is activated by all of depolarization is approximately one-half that for the remaining
the antidromically propagating motor impulses. An explanation portions of the motor n e u r o n . Dendrites are the remaining
11198
for this assertion is obviously required. Renshaw observed that projections from the soma. In excess of 6,000 synapses with
the ventral horn of the spinal cord to synapse with inhibitory in-
terneurons known as Renshaw cells (Fig. 6-10). Renshaw cell
activation tends to suppress activation of motor neurons they
synapse with by generating inhibitory presynaptic potentials
(IPSPs). - As an antidromic impulse traverses the axon
45 198
toward the ventral horn, the axon collateral conveys an action

potential to the Renshaw cell, which in turn tends to suppress
the motor neurons it synapses with.
The final level of excitability of the motor neuron pool, there-
fore, is dependent upon multiple excitatory and inhibitory influ-
ences from various aspects of the central and peripheral nervous
systems. When a mixed peripheral nerve is stimulated with
176177
a supramaximal stimulus, the large number of antidromic motor

action potentials enter the ventral horn to find the resting mem-
brane potentials of their respective motor neurons' soma at vari-
ous levels. Whether a particular motor neuron generates a
recurrent discharge depends upon the level of depolarization of
the soma and its dendrites. Let us assume that the resting mem-
brane potential of the axon hillock favors depolarization of this
Figure 6-10. Renshaw cell activation.The alpha m o t o r neurons region, thus facilitating action potential propagation into the
possess a recurrent- collateral portion of t h e axon just distal t o t h e motor neuron soma from an antidromically induced impulse.
axon hillock, which extends t o inhibitory interneurons known as This action potential then propagates into not only the main por-
Renshaw cells (R). O n c e the r e c u r r e n t collaterals activate t h e tion of the soma but also into the various expanses of the alpha
Renshaw cell, it in turn synapses with alpha m o t o r neurons t o gener- motor neuron's dendrites (Fig. 6-11). By the time the depolar-
ate inhibitory postsynaptic potentials ( - ) , which suppress firing of ization has reached the distal portions of the dendrites, the axon
these neurons. hillock has undergone repolarization and is no longer in its re-
fractory period (about 1 ms). - The negative sinks of the den-
46 4 74 8
drites are causing the ions surrounding the axon hillock to serve
other dendrites occur over the motor neuron's soma and gener- as a current source for the dendritic depolarization. This tends
ate either excitatory or inhibitory i m p u l s e s . The net summa-
11,98
to alter the ionic distribution around the axon hillock by de-
tion of excitatory and inhibitory potentials determines the creasing the positive charge on its surface. The transmembrane
overall excitability of the motor neuron and whether it generates voltage alteration may induce an action potential to occur at this
a depolarization impulse of sufficient magnitude to excite the portion of the axon, thus generating the recurrent backfiring of
axon hillock region producing a propagating action potential. the motor neuron begetting the subsequently observed F-wave-
Within a short distance distal to the axon hillock, a number of (Fig. 6-11). The critical time period or "window of opportunity"
spinal motor neurons possess a recurrent collateral, which is a between repolarization of the axon hillock coinciding with
neural branch given off from the axon that proceeds back into soma/dendritic local circuit currents is about 10-30 p s . 195
A B C
Figure 6-11. Motor neuron "backfiring." Proposed mechanism of the so-called alpha m o t o r neuron's "backfiring" t o generate an F-wave. A,
Initially t h e action potential enters the axon hillock region and begins depolarization of the anterior horn cell's soma. Solid arrows are the sodium
ions carrying the inwardly directed current while dotted arrows are the internally directed current. B,This depolarization then extends into the
dendritic extensions of the motor neuron while the axon hillock is refractory. Because the m o t o r neuron's dendrites are depolarizing similar t o an
unmyelinated nerve, i.e., continuous and n o t saltatory, the axon hillock exits its refractory period while depolarization is still occurring in t h e den-
drites.The dendrites regions of depolarization act as a current sink while the sodium ions surrounding the axon hillock serve as a current source.
C,A source current from the region of the axon hillock alters t h e transmembrane voltage (less positive extracellular) and this tends t o depolarize
the axon hillock generating an impulse propagating toward t h e periphery, i.e., an F-wave is then detected.
Should the soma's membrane be depolarized to an extent ex- the proximal segment of the peripheral nervous system or depo-
ceeding that previously described, it and the dendrites will de- larize the soma preventing repolarization. Finally, there is a
81136
polarize comparatively early and generate a local circuit current greater chance of shortening the depolarization time of the soma
during the refractory period of the axon hillock. This situation in smaller motor neurons, possibly because of suprasegmental
results in the failure of F-wave production. On the other hand, influences lowering the resting membrane threshold of the
the transition between the myelinated portion of the axon and smaller motor neuron soma. - It is known that smaller motor
10 182
the axon hillock does not favor conduction into the soma be- neurons fire at lower thresholds than larger motor neurons giving
cause the current distribution is diluted over the nonmyelinated rise to the orderly recruitment of motor neurons, i.e., the
portion of the axon hillock. In other words, the current distribu- Hennemann size principle. - Smaller motor neurons, therefore,
91 92
tion is not concentrated at a node of Ranvier but spread out over may have resting membrane levels closer to the depolarization
the surface of the axon hillock. If the resting membrane poten- threshold compared to larger ones. This may be an important
tial of the axon hillock is relatively hyperpolarized because of mechanism of recurrent discharge inhibition in smaller motor
segmental and suprasegmental influences, action potentials will neurons because recurrent collaterals are found in approximately
not cross this region to invade the soma and dendrites. In this 70-80% of them, leaving 20-30% without the possibility of re-
case, an F-wave is not produced. The reason only a small current inhibition. Should the threshold be lowered in smaller
194
number of F-waves are observed, therefore, is because of the re- motor neurons, they will depolarize rather quickly, generating an
quired convergence of a number of excitatory and inhibitory in- action potential in the soma-dendrite region and creating a local
fluences favoring action potential conduction across the axon circuit current incapable of exciting the axon hillock because it is
hillock with an appropriate temporal delay across the soma and still refractory. These three mechanisms or some combination
dendrites favoring reactivation of the axon hillock. This situa- may be the reason why there is preferential activation of rela-
tion changes from moment to moment, thereby resulting in a tively larger motor neurons generating the detectable F-waves.
different subpopulation of motor neurons amenable to depolar-
ization by an antidromic means with each ensuing stimulus. Diagnostic F-Wave Techniques
The variable latency of sequentially elicited F-waves may be A number of investigators have developed several interesting
understood if one considers the motor neuron population pro- methodologies in which the F-wave can be used diagnostically.
ducing the individual F-waves. Investigations in both humans The basic parameter used by all investigators is the F-wave la-
I and animals reveal that there is a greater chance of F-waves tency. Because sequential F-wave latencies are variable, innova-
being generated by comparatively larger motor n e u r o n s . - 81 124169
tive strategies have been developed to address this potential
' Larger motor neurons give rise to relatively large axons that problem. Some of the techniques discussed include mean F -
have faster conduction velocities than smaller axons from the wave latencies over various body segments, latency ranges, F-
smaller motor neurons. Also, larger motor neurons innervate wave conduction velocities, and F-wave latency ratios. Only a
more muscle fibers, thus creating larger motor units with larger- few investigations, however, have addressed amplitude for diag-
magnitude F-waves. The resultant F-waves detected, therefore, nostic purposes.
preferentially arise from the faster-conducting axons that have a An F-wave may be obtained from essentially any muscle pro-
certain diameter range. This diameter distribution yields axons vided a supramaximal stimulation is used and the amplifier's
conveying F-waves with slightly different conduction velocities. sensitivity is sufficient to detect the response. Because of the
Since there are very few F-waves that repeat with sequential relative long duration of the CMAP the F-wave may be un-
stimulation (19.5%), a large number of different F-waves are recordable in short nerve segments. The amplifier should be set
observed from the available pool of motor neurons. The vari- 81
at approximately 100-200 pV/div to ensure observation of the
able latency of F-waves represents the distribution of conduc- F-wave. Of course, a sensitivity of this magnitude does not
tion velocities of axons mediating the recurrent responses. permit one to simultaneously observe the entire CMAP. The
Of the 1-2% of motor neurons capable of producing an F- rather delayed latency of the F-wave with respect to the CMAP
wave secondary to segmental and suprasegmental inhibitory in- requires a sweep speed of 5 ms/div and 10 ms/div in the upper
fluences, one may attempt to understand the preferential bias and lower limbs, respectively. The easiest muscles to record F-
toward larger motor neurons generating F-waves. First, there is a waves from are the small intrinsic hand and foot muscles.
greater chance of larger motor neurons, through axon collaterals Because of this, the majority of reference data available pertain
activating Renshaw cells, inhibiting smaller ones. - This is be-
46 81
to the following muscles: abductor pollicis brevis, abductor
cause of the faster conduction velocity of antidromic impulses in digiti minimi, extensor digitorum brevis, and abductor hallucis.
larger axons (bigger motor neurons) reaching the Renshaw cells Routine recording techniques previously described are used. It
before the smaller axons (smaller motor neurons) and exerting a is not necessary to relocate the anode distal to the cathode when
blocking influence on the smaller motor neurons. In a sense, the exciting the nerve as anodal block most likely does not occur.
faster axons compete for optimal levels of resting membrane po- When stimulating the nerve, an optimal stimulus rate is 1 Hz or
tentials for recurrent motor neuron excitation with each other, less.150
whereas the smaller motor neurons have a reduced chance of Slight contraction of the muscle under investigation can facil-
generating an F-wave. Also, it is easier for Renshaw cells to in- itate the observation of F-waves should one note a decreased
hibit smaller motor neurons as there is less soma membrane to ability to record them. This is most likely mediated through in-
be affected. - Secondly, the afferent fibers of smaller motor
47 79
creased motor neuron pool excitability. Caution is required
neurons conduct slightly faster than their corresponding motor when attempting to facilitate the F-wave because amplitudes
fibers. These afferent impulses may reach the spinal cord prior to may be increased and an H-reflex (see below) may contaminate
the antidromically excited motor fibers setting up a reflex re- the desired responses. The effect of facilitation, however, is not
sponse in which the small motor neuron is reflexively excited. consistent. Although difficult to quantify, a reduction in the
195
The reflex-induced action potential from the small motor neuron numbers of F-waves following supramaximal stimulation may
would then collide and cancel the antidromic action potential in indicate pathology.
The rationale for attempting to record the F-wave is multifac- soleus: 2.8 ± 1.1 . - - - - - The major limiting factor in
ms
59 169 170 171 176 177
torial. Recall that the F-wave is a potential that represents con- performing the F chronodispersion technique is that 100 F-
duction from the site of stimulation to the motor neuron and waves must be acquired in order to obtain a large distribution of
back to the recording electrode, i.e., both the proximal and latency differences. Patient tolerance and the time required to
distal regions of the peripheral nervous system. As demon- calculate these data are major drawbacks to routinely using this
strated above, it is relatively easy to examine conduction in the technique despite its reported sensitivity to pathology. 170
distal portions of the peripheral nerves. Assessment of proximal Occasionally, one may wish to calculate the F-wave latency
conduction becomes technically more demanding and subject to over a localized proximal segment such as the brachial plexus.
volume conduction effects. The F-wave impulse, however, orig- Obviously, stimulating the median or ulnar nerve at the wrist in-
inates distally where there is less ambiguity of the nerve ex- cludes the entire nerve segment from wrist to spinal cord and
cited; it also is recorded distally, with little interference from back to the muscle. By subtracting the CMAP distal motor la-
neighboring muscles. Theoretically, the F-wave appears to be tency to wrist stimulation from the shortest F-wave latency and
the ideal parameter to use to assess proximal conduction. then subtracting an additional 1 ms, a conduction time for the
Although there are a number of limitations regarding the F- fastest conducting F-wave from wrist to spinal cord and back to
wave (see below), this concept is generally correct. the wrist is obtained. It is necessary to reduce the conduction
time by 1 ms because this is believed to represent the turnaround
F-Wave Latency time for motor neuron reactivation in the spinal cord. It is impor-
As previously stated, the F-wave latency varies from one tant to note that this presumed 1-ms turnaround time has never
stimulus to the next (Fig. 6-9). It is important to record a suffi- been documented and obviously presents itself as a potential
cient number of F-waves to ensure analysis of a representative complicating factor in various techniques using this time frame.
sample of the total available pool of motor neurons producing Further, dividing this latency by 2 allows one to determine the
F-waves. The exact number of F-waves necessary to produce a conduction time from wrist to spinal cord, the central conduc-
representative number is unknown. The practicality of available tion time. In other words, the equation representing this latency
time for F-wave collection during an electrodiagnostic medicine is: central conduction time = (F-wave latency - DML - 1 ms)/2.
examination also must be considered. A number of investigators The problem with this method is that a small lesion in a proximal
recommended obtaining between 10 and 20 F-waves per stimu- portion of the peripheral nervous system could be diluted out
lus site. Although gathering still larger numbers of F-waves may over the spinal cord to wrist distance, thereby reducing the sensi-
yield a few with shorter latencies, the diagnostic utility versus tivity of this technique. An alternative method is to stimulate the
time consumed becomes prohibitive. F-wave reference data are median or ulnar nerve in the axilla and measure the F-wave over
somewhat variable from one investigator to the next not only this comparatively shorter segment. Unfortunately, the CMAP
because of the inherent variability of the response itself, but also and F-wave occur at about the same time, thus obliterating the F- \
because the latency depends on the stimulus site. As there are wave. A second stimulation applied at the wrist simultaneously
no universally accepted standards with respect to distance be- with axillary excitation collides with the orthodromic axillary
tween the cathode and recording electrodes, the F-wave demon- impulses permitting detection of the axillary F-wave through a-
strates slightly different mean values from one laboratory to the collision technique. A simpler method to examine the proximal
60
next. For the F-wave latencies reported, the median and ulnar F-wave latency is to stimulate the desired nerve in the axilla 25
nerves are excited just proximal to the distal wrist crease, cm from the sternal notch with the arm abducted 90° and the
whereas the tibial nerve is activated posterior to the superior forearm supinated. The shortest F-wave latency from the wrist
97
margin of the medial malleolus and the peroneal nerve just is then added to the previously obtained CMAP DML from
above the ankle region (Table 6-6). The previous locations of an which is subtracted the axillary CMAP latency multiplied by 2
8 cm standard distance should result in similar mean F-wave la- and is called the axillary F-Ioop latency (AFLL): AFLL = (F-
tencies. Recommended side-to-side differences for both short- wave + DML) - 2 axillary latency. An axillary F-loop latency in
est latency and mean latency are 2.0 ms in the upper limb and excess of 11.0 ms is considered abnormal. - Because this tech-
97 240
4.0 ms for lower limb intrinsic muscle studies. In general, F- 63

nique involves the fastest F-waves, an attempt was made to in-
wave latencies are directly related to height and limb length as crease the sensitivity by averaging 32 F-waves and measuring
anticipated given the length of the neural pathway, but there is the averaged F-wave peak latency and inserting this value into
minimal correlation to age and gender. - - 59 126 176177
the previously defined AFLL equation. Normal values for the
In the above technique, the shortest F-wave latency may be median and ulnar nerves were reported as 14.12 ± 0.88 ms and
used to determine pathology if a given nerve is injured. The dif- 13.97 ± 0.9 ms, respectively. 98
ficulty in using the shortest F-wave is that the study is biased

toward one nerve fiber. If there is significant damage to the pe- F-Wave Conduction Velocity
ripheral nervous system but one or a few of the fastest-conduct- Once the shortest F-wave of a series is obtained, it is possible
ing fibers survive, then a normal study is declared. A more to convert this latency into a conduction velocity. There are 115116
rational approach is to consider the mean value of a group of F- two major assumptions involved in using F-wave conduction
waves recorded. The mean onset latency of 10 or more F-
212
velocities. The first assumption is that the shortest F-wave is de-
waves is believed to be more sensitive than only considering the tected within the limited number of responses obtained, less
fastest F-wave. - -
52 39 98
than 20, and these correspond to the motor fibers producing the
One also may attempt to measure the latencies of a large onset latency for the CMAP detected with distal stimula-
number of F-waves, 100 or more, and calculate the time differ- tion. - It has been clearly demonstrated that the shortest F-
130 242
ence between the shortest and longest F-waves. This technique wave does not always occur within the first 20 potentials, but
has been referred to as F chronodispersion. F chronodisper-
169
may require up to 100 or more responses. The second as- 169
sion reference values for a number of muscles are known: APB: sumption requires an accurate measurement of the conducting
3.6 ± 1.2 ms; ADM: 3.3 ± 1.1 ms; EDB: 6.4 ± 0.8 ms; and pathway traversed by the impulses generating the F-wave. This
is rather easy for the limb, but the difficulty arises when proxi- F-Wave Ratio
mal segments across the brachial or lumbosacral plexi are in- Because of the potential for distance measurement errors in
volved. It has been determined in a very limited number of calculating F-wave conduction velocities, an alternative F-wave
anatomic specimens that measuring from the stimulus site, technique was developed that does not involve distance. - It 49 50
ankle or popliteal fossa, to the T12 spinous process by way of was determined that if the median or ulnar nerve was stimulated
the greater trochanter approximates rather well the true at the elbow region, the time of conduction for the F-wave to the
anatomic length of the tibial nerve. The same anatomic verifi-
130
spinal cord was very similar to the latency for direct motor
cation, however, has not been determined for the upper limb. nerve activation from the same site to the muscle, i.e., C M A P
Although F-wave conduction velocities have been criticized be- onset latency. In other words, the F r a t i o is close to unity.
cause of the unnecessary addition of a potentially large error Similar findings were noted for tibial and peroneal nerve stimu-
due to less than accurate distance m e a s u r e m e n t s , conduc- 130243
lation while recording from the intrinsic foot muscles (Table 6¬
tion velocities nevertheless continue to be used. The use of F- 5) ii9.i2o.i2i Tftg equation used to determine F ratios is:
wave conduction velocities has been justified on the basis of
noting that the difference in latencies between stimulating the _ (F-wave latency - CMAP latency - 1 ms)/2
F ratio =
peroneal nerve at the ankle and knee while recording from the CMAP latency
EDB correspond to the differences in F-wave latencies from
these two sites, i.e., 6.5 ms and 6.4 ms, respectively. The impli-
cation of this finding is that the shortest-latency motor fibers de- - (F-wave latency - CMAP latency) - 1 ms
F ratio
termining CMAP onset latency correspond to similar fast fibers CMAP latency x 2
mediating the shortest F - w a v e . In calculating F-wave con-
121124
Although it is possible to calculate F ratios with either more
duction velocities for upper limb examinations, the distance proximal or distal stimulation sites, the variability of data is
from the point of stimulation is measured to the C7 spinous minimal with elbow and popliteal fossa excitation. Motor nerve
process with the arm abducted 90°. The equation used to calcu- and F-wave conduction velocities may both be abnormal yet the
late F-wave velocities for both intrinsic hand and foot muscles F ratio can be within normal limits. This suggests that not only
is: are the peripheral nerves conducting slowly over both the distal
and proximal segments, but they are slowed to a similar degree.
F-wave C V (m/s) = (distance to T12 o r C7 in m m ) x 2
(F-wave latency - CMAP latency - 1 ms) F-Wave Amplitudes and Persistence
Normal values for both upper and lower limb nerves at multi- In disorders in which the central excitability of the motor
ple stimulation sites are provided (Table 6 - 6 ) . The F-wave 123
neuron pool is decreased, one could anticipate both a reduced
conduction velocity has been reported to be of value in detect- number and smaller amplitude of F-waves. This has been found
ing proximal slowing in various disease states affecting the pe- to be the case in patients examined immediately following a
ripheral nervous s y s t e m . There is some suggestion that
5 8 1 1 4 1 1 5
unilateral stroke. Excitation of the cerebellum also can d e -
58
using an averaged F-wave latency to calculate F-wave conduc- crease F-wave amplitude and persistence. - On the other hand,
58 66
tion velocities may be of greater sensitivity in detecting abnor- in patients with chronic myelopathies and spasticity, F-wave
mality compared to the shortest F-wave latency. - A 59 98
persistence and magnitude are increased commensurate with the
modification of the F-wave chronodispersion using the distrib- elevated excitability of the motor neuron pool. The latencies
51
ution of F-wave conduction velocities ( F tacheodispersion) is of F-waves in patients with upper motor neuron lesions, how-
believed to be a sensitive method of defining peripheral nerve ever, may be prolonged secondary to the unmasking of smaller
conduction abnormalities but more studies are required to fully motor neurons (slower peripheral conduction) while the larger
evaluate this technique. 30
ones are blocked secondary to rapid depolarization. - It is possible
60 61
T a b l e 6-6. F-Wave Reference Values 1 2 3
FWCV from
MNCV between Cord to
Number of Site of M latency F latency F ratio F ratio (R) Two Stimulus Stimulus
Nerves Tested Stimulation (msec) (msec) (F - M - l)/2M F ratio (L) Sites (m/sec) Site (m/sec)
66 Median nerves 3
Wrist 3.5 ± 0.5 29.1 ± 2 . 3 52.9 ± 3.9
Elbow 7.8 ± 0.8 24.8 ± 2.0 1.04 ± 0.09 1.01 ± 0.07 56.0 ± 5.0 62.2 ± 5.2
Axilla 1 1.3 ±1.0 21.7 ± 2 . 8 63.3 ± 6.0 64.3 ± 6.4
66 Ulnar nerves b
Wrist 2.9 ±0.5 30.5 ± 3.0 56.7 ± 2.9
Below elbow 6.7 ± 0.7 26.0 ± 2.0 " 1.40 ± 0 . 1 1 0.99 ± 0.09 55.9 ± 5.1 58.2 ± 2.9
Above elbow 9.2 ± 0.9 23.5 ± 2.0 56.9 ± 4.6 61.1 ±5.4
Axilla l l . 2 ± 1.0 21.9 ± 1.9 61.3 ± 6.8 63.0 ± 5.9
66 Peroneal nerves Ankle 4.5 ± 0.9 51.3 ± 4 . 7 53.3 ± 3.7
Knee 12.9 ± 1.4 42.7 ± 4.0 I.I 1 ± 0.09 1.02 ± 0 . 0 9 49.4 ± 3.8 56.3± 4.9
66 Tibial nerves Ankle 4.1 ± 0 . 6 52.3 ± 4.3 51.3 ± 2 . 9
Knee 12.8 ± 1.3 12.8 ± 1.3 I.I7±0.I0 1.00 ± 0 . 1 0 46.8 ± 3.4 54.4 ± 3.6
a
F wave was elicited by axillary stimulation in 42 of 66 nerves.
b
Middle segment across elbow was tested in 34 of 66 nerves.
2 4 4 — P A R T II BASIC A N D ADVANCED TECHNIQUES
to calculate the ratio of the F-wave to that of the corresponding process affecting at least the motor aspects of the peripheral ner-
CMAP (F/M ratio) in an attempt to measure the amount of the vous system.
motor neuron pool activated. Because of the variability of the F- In addition to a mild diffuse peripheral nerve process, a prox-
wave amplitude, mean amplitudes calculated from a series of F- imal lesion in the neighborhood of the brachial plexus or more
waves appears to be the most reasonable method. '- - - - The
5 58 62 66 67
rostral (root level) may be amenable to diagnosis by the use of
clinical utility of F/M measurements in routine electrodiagnos- F-waves. These "proximal" lesions are limited to two disorders,
tic medicine examinations remains to be demonstrated. and under specific conditions. The first is Guillain-Barre syn-
drome, in which a reduced number or absence of F-waves may
F-Wave Clinical Utility be observed secondary to an early and significant blockade of
The clinical utility of various F-wave techniques is by no action potential propagation across the root region. This may
171
means universally agreed upon by even a minority of practition- be the only abnormality noted early on in some but not all pa-
ers. - As a result, the authors will exercise their prerogative
65 185
tients with this disorder. We do not mean to imply that this is the
based on clinical experience and a review of the literature that most sensitive technique for diagnosing this entity, but rather
some readers may disagree with. that the practitioner should not forget to address these regions of
As noted above, the F-wave is a very long conduction path- the nervous system in a patient that may be presenting in an
way with a variable response latency from one stimuli to the atypical manner. A second possible disorder in which F-waves
next. This is clearly a physiologic disadvantage in attempting to may be of assistance in alerting the clinician to a particular dis-
localize a lesion to a focal region of the nervous system, partic- ease is multifocal motor neuropathy with conduction block. 171
ularly in mild disease. This so-called disadvantage, however, Again, it should not be concluded that an absence or reduced
can be used to an advantage in some disorders. In our opinion, number of F-waves is diagnostic of this disease, but rather that-
the very fact that the F-wave is traversing the peripheral nerve many practitioners do not routinely study the peripheral nervous
twice can be used to a diagnostic advantage in detecting an early system from the root to the axilla in all patients presenting with
disease process that is diffusely distributed along the nerve and limb weakness. Conduction studies of the arm and forearm may
may not be detected by assessing a focal neural segment. One be normal, whereas the F-waves may be reduced in number or
such disease entity is diabetic neuropathy. Although appro-
5164
absent. This finding should suggest that the region between the
priate reservations have been raised regarding the true sensitiv- root and axilla should be studied. No doubt fibrillation poten-
ity of F-waves in diagnosing early peripheral nerve disease, 232
tials may be detected in distal muscles, but the F-wave can help
there is a sound physiologic basis for considering the use of F- direct a more "focal" exam of the relatively proximal neural
waves in attempting to define if there is a generalized mild segments. Having said the above, not all proximal lesions are
particularly amenable to F-wave studies. For example, cervical
and lumbosacral radiculopathies certainly can result in abnor-
mal F-wave studies; - however, most of these patients have
222 237
more localizing findings on needle electromyography. This is

understandable since the majority of muscles are innervated by
more than one root and most radiculopathies do not produce
complete obliteration of all the nerve fibers in a nerve root. It is
not suprising, therefore, that F-waves are usually abnormal in
radiculopathies only when significant unilevel or multilevel root
disease is present. Although F-waves can be used to study focal
demyelinating lesions, it is not our contention that focal en-
64
trapment neuropathies such as carpal tunnel syndrome should

be routinely assessed by F-wave studies. As always, it is in-
159
cumbant upon the practitioners to become familiar with a par-

ticular technique and its available literature, and then assess
whether the technique is of value in their patient population.
H-REFLEX
A stimulus applied to the tibial nerve with a magnitude that is
subthreshold for a direct motor response usually produces a late
IOOO^V response when recording from the gastrocnemius-soleus mus-
cles in the neighborhood of 30 ms (Fig. 6-12). This potential was
first described, by Hoffmann in 1918. Magledary and 95
McDougal performed in-depth electrophysiologic investigations

of this late response by the 1950s and they designated this poten-
tial the H-reflex in honor of Hoffmann. 145
In studying both the
146
Figure 6-12. H-reflex.An H-reflex evoked from the gastrocnemius- H-reflex and the F-wave, Magledary and McDougal defined a
soleus muscle following tibial nerve stimulation. Note that as the cur- number of ways to distinguish between these two responses with
rent intensity of t h e stimulus is slowly increased (lower t o upper similar latencies (Table 6-7). A number of clinical applications
trace) the magnitude of the H-reflex increases to a maximum (third have been developed using the H-reflex. Prior to discussing how
trace from bottom) and then decreases with continued elevations in the H-reflex may be employed in the diagnosis of potential
current strength. W i t h a supramaximal stimulus, the H-reflex is re- pathology affecting the peripheral and central nervous systems,
placed with an F-wave. it is necessary to discuss the physiology of the H-reflex.
Physiology of the H-Reflex Table 6-7. Distinguishing Characteristics of F-Wave

The H-reflex is believed to be a CMAP arising from an elec- vs. H-Reflex
trical afferent activation of a monosynaptic reflex a r c . The95145
F-Wave H-Reflex
afferent pathway of the H-reflex involves electrical activation of
Presumed response Motor neuron Monosynaptic reflex
the large la afferent nerve fibers originating from muscle. After
"backfiring" arc
entering the dorsal horn of the spinal cord, the la afferents
synapse with the alpha motor neurons innervating that muscle. Afferent path a m o t o r fibers la afferents
This afferent motor impulse traverses the motor nerves to result Efferent path a m o t o r fibers a m o t o r fibers
in a CMAP. The complete reflex arc, therefore, is mediated by Gastrocnemius-soleus
Muscle* All skeletal muscles
orthodromic sensory and orthodromic motor neural conduction.
Flexor carpi radialis
The H-reflex is most easily elicited by stimulating the tibial
Stimulus threshold Supramaximal Low
nerve at the popliteal fossa with a relatively long-duration stimulus
and an intensity that is subthreshold for motor nerve stimulation. (compared to maximal
Recordings are typically performed from the gastrocnemius- CMAP)
soleus muscles. The intensity of the current is initially set at zero. Morphology Variable Constant at low
As the stimulus intensity is slowly increased, the H-reflex is first stimulus rates
noted to appear with a small amplitude and duration approximat- Magnitude <5% Same o r smaller
ing 30 ms (Fig. 6-12). Continued elevation of the stimulus re-
145
(compared to maximal
sults in a progressively larger-amplitude H-reflex. The magnitude CMAP)
of the H-reflex usually peaks at or just prior to the observation of a
Jitter > CMAP; < H-reflex » F-wave
direct CMAP or M-response from the gastrocnemius-soleus mus-
cles. Further increases in the current intensity results in a continu- Response to increasing More persistent Disappears
ally increasing M-response but a steadily declining H-reflex stimulus intensity
amplitude. When the M-response approaches a maximum and its Response to agonist Slight increase Appears in muscles
amplitude no longer increases, the H-reflex is usually replaced by not displaying H-
an F-wave. reflex at rest
The recommendation of a stimulus duration between 0.5 ms * Refers to muscle at rest where response can usually be elicited. Modified from
and 1.0 ms is made because the relatively longer current dura- Lachman et al. and Magledary and McDougal.
130 145
tions are believed to preferentially activate the large sensory

compared to somewhat smaller motor f i b e r s . - Excitation
173174 233
of the large la afferent fibers is desired in order to initiate the potentials (EPSPs). The rise time of the EPSP approximates 3.6
reflex arc. An alternative explanation uses an anatomic location m s.23.24.36 j j necessary for multiple la afferents to each con-
s t s
of sensory compared to motor fiber within the nerve. Selective tribute an EPSP in a rather synchronous volley given the above-
activation of the anterior as opposed to posterior aspects of the noted rise time to sufficiently depolarize the motor neuron to
tibial nerve in both the human and cat reveal that the motor threshold to generate an efferent motor impulse. Single-fiber
fibers are primarily located anteriorly. Additionally, the motor
153
electromyography studies reveal that the amount of variability
and sensory fibers are found to be approximately of the same between successive H-reflex responses is rather large and re-
size and should have the same threshold of activation. Sensory quires an interposed synapse in addition to the neuromuscular
fibers are activated first because they are more superficial and junction. - - The amount of afferent stimulus required to
107 228 229
located closer to the cathode. Although this may be partially regenerate the reflex contraction of the homonymous muscle (H-
sponsible for activation of the sensory fibers prior to motor reflex) is dependent upon several factors.
fibers, it does not address the capability of comparatively Slowly increasing the magnitude of the electrical stimulus re-
longer-duration impulses at the same location to selectively ac- sults in a progressively larger H-reflex (Fig. 6-12). This is the
tivate sensory before motor fibers. The most likely explanation result of activating more la afferents with each impulse, thereby
is probably a combination of lower threshold levels in sensory recruiting more alpha motor neurons. This process continues
fibers to longer current durations and the more posterior or su- until the H-reflex begins to decline. The maximum H-reflex am-
perficial position of sensory fibers in the tibial nerve at the plitude may be compared to the maximum CMAP response ob-
popliteal fossa. Following activation of the large la afferents at tained by direct excitation of the peripheral nerve to estimate
the knee, the impulse propagates superiorly to enter the dorsal the percentage of the motor neuron pool activated through the
horn of the spinal cord to synapse with the alpha motor neurons. reflex response. It has been determined that 2 4 - 1 0 0 % of the
As previously stated, the H-reflex is believed to be a monosy- motor neuron pool may participate in the H-reflex. - This 207 208
naptic reflex arc. This belief is founded upon several experi- suggests that the H-reflex amplitude is quite variable and sub-
mental observations. Sectioning of the dorsal root permanently ject to multiple factors. The percentage of motor neurons con-
obliterates any trace of the H-reflex. Investigations regarding
153
tributing to the H-reflex response can be increased by various
the time necessary to conduct through the possible central con- degrees of mild voluntary contraction of the muscle under in-
nections between the dorsal and ventral root revealed only vestigation. The suprasegmental facilitatory influences can alter
enough time for one synapse, i.e., between 0.5 and 1.0 the magnitude of the H-reflex by lowering the alpha motor
m s136-138.145 y^g | afferents that synapse with a homonymous
a
neuron's threshold, thus making it easier for the la afferents to
alpha motor neuron tend to depolarize the soma's resting mem- recruit more alpha motor neurons at lower thresholds. A possi- 1
brane potential through a transmitter substance by increasing ble mechanism may be that the suprasegmental drive for mild
the permeability of small cations such as N a and K . The re-
+ +
voluntary contraction depolarizes the motor neuron's soma rest-
lease of a transmitter with the capability of depolarizing the ing membrane potential closer to threshold. When the la afferent
soma's membrane is said to generate an excitatory postsynaptic generates its EPSP, fewer are required to facilitate attainment of
the threshold level. Additionally, there may be motor neurons facilitatory and inhibitory influences. The progressive decline
that were not previously depolarized that can now fire because of the H-reflex and replacement by an F-wave at sequentially
less EPSP summation is required. The suprasegmental influence greater current intensities is likely a combination of collision,
of central facilitation introduces a potentially significant vari- refractory alpha motor neurons, and Renshaw inhibition of the
able when attempting to quantitate the H-reflex amplitude for motor neuron pool (6-12). - 234 235
diagnostic purposes. Each individual possesses his or her own

level of central nervous system activity with respect to the nec- Factors Affecting t h e H-reflex
essary amount of transmitter directed EPSPs required to poten- As previously stated, the magnitude of the H-reflex is subject
tiate alpha motor neuron discharge. to the amount of current delivered to the nerve, ratio of la affer-
The progressive reduction and eventual disappearance of the ents to motor fibers activated, interpersonal differences in
H-reflex with continued elevation in stimulus strength is poorly suprasegmental facilitatory influences, and level of muscle con-
understood, but may be a result of several factors. A mechanism traction. The H-reflex is commonly observed only in the gas-
183
initially proposed relies upon the H-reflex impulses colliding trocnemius-soleus and flexor carpi radialis muscles. The
with the antidromic action potentials propagating in the motor detection of H-reflexes in these two muscles may not be ob-
nerves just distal to the alpha motor neurons. - Remember
145 239
served in all normal individuals, particularly in the elderly.
that action potentials are generated at the same location in both Contraction of the muscle from which the H-reflex is recorded
sensory and motor fibers, e.g., the popliteal fossa for an H- and its agonists may allow an H-reflex to be recorded from
reflex recorded from the gastrocnemius-soleus muscle. If the H- where it was previously absent. Again, this is because of facili-
reflex impulses are to collide with the antidromic motor action tatory influences decreasing the difference between the alpha
potentials somewhere at the level of the ventral roots or distally, motor neuron's resting membrane and threshold levels. Muscle
sensory action potentials obviously have to conduct signifi- contraction is found to not alter the latency of the response sig-
cantly faster than the motor impulses in order to traverse and nificantly. Contracting the antagonist muscles tends to sup-
25
exit the central nervous system. Recall that about 4 ms is re- press the appearance of the H-reflex. - - - It is also possible 77 78 95 153
quired for the EPSP to reach its peak plus an additional 1 ms for through contraction of a muscle to observe H-reflexes where
synaptic transmission. A difference of about 5 ms is the required they are not routinely detected. For example, through muscu-
172
separation between the sensory and motor conduction times. lar contraction, H-reflexes have been detected in the tibialis an-
Histologic examination of the peripheral nerves concerned has terior, abductor pollicis brevis, extensor digitorum communis,
revealed similar diameters for both motor and sensory nerves foot intrinsic muscles, and flexor digitorum profundus. - - - 25 54 70 174
suggesting that they have similar conduction velocities. 153

Patients with spasticity resulting from various upper motor
Despite the well-accepted opinion that sensory fibers generally neuron lesions may produce H-reflexes in muscles other than
conduct faster than motor fibers, this has not been found in care- the soleus m u s c l e s . - Also, in newborns, an H-reflex may i
131146 215
ful investigation of afferent and efferent conduction veloci- often be seen in the small muscles of the hands and feet, and
ties. - - - - Although it may be possible for fast-conducting
40 41 52 153 217
usually disappears by 1 year of age. - - 13 94 216
la afferents that have activated the lower threshold and rela- In addition to contraction of the agonist muscles, it is also,
tively smaller slower-conducting alpha motor neurons to collide possible to facilitate the H-reflex with a Jendrassik maneuver
at the ventral root region, this may not be the complete explana- by asking the patient to forcefully make a fist. This is signifi-
tion for all motor neurons. The fast-conducting antidromic cantly less effective in facilitating an H-reflex than a contraction
motor potentials reach the ventral root at essentially the same of the agonist muscles. Significant potentiation of the H-
153
time the afferent impulses arrive at the dorsal root. Additionally, reflex can occur with post-tetanic stimulation. It is possible to 135
by strongly contracting the muscle, an H-reflex reappears degenerate an H-reflex with a subthreshold response after a
spite a strong stimulus, thereby proving collision is not a major tetanus of 100-500 Hz for 20 seconds lasting about 10-40 sec-
component of H-reflex suppression. Therefore another explana- onds. The mechanism is unclear but is believed to involve
tion of this phenomenon is necessary. presynaptic facilitation. 83
Let us assume that a suprathreshold stimulus is delivered to H-reflex amplitude may be affected by factors other than an-
the tibial nerve in the popliteal fossa, thus activating the large la tagonist muscle contraction. Forceful active or passive ankle
172
afferents as well as the large motor fibers conducting at similar flexion and contraction also can markedly reduce the tibial I
velocities. In this case, the alpha motor neurons are activated by nerve's H-reflex m a g n i t u d e . Mild passive stretching of the
53153
the antidromic impulse as discussed previously for the F- gastrocnemius-soleus muscle can either facilitate or inhibit the
wave. Additionally, the recurrent collaterals mediate Renshaw
112
H-reflex. - Vibration is an effective way to suppress the H-
214 236
cell inhibition of the alpha motor neuron pool. By the time the reflex. - - Applying a vibrating stimulus to the Achilles
9 20 1 5 231
EPSP and synaptic transmission of H-reflex impulses have con- tendon in the limb under investigation results in depression of
verged on the alpha motor neuron, it is most likely no longer ca- the H-reflex that may outlast the duration of the vibration by
pable of depolarizing the previously activated anterior horn several hundred milliseconds. The mechanism of H-reflex 205
cells at lower stimulus levels because they are either in a refrac- suppression is unknown but may involve presynaptic inhibition
tory state from F-wave generation or under the influence of through primary spindle afferent firing or neurotransmitter de-
Renshaw inhibition. It is possible to detect an H-reflex follow- pletion. - It is also possible to suppress the H-reflex with
39 214
ing inhibition through a supramaximal stimulus if the patient stimuli delivered at several Hz. 25
contracts the agonist muscles. This finding strongly argues

against collision as a reason for obliteration of the H-reflex. Diagnostic H-Reflex Techniques
Muscle contraction should not affect whether antidromic im- There are two major clinical applications of the H-reflex.
pulses collide with the orthodromic H-reflex. It appears that First, the H-reflex may be used to evaluate the status of the pe-
the major component of H-reflex suppression with supramaxi- ripheral nervous system with respect to proximal peripheral
mal stimulus delivery is based on the balance between central nerve conduction and potential entrapment of nerve roots, e.g.,
radiculopathies. Secondly, it is possible to examine the intricate Some investigators prefer to use a needle cathode, requiring less
interactions of segmental and suprasegmental facilitatory and stimulus. In this instance, a large electrode is secured to the patella
inhibitory influences on the two-neuron reflex arc designated to serve as the anode. Using a similar pulse width to that noted
the H-reflex. The H-reflex, therefore, can help us assess both above should not damage the nerve as a subthreshold stimulus is
the peripheral and central nervous systems in both health and used. It is our experience that should surface stimulation fail to
disease. elicit an H-reflex, needle excitation often results in a demonstrable
response. Further study, however, is required comparing the opti-
Peripheral Nervous S y s t e m Applications mal stimulus parameters for needle and surface stimulation.
Perhaps the most common use of the H-reflex is to assess the As stated above, the stimulus is slowly increased until a re-
conduction properties of the S1 nerve root in the neural foramen sponse with a latency approximating 30 ms is detected. The cur-
region to investigate the possibility of nerve root compromise. rent level is increased in small increments until a motor
A number of studies have confirmed that the primary nerve root response is just noted. One should then optimize the H-reflex by
level mediating the H-reflex is SI and although L 5 may con- decreasing or minimally increasing the current intensity until
tribute to the neural supply of the gastrocnemius-soleus muscles, it the H-reflex magnitude is maximized. Several responses are ob-
does not significantly participate in the H-reflex. ' ' - ' - ' -
, 2 3 184 ,9, 192 m226 227
served at this stimulus level to ensure a reproducible and stable
The H-reflex from the lower limb is also used to assess the con- response. The latency is then recorded to the initial departure of
duction of proximal afferent nerves not accessible to routine the H-reflex from the baseline. This is typically although not
evaluation. Specifically, the H-reflex may demonstrate compro- always a positive deflection, most likely because the electrode
mise of the proximal afferent pathways in various polyneu- is preferentially detecting the response from the soleus muscle
ropathies and allow one to calculate conduction velocities in and not over its motor point. A characteristic appearance of the
these regions. - - ' ' One also may examine the H-reflex to
80 86 224 225 238
H-reflex is a triphasic initially positive potential (Fig. 6-12). It
the flexor ca~pi radialis (FCR) muscle for upper limb peripheral is not unusual, however, to observe an initially negative H-
nerve lesions. It is recommended to adopt a standardized reflex. In either case, the initial baseline departure denotes the
method of applying electrodes in order to minimize potential correct latency measurement. An H-reflex demonstrates a very
errors introduced by variability of anatomic landmarks. stable onset latency from one response to the next. This is quite
different than the rather variable F-wave latency with each con-
H-Reflex: Gastrocnemius-Soleus Muscle secutive stimulus. Continued elevation of the current strength
Recording Electrodes. Surface recording electrodes are characteristically results in the disappearance of the H-reflex
preferred for this technique as they were used for the develop- and the appearance of an F-wave (Fig. 6-12). The H-reflex
ment of the accompanying reference data. Because amplitudes should not be recorded if it is smaller than the M-response or
are not used, needle recordings to document H-reflex latencies demonstrates a variable latency and morphology.
are acceptable provided the same technique is used for both left
and right measurements.
The patient is positioned comfortably in the prone position 50-r- E40.I4 •80
49"
with the feet off the edge of the plinth. It is often helpful to
48-¬ -39
place a pillow beneath the legs to cause slight knee flexion. 47-¬ •75
-38
Both left and right H-reflexes can be obtained in this position 46" -70
with little difficulty. It is necessary to record the H-reflexes 45 — -37
from both lower limbs as left/right latency comparisons often 44-¬ -36 -J-65
43"
are quite helpful.
42" -35
E-L The E-l recording electrode is located by first flexing
41 — -34 •60
the leg and drawing a line across the popliteal fossa. - A line 40-¬
19 20
connecting the mid-popliteal fossa with the proximal flare of 39" -33 •55
the medial malleolus is bisected for the E-l electrode location. 38" H-32
This site typically approximates the musculotendinous junction 37 — •31 -f-50
of the gastrocnemius muscle. 36 —
35 — -30 •45
E-2. An E-2 electrode is secured to the distal portion of the
34 — -29
Achilles tendon just proximal to its insertion on the calcaneus. 33" -40
Ground. A ground electrode is positioned just proximal to 32 — -28
E-l. 31 — -27 •35
Stimulation. The cathode is placed in the mid-popliteal fossa 30"
29 — •26
with the anode distal. A pulse duration of between 0.5 ms and 1.0
28 — -25 -^-30
ms is recommended. The current intensity is slowly increased
27-¬ -24
until the stimulus just activates the large la afferent fibers without 26 — •25
concomitant activation of the motor fibers or is just threshold for
25 J •22.64
only a few motor fibers. The stimuli should be delivered at a rate LATENCY AGE
•20
LEG LENGTH«
of 1 stimulation every 2 - 3 seconds to avoid suppressing the H- (cm) (msec) (yrs)
reflex through central mechanisms. It is often necessary to move
the stimulating electrodes either slightly medially or laterally to Figure 6 - / 3 . H - r e f l e x n o m o g r a m . A nomogram for predicting t h e
optimize the cathode directly over the tibial nerve. Care should be H-reflex provided the patient's age and leg length (see t e x t for proper
taken to avoid proceeding too far laterally as the peroneal nerve measurement) are known. (From Braddom RL, Johnson EW:
may be excited. This is relatively easy to define as the foot no Standardization of H reflex and diagnostic use in SI radiculopathy.Arch
longer plantarflexes but dorsiflexes with each stimulus. Phys Med Rehabil 1974;55:161 - 1 6 6 , with permission.)
Instrumentation Parameters. A sweep speed of 10 ms/div right responses) smaller than 0 . 4 or 0.5 is considered an ab-
108 89
is optimal for lower limb H-reflex examinations. Although a normal finding and suggestive of possible pathology affecting
sweep speed of 5 ms/div can be used, it is possible for this re- the fibers conveying the response. Another study found an am-
sponse to be greater than 50 ms in a particularly tall individual plitude ratio (right side/left side) of greater than 1.8 (unaffected
or in a person with a peripheral neuropathy. An amplifier sensi- side/affected side) to indicate an abnormality. In our opinion,
163
tivity of 200 pV/div to 500 pV/div usually suffices for most re- a shortcoming of these studies is a failure to fully explore the
sponses. Filter settings commensurate with routine motor effect of central facilitation on the "normal" side-to-side ampli-
studies are recommended. tude range. Specifically, in persons with a considerable side-to-
Reference Values. The technique of Braddom and Johnson side difference no apparent attempt was made to facilitate the
is suggested as it represents a standardized method for obtain- response and see if the side-to-side difference would diminish.
ing H-reflex latencies. - H-reflexes are found to correlate
19 20
This is critical since patients with back pain and no radicular
highly with both age and leg length. A regression equation may axonal/demyelinating lesion may splint the painful side and in-
be used to predict the optimal H-reflex latency: H-reflex (ms) = advertantly contract the affected side's foot dorsiflexors ever so
9.14 + 0.46(leg length-cm) + O.l(age-yrs). A nomogram based slightly, resulting in diminished H-relfex a m p l i t u d e . As69156
upon this equation also may be used in the clinical setting as a noted above, the effect of antagonist muscles on the H-reflex is
quick reference (Fig. 6-13). When using the nomogram, a line well documented. Therefore, side-to-side amplitude differences
connecting the patient's age and leg length is constructed and for the H-reflex may be of diagnostic use, but the presently
where it crosses the middle set of values predicts the H-reflex available studies are unconvincing regarding the most appropri-
latency. A mean H-reflex latency of 29.8 ± 2.74 ms is found for ate amplitude values to use clinically. Obviously, any side-to-
a normal population. In this study a side-to-side difference of side amplitude comparisons are of questionable validity in
1.5 ms is predictive of an SI radiculopathy. Some investigators bilateral disease.
use as little as a 1.0 ms difference. In persons aged 60-88 years, When using left/right criteria as sensitive as 1.1 m s , it is
8 9 1 6 3
an H-reflex can be obtained in up to 92% of persons with a side- advisable to always use standardized distances to assist in re-
to-side latency difference of 1.8 m s . The normal side-to-side
57
producibility. Using anatomic landmarks may result in undue
amplitude differences in persons between 21 and 67 years can latency differences and predispose one to false-positive or false-
reach 60% (see Additional Comments for further discussion). 108
negative results.
Additional Comments. Should one have difficulty eliciting Just as for F-waves, it is possible to calculate the H-reflex con-
an H-reflex, slight voluntary contraction of the muscle exam- duction velocity for the proximal tibial nerve segment. A maxi-
ined may facilitate its detection. Although this should not sig- mal H-reflex is obtained using the above noted technique. One
nificantly affect the latency, it may be advisable to also use then elicits a maximal M-response for supramaximal excitation
slight voluntary contraction on the contralateral side, particu- of the tibial nerve from the same site of stimulation used for the ,
larly if the left/right differences approach significance. H-reflex, i.e., the anode is rotated proximally. A time of 1 ms is
Caution should be exercised in attempting to use the H-reflex subtracted to account for central synaptic delay. The distance
amplitude for diagnostic purposes as it is highly variable and measured is from the popliteal fossa stimulation site to the T l 1
subject to central nervous system influences. The amplitude is spinous process. The formula used to calculate this proximal con-
also quite sensitive to electrode placement and may noticeably duction over afferent sensory and efferent motor fibers is: H-
change within just a few centimeters. - If side-to-side ampli-
147 148
reflex CV (m/s) = (distance popliteal fossa to Tl 1 x 2)/(H-reflex
tudes are to be compared, it is very important to exactly repro- latency - M latency - 1 ms). This technique can document slow-
duce the electrodes' locations since different electrode positions ing of proximal conduction velocities in various peripheral neu-
can result in considerable amplitude variations. As noted 27
ropathies, such as diabetic polyneuropathy - and uremic
225 226
above, several investigations have attempted to define the neuropathy. Of course, the same precautions regarding distance
86
degree of side-to-side amplitude variation in reference popula- measurements noted for the F-wave also apply for the H-reflex.
tions. When comparing the smaller of the two responses to the
larger, an amplitude ratio (smallest/largest between left and H-Reflex: SI Central Loop
The traditionally performed H-reflex to the lower limb uses a
very long pathway reducing its ability to localize a lesion strictly |
to the SI nerve root. As result, an attempt has been made to
reduce the pathway over which an H-reflex can be obtained in
the hopes of localizing a lesion to the SI nerve root. This tech-
175
nique is promising, but requires larger studies to fully assess the

sensitivity and specificity of the SI central loop latency.
Recording Electrodes. The recording and ground elec-
trodes are positioned in the same manner as for the H-reflex ob-
tained through tibial nerve stimulation (see above).
Stimulation. A monopolar needle electrode serves as the
cathode and is inserted 1 cm medial to the posterior superior
iliac spine perpendicular to the patient's frontal plane, and a sur-
Figure 6-14. C e n t r a l SI l o o p l a t e n c y . T h e two peaks of the direct face anode is affixed to the anterior superior iliac spine. The
motor (M) response and the H-reflex (H) following SI nerve root stim- cathode is gently inserted until it contacts the sacrum and is then
ulation are shown. The difference between the M and H peak latencies withdrawn slightly. A stimulus pulse duration of 1 ms is used
should be less than 8.0 ms. (From Pease W S , Kozakiewicz R, Johnson with a maximal pulse delivery of 0.5 Hz. The current is slowly
EW: Central loop of the H reflex: Normal value and use in SI radicu- increased until a combined direct motor (M) and indirect H-
lopathy. Am J Phys Med Rehabil 1997;76:182-184, with permission). reflex (H) is obtained.
Instrument Parameters. The instrument's sweep speed is Reference Values. Onset H-reflex latencies to the rectus
set to 5 ms/div with high- and low-frequency filter settings of 10 femoris, vastus medialis, and vastus lateralis muscles are de-
kHz and 20 Hz. A gain of 0.5 to 1.0 mV/div is used to best visu- scribed as 17.7 ± 1.8 ms, 18.4 ± 1.8 ms and 18.1 ± 1.7 ms, re-
alize the response. spectively. It is important to note that this response may not
109
Reference Values. An H to M latency difference of 7 ± 0.3 be obtainable even in normal individuals. Slight voluntary con-
ms is anticipated in healthy individuals (Fig. 6-14). A latency of traction of the quadriceps and a Jendrasik maneuver may assist
8.0 ms or greater is considered indicative of a lesion affecting in the facilitation of the response.
the proximal S1 conducting pathway.
Central Nervous S y s t e m Applications
H-Reflex: Flexor Carpi Radialis Because a portion of the H-reflex involves the central ner-
Recording Electrodes. It is possible to record an H-reflex vous system, it is subject to both segmental and suprasegmen-
from the FCR in most normal individuals. This muscle is inner- tal influences. - -
6 58
This is best demonstrated by facilitation
71135
vated by C6 and C7 nerve roots and thus the H-reflex may be of of an H-reflex with contraction of the muscle under investiga-
assistance in assessing the neurophysiologic status of these two tion and inhibition with antagonist muscle contraction. It
nerve roots and possibly conduction through the brachial should be possible, therefore, to indirectly investigate various
plexus. - - -
38 167 196 197
aspects of the central nervous system with reflex responses by
E-l. A surface E-l electrode is positioned over the belly of studying the H-reflex in both health and d i s e a s e . 149158
One
the FCR. This site is located one-third the distance from the common method of using the H-reflex for this purpose involves
medial epicondyle to the radial styloid. 106
a technique of conditioning and test stimuli similar to refrac-
E-2. An E-2 is positioned over the brachioradialis muscle. tory period experiments.
Ground. The ground should be secured to the skin just prox- The central motor neuron pool excitability can be investi-
imal to E-l. gated by using a dual stimulation technique to document the H-
Stimulation. The subject may be either supine or sitting reflex excitability or recovery curve. The recording electrodes
with the elbow slightly flexed. The cathode is located over the are positioned as noted above for peripheral nerve techniques. A
median nerve at the antecubital fossa with the anode distal. A stimulator is secured to the tibial nerve at the popliteal fossa,
pulse width of 0.5-1.0 ms is optimal and the current intensity is which has the capability of delivering square wave stimuli 1.0
slowly increased until the H-reflex is maximized with an absent ms in duration with a variable interval between successive im-
or minimally present FCR CMAP. The presence of an H-reflex pulses. The initial or conditioning stimulus is delivered at or just
should be verified by increasing the stimulus intensity and ob- below threshold for eliciting an H-reflex. This impulse gener-
serving for a disappearance of the response and replacement by ates a number of la afferent action potentials that enter the cen-
an F-wave. The stimuli delivery should not exceed 0.5 Hz. tral nervous system over the reflex arc to condition the alpha
Slight contraction of the FCR may be necessary to detect an H- motor neuron pool without causing a motor impulse. A second
reflex in some individuals. or test stimulus is then given through the same stimulating elec-
Instrumentation Parameters. A sweep speed of 5 ms/div trodes at a level sufficient to evoke a minimal direct M-re-
with a sensitivity of 0.5-1.0 mV/div and routine motor conduc- sponse. Of course, this neural activation also should yield
tion filters should produce clearly recognizable H-reflexes. enough current to activate a large number of la afferents with
Reference Values. One may anticipate a latency to the ini- the capability of generating an H-reflex independent of the con-
tial baseline deflection of 15.9 ± 1.5 m s . A side-to-side differ-
106
ditioning impulse. The amplitude of the H-reflex produced by
ence of 0.4 ± 0.3 ms is expected. the test stimulus at increasing time intervals from the condition-
ing stimulus is then plotted for each sequential time interval and
H-Reflex: Quadriceps Muscle describes a characteristic shape reflecting central nervous
Recording Electrodes. Recording the H-reflex from the system segmental and suprasegmental interactions.
quadriceps muscle is somewhat more challenging than either Within a few milliseconds of the conditioning response, the
the gastrocnemius-soleus or FCR muscles. The femoral nerve is H-reflex magnitude is maximal (Fig. 6-15). Increasing the
209
a rather deep structure and difficult to activate with surface time interval between conditioning and test stimulus results in a
stimulation. Because of this, the current intensity is hard to in- progressive amplitude decline of the H-reflex to a minimum
crementally deliver and there is little gradation between an value at approximately 75 ms. At interstimulus intervals between
absent and present direct motor response. The utility of a 100 and 200 ms the amplitude of the H-reflex again increases,
quadriceps H-reflex may be of assistance in the L3/L4 nerve peaking at about 150-200 ms. As the time interval between the
root compromise. 3162
conditioning and test stimulus continues to increase, the H-reflex
E-l. An E-l electrode may be located over the main muscle amplitude demonstrates a second decline, reaching a minimum
bulk of the vastus medialis, vastus lateralis, or rectus femoris. value at 200-400 ms. The H-reflex amplitude reveals a slow and
E-2. The patella is a convenient site for E-2. progressive increase as the interstimulus time delay approaches
Ground. This electrode should be situated just proximal 4o 1900 ms. This H-reflex recovery curve is a graphic representa-
E-l. tion of the various central nervous system interactions all con-
Stimulation. The femoral nerve is excited in the inguinal verging upon the alpha motor neuron pool and reflects the CNS's
region just distal to the inguinal ligament about the region of the normal physiologic state. Although the exact mechanism gener-
femoral artery. Cathodal placement is distal to the anode. A ating the above noted curve is unknown, an assumption based
pulse duration of 0.5-1.0 ms is delivered at less than 0.5 Hz upon what little is known about the peripheral/central nervous
with an intensity capable of producing an H-reflex with little, if system provides some insight into the various interactions result-
any, M-response. ing in the detected H-reflex recovery curve.
Instrumentation Parameters. The same instrumentation The initial H-reflex response is believed to result from the
setup previously described for the FCR muscle can be used. conditioning stimulus' EPSPs generated by the la afferent input.
slightly depolarize the motor neuron pool when the condition-

ing stimulus of 200 ms reaches the anterior horn cells. Once the
time course of the long-loop reflex EPSPs has been exceeded,
the neurotransmitter depletion effect of the direct la afferent
pathway again becomes manifest. The H-reflex then does not
recover until the neurotransmitter from the initially depleted fa-
cilitory inputs have been replenished, i.e., greater than 1 second.
This technique has been used to investigate the recovery curves
in normal newborns and patients with Parkinson's disease,
spinal cord injury, and dystonia. Continued work is neces-
155J6S
sary to clearly elucidate the pathways involved in the H-reflex

covery curve and its potential clinical applications.
Tendon Reflex
It is possible to record the electrical activity associated with
nl v
0 100 200 300 400 500 600 700 the reflex contraction of a muscle induced by acutely stretching
msec a tendon through percussion. A modified reflex hammer is re-
quired with the ability to initiate the instrument's cathode ray
Figure 6 - / 5 . Postulated pathways mediating the H-reflex's tube sweep. Surface or needle electrodes can then record the
magnitude as elicited with conditioning stimuli. A, Solid line signifies reflex contraction of the muscle following the mechanical
that 2 of 4 la afferent fibers were excited that in turn monosynaptically energy delivered to the muscle's tendon. In brief, the tendon tap
synapse with a m o t o r neuron (MN) but w e r e subthreshold for evok- stretches its muscle, which in turn activates the muscle spindle.
ing a response.This same la volley simultaneously proceeded rostral- The muscle spindle is a specialized structure consisting of a
ward up t h e dorsal column tracts (DSCT) t o t h e various structures connective tissue capsule surrounding several types of muscle
depicted in the circle.The vestibulospinal t r a c t (VST) then conveys fibers referred to as intrafusal muscle f i b e r s . The intrafusal
I,J9S
these volleys t o activate the m o t o r n e u r o n , forming t h e long-loop muscle fibers are approximately 15-30 pm in diameter and 4-7
reflex. B,The continuous line reveals the time course of the condition- mm long as opposed to the commonly thought-of muscle tissue,
ing of the observed H-reflex by the subthreshold H-reflex stimulus. It extrafusal muscle fibers, which are 50-100 pm in diameter and
is suggested that the H-reflex amplitude curve is a composite of three several millimeters to many centimeters in length. Sensory in-
conditioning factors, i.e., (I) local EPSP (see text) on motoneuron, (2) nervation to the intrafusal muscle fibers is provided by one
long-loop facilitation as shown above, and (3) depletion of transmitter larger myelinated la nerve that wraps around the center of the i
in the synapses activated by the conditioning stimulus. (From muscle fibers several times and is called the annulospiral
Taborikova H, Sax DS: Conditioning of H-reflexes by a preceding sub- ending or primary sensory ending. Stretching of the extrafusal
threshold H-reflex stimuls. Brain 1969;92:203-212, with permission.) muscle leads to a concomitant lengthening of the intrafusal •
fibers, which in turn activates the annulospiral ending sending
Specifically, the effects of the conditioning stimulus cause the impulses toward the central nervous system along the la affer-
la afferent-induced EPSPs to occur that have a rise time of sev- ents. The la afferents cause EPSPs to be produced in the
eral milliseconds. The duration of the rise time produces a facil- homonymous alpha motor neurons, resulting in a contraction of
itatory effect that is still present when the test stimulus' la the muscle stretched. The intensity of muscle contraction de-
afferents arrive. They find the alpha motor neuron pool already pends upon the number and degree to which the annulospiral
in a slightly depolarized state, thereby facilitating an H-reflex endings have been activated. The above description is the clas-
from the test stimulus. It is important to realize, however, that sic reflex arc familiar to all. Intrafusal muscle fibers also are in-
the subthreshold conditioning stimulus has been depleting a la nervated by a second class of sensory fibers, group II fibers,
afferent subpopulation of neurotransmitter substance. Experi- which form secondary muscle spindle endings or flower spray
mental results have demonstrated that the subthreshold stimulus endings. Because they do not directly participate in the reflex
activates about 50% of the la afferents, hence accounting for the arc the details of their function are not discussed. The intrafusal i
50% reduction in the H-reflex by 75 m s . Once the stimulus
209
muscle fibers also possess a motor innervation by small anterior
interval has increased beyond the facilitatory effect of the la af- horn cells referred to as gamma motor neurons. Unlike the
ferents, the full effect of the transmitter depletion becomes de- alpha motor neurons with peripheral nerve fibers in the range of
tectable, reaching a maximum at about 75 ms. Despite the 12-21 pm, the gamma efferent diameters approximate 2-8 pm.
above-noted depletion of neurotransmitter, a second elevation in Activation of the gamma fibers also can cause the annulospiral
the H-reflex is noted peaking at about 200 ms. A long-loop fa- ending to contract, thus potentiating a contraction of the extra-
cilitatory pathway traversing up to and back down the brainstem fusal muscle fibers. The gamma fibers, therefore, exert a power-
has been proposed to account for this increase in H-reflex mag- ful influence on the output of the muscle spindle, which
nitude.(Fig. 6-15) Some portion of the la afferent input to the modulates the type of contraction resulting from a tendon tap.
central nervous system not only directly synapses with the The electrical activity associated with muscle contraction
motor neuron pool, but also ascends through the dorsal spinal arising from a tendon tap can be recorded and displays various
cerebellar tract to involve the cerebellum and reticular sub- latencies depending upon the length of the afferent and efferent
stance. From these structures descending fibers may be con- pathway. The impulse traverses the la afferents and alpha motor
veyed through the vestibulospinal tract to also facilitate firing of neurons noted above. The H-reflex has been considered the
the motor neuron pool. It is proposed that the long loop reflex electrical equivalent of the tendon reflex. An important distinc-
time course coincides with and accounts for the second elevation between the H-reflex and the muscular contraction of a ten-
tion in the H-reflex magnitude because it provides EPSPs that don tap is that the H-reflex directly activates the large la afferents
Tibial Nerve
Figure 6-/6. A x o n r e f l e x . A , Series of axon reflex r e -

sponses designated I, 2, and 3 demonstrating a c o n s t a n t la-
tency. N o t e that the axon reflex can either proceed o r follow
t h e F-wave. B, Diagrammatic explanation of t h e axon reflex
(see t e x t ) . (From Kimura J: Electrodiagnosis in Diseases of
N e r v e and Muscle: Principles and Practice. Philadelphia, RA.
Davis, 1989, with permission.)
in the nerve and bypasses the intrafusal muscle fibers. This is referred to as an axon reflex or A-wave. This intermediate
68
view, however, may be an oversimplification. The above-noted potential is usually elicited with a submaximal stimulus and dis-
"classic" description of the interaction between the la afferents plays a constant latency, unlike the varying F-wave. The pre-
and alpha/gamma motor system may be less well understood sumed physiology of the axon reflex involves some type of
than previously thought. Human studies reveal that the fusimo- neural damage with a collateral sprout from a proximal aspect
tor drive of the gamma system is not necessary to elicit a tendon of the nerve to the muscle. A submaximal stimulus distal to the
reflex. The ease with which a tendon reflex can be obtained may collateral's branch point activates a significant portion of the
be more related to the "central excitability" state of the alpha nerve but spares some fibers. The orthodromic impulses results
motor neurons than the gamma system's influence on the in a less than maximal C M A P while the antidromic impulse
muscle spindle. - - The muscle's response to a constant
21 22 24
proceeds proximally to encounter the branch point. A portion of
tendon tap varies, suggesting that the central nervous system's the electrical impulse proceeds distally along the collateral
segmental and suprasegmental influences are important and while the remainder of the impulse continues into the CNS. The
necessary factors affecting the reflex response. The H-reflex action potentials in the collateral sprout then activate a small
demonstrates the same independence of the fusimotor system portion of the muscle resulting in the axon reflex response. The
that the tendon reflex does. Additionally, the afferent volley in- major antidromic impulse backfires in the motor neuron pool to
duced by electrical stimulation for the H-reflex is a synchro- then yield the F-wave. Increasing the stimulus intensity elimi-
nized volley, whereas the tendon tap afferent volley is nates the axon reflex because the entire nerve is now depolarized,
considerably more dispersed in time. EPSP rise times from including the aberrant collateral branch, thus resulting in a syn-
tendon percussion are 8.3 ± 2.5 ms, whereas those from the H- chronous activation of the entire nerve. The antidromic impulses
reflex are 3.5-5.5 ms. The combination of dispersed action po- are induced in both the collateral branch and the main nerve to
tential volleys and longer rise times for the tendon tap reflex collide proximally, thus eliminating the delayed axon reflex
suggest that it may not travel the same central pathways to action potential. Rarely, one may note that the axon reflex per-
achieve the muscle contraction observed with the H-reflex. sists despite a supramaximal stimulus. In this instance the
There is sufficient time in tendon tap reflex activity, temporal mechanism for the axon reflex is obscure but may occur be-
dispersion of action potential volleys combined with relatively cause the branch generating the axon reflex is somehow isolated
long EPSPs, to traverse disynaptic and trisynaptic pathways from the impulse either through connective tissue, electrical
centrally. This also applies to a lesser degree to the H-reflex. "shielding" from muscle or bone, or some other poorly under-
These findings suggest that the classic acceptance of a monosy- stood reason. 125
naptic reflex arc for all fibers activated by either a tendon tap or It is also possible to localize the site of the collateral sprout
H-reflex traveling a single synapse may be inaccurate. - - - 21 22 24 231
producing the axon reflex by slowly moving the site of neural
A more plausible explanation suggests that the motor neurons depolarization to sequentially more proximal locations. As the
of lowest threshold may be activated by a monosynaptic arc site of neural activation moves more proximally, the latency of
through the fastest la afferents, whereas motor neurons of the CMAP increases, whereas the axon reflex latency decreases
higher threshold fire through several synapses initiated by the because the branch point is closer to the cathode. When the axon
fastest la afferents and single synapses of relatively slower la reflex disappears, the stimulus is now just proximal to the branch
afferents. This central combination of synapses and rate of af- point. Should the branch point depart the damaged nerve at a site
ferent conduction serve to yield a more synchronous output in- more distal than the lesion, localization is not possible. This is
dependent of the fusimotor system. rather obvious as more proximal stimulation does not result in an
absent axon reflex, indicating a more distal branching. In this
Axon Reflex case, one can determine the collateral sprout's conduction veloc-
A response may occasionally be observed with a constant la- ity. Of course, should an immature collateral be present, its
68
tency between that of the CMAP and the F-wave or exceeding conduction velocity can be considerably slower than that of the
the F-wave, particularly from the small muscle of the hand and main portion of the nerve. In this instance, it is possible for the
rarely from the intrinsic foot muscles (Fig. 6-16). The potential axon reflex to follow instead of precede the F-wave. The presence
of an axon reflex is simply a nonspecific indication that the nerve nerves influencing cutaneous blood vessels and sweat glands,
has most likely experienced some type of chronic insult that re- the technique is quite demanding and limited primarily to re-
sulted in a collateral sprout. An axon reflex can be seen in multi- search facilities. - - A simple method of evaluating sympa-
26 84 230
ple chronic neuropathies such as radiculopathies, plexopathies, thetic skin activity is the galvanic skin r e s p o n s e . Following
131178
peripheral neuropathies, and motor neuron disease. It is pos-

68193
an emotional or noxious stimulus the sudomotor activity medi-
sible to observe axon reflexes in some normal persons with rou- ated by the sympathetic nervous system results in an alteration
tine nerve conduction studies because of possible subclinical in the skin's resistance to an electrical current. It is possible to
nerve injury or some other ill-defined reason. Axon reflexes, use commonly available electrodiagnostic medicine equipment
however, are a common occurrence in normal individuals when to examine a similar response mediated by the sympathetic
examined with single-fiber electromyography. Caution should system, i.e., the sympathetic skin response.
be exercised when a late potential is observed. In addition to the Recording Electrodes. The sympathetic skin response is
possibility of detecting an F-wave, H-reflex, or axon reflex, there relatively easy to perform and should be tried by beginning
are a number of potentials following a stimulus that are none of practitioners. Commonly available surface recording elec-
82
the above. Additional "late" potentials can be seen with neural trodes are used. Only a few of the possible nerve techniques are
stimulation because of repetitive firing of the nerve trunk, slowly described but essentially any aspect of an limb may be used.
conducting poorly myelinated nerves, ephaptic conduction of in- E-l. For upper limb median nerve studies an E-l recording
jured nerves, intramuscular ephaptic conduction loops (complex electrode is secured to the palmar surface of the hand. In the 202
repetitive discharge), an axon loop, or other unexplained electri- lower limb, an E-1 electrode can be positioned to the dorsum of
cal pathways. ' -i9o.2oi
88
the foot for peroneal nerve stimulation and on the foot's plantar
An A-wave can be observed in many different types of lesion aspect for tibial nerve excitation.
including polyneuropathy, radiculopathy, motor neuron disease, E-2. In the upper limb, E-2 is located on the dorsum of the
Guillain-Barre syndrome, plexopathies, myopathies, and focal hand. For peroneal nerve stimulation, the plantar aspect of the
nerve injuries. It is rare, however, to observe an axon reflex re-
12
foot is used, whereas the foot's dorsum is appropriate for tibial
sponse in normal individuals with the exception of an occa- nerve activation.
sional A-wave detected in the foot intrinsic muscles innervated Ground. This electrode is positioned just proximal to the E-
by the tibial nerve. Therefore, documenting an A-wave in any 1 electrode with respect to the cathode's location.
response other than from the tibial nerve is considered indica- Stimulation. The stimulator is positioned in the routine
tive of pathology by some authors. 12
manner over the desired median nerve at the wrist and peroneal
and tibial nerves at the ankle. A stimulus is applied to each
Sympathetic Skin Response nerve with a pulse width of 0.1 ms and a current intensity ap-
All of the previously described nerve conduction techniques proximating 10-20 mA or enough to elicit a slightly uncomfort-
examine either sensory or motor nerve fibers or both. The pe- able sensation. Because the response readily habituates, the
180
ripheral nervous system, however, not only contains sensory stimuli are delivered at irregular intervals over several minutes.
and motor fibers, but autonomic nerves as well. The majority of There should be a pause of several seconds between successive
nerve conduction studies only evaluate the fastest-conducting stimuli. About 10 stimuli are recommended as each response
fibers and do not consider the more slowly conducting myeli- can vary somewhat and only responses that are consistent are
nated or unmyelinated fibers, e.g., the sympathetic fibers con- selected for analysis.
tained within the peripheral nerves. It is possible to investigate the conduction velocity of the
Initially, the autonomic nervous system could only be investi- fibers mediating the sympathetic skin response (Fig. 6-17). A
gated electrophysioiogically with the insertion of a fine needle second stimulus site at a more proximal location is selected
recording electrode into the substance of a peripheral autonomic once a satisfactory distal potential is obtained. The same para-
nerve and recording the ensuing electrical activity, i.e., mi- meters are used for proximal stimulation as described above.
croneurography. Although microneurography is capable
2684 2 2 0
Instrumentation Parameters. A sweep speed of 500
of distinguishing between the sympathetic fibers conveying im- ms/div is optimal as this response is mediated by slowly con-
pulses controlling intramuscular blood supply and sympathetic ducting C fibers and substantial time is required to resolve the
sympathetic skin response. Of paramount importance is the
bandpass selected for this technique. A low-frequency filter of
0.5 Hz or less is necessary as the recorded potential has signifi-
cant low frequency components. The high frequency filter
should approximate 2,000 Hz. An amplifier sensitivity of
200-1,000 pV/div usually suffices for most individuals.
Reference Values. The onset latencies for upper and lower
limb studies are 1.39 ± 0.07 seconds and 1.88 ± 0.11 seconds,
respectively. Conduction velocities in the upper limb approxi-
202
mate 1.57 ± 0.11 m/s and for the lower limb are 1.02 ± 0.07 m/s.
1000/xV Amplitudes are 806 ± 322 pV and 640 ± 276 pV in the upper
and lower extremities, respectively.
1000ms Additional Comments. Although the actual technique of
eliciting a sympathetic skin response is relatively easy, the
Figure 6-17. S y m p a t h e t i c skin r e s p o n s e . Sympathetic skin re- actual response can be quite variable. - - Habituation of the
7 37 221
sponse from the palm of the right hand following right median nerve response also can be a problem. The practitioner should be
221
stimulation at the wrist (upper trace) and elbow.The conduction ve- prepared for considerable amplitude variation from one re-
locity of the sympathetic fibers is determined to be 1.3 m/s. sponse to the next. It is important to have the patient relaxed and
Chapter 6 SPECIAL NERVE C O N D U C T I O N T E C H N I Q U E S — 253
respond to as little external stimuli as possible between succes- 16. Borg J: Refractory period of single motor nerve fibers in man. J Neurol
Neurosurg Psychiatry 1984;47:344-348.
sive stimuli. Temperature can have a significant effect on the re-
17. Borg J: Conduction velocity and refractory period of single motor nerve fibres
sponse and should be monitored throughout the procedure with in motor neuron disease. J Neurol Neurosurg Psychiatry 1984;47:349-353.
a recommended temperature for upper and lower limb studies 18. Borg J: Conduction velocity and refractory period of single motor nerve
similar to that of routine nerve conduction studies. A correction fibres in antecedent poliomyelitis. J Neurol Neurosurs Psychiatry 1987;
50:443-446.
factor of 0.088 seconds/°C is recommended. Preliminary in- 35
19. Braddom RL, Johnson EW: Standardization of H reflex and diagnostic use in
vestigations in patients with various peripheral neuropathies SI radiculopathy. Arch Phys Med Rehabil 1974;55:161-166.
suggest that the sympathetic skin response may be of some as- 20. Braddom RL, Johnson EW: H reflex: Review and classification with suggested
sistance in evaluating autonomic fibers in these diseases. 202 clinical uses. Arch Phys Med Rehabil 1974;55:412Û7.
21. Burke D, McKeon B, Skuse NF: The irrelevance of fusimotor activity to the
Continued studies with this technique are required to adequately Achilles tendon jerk of relaxed humans. Ann Neurol 1981;10:547-550.
determine its clinical utility. 22. Burke D, McKeon B, Skuse NF: Dependence of the Achilles tendon reflex on
the excitability of spinal reflex pathways. Ann Neurol 1981;10:551-556.
23. Burke D, Gandevia SC, McKeon B: The afferent volleys responsible for spinal
proprioceptive reflexes in man. J Physiol 1983;339:535-552.
CONCLUSION 24. Burke D, Gandevia SC, McKeon B: Monosynaptic and oligosynaptic contribu-
tions to human ankle jerk and H-reflex. J Neurophysiol 1984;52:435-448.
25. Burke D, Adams RW, Skuse NF: The effects of voluntary contraction on the H
The specialized nerve conduction techniques discussed in reflex of human limb muscles. Brain 1989; 112:417-433.
this chapter are complementary to those described in the previ- 26. Burke D: Microneurography, impulse conduction, and paresthesias. Muscle
ous chapter. Although a number of the more specialized meth- Nerve 1993;16:1025-1032.
ods of investigating the nervous system may not be used in most 27. Burke JR: Multielectrode recordings of tibial nerve H-reflexes at various triceps
surae muscle sites in the right and left legs. Electromyogr Clin Neurophysiol
patients routinely, it is nevertheless important for the practi- 1997;37:277-286.
tioner to be familiar with both the methodology and utility of 28. Cherington M: Accessory nerve: Conduction studies. Arch Neurol
these techniques in order to know when their use is most effica- 1968;18:708-709.
29. Cherington M: Long thoracic nerve: Conduction studies. Dis Nerv System
cious. These techniques should be practiced until proficiency is
1972;33:49-51.
achieved so that an absent response can be confidently declared 30. Chroni E, Panayiotopoulos CP: F tacheodispersion: Quantitative analysis of
abnormal as opposed to its absence being a result of technical motor fiber conduction velocities in patients with peripheral neuropathies.
difficulty or a lack of skill. The actual utility of the special nerve Muscle Nerve 1993;16:1302-1309.
31. Cragg BG, Thomas PK: Changes in nerve conduction in experimental allergic
conductions is discussed throughout the remainder of this text neuritis. J Neurol Neurosurg Psychiatry 1964;27:106-115.
with respect to specific disorders. 32. Davis FA: Impairment of repetitive impulse conduction in experimentally de-
myelinated and pressure injured nerves. J Neurol Neurosurg Psychiatry
1972;35:537-544.
33. DeLisa JA, Mackenzie K, Baran EM: Manual of Nerve Conduction Velocity
REFERENCES and Somatosensory Evoked Potentials, 2nd ed. New York, Raven Press, 1987.
34. Dawson GD, Merton PA: "Recurrent" discharges from motoneurones. XX
1. Abbruzzese M, Reni L, Minatel C, Favale E: Presynaptic and postsynaptic International Congress of Physiology, Brussels, Abstract of Communications,
mechanisms underlying H-reflex changes produced by a selective voluntary 1956, pp 221-222.
contraction. Muscle Nerve 1998;21:439-453. 35. Deltombe T, Hanson P, Jamart J, Clerin M: The influence of skin temperature
2. Aiello I, Rosati G, Serra G, et al: The diagnostic value of H index in SI root on latency and amplitude of the sympathetic skin response in normal subjects.
compression. J Neurol Neurosurg Psychiatry 1981;44:171-172. Muscle Nerve 1998;21:34-39.
3. Aiello I, Serra G, Rosati G, et al: A quantitative method to analyze the H reflex 36. Delwaide PJ: Human monosynaptic reflexes and presynaptic inhibition. In
latencies from vastus medialis muscle: Normal values. Electromyogr Clin Desmedt JE (ed): New Developments in Electromyography and Clinical
Neurophysiol 1982;22:251-254. Neurophysiology. Basel, Karger, 1973, pp 444-462.
4. Alderson MK, Petajan JH: Relative refractory period: A measure to detect early 37. Denislic M, Meh D: Reproducibility of sympathetic skin response. Muscle
neuropathy in alcoholics. Muscle Nerve 1987;10:323-328. Nerve 1997;20:1332-1333.
5. Andersen H, Stalberg E. Falck B: F-wave latency, the most sensitive nerve con- 38. Deschuytere J, Rosselle N, De Keyser C: Monosynaptic reflexes in the superfi-
duction parameters in patients with diabetes mellitus. Muscle Nerve cial forearm flexors in man and their clinical significance. J Neurol Neurosurg
1997;20:1296-1302. Psychiatry 1976:39:555-565.
6. Angel RW, Hofmann WW: The H reflex in normal, spastic and rigid subjects. 39. Dindar F, Verrier M: Studies on the receptor responsible for vibration-induced
Arch Neurol 1963;8:21-26. inhibition of monosynaptic reflexes in man. J Neurol Neurosurg Psychiatry
7. Aramaki S, Kira Y, Yasusuke H: A study of the normal mean values and habitu- 1975;38:155-160.
ation phenomenon of sympathetic skin response. Am J Phys Med Rehabil 40. Dorfman LJ. Cummins KL, Abraham GS: Conduction velocity distributions of
1997;76:2-7. the human median nerve: Comparison of methods. Muscle Nerve
8. Ashby P, LaBelle K: Effects of extensor and flexor group I afferent volleys on 1982;5:S148-S153.
the excitability of individual soleus motoneurons in man. J Neurol Neurosurg 41. Dorfman LJ: The distribution of conduction velocities (DCV) in peripheral
Psychiatry 1977;40:910-919. nerves: A review. Muscle Nerve 1984;7:2-11.
9. Ashby P. Verrier M, Carleton S, et al: Vibratory inhibition of the monosynaptic 42. Dumitru D, Walsh NE: Practical instrumentation and common sources of error.
reflex and presynaptic inhibition in man. In Felman RG, Young RR, Koella WP Am J Phys Med Rehabil 1988;67:55-65.
(eds): Spasticity: Disordered Motor Control. Chicago, Year Book Publishers. 43. Dumitru D, Walsh NE, Porter LD: Electrophysiologic evaluation of the facial
1980. pp 335-344. nerve in Bell's palsy. Am J Phys Med Rehabil 1988;67:137-144.
10. Barakan TH, Dowman CBB, Eccles JC: Electric potentials generated by an- 44. Dumitru D, Walsh NE, Weber CF: Electrophysiologic study of the Riche-
tidromic volleys in quadricep and hamstring motoneurones. J Neurophyiol Cannieu anomaly. Electromyogr Clin Neurophysiol 1988;28:27-31.
1949:12:393-424 45. Eccles JC, Fatt P, Koketsu K: Cholinergic and inhibitory synapses in a pathway
11. Barr ML, Kiernan JA: The Human Nervous System: An Anatomical Viewpoint. from motor-axon collaterals to motoneurones. J Physiol 1954;126:524-562.
5th ed. Philadelphia, J.B. Lippincott Co., 1988. 46. Eccles JC: The central action of antidromic impulses in motor nerve fibers.
12. Bischoff C, Stalberg E, Falck B, Puksa L: Significance of A-waves recorded in Pflugers Arch 1955;260:385-415.
routine motor nerve conduction studies. Electroencephalogr Clin Neurophysiol 47. Eccles JC, Eccles RM, Iggo A, et al: Distribution of recurrent inhibition among
1996;101:528-533. motoneurones. J Physiol 1961; 159:479-^99.
13. Blom S, Hagbarth KE, Skoglund S: Post-tetanic potentiation of H-reflexes in 48. Eccles JC: The inhibitory control of spinal reflex action. Electroencephalog
humans and infants. Exp Neurol 1964;9:198-211. Clin Neurophysiol 1967;25:20-34.
14. Borg J: Axonal refractory period of single short toe extensor motor units in 49. Eisen A: Schomer D. Melmed C: An electrophysiological method for examin-
man. J Neurol Neurosurg Psychiatry 1980;43:917-924. ing lumbosacral root compression. Can J Neurol Sci 1977;4:117-123.
15. Borg J: Axonal refractory period of single short toe extensor motor units in neu- 50. Eisen A. Schomer D, Melmed C: The application of F-wave measurements in
ropathies and neuromuscular diseases. J Neurol Neurosurg Psychiatry 1981;44: the differentiation of proximal and distal upper limb entrapments. Neurology
1136-1140. 1977;27:662-668.
51. Eisen A, Odusote K: Amplitude of the F wave: A potential means of document- 90. Haymaker W: Bing's Local Diagnosis in Neurological Diseases. Saint Louis.
ing spasticity. Neurology 1979;29:1306-1309. C.V. Mosby, 1969.
52. Eisen A, Hoirch M, White J, et al: Sensory group la proximal conduction veloc- 91. Henneman E, Somjen G, Carpenter DO: Functional significance of cell size in
ity. Muscle Nerve 1984;7:636-641. spinal motoneurones. J Neurophysiol 1965;28:560-580.
53. Ekholm J, Skoglund S: Possible factors influencing the demonstration of post- 92. Henneman E, Somjen G, Carpenter DO: Excitability and inhibitibility of mo-
tetanic potentiation of the H-reflex as studied in the cat. Exp Neurol 1964:9: toneurons of different sizes. J Neurophysiol 1965;28:599-620.
183-197. 93. Hodes R: Selective destruction of large motor units by poliomyelitis virus. I.
54. Ellrich J, Steffens H, Treede R-D, Schomberg EK: The Hoffman reflex of Conduction velocity of motor nerve fibers of chronic poliomyelitis patients. J
human plantar foot muscles. Muscle Nerve 1998;21:732-738. , Neurophysiol 1948;12:257-266.
55. Esslen E: Electromyography and electroneuronography. In Fisch U (ed): Facial 94. Hodes R. Gribetz J: Normal electrically induced reflexes (H-reflexes) in a 2-
Nerve Surgery. Birmingham. Aesculapius Pub. Co., 1977. week-old infant with acrania and cerebral dysgenesis. Electroencephalogr Clin
56. Esslen E: The Acute Facial Palsies. New York. Springer Verlag, 1977. Neurophysiol 1963;15:1033-1036.
57. Falco FJE, Hennessey WJ, Goldberg G, et al: H reflex latency in the healthy el- 95. Hoffmann P: Uber die beziehungen der sehnenreflexe zur willkurlichen bewe-
derly. Muscle Nerve 1994;17:161-167. gung und zum tonus. Z Biol 1918;68:351-370.
58. Fisher MA, Shahani BT, Young RR: Assessing segmental excitability after 96. Hollinshead WH: Anatomy for Surgeons: The Back and Limbs. 3rd ed.
acute rostral lesions. I. The F response. Neurology 1978;28:1265-1271. Philadelphia. Harper & Row, 1982.
59. Fisher MA: F response latency determination. Muscle Nerve 1982;5:730-734. 97. Hong C, Joynt RL. Lin JC, et al: Axillary F-loop latency of ulnar nerve in
60. Fisher MA: F response analysis of motor disorders of central origin. J Neurol normal young adults. Arch Phys Med Rehabil 1981;62:565-569.
Sci 1983;62:13-22. 98. Hong C, Cheng H, Wang L, et al: Averaged F-wave conduction velocity of per-
61. Fisher MA: F response latencies and durations in upper motor syndromes. oneal nerve. Am J Phys Med Rehabil 1988;67:166-170.
Electromyogr Clin Neurophysiol 1986;26:327-332. 99. Hopf HC: Topodiagnostic value of brain stem reflexes. Muscle Nerve
62. Fisher MA: F/M ratios in polyneuropathy and spastic hyperreflexia. Muscle 1994; 17:457^84.
Nerve 1988;11:217-222. 100. Hopf HC: Untersuchunger liber die unterschiede in der leitgeschwindigkeit rno-
63. Fisher MA: Minimonograph #13: H reflex and F wave: Physiology and clinical torischer nervenfasern beim menschen. Dtsch Z Nervenheil 1962;183:579-588.
indications. Muscle Nerve 1992;15:1223-1233. 101. Hopf HC, Eysholdt M: Impaired refractory periods of peripheral sensory nerves
64. Fisher MA, Hoffen B: F-wave analysis in patients with carpal tunnel syndrome. in multiple sclerosis. Ann Neurol 1978;4:499-501.
Electromyogr Clin Neurophysiol 1997;37:27-31. 102. Horning MR, Kraft GH, Guy A: Latencies recorded by intramuscular needle
65. Fisher MA: The contemporary role of F-wave studies. Muscle Nerve electrodes in different portions of a muscle: Variation and comparison with sur-
1998;21:1098-1101. face electrodes. Arch Phys Med Rehabil 1972;53:206-211.
66. Fox JE, Hitchcock ER: Changes in F wave size during dentatotomy. J Neurol 103. Hughes GB, Josey AF, Glassock ME, et al: Clinical electroneurography: Statis-
Neurosurg Psychiatry 1982;45:1165-1167. tical analysis of controlled measurements in twenty-two normal subjects.
67. Fox JE, Hitchcock ER: F wave size as a monitor of motorneuron excitability: Laryngoscope 1981;91:1834-1846.
The effect of deafferentation. J Neurol Neurosurg Psychiatry 1987;50:453-459. 104. Hughes GB, Nodar RH, Williams GW: Analysis of test-retest variability in
68. Fullerton PM, Gilliatt RW: Axon reflexes in human motor nerve fibres. J Neurol facial electroneurography. Otolaryngol Head Neck Surg 1983;91:290-293.
Neurosurg Psychiatry 1965;28:1-11. 105. Jabre JF: The superficial peroneal sensory nerve revisited. Arch Neurol
69. Funase K, Miles TS: Observations on the variability of the H reflex in human 1981;38:666-667.
soleus. Muscle Nerve 1999;22:341-346. 106. Jabre JF: Surface recording of the H-reflex of the flexor carpi radialis. Muscle
70. Garcia HA, Fisher MA. Gilai A: H reflex analysis of segmental reflex excitabil- Nerve 1981;4:435-438.
ity in flexor and extensor muscles. Neurology 1979;29:984-991. 107. Jabre JF, Stalberg EV: Single fiber EMG study of the flexor carpi radialis H
71. Garcia-Mullin R, Mayer RF: H reflexes in acute and chronic hemiplegia. Brain reflex. Muscle Nerve 1989;12:523-527.
1972;95:559-572. 108. Jankus WR, Robinson LR, Little JW: Normal limits of side-to-side H-reflex
72. Gassel MM: A test of nerve conduction to muscles of the shoulder girdle as an amplitude variability. Arch Phys Med Rehabil 1994;75:3-7.
aid in the diagnosis of proximal neurogenic and muscular disease. J Neurol 109. Kameyama O, Hayes KC, Wolfe D: Methodological considerations contribut-
Neurosurg Psychiatry 1964;27:200-205. ing to variability of the quadricepts H-reflex. Am J Phys Med Rehabil
73. Gassel MM, Wiesendanger M: Recurrent and reflex discharges in plantar mus- 1989;68:277-282.
cles of the cat. Acta Physiol Scand 1965;65:138-142. 110. Kaplan P, Sahgal V: Residual latency: New applications of an old technique.
74. Gavilon J, Gavilon C, Sarria MJ: Facial electroneurography: Results on normal Arch Phys Med Rehabil 1978;59:24-27.
humans. J Laryngol Otol 1985;99:1085-1088. 111. Kaplan PE: Electrodiagnostic confirmation of long thoracic nerve palsy. J
75. Gilliatt RW, Willison RG: The refractory and supernonnal periods of the human Neurol Neurosurg Physchiatry 1980;43:50-52.
median nerve. J Neurol Neurosurg Psychiatry 1963;26:136-147. 112. Katz R, Morin C, Pierrot-Deseilligny E. et al: Conditioning of H reflex by a
76. Gilliat RW, Hopf HC, Rudge P, et al: Axonal velocities of motor units in the preceding subthreshold tendon reflex stimulus. J Neurol Neurosurg Psychiatry
hand and foot muscles of the baboon. J Neurol Sci 1976;29:249-258. 1977;40:575-580.
77. Gottlieb GL, Agarwal GC: Effects of initial conditions on the Hoffman reflex. J 113. Kimura J, Powers JM, Van Allen MW: Reflex response of orbicularis oculi
Neurol Neurosurg Psychiatry 1971;34:226-230. muscle to supraorbital nerve stimulation: Study in normal subjects and in pa-
78. Gottlieb GL, Agarwal GC: Extension of the Hoffmann reflex by antidromic ripheral facial paresis. Arch Neurol 1969;21:193-199.
conduction. Electroencephalogr Clin Neurophysiol 1976;41:19-24. 114. Kimura J: F-wave velocity in the central segment of the median and ulnar
79. Granit R, Pascoe JE, Steg G: The behavior of tonic a and g motoneurones as nerves: A study in normal subjects and in patients with Charcot-Marie-Tooth
measured by recurrent inhibition. J Physiol 1957;138:381-400. disease. Neurology 1974;24:539-546.
80. Guiheneuc P, Bathien N: Two patterns of results in polyneuropathies investi- 115. Kimura J, Butzer JF: F-wave conduction velocity in Guillain-Barre syndrome:
gated with the H reflex. J Neurol Sci 1976;30:83-94. Assessment of nerve segment between axilla and spinal cord. Arch Neurol
81. Guiloff RJ, Moderres-Sadeghi H: Preferential generation of recurrent responses 1975;32:524-529.
by groups of motor neurons in man. Brain 1991:114:1171-1801. 116. Kimura J, Bosch P. Lindsay GM: F-wave conduction velocity in the central seg-
82. Gutrecht JA: Sympathetic skin response. J Clin Neurophysiol ment of the peroneal and tibial nerves. Arch Phys Med Rehabil 1975;56:492-497.
1994; 11:519-524. 117. Kimura J: A method for estimating the refractory period of motor fibers in the
83. Hagbarth KE: Post-tetanic potentiation in myotatic reflexes in man. J Neurol human peripheral nerve. J Neurol Sci 1976;28:485-490.
Neurosurg Psychiatry 1962;25:1-10. 118. Kimura J: Collision technique: Physiologic block of nerve impulses in studies
84. Hagbarth KE: Microneurography and applications to issues of motor control. of motor nerve conduction velocity. Neurology 1976;26:680-682.
Muscle Nerve 1993;16:693-705. 119. Kimura J: A method for determining median nerve conduction velocity across
85. Hakamada S, Kumagai T, Watanabe K, et al: The conduction velocity of slower the carpal tunnel. J Neurol Neurosurg Psychiatry 1978;41:1-10.
and the fastest fibres in infancy and childhood. J Neurol Neursurg Psychiatry 120. Kimura J. Yamada T, Rodnitzky RL: Refractory period of human motor nerve
1982;45:851-853. fibres. J Neurol Neurosurg Psychiatry 1978;41:784-790.
86. Halar EM. Brozovich FV. Milutinovic J, et al: H-reflex latency in uremic neu- 121. Kimura J: Clinical value and limitations of F-wave determinations: A comment.
ropathy: Correlation with NCV and clinical findings. Arch Phys Med Rehabil Muscle Nerve 1978;1:250-252.
1979;60:174-177. 122. Kimura J: Refractory period measurement in the clinical domain. In Waxman
87. Hall SM, Williams PL: Studies on the incisures of Schmidt and Lanterman. J SG, Ritchie JM (eds): Demyelinating Disease: Basic and Clinical Electro-
Cell Sci 1970;6:767-792. physiology. New York, Raven Press. 1981, pp 239-265.
88. Hallin RG, Torebjork HE: Methods to differentiate electrically induced afferent 123. Kimura J: F-was'e determination in nerve conduction studies. In Desmedt JE
and sympathetic C unit responses in human cutaneoeous nerves. Acta Physiol (ed): Motor Control Mechanisms in Health and Disease. New York, Raven
Scand 1974;92:318-331. Press, 1983, pp 961-975.
89. Han TR, Paik NJ: A study on new diagnostic criteria of H reflex. Electromyogr 124. Kimura J, Yanagisawa H, Yamada T, et al: Is the F wave elicited in a select
Clin Neurophysiol 1997:37:241-250." group of motoneurons? Muscle Nerve 1984;7:392-399.
125. Kimura J: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and 162. Mongia SK: H reflex from quadriceps and gastrocnemius muscles.
Practice. Philadelphia, F.A. Davis, 1989. Electromyogr Clin Neurophysiol 1972;12:179-190.
126. Komori T, Watson BV, Brown WF: Characteristics of single "F motor units at
1
163. NishidaT, Kompoliti A, Janssen I, Levin KF: H reflex in S-l radiculopathy: La-
different stimulus intensities. Muscle Nerve 1991; 14:875. tency versus amplitude controversy revisited. Muscle Nerve 1996;19:915-917.
127. Kopec J, Delbeke J, McComas AJ: Refractory period studies in a human neuro- 164. Ohgaki K, Nakano K, Shigeta H, et al: Ratio of motor nerve conduction veloc-
muscular preparation. J Neurol Neurosurg Psychiatry 1978;41:54-64. ity to F-wave conduction velocity in diabetic neuropathy. Diabetes Care
128. Kraft GH: Axillary, musculocutaneous and suprascapular nerve latency studies. 1998;21:615-618.
Arch Phys Med Rehabil 1972;53:383-387. 165. Olsen PZ, Diamantopoulos E: Excitability of spinal motor neurones in normal
129. Kraft GH, Halvorson GA: Median nerve residual latency: Normal value and use subjects and patiens with spasticity, parkinsonian rigidity, and cerebellar hypo-
in diagnosis of carpal tunnel syndrome. Arch Phys Med Rehabil 1983;64: tonia. J Neurol Neurosurg Psychiatry 1967;30:325-331.
221-226. 166 Ongerboer de Visser BW, Cruccu G: Neurophysiologic examination of the
130. Lachman T, Shahani BT, Young RR: Late responses as aids to diagnosis in pe- trigeminal, facial, hypoglossal and spinal nerves in cranial neuropathies and
ripheral neuropathy. J Neurol Neurosurg Psychiatry 1980;43:156-162. brain stem disorders. In Brown WF, Bolton CF (eds): Clinical Electromygraphy,
131. Lader MH, Montagu JD: The psycho-galvanic reflex: A pharmacologic study 2nd ed. New York, Butterworth-Heinemann, 1993, pp 61-92.
of the peripheral mechanism. J Neurol Neurosurg Psychiatry 1962;25:126-133. 167. Ongerboer De Visser BW, Schimsheimer RJ, Hart AAM: The H-reflex of the
132. Languth HW, Teasdall RD, Magladery JW: Electrophysiological studies of flexor carpi radialis mucle: A study in controls and radiation-induced brachial
reflex activity in patients with lesions of the nervous system. Bull Johns Hopk plexus lesions. J Neurol Neurosurg Psychiatry 1984;47:1098-1101.
Hosp 1952;91:257-266. 168. Paintal AS: Conduction in mammalian nerve fibres. In Desmedt J (ed): New
133. Liberson WT: Monosynaptic reflexes and their clinical significance. Developments in Electromyography and Clincal Neurophysiology, Vol. 2.
Electroencephalogr Clin Neurophysiol 1962(Suppl 22):79-89. Basel, Karger, 1973, pp 19-^1.
134. Liberson WT: Sensory conduction velocities in normal individuals and in pa- 169. Panayiotopoulos CP, Scarpalezos S, Nastas PE: F-wave studies on the deep per-
tients with peripheral neuropathies. Arch Phys Med Rehabil 1963;44:313-320 oneal nerve. J Neurol Sci 1977;31:319-329.
135. Little JW, Halar EM: H-reflex changes following spinal cord injury. Arch Phys 170. Panayiotopoulos CP: F chronodispersion: A new electrophysiologic method.
Med Rehabil 1985,66:19. Muscle Nerve 1979;2:68-72.
136. Lloyd DPC: Neuron patterns controlling transmission of ipsilateral hind limb 171. Panayiotopoulos CP, Chroni E: F-waves in clinical neurophysiology: A review,
reflexes in cat. J Neurophysiol 1943;6:293-315. methodological issues and overall value in peripheral neuropathies.
137. Lorente de No R: The synaptic delay of the motoneurons. Am J Physiol Electroencephalogr Clin Neurophysiol 1996;101:365-374.
1935;111:272-282. 172. Panizza M, Lelli S, Hallett M: H-reflex in nonhomonymous muscles in the
138. Lorente de No R: Limits of vartiation of the synaptic delay of motoneurons. J human forearm. Neurology 1989;39:785-788.
Neurophysiol 1938;1:187-194. 173. Panizza M, Nilsson J, Hallett M: Optimal stimulus duration for the H reflex.
139. Louis AA, Hotson JR: Regional cooling of human nerve and slowed Na inacti-
+
Muscle Nerve 1989;12:576-579.
vation. Electroencephalogr Clin Neurophysiol 1986;63:371-375. 174. Panizza M, Lelli S. Nilsson J, et al: H-reflex recovery curve and reciprocal inhi-
140. Ma DM, Kim SH, Spielholz N, et al: Sensory conduction study of distal radial bition of H-reflex in different kinds of dystonia. Neurology 1990;40:824-828.
nerve. Arch Phys Med Rehabil 1981;62:562-564. 175. Pease WS, Kozakiewicz R, Johnson EW: Central loop of the H reflex: Normal
141. Ma DM, Liveson JA: Nerve Conduction Handbook. Philadelphia, F.A. Davis, value and use in SI radiculopathy. Am J Phys Med Rehabil 1997;76:182-184.
1983. 176. Peioglou-Harmoussi S, Howel D, Facett PRW, et al: F-response behaviour in a
142 Macdonell RAL, Cros D, Shahani BT: Lumbosacral nerve root stimulation com- control population. J Neurol Neurosurg Psychiatry 1985;48:1152-1158.
paring electrical with surface magnetic coil techniques. Muscle Nerve 177. Peioglou-Harmoussi S, Fawcett PRW, Howel D, et al: F-response: A study of
1992;15:885-890. frequency, shape, and amplitude characteristics in healthy control subjects. J
143. Mackenzie K, DeLisa JA: Distal sensory latency measurement of the superficial Neurol Neurosurg Psychiatry 1985;48:1159-1164.
radial nerve in normal adult subjects. Arch Phys Med Rehabil 1981;62:31-34. 178. Prout BJ: Independence of the galvanic skin reflex from the vasocontrictor
144. MacLean IC: Spinal nerve stimulation. In AAEM Course B: Nerve Conduction reflex in man. J Neurol Neurosurg Psychiatry 1967;30:319-324.
Studies—A Review. AAEM, Rochester, MN, 1988. 179. Rasminsky M: The effects of temperature on conduction in demyelinated single
145. Magladery JW, McDougal DB: Electrophysiological studies of nerve and reflex nerve fibers. Arch Neurol 1973;28:287-292.
activity in normal man. I. Identification of certain reflexes in the electromyo- 180. Ravits JM: AAEM Minimonograph #48: Autonomic nervous system testing.
gram and the conduction velocity of peripheral nerve fibres. Bull Johns Muscle Nerve 1997,20:919-937.
Hopkins Hosp 1950;86:265-290. 181. Renshaw B: Activity in the simplest spinal reflex pathways. J Neurophysiol
146. Magladery JW, Teasdall RD: Stretch reflexes in patients with spinal cord le- 1940;3:373-387.
sions. Bull Johns Hopk Hosp 1958;103:236-241. 182. Renshaw B: Influence of discharge of motoneurons upon excitation of neigh-
147. Maryniak O, Yaworski R: H-reflex: Optimum location of recording electrodes. boring motoneurons. J Neurophysiol 1941;4:167-183.
Arch Phys Med Rehabil 1987;68:798-802. 183. Rexed B: Contributions to the knowledge of the postnatal development of the
148. Maryniak O, Yawarski R, Hayes KC: Intramuscular recording of H-reflexes peripheral nervous system in man. Acta Psychiatr Scand 1944(Suppl 33): 1-206.
from muscles of the posterior compartment of the lower leg. Am J Phys Med 184. Rico RE, Jonkman EJ: Measurement of the Achilles tendon reflex for the diag-
Rehabil 1991;70:34-39. nosis of lumbosacral root compression syndromes. J Neurol Neurosurg
149. Masland WS: Facilitation during the H-reflex recovery cycle. Arch Neurol Psychiatry 1982;45:791-795.
1972:26:313-319. 185. Rivner MH: F-wave studies: Limitations. Muscle Nerve 1998;21:1101-1104.
150. Mastaglia FL, Carrol WM: The effects of conditioning stimuli on the F-re- 186. Rosier KM, Hess CW, Schmid UD: Investigation of facial motor pathways by
sponse. J Neurol Neurosurg Psychiatry 1985;48:182-814. electrical and magnetic stimulation: Sites and mechanisms of excitation. J
15!. Maurer K, Hopf HC, Lowitzsch K: Hypokalemia shortens relative refractory Neurol Neurosurg Psychiatry 1989:52:1149-1156.
period of peripheral sensory nerves in man. J Neurol 1977;216:67-71. 187. Rossi B, Sartucci F, Stefanini A: Measurement of motor conduction velocity
152. May M: The Facial Nerve. New York, Thieme Inc., 1986. with Hopf's technique in the diagnosis of mild peripheral neuropathies. J
153. Mayer RF, Mawdsley C: Studies in man and cat of the significance of the H- Neurol Neurosurg Psychiatry 1981;44:168-170.
wave. J Neurol Neurosurg Psychiatry 1965;28:201-211. 188. Roth G: Intranervous regeneration of lower motor neuron—1. Electromyogr
154. Mayer RF, Feldman RG: Observations on the nature of the F wave in man. Clin Neurophysiol 1978:18:225-288.
Neurology 1967;17:147-156. 189. Roth G, Egloff-Baer S: Ephaptic response in man. Eur Neurol
155. Mayer RF, Mosser RS: Excitability of motoneurons in infants. Neurology 1979:18:261-266.
1969;19:932-945. 190. Roth G, Egloff-Baer S: Motor axon loop: An electroneurographic response.
156. McHugh DJ, Reeser JC. Johnson EW: H reflex amplitude: Effect of leg muscle Muscle Nerve 1984;7:294-297.
activity. Am J Phys Med Rehabil 1997;76:185-187. 191. Sabbahi MA, Khalil M: Segmental H-reflex studies in upper and lower limbs of
157. McLeod JG, Wray S: An experimental study of the F wave in the baboon. J healthy subjects. Arch Phys Med Rehabil 1990;71:216-222.
Neurol Neurosurg Psychiatry 1966;29:196-200. 192. Sabbahi MA, Khalil M: Segmental H-reflex studies in upper and lower limbs of
158. McLeod JG: H reflex studies in patients with cerebellar disorders. J Neurol patients with radiculopathy. Arch Phys Med Rehabil 1990;71:223-227.
Neurosurg Psychiatry 1969;32:21-27. 193. Sawhney BB, Kayan A: A study of axon reflexes in some neurogenic disorders.
159. Menkes DL, Hood DC. Bush AC: Inversion of the F-waves in median neuropa- Electromyogr Clin Neurophysiol 1970;10:297-305.
thy at the wrist (carpal tunnel syndrome): An adjunctive electrodiagnostic 194. Scheibel ME, Scheibel AB: Spinal motoneurons, interneurons and Renshaw
method. J Contemp Neurol 1997; 1:1 -10. cell: A Golgi study. Arch Ital Biol 1966;104:328-353.
160. Miglietta OE: The F-response after transverse myelotomy. In Desmedt JE (ed): 195. Schiller HH, Stalberg E: F responses studied with stingle fibre EMG in normal
New Developments in Clinical Neurophysiology. Basel. Karger, 1973, pp subjects and spastic patients. J Neurol Neurosurg Psychiatry 1978;41:45-53.
323-327. 196. Schimsheimer RJ, Ongerboer De Visser BW, Kemp B: The flexor carpi radialis
161. Mihelin M, Trontelj JV, Stalberg E: Muscle fiber recovery functions studied H-reflex in lesions of the sixth and seventh cervical nerve root. J Neurol
with double pulse stimulation. Muscle Nerve 1991;14:739-747. Neurosurg Psychiatry 1985;48:445-449.
197. Schimsheimer RJ, Ongerboer De Visser BW, Kemp B, et al: The flexor carpi ra- 219. Tomasulo RA: Aberrant conduction in human peripheral nerve: Ephaptic trans-
dialis H-reflex in polyneuropathy: Relations to conduction velocities of the mission? Neurology 1982;32:712-719.
median nerve and the soleus H-reflex latency. J Neurol Neurosurg Psychiatry 220. Torebjork E: Human microneurography and intraneural microstimulation in the
1987;50:447^52. study of neuropathic pain. Muscle Nerve 1993;16:1063-1065.
198. Schmidt RF: Fundamental of Neurophysiology. New York, Springer-Verlag, 221. Toyokura M, Murakami K: Reproducibility of sympathetic skin response.
1985. Muscle Nerve 1996;19:1481-1483.
199. SchrieferTN, Mills KR, Murray NMF, Hess CW: Evaluation of proximal facial 222. Toyokura M, Murakami K: F-wave study in patients with lumbosacral radicu-
nerve conduction by transcranial magnetic stimulation. J Neurol Neurosurg lopathies. Electromyogr Clin Neurophysiol 1997;37:19-26.
Psychiatry 1988;51:60-66. 223. Trojaborg W: Motor and sensory conduction in the musculocutaneous nerve. J
200. Schuchmann JA: H reflex latency in radiculopathy. Arch Phys Med Rehabil Neurol Neurosurg Psychiatry 1976;39:890-899.
1978;59:185-187. 224. Troni W: Analysis of conduction velocity in the H pathway: Part 1.
201. Serra G, Aiello 1, De Grandis D, et al: Muscle-nerve ephaptic excitation in some Methodology and results in normal subjects. J Neurol Sci 1981;51:223-233.
repetitive after-discharged. Electroencephalogr Clin Neurophysiol 225. Troni W: Analysis of conduction velocity in the H pathway: Part 2. An electro-
1984;57:416-422. physiological study in diabetic polyneuropathy. J Neurol Sci 1981;51:235-246.
202. Shahani BT, Halperin JJ, Boulu P, et al: Sympathetic skin response—a method 226. Troni W: The value and limits of the H reflex as a diagnostic tool in SI root
of assessing unmyelinated axon dysfunction in peripheral neuropathies. J compression. Electromyogr Clin Neurophysiol 1983;23:471-480.
Neurol Neurosurg Psychiatry 1984;47:536-542. 227. Troni W, Cantello R, Rainero E: The use of the H reflex in serial evaluation of
203. Shanthaveerappa TR, Bourne GH: The "peripheral epithelium," a metabolically nerve conduction velocity. Electroencephalogr Clin Neurophysiol
active, continuous, protoplasmic cell barrier surrounding peripheral nerve fasci- 1983;55:82-90.
culi. J Anat 1962;96;527-537. 228. Trontelj JV: H-reflex of single motoneurones in man. Nature 1968;220:
204. Smith RS: The short-term accumulation of axonally transported organelles in 1043-1044.
the region of localized lesions of single myelinated axons. J Neurocytol 229. Trontelj JV: A study of the H-reflex by single fibre EMG. J Neurol Neurosurg
1980;9:39-65. Psychiatry 1973;36:951-959.
205. Somerville J, Ashby P: Hemiplegic spasticity: Neurophysiological studies. Arch 230. Vallbo AB, Hagbarth KE, Torebjork HE, et al: Somatosensory, proprioceptive,
Phys Med Rehabil 1978;59:592-596. and sympathetic activity in human peripheral nerves. Physiol Rev 1979;59:
206. Stalberg E, Trontelj JV: Single Fiber Electromyography. Old Woking, UK, 919-57.
Mirvalle Press, 1979, pp 1-244. 231. Van Boxtel A: Differential effects of low-frequency depression, vibration-in-
207. Taborikova H: Fraction of the motoneurone pool activated in the monosynaptic duced inhibition, and posttetanic potentiation on H-reflexes and tendon jerks in
H-reflexes in man. Nature 1966;209:206-207. the human soleus muscle. J Neurophysiol 1986;55:551-568.
208. Taborikova H, Sax DS: Motoneurone pool and the H-reflex. J Neurol Neurosurg 232. Van Dijk JG, Ferrari MD: Nerve conduction sensitivity in diabetic neuropathy
Psychiatry 1968;31:354-361. (letter to the editor). Muscle Nerve 1999;22:140.
209. Taborikova H, Sax DS: Conditioning of H-reflexes by a preceding subthreshold 233. Veale JL, Mark RF, Sandra R: Differential sensitivity of motor and sensory
H-reflex stimulus. Brain 1969;92:203-212. fibres in human ulnar nerve. J Neurol Neurosurg Psychiatry 1973;36:75-86.
210. Tackmann W, Lehmann HJ: Refractory period in human sensory nerve fibres. 234. Veale JL, Rees S: Renshaw cell activity in man. J Neurol Neurosurg Psychiatry
Eur Neurol 1974;12:277-292. 1973;36:674-683.
211. Tackmann W, Ullerich D, Cremer et al: Nerve conduction studies during the rel- 235. Veale JL, Rees S, Mark RF: Renshaw cell activity in normal and spastic man. In
ative refractory period in sural nerves of patients with uremia. Eur Neurol Desmedt JE (ed): New Developments in Electromyography and Clinical
1974;12:331-339. Neurophysiology, Basel, Karger 1973, pp 523-537.
212. Taniguchi MH, Hayes J, Rodriguez A: Reliability determination of F mean re- 236. Verrier MC. Alterations in H reflex magnitude by variations in baseline EMG
sponse latency. Arch Phys Med Rehabil 1993;74:1139-1143. excitability. Electroencephalogr Clin Neurophysiol 1985;60:492-499.
213. Tasaki I: Conduction of the nerve impulse. In: Field J, Magoun HW, Hall VE 237. Voulgaris AN, Constantinidis TS: P-perisistence in L5-SI radiculopathies.
(ed): Handbook of Physiology, Vol I. Washington, American Physiological Electromyogr Clin Neurophysiol 1996;36:387-393.
Society, 1959, pp 75-121. 238. Wager EW, Buerger AA: A linear relationship between H-reflex latency and.
214. Taylor S, Ashby P, Verrier M: Neurophysiological changes following traumatic sensory conduction velocity in diabetic neuropathy. Neurology
spinal lesions in man. J Neurol Neurosurg Psychiatry 1984;47.T 102-1108. 1974;24:711-714.
215. Teasdall RD, Park AM, Languth HW, et al: Electrophysiological studies of 239. Wang F-C, Delwaide PJ: Single motor unit H reflexes recorded in thenar mus-
reflex activity in patients with lesions of the nervous system. Bull Johns Hopk cles at rest. Muscle Nerve 1999;22:291-292.
Hosp 1952;91:245-256. 240. Wu Y, Kunz JRM, Putnam TD, et al: Axillary F central latency: Simple electro-
216. Thomas JE, Lambert EH: Ulnar nerve conduction velocity and H-reflex in in- diagnostic technique for proximal neuropathy. Arch Phys Med Rehabil
fants and children. J Appl Physiol 1960;15:1-9. 1983;64:117-120.
217. Thomas PK, Sears TA, Gilliatt RW: The range of conduction velocity in normal 241. Wuerker RB, Kirkpatrick JB: Neuronal microtubules, neurofilaments, and mi-
motor nerve fibres to the small muscles of the hand and foot. J Neurol crofilaments, lnt Rev Cytol 1972;33:45-75.
Neurosurg Psychiatry 1959;22:175-181. 242. Yates SK, Brown WF: Characteristics of the F response: A single motor unit
218. Thorne J: Central responses to electrical activation of the peripheral nerves sup- study. J Neurol Neurosurg Psychiatry 1979;42:161-170.
plying the intrinsic hand muscles. J Neurol Neurosurg Psychiatry 1965:28: 243. Young RR, Shahani BT: Clinical value and limitations of F-wave determina-
482-495. tion. Muscle Nerve 1978;1:248-249.
Chapter 7
Needle Electromyography
CHAPTER OUTLINE
Preparation for the Needle Examination Muscle Generators of Abnormal Spontaneous

The Patient • The Examiner • The Equipment Potentials
Insertional Activity • Fibrillation Potentials • Positive Sharp
The Art of the Needle Examination Waves • Fibrillation/Positive Sharp Waves: Clinical Findings
• Complex Repetitive Discharge • Myotonic Discharge
Performing the Needle Electromyographic
Examination Neural Generators of Abnormal Spontaneous
Muscle at Rest • Insertional Activity • Minimal t o Moderate Potentials
Contraction/Recruitment • Information Synthesis Fasciculation Potentials • Myokymic Discharge * Continuous
• Impression Formulation Muscle Fiber Activity • Cramps • Multiple Discharges
• Tremor
Electrical Activity in Muscle
Electrophysiology of Muscle • Action Potential Generation Muscle Generators of Abnormal Voluntary Activity
* Local Circuit Currents • Local Circuit Model of Waveform Neural Loss • Motor Unit • MUAP Findings Following
Morphology Denervation/Reinnervation
Electrical Potentials in Muscle Muscle Loss

Normal Potentials in Muscle • Needle Insertional Activity Motor Unit • MUAP Findings in Primary Muscle Disease
• Spontaneous Activity • Miniature Endplate Potentials • MUAP Findings in Neuromuscular Junction Disorders
(Monophasic Waveform) • Endplate Spikes (Biphasic • Abnormal MUAP Recruitment • Neurogenic Recruitment
Negative/Positive Waveform) • Myogenic Recruitment • Recruitment in Stroke • Additional
Waveforms
Muscle Generators of Normal Voluntary
Activity Needle Electromyography: Relevant Issues
Single Muscle Fiber (Triphasic Waveform) • Motor Unit Pain on Needle Examination • Contraindications and
Action Potentials • Anatomy • Physiology • Recruitment Complications Related t o Needle Examination • Sterilization of
Principles • Parameters Needle Electrodes • Muscle Biopsy/Serum CPK Level
Inserting a needle electrode into muscle tissue and recording PREPARATION FOR THE NEEDLE
the ensuing electrical activity is referred to as needle elec-
tromyography (EMG). This is one of the fundamental portions
EXAMINATION
of the electrodiagnostic medicine consultation. Clinical elec- THE PATIENT
tromyography can be thought to originate with the introduction
of the concentric needle electrode by Adrian and Bronk in Prior to discussing the actual needle examination, several
1929. Technological advances in electronics and microcomput-
2
general precepts must first be recognized. It is important to re-
ers have enabled the development of multiple data acquisition member that the needle electromyographic examination can be
methods leading to additional ways of investigating the electri- at best somewhat uncomfortable and at worst barely tolerable.
cal activity generated by the muscle fiber. We will explore A relatively pleasant and productive needle examination can be
normal and abnormal muscle electrical activity that can be de- experienced by both the patient and physician, provided the
tected with a needle recording electrode placed within muscle physician considers several simple but worthwhile principles. 112
tissue. One of the most important goals the physician can accomplish
257
is to gain the confidence and cooperation of the patient. In addi- monopolar and standard concentric needle electrodes, obviating
tion to a professional demeanor, communication is the single sterilization for most routine needs. It is still acceptable to use
most valuable factor in achieving meaningful data. The physi- non-disposable needle electrodes, provided they are soaked in
cian should inform the patient as to the reason for performing standard bleach solution (sodium hypochlorite) for approxi-
the needle examination, how the information gained may help mately 15-20 minutes, rinsed clean, and then sterilized. During
establish a diagnosis, and what may be experienced during the needle insertion, it may be of benefit to use distraction tech-
examination. Informed patients are usually cooperative because niques such as applying pressure about the needle insertion site
they are participating in their care as opposed to having some- with the free hand, talking to the patient about his or her various
thing done to them. interests, or other helpful maneuvers to keep the patient's mind
The patient's anxiety level may be diminished by accurately, off the needle insertions.
but not graphically, explaining that some tolerable discomfort
may at times be felt with needle insertion and movement. THE EQUIPMENT
Attempting to convince the individual that pain will not be expe-
rienced is inappropriate and reduces the examiner's credibility The specific electrodiagnostic instrument one uses is usually
when pain is indeed noted. Most patients will tell the physician based more on preference than actual capabilities. Most of the
that they would like to be informed just prior to needle insertion equipment manufacturers offer virtually the same options which
in order to prepare themselves. Giving the patient the option of are adequate to perform a routine needle electromyographic ex-
preparation offers some modicum of control which is often ap- amination. The added expense of multicolored graphic displays
preciated. Displaying the needle electrode is usually inadvisable and various data manipulation options are often unnecessary and
unless requested because needle length is frequently translated rarely used by most practitioners. Before purchasing an instru-
into pain. Many physicians refer to the needle electrode as a ment, the clinician needs to inquire about the following: (1) am-
"pin" or similar term to diminish the anxiety associated with the plifier input impedance and common mode rejection ratio, (2)
designation "needle." Gauze or tissue within easy reach com- variable filters, (3) analog-to-digital conversion, (4) trigger and
bined with firm pressure over the needle site following needle delay, (5) cathode ray tube (CRT) resolution, and (6) service.
electrode withdrawal is usually quite helpful in controlling pos- The instrument's amplifier is one of its most important com-
sible blood loss. Bloody tissues and pads should be kept from ponents. The input impedance should be greater than or equal to
patient view and disposed of efficiently to minimize anxiety. It 10 MiQ thereby ensuring most of the voltage detected will be
may be difficult to tell the patient how many more muscles will amplified and not lost to the recording electrode. - - 75 77 218
need to be examined when requested to do so, but an approxi- Differential amplification quality is also crucial to maximizing
mate number with an explanation for the uncertainty often the signal while eliminating common mode noise. This is usu-
allows the patient to prepare adequately. Finally, the patient ally measured by the common mode rejection ratio, which
should be informed that muscle soreness for a day or two occa- should exceed 10,000:1. The practitioner must be familiar with
sionally accompanied by a small degree of bruising around the filters and their ability to help or confound the observed results.
needle site can occur. Mild over-the-counter analgesics typi- Optimal filter settings for the routine (qualitative and not quan-
cally suffice until the pain spontaneously resolves.
112
titative) needle examination are 10-30 Hz for the low-frequency
(high-pass) filter and 10,000-30,000 Hz for the high-frequency
THE EXAMINER (low-pass) filter. Beware of instruments that have preset
75 2 1 8
filter settings as they may be inappropriate for the test per-

The practitioner performing the needle electromyographic formed and compromise flexibility in altering the recording pa-
examination is a medical consultant and should use the history rameters, particularly if there is a desire to reproduce another
and physical examination combined with the needle examina- laboratory's conditions. Once the analog signal is amplified and
tion's electrical findings to arrive at an electrodiagnostic medi- filtered, most modern instruments convert the analog signal into
cine diagnosis. Upon completion of the history and physical a digital representation (analog-to-digital conversion) so that it
examination, the electrodiagnostic medicine consultant should can be "frozen" on the screen and analyzed in detail. The
have a clear idea regarding which muscles should be examined process of converting an analog signal to a digital one requires
to arrive at a correct diagnosis. Informing the patient of the in- an analog-to-digital converter, which effectively takes a real
tended plan is usually helpful in gaining cooperation. It is help- time signal and translates it into binary code. During the needle
ful to keep in mind, however, that the electrodiagnostic examination, it is often helpful to position a motor unit action
medicine examination is a dynamic process and may require potential in the same place on the CRT screen each time it fires.
slight alterations in the exact muscles explored as information is This allows one the ability not only to observe the potential's
acquired. This is an important concept to convey to the patient morphology, but also look for time locked potentials (satellite
particularly when several more muscles than originally antici- potentials) that are not part of the main complex. These time
147
pated require examination. If one is unsure of the reason for ex- locked potentials can be misinterpreted as separate potentials on
amining the patient, the referring physician should be contacted a free running CRT screen. The trigger and delay line triggers
to avoid an unnecessary repeat study at a later time. Discussing on a particular portion of the desired potential and delays the
the patient with the referring physician can be quite productive signal while the CRT trace advances a preset amount, resulting
in considering alternative diagnostic strategies and is often ap- in the potential repetitively appearing in the same place on the
preciated as a recognition of thoroughness. CRT screen each time it fires (see Chapter 3). The CRT contains
All necessary equipment should be organized for efficient discrete points of resolution to define the waveform; approxi-
use. Searching for supplies while the patient is waiting to be ex- mately 500 points across the screen suffices for all routine pur-
amined can be quite disconcerting for everyone. It is a good poses. One of the most important considerations regarding any
practice to use disposable gloves during the needle examination. piece of equipment is the service provided by the factory. This
Most manufacturers now offer technically acceptable disposable knowledge is usually obtained by talking to colleagues who
Chapter 7 NEEDLE ELECTROMYOGRAPHY — 2 5 9
have worked with the company in question. Finally, it is impor- necessary depending upon the individual patient. When examin-
tant to remember that the skill required to perform a competent ing the muscle at rest, the needle should be inserted along a par-
needle examination is in the "hands" of the practitioner and not ticular line and then withdrawn to the subcutaneous position
the instrument, irrespective of its cost. and redirected along another path while still in the same muscle.
Perhaps one of the most discussed aspects regarding the Performing this several times allows one to explore rather large
needle examination is the type of needle electrode used. Most portions of the muscle while using only one skin puncture site.
routine needle examinations will be performed with either a Once the needle is within muscle tissue, pain can also.be re-
monopolar or concentric needle e l e c t r o d e .
77167180
Significant duced by avoiding structures such as the endplate region, which
energy is expended at conferences and in publications regarding yields characteristic potentials and sounds (see below), perios-
the benefits and detriments of these two needle types. Unfor- teum, nerves, vessels, and tendons. 112
tunately, this issue typically becomes rather emotional and is When voluntary motor units are to be assessed, the patient
based more on what type of electrode the individual trained with must contract the muscle. It is usually best to withdraw the
as opposed to objective documentation of the electrode's merits needle to the subcutaneous position to avoid the muscle bending
or faults. It is the authors' opinion that each electrode has partic- the needle or the needle tearing muscle tissue." Isometric
2
ular characteristics that one may wish to employ for specific in- muscle contractions are tolerated rather well and the needle can
dications and that the competent electrodiagnostic medicine be reinserted into muscle tissue from the subcutaneous location
practitioner should be familiar with and use both electrodes. into actively contracting muscle with relatively little muscle
The reader is referred to Chapter 3 for a detailed discussion of injury. The paraspinal muscles may occasionally present a chal-
the electrodiagnostic medicine equipment. lenge to the practitioner regarding relaxation. With the patient
in the prone position, placing several pillows under the chest so
that the forehead barely touches the plinth (cervical paraspinal
THE ART OF THE NEEDLE EXAMINATION muscles) or under the abdomen so the lumbosacral region is
rather arched (lumbosacral paraspinal muscles) may help. In
The needle electromyographic examination is predicated both instances, the author has found that having the patient
upon a great deal of neuroscience, but there is also considerable place both arms off the plinth and hang limply facilitates
technical skill "in the hands" of the examiner, i.e., the "art of the paraspinal muscle relaxation. Should this not produce electrical
needle examination." There is more to the needle examination silence with the needle in the cervical paraspinal muscles,
than placing needles into muscles and mastering the knowledge having the patient gently push the forehead against one of the
regarding anatomy and disease pathophysiology. It takes years examiner's hands may help in that the firing of the neck flexors
to gain the sense of how best to examine a particular muscle, can relax the neck extensors. In the lumbosacral region, the pa-
minimize pain upon needle insertion, gain the patient's confi- tient can be asked to tighten the stomach muscles against the ex-
dence, possess a "feel" for what muscles will be most revealing aminer's hand to accomplish the same effect described for the
for a specific lesion in any given patient, etc. Diligent practice n e c k . It may also help to place pillows under the patient's
128
and keeping informed about recent advances in electrodiagnos- legs, producing slight knee flexion. The patient can then be told
tic medicine will reward the practitioner with a special ability to to gently press the knees into the plinth in an attempt to relax
assist in the diagnosis and treatment of patients afflicted with the lumbosacral paraspinal muscles. Paraspinal muscle relax-
diseases of the central and peripheral nervous systems. Several ation can also be achieved at times with the patient in the side-
"hints" gained from experienced electrodiagnostic medicine lying position. Muscles on either side of the spine may be
practitioners can help the relatively less seasoned examiner examined, as relaxation of a particular side is highly individual-
enjoy a more productive needle examination. ized. Extensor muscles, especially of the forearm, are also often
An initial place to begin is to implement the general guide- difficult to relax. In the authors' experience these motor units
lines discussed above regarding the patient and examiner. A few often stop firing when the forearm is switched from neutral to a
specifics with respect to the needle examination may also be more supinated position, such that the muscles do not act any
beneficial in minimizing patient discomfort and obtaining re- longer as antigravity muscle. Additionally, a towel roll may be
vealing information. placed under the patient's wrist with the forearm supinated
It may be worthwhile in a particularly anxious patient to per- thereby permitting relaxation of the forearm extensor muscles.
form the examination in a quiet location, dim the room lights, These are just a few of the helpful hints in performing the
speak in a calming tone, and keep the patient comfortable re- needle examination that the authors have found useful over the
garding room temperature. Prior to piercing the skin over a se- years. Of course, none of the techniques described works all the
lected muscle, it is often possible to find an area that is time in all patients. Discussing solutions to these issues with
relatively less painful than others by gently touching the skin one's colleagues can be rather helpful in trying to solve various
surface with the needle tip in several locations. This is an at- problems regarding the art of the needle examination.
tempt to find a region of skin that is less well innervated and
minimize pain upon piercing the skin. During the actual punc-
turing of the skin, pinching or applying firm pressure near the PERFORMING THE NEEDLE ELECTRO-
needle site while simultaneously inserting the needle electrode MYOGRAPHIC EXAMINATION
can reduce discomfort. The author prefers to stretch the region
of skin about to be penetrated with the first and second digits of In this section, the needle electromyographic examination
the hand not holding the needle to facilitate needle insertion. refers to the qualitative needle exploration of muscle most practi-
Either before the needle is inserted or while it is in the subcuta- tioners think of when using the term electromyography or needle
neous tissue, the limb should be optimally positioned to en- EMG. The beginning practitioner may wish to divide the needle
hance complete relaxation. Flexion or extension of the major electromyographic examination into several subcomponents to
joint(s) associated with the muscle under investigation may be assist in mastering the procedure as a whole. One of the most
ardent proponents of this concept developed what the author

shall refer to as "Johnson's five steps to the needle examina-
tion." Doctor Johnson categorized the manner in which most
128
practitioners of the needle electromyographic examination per-

form most if not all of the test. The five steps are: (1) muscle at
rest, (2) insertional activity, (3) minimal muscle contraction, (4)
maximal muscle contraction, and (5) exploration (authors'
term). The author has slightly modified these five steps to arrive
at a slightly different step-wise manner in which to apply the
needle examination to patients with possible pathology: (1)
muscle at rest, (2) insertional activity, (3) minimal to moderate
isometric muscle contraction, (4) information synthesis, and (5)
impression formulation.
MUSCLE AT REST
Prior to inserting the needle electrode into the patient's muscle,
the instrument's amplifier settings must be optimized. The ampli-
fier's sensitivity is adjusted to 50 pV/div, while the sweep speed
is usually 10 m s / d i v . - Sensitivities of 100 pV/div can be
75128 218
used, but care must be taken as some potentials may be rather

small and, therefore, not detected. A sweep speed of 5 ms/div has
been recommended; however, for most routine needle examina-
tion, 10 ms/div is sufficient. Low filter settings of 10-30 Hz and
high filter settings of 10,000-30,000 Hz will avoid distortion of
most potentials likely to be detected. A 60-Hz notch filter should
be avoided if at all possible to preclude signal distortion. Figure 7-1. N e e d l e i n s e r t i o n . A needle recording electrode is in-
The patient is usually placed either supine or prone, depend- serted into four separate regions of muscle through one skin insertion
ing upon the muscle to be examined, and completely relaxed. site. Three successive depths (S) are sampled for each along each side
The individual muscle to be investigated is positioned to facili- of the pyramid. (From Cohen HL, Brumlik J: A Manual of Electro-
tate relaxation. If a monopolar needle is used, the reference neuromyography. New York, Harper & Row Publishers, 1968, p 40, with
permission.)
electrode is located relatively close to the site of insertion and a
ground is placed on the patient in a convenient location. A stan-
dard concentric needle only requires a separate ground as the is purposefully directed through the muscle to induce electrical
cannula serves as the reference electrode. The needle should be activity that is not present spontaneously. During this portion of
quickly inserted through the skin to minimize patient discom- the investigation, the instrument's settings remain unchanged
fort. As noted above, it may be helpful to spread the skin overly- and the needle is sequentially inserted through the muscle in
ing the insertion site with the hand not holding the needle to 0.5-2 mm increments with a several second pause between each
make the skin taut. Loose skin tends to be dragged by the needle insertion. The exact number of serial insertions along one direc-
and this impedes a quick and relatively painless insertion. As tion depends upon the muscle's thickness. Once the needle elec-
the needle passes through the subcutaneous tissue, little electri- trode has been positioned sufficiently deep in the muscle
cal activity is noted. A slight amount of resistance may be felt (several centimeters for limb and paraspinal muscles and sev-
when the connective tissue surrounding the muscle is encoun- eral millimeters for facial and sphincter muscles), the electrode
tered. A gentle thrust of the needle easily pierces this barrier and is withdrawn to the subcutaneous tissue and redirected along a
a burst of electrical activity (Insertional activity: see below) is different axis. This procedure is replicated multiple times until
typically seen signifying that the needle is now in muscle tissue. a somewhat pyramidal region of muscle has been examined
Healthy muscle is electrically silent at rest, provided the needle (Fig. 7-1). Note that only one skin insertion site is necessary to
electrode is not in the endplate region (see below). During this explore a relatively extensive portion of the muscle. Upon com-
portion of the needle examination, the practitioner attempts to pleting this portion of the investigation, it is good practice to ex-
observe spontaneous activity, i.e., electrical activity not under amine other regions of the same muscle through one or two
voluntary control or induced by the examiner. It is often helpful separate needle insertion sites, particularly if pathology is sus-
during this portion of the examination if the beginning practi- pected. Alternatively, the examiner performs still more redirec-
tioner releases the needle electrode to avoid subtle movements tions using the same insertion. In this respect it should be noted
that may inadvertently induce electrical activity. One can then that the viewing field of the needle electrode is very limited.
be certain that any observed electrical activity is either sponta- Healthy muscle tissue will produce bursts of electrical poten-
neous or under voluntary control (lack of relaxation) and not ar- tials (insertional activity) with each needle insertion that persists
tifactually induced by the examiner. Both normal and abnormal for only a short time following cessation of needle move-
muscle activity at rest will be discussed separately. m e n t . - The electrical potentials detected are believed to
232 234
result from the needle electrode mechanically depolarizing the

INSERTIONAL ACTIVITY muscle fibers it contacts while moving through the muscle.
These insertional potentials are also referred to as "injury poten-
The second step of the needle electromyographic examination tials." - The electrical activity associated with needle move-
232 233
is referred to as insertional activity because the needle electrode ment may fall in the normal range (normal insertional activity).
persist following needle movement (increased insertional activ- medicine consultation), data acquisition may alter the initial
ity), or barely be detected and possibly approach electrical si- plan of study. It is important for the practitioner to recognize
lence (decreased insertional activity). These three possibilities that the electrodiagnostic medicine examination is dynamic and
and their clinical significance will be explored separately. can flow in a number of directions as the investigation unfolds.
One may need to modify the order of the examination, number
MINIMAL TO MODERATE CONTRACTION/ of muscles studied, or number of limbs assessed at any point
RECRUITMENT during the consultation. It is impossible, therefore, for someone
other than the practitioner to simply perform "routine" tests and
Following the muscle at rest and insertional activity assess- present them for analysis at a later time. The electrodiagnostic
ment, it is then necessary to evaluate the electrical potentials medicine examination is very different from reading an electro-
generated by the voluntary activation of motor units. The 211
cardiogram in isolation from the patient.
patient is instructed to gently activate the muscle by just During the course of the electrodiagnostic medicine consulta-
"thinking" of a small contraction. The electrical potentials tion, needle electromyographic findings may begin to suggest a
produced by all of the single muscle fibers innervated by one particular diagnosis. The practitioner must be thoroughly famil-
nerve summate to form motor unit action potentials (MUAPs) iar with both neuromusculoskeletal anatomy and pathophysiol-
with characteristics that can be assessed such as morphology ogy to consider multiple avenues of exploration. Abnormalities
and r e c r u i t m e n t .
33131
A few MUAPs regularly firing on the found in a group of muscles must be characterized as belonging
CRT screen is the desired result. The MUAPs should not vary to a particular distribution suggestive of pathology affecting a
in their amplitude from one firing to the next. The instrument root, peripheral nerve, or a more generalized process for exam-
sensitivity may need to be reduced to 500 pV/div or less to vi- ple. The examination is complete when the acquired data c a n
sualize the entire potential on the screen. In the performance explain not only the patient's symptoms, but also any other ab-
of routine qualitative needle electromyography, one is primar- normalities uncovered during the electrodiagnostic medicine
ily looking for gross abnormalities of MUAPs such as: ampli- examination. The electrical data obtained from the needle elec-
tudes approaching 5 mV or more, MUAPs with more than tromyography (and nerve conduction studies) combined with
four phases (polyphasic MUAPs), MUAP durations abnor- the history and physical examination is synthesized, based upon
mally long or short, and recruitment disturbances. If the the physician's expertise, to formulate an electrodiagnostic
above abnormalities are suspected, then quantitative needle medicine impression.
electromyography is recommended to accurately assess and
quantify the MUAPs' parameters. IMPRESSION FORMULATION
The motor units activated at low levels of contraction are type
I and produce the MUAPs routinely analyzed. With increas-
35183
Once the data has been gathered and analyzed in the context
ing force of contraction, the initially present MUAPs increase of the patient's presenting complaint, an electrodiagnostic med-
their firing rate, and additional MUAPs are recruited. A moder- icine impression can be formulated. The impression may be
ate muscle contraction usually results in a significant number of simple (e.g., median nerve entrapment in the carpal tunnel) or
MUAPs on the CRT screen limiting individual MUAP analysis. complex (e.g., peripheral neuropathy combined with a radicu-
Moderate to maximal contraction is recommended by some in- lopathy) and should not only explain the patient's symptoms,
vestigators to evaluate the resulting overlap of multiple MUAPs but assist in the patient's management. This impression is a
known as the interference pattern. The information gained
128
result of the practitioner's expertise regarding the electric find-
from an interference pattern is minimal in the authors' opinion. ings and their association with specific disease entities affecting
The extent to which an individual can obliterate the CRT base- nerve and muscle tissue. The electrodiagnostic medicine practi-
line is dependent upon multiple factors: patient effort, pain, tioner's impression should also be supported with appropriate
ability of the examiner to resist the muscle's contraction, and recommendations regarding prognosis, and if requested, treat-
possibly other factors. It is the authors' belief that all of the nec- ment options.
essary information regarding MUAP morphology and recruit-
ment (see "MUAP Recruitment [Normal Muscle]" below) can
be evaluated at minimal to moderate muscle contraction for the ELECTRICAL ACTIVITY IN MUSCLE
majority of conditions. One exception to this statement would
be a type II fiber atrophy (e.g., steroid myopathy) in which nor- Performance of the needle electromyographic portion of the
mally large MUAPs would be absent at maximal contraction. electrodiagnostic medicine evaluation requires a prerequisite
This finding could not be observed if only minimal to moderate knowledge of muscle tissue electrophysiology. Action potential
contractions were investigated and is essentially the only time generation in a volume conductor with its associated extracellular
the authors evaluate maximal contraction. If one pursues
128
waveform morphology is the fundamental knowledge necessary
moderate to maximal muscle contraction, however, it is neces- to appreciate normal and abnormal needle electromyographic
sary to ensure superficial needle placement in the muscle-to findings. Once single muscle fiber waveform generation is un-
avoid bending the needle within the muscle and concomitant derstood, voluntary and spontaneous (nonvoluntary) normal and
tissue trauma. 112
pathologic electrical activity arising from single muscle fibers
can then be appropriately interpreted. Mastery of this information
INFORMATION SYNTHESIS forms the foundation upon which to formulate an appropriate
diagnosis, prognosis, and treatment plan. The electrophysiology
The needle electromyographic investigation is only one part of of muscle and single muscle fiber action potential morphology
the electrodiagnostic medicine evaluation and is directed by the in volume conductors will be briefly reviewed below. T h e
history and physical examination. As one proceeds through the reader is referred to Chapter 2 for a more detailed explanation
needle examination (and all other portions of the electrodiagnostic of volume conductor theory.
Direction of Travel Although relatively impermeable, small quantities of N a do +
leak into the cell and alter the resting membrane potential
slightly. A steady-state between N a entering and K exiting a + +
cell with a constant resting membrane potential is maintained

by the sodium-potassium p u m p . The steady state of the rest-
13130
ing membrane potential's voltage approaches the negative equi-

librium potential of K with a slight modification due to the
+
influence of the positive Na equilibrium potential.+
Muscle cells are known as excitable tissues because they can

sustain an induced self-propagating action potential. If the rest-
ing membrane potential is depolarized from - 8 0 mV to about
- 5 0 to - 6 0 mV, threshold level, following an appropriate stimu-
lus, an action potential is produced (Fig. 7-2A). - - The 13 21 130
action potential arises from an alteration in voltage-gated chan-

nels controlling the permeability of N a and K . At threshold, + +
the sodium voltage-gated channels open allowing N a to enter +
Potassium Eft the excitable tissue and further depolarize the cell thereby in-
ducing more sodium gates to open in a positive feedback loop
that approaches the equilibrium potential of Na (+55 mV). This +
process of depolarization lasts about 0.5 ms in mammalian

muscle cells and propagates along the muscle fibers constituting
an action potential. - The cell's resting membrane potential
174 175
depolarizes from - 8 0 mV to approximately +40 m V . A de- 130
layed increase in K permeability lasting about 2.0 ms in mam-

+
malian muscle coupled with inactivation of the sodium gates

subsequently restores the resting membrane potential back to
-80 mV. il4J74
- The cell, therefore, produces a positive mono-
222
phasic intracellular action potential beginning at - 8 0 mV, in-

creasing to about +40 mV, and returning to - 8 0 mV (Fig. 7-2A). 213
The action potential's appearance as characterized by its ampli-

tude, duration, rapid rise and relatively slow baseline return is
Figure 7-2. A, Intracellular action potential arising from an increase directly dependent upon the temporal aspects of Na and K ac- + +
in sodium (gNa ) and potassium (gK ) conductance. B, Region of

+ +
tivation and inactivation. 169
t r a n s m e m b r a n e ion flow associated with opening of voltage-depen-

dent ion gates. C, Local circuit currents generated by the opening of Local Circuit Currents
sodium gates. D, Extracellular triphasic single muscle waveform gener- The impermeable protein anions located within excitable tis-
ated by the monophasic intracellular action potential. sues attract extracellular positive ions, primarily Na because of +
its abundance, and cause them to align along the extracellular

ELECTROPHYSIOLOGY OF MUSCLE surface of that cell's membrane. In effect, there is a series of
negative and positive charges (dipoles) along the length of the
Action Potential Generation cell separated by the plasma membrane (Fig. 7-2B). During the
All living cells possess the unique characteristic of a resting course of an action potential, inwardly flowing positive N a pro- +
membrane potential with the intracellular region electrically ceed bi-directionally along the axon's interior. These additional
negative (approximately - 8 0 mV for muscle tissue) compared intracellular positive charges balance a portion of the intracellu-
to the extracellular space. - The cell's membrane is respon-
13110 222
lar anions attracting the extracellular Na . The transmembrane +
sible for developing and sustaining a transmembrane voltage electrical attraction for the Na is subsequently reduced, allow-
+
difference because of its semipermeable nature. Most cells con- ing these ions an increased degree of freedom to move away
tain a high intracellular concentration of positive potassium ions from the immediate vicinity of the membrane. The increased
(cations) and negative protein ions (anions). In the resting state, permeability of the depolarized membrane region with its open
the semipermeable membrane allows intracellular potassium sodium gates permits those more mobile Na to now enter the +
ions (K ) to exit the cell down a high intracellular to low extra-

+
cell through the open Na channels. This sequence of events is
+
cellular concentration gradient. This process continues until

13130
referred to as a local circuit current; it specifically encompasses
it is balanced by an inward electrical attraction from the intra- extracellular N a entering the cell and neutralizing the trans-
+
cellular protein anions that cannot exit the cell. A net negative membrane attraction toward additional extracellular Na* thereby
charge results when a balance is established between the forces allowing these ions to move away from the membrane's surface
driving K out of the cell (concentration gradient) and an in-
+
and also enter the cell through the open sodium c h a n n e l s . ° 13J3
wardly directed force (electrical gradient) from the anions at- The extracellular Na that are no longer tightly bound to the mem-
+
tracting the K into the cell. The resting membrane potential is

+
brane's surface can now enter the cell through the local circuit cur-
also influenced to a small degree by the relatively impermeable rent and are referred to as a current source. The region of membrane
positive sodium ions (Na ). There are two strong forces driving
+
through which the sodium ions enter the cell (open Na voltage- +
sodium into the cell. The first is the concentration gradient, high dependent gates) is called a current s i n k . The source current 13130
extracellular and low intracellular N a concentration, and the

+
will tend to depolarize the adjacent membrane regions surround-
second is the negative intracellular potential or electrical gradient. ing the current sink by reversing the intracellular-to-extracellular
Chapter 7 NEEDLE ELECTROMYOGRAPHY — 263
Figure 7-3. Isopotential lines gen-

erated by an action potential in a
volume conductor.The isopotential lines
associated with the central current sink
(negative region between O) and 0 ) 2
and t w o surrounding c u r r e n t sources

(positive regions) form a potential gra-
dientThe two zero isopotentials 0 and(
Electrode
position
0 define the negative current sink and
2
positive current sources.The triphasic

waveform detected by an electrode
passing through the action potential's Leading current source Current sink Trailing current source
corresponding field lines is depicted
below. (Modified from Lorente de N o
R: Analysis of the distribution of action
currents of nerve in volume conductors. 1
Stud Rockefeller Inst Med Res 1947;
132:384-477.)
i I I \ I
voltage relationship. The self-sustaining nature of these processes negative sink, but still within the region of the current source, now
results in action potential propagation. detects isopotential lines with less positive magnitude than pre-
The current sink from a propagating action potential permits viously measured. The CRT trace demonstrates a corresponding
positive sodium ions to extend not only into that portion of less positive deflection and begins heading toward the baseline.
nerve or muscle previously depolarized, but also into the region When the first zero isopotential (O Fig. 7-3) line demarcating
l
of the tissue about to be depolarized. Extracellularly, the sodium the leading current source and sink is reached, the CRT trace
ions are moving into the sink from portions of the volume con- crosses the baseline, denoting zero potential difference.
ductor both preceding and following the traveling action poten- Placing the electrode just to the right of the first zero isopoten-
tial. In effect, there are two local circuit currents associated with tial line (Fig. 7-3) now allows it to observe isopotential lines of
the current s i n k . This observation leads to the common de-
13130
rather low magnitude and negative potential. The CRT trace then
scription of an action potential consisting of a "source-sink- proceeds above the baseline now describing a negative potential
source" (+ - +) arrangement. 194
(sodium activation/depolarization). As this region contains a
high spatial gradient of field lines, slight electrode progression
Local Circuit Model of W a v e f o r m Morphology measures field lines of rapidly increasing negativity (Fig. 7-3).
Lorente de N o provided us with a detailed mapping of the
155
The CRT trace quickly rises above the baseline. Again, as the
field lines surrounding a propagating nerve action potential in a middle region of the negative sink is reached, there is a region
volume conductor (Fig. 7-3). For this discussion, the similari- where little change in the isopotential magnitude occurs over
ties between local circuit currents of nerve and muscle are more time and the CRT trace now defines the second inflection point
alike than different. The primary distinction between muscle or maximum negative amplitude (Fig. 7-3). Continued progres-
and neural tissue is that depolarization/repolarization is approx- sion allows the electrode to measure isopotential lines with a
imately 5 times longer in muscle. - - - - - - To assist in
89 90 113 114 156 174 175
declining negative magnitude that is reflected in the CRT trace
the understanding of extracellular waveform morphology gen- returning to baseline (sodium inactivation and potassium activa-
eration, we may assume that an action potential and the isopo- tion: repolarization). The second baseline crossing occurs when
tential lines (regions of equipotential extending into the volume the electrode encounters the second zero isopotential line ( 0 2
conductor) associated with the local circuit currents on the sur- Fig. 7-3), noting the negative sink's terminal portion.
face of a single muscle fiber are stationary or "frozen" in time Upon entering the trailing current source (positive region),
(Fig. 7-3). - Note the extracellular potential field lines consti-
5 76
isopotential lines of increasing positivity produce a second
tute a central current sink (negative) flanked by a leading and downward deflection of the CRT trace below the baseline (Fig. 7-3).
trailing current source (positive) for the familiar tripole (+ - +). - 6 71
The waveform's third inflection point is described as the rela-
A recording electrode can then be sequentially moved through tively extended region of positive potential and demarcates the
the isopotential field lines associated with the muscle's action maximum amplitude of the potential's third phase. Note that the
potential. Initially (Fig 7-3, far left), the electrode encounters spatial gradient is rather low and spreads out over a compara-
positive isopotential lines with rather low voltage and low spa- tively longer spatial extent than the initial source and subsequent
tial gradient (lines per unit distance) of the leading current sink (Fig. 7-3). The amplitude of the third phase is correspond-
source. The positive potential results in a downward (positive) ingly smaller and longer in duration. Finally, as the electrode
motion of the cathode ray tube (CRT) trace (Fig. 7-3). Con- reaches the terminal portion of the trailing current source, posi-
tinued electrode movement detects isopotential lines of increas- tive isopotential lines of low magnitude result in the CRT trace
ing positive magnitude and the CRT trace proceeds in the more returning to baseline. The net result of this negative current sink
positive direction. At about the middle of the leading current flanked by two current sources in a volume conductor such as
source, the isopotential region is now rather constant and the the body is a triphasic potential (Figs. 7-2D and 7-3). Single
CRT trace maintains the same level for a relatively brief time muscle fibers should, therefore, produce triphasic waveforms in
describing the waveform's first inflection point and demarcat- good volume conductors such as the human body. An alterna-
ing the maximum amplitude of the first or initial positive de- tive and more detailed explanation of action potential morphol-
flection. Displacing the recording electrode still closer to the ogy in volume conductor is provided in Chapter 2.
Table 7-1. Electrical Potentials

1. Insertional Activity
A. Normal
B. Increased
C. Decreased
II. Spontaneous Activity
100/xV
A. Muscle generator B. Neural generator
1. Fasciculation 1. Fasciculation
2. Fibrillation 2. Myokymic discharge
3. Positive sharp wave 3. Continuous motor unit activity
4. Myotonia 4. C r a m p
5. Complex repetitive 5. Tremor
discharge 6. Multiplet
Voluntary Activity
J
A. Normal MUAPs
200/xV
B. Polyphasic MUAPs
C. MUAPs with increased/decreased duration
50ms
D. MUAPs with increased/decreased amplitude
E. Multiplet MUAPs
F. MUAPs with variable amplitudes
G. Abnormal MUAP recruitment
ELECTRICAL POTENTIALS IN MUSCLE

NORMAL POTENTIALS IN MUSCLE
A number of distinct electrical potentials can be observed
when a needle recording electrode is placed in muscle tissue.
Electrical potentials detected in muscle can be categorized into
three main types and several subcategories: (1) insertional activ-
ity, (2) spontaneous activity, and (3) voluntary activity (Table 7-1).
Recognition of these potentials coupled with an understanding
of their pathophysiologic basis will assist the clinician in for-
mulating an electrodiagnostic medicine impression.
NEEDLE INSERTIONAL ACTIVITY Figure 7-4. Insertional activity. A, Normal insertional activity re-
sulting from a brief needle electrode insertion. B, Decreased inser-
Placing a needle (monopolar or standard concentric) recording tional activity noted in fibrotic muscle tissue. C, Increased insertional
electrode into healthy muscle tissue and advancing it in quick, activity as routinely described by m o s t practitioners. A large burst of
but short intervals results in brief bursts of electrical potentials electrical activity is associated with needle movement immediately fol-
(Fig. 7-4A). A crisp sound is associated with needle movement lowed by florid positive sharp waves and fibrillation potentials that
as a series of negative and positive spikes are produced. These eventually result in a quiet baseline over the course of several hundred
waveforms are referred to as insertional activity and this term milliseconds t o seconds.
was first introduced into the standard needle electromyographic
examination by Weddell et al. The observed electrical activity
231
is the basis for the synonymous term of "injury potentials." The
is believed to result from the needle electrode mechanically de- duration of insertional activity was quantified in the needle ex-
polarizing the muscle fibers surrounding its leading edge as it amination performed by clinicians practicing electrodiagnostic
pierces and pushes aside muscle tissue. Minimal and localized
140
medicine. The total time of activity, from the initiation of
233
muscle tissue damage may occur from direct needle trauma and electrical potentials and extending beyond needle movement
cessation, is less than 230 ms for monopolar needle electrode
Table 7-2. Insertional Activity 154
insertion distances of 1-5 mm. The time of electrical activity
persisting after needle cessation has a mean of 48 ± 18 m s . 232-234
Type Duration Shape Etiology

The total time of insertional activity for standard concentric
Normal < 300 ms Spikes Muscle depolarization needles is reported to be less than 300 ms (Table 7-2). 109
Increased > 300-500 ms Spikes Normal variant The purpose of including insertional activity analysis as part
Positive waves Normal variant of the electromyographic examination is that the probing needle
Denervation may provoke transient and/or sustained membrane instability
Myopathy (fibrillation potentials and positive sharp waves: see below) prior
to these potentials being present at rest. Wiechers et al. clearly
232
Decreased Absent o r Absent/spikes Fat

demonstrated in a denervated animal model that positive sharp
< 300 ms Fibrosis
waves are the first sign of membrane instability followed by fib-
Periodic paralysis
rillation potentials. This early membrane instability resulted only
from needle insertion while the denervated muscle was quiet at A — — —

rest. These findings experimentally substantiated the clinical
impression that needle probing can induce membrane instability
prior to it being spontaneously present at rest. B
A form of "normal" insertional activity described as "snap,
crackle, and pop" has been reported in healthy young men pri-
marily in the gastrocnemius muscle. These potentials persist
237
for a variable length of time after needle movement ceases.

Another apparently normal variant of insertional activity con-
sists of runs of positive sharp waves without or with only a very c
few fibrillation potentials. - - The positive waveforms are
I73 235 240
induced by needle movement, eventually disappear, have vari-

able abundances depending upon the muscle examined and time
of investigation, and demonstrate an autosomal dominant inher-
itance pattern. Given these findings, one must conclude that the
observation of positive sharp waves does not always signify
pathology or denervation. The terms "increased insertional ac-
D
tivity" and "decreased insertional activity" have been used as
qualitative criteria to indicate muscle pathology; however, the
existence of increased insertional activity has been questioned
(see below),
Figure 7-5. Endplate activity. A, Monopolar needle electrode lo-
SPONTANEOUS ACTIVITY cated in relaxed muscle tissue. B, Miniature endplate potentials (MEPPs)
recorded with t h e same electrode needle in an endplate zone.These
We may define spontaneous activity as electrical waveforms monophasic negative potentials primarily present as baseline irregular-
not under voluntary control. Placing a needle in healthy ities because of their small amplitude. N o t e t h e change in the baseline
muscle tissue at rest usually results in complete electrical si- between traces A and B. C, Endplate spikes recorded only within t h e
lence, provided the needle is not located in the endplate muscle's endplate region.These a r e biphasic initially negative single
region. The endplate is that specialized portion of a single muscle fiber action potentials recorded from their site of origin, hence
muscle fiber where the terminal axon and muscle fiber form a t h e initial negative deflection. D, Slight repositioning of the needle elec-
neuromuscular junction. Two waveforms, miniature endplate t r o d e results in t h e detection of both endplate spikes and MEPPs.
potentials (MEPPs) and endplate spikes, may be concomi-
tantly or independently observed with a needle electrode in the be observed despite a documented firing frequency of 0.2 H z
endplate portion of muscle tissue. Neither of these potentials because the size of the recording electrode may straddle several
can be elicited following denervation because of Wallerian de- endplates simultaneously and one actually observes the electri-
generation of the terminal axon and its presynaptic portion of cal activity resulting from multiple endplates spontaneously re-
the endplate. - - -
20 21 200 236
leasing ACh. Not infrequently, the observation of MEPPs is
associated with a complaint of increased pain from the patient.
Miniature Endplate Potentials Because of the uncomfortable sensations accompanying needle
(Monophasic W a v e f o r m ) insertion in the endplate, it is recommended to avoid this area of
As noted above, an electrode located in the endplate region the muscle. If the needle electrode is found to be placed in this
can record two distinct potentials. One of the observed poten- region, a quick thrust of the needle electrode through the end-
tials may be a short duration (0.5-2 ms), irregularly occurring plate zone eliminates discomfort and interpretive errors. Of
(once every 5 seconds per axon terminal), small (10-40 pV), note, following complete denervation of muscle tissue, MEEPs
monophasic negative waveform (Fig. 7-5, Table 7-3). These 32
disappear because the endplate degenerates. If MEPPs are ob-
potentials are monophasic and negative because the subthresh- served following sufficient time for axons to completely degen-
old depolarizations do not propagate away from the electrode erate (8-10 days after nerve transection), the muscle is likely
(see Chapter 2). The examiner is usually made aware of MEPPs not completely denervated.
by an abrupt increase in the irregularity of a previously quiet
CRT baseline. Table 7-3. Endplate: Normal Spontaneous Activity
A needle recording electrode placed in proximity to the
MEPPs Endplate Spikes
muscle's endplate region is believed to record the spontaneous
release of acetylcholine (ACh) from the presynaptic nerve ter- Morphology Monophasic negative Biphasic negative/positive
minal. - Once the ACh binds with the acetylcholine receptdrs
89 91
Firing pattern Irregular Irregular
on the muscle, opening of the sodium and potassium gates re- Amplitude 10-50 uV 100-200 uV
sults in a small current flow. This current flow is detected by the
Duration 0.5-2.0 ms 3.0-4.0 ms
extracellularly located needle electrode. The needle electrode
Origin site Endplate Single muscle fiber depolariza-
may also "irritate" the endplates directly to release additional
tion from needle electrode
quanta of ACh, thus increasing the frequency of observed
generating a suprathreshold
MEPPs. The firing rate of MEPPs is irregular as the ACh re-
20
endplate potential
lease is spontaneous and random. The sound associated with
this random release of ACh is a high-pitched hiss called end- Denervation Disappear Disappear
plate noise or seashell murmur. A large number of MEPPs may
236
Modified after Brown 21
for MEPPs, the irritation produces an irregular release of acetyl-

choline and hence endplate spike firing is also very irregular.
END-PLATE
Once again, the location of the needle electrode is associated
ZONE
.050 mV with patient discomfort, and pushing the needle through this
area usually results in considerably less pain. It is possible to
mistake endplate spikes for fibrillation potentials (see below),
which are also single muscle fiber discharges. The highly irreg-
AWAY
ular firing rate of endplate spikes usually distinguishes these po-
tentials from fibrillation potentials. The cannula's tip of a
JL
standard concentric needle electrode placed in the endplate zone
100 200 may produce endplate spikes with an inverted polarity (bipha-
sic: positive/negative; Fig. 7-6) possibly leading to confusion
T I M E (ms)
with positive sharp waves (see below). Again, the highly ir-
1282 1 7
Figure 7-6. E n d p l a t e e l e c t r o d e a r t i f a c t . MEPPs recorded with a regular firing rate should alert one to the possibility of endplate
standard concentric needle electrode ( t o p trace) may b e c o m e in- spikes. As previously stated, the needle recording electrode
verted with respect t o polarity if these same potentials are preferen- should be relocated if it is in the endplate zone to avoid patient
tially recorded with t h e needle's cannula (reference electrode). (From discomfort and waveform morphology confusion. Endplate
Brown WF:The Physiological and Technical Basis of Electromyography. spikes will also not be detected following denervation for simi-
Boston, Butterworth, 1984, pp 317-368, with permission.) lar reasons to those given for MEPPs.
Several investigators have suggested that endplate spikes may
Endplate Spikes (Biphasic Negative/Positive arise from the intrafusal muscle fibers (muscle spindles) inner-
Waveform) vated by gamma motor neurons. Although it is certainly pos-
179
A needle recording electrode situated in the endplate region sible that intrafusal muscle fibers may be a source of
can detect a second spontaneous potential with a different mor- spontaneous electrical activity, it is doubtful that they produce
phology than the MEPR This second potential is biphasic with an the commonly encountered endplate spikes associated with
initial negative deflection and referred to as an endplate spike (see MEPPs. The reasoning for this conclusion is that endplate
Chapter 2 for a more complete discussion of this potential's mor- spikes and MEPPs are frequently found in combination. There 20
phology). Compared to the MEPP, the endplate spike is longer is no question that MEPPs are generated in the endplate. Slight
in duration (3-5 ms), of moderate amplitude (100-200 pV), and needle electrode movement typically produces endplate spikes,
fires irregularly (Fig. 7-5, Table 7-3). It is important to note that
32
strongly suggesting that both potentials are produced from the
endplate spike potentials may also have a triphasic initially pos- same location, i.e., extrafusal and not intrafusal muscle fibers.
itive shape, a biphasic initially positive morphology and even Also, investigators have not detected an alteration in the firing
rather complex configurations resembling a combination of the of endplate spike potentials following passive stretch or volun-
two previously noted waveforms.The first of these less proto- tary activity which should occur if the muscle spindle produces
typical appearances may be mistaken for a fibrillation potential endplate spikes. 21
while the second may be misinterpreted as a positive sharp

wave. Additionally, on rare occasions irritation of a single or a
80
few terminal nerve sprouts may induce the terminal nerve's MUSCLE GENERATORS OF NORMAL
parent motor unit to fire, revealing an irregular firing motor unit VOLUNTARY ACTIVITY
potential. The key to properly interpreting these waveforms is
paying particular attention to the irregular firing rate. SINGLE MUSCLE FIBER (TRIPHASIC WAVEFORM)
The endplate spike is thought to be a suprathreshold single
muscle fiber depolarization that propagates away from the Depolarization of a single muscle fiber results in a triphasic
recording electrode. It originates from the endplate zone and re- extracellular waveform when recorded outside of the endplate
sults from the recording needle electrode irritating the single zone (Fig. 7-7). - This triphasic potential may appear biphasic
64 87
muscle fiber's terminal axon or the presynaptic terminal (negative/positive) if the needle recording electrode is located
itself. - This irritation produces substantially more acetyl-
32 237
within the endplate region. A biphasic positive/negative
208
choline release than in the generation of MEPPs resulting in a single muscle fiber potential may also be observed under three
suprathreshold depolarization of the muscle fiber. Just as 20236
conditions. The first condition producing a biphasic posi-
tive/negative single muscle fiber potential arises when the termi-
nal positive phase is rather small and lost in the baseline noise of
the recording. In reality, the potential is triphasic, but the third
87
200fN
phase is prolonged and small enough to be mistaken as part of
the baseline irregularity. A recording electrode within the mus-
culotendinous region is the second condition where an initially
positive biphasic single muscle fiber potential can be recorded. 129
Finally, if the cannula's tip of a standard concentric needle elec-

trode is placed in the endplate zone, a suprathreshold initially
negative biphasic potential following endplate depolarization is
inverted to produce a positive/negative w a v e f o r m .
128135
A de-
Figure 7-7. Two triphasic single m u s c l e fiber w a v e f o r m s belong- tailed volume conduction description of the above potentials is
ing t o t h e same m o t o r unit recorded from a normal extensor digito- provided in Chapter 2. A separate component of the electrodiag-
rum communis muscle. nostic medicine consultation which exclusively focuses on the
single muscle fiber action potential, single fiber electromyogra-

phy (SFEMG) will be discussed later (see Chapter 8).
MOTOR UNIT ACTION POTENTIALS (MUAPS)

The electrical activity typically recorded by a needle elec-
trode placed in a voluntarily contracting muscle is the summa-
tion of action potentials resulting from single muscle fibers
innervated by one anterior horn cell. This summated electrical
activity gives rise to electrical waveforms called motor unit
action potentials (MUAPs). Prior to detailing the characteristics
of the MUAP, the anatomy and physiology of the motor unit
must be understood.
Anatomy
The concept of the motor unit was introduced by Sherrington Figure 7-8. Two m o t o r units (type I and II) a r e depicted. N o t e
over 60 years ago. The motor unit is the fundamental or quan-
202
how fibers from o n e m o t o r unit are interspersed with those from an-
tal element which forms the foundation upon which the nervous o t h e r m o t o r unit. (From O h SJ: Clinical Electromyography: N e r v e
system exercises control over muscle tissue to produce move- Conduction Studies, 2nd ed. Baltimore, Williams & Wilkins, 1993, with
permission.)
ment. The motor unit is defined as one anterior horn cell (mo-
toneuron), its peripheral nerve, and all of the individual muscle
fibers innervated by that individual motoneuron (Fig. 7-8). innervated by only one motoneuron, i.e., there is no polyneu-
There are three types of motor neurons identified: alpha (skele- ronal innervation in adult humans. Occasionally two neuromus-
tomotor) motor neurons, beta (skeletofusimotor) motor neurons, cular junctions may be observed on a single muscle fiber;
and gamma (fusimotor) motor neurons. The alpha motor neu-
161
however, both endplates originate from the same motor axon. It
rons consist of large cell bodies and fast conducting axons that is possible to estimate the number of muscle fibers innervated
only innervate the skeletal (extrafusal) muscle fibers. Alpha by one motoneuron, i.e., the innervation ratio. 35
motor neurons are located in the ventrolateral horn (lamina IX, The innervation ratio can be calculated by painstakingly
Rexed classification) of the gray matter within the spinal counting the total number of muscle fibers in a particular
cord. The motor neurons innervating a particular muscle are
192
muscle and dividing by the number of large motor axons inner-
clustered into vertical columns extending for a finite distance vating that muscle. The number of large motor axons in a p e -
over one spinal segment. Additionally, the motor neurons inner- ripheral nerve associated with a muscle is approximately 5 0 %
vating the limbs are situated relatively laterally in the spinal with the remainder primarily consisting of afferent fibers. - - 81 82 239
cord's cervical and lumbar enlargements compared to the motor Muscle fiber numbers in motor units vary within one anatomi-
neurons innervating the neck and trunk muscle. It has been
14
cally defined muscle and between different muscles. Generally,
demonstrated that the cell soma of the alpha motoneuron is
completely covered with synaptic connections to other cells
except for the axon hillock which gives rise to the axon. Axon
collaterals can also arise from the alpha motoneuron to synapse
with Renshaw interneurons and other alpha motor neurons. 82
The alpha motor neurons exhibit a two-fold variation in cell

body size from approximately 40-80 pm (Fig. 7 -9 ) - - - 43 44 45 46 3
Gamma (fusimotor) motor neurons are smaller cells (20-40

pm) located in the ventral gray matter of the spinal cord and in-
nervate the muscle spindle (intrafusal muscle fibers) stretch re-
ceptors (Fig. 7-9). - ^ The gamma motor neurons are not only
41 43 6
about half the size of the alpha motor neurons, but also give rise o
to axons that are smaller and conduct comparatively more
slowly. The beta (skeletofusimotor) motor neurons innervate
46
both intra- and extra-fusal muscle fibers. Because the needle E

3
electromyographic examination records only the voltage associ-
ated with depolarization of the extrafusal muscle fibers, this dis-
cussion will focus only on the alpha motor units.
Skeletal muscle consists of muscle fiber bundles surrounded
by connective tissue to form a fascicle. There are approximately
20-60 muscle fibers per fascicle and 2 - 3 motor units within a
fascicle. One motor unit may be distributed over 100 fascicles.
31 Average Soma Diameter (pm)
A single anterior horn cell, however, innervates muscle fibers
Figure 7-9. Histogram of average soma diameter innervating
contained only in one muscle such that one motor unit does not the medial gastrocnemius (heavy line; open area) and soleus (shaded
extend between anatomically distinct muscles. - Upon reach-
35 46
area) of t h e cat. N o t e t h e bimodal distribution of t h e t w o m o t o r

ing a skeletal muscle the motor axon begins to subdivide repeat- neuron pools. (From Burke RE, Strick PL, Kanda K, e t al: Anatomy of
edly to form one neuromuscular junction for each of the muscle medial gastrocnemius and soleus m o t o r nuclei in cat spinal cord. J
fibers contained within its motor unit. A single muscle fiber is Neurophysiol 1977;40:667-680, with permission.)
Table 7-4. Motor Units in Human Muscle

Muscle No. of Muscle Fibers/ Motor Unit
Motor Units Motor Units Territory (mm) 31
External rectus 96
2,970 9
Tensor tympani 226
55 20
Platysma 96
1,096 25
Cricothyroideus 96
112 41
First lumbrical 96
96 108
FDI 96
119 340
Brachioradialis 96
330 390
Biceps brachii - 35 54
774 750 5.1
Deltoid 35
6.7
EDC 35
5.5
Opponens pollicis 35
7.4 Figure 7-10. Cross-section of a rat anterior tibial muscle.
Tibialis anterior - 54 96
445 562 7.Q N o t e t h e random distribution of muscle fibers belonging t o one
m o t o r unit (lighter fibers-glycogen depletion technique: PAS) and how
Rectus femoris - 35 54
1,000 305 10.0
they are interspersed with muscle fibers from o t h e r m o t o r units.
Gastrocnemius 96
579 1,934 (From Kugelberg E: Properties of t h e rat hind-limb m o t o r units. In:
EDB 35
11.3 Desmedt JE (ed): New Developments in Electromyography and Clinical
Neurophysiology. Karger, Basel, 1973, pp 2 - 1 3 , with permission.)
Abbreviations: APB = abductor pollicis brevis; FDI = first dorsal interosseus;
EDC = extensor digitorum communis; EDB = extensor digitorum brevis.
glycogen is depleted yields interesting results. - All of the
12 84
motor units that control large muscle groups requiring little dex- muscle fibers belonging to that axon (motor unit) can then be
terity (e.g., gastrocnemius: 2,000 fibers/motor unit) "have more histologically identified through the periodic-acid Schiff re-
muscle fibers per motor unit compared to small, highly dexter- action (PAS) in which muscle tissue depleted of glycogen
ous muscles (e.g., superior rectus: 23 fibers/motor unit). - 31 32
stains differently and subsequently "stands out" from the sur-
Data concerning the size of motor units have been acquired in rounding muscle tissue (Fig. 7-10). In essence, one motor
humans (Table 7-4). unit's muscle fibers can be localized within a muscle. This
In addition to the anatomic technique noted above for esti- technique revealed that the muscle fibers of a single motor
mating the number of muscle fibers per motor unit, it is also unit are randomly distributed over a circular region approach-
possible to electrophysiologically count the number of motor ing 2 0 - 3 0 % of the muscle's total volume. Although the dis-
84
units in a m u s c l e . This incremental motor unit counting

49,162163
tribution of motor unit muscle fibers is rather uniform, there
technique is performed by stimulating a peripheral nerve and is a slightly higher density of fibers near the center of the
recording from the endplate zone of a skeletal muscle (see motor unit. - - - Fibers from 2 0 - 5 0 different motor units
28 31 35 84
Chapter 8). The stimulus intensity is gradually increased from a may share this same volume of muscle. Additionally, there
43
subthreshold level. At some minimal value, the stimulus evokes are few muscle fibers of the same motor unit lying side by
a small compound muscle action potential. A further increase in side.
the stimulus intensity results in a slightly larger response. This The physical distribution of single muscle fibers associated
process is continued and incremental "jumps" in the amplitude with one motor unit can be investigated not only histologically
of the response are observed. It is assumed that each sequential but also electrically. A single needle electrode with 12 or more
increase in amplitude results from the activation of an additional individual leads along its length can be inserted into a skeletal
motor unit/motor axon. When a supramaximal response is ob- muscle, e.g., biceps brachii. - It can then be positioned during
27 28
tained, the mean amplitude of the incremental responses is cal- minimal contraction to optimally record the electrical activity
culated. The maximal compound muscle action potential originating from the single muscle fibers of one motor unit.
amplitude is then divided by the mean incremental amplitude to Because the distance between individual leads is known, it is
estimate the total number of motor units/motor axons associated possible to estimate the distance across a motor unit by noting
with the muscle examined. A number of muscles have been ex- which leads record time-locked electrical spikes. Upper limb
amined with this method and comparable results with histologic muscles demonstrate roughly circular to oval motor unit terri-
techniques have been reported. The incremental motor unit 163
tories of 5-7 mm, while lower limb muscles have slightly
counting method, however, has been criticized. - Sources 1819 95176
larger motor unit territories of 7 - 1 2 mm (Table 7-4). These 31
of error and limitations of the technique include (1) variable findings are compatible with the motor unit distribution as
morphology and amplitude of the compound muscle action po- measured by histologic methods. Electrical studies indicate
tential, (2) excitability variations reduce the specificity of dif- that the muscle fiber density of a single motor unit is somewhat
ferent axon activation thresholds, ( 3 ) axons with identical greater in the center of the circular territory than near the pe-
thresholds diminish the orderly addition of sequentially acti- riphery. This finding has also been substantiated by the glyco-
vated motor units, and (4) the incremental amplitude of the re- gen depletion method.
sponses may not be a representative sample of the muscle's
motor units. Physiology
Repetitively activating a single motor axon with an electri- The muscle fibers comprising a single motor unit can be stud-
cal stimulus for a prolonged period of time such that its intrafiber ied anatomically and physiologically. It is possible to stimulate
FLEXOR HALLUCIS
LONGUS
6.3 Hz 10 Hz
500 ms
L. Figure 7-12. Tetanic tension generation at increasing frequen-

100 ms cies of muscle excitation. As t h e stimulation frequency increases, t h e
separate muscle contractions fuse into a rapid and smooth increase in
Figure 7-11. Isometric twitch times recorded from fast-twitch tension, i.e., tetanus. (From Buller AJ:The Physiology of the motor unit.
(flexor hallycis longus) and slow-twitch (soleus) muscles of t h e cat. In: Walton JN (ed): Disorders of Voluntary Muscle, 3rd ed. Edinburgh,
T h e s e t w o muscles clearly d e m o n s t r a t e t h e obvious difference in Churchill Livingstone, 1974, pp 20-30, with permission.)
twitch times between t w o anatomically distinct muscle. (From Buller
AJ:The Physiology of t h e m o t o r unit. In:Walton JN (ed): Disorders of
Voluntary Muscle, 3rd ed. Edinburgh, Churchill Livingstone, 1974, pp
20-30, with permission.) muscle demonstrates that "slow" muscles tend to be red while
"fast" muscles are more pale or white in appearance. T h i s 204
color distinction is primarily because of the myoglobin c o n -

a single motor axon and record the time required for its muscle tent of the tissue. Red muscles have a significantly higher
fibers to reach their peak isometric tension, i.e., the twitch time myoglobin content than white muscle. This finding supports
(Fig. 7-11).38.39.4i.24i T j technique has revealed that motor
n s the concept that motor units may have different physiologic
units (their muscle fibers) can be classified into two major cat- properties.
egories: fast-twitch and slow-twitch. The slow-twitch motor In addition to the twitch time of motor units, it is also possi-
units generate small forces and are innervated by slow conduct- ble to investigate their resistance to fatigue using a sustained
ing nerves compared to fast-twitch motor units which produce stimulus at high rates. For our purposes, fatigue may be defined
relatively more force and are innervated by faster conducting as: " . . . the condition in which muscles fail to maintain tension,
nerves. 41241
Fast and slow are relative terms as the speed of or fail to develop constant tension, in response to stimulation at a
contraction varies from muscle to muscle and from species to constant frequency." Normal muscle responds in a stereotypical
204
species. Specifically, in the mouse a fast-twitch motor unit way to repetitive stimulation. By securing the ends of a muscle to
reaches its peak tension in 6.0 ms while in man a fast motor fixed points, its tension can be measured when the innervating
unit achieves peak tension in 35.0 m s . On the other hand, a 204
nerve is excited. - At stimulation frequencies of < 5 Hz, indi-
38 39
mouse slow-twitch muscle requires 20.0 ms to reach peak ten- vidual muscle contractions can be noted with respect to tension
sion but a human slow-twitch muscle takes 90.0 ms for peak (Fig. 7-12). At rates exceeding 5 Hz, subsequent twitches sum-
tension. Although human muscle usually displays a continuum mate to result in a greater maximum tension. When the stimula-
of twitch times, muscle nevertheless continues to be classified tion frequency reaches 20 Hz or more, the individual responses
into two general categories with respect to twitch times, i.e., fuse to form a tetanus, i.e., a smooth tension increase reaching a
fast-twitch and slow-twitch. - The speed of contraction for
46 204
stable plateau (Fig. 7-12). The duration over which a motor unit
muscle fibers is primarily dependent upon two factors: (1) spe- can maintain a tetanus to repetitive stimulation defines its fa-
cific contractile proteins present in the muscle, and (2) rate of tigue characteristics. Motor units studies such as those d e -
calcium uptake. Actin and myosin can be present in several
186
scribed above reveal that there are two basic types of motor
isoforms which are different in slow and fast muscle tissue. units: fatigue resistant and fatigue sensitive. - 42 43
The ATPase of slow myosin is approximately half that of fast One may graphically plot motor units based upon the fol-
myosin, which allows histochemieal techniques to test for lowing criteria: (1) tetanus tension, (2) twitch contraction
these two enzymes by using monoclonal antibodies. As- a time, and (3) fatigue sensitivity (Fig. 7-13). Four subpopula-
43
result, the speed of contraction is directly influenced by the tions of motor units appear through this graphic technique.
type of contractile protein present. Muscle relaxation time There is a cluster of motor units, termed fast-fatigue (FF), that
(rate-limiting factor for speed of repetitive stimulation) is re- fatigue readily, produce relatively high tetanus tensions, and
lated to the rate of calcium uptake. Studies have shown that possess very fast twitch contraction times. A second popula-
fast-twitch fibers have a more extensive sarcoplasmic reticu- tion of motor units, termed slow (S), have low tetanus tension,
lum and are richer in C a - M g A T P a s e (modulates C a
2+ 2+ 2+
long contraction times, and are resistant to fatigue. The third
uptake) than slow-twitch fibers. - 16 86186
group of motor units to appear produce moderate tetanic ten-
Muscles that have a predominance of fast- or slow-twitch sions, but have fast twitch times and are relatively fatigue-re-
fibers tend to be either fast or slow. Histologic examination of sistant (FR). A fourth small intermediate population of motor
Recruitment Principles
A physiologic property of motor units closely related to those
noted above is recruitment. Recruitment may be defined as: "The
successive activation of the same and additional motor units
with increasing strength of voluntary muscle contraction." The 1
central nervous system can increase the strength of muscle con-

traction in three ways: altering the number of active motor units,
changing the rate at which individual motor units fire to optimize
the summated tension generated, or combining the first two
mechanisms. - - Within rather general terms, motor units are
35 46 204
recruited in a relatively constant manner according to their size

and the amount of tension they can generate. When a muscle is
initially activated, the first motor units to fire are usually the
weakest as defined by the tetanus tension. With an increasing
strength of contraction, larger motor units are called upon to
fire. The result is a not only an orderly addition of sequentially
larger/"stronger" motor units but also a smooth increase in
strength. The orderly recruitment of sequentially larger motor
units is referred to as the Henneman size principle described by
Henneman et a l . ' The order in which individual motor units
1 6 1 1 7
are recruited is not fixed, but can vary depending upon the type
of movement induced. Although it appears that the order of re-
65
cruitment may be rather stereotypical for the same movement,

Figure 7-13. Three physiologic properties of motor units are the tension for specific motor unit firing varies with the move-
graphically depicted. N o t e that motor units with relatively large tetanic
ment's s p e e d . As recruitment is one of the major portions of
100111
tension (> 20 gm) may either fatigue rapidly (FF: near 0.0 on fatigue
the needle electromyographic examination, it will be reviewed
scale) o r display resistance t o fatigue (FR: near 1.0 on fatigue scale).
Both of t h e s e m o t o r units have short-twitch tension times. A third in detail with respect to normal and abnormal findings below.
m o t o r unit type can have low tetanus tensions (< 20 gm) but is resis-
tant to fatigue (S: 1.0 on fatigue scale). A few m o t o r units, F(int), have Parameters
tetanus tensions > 20 gm, fast twitch times, and intermediate fatigue The MUAP recorded with a needle electrode placed in a vol-
properties. (From Burke RE, Levine DN.Tsairis P, et al: Physiological untarily contracting muscle arises from the summated electrical
types and histochemical profiles in m o t o r units of the cat gastrocne- activity of all the single muscle fibers belonging to a particular
mius.] Physiol 1973;234:723-748, with permission.) motor unit. For discussion purposes, a single motor unit within
the biceps brachii may be examined (Fig. 7-14). A single muscle
units between FF and FR is denoted F(int). Because contractile fiber in the biceps muscle extends from the origin to the inser-
and metabolic properties are related, histochemical techniques tion. Each muscle fiber is innervated by only one terminal
30
reveal that mitochondrial enzymes, succinate dehydrogenase,

have higher activity in fibers that are fatigue resistant while Table 7-5. Motor Unit Characteristics
phosphorylase activity (myosin-ATPase) are high in fibers re-
liant on glycolysis, i.e., fatigue sensitive (Table 7-5). The inter-
74 Fast-Twitch Slow-Twitch Fast-Twitch
mediate fibers display metabolic pathways capable of using Oxidative Oxidative Glycolytic
both enzyme systems. (SO) Glycolytic (FOG) (FG)
As previously noted, it is possible to tease out a single motor General Features
axon and repetitively stimulate it to metabolically drive the Red coloration Dark Dark Pale
muscle fibers belonging to that particular motor u n i t . 1 2 1 4 1 1 4 2
Myoglobin High High Low
Staining these fibers for glycogen by using the PAS technique Capillaries High High Low
reveals that all of the muscle fibers innervated by the stimulated Mitochondria High High Low
nerve are depleted of glycogen. This finding suggests that Fatigue Very resistant Resistant Sensitive
muscle fibers of different motor units may differ in their meta- Twitch speed Low Intermediate High
bolic requirements. Indeed, fast-twitch motor unit muscle fibers Tetanic force Low Intermediate High
rely on glycolysis for energy while slow-twitch fibers depend Fiber diameter Small Intermediate Large
upon oxidative mechanisms. - An intermediate muscle type
43 46
Staining Quality
has been recognized that uses a combination of both oxidative
PAS (glycogen) Low High High
and glycolytic pathways to sustain its metabolic needs. These
Phosphorylase Low High High
motor unit metabolic characteristics have resulted in a number
Myosin ATPase Low High High
of classification systems (Table 7-5). - Slow-twitch oxidative
15 185
SDH Intermediate High Low

(SO) fibers are also referred to as type I or B fibers. Fast-twitch
oxidative glycolytic (FOG) motor units may also be called type Other Classifications
IIA, type II, or C. Finally, fast-twitch glycolytic (FG) are also Dubowitz/Pearse 73
1 II II
known as type IIB, type II, or A. Fast-twitch fibers are sup- Brooke/Kaiser 15
1 HA IIB
ported by few capillaries while slow-twitch motor units are en- Stein/Padykula 214
B C A
veloped by a rich capillary supply consistent with their Burke 43
S FR FF
metabolic requirements. Modified after M c C o m a s 163
with the needle electrode is the summation of all of the electri-

cal activity produced from that motor unit's muscle fibers.
Because the fibers are not located next to each other, the
MUAP morphology varies with the needle electrode's location
(Fig. 7-14). This MUAP shape variability results from the fil-
tering effect of neighboring muscle fibers, number of active
single muscle fibers less than 500 pm from the recording sur-
face, and location of the electrode with respect to different
6 mm muscle fibers' arrangement about the endplate (see Chapter 2).
The various MUAP parameters to consider are (Fig. 7-15): (1)
amplitude, (2) rise time, (3) duration, (4) phases, (5) variability
and (6) recruitment.
Amplitude/Rise T i m e . A MUAP's amplitude is defined as
Figure 7-/4. A , M o t o r u n i t t e r r i t o r y within a human biceps brachii that portion of the waveform between the sequential most posi-
muscIe.Three different monopolar needle electrode locations within tive peak to most negative peak (Fig. 7-15). As noted above,
1,212
the same motor unit territory yield different MUAPs at each record- this parameter is primarily influenced by the number of single
ing site. (From Dumitru D, DeLisa JA: AAEM Minimonograph # 1 0 : muscle fibers (possibly 1 and certainly fewer than 12) - - 87 224 227
Volume conduction. Muscle Nerve 1991; 14:605-624,with permission.) less than 300-500 pm from the recording surface.27.28.105-107
Slightly displacing the recording electrode alters the spatial re-
axon. As previously noted, a peripheral nerve separates into mul- lationship of the needle electrode with respect to a particular set
tiple terminal axons upon reaching muscular tissue. Each termi- of neural generators. Recording from locations separated by as
nal axon forms a neuromuscular junction with one single muscle little as 500 pm places the needle electrode in a completely dif-
fiber. The area over which a single motor unit is distributed is ferent and unique portion of the muscle with respect to its elec-
circular to oval in shape with a diameter of 4 - 6 mm. The single trical characteristics. - A different number of active muscle
27 28
muscle fibers of one motor unit are randomly distributed in this fibers associated with new surrounding low pass filter proper-
area with a slight tendency for a greater density of muscle fibers ties leads to a change in the amplitude and possibly shape of the
near the geometric center of the circular region. - Within 34 8 5 209
main spike of the MUAP. The MUAP rise time is simply the
this motor unit territory are believed to be located an additional time it takes for the MUAP to describe its maximum positive to
10 and possibly more motor units interspersed in the same negative displacement (Fig. 7-15).
random fashion. The estimate of 10 motor units is determined D u r a t i o n . The MUAP duration is defined as the point of
by electrical means in which the first several motor units acti- initial departure from, and final return to, the CRT baseline
vated with minimal contraction are examined. As a result, these (Fig. 7-15). - This time represents the total depolarization of all
1 215
10 motor units are most likely type I. This limitation of only of the single muscle fibers forming one motor unit. - The du-27 28
being able to study the first recruited motor units accounts for ration is longer than a single muscle fiber's action potential be-
the discrepancy between electrical data and histologic prepara- cause the multiple single muscle fibers comprising a motor unit
tions in which more than 30 motor units (type I and type II) may do not all depolarize simultaneously, but depend upon the end-
share a similar region within the muscle. - - - 27 28 35 46
plate separation and terminal axon conduction velocities previ-
Needle electromyographic recordings of muscle tissue can ously noted. The first muscle fiber to depolarize is the one with
reveal several parameters of MUAPs that can be accurately the shortest distance from the terminal axon's formation to its
measured. Although the actual quantification of these parame- respective neuromuscular junction. Similarly, the last muscle
ters are best documented through quantitative needle elec-
tromyography (see below), the theoretical explanation for
MUAP morphology is presented in this section. Several
anatomic and physiologic factors regarding the motor unit must
be considered to gain an appreciation of why MUAPs appear as EU Phases
they do. As previously stated, single muscle fibers constituting sfs Turns
one motor unit are interdigitated with those fibers from other
motor units. - ' - - This places potentially inactive muscle
12 26 35 46 141
tissue between fibers that happen to be electrically active. Mus-

cle tissue acts as a low pass filter in that it significantly reduces
the spike component of a single muscle fiber action poten-
tial. - This filter effect is capable of reducing over 90% of the
87 207
spike's amplitude in just 200-500 pm. - - - -' 9 27 28 105 07
The time necessary to activate all of the muscle fibers of one

motor unit to form a MUAP is dependent upon the (1) distance
to individual neuromuscular junctions from which a particular
terminal axon separates from the parent nerve, (2) spatial sepa-
ration of the endplate region extremes, and (3) conduction ve-
locity along the various terminal axons and muscle fibers. - - 24 27 28
Once an anterior horn cell fires, all of the single muscle fibers it Figure 7 - / 5 . M U A P p a r a m e t e r s . A single MUAP waveform with
innervates generate an action potential. These multiple action its morphologic parameters delineated. (From Johnson EW:The EMG
potentials are somewhat spatially and temporally asynchronous examination. In: Johnson EW (ed): Practical Electromyography.
because of the above noted factors. The net MUAP detected Baltimore, Williams & Wilkins, 1988, pp I - 2 1 , with permission.)
fiber to be activated has the longest terminal axon to neuromus- fibers' action potentials will result in a coincident reduction in
cular junction length. Recall that the conduction velocity of the MUAP spike amplitude. Additionally, larger muscle fibers pro-
terminal axon is considerably faster than muscle action poten- duce a larger voltage associated with depolarization compared
tial conduction. Even though a particular single muscle fiber to smaller fibers with corresponding MUAP amplitudes. - A 27 28
may be activated prior to another because of a relatively short number of instrumentation factors can also affect the MUAP's
terminal axon, its muscle action potential propagates much amplitude (see Chapter 3). Specifically, a reduction in MUAP
slower than a longer neighboring terminal axon. A second ter- amplitude can be anticipated if: (1) the high-frequency filter is
minal axon, therefore, may proceed somewhat further to its neu- reduced, (2) concentric as opposed to monopolar needles are
56-58
romuscular junction which is located at the edge of the endplate used, - - - (3) comparatively larger electrode lead off surface
55 58 124 178
zone for example. If a needle electrode is located near the end- is used, (5) a concentric needle electrode is rotated (directional
88
plate's perimeter and closer to the second terminal axon/muscle recording properties), - - and (6) Teflon from a monopolar
24 25 47
fiber combination, that portion of the MUAP resulting from this needle is removed. The effect of temperature on MUAP ampli-
59
particular single muscle fiber action potential will be observed tude is somewhat controversial. Both amplitude reduction 57124
prior to that of the muscle fiber excited earlier. This is because and elevation have been reported with a decrease in muscle tem-
the terminal axon's conduction velocity is faster than that of ini- perature. - - It is the authors' opinion that MUAP amplitude
I0 25 61
tially activated single muscle fiber which is further from the increases with temperature reductions initially, but as the tem-
recording electrode. In summary, the net MUAP duration is the perature continues to decline, action potential failure occurs re-
electrical summation of all muscle fibers' action potentials com- sulting in MUAP amplitude reductions. MUAPs recorded in
prising one motor unit. persons greater than 65 years in general are slightly larger than
Phase. The MUAP phase is simply defined as that portion of those recorded in younger individuals. - 25 55
a potential between CRT trace baseline departure and return MUAP duration is directly proportional to the width of the
(Fig. 7-15). - The number of CRT baseline crossing plus one
1 212
endplate zone and to a lesser extent the terminal axon conduc-
yields the total number of MUAP phases. For example, if the tion velocity. - - Elevating the low-frequency filter will
25 27 28
MUAP crosses the CRT baseline twice, the MUAP is triphasic. remove the low-frequency content of the MUAP, initial and ter-
This is the usual appearance of most MUAPs. Normal MUAPs minal phases, thereby reducing the duration. Concentric 58
have 4 or fewer phases and MUAPs with 5 or more phases are needle electrodes may reduce the MUAP duration somewhat,
said to be polyphasic. A small percentage of the total number
1
although this is debatable. Cooling of the muscle produces an
recorded normal MUAPs are polyphasic depending upon the increase in MUAP duration. - - 10 25 61
needle electrode: monopolar (up to 30%) or concentric (up to The number MUAP phases is contingent upon the motor
15%). As previously stated, the main spike component of the unit's single muscle fiber action potential's degree of asynchro-
MUAP is believed to arise from one and possibly fewer than 12 nous passage past the recording needle's active surface. The
single muscle fibers in very close (< 500 pm) proximity to the greater the degree of synchronicity, the fewer the number of
needle electrode's recording s u r f a c e . - - ' The remaining
27 28105 07
phases, while greater temporal dispersion among the single
single muscle fiber action potentials located at some distance muscle action potentials increases the number of MUAP phases.-
(> 500 pm) will be propagating toward, and away from the Factors such as aging and temperature reduction will increase
needle electrode prior to and following the formation of the the MUAP's number of phases. Elevating the low-frequency
MUAP spike. Recall from volume conduction that (1) muscle filter may also result in a greater degree of phasicity. Mono- 58
tissue acts as a low pass filter and (2) single muscle fiber action polar needle recordings may have MUAPs with a higher per-
potentials are triphasic. - The muscle tissue surrounding the
87 88
centage of polyphasic potentials. 55
needle electrode filters out most, but not all, of the negative Recruitment (Normal Muscle). The physiologic principles
spike components of the action potentials arising from neigh- governing the orderly recruitment of motor unit action potentials
boring muscle fibers belonging to the activated motor unit. has been briefly described a b o v e . We will now turn to the 116117
Following filtering, what remains is the low frequency initial practical assessment of MUAP recruitment during the electro-
and terminal positive phases and a small negative spike of the diagnostic medicine examination. It is important for the practi-
triphasic action potentials. These "distant" potentials' phases tioner to keep in mind that motor unit recruitment is a poorly
help form the initial and terminal positive phases and, to a small understood process and only a very simple explanation will be
degree, the negative phase of the nearby spike potential. The provided in this section as it pertains to the electrodiagnostic
spike portion of the MUAP also has its own initial and terminal medicine consultation. Several assumptions will be made to
positive phases. In the volume conductor, the net summation of both simplify the discussion and maintain its relevance to practi-
these positive and negative waveforms depends upon the loca- cal needle electromyography. Specifically, we will assume that:
tion of the needle electrode with respect to the single muscle (1) the first recruited MUAPs arise from the small and relatively
fibers; they all contribute to forming the various phases of the slow conducting type I motor units (see above) exclusively, - 35 4 6 183
MUAP (see Chapter 2). (2) only minimal to moderate effort will be expended by the pa-
Variations of Parameters. The amplitude, duration, and tient (depending upon the strength of the muscle), (3) later re-
phases of normal MUAPs can vary depending upon several non- cruited motor units (type II) will simply not be capable of being
pathologic variables. Maximum MUAP spike amplitude is pri- analyzed, (4) subtle pathology affecting only a small portion of
marily influenced by the distance between the electrode's the muscle may be missed, (5) the first motor unit to fire con-
recording surface and the electrical generator. As previously tinues to increase its firing rate with more force generation, and
noted, 90% or more of a potential's amplitude may be dimin- (6) slow and constant contraction will be sufficient to demon-
ished by displacing the electrode more than 500 pm away from strate motor unit recruitment abnormalities. 182-184
the active tissue's surface. " Amplitude is also directly re-

105 107
Equipment. There are no generally accepted parameters for
lated to the total number of synchronously active fibers near the evaluating MUAP recruitment. These authors use amplifier low-
electrode. Decreasing the synchronous arrival of single muscle frequency filter settings of 10-30 Hz and a high-frequency filter
range of 10,000 Hz or more. A sensitivity capable of resolving Table 7-7. Human Muscle Recruitment
the entire potential is typically used and ranges from 100 to Onset Onset Recruitment Recruitment
1000 pV/div. A sweep speed of 20 ms/div suffices for most pur- Interval Frequency Interval Frequency
poses. Either a monopolar or standard concentric needle elec- Muscle (ms) (Hz) (ms) (Hz)
trode can be used.
Technique. In attempting to analyze MUAP recruitment one Facial Muscles
Frontalis 102 ± 2 9 9.8 46 ± 17 21.7
must realize that the main goal is to identify the first few re-
Orbicularis oris 70 ± 19 14.3 34 ± 10 29.4
cruited motor unit's action potentials and measure their firing
rate. Although this may sound like a simple task, some practice All facial muscles 86 ± 29 11.6 40 ± 16 25.0
and patient cooperation are required to assess MUAP firing fre- Upper Limb Muscles
quencies in a time-efficient manner. The beginning practitioner Deltoid 116 ±23 8.6 84 ± 16 11.9
should practice this analysis on a limited basis in normal muscle Bicep 124 ±21 8.1 86 ± 14 11.6
tissue prior to investigating pathological muscle. Tricep 132 ± 16 7.6 84 ± 17 11.9
MUAP recruitment may be examined before the needle elec- Brachioradialis 116 ±22 8.6 78 ± 18 12.8
trode is withdrawn from the muscle under investigation just fol- Pronator teres 132 ±38 7.6 88 ± 19 11.3
lowing the routine needle study. The patient is instructed that 1 st dorsal 142 ±39 7.0 98 ±21 10.2
only a gentle contraction of the muscle is required and state- interosseous
ments such as " . . . just think about contracting the muscle . . ." Paraspinal Muscle
are often used. One would like to detect just one motor unit Multifldus 1 52 ± 33 6.6 102 ± 2 0 9.8
firing regularly. Motor units usually begin firing irregularly at
Lower Limb Muscles
2 - 3 Hz and then achieve a stable and regular firing rate at 5-7
Gluteus maximus 1 28 ± 30 7.8 88 ± 16 11.4
j | 35.72,182-184 patient is then asked to increase the force of
Vastus lateralis 126 ±30 7.8 88 ± 18 11.4
z
contraction minimally. The first recruited MUAP (MUAP A)

Biceps femoris 132 ± 29 7.6 92 ± 16 10.9
should then increase its rate of firing to between 6 Hz and 10 Hz
Tibialis anterior 124 ±26 8.1 90 ± 13 II.1
(Table 7-6). More force is usually generated initially by increas-
Medial gastroc- 156 ±29 6.4 110 ±23 9.1
ing the firing frequency of the first recruited motor units. With a
nemius
continued modicum of increased force, the first recruited
EDB 138 ± 2 9 7.2 98 ± 13 10.2
MUAP eventually achieves a firing rate of about 10 Hz. At this
All limb 132 ± 3 2 7.6 90 ± 19 II.1
frequency; a second motor unit is recruited (MUAP B ; Tables 7-6
muscles
and 7-7). The frequency of MUAP A when MUAP B just begins
to fire regularly is called the recruitment frequency. Although it
1
Modified after Petajan. -
182 184
takes some practice, the physician can recognize this event by

(1) observing the screen for a second motor unit to appear and its morphology may be altered. Care must be exercised, there-
(2) listening for the change in sound associated with two fore, to ensure that the same MUAP from one motor unit is cor-
MUAPs firing. The authors find the sound of the second motor rectly identified for time marker placement. The examiner and
unit to be most helpful. Once this sound is detected, the instru- patient may require practice before mutual understanding arises
ment's footswitch is depressed to "freeze" the screen. The la- as to what is needed to measure MUAP recruitment intervals
tency markers are positioned on two sequential MUAPs from and frequencies. Several (4-5) regions of the muscle under ex-
MUAP A just before MUAP B appeared. This time difference is amination may need to be studied to gain a representative
called the recruitment interval, the reciprocal of which yields
1
sample of recruitment frequencies.
the recruitment frequency. Most normal muscles in humans To simplify the discussion regarding motor unit recruitment
have a recruitment interval of 100 ms with a recruitment fre- in the normal situation and to lay the groundwork for abnormal
quency approximating 10 Hz (1 motor unit/100 ms = 10 Hz; recruitment, we will assume that the first several motor units
Table 7-7). Facial muscles have a somewhat shorter recruitment follow the above described simple rule of increasing firing rate.
interval and corresponding higher recruitment frequency. Let us return to our patient in which the first recruited MUAP
The "trick" in the above relatively straightforward procedure (MUAP A) is firing at 10 Hz and the second (MUAP B) is firing
is to accurately identify sequentially firing motor units. With a at 5 Hz (Table 7-6). If the patient is requested to increase the
sweep of 20 ms/div, the MUAP is somewhat "compressed" and force production slightly, MUAP A will incrementally increase
its firing rate as will MUAP B . When MUAP A reaches 15 Hz
Table 7-6. Normal Recruitment and MUAP achieves a firing rate of 10 Hz, a third motor unit
(MUAP C) begins to fire at about 5 Hz. To continue this process
Motor Unit Recruited one more step, a further increase in muscle force production re-
1st (A) 2nd (B) 3rd (C) 4th (D) sults in the following firing frequencies: (1) MUAP A: 20 Hz,
A (5 Hz) (2) MUAP B : 15 Hz, (3) MUAP C: 10 Hz, and (4) the addition
A (10 Hz) B (5 Hz) of a fourth motor unit represented electrically by MUAP D
firing at 5 Hz (Table 7-6). Again, the sequential addition of
A (15 Hz) B(IOHz) C (5 Hz)
motor units is considerably more complex than illustrated above
A (20 Hz) B(l5Hz) C(IOHz) D (5 Hz) and may not exactly follow this sequence. However, normal and
MUAP A begins firing stably at about 5 Hz. With a minimal increase in force of pathologic physiologic conditions can be conceptualized
muscle contraction, MUAP A increases its firing rate to 10 Hz and MUAP B through this simple example.
begins firing stably at 5 Hz.The recruitment frequency for this muscle is 10 Hz.
A still further increase in contraction causes MUAP A to fire at 15 Hz, MUAP B Although the firing frequencies noted above are rough ap-
fires at 10 Hz, and a new motor unit (MUAP C) is activated.This process con- proximations, they can be used to formulate a few simple con-
tinues as noted for MUAP D. cepts that may be used to characterize the normal recruitment of
MUAPs identified on the CRT screen. The orderly recruitment abnormalities as observed in attacks of periodic paralysis
of the four motor units above can be discussed in terms of the (Table 7-2). Necrotic muscle resulting from profound ischemia
rules of five, i.e., motor units begin firing at stable rates of
63
also produces little electrical activity. The author has observed
5 Hz. When the first motor unit to fire reaches 10 Hz, a second that improper electrode amplifier connections may simulate de-
motor unit is activated and fires at 5 Hz. Each time a motor unit creased insertional activity. Inadvertently connecting the needle
is recruited, 5 Hz is serially added to each MUAP firing fre- electrode lead to the ground port while the ground and reference
quency already present (Table 7-6). Using the MUAP firing rate leads are placed into the active and reference amplifier ports re-
and number of different motor units on the CRT screen, one can spectively will result in a satisfactory baseline, but on inserting
formulate a recruitment ratio. Specifically, the frequency of the the needle minimal insertional activity is observed, erroneously
fastest firing MUAP divided by the number of different MUAPs producing the appearance of fibrotic muscle. A quick check on
on the screen should result in a number close to 5 (fastest the electrode connections will rectify this instrumentation
MUAP/# MUAPs = 5). In the above example, 4 separate motor
63
error.
units were activated when the highest MUAP frequency ap- Increased Insertional Activity. Following the insertion of a
proached 20 Hz. The recruitment ratio is 5 (20/4 = 5). If this needle electrode into normal muscle tissue, electrical activity
number approaches 10, then there are too few motor units for the persists for approximately 50 ms after the cessation of needle
greatest firing frequency and force produced. Similarly, if this movement. Electromyographers have noted that inser-
232-234
number is reduced below 4 or 5, then there are too many motor tional activity may appear to be prolonged in a number of patho- j
units for the highest firing rate. The recruitment ratio can also be logical conditions. This finding has led to the term "increased
used to predict the number of MUAPs that should be present insertional activity" which is believed to designate abnormal
given the firing rate of the fastest MUAP. Simply, suppose a potentials after needle movement has stopped but before the
MUAP's frequency is 15 Hz. Dividing this number by 5 suggests occurrence of sustained spontaneous potentials (Fig. 7-4C,
that three separate MUAPs should be present on the CRT screen. Table 7 - 2 ) . - In disease states where the muscle is no longer
109 m
A last concept to consider is predicting the firing rate of innervated, the increased insertional activity supposedly com-
MUAPs by looking at the CRT screen if the sweep speed is pletes a temporal continuum from the previously normal inser-
known. In the above examples, how many times will a motor tional activity to the development of sustained membrane
unit firing at 20 Hz appear on the CRT screen? If the sweep instability. A controlled study investigating the duration of
speed is 10 ms/div, there are 100 ms across the entire screen (10 muscle depolarization following needle movement in dener-
ms/div x 10 div/screen = 100 ms/screen). A motor unit firing at vated muscle did not substantiate abnormalities in insertional
20 Hz means that it is firing 20 times/1 second or 20 times/1,000 activity. The time period of 50 ms was the same for both the
232
ms. If it is firing at 20 times/1,000 ms, this is the same as 1 denervated and control muscle tissue despite significant mem-
time/50 ms. If the screen has 100 ms, this motor unit should brane instability. It appears that the concept of increased inser- |
appear twice every 100 ms or 2 MUAPs per CRT screen (1 tional activity, strictly speaking, is erroneous. Therefore, it may
MUAP/50 ms x 100 ms/screen = 2 MUAP/screen). Remember be more accurate to state that needle movement actually pro-
that this is the same MUAP firing at a frequency that allows it to vokes either sustained or unsustained abnormal spontaneous po^
appear twice during the one sweep of 100 ms. As long as the tentials when addressing insertional activity.
sweep is 10 ms/div, one only has to multiply the number of
times the same potential is present on the CRT screen by a Fibrillation Potentials
factor of 10 to arrive at its correct firing frequency. If the sweep In vitro observations reveal that following denervation for
speed is changed to 5 ms/div, the number of times the same 4-6 days or longer, the muscle fiber's resting membrane poten-
MUAP appears on the CRT is multiplied by a factor of 20 to de- tial begins to approach a less negative level of - 6 0 mV com-
termine the MUAP's firing rate. Note that calculating the firing pared to the normal value of - 8 0 mV. - - Additionally, the
82 221 223
frequency for a particular motor unit is not the same as deter- resting membrane potential begins to oscillate. As the "thresh-
mining the recruitment frequency. old level" (about - 6 0 mV) producing the all-or-none self-sus-
taining action potential and the new resting membrane potential
are now similar, an oscillating membrane potential can eventu-
MUSCLE GENERATORS OF ABNORMAL ally reach the membrane's depolarization threshold level. Once |
SPONTANEOUS POTENTIALS threshold is achieved, a propagating action potential is induced '
and referred to as a fibrillation potential. The repolarization
32
INSERTIONAL ACTIVITY phase of denervated muscle results in a temporarily more hyper-

polarized (-75 mV or more) level than the new resting mem-
Decreased Insertional Activity. Although not examined in brane potential of - 6 0 mV. Additionally, the muscle's
a quantifiable manner, the designation "decreased insertional repolarization then affects the threshold by hyperpolarizing it
activity" may be justified. Muscle tissue that has undergone fi- below - 6 0 mV. As the hyperpolarized membrane level (-75 mV
brosis will no longer be capable of electrical activity and hence or more) begins to return toward its resting membrane level of
a decrease in needle insertional activity is noted (Fig. 7-4B). - 6 0 mV (now less negative than the new threshold level e.g.,
The result is little if any electrical potentials detected following - 6 5 mV), an action potential is again produced. This process
needle movement. This tissue may also have a "gritty" or fi- regularly repeats on a time interval dependent upon the repolar-
brous feel during needle insertion. Obviously, one must ensure ization-to-threshold turnaround time. - - The regularly oc-
66 220 221
that the recording electrode is located in muscle tissue. A needle curring fibrillation potentials fire in a cyclical pattern. That is,
electrode inserted into subcutaneous adipose tissue or tendon they tend to suppress themselves, creating relatively quiescent
will demonstrate minimal electrical activity associated with periods. These intervals of quiescent muscle may
1 5 3 1 5 4 1 8 7 1 8 8 1 8 9
needle insertions. One may also detect decreased insertional ac- explain in part why fibrillation potentials and positive sharp
tivity in electrically silent tissue due to metabolic/electrolyte waves may be found in abundance at some times and and absent
at other times even in the same muscle. The difference between

the resting membrane potential and threshold level is less in the
former endplate region than along the muscle fiber.206,222,223 x j n s
may be the reason why fibrillation potentials appear to arise

more commonly from the previous endplate than at other por-
tions of the muscle. Fibrillation potentials may be reduced in
number following a decrease in temperature, - muscle is- 34 37
chemia, or D-tubocurarine administration. Irregularly firing

120 11
fibrillation potentials are less well understood, but thought to

arise from spontaneous depolarizations within the transverse
tubule system. 205
Because a fibrillation potential is the spontaneous depolariza-

tion of a single muscle fiber, approximately 30% will have a
morphology consistent with that previously described for this
tissue, i.e., triphasic: positive-negative-positive (Fig. 7-16A). 32
It is not unusual, however, for fibrillation potentials to also be Figure 7-/6. Fibrillation potentials and positive sharp waves.
biphasic. Biphasic negative/positive fibrillation morphologies
157 A , Biphasic and triphasic fibrillation potentials (f) recorded in dener-
may be observed if a fibrillation potential is recorded from the vated muscle. B , Positive sharp waves (p) r e c o r d e d from the s a m e
muscle in A.
endplate zone. Biphasic positive/negative fibrillation potentials
are possibly more commonly observed than triphasic fibrilla-
tions because the third phase is believed to become lost in the Positive Sharp W a v e s ( P S W s )
baseline (Fig. 7-16A). A fibrillation potential has a duration less The positive sharp wave's name is characteristic of its mor-
than 5 ms and an amplitude under 1 mV (Table 7-8). Fibrillation phology. Specifically, this potential consists of a primary
potentials may fire regularly approximately 50% of the time and initial positive phase followed by a return to baseline (mono-
irregularly for the remaining 50%. The irregular firing is not as phasic) or a small negative deflection (biphasic) (Fig. 7-16B).
irregular as that of the endplate spike, however. The regular Jasper and Ballem are credited with first describing this po-
127
high-pitched sound associated with fibrillation potentials has tential. Similar to fibrillation potentials, PSWs may fire regu-
j been likened to the crackling of cellophane or rain falling on a larly or irregularly. These potentials have durations of several
I tin roof. As the needle electrode is inserted into muscle tissue milliseconds to 100 ms and fire regularly at several hertz. The
with an unstable resting membrane potential, fibrillation poten- sound of the PSW is similar to a dull thud. Most investigators
tials may be seen to fire spontaneously or arise secondary to believe that the positive sharp wave and fibrillation potential
provocation of the recording electrode. have the same clinical significance. The presumed theory
32145
of PSW origin revolves around a blocked fibrillation potential

Table 7-8. Characteristics of Fibrillation Potentials 154
in which the negative current sink approaches, but does not
pass, the recording e l e c t r o d e . This concept has been chal-
32129
Appearance: a. Biphasic spike 1-5 ms in duration

lenged because of several clinical observations. First, PSWs
137
b. Positive wave with negative phase

appear earlier than fibrillation potentials in animal and human
Rhythm: Usually regular but can be somewhat irregular muscle deprived of its nerve supply. If PSWs and fibrillation
Frequency: 0.5 t o 15 Hz potentials are the same, their clinical appearance with respect
Amplitude: 20 t o 1000 pV to timing should be the same. Purely demyelinating diseases
have been documented to yield PSWs but not fibrillation p o -
Stability: Stable
tentials. It is the authors' opinion that caution should be o b -
Observed in: a. Muscle disorders served in interpreting these findings as some axonal loss may
Inflammatory myopathies have occurred even though it was not seen histologically on
Inclusion body myositis nerve biopsy. The interested reader is referred to the literature
Congenital myopathies for a detailed discussion of the controversy surrounding the
Some muscular dystrophies PSW and its relationship to fibrillation p o t e n t i a l s . 78137138
Hyperkalemic periodic paralysis Potentials with similar waveforms may be confused with
Rhabdomyolysis PSWs as detailed in Chapter 2.
Muscle trauma following muscle biopsies
Trichinosis Fibrillation/Positive Sharp W a v e s :
b. Neurogenic disorders Clinical Findings
Anterior horn cell disorders
Fibrillation potentials and positive sharp waves may be
Radiculopathies
recorded in both neurogenic and myopathic disease states and
Plexopathies
commonly observed together (Table 7-8). They may be graded
Mononeuropathies
according to their number and ease of observation based upon
Peripheral neuropathies
the number of sites examined (Table 7-9). Clinically, the devel-
Entrapment neuropathies
opment of membrane instability following neural injury may
Upper m o t o r neuron disorders (stroke, head
occur within one to three weeks. - In the case of denervation
32 36
injury, and spinal cord injury)

induced potentials, the appearance of positive sharp waves and
c. Neuromuscular junction disorders
fibrillations depends upon the length of nerve segment between
Myasthenia gravis the site of injury and muscle tissue. The longer the nerve
221
Botulism segment, the more time from axonal loss to the first sign of
Table 7-9. Grading of Fibrillation Potentials 61

nerve stimulation by showing an absent response above but
Grading Characteristics not below the presumed lesion site.
Fibrillation potentials and positive sharp waves have also been
0 N o fibrillation potentials documented in primary muscle diseases (Table 7-8). - 144 171
1+ Persistent/unsustained single trains in at Although the exact mechanism of membrane instability produc-
least two muscle regions tion is poorly understood, a number of proposals have been sug-
2+ Moderate numbers in three or more gested. Essentially any muscle fiber deprived of its innervation
muscle areas will fibrillate. - - If the muscle undergoes segmental necrosis,
69 70 222
3+ Many in all muscle regions portions of the muscle fiber may no longer be connected to its ter-
4+ CRT baseline obliterated with fibrillation minal axon. In effect, those muscle regions are now denervated.
potentials in all areas of muscle examined This process has been documented in a number of primary
muscle diseases. Also, it may be possible for inflammatory cells,
muscle fibrosis, and fiber splitting to result in the terminal axon
spontaneous muscle fiber depolarization. For example, in the becoming separated from the muscle fiber resulting in a dener-
case of a radiculopathy, membrane instability may be ob- vated muscle. One can also detect fibrillations in familial hyper-
served in the paraspinal muscles within 7-10 days. These kalemic periodic paralysis during attacks of paralysis. Fibrillation
same abnormal potentials are then seen in the corresponding potentials are either absent or markedly reduced between at-
myotomal limb regions in approximately 21 days. If the lesion tacks. The etiology of fibrillation potentials in this disease is
102
resolves, the membrane instability will first resolve in the most likely a metabolic abnormality of K and/or Cl" conduc-
+
paraspinal muscles to be followed by resolution in the limb tance. Any metabolic abnormality adversely affecting the sodium,
muscles. Positive sharp waves and fibrillations may persist for potassium, and chloride ion gates may lead to a situation where
years following an injury resulting in axonal l o s s . The 136
the resting membrane potential may become unstable to such a
mechanism whereby this finding occurs is poorly understood. degree as to result in positive sharp waves and fibrillation poten-
It may be that some of the denervated muscle fibers become tial generation even in the absence of denervation. Positive sharp
sequestered by connective tissue or other denervated muscle waves and fibrillation potentials are abnormal, but they must be
associated with the denervation process thereby impeding viewed in the context of all the data collected in combination with
reinnervation. "Walled o f f regions of muscle surrounded by the history and physical examination.
fibrous tissue are created through which a nerve sprout is A single investigation has suggested that it is easier to detect
prevented from entering. As long as the muscle's vasculature fibrillations potentials and positive sharp waves with a concen-
survives, it may continue to fibrillate. Of course, these se- tric compared to monopolar needle electrode. This has not
201
questered muscle fibers will atrophy. Atrophic muscle pro- been the experience of the author. On the contrary, it is the au-
duces less of a voltage following depolarization, and one thors' experience that fibrillation potentials and positive sharp
could anticipate that these fibrillations and positive sharp waves are detected equally with either needle type, provided one
waves should decrease in amplitude over time. This has indeed is paying close attention to the loudspeaker. It is rather typical to •
been found to be the case. The mean fibrillation amplitude hear these potentials prior to seeing them on the CRT. However,
during the first two months following nerve injury approxi- one may hear and not see the fibrillation potential/positive sharp
mated 600 pV. By the end of the third month the amplitude
136
wave with a concentric electrode while being able to see and
had declined to 500 pV and reached 300 pV at 6 months. After hear it with the monopolar electrode. As stated elsewhere in this
one year, fibrillation amplitude did not exceed 100 pV. In the book, one needle is not superior to the other; rather they each
authors' opinion, caution should be exercised in using fibrilla- have unique recording characteristics which should be taken ad-
tion potential following injury to date the implied lesion onset. vantage of.
Considerable more work is required to more fully substantiate
the relationship between fibrillation potential amplitude and C o m p l e x Repetitive Discharge
the lesion's age. Buchthal has attempted to roughly quanti-
79 36
A complex repetitive discharge (CRD) is a spontaneously

tate the amount of fibrillation potentials based on the patient's firing group of action potentials formerly called bizarre high-
strength (1-5 scale of Medical Research Council, 1976). The frequency discharges or pseudomyotonic discharges. - These92 216
incidence of membrane instability was greater in muscle with potentials are not observed by direct visual inspection, but re-
significant weakness (force < 4) than muscle with a grade of quire a needle recording electrode be placed into the muscle.
4+ or more. Additionally, in muscle with equal degrees of The morphology of these potentials on needle electromyo-
strength there are differences in the amount of membrane in- graphic examination are continuous runs of simple or complex
stability depending upon the type of neurogenic lesion. spike patterns that regularly repeat at 0.3-150 Hz (Fig. 7-17;
Attempts to further quantify the degree of fibrillation poten- Table 7-10). The repetitive pattern of spike potentials has the same
tials with the amount of nerve loss are not substantiated by ex- appearance with each firing and bears the same relationship
perimental data and should not be attempted. Four plus (4+) with its neighboring spikes. A distinct sound likened to heavy
fibrillation potentials do not of themselves suggest complete machinery or an idling motorcycle is produced by the firing of
muscle denervation, but only that a specific region of muscle CRDs. In addition to the sound and repetitive pattern, a hall-
surrounding the needle electrode is deprived of its nerve mark of these waveforms is that they start and stop abruptly. - 92 228
supply or metabolically affected in an adverse manner. It is Complex repetitive discharges may begin spontaneously or be
also necessary to attempt to record MUAPs and examine mul- induced by needle movement, muscle percussion, or muscle
tiple areas of muscle prior to concluding whether a muscle is contraction. - - Single-fiber electromyography has provided
208 210 228
completely denervated or not. The possibility that the absence insight into CRD production.
of motor units is due to a complete proximal conduction block Nerve block and curare do not abolish CRDs suggesting that
should alway be contemplated and can be ruled out with electrical these potentials originate in muscle tissue. 216
Single-fiber
Table 7-10. Complex Repetitive Discharge Characteristics' !
Appearance: May take any form but this form is constant

from o n e potential complex t o the next
Rhythm: Regular
1 Frequency:
Amplitude:
Stability:
10 t o 100 Hz
50 t o 1000 uV
Abrupt onset and cessation
Observed in: a. Myopathies
Polymyositis
Limb-girdle dystrophy
Myxedema
Schwartz-Jampel syndrome
b. Neuropathies
Poliomyelitis
Spinal muscular atrophy
Amyotrophic lateral sclerosis
Hereditary neuropathies
Chronic neuropathies
Carpal tunnel syndrome
c. "Normal"
Iliopsoas
200 V
Biceps brachii
M
Figure 7 - / 7 . Three examples (A-C) of complex repetitive dis- Additional fibers excited before the principal pacemaker is re-
charges. N o t e how t h e same potentials appear in t h e same repetitive activated form the remaining spike potentials of the CRD. A
groups.The minor differences between discharges is due t o baseline
CRD, therefore, consists of serially activated single muscle
irregularities.
fibers that fire in the same pattern time after time. The C R D
continues to fire until either the principal or co-pacemaker alters
electromyography jitter variability between the individual spike the intra/extracellular ionic ratios sufficiently to block further
potentials is usually less than 5 p s . When jitter values of such
228
action potential generation, at which time abrupt cessation of
low magnitude are found, neuromuscular transmission is not in- firing occurs. It is also conceivable that a discharge from an-
volved in the generation of these potentials. It is postulated that other impulse-generating site blocks the pacemaker's action p o -
cross-talk or ephaptic conduction between neighboring single tential through a loop collision.
muscle fibers forms the basis of the observed repetitive firing Complex repetitive discharges may be seen in a variety of
pattern. The proposed mechanism of CRD generation is as fol- mostly longstanding diseases and often accompany fibrillating
lows (Fig. 7-18): (1) The action potential of a spontaneously fib- potentials (Table 7-10). These potentials have been observed in
rillating single muscle fiber, principal pacemaker, ephaptically myopathies such as polymyositis and various forms of muscular
activates one or several adjacent muscle fibers at a low thresh- dystrophy. - There have also been reports of finding CRDs in
92 220
old region on their membranous surface; (2) One of these adja- chronic denervation states such as motor neuron disease, radicu-
cent muscle fibers re-activates the principal pacemaker, causing lopathies, and polyneuropathies. 92
it to fire slightly earlier than it would because of its inherent fib-

rillating rate. The muscle fiber that reactivates the principal Myotonic Discharge
pacemaker is called the co-pacemaker and forms a closed loop The phenomenon of delayed muscle relaxation following muscle
representing the cycle time or repetition rate of the CRD; ( 3 ) contraction is referred to as myotonia or action myotonia. - I93 219
Figure 7-18. Proposed mecha-

nism of a complex repetitive
discharge. A principal pacemaker
(fiber A) initiates the complex repet-
itive discharge (starburst fiber A).
This fibrillation o r single muscle fiber
action potential propagates along
fiber A until it ephaptically activates
fiber B and C. Fiber B (co-pace-
maker) reactivates fiber A t o repeat
the cycle. Fibers C and D contribute
their waveforms t o t h e complex
repetitive discharge. (From Trontelj J,
Stalberg EV: Bizarre repetitive dis-
charges recorded with single fiber
EMG.J Neurol Neurosurg Psychiatry
1983; 46:310-316, with permission.)
Table 7-11. Myotonic Disorders

Syndrome Heredity Defect
Myotonic dystrophy Dominant Cation channel o r
lipid membrane
Myotonia congenita Dominant Reduced chloride
conductance
Type I: Thompson
Type II: Myotonia with painful spasm
Type 111: Myotonia with cold sensitivity
Recessive generalized myotonia Recessive Reduced chloride
conductance Figure 7-19. Myotonic potentials demonstrating an alteration in
firing rate and declining amplitude.The spike form (top trace) and pos-
Paramyotonia congenita Dominant Possible cation
itive potential form (bottom trace) are pictured. (From Streib EW:
(Eulenburg) channel AAEM Minimonograph No. 27: Differential diagnosis of myotonic syn-
Hyperkalemic periodic Dominant Possible ionic dromes. Muscle Nerve 1987;10:603-615, with permission.)
paralysis with myotonia channel
Drug-Induced Acquired the action potential propagates toward the needle but does not
Azacholesterol Membrane matrix propagate past the electrode. ° This is because the record-
W 2 I 2 U 2 ,
Monocarboxylic acids unknown ing needle is thought to have damaged that portion of the muscle
Other Acquired Unknown fiber in which it is in contact. Secondly, myotonic potentials may
Hypothyroid also appear as a series of rapidly firing biphasic or triphasic
Polymyositis single muscle fiber potentials following muscle contraction. The
Acid maltase deficiency myotonic discharge has a characteristic sound likened to a dive
bomber and is easily recognized once identified. This distinct
Modified from Brown and Strieb.
21 215
sound quality is produced from the waxing and waning firing

pattern of the single muscle fibers generating the potentials. 22
The finding of delayed muscle relaxation after reflex activation or Their amplitude may range from 10 pV to 1 mV and fire be-
induced by striking the muscle belly with a reflex hammer is tween 20 Hz and 100 Hz. A quantitative description of myotonic
called percussion myotonia. Clinical myotonia is usually accentu- discharges has been suggested as follows: 1+: discharges should
ated by energetic muscle activity following a period of rest. last at least 500 ms and be obtained in 3 regions outside of the
Continued muscle contraction lessens the myotonia and is known endplate zone; 2+: myotonic discharges found in greater than
as the "warm up." There are multiple hereditary and several ac- one-half of all needle sites; 3+: myotonic discharges noted in all
quired syndromes involving myotonia with additional subcate- areas examined secondary to needle movement. 219
gories (Table 7-11). It is believed that cooling the muscle worsens Myotonic discharges can occur with or without clinical my-
myotonia but this finding has been objectively documented only in otonia. The observation of these potentials requires needle move-
paramyotonia congenita. The severity of myotonia in some pa-
191
ment or muscle contraction. These potentials persist after
219
tients apparently parallels the serum potassium level. 152

nerve block, neuromuscular block, or frank denervation, sug-
Electrically, a myotonic discharge may present in one of two gesting that their site of origin resides in the muscle membrane
forms (Fig. 7-19, Table 7-12). T h e myotonic potential in-
219
itself. Although the exact mechanism of myotonic discharge pro-
duced by needle electrode insertion usually assumes a morphol- duction remains unclear, it is proposed that decreased chloride
ogy similar to that of a positive sharp wave. It is believed that conductance is responsible at least in part for the findings in my-
the needle movement induces a repetitive firing of the unstable otonia congenita. - In addition to the syndromes noted above,
8 9
membrane belonging to multiple single muscle fibers in which myotonic discharges may be detected in acid maltase deficien-
cies, polymyositis, and other diseases (Table 7-12). - 93 219
Table 7-12. Characteristics of Myotonic Discharges 154
Appearance: Brief spikes/positive waveform

NEURAL GENERATORS OF ABNORMAL
Rhythm: Wax and wane
SPONTANEOUS POTENTIALS
Frequency: 20 t o 100 Hz
Amplitude: Variable (20uV t o 1 mV) FASCICULATION POTENTIALS
Stability: Firing rate alterations
The visible spontaneous intermittent contraction of a portion
Observed in: a. Myopathies
of muscle is referred to as a fasciculation. When these contrac-
Myotonic dystrophy
tions are observed with an intramuscular needle recording elec-
Myotonia congenita
trode they are called fasciculation potentials. A fasciculation
21
Paramyotonia
potential is the electrically summated voltage of depolarizing
Polymyositis
muscle fibers belonging to one motor unit. Either an entire
Acid maltase deficiency
motor unit or only a portion of the muscle fibers belonging to it
Hyperkalemic periodic paralysis
may fire to constitute a fasciculation potential. Occasionally
b. Other
fasciculation potentials may be documented only with needle
Chronic radiculopathy
electromyography because they lie deep in the muscle, thereby
Chronic peripheral neuropathy obscuring clinical observation.
Figure 7-20. Multiple fascicula-

tion potentials recorded from a
patient with amyotrophic lateral
sclerosis.The firing frequencies, al-
though irregular, range from 0.005
t o 0.01 Hz. (From Brown W F : T h e
Physiological and Technical Basis of
Electromyography. Boston, Butter-
w o r t h , 1984, pp 317-368, with per-
mission.)
50 s
Fasciculation waveform morphology has the same characteris- with these potentials is a type of sputtering often heard with a
tics as that of simple motor unit or polyphasic action potentials low-powered motor boat engine. The actual discharge may be
with respect to amplitude, duration, and phases (Fig. 7-20). 227
distinguished from CRDs in that myokymic discharges do not
Their discharge rate (1 Hz to many per minute) is irregular and display a regular pattern of spikes from one burst to the next,
not under voluntary control; nor are they influenced by mild con- nor do they typically start and stop abruptly. Physiologically,
traction of the agonist or antagonist muscles (Table 7-13). The myokymic discharges are groups of motor units, while C R D s
site of origin of fasciculation potentials remains unclear, although are groups of single muscle fibers. The groups of motor units
210
it appears that the spontaneous discharge may arise from within within a burst may fire only once or possibly several times.
the spinal cord (anterior horn cell), along the entire peripheral
195
Also, the sputtering bursts of myokymic discharges sound quite
nerve (particularly the terminal portion), or at times within the
149
different from the continuous drone of a CRD.
muscle itself. Anterior horn cell, peripheral nerve, or terminal
210
The documentation of myokymic discharges can be found in
nerve membrane variations in excitability have been suggested as normal individuals where the orbicularis oculi may twitch for
possible causes for the spontaneous motor unit discharges. The days. Myokymic discharges may appear in several patterns: -
3 3 62
majority of fasciculation potentials observed in normal individu- focal (one muscle), segmental, or generalized. Focal myokymic
I als occur in the foot intrinsic muscles or gastroc-soleus muscle. 166
potentials can be observed exclusively in the face (facial myo-
One may detect fasciculation potentials in a variety of dis- kymia) arising from: fatigue, multiple sclerosis or a brainstem
eases as well as in normal individuals, particularly following neoplasm (Table 7-14). Segmental myokymic discharges can be
tension or anxiety, fatigue, heavy exercise, coffee, smoking, or noted in syringomyelia or radiculopathies. Finally, generalized
no known associated factors (Table 7-13). - - - Typical dis-
67 68 166 190
myokymic discharges have been detected in uremia, thyrotoxi-
eases in which fasciculation potentials may be found include: cosis, inflammatory polyradiculoneuropathy and rare hereditary
motor neuron disorders, radiculopathies, entrapment neu- disorders such as familial paroxysmal kinesigenic ataxia. 23
ropathies, or cervical spondylotic myelopathy. Fasciculation po- Limb myokymic discharges have also been described associated
tentials have also been described in metabolic disturbances primarily with radiation plexopathy. The site of origin of the
3
including: tetany, thyrotoxicosis, and anticholinesterase over- myokymic discharge remains unknown but has been postulated
doses. Studies have attempted to distinguish between benign
21
to arise anywhere along the motor neuron from the cell body to
and pathological fasciculation potentials. To date, there
1181462 2 7
the terminal axon. One theory suggests that transaxonal ephap-
is no reliable way to categorize whether fasciculation potentials tic activation of neighboring axons results in the characteristic
herald a disease state by solely considering their inherent char- discharges seen in myokymia. 189
acteristics based on routine needle electromyography. Fascic-

ulations with increased jitter and/or blocking (failure of Table 7-13. Characteristics of Fasciculation Potentials 154
neuromuscular junction transmission) on single-fiber elec-

tromyography may, however, be considered abnormal. The best Appearance: Variable: normal or complex MUAPs
way to evaluate fasciculation potentials is to also analyze the Rhythm: Irregular
"company they keep." That is, a careful analysis of voluntary Frequency: 0.1 t o 10 Hz
motor unit action potential morphology combined with a search Amplitude: > 300 uV
for abnormal spontaneous potentials is required prior to con-
Stability: Stable
cluding that fasciculation potentials are an abnormal finding.
Observed in: a. Normal individuals
Spontaneous
MYOKYMIC DISCHARGE Following exercise
b. Lower motor neuron disorders
A readily observable vermicular (bag of live worms) or con-
Amyotrophic lateral sclerosis
tinuous rippling movement of the skin, myokymia, is usually
Creutzfeldt-Jakob disease
associated with myokymic discharges. - A myokymic dis-
3 62
Radiculopathy
charge is recorded when a needle recording electrode is placed
Peripheral neuropathy
into a muscle with the above noted finding. Bursts of normal ap-
pearing motor units with interburst intervals of silence firing at Entrapment neuropathy
0.1-10 Hz in a semirhythmic pattern form the basis of a myo- c. Metabolic disorders
kymic discharge (Fig. 7-21; Table 7-14). Two to ten potentials Thyrotoxicosis
within a single burst may fire at 20-150 H z . These potentials
62 Tetany
are not affected by voluntary contraction. The sound associated Anticholinesterase medication
I'M; M i
Figure 7 - 2 / . Multiple examples of myokymic poten-
flrtr tials from patients with radiation plexopathy. N o t e h o w

each burst of m o t o r units is relatively regular, but t h e
MUAP content of each burst is variable. (From Albers JW,
Allen AA, Bastron JD, e t al: Limb myokymia. Muscle Nerve
1981 ;4:494-504, with permission.)
Hi Mj|
CONTINUOUS MUSCLE FIBER ACTIVITY usually begins in the lower limbs in the late teens, progressing
to all skeletal muscles. Acquired neuromyotonia is a result of
A number of relatively rare syndromes producing continuous autoantibodies directed against the peripheral nerve's voltage
muscle fiber activity have been r e p o r t e d . ' Portions of
110 134,197
gated potassium channels rendering the nerve hyperexcitable
both the central and peripheral nervous system have been impli- and prone to repetitive firing. - A rippling of the skin
115 168,203
cated in generating the sustained firing of motor units. One form (myokymia) is noted clinically, with patients usually sweating
of continuous muscle fiber activity is known as stiff-man syn- and possibly in pain from the constant muscle activity. Needle
drome or stiff-person syndrome. The motor unit discharges are recordings demonstrate a continuous type of motor unit firing
believed to have a central origin as they are abolished or attenu- with occasional bursts of motor unit potentials resembling
ated by peripheral nerve block, neuromuscular block, spinal myokymic discharges. Not infrequently, multiple discharges of
block, general anesthesia, and sleep. Additionally, the motor
159
the same motor unit (doublets or multiplets) can be seen in the
unit firing is diminished with diazepam but not phenytoin or above burst type of pattern. This type of discharge is to be dis-
carbamazepine. Progressive muscle stiffness involving all mus-
7
tinguished from the so-called neuromyotonic discharge where
cles including the chest wall and pharynx eventually occurs re- motor units discharge with frequencies up to 300 Hz associated,
sulting in contractures and profound disability. A needle with a characteristic "pinging" sound. These discharges may be
electrode recording reveals normal motor unit potentials pro- seen in Isaacs syndrome or rarely in other types of neurogenic
ducing a sustained interference pattern in both the agonists and disorders (see below). The neuromyotonic discharge is not the
antagonists. characteristic discharge observed in neuromyotonia despite its
A "peripheral" form of continuous motor unit activity is re- name, which can lead to some confusion. In the discharges re-
ferred to as Isaacs syndrome or neuromyotonia. The motor
125126
ferred to as "neuromyotonia," the amplitude of the firing motor
unit activity is eliminated by neuromuscular block but not pe- unit potentials eventually declines as single fibers fail to fire
ripheral nerve block, spinal or general anesthesia, or sleep. It from exhaustion (Fig. 7-22). These potentials are not influenced
Table 7-14. Characteristics of Myokymic Discharges 154
Appearance: Normal MUAPs

Rhythm: Regular
Frequency: 0.1 t o 10 Hz
Burst frequency: 20 t o 250 Hz
Stability: Persistent firing/occasional abrupt cessation
Observed in: a. Facial
Multiple sclerosis
ÛUlll^llll»i.iM.i^M.<.iVi H-U i^M,»ii,,| |,-,tMi, { u i U U i
1 l 1
Brainstem neoplasm
Polyradiculopathy
Bell's palsy
I — 1
Normal 100 ms
b. Extremity
Radiation plexopathy Figure 7-22. Neuromyotonic discharges. Recording of MUAPs
Chronic nerve compression from a patient with Isaacs syndrome. The continuous MUAP firing
Carpal tunnel syndrome eventually declines in number and amplitude as the muscle becomes
exhausted. (From Daube JA:AAEM Minimonograph N o . I I: N e e d l e
Radiculopathy
Examination in Electromyography. Rochester, MN, American Asso-
Rattlesnake venom
ciation of Electrodiagnostic Medicine, 1979, with permission.)
by voluntary contraction and may be induced by ischemia or

electrical nerve stimulation. Similar neuromyotonic discharges
may be observed in tetany, anticholinesterase overdose, and
spinal muscular atrophy.
CRAMPS
Cramps are sustained and possibly painful muscle contrac-
tions of multiple motor units lasting seconds or minutes that
may appear in normal individuals or specific disease states (Fig.
7-23). In healthy subjects, a cramp usually occurs in the calf w^-u— y~^-^p~^Y~*~~^f — y —-^v-—K~~V^— — k — 1 m'
muscles or other lower limb muscles following exercise, abnor- 1 *
100 ms
mal positioning, or maintaining a fixed position for a prolonged
period of t i m e . Cramps may also be induced by hypona-
148
Figure 7-23. Muscle cramp with MUAP frequencies between 3 0
tremia, hypocalcemia, vitamin deficiency, or ischemia. One 67
and 50 Hz. (From Daube JA:AAEM Minimonograph N o . I I: N e e d l e
may also note the occurrence of cramps in early motor neuron Examination in Electromyography. Rochester, MN, American A s s o -
ciation of Electrodiagnostic Medicine, 1979, with permission.)
disease and peripheral neuropathies. Familial syndromes in-
volving fasciculations and cramps; alopecia, diarrhea, and
cramps; and simply autosomal dominantly inherited cramps
198
MUSCLE GENERATORS OF ABNORMAL
have been described. 150
VOLUNTARY ACTIVITY
A needle recording electrode placed into a cramping muscle
will reveal multiple motor units firing synchronously between NEURAL LOSS
40 and 60 Hz and occasionally reaching 200-300 H z . A large172
portion of the muscle is simultaneously involved in a cramp as Motor Unit

opposed to the asynchronous excitation of motor units during If a muscle is completely denervated by a lesion affecting the
voluntary activation. Cramps are believed to arise from a pe- motor neurons or the peripheral nerve, there will be a complete
ripheral portion of the motor unit. A cramp accompanied by absence of MUAPs. Fibrillation potentials and PSWs will appear
electrical silence is a pathological contracture as seen in Mc- in great abundance following an appropriate period of time. A
Ardle's disease/ 48
partial nerve lesion, however, may result in profuse fibrillations
and PSWs, with preservation of some MUAPs. A muscle that is
MULTIPLET DISCHARGES totally denervated can be reinnervated only by regrowth of the
peripheral nerve along its original course. A partially denervated
A clinical syndrome manifested by spontaneous muscle muscle will consist of denervated motor units and intact motor
twitching, cramps and carpo-pedal spasm is known as tetany. units. The motor units deprived of their innervation may be rein-
This entity usually results from peripheral and/or central nervous nervated by one of two mechanisms. The first process involves
system irritability associated with systemic alkalosis, hypocal- a regrowth of axons along the previous neural pathways. 118
cemia, hyperkalemia, hypomagnesemia, or local ischemia. 139

Neural regrowth occurs at approximately 3 - 4 mm per day. - 9 98
Clinically one may induce tetany by tapping the facial nerve The second manner in which denervated muscle fibers can b e
(Chvostek's sign), the peroneal nerve at the fibular head (per- reinnervated is through collateral sprouting. That is, terminal
238
oneal sign), and inducing limb ischemia (Trousseau's sign). 108

axons from intact neighboring motor units sprout additional ter-
In the above conditions, characteristic motor unit potentials minal axons at their branch point nodes of Ranvier. Denervated
may be observed. A single motor unit potential may fire rather muscle fibers somehow direct the growth of new terminal
rapidly with an interdischarge interval of 2-20 ms twice (dou- sprouts through the remaining Schwann cell sheaths (bands of
blets) or three times (triplets) or more (multiplets) (Fig. 7-24). - I33 210
Bunger) resulting in muscle reinnervation. The motor unit ar-
83
These potentials can be seen following voluntary contraction or chitecture of the muscle is permanently and irreversibly altered
be observed spontaneously from induction maneuvers noted in this manner and present during the rest of this person's life.
above in which MUAPs may fire in long trains or short bursts of
II II !
5 - 3 0 Hz (tetany). As opposed to doublets and triplets, paired 1
discharges have an interspike interval of 20-80 ms and can arise
in similar states as previously described. 1 in "i fjr - f • r rJ
J)
TREMOR
Involuntary activation of multiple motor units arising from
the central nervous system in a semirhythmic pattern is called a
tremor. The individual motor units within a tremor burst do not
possess a consistent relationship of firing from one burst to the
next. The electrical activity produced from a tremor complex,
therefore, continuously changes with respect to the motor unit Figure 7-24. Doublet (top trace) and multiplets (bottom trace)
morphology contained in successive firings. - The sponta-
61 133
MUAPs resulting from voluntary contraction. (From Daube JA: AAEM

neous activity associated with a tremor may render a search for Minimonography N o . I I: N e e d l e Examination in Electromyography.
subtle abnormalities of fibrillations and positive sharp waves Rochester, MN, American Association of Electrodiagnostic Medicine,
extremely difficult. 1979, with permission.)
particular motoneuron now take on the characteristics of those

muscle fibers comprising the original motor unit. The muscle 141
-vV- fiber type matches the motoneuron irrespective of what it was

600 M-V
before. If type I muscle fibers are reinnervated by a type II mo-
toneuron through collateral sprouting, the type I fibers are even-
tually converted to type II fibers. Histologically staining the
muscle reveals the absence of the expected mosaic pattern and
— A - shows groups of muscle fibers all of the same histologic type,
600 n-V
i.e., "fiber type grouping." After the establishment of functional
neuromuscular junctions through the reinnervation process, it
takes about two weeks for muscle fiber type conversion to
occur. Following reinnervation, the muscle fibers comprising
230
1200 p.V a single motor unit may all be adjacent to each other compared
to the random and nonadjacent normal situation.
Electrophysiologic studies of neurogenic disease process
have demonstrated an interesting finding regarding the motor
2-6 MONTHS unit's territory. Recall that the human biceps muscle has a motor
POST-REINN
7 0 0 0 »JLV
unit territory approximating a 6 mm oval. - - After denerva-
27 28 207
tion and reinnervation, the motor unit territory remains essen-

tially unchanged. The number of muscle fibers within this
199
territory increases significantly. This suggests that although a

motor unit has a large capacity to support multiple muscle
6MO - 2 YR
POST-REINN fibers, its extent of reinnervation over distance is limited. These
findings also substantiated the observations that reinnervated
muscle fibers are all of the same physiologic type.
TOTAL DENERVATION
MUAP Findings Following Denervation/Reinnervation
100 uV
For discussion purposes, let us consider a situation in which a
Figure 7-25. M o t o r u n i t r e m o d e l i n g . Two motor units (type I light skeletal muscle has experienced an insult resulting in partial
circles and type II dark circles) are represented. Depolarization of one denervation. Within a week or two, depending upon the distance
m o t o r unit results in a 6 0 0 uV MUAP. Following degeneration of the between the lesion and muscle, fibrillation potentials and posi-
type II m o t o r unit ( 2 - 3 weeks), the intact MUAP still yields a 6 0 0 JJV tive sharp waves will be observed. Initially, there continues to
potential. Within 1 - 2 months the type II muscle fibers have atrophied be two distinct fiber types. The denervated muscle fibers will
and collateral sprouting is beginning t o reinnervate t h e m . T h e type I begin to atrophy, reducing the distance between fibers of the
m o t o r unit t e r r i t o r y has subsequently collapsed, generating a larger
same motor unit up to 30% within the first two months (Fig. 7-25).
MUAP ( 1 2 0 0 uV).As conduction occurs through the newly formed
terminal axons, the MUAP demonstrates a further increase in ampli-
Voluntary activation of the more closely packed muscle fibers
tude ( 7 0 0 0 pV) and number of phases. By 6 months, all muscle fibers comprising the intact motor units leads to increased spatial sum-
belonging t o the now larger motor unit are of the same fiber type, i.e., mation. The improved spatial summation of the single muscle
the type II fibers have been converted t o type I fibers.The MUAP may fibers' electrical activity results in MUAPs with larger amplitudes
decrease its amplitude and number of phases as the collateral sprouts and normal number of phases. Within 2-6 weeks, immature ter-
conduct action potentials m o r e rapidly than previously. (From minal sprouts are becoming functional thereby increasing the
Herbison GJ: Neuropathic Needle Examination. AAEM C o u r s e A: number of muscle fibers comprising the intact motor units. With
Fundamentals of EMG. Rochester, MN, American Association of reinnervation, the numbers of fibrillation potentials and positive
Electrodiagnostic Medicine, 1984, pp 2 1 - 2 5 , with permission.) sharp waves decreases until they have all disappeared with com-
plete reinnervation.
Recall that normal muscle tissue is a mosaic of muscle fibers Over the next several weeks, the reinnervated muscle fibers
from different motor units randomly distributed such that mus- will begin to take on the physiologic characteristics of its moto-
cle fibers from the same motor unit rarely contact each neuron. These reinnervated muscle fibers will increase the
230
other. - Following partial denervation, an intact motor unit

12 141
density of muscle fibers belonging to one motor unit for a given
may be surrounded by muscle fibers from other motor units that area. At first, the immature terminal sprouts and denervated
are no longer innervated. These denervated muscle fibers will muscle fibers conduct action potentials poorly producing a com-
induce the intact motor unit's terminal axons to sprout and paratively asynchronous summation of electrical activity within
become innervated. It is possible for a single motoneuron to in- the motor unit. It may also be possible for the conduction along
crease the number of muscle fibers it supplies between 4 and 5 an immature terminal axon to block completely. - The MUAP
199 208
times. - The minimum number of surviving motor units theo-

17 225
recorded by the needle electrode will reflect the asynchronous
retically required to reinnervate the muscle completely, there- electrical summation and have a morphology that is both larger
fore, is 20% (i.e., 80% of motor units lost). This may be (increased fiber density: more electrical activity) and more
understood if a muscle containing 100 motor units loses 80 polyphasic (increased temporal dispersion of action potentials)
motor units. The remaining 20 motor units can each reinnervate than normal. One may also observe a MUAP to vary in its ampli-
4 additional motor units for a total of 80 motor units thereby tude or possibly morphology. Remember that a normal MUAP
yielding a total of 100 motor units. The cytoarchitecture of the displays the same configuration from one firing to the next. If
reinnervated muscle is significantly altered compared to the an immature terminal axon blocks from conduction failure, a
normal situation. All of the muscle fibers reinnervated by a number of muscle fibers will not be activated to contribute their
electrical activity to the MUAP, and its morphology will reflect

this blocking by a decrease in MUAP amplitude. Over the en-
suing months the newly formed terminal sprouts mature and
conduction velocity increases and stabilizes. The electrical sum-
mation of all motor unit action potentials increases with an as-
sociated decrease in the number of phases detected in the
reinnervated MUAP. The configuration of the MUAP will also
stabilize with an associated disappearance of MUAP variability.
Reinnervation following profound or complete denervation
initially results in the newly formed terminal sprouts depolariz-
ing only a limited number of muscle fibers per motor unit.
These first few motor units yield MUAPs that are temporally
dispersed because of less than optimal terminal axon conduc- 10 ms
tion velocity (immature myelination). The asynchronous elec- Figure 7-26. Satellite potential. Four tracings of the same MUAP
trical summation of the relatively few muscle fibers that are shown with an associated satellite potential (arrow).The satellite
depolarize produce initially small highly polyphasic, and occa- potential is time-locked t o t h e main MUAP. It is possible for these p o -
sionally long-duration MUAPs. A characteristic sputtering tentials either t o precede o r follow the main MUAP.The satellite p o -
sound results because of the temporally dispersed single muscle tentials have been called: late components, parasite potentials, linked
fibers action potentials within the motor unit. These potentials potentials, and coupled discharges. (From AAEM Glossary of Terms in
were initially called nascent potentials but this term is discour- Clinical Electromyography. Muscle N e r v e 1987; I O(Suppl), 8S, with
aged in favor of simply describing this potential's morphology. permission.)
MUAP Findings in Primary Muscle Disease

MUSCLE LOSS Based upon the above findings in primary muscle disease, it
is possible to predict what might be observed on routine needle
MOTOR UNIT electromyography. A random loss of muscle fibers would
reduce the total number of muscle fibers comprising a motor
A number of primary muscle diseases are characterized by a unit. This loss of muscle fibers is reflected by a reduction ap-
disease process that randomly affects muscle fibers throughout proaching 30% in the motor unit territory as measured electro-
the muscle as a whole. - - The effect is to reduce the number
34 37 40
physiologically with a multilead electrode. One aspect of the
30
of muscle fibers comprising each motor unit. Muscle fibers still MUAP's duration arises from the extremes of the endplate
innervated also demonstrate an increase in variation of muscle fibers forming the initial and terminal positive phases of the
fiber diameter. The loss of muscle fibers combined with fiber MUAP. When these aspects of the MUAP are removed or re-
size changes may lead to a reduced motor unit territory. This duced in amplitude so as to be lost in baseline noise, the total
motor unit territory reduction is somewhat different from that duration of the measured MUAP can be expected"to decrease.
described for neurogenic lesions. There is an initial spatial col- The random loss of muscle fibers would diminish the smooth
lapse of the motor unit territory whereby intact muscle fibers summation of electrical potentials that normally produce less
from one motor unit are brought into closer proximity thereby than 4 MUAP phases. This loss of MUAP electrical subcompo-
yielding a larger amplitude potential in some neurogenic dis- nents could lead to an increase in the number of phases. Also an
eases. In muscle diseases, the muscle loss is apparently uni- alteration in fiber size combined with immature terminal axons
form enough so that even though muscle fibers may be may decrease the temporal summation of single muscle fibers
approximated, the reduced number of muscle fibers from all of thus also increasing the number of phases. Fiber splitting and
the motor units is more than enough to offset the electrical slow conduction along immature terminal sprouts have been
summation. Although the muscle fibers are closer together, suggested as a cause for the increased number of "linked" or
there are fewer muscle fibers remaining and an equivalent re- satellite potentials (Fig. 7-26). - - These are small poten-
147177 186 196
duction in the net electrical activity. These findings clinically tials, possibly single or just a few muscle fibers, time-locked to
manifest as muscle weakness (usually proximal), normal deep the main MUAP but appearing several milliseconds after the
tendon reflexes, preservation of sensory modalities, and normal primary MUAP. Satellite potentials may be detected in up to
limb tone. 40
10% of MUAPs in normal muscles and in 12% of neuropathic
Segmental necrosis of muscle tissue and fiber splitting have muscle. In some myopathies, however, 4 5 % of MUAPs may be
been reported in muscular dystrophies and polymyositis. - The
69 70
preceded or followed by satellite potentials. If one were to in-
segmental necrosis results in a portion of the muscle fiber being clude the satellite potential in the overall duration of an MUAP,
separated from the neuromuscular junction, effectively render- durations approaching 60 ms can be found. Thus, one can con-
ing that muscle segment denervated. This denervated aspect of clude that some myopathic processes produce MUAPs with ab-
the single muscle fiber will induce nearby terminal axons to normally long and not short duration MUAPs. Some investigators
sprout collateral nerves and result in reinnervation similar to have suggested that the diagnostic yield of the needle elec-
that described above for neurogenic processes. Also, fiber type tromyographic examination can be increased by carefully look-
grouping, which is suggestive of a neurogenic lesion, can be de- ing for satellite potentials since they are found so infrequently
tected in primary muscle diseases. Fiber splitting will yield two in healthy individuals. 97
muscle fibers belonging to the same motor unit in very close A reduction in MUAP amplitude has been postulated to occur
proximity. Both of these above events increases the number of because of the decrease in the total number of muscle fibers. 37
muscle fibers innervated by the same motoneuron in a relatively The suggested reduction of the motor unit territory may not
small area, i.e., an increase in motor unit fiber density. fully explain a decreased MUAP. Recall that the spike of the
MUAP arises from just a few single muscle fibers less than 500
urn from the needle electrode. Why should a total reduction in
the number of muscle fibers, therefore, result in a decreased
MUAP amplitude? The answer to this question has not been
fully explained but may be partially addressed. It is possible that |100MV
the spike of the motor unit does have some contribution from all 200 ms
or some portion of the muscle fibers in the motor unit. This
would explain why a random loss of muscle fibers could pro- I
i i
duce a small reduction in the MUAP amplitude. The histologic
I I fj
Mt
finding of fiber size variation may also play some role in MUAP
f (
amplitude. A smaller muscle fiber could be expected to generate
less voltage upon depolarization, producing a smaller MUAP ft 2mV
spike. Depending upon the size of the fiber nearest the record- 200 ms
ing electrode, which is the primary influence upon MUAP am-
plitude, one could anticipate both normal and abnormally small Figure 7-27. M U A P v a r i a b i l i t y . A,An MUAP is recorded from a
MUAPs. As the severity of the disease progresses, an increased healthy individual. N o t e the identical appearance of t h e MUAP with
number of smaller diameter fibers may occur, enlarging the per- each successive discharge. B.The same MUAP recorded from a person
centage of small MUAPs. Finally, an increase in electrically with Lambert-Eaton syndrome. N o t e the MUAP's highly variable mag-
measured fiber density can arise because of some physical col- nitude with each successive discharge.
lapse of the motor unit following muscle fiber atrophy or loss,
fiber spitting, and reinnervation secondary to segmental necro- total voltage associated with the observed MUAP. A slightly dif-
sis. It is also possible to see normal or even large amplitude ferent voltage for each sequential MUAP firing means that the
MUAPs. Locating the needle electrode tip immediately adjacent MUAP's amplitude will vary somewhat with each discharge,
to a hypertrophic muscle fiber may result in a relatively large thereby generating the so-called MUAP variability (Fig. 7-27).
MUAP spike. This situation is completely different for the normal motor unit,
A needle recording electrode placed in a muscle affected by a which has the same appearance with each successive discharge.
myopathic disease can show all of the above noted possibilities.
Small amplitude and short duration MUAPs with an increase in ABNORMAL MUAP RECRUITMENT
the number of phases accompanied by satellite potentials have
all been documented. Most myopathies reveal both normal and Following alteration of the motor unit secondary to pathol-
short duration MUAPs where the number of short duration ogy, one can expect to note abnormalities in MUAP recruit-
MUAPs increase with disease progression. - - The special
34 37 40
ment. Distinct recruitment patterns may be observed depending
technique of single-fiber EMG has demonstrated an increase in upon the type of disorder or which portion of the motor unit is
muscle fiber density. Of course, the segmental necrosis could
171
affected. Two major abnormal recruitment patterns may be de-
lead to the observation of fibrillation potentials and positive sharp tected. If a lesion damages the neural portion of the motor unit,
waves. - The designations "myopathic potentials" and "brief
69 70
a so-called "neurogenic recruitment" may be observed depend-
small abundant polyphasic potentials" (BSAPPs) have been dis- ing upon how many motor units are injured. On the other hand,
couraged in favor of more descriptive MUAP terminology re- a random loss of muscle tissue (myopathy) can result in a "my-
garding MUAP morphologic characteristics. This is because opathic recruitment" pattern. Although the sensitivity of recruit-
MUAPs are not pathognomonic of any disease in particular, but ment patterns in various disease states is most likely less than
more accurately describe possible disease processes in general. quantifying MUAP duration, - discussion regarding potential
72 183
findings is useful in understanding how different lesions can

MUAP Findings in Neuromuscular generate distinct recruitment findings.
Junction Disorders
There are three anatomic regions of the neuromuscular junc- Neurogenic Recruitment
tion that can be considered primary areas where a disorder may If the motoneuron's neural elements (anterior horn cell, pe-
occur: (1) presynaptic terminal (e.g., botulism, Lamber-Eaton ripheral nerve, terminal axons, and presynaptic aspect of the
syndrome), (2) synaptic space (e.g., organophosphate poisons, neuromuscular junction) are permanently affected by disease
congenital disorders), and (3) postsynaptic membrane (e.g., and Wallerian degeneration occurs, then all of the muscle fibers
medication, myathenia gravis). Diseases affecting the first and innervated by those injured axons will be denervated. As previ-
second anatomic regions are more likely to be encountered clin- ously noted, collateral sprouting will reinnervate these muscle
ically and can give rise to a particular kind of MUAP abnormal- fibers comparably increasing the number of muscle fibers be-
ity, i.e., motor unit variability similar to that previously longing to one motor unit. - When the patient calls upon the
29 238
described for the newly reinnervated MUAP. Normally, all of neuromuscular system to function, the interplay between the
the muscle fibers comprising a single motor unit all depolarize central nervous system and peripheral nerve-muscle compo-
and summate their individual electrical signals to generate an nents through the "servo-control" mechanism of the muscle
MUAP with the same configuration and magnitude from one spindles is inextricably altered. In effect, there are fewer motor
discharge to the next. In diseased or immature neuromuscular units available with larger complements of muscle fibers com-
junctions, failure of several individual muscle fibers comprising pared to the previous normal situation. The manner in which the
a motor unit to discharge with each firing of the anterior horn electrodiagnostic medicine evaluation recognizes this altered
cell results in a slightly different number of muscle fibers elec- state is through observing the MUAP recruitment pattern.
trically active at any particular time for that motor unit. As We must assume that the damage inflicted upon the moto-
result, each time the anterior horn cell fires, there is a different neuron is severe enough to compromise a sufficient number of
motor units to display an abnormality in motor unit recruitment. Table 7-15. Neurogenic Recruitment
The exact number of motor units necessary to meet this assump- Motor Unit Recruited
tion is unknown. For discussion purposes, let us suppose that in 1st (A) 2nd (B) 3rd (C) 4th (D)
our 4 motor unit example under "MUAP Recruitment (Normal
A (20 Hz)
Muscle)," motor units B and C are no longer present (Table 7-15).
Placing a needle recording electrode into the affected portion of A (25 Hz) 0 0
muscle will demonstrate a characteristic recruitment order for a A (30 Hz) 0 0 D (20 Hz)
minimal muscle contraction. Initially, motor unit A will begin Motor unit A begins firing at 20 Hz because motor units B and C are not pre-
firing at about 20 Hz and increase its firing frequency with an sent. When motor unit Afiresat 30 Hz, motor unit D finally becomes active at
increase in muscular e f f o r t .
72183
Normally, when motor unit A 20 Hz.The recruitment pattern is altered, and fewer motor units are firing at
reaches 10 Hz, a second motor unit, i.e., motor unit B should higher than anticipated rates.
begin firing. In this instance, however, motor unit B, is absent
because of disease, and motor unit A has already proceeded to a lesions, the number of motor units are unaffected, but the
rapid firing frequency to generate the force requested. The force muscle fiber content of each motor unit is reduced. In essence,
required by the individual must be met by the only physiologic the net force output from each motor unit is diminished com-
mechanism remaining, which is an increase in motor unit A's pared to normal. The manner in which the central-peripheral in-
firing rate. The first motor unit increases it firing rate still fur- teraction compensates for this situation is for multiple motor
ther approaching 25 Hz. Note that at 20 Hz, the recruitment units to begin firing simultaneously at high rates. This may be
ratio predicts that 4 motor units should be present. Under better understood if we return to our simple motor unit example.
normal conditions, when motor unit A reaches about 15 Hz, If the patient is requested to minimally contract a muscle af-
motor unit C should have become activated. Unfortunately, fected by a myopathic process, the initial force output for the first
motor unit C is also affected by a disease process and cannot recruited motor unit is less than normally anticipated. Motor unit
fire. At this point, the recruitment ratio states that 3 motor units A firing at 5 Hz produces insufficient force to satisfy that desired
should be present (15 Hz/5 = 3 motor units). As the patient tries by the patient. As a result, additional motor units are immediately
to produce still more force, motor unit A eventually reaches 30 added so they may contribute their muscle fibers to the net force
Hz, and motor unit D now fires. It also begins firing at an in- output. The result is 5 separate motor units (Table 7-16) for exam-
j creased rate of 20 Hz to meet the force demanded of the muscle. ple, firing at relatively high rates (15 H z ) . It is difficult for the
71183
We now have two separate motor units firing and the recruit¬ patient to recruit only one motor unit. The motor unit ratio states
' ment frequency is 30 Hz. That is, the firing rate of the first that three motor units should be present only if the first recruited
motor unit when the second motor unit became active is 30 Hz. motor unit is firing at 15 Hz. In our example, however, we have 5
There are a decreased number of motor units active for the fre- motor units firing at 15 Hz. The final observation on the CRT
quency of the first motor unit's firing rate. This is the so-called screen is that there are too many, or an increased number of,
neurogenic recruitment pattern also referred to as "decreased motor units firing for the amount of contraction requested, i.e.,
recruitment" or "reduced recruitment." A recruitment ratio of 15 motor units B-E fire earlier than normal leading jto the terms
(30/2 = 15) is noted in this example and is clearly abnormal as it "early or increased recruitment." This process differs from the
is significantly greater than 5. The loss of motor units may be so "neurogenic" scenario in that additional motor units with fewer
profound in some motor neuron or peripheral nerve disorders muscle fibers per motor unit are called upon for their contribution
that muscle activation results in only one motor unit firing at to the net force output. Because the motor units contain fewer
30-40 Hz. muscle fibers and produce less force than normal, the only strat-
A decreased number of motor units firing at abnormally high egy left is to have them fire both earlier and at faster rates. In our
frequencies may also be seen in the presence of normal motor example, a firing rate of 15 Hz for 5 motor units results in a re-
units whose neural conduction may be temporarily blocked. A cruitment ratio of 3, which is less than a normal ratio of 5.
potentially reversible conduction block along a nerve, neu- Myopathic processes, therefore, yield recruitment ratio values
rapraxia, may occur secondary to trauma, ischemia, and other less than 5 while neurogenic diseases generate recruitment ratio
conditions. Neurapraxia may also coexist with axonal loss, values greater than 5. Of course, 5 is a rough guide and some vari-
yielding a mixed lesion. A rather "pure" form of neurapraxia is ability (e.g., greater than 3 but less than 8) may be normal. This is
the common experience when a "foot goes to sleep" after the observed electrically as a rapid obliteration of the baseline sec-
crossing of legs for prolonged periods. The peroneal nerve fails ondary to mulitple MUAPs being activated with just a slight in-
to conduct efferent motor and afferent sensory impulses for the crease in attempted force production.
time necessary to reverse the ischemic block of neural impulses. It must be emphasized that the above description of motor unit
For the time of conduction failure, however, a number of motor recruitment is an oversimplification for illustrative purposes and
units are rendered nonfunctional. The muscle is effectively de- most likely does not accurately reflect the complexities of the in-
prived of multiple motor units. A recruitment pattern under con- teraction between the central nervous and peripheral neuromus-
ditions of neurapraxia is very similar to that seen in actual cular systems. The observations noted above can be practically
motor unit loss, i.e., decreased numbers of MUAPs firing at in-
creased rates. The recruitment pattern then returns to normal Table 7-16. Myopathic Recruitment
when all of the nerve fibers are capable of conducting electrical
Motor Unit Recruited
impulses to and from the muscle.
1st (A) 2nd (B) 3rd (C) 4th (D) 5th (E)
Myogenic Recruitment A (15 Hz) B(l5Hz) C(l5Hz) D(l5Hz) E(l5Hz)
The random loss of muscle fibers in diseases such as poly- A random loss of muscle fibers results in each motor unit containing a smaller
myositis or various muscular dystrophies will also yield charac- complement of musclefibers.For a given force output, therefore, more individ-
teristic recruitment patterns. Unlike the above "neurogenic" ual motor units must fire earlier and faster than normal.
applied to the first few recruited motor units to arrive at some from intramuscular pain receptors and subjective patient fac-
165
insight regarding possible lesions affecting the motor unit. The tors. Some of these factors impacting upon the patient's
I 0 3 1 3 2 2 0 6
sensitivity of motor unit recruitment analysis is unknown and view of the examination include: warnings about the test from
most likely does not precede changes in MUAP duration for as- well-meaning friends or poorly informed medical personnel,
sessing pathology of the motor unit. fear of needles, sound emanating from the instrument's speaker,
fear of an unfamiliar test, etc. Studies attempting to examine
206
RECRUITMENT IN STROKE some of these factors demonstrated that a patient's anxiety

level, perception of his or her own pain, and female gender were
There are few data examining in detail the differences in predictors of individuals who experience the most pain during
MUAP recruitment between healthy persons and those sustain- an examination. - Although the number of examination sites,
103 132
ing a stroke. The little information present suggests that the patient's age, length of examination or wait, day of week, previ-
motor units begin to fire at a lower rate (7 Hz vs. 8 Hz) in the ous examination, or time of day does not appear to adversely
hemiplegic limb, while the second recruited motor unit begins affect pain perception, several body areas were found to be
firing earlier (10 Hz vs. 15 Hz) than anticipated, given the first more uncomfortable than others. Specifically, cervical
motor unit's firing rate." These findings are more pronounced paraspinals were the most painful region of the body examined
in distal compared to proximal limb muscles. In all muscle followed by the lumbosacral paraspinal muscles and the hand
groups on the affected side, a diminished degree of voluntary intrinsic muscles. 206
control is found, as would be anticipated. Although the examiner cannot change the individual's per-
ception of pain, it may be advisable to attempt to reduce the
ADDITIONAL WAVEFORMS anxiety level. The use of analgesics has been recommended; 143
however, we do not use medication during the electrodiagnostic

Three additional waveforms that may be observed during the medicine consultation. Audio analgesia in the form of music or
needle electromyographic examination are discussed. It is white noise played through headphones may be of some assis-
relatively common to observed a sinusoidal wave representing tance in reducing the patient's perception of pain. As has been
206
60 Hz interference in the U.S. and 50 Hz interference in other previously noted, information delivered in a calming and reas-
parts of the world. Detection of this waveform usually means suring voice might help some persons through the consultation.
there is less than optimal impedance matching between the elec- It is the belief of these authors and others that monopolar nee-
201
trodes or even an open circuit. When this waveform is recorded, dles cause less pain than standard concentric needle electrodes
all electrodes must be checked to ensure that none have de- probably because Teflon reduces the needle shaft's drag on the
tached from the patient and that there is adequate electrode tissue.
paste between the electrodes and the patient. Also, a broken
wire may be present beneath the plastic coating of the elec- CONTRAINDICATIONS AND COMPLICATIONS
trodes. After these obvious sources have been checked, the in- RELATED TO NEEDLE EXAMINATION
strument's preamplifier should be positioned close to the patient
and away from the instrument or other potential sources of in- In the authors' opinion there are no absolute contraindica-
terference (monitors, printers, medical equipment, etc.). Placing tions to performing a well-directed and selective needle elec-
an ungloved hand on the patient can help reduce this type of in- tromyographic examination. A relative contraindication to
terference. Additional sources of interference are discussed in inserting electromyographic needles into muscle tissue is most
the chapter dealing with instrumentation. assuredly persons who have a coagulopathy either medically in-
When recording with a needle electrode at essentially any loca- duced or otherwise acquired. - It is certainly possible for indi-
4 48
tion on the patient's torso, a regular occurring waveform that is viduals on a sufficient dosage of sodium warfarin (Coumadin)
relatively silent may be observed. Slowing the instrument's to bleed significantly following a needle study. The propensity
sweep speed will reveal that the firing rate of this potential is about to bleed excessively following needle electromyography is in-
1 Hz, i.e., the waveform is the patient's EKG. This waveform can creased in patients with a platelet count below 50,000/mm , a 3
be particularly prominent when recording from the chest or back prothrombin time 1.5-2 times control values, and intravenous
when examining the shoulder rotators or the serratus anterior mus- heparin with a concomitant partial thromboplastin time above
cles. A related waveform with a similar firing rate but of larger 1.5-2 times the control level. In these patients, the referring
4
magnitude results when a patient has a pacemaker. This wave- physician and electrodiagnostic consultant must weigh the risk-
form typically has a loud high-pitched sound associated with it. benefit ratio of examining the patient with needle electrodes.
Should a study be deemed necessary, a limited number of elec-
trode insertions should be performed in a well-selected popula-
NEEDLE ELECTROMYOGRAPHY: tion of muscles to yield the greatest amount of information with
RELEVANT ISSUES the least number of needle punctures. It is also prudent to apply
localized pressure to the puncture site following needle re-
moval. A needle examination in hemophiliacs should be de-
PAIN ON NEEDLE EXAMINATION
ferred unless their clotting factors are optimized.
Unfortunately the electrodiagnostic medicine consultation There may be some hesitancy in performing a needle exami-
can be an uncomfortable experience for the patient. The pain nation on a limb with significant lymphedema because of an in-
perceived during a study can limit the investigation and the creased risk of infection. Again, the risk-benefit ratio must be
amount of information gained. Pain perception does not solely consider prior to engaging in a needle investigation. In the au-
arise from the insertion of a needle electrode. To be sure, punc- thors' opinion, one should proceed with caution in such patients
turing the skin and particularly the superficial fascia accounts and perform only a limited number of needle insertions after thor-
for most of the examiner-delivered pain with some contribution oughly preparing the skin insertion site with an iodine solution.
Soft tissue infections secondary to needle electromyography having Jakob-Creutzfeldt disease or other disorders suspected
have been reported but are rare. Five cases of a nontuberculous of being tranmitted through slow viruses should be discarded
mycobacteria (Mycobacterium fortuitum) were postulated to following autoclaving for 1-1.5 hours at 120°C at 20 pounds
have been the result of a needle electromyographic examination per square i n c h . It is the authors' personal preference to use
101160
performed by a single individual. Although the sterilization

170
disposable needle electrodes on all patients receiving a routine
procedures for the nondisposable needle were adequate, the electrodiagnostic medicine consultation. This allows one the
report appeared to implicate the needle examination as the cul- ability to inform the patient that the needle about to be inserted
prit for the infections. Some electrodiagnostic medicine practi- is sterile, never been used before, and will be discarded once the
tioners clean the needle insertion sites with alcohol prior to examination is completed. Most persons appreciate this consid-
inserting the needle electrode. The value of this practice is ques- eration. Standard concentric and monopolar needle electrodes
tionable if the intent is to sterilize the skin and may result in are now manufactured with sufficient quality control to provide
more pain upon needle insertion if the alcohol is driven into the high quality electrodes for acquiring routine data, obviating the
subcutaneous tissues when not allowed to dry. need to use nondisposable needle electrodes. Unfortunately, this
It is certainly a good practice for practitioners to employ the cannot be said of special needle electrodes such as single-fiber
use of nonsterilized gloves to protect themselves from blood and macro EMG needles, which require sterilization procedures
products and bloodborne pathogens contaminating their hands as they are rather expensive to discard after one use. Most rou-
following withdrawal of the needle electrode. - Once the
50 51
tine sterilization procedures used in hospital settings are suffi-
gloves have been used, they should be appropriately discarded. cient to destroy bloodborne pathogens including the HIV virus
It is inappropriate to wash or otherwise attempt to disinfect the and can be used for such needles with the exception noted
gloves as this may cause undetected disintegration of their in- above. - Solutions of sodium hypochlorite (household bleach)
50 51
tegrity. Use of gloves is an inconvenience as the "feel" of the in dilutions of 1:100 to 1:10 have been deemed effective as a
needle electrode penetrating various tissue planes or encounter- germicide.
ing fibrous tissue is dulled, not to mention decreased manual The electrodiagnostic medicine practitioner is certainly at
dexterity and adhesive tape sticking to the gloves. Given all of risk for needle punctures. This is likely to occur when attempt-
the preceding reasons not to use gloves, one can accommodate ing to resheathe the needle electrode either between insertions
the sensation of performing a needle examination with gloves. or when the examination is complete. A good practice is to
60
In any event, even though the risk of human immunodificiency have some type of simple mechanical device capable of holding
virus (HIV) after skin contact with infected blood is less than the needle's protective covering so that the practitioner does not
the 0.5% for needle puncture exposure, - one should never
50 51
have to hold this covering and risk a self-inflicted puncture. A
apply pressure to a bleeding needle site with unprotected skin. It large sterile foam block may also serve this purpose as one can
is also good practice to wash one's hands with an appropriate readily impale the needle into the foam and move it to the next
disinfectant soap prior to and following a needle examination. needle location. Following completion of the needle study, the
Various complications have been documented to result from needle and foam block can be discarded.
the needle electromyographic examination. Urticaria changing
into vitiligo has been reported in a hysterical individual. More MUSCLE BIOPSY/SERUM CK LEVEL
229
serious complications have been the production of pneumotho-

races following needle studies. Reports document that it is pos- Inserting a hypodermic needle or electromyographic needle
sible to induce a pneumothorax after inserting a needle recording electrode into muscle tissue results in fiber damage or
electrode too far superiorly above the scapular margin when at- so-called "needle myopathy." In both humans and animals,
94
tempting to examine the supraspinatus muscle. Placing the

191
placing a needle electrode into various muscles produced intra-
needle too far laterally from the midline when investigating cer- muscular evidence of muscle fiber destruction. Muscle biopsies
vical paraspinal muscles can also penetrate the pleural performed parallel to the plane of needle insertion site revealed
cavity. 122123
Concern has also been expressed about serratus an- evidence of muscle fiber necrosis within 1-24 hours and phago-
terior (possible pneumothorax) and abdominal muscle (peri- cytosis at these regions within 1-3 days. After 3 days biopsy
toneal cavity insertion) needle examination as well as specimens demonstrated large nuclei with prominent nucleoli
diaphragmatic investigation. It is recommended that one should
4
and accompanying basophilic cytoplasm suggesting muscle
stay rather close to the midline when studying paraspinal mus- fiber regeneration. A biopsy specimen taken perpendicular to
cles, particularly in the cervical region, and place the needle the plane of the needle's tract would reveal well-localized
toward the medial border of the scapula for the supraspinatus region of muscle destruction and phagocytosis. This appearance
muscle. The authors suggest that when examining any muscle, may be mistaken for inherent muscle disease and lead to an er-
particularly those prone to result in adverse consequences, the roneous diagnostic conclusion. It is recommended that muscle
amplifier should be activated immediately upon entering the biopsies not be performed on muscle irrespective of their clini-
subcutaneous tissue. Once insertional activity is detected, the cal involvement if an electromyographic examination has been
plane of exploration should be altered to avoid any potentially carried out. The biopsy should instead be taken from a similarly
hazardous structures. If insertional activity disappears during involved muscle not studied with a needle electrode.
the examination, the needle should be withdrawn toward the If the patient is to undergo an electrodiagnostic medicine ex-
surface and carefully repositioned. amination prior to receiving a muscle biopsy, only one side of
the body should be investigated. The practitioner should then
STERILIZATION OF NEEDLE ELECTRODES clearly state what muscles were examined on which side of the
body and communicate that these limbs should not be used for
Although there are guidelines for properly sterilizing reusable muscle biopsy. On the other hand, if a needle examination is
needle electrodes through steam autoclaving, those needles performed following a muscle biopsy, the needle should not be
known or suspected to have come in contact with patients located in muscle tissue surrounding the biopsy site. Membrane
instability can be detected in muscle tissue separated from its 8. Barchi RL: A mechanistic approach to the myotonic syndromes. Muscle Nerve
1982;5:S60-S63.
nerve supply as a result of the muscle biopsy. 178
9. Berenberg RA, Forman DS, Wood DK, et al: Recovery of peripheral nerve
As previously stated, needle insertion causes rather obvious function after axonotomy. Effect of triiodothyronine. Exp Neurol 1977;57:
muscle fiber disruption on microscopic examination. It is rea- 349-363
sonable to conclude that interrupting the muscle membrane 10. Bolton CF, Sawa GM, Carter K: The effects of temperature on human com-
pound action potentials. J Neurol Neurosurg Psychiatry 1981;44:407^413.
could result in serum elevations of creatine kinase (CK). The 11. Bowman WC, Raper C: Spontaneous fibrillatory activity of denervated muscle.
level of serum CK is a reflection of this substance being re- Nature 1964;201:160-162.
leased from various body tissues of which the highest content is 12\ Brandstater ME, Lambert EH: Motor unit anatomy. Type and spatial arrange-
ment of muscle fibers. In Desmedt JE (ed): New Developments in Electro-
found in decreasing order: skeletal muscle, heart muscle, and
myography and Clinical Neurophysiology. Vol 1, Basel, Karger, 1973, pp
brain tissue. As CK is a measure of muscle function, an eleva-
53
14-22.
tion following an electromyographic examination could lead to 13. Brazier MAB: Electrical Activity of the Nervous System. London, Pitman
the diagnosis of muscle disease when the true cause may be the Medical Publishing Co., 1977, pp 51-74.
14. Brodal A: Neurological Anatomy in Relation to Ciinical Medicine. New York,
needle study. Multiple human and animal studies reveal that Oxford Univ Press, 1969, pp 117-150.
there is a transient elevation of the CK level following the 15. Brooke MH, Kaiser KK: Muscle fiber types: How many and what kind? Arch
needle i n v e s t i g a t i o n . Specifically in normal individu-
52J5818K215
Neurol 1970;23:369-379.
als, a peak CK level of 1.5 times normal (still within normal 16. Brooke MH, Kaiser KK: The use and abuse of muscle histochemistry. Ann N Y
Acad Sci 1974;228:121-144.
limits) is detected 6 hours post-needle study with a return to 17. Brown MC, Ironton R: Sprouting and regression of neuromuscular synapses in ,
baseline values by 48 hours. This time course and magnitude
53
partial denervated mammalian muscles. J Physiol 1978;278:325-348.
of elevation are similar to those found for post-exercise mea- 18. Brown WF: A method for estimating the number of motor units in the thenar
muscles and the changes in motor unit count with aging. J Neurol Neurosurg
surements. At no time did the CK level reach an abnormal Psychiatry 1972;35:845-852.
range. There is a normal diurnal variation of CK which may 19. Brown WF, Milner-Brown HS: Some electrical properties of motor units and
combine with exercise and a needle examination to yield a their effects on the methods of estimating motor unit numbers. J Neurol
false-positive result. Needle electromyography should not by Neurosurg Psychiatry 1976;39:249-257.
20. Brown WF, Varkey GP: The origin of spontaneous electrical activity at the end-
itself result in abnormal CK values. An elevation of the CK level plate zone. Ann Neurol 1981;10:557-560.
into the abnormal range following a needle study may result if 21. Brown WF: The Physiological and Technical Basis of Electromyography.
the individual has a neuromuscular disease and should raise the Boston, Butterworth Publishers, 1984, pp 317-368.
22. Brumlik J, Drechsler B, Vannin TM: The myotonic discharge in various neuro-
suspicion of possible pathology.
logical syndromes: A neurophysiologic analysis. Electromyography 1970; 10:
369-383.
23. Brunt ER, van Weerden TW: Familial paroxysmal kinesigenic ataxia and con-
tinuous myokymia. Brain 1990;113:1361-1382.
CONCLUSION 24. Buchthal F, Guld C, Rosenfalek P: Action potential parameters in normal
human muscle and their dependence on physical variables. Acta Physiol Scand I
The needle electromyographic examination is a powerful 1954;32:200-218.
25. Buchthal F, Pinelli P, Reosenfalck P: Action potential parameters in normal
method for exploring the dynamic physiologic status of the cen-
human muscle and their physiological determinants. Acta Physiol Scand
tral/peripheral neuromuscular system. Characteristic normal 1954;32:219-229.
and abnormal waveforms in combination with the patient's his- 26. Buchthal F, Rosenfalek P: Action potential parameters in different human mus-
tory and physical examination can assist the practitioner to ac- cles. Acta Psychiatr Scand 1955;30:125-131.
27. Buchthal F, Guld C, Rosenfalek P: Volume conduction of the spike of the motor
curately diagnose multiple potential diseases affecting the unit potential investigated with a new type of multielectrode. Acta Physiol
motor unit. To a large degree, the results obtained through this Scand 1957;38:331-354.
portion of the electrodiagnostic medicine consultation are an 28. Buchthal F, Guld C, Rosenfalek P: Multielectrode study of the territory of a
extension of the physical examination on a cellular level. It is motor unit. Acta Physiol Scand 1957;39:83-104.
29. Buchthal F, Rosenfalek P: Rate of impulse conduction in denervated human
important to remember that the various potentials described in muscle. Electroenceph Clin Neurophysiol 1958;10:521-526.
this chapter generally do not point to a specific diagnosis, but 30. Buchthal F, Rosenfalek P, Ermino F: Motor unit territory and fiber density in
only provide insights into the underlying pathophysiologic myopathies. Neurology 1960; 10:389^*08, 1960.
31. Buchthal F: The general concept of the motor unit. Res Publ Assoc Res Nerv
process affecting either the nerve, muscle, or both. The knowl- Ment Dis 1961;38:3-30.
edge and experience of the practitioner combined with the in- 32. Buchthal F. Rosenfalek P: Spontaneous electrical activity of human muscle. {
formation gained in light of the patient's presenting complaint Electroenceph Clin Neurophysiol 1966;20:321-336. j
are essential to formulating a correct diagnosis and prognosis. 33. Buchthal F: Electromyography. In Remond A (ed): Handbook of Electro-
encephalography and Clinical Neurophysiology. Amsterdam, Elsevier, 1976.
34. Buchthal F: Electrophysiologic signs of myopathy as related with muscle
biopsy. Acta Neurol 1977;32:1-29. "
REFERENCES 35. Buchthal F, Schmalbruch H: Motor unit of mammalian muscle. Physiol Rev
1980;60:90-142.
1. AAEM Glossary of Terms in Clinical Electromyography. Muscle Nerve 36. Buchthal F: Fibrillations: Clinical electrophysiology. In Culp WJ, Ochoa J
1987;10:(Suppl), 8S. (eds): Abnormal Nerves and Muscle Generators. New York, Oxford University
2. Adrian ED, Bronk DV: Discharge of impulses in motor nerve fibers. J Physiol Press, 1982, pp 632-662.
1929;67:i 19-151 37. Buchthal F: Electromyography in evaluation of muscle diseases, in Aminoff
3. Albers JW, Allen AA, Bastron JD, et al: Limb Myokymia. Muscle Nerve MJ (ed): Neurologic Clinics. Philadelphia, WB Saunders Vol 3, 1985, pp 573¬
1981;4:494-504. 598.
4. AAEM Professional Standards Committee. Guidelines in Electrodiagnostic 38. Buller AJ, Lewis DM: Factors affecting the differentiation of mammalian fast
Medicine. Rochester MN, American Association of Electrodiagnostic Medicine and slow muscle fibers. In Gutmann E, Hink P (eds): The effect of use and
1984, pp 1-12. disuse on neuromuscular functions. Prague, Publishing House of the
5. Andreassen S, Rosenfalek A: Recording from a single motor unit during strong Czechoslovak Academy of Sciences, 1963, pp 149-159.
effort. IEEE Trans Biomed Eng 1978;25:501-508. 39. Buller AJ: The physiology of the motor unit, in Walton JN (ed): Disorders of
6. Andreassen S, Rosenfalek A: Relationship of intracellular and extracellular Voluntary Muscle (3rd ed). Edinburgh. Churchill Livingstone, 1974, pp 20-30.
action potentials of skeletal muscle fibers. CRC Crit Rev Bioeng 1981 ;6: 40. Burke JM: Electromyography in myopathies, in AAEM Course A:
267-306. Fundamentals of Electromyography. Rochester, Minn, 1989, pp 41-50.
7. Ashizawa T, Butler JJ, Harai Y. et al: A dominantly inherited syndrome with 41. Burke RE: Motor unit types of cat triceps surae muscle. J Physiol 1967;
continuous motor neuron discharges. Ann Neurol 1983;13:285-290. 193:141-160.
42. Burke RE, Nelson PG: Accommodation current ramps in motor neurons of fast 75. Dumitru D, Walsh NE: Practical instrumentation and common sources of error.
and slow-twitch motor units. Int J Neurosci 1971;1:347-356. Am J Phys Med 1988;67:55-65.
43. Burke RE, Levine DN, Tsairis P, et al: Physiological types and histochemical 76. Dumitru D: Volume conduction: Theory and application. In Dumitru D (ed) :
profiles in motor units of the cat gastrocnemius. J Physiol 1973;234:723-748. Clinical Electrophysiology. Philadelphia, Hanley & Belfus, 1989, pp 665-681.
44. Burke RE, Tsairis P: Anatomy and innervation ratios in motor units in cat gas- 77. Dumitru D, Walsh NE: Electrophysiologic instrumentation. In Dumitru D (ed):
trocnemius. J Physiol 1973;234:749-765. Clinical Electrophysiology. Philadelphia, Hanley & Belfus, 1989, pp 683-699.
45. Burke RE, Strick PL, Kanda K, et al: Anatomy of medial gastrocnemius and 78. Dumitru D: Single muscle fiber discharges (insertional activity, end plate poten-
soleus motor nuclei in cat spinal cord. J Neurophysiol 1977;40:667-680. tials, positive sharp waves, and fibrillation potentials): A unifying proposal.
46. Burke RE: Motor units in mammalian muscle. In Sumner AJ: The Physiology Muscle Nerve 1996; 19:221 -226, 227-230.
of peripheral nerve disease. Philadelphia, WB Saunders, 1980, pp 133-194. 79. Dumitru D. King JC: Fibrillation potential amplitude after denervation. Am J
47. Butler BP, Ball RD, Albers JW: Effect of electrode type and position on motor Phys Med Rehabil 1998;77:483-489.
unit action potential configuration. Muscle Nerve 1982;5:S95-S97. 80. Dumitru D, King JC, Stegeman DF: Endplate spike morphology: A clinical and
48. Butler ML. Dewan RW: Subcutaneous hemorrhage in a patient receiving anti- simulation study. Arch Phys Med Rehabil 1998;79:634-640.
coagulant therapy: An unusual EMG complication. Arch Phys Med Rehabil 81. Eccles JC, Sherrington GS: Number and contraction values of individual motor
1984;65:733-734. units examined in some muscles of the limb. Proc Roy Soc London (Biol)
49. Campbell MJ, McComas AJ, Petito F: Physiological changes in aging muscle. J 1930;106:326-357.
Neurol Neurosurg Psychiatry 1973;36:174-182. 82. Eccles JC, Fatt P, Kokitser K: Cholinergic and inhibitory synapses in a pathway
50. Center for Disease Control. Recommendations for prevention of HIV transmis- from motor-axon collaterals to motoneurones. J Physiol 1954;126:524-562.
sion in health care settings. MMWR 1987;36(Suppl 2S). 83. Edds MV: Collateral regeneration of residual motor axons in partially dener-
51. Center for Disease Control. Update: Universal precautions for prevention of vated muscle. J Exp Zool 1950; 113:507-552.
transmission of human immunodificiency virus, hepatitis B virus, and other 84. Edstrom L, Kugelberg E: Histochemical composition distribution of fibers and
bloodborne pathogens in health care setting. MMWR 1988;37:377-382, fatiguability of single motor units. Anterior tibial muscle of the rat. J Neurol
387-388. Neurosurg Psychiatry 1968;31:424-438.
52. Cherington M, Lewin E, McCrimmon A: Serum creatine phosphokinase changes 85. Eisen A, Karpati G, Carpenter S, et al: The motor unit profile of the rat soleus in
following needle electromyographic studies. Neurology 1968;18:271-272. experimental myopathy and reinnervation. Neurology 1974;24:878-884.
53. Chrissian SA, Stolov WC, Honglodarom T: Needle electromyography: Its effect 86. Eisenberg BR, Kuda AM: Discrimination between fiber populations in mam-
on serum creatine phosphokinase activity. Arch Phys Med Rehabil 1976;57: malian skeletal muscle using ultrastructural parameters. J Ultrastruc Res
114-119.' 1976;54:76-88.
54. Christensen E: Topography of terminal motor innervation in striated muscles 87. Ekstedt J: Human single muscle fiber action potentials. Acta Physiol Scand
from stillborn infants. Am J Phys Med 1959;38:17-30. 1964;(Suppl 226): 1-96.
55. Chu-Andrews J, Johnson RJ: Electrodiagnosis: An Anatomical and Clinical 88. Ekstedt J, Stalberg E: How the size of the needle electrode leading-off surface
Approach. Philadelphia, JB Lippincott Co., 1986, pp 199-244. influences the shape of the single muscle fiber action potential in electromyog-
56. Chu J, Chan R-C: Changes in motor unit action potential parameters in raphy. Comp Prgms Biomed 1973;3:204-212.
monopolar recordings related to filter settings of the EMG amplifier. Arch Phys 89. Elmqvist D, Johns TR, Thesleff S: A study of some electrophysiological prop-
Med Rehabil 1985;66:601-604. erties of human intercostal muscle. J Physiol 1960;154:602-607.
57. Chu J, Bruyninckx F, Chan R-C: Significance of motor unit action potential pa- 90. Elmqvist D, Hofmann WW, Kugelberg J, Quastel DMJ: An electrophysiologi-
rameters in normal and neurogenic situations. Electromyogr Clin Neurophysiol cal investigation of neuromuscular transmission in myasthenia gravis. J Physiol
1986;26:405-479. 1964;174:417-434.
58. Chu J, Chan R-C, Bruyninckx F: Effects of the amplifier filter settings on the 91. Elmqvist D, Quastel DMJ: A quantitative study of endplate potentials in iso-
motor unit action potential parameters recorded with concentric and monopolar lated human muscle. J Physiol 1965;178:505-529.
needles. Electromyogr Clin Neurophysiol 1986;26:629-639. 92. Emeryk B, Hausmanowa-Petrusewicz 1, Nowak T: Spontaneous volleys of
59. Chu J, Chan R-C, Bruyninckx F: Progressive Teflon denudation of the monopo- bizarre high frequency potentials in neuromuscular diseases. Electromyogr Clin
lar needle: Effects on motor unit potential parameters. Arch Phys Med Rehabil Neurophysiol 1974;14:339-354.
1987;62:36-40. 93. Engel AG, Gomez MR, Seybold ME, et al: The spectrum of diagnosis of acid
60. Dalakas MC, Pezeshkpour GH. Neuromuscular diseases associated with the maltase deficiency. Neurology 1973;23:95-106.
human immunodificiency virus infection. Ann Neurol 1988;23:S38-S48. 94. Engel WK: Focal myopathic changes produced by electromyographic and hy-
61. Daube JA: AAEM Minimonograph #11: Needle Examination in Electro- podermic needles: "needle myopathy." Arch Neurol 1967;16:509-511.
myography. AAEM Rochester, Minn, 1979. 95. Feasby TE, Brown WF: Variation of motor unit size in the human extensor dig-
62. Daube JA, Kelly JJ, Martin RA: Facial Myokymia with polyradiculoneuropa- itorum brevis and thenar muscles. J Neurol Neurosurg Psychiatry 1974;37:
thy. Neurology 1979;29:662-669. 916-926.
b3. Daube JA: AAEM Minimonograph #18: EMG in Motor Neuron Disease. 96. Feinstein B, Lindegard B, Nyman E, et al: Morphologic studies of motor units
Rochester, Minn, 1982. in normal human muscles. Acta Anat 1955;23:127-142.
64. De Lisa JA Brozovich FV: AAEM Minimonograph #10: Volume conduction in 97. Finsterer J, Mamoli B: Satellite potentials as a measure of neuromuscular disor-
electromyography. American Association of Electrodiagnostic Medicine, ders. Muscle Nerve 1997;20:585-592.
Rochester, Minn, 1979. 98. Forman DS, Berenberg RA: Regeneration of motor axons in the rat sciatic
65. Denier Van der Gan JJ, Gielen CCAM, ter Hgar Romeny BM: Changes in re- nerve studied by labeling with axonally radioactive proteins. Brain Res
cruitment threshold of motor units in the human biceps muscle. J Physiol 1978;156:213-225.
1982:328:28P. 99. Frascarelli M, Mastrogregori L. Conforti L: Initial motor unit recruitment in pa-
66. Denny-Brown D, Pennybacker JB: Fibrillation and fasciculation in voluntary tients with spastic hemiplegia. Electromyogr Clin Neurophysiol 1998;38:
muscle. Brain 1938;61:311-332. 267-271.
67. Denny-Brown D, Foley JM: Myokymia and the benign fasciculation of muscu- 100. Freund HJ: Motor units and muscle activity in voluntary motor control. Physiol
lar cramps. Trans Assoc Am Phys 1948;61:88-96. Rev 1983;63:387-436.
68. Desai J, Swash M: Fasciculations: What do we know of their significance. J 101. Gajdusek DC, Gibbs CJ, Asher DM, et al: Precautions in medical care of han-
Neurol Sci 1997; 152(Suppl 1 ):S43-S48. dling materials from patients with transmissable virus dementia (Creutzfeldt-
69. Desmedt JE. Borenstein S: Regenerative phenomena in muscular dystrophy and Jakob disease). N Engl J Med 1977;297:1256-1258.
polymyositis: An electromyographic study. In Milhorat AT (ed): Exploratory 102. Gamstorp I: Adynamia episodica hereditaria. Acta Paediat 1956;(Suppl
Concepts in Muscular Dystrophy. No 333, International Congress Series. 108): 1-126.
Amsterdam, Excerpta Medica. 1974, pp 555-559. 103. Gans BM. Kraft GH: Pain perception in clinical electromyography. Arch Phys
70. Desmedt JE. Bomstein S: Relationship of spontaneous fibrillation potentials, to . Med Rehabil 1977;58:13-16.
muscle fiber segmentation in human muscular dystrophy. Nature 1975;258: 104. Gassel MM: Sources of error in motor nerve conduction studies. Neurology
531-534. 1964;14:825-835.
71. de Weerd JPC: Volume conduction and electromyography. In Notermans SLH 105. Gath I, Stalberg E: Techniques for improving the selectivity of electromyo-
(ed): Current Practice of Clinical Electromyography. Amsterdam, Elsevier, graphic recordings. IEEE Trans Biomed Eng 1976;23:467-472.
1984. pp 9-28. 106. Gath I, Stalberg E: On the volume conduction in human skeletal muscle: In situ
72. Dorfman LJ, Howard JE, McGill KC: Motor unit firing rates and firing variability measurements. Electroenceph Clin Neurophysiol 1977;43:106-110.
in the detection of neuromuscular disorders. Electroenceph Clin Neurophysiol 107. Gath I, Stalberg E: The calculated radial decline of the extracellular action po-
1989;73:215-224. tential compared with the in situ measurements in the human brachial biceps.
73. Dubowitz V, Pearse AGE: Reciprocal relationship of phosphorylase and oxida- Electroenceph Clin Neurophysiol 1978;44:547-552.
tive enzymes in skeletal muscle. Nature 1960;185:701-702. 108. Gilliat RW: Paresthesiae. in Culp WJ, Ochoa J (ed): Abnormal Nerves and
74. Dubowitz V, Brooke MH: Muscle Biopsy—A Modern Approach. London, WB Muscles as Impulse Generators. New York, Oxford University Press, 1982, pp
Saunders, 1973, pp 5-253. 477-489.
109. Goodgold J, Eberstein A: Electrodiagnosis of Neuromuscular Diseases (3rd ed). 145. Lambert EH: Studies of the origin of the positive sharp wave in electromyography.
Baltimore, Williams & Wilkins, 1983, pp 65. Newsletter, American Association of Electromyography and Electrodiagnosis;
110. Gordon EE, Januszko DM, Kaufman L: A critical survey of the stiff-man syn- Rochester, Minnesota, 1957;3:3.
drome. Am J Med 1967;42:582-599. 146. Lambert EH: Electromyography in amyotrophic lateral sclerosis. In Norris FH
111. Grimby L, Hannerz J: Recruitment order of motor units on voluntary contrac- Jr, Kurland LT (eds): Motor Neuron Diseases. New York, Grune & Stratton,
tion: Changes induced by proprioceptive afferent activity. J Neurol Neurosurg 1969, pp 135-153.
Psychiatry 1968;31:565-573. 147. Lang AH, Partanen J: "Satellite potentials" and the duration of potentials in
112. Gross PT: How to perform a comfortable but revealing needle examination. nonnal, neuropathic and myopathic muscles. J Neurol Sci 1976;27:513-524.
AAEM Course C: Standard Needle Electromyography of Muscles. Rochester, 148. Layzer RB, Rowland LP: Cramps. N Engl J Med 1971;285:30-31.
Minn, 1988, p p 9 - 1 3 . 149. Layzer RB: The origin of muscle fasciculations and cramps. Muscle Nerve
113. Hakansson CH: Action potentials recorded intra-and extracellularly from the 1994;17:1243-1249.
isolated frog muscle fibre in Ringer's solution and in air. Acta Physiol Scand 150. Lazaro RP, Rollinson RD, Fenichel GM: Familial cramps and muscle pain.
1957;39:291-312. Arch Neurol 1981;38:22-24.
114. Hakansson CH: Action potential and mechanical response of isolated cross stri- 151. Lenman JAR, Ritchie AE: Clinical Electromyography. Philadelphia, Lippincott,
ated frog muscle fibres at different degrees of stretch. Acta Physiol Scand 1970, pp 100.
1957;41:199-216. 152. Le'yburn P, Walton JN: The effect of changes in serum potassium upon myoto-
115. Hart IK, Waters W, Vincent A, Newland C, et al: Autoantibodies directed to ex- nia. J Neurol Neurosurg Psychiatry 1960;23:119-126.
pressed K+ channels are implicated in neuromyotonia. Ann Neurol 1977;41: 153. Li C, Shy G, Wells J: Some properties of mammalian skeletal muscle fibres
238-246. with particular reference to fibrillation potentials. J Physiol 1957;135:522-535.
116. Henneman E, Somjen G, Carpenter DO: Functional significance in cell size in 154. Litchy WJ: A practical demonstration of EMG activity. In AAEM Course C:
spinal motor neurons. J Neurophysiol 1965;28:560-580. Standard needle electromyography of muscles. Rochester Minn, 1988, pp i
117. Henneman E, Olsen CB: Relation between structure and function in the design 23-33. !
of skeletal muscle. J Neurophysiol 1965;28:581-598. 155. Lorente de No R: Analysis of the distribution of action currents of nerve in
118. Herbison GJ: Neuropathic Needle Examination. AAEM Course A: Funda- volume conductors. Stud Rockefeller Inst Med Res 1947;132:384-477.
mentals of EMG. Rochester, Minn, 1984, pp 21-25. 156. Ludin HP: Microelectrode study of normal human skeletal muscle. Eur Neurol
119. Hjorth RJ, Walsh JC, Willison RG: The distribution and frequency of sponta- 1969;2:340-347.
neous fasciculations in motor neuron disease. J Neurol Sci 1973;18:469-474. 157. Ludin HP: Electromyography in Practice. Stuttgart, Georg Thieme Verlag,
120. Hnik P, Skorpil V: The significance of blood circulation for fibrillation activity 1980, pp 48-143.
in denervated muscles of the rat. Physiol Bohem 1959;8:81-86. 158. Maeyens E, Pitner S: Effect of electromyography on CPK and aldolase levels.
121. Hofmann WW, Alston W, Rowe G: A study of individual neuromuscular junc- Arch Neurol 1968;19:538-539.
tions in myotonia. Electroenceph Clin Neurophysiol 1966;215:521-537. 159. Mamoli B, Heiss WD, Maida E, et al: Electrophysiological studies on the "stiff-
122. Honet JE, Honet JC, Cascade P: Pneumothorax after electromyographic elec- man" syndrome. J Neurol 1977;217:111 - 1 2 1 .
trode insertion in the paracervical muscles: Case report and radiographic analy- 160. Manuelidis EE, Angelo JN, Gorgacz EJ, et al: Experimental Creutzfeldt-Jakob
sis. Arch Phys Med Rehabil 1986;67:601-603. disease transmitted via the eye with infected cornea. N Engl J Med 1977;296:
123. Honet JC: Pneumothorax and EMG. Arch Phys Med Rehabil 1988;69:149. 1334-1336.
124. Howard J, McGill KC, Dorfman LJ: Properties of motor unit action potentials 161. Mathews PB: Mammalian Muscle Receptors and Their Central Actions.
recorded with concentric and monopolar needle electrodes: ADEMG analysis. Baltimore, Williams & Wilkins, 1972, pp 1-95.
Muscle Nerve 1988;11:1051-1055. 162. McComas AJ, Fawcett PRW, Campbell MJ, et al: Electrophysiological estima-
125. Isaacs H: Continuous muscle fiber activity. J Neurol Neurosurg Psychiatry tion of the number of motor units within a human muscle. J Neurol Neurosurg .
1961;24:319-325. Psychiatry 1971;34:121-131. |
126. Isaacs H: Continuous muscle fiber activhy in an Indian male with additional ev- 163. McComas AJ: Neuromuscular Function and Disorder. London, Butterworths,
idence of terminal motor fiber abnormality. J Neurol Neurosurg Psychiatry 1977, pp 47-62.
1967;30:126-133. 164. McLaughlin BJ: The fine structure of neurons and synapses in the motor nuclei
127. Jasper H, Ballem G: Unipolar electromyograms of normal and denervated of the cat spinal cord. J Comp Neurol 1972; 144:429^60.
human muscle. J Neurophysiol 1949;12:231-244. 165. Meadows JC: Observation on muscle pain in man with particular reference to
128. Johnson EW: The EMG examination, in Johnson EW (ed): Practical pain during needle electromyography. J Neurol Neurosurg Psychiatry
Electromyography. Baltimore, Williams & Wilkins, 1988, pp 1-21. 1970;33:519-523.
129. Katz B, Miledi R: Propagation of electric activity in motor nerve terminals. 166. Mitsikostas DD, Karandreas N, Coutsopetras P, et al: Fasciculation potentials in
Proc Soc Lond 1965; 161:453^82. healthy people. Muscle Nerve 1998;21:533-535.
130. Katz B: Nerve, Muscle, Synapse. New York, McGraw-Hill, 1966, pp 11-72. 167. Nandedkar SD, Sanders DB: Recording characteristics of monopolar EMG
131. Kelly JJ, Stolov WC: Motor unit potential quantitation: An AAEM workshop. electrodes. Muscle Nerve 1991;14:108-112.
Rochester, Minn, 1984. 168. Newsom-Davis J, Mills KR: Immunological associations of acquired neuromy-
132. Khoshbin S, Hallet M, Lanbeck R: Predictors of patients' experience of pain in otonia (Isaacs' syndrome). Brain 1993:116:453-469.
EMG. Muscle Nerve 1987;10:629-632. 169. Noble D: Applications of Hodgkin-Huxley equations to excitable tissues.
133. Kimura J: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Physiol Rev 1966;46:1-50.
Practice (2nd ed). Philadelphia, FA Davis Co., 1989, pp 270. 170. Nolan CM, Hashisaki PA, Dundas DF: An outbreak of soft-tissue infections due
134. Kingham S, Wilson J, Hart EW: Hereditary stiff-baby syndrome. Am J Dis to myocobacterium fortuitum associated with electromyography. J Infect Dis
Child 1981;135:909-911. 1991;163:1150-1153.
135. Kosarov D, Gydikov A: The influence of volume conduction on the shape of the 171. Norris FH, Chatfield PO: Some electrophysiologic aspects of muscular dystro- !
action potential recorded by various types of needle electrodes in normal human phy. Electroenceph Clin Neurophysiol 1955;7:391-397.
muscle. Electromyogr Clin Neurophysiol 1975;15:319-335. 172. Norris FH, Gasteiger EL, Chatfield PO: An electromyographic study of induced
136. Kraft GH: Fibrillation potential amplitude and muscle atrophy following pe- and spontaneous muscle cramps. Electroenceph Clin Neurophysiol I957;9:
ripheral nerve injury. Muscle Nerve 1990;13:814-821. 139-147.
137. Kraft GH: Fibrillation potentials and positive sharp waves: Are they the same? 173. Nutter P, Collins K: Diffuse positive waves: Case report. Arch Phys Med
Electroenceph Clin Neurophysiol 1991 ;81:163-166. Rehabil 1988;69:295-296.
138. Kraft GH: Are fibrillation potentials and positive sharp waves the same? No. 174. Paintal AS: The influence of diameter of medullated nerve fibers of cats on the
Muscle Nerve 1996;19:216-220. rising and falling phases of the spike and its recovery. J Physiol 1966; 184:
139. Kugelberg E: Activation of human nerves by hyperventilation and hypocal- 791-811.
cemia. J Neurol Neurosurg Psychiatry 1948;60:153-164. 175. Paintal AS: Conduction in mammalian nerve fibers. In Desmedt JE (ed): New
140. Kugelberg E: "Insertion activity" in electromyography. J Neurol Neurosurg Developments in Electromyography and Clinical Neurophysiology. Basel,
Psychiatry 1949;129:268-273. Karger, 1973, pp 19-41.
141. Kugelberg E, Edstrom L, Abruzzese M: Mapping of motor units in experi- 176. Panayiotopoulos CP, Searpalezos S: Electrophysiological estimation of motor
mentally innervated rat muscle. J Neurol Neurosurg Psychiatry 1970.33: units in Duchenne muscular dystrophy. J Neurol Sci 1974;23:89-98.
319-329. 177. Partanen JV, Lang AH: EMG dynamics in polymyositis. A quantitative single
142. Kugelberg E: Properties of the rat hind-limb motor units. In Desmedt JE (ed): motor unit potential study. J Neurol Sci 1974;57:221-234.
New Developments in Electromyography and Clinical Neurophysiology. 178. Partanen JV, Danner R: Fibrillation potentials after muscle injury in humans.
Karger, Basel, 1973, pp 2-13. Muscle Nerve 1982;5:S70-S73.
143. LaJoie WJ: Analgesia in electromyography. Arch Phys Med Rehabil 1963; 179. Partanen JV, Nousiainen U: End-plate spikes in electromyography are fusimo-
44:42-44. tor unit potentials. Neurology 1983;33:1039-1043.
144. Lambert EH, Sayer GP, Eaton LM: Electrical activity of muscle in polymyosi- 180. Pease WS, Bowyer BL: Motor unit analysis: Comparison between concentric
tis. Trans Am Neurol Assoc 1954;79:64-69. and monopolar electrodes. Am J Phys Med Rehabil 1988;67:2-6.
181. Pedinoff S, Sandhu RS: Electromyographic effects on serum creatine phospho- 212. Stalberg E, Andreassen S, Falck B, et al: Quantitative analysis of individual
kinase in normal individuals. Arch Phys Med Rehabil 1978;59:27-29. motor unit potentials: A proposition for standardized terminology and criteria
182. Petajan JH: Frequency control of motor unit action potentials. Electroenceph for measurement. J Clin Neurophsyiol 1986;3:313-348.
Clin Neurophysiol 1969;27:66-72. 213. Stegeman DF, de Weerd JPC: Modelling compound action potentials of periph-
183. Petajan JH: Clinical electromyographic studies of diseases of the motor unit. eral nerves in situ I: Model description: Evidence for nonlinear relation between
Electroenceph Clin Neurophysiol 1974;36:395-401. fibre diameter and velocity. Electroenceph Clin Neurophysiol 1982;54:436-448.
184. Petajan JH: Motor unit frequency control in normal man. In Desmedt JE (ed): 214. Stein JM, Padykula HA: Histochemical classification of individual skeletal
Progress in Clinical Neurophysiology Vol 9. Basel, Karger, 1981, pp 184-200. muscle fibers of the rat. Am J Anat 1962;110:103-123.
185. Peter JB, Barnard RJ, Edgerton VR, et al: Metabolic profiles of three fiber types 215. Steiss JE, Forsyth G: Effect of electromyography on serum creatine kinase
of skeletal mammalian muscle in guinea pigs and rabbits. Biochem values in clinically normal dogs and horses. Am J Vet Res 1984;45:1199-1200.
1972;11:2627-2633. 216. Stoehr M: Low frequency bizarre discharges. Electromyogr Clin Neurophysiol
186. Pette D, Vrbova G: Natural control of phenotypic expression in mammalian 1978;18:147-156.
muscle fibers. Muscle Nerve 1985;8:676-689. 217. Stohr M: Benign fibrillation potentials in normal muscle and their correlation
187. Purves D, Sakmann B: The effect of contractile activity on fibrillation and ex- with endplate and denervation potentials. J Neurol Neurosurg Psychiatry
trajunctional acetylcholine sensitivity in rat muscle maintained in organ culture. 1977;40:765-768.
J Physiol 1974;237:157-182. 218. Stolov WC: AAEM Minimonograph #16: Instrumentation and measurement in
188. Purves D, Sakmann B: Membrane properties underlying spontaneous activity electrodiagnosis. Rochester, MN, American Association of Electrodiagnostic
of denervated muscle fibers. J Physiol 1974;239:125-153. Medicine, 1981.
189. Rasminski M: Ectopic generation of impulses and cross-talk in spinal nerve 219. Streib EW: AAEM Minimonograph #27: Differential diagnosis of myotonic
roots of dystrophic mice. Ann Neurol 1978;3:351-357. syndromes. Muscle Nerve 1987;10:603-615.
190. Reed DM, Kurland LT: Muscle fasciculation in a healthv population. Arch 220. Swash M, Schwartz MS: Neuromuscular Diseases: A Practical Approach to Di-
Neurol 1963;9:363-367. agnosis and Management. New York, Springer- Verlag, 1981, pp 57-65.
191. Reinstein L, Twardzik FG, Mech KF: Pneumothorax: A complication of needle 221. Thesleff S: Physiological effects of denervation of muscle. Ann New York Acad
electromyography of the supraspinatus muscle. Arch Phys Med Rehabil Sci 1974;228:89-103.
1987;68:561-562. 222. Thesleff S, Ward MR: Studies on the mechanism of fibrillation potentials in
192. Rexed JB: The cytoarchitectonic organization of the spinal cord in the cat. J denervated muscle. J Physiol 1975;244:313-323.
Comp Neurol 1952;96:415-496. 223. Thesleff S: Fibrillation in denervated mammalian muscle. In Culp WJ, Ochoa J
193. Ricker K, Hertel G, Langscheid K, Stodieck G: Myotonia not aggravated by (eds): Abnormal Nerve and Muscle as Impulse Generators. New York, Oxford
cooling. J Neurol 1977;2T6:9-20. University Press, 1982, pp 678-694.
194. Rosenfalek P: Intra- and extracellular potential fields of active nerve and muscle 224. Thiele B, Bohle A: Number of spike components contributing to the motor unit
fibers. Acta Physiol Scand 1969;(Suppl 321): 1-168. potential. Z EEG-EMG 1978;9:125-130.
195. Roth G: The origin of fasciculations. Ann Neurol 1982;12:542-547. 225. Thompson W, Jansen JKS: The extent of sprouting of remaining motor units in
196. Rowinska-Marcinska K, Ryniewicz B, Hausmanowa-Petrusewicz I, Karwanska partially denervated immature and adult soleus muscle. Neuroscience
A: Diagnostic value of satellite potentials in clinical EMG. Electromyogr Clin 1977;2:523-535.
Neurophysiol 1997;37:483-489. 226. Torre M: Nombre et dimensions des unites matrices dans les muscles extrinse-
197. Sander JE, Layzer RB, Goldsobel AB: Congenital stiff-man syndrome. Ann ques de 1'ocil et en general dans les muscles squelettiques relies a des organes
Neurol 1980;8:195-197. des sens. Arch Suisses Neurol et Psychiat 1953;72:362-376.
198. Satoyoshi E: A syndrome of progressive muscle spasm, alopecia, and diarrhea. 227. Trojaborg W, Buchthal F: Malignant and benign fasciculations. Acta Neurol
Neurology 1978;28:458-471. Scand 1965;41(Suppl 13):251-254.
199. Schwartz MS, Stalberg E, Schiller HH, et al: The reinnervated motor unit in 228. Trontelj J, Stalberg EV: Bizarre repetitive discharges recorded with single fiber
man. J Neurol Sci 1976;27:303-312. EMG. J Neurol Neurosurg Psychiatry 1983;46:310-316.
200. Sellin LC, Libelius R, Lundquist I, et al: Membrane and biochemical alterations 229. Walpin LA, Reiss F: Urticaria changing lo vitiligo provoked by electromyo-
after denervation and during reinnervation of mouse skeletal muscle. Acta graphic examination. Acta Derm Venereol 1966;46:242-244.
Physiol Scand 1980,110:181 -186. 230. Warszaueski M, Telerman-Tappet N, Durdu J, et al: The early stages of neuro-
201. Sherman HB, Walker FO, Donofrio PD: Sensitivity for detecting fibrillation po- muscular regeneration after crushing the sciatic nerve in the ra*t—electrophysio-
tentials: A comparison between concentric and monopolar needle electrodes. logical and histochemical study. J Neurol Sci 1975;24:21-32.
Muscle Nerve 1990;13:1023-1026. 231. Weddell G, Feinstein B, Pattle RE: Clinical application of electromyography.
202. Sherrington CS: Some functional problems attaching to convergence. Proc R Lancet 1943;1:236-239.
Soc Lond 1929;105:332-362. 232. Wiechers DO: Mechanically provoked insertional activity before and after
203. Shillito P, Molenaar PC, Vincent A, Leys, K, et al: Acquired neuromyotonia: nerve section in rats. Arch Phys Med Rehabil 1977;58:402-^05.
Evidence for autoantibodies directed against K+ channels of peripheral nerves. 233. Wiechers DO, Stow R, Johnson EW: Electromyographic insertional activity
Ann Neurol 1995;38:714-722. mechanically provoked in the biceps brachii. Arch Phys Med Rehabil
204. Slater CR, Harris JB: The anatomy and physiology of the motor unit. In Walton 1977;58:573-578.
JW (ed): Disorders of Voluntary Muscle (3rd ed). London, Churchill 234. Wiechers DO: Electromyographic insertional activity in normal limb muscles.
Livingstone, pp 1-26. Arch Phys Med Rehabil 1979;60:359-363.
205. Smith JW, Thesleff S: Spontaneous activity in denervated mouse diaphragm 235. Wiechers DO, Johnson EW: Diffuse abnormal electromyographic insertional
muscle. J Physiol 1976;257:171-186. activity. A preliminary report. Arch Phys Med Rehabil 1979;60:419-422.
206. Spence WR, Guyton JD: Control of pain during electromyography. Arch Phys 236. Wiederholt WC: "End-plate noise" in electromyography. Neurology 1970;
Med Rehabil 1966;47:771-774. 20:214-224.
207. Stalberg E, Schwartz MS, Thiele B, et al: The normal motor unit in man. J 237. Wilboum AJ: An unreported, distinctive type of increased insertional activity.
Neurol Sci 1976;27:291-301. Muscle Nerve; 1982:S 101 -S105.
208. Stalberg EV, Trontelj J: Single Fibre Electromyography. United Kingdom, 238. Wohlfart G: Collateral regeneration in partially denervated muscles. Neurology
Mirvalle Press, 1979. 1958;8:175-180.
209. Stalberg EV, Antoni L: Electrophysiological cross section of the motor unit. J 239. Wray SH: Innervation ratios for large and small limb muscles in the baboon. J
Neurol Neurosurg Psychiatry 1980;43:469-474. Comp Neurol 1969; 137:227-250.
210. Stalberg EV, Trontelj JV: Abnormal discharges generated within the motor unit 240. Wright KC, Ramsey-Goldman R. Nielson VK. et al: Syndrome of diffuse ab-
as observed with single-fiber electromyography. In Culp WJ, Ochoa J (eds): normal insertional activity: Case report and family study. Arch Phys Med
Abnormal Nerves and Muscles as Impulse Generators. New York, Oxford Rehabil 1988;69:534-536.
University Press, 1982. pp 443-474. 241. Wuerker RB, McPhedran AM, Henneman E: Properties of motor units in a het-
211. Stalberg EV, Chu J, Bril V, et al: Automatic analysis of the EMG interference erogeneous pale muscle (m. gastrocnemius) of the cat. J Neurophysiol
pattern. Electroenceph Clin Neurophysiol 1983:56:52-62. 1965;28:85-99.
\
Chapter 8
Quantitative EMG
Sanjeev D. Nandedkar, Ph.D.
Erik V. Stalberg, M.D., Ph.D.
Donald B. Sanders, M.D.
CHAPTER OUTLINE
The Motor Unit Special Recording Techniques

Routine Needle EMG Examination • Electrophysiology Single-Fiber EMG • Fiber Density * Jitter * Relevance t o
of a Motor Unit • Quantitative Analysis • Manual Routine EMG Studies • Surface EMG and MU Number
Measurements • MU Architecture and MUAP Features Estimation • Recording Characteristics of Surface EMG
• Concentric vs. Monopolar Needle Electrode Electrodes • MUNE Technique • O t h e r Techniques • MUNE
• Interference Pattern • Interference Pattern Analysis at and Disease Progression • Macro EMG
Maximal and Minimal Effort • Power Spectrum Analysis
• The Turns and Amplitude (TA) Method * Expert Putting it Together: Quantitative Analysis and a Model
Quantitative IP (EQIP) Analysis for Disease Processes
• Decomposition Scanning EMG • O t h e r Quantitative Methods
The needle electromyographic (EMG) examination is a EMG examination to better identify and document signal abnor-
powerful procedure in the diagnosis of nerve and muscle dis- malities. The integration of QA and the routine examination
eases. Signals are recorded using a concentric needle or leads to "objective EMG." This approach helps not only to
139
monopolar needle electrode. The intramuscularly recorded detect abnormalities, but in some instances to assess their sever-
EMG waveform is assessed subjectively from its appearance on ity, status (active versus inactive), duration (acute versus
the display screen and its corresponding sound on an audio chronic) and perhaps prognosis. By practicing QA techniques,
monitor. In quantitative analysis (QA), a statistically valid one may train the eyes and ears to recognize subtle abnormalities
sample of EMG signals is quantified manually or using auto- that may otherwise be missed. Obviously, one cannot practice
mated computer-based methods. One can also use specialized every available QA technique. However, a basic understanding
electrodes to facilitate QA. Individual and mean waveform pa- of these methods will allow the practitioner to better decide
rameter values are compared to appropriate reference values to when to order or perform a QA procedure and how to interpret
identify the underlying disease processes. When EMG abnor- the results.
malities are obvious on the routine needle evaluation, QA may Finally, many QA techniques have recently become user-
not add much to the electrodiagnostic medicine evaluation. In friendly and more automated. Additionally, the cost of instru-
fact, QA is still performed infrequently in most clinical EMG ments that support such analyses has decreased dramatically.
settings. As a result, many electromyographers may consider The statistical treatment of data has also changed and become
QA as a specialty within the practice of EMG having little or no automatic for many applications. As a result, QA in one muscle
role in day-to-day clinical practice. In this chapter our goal i§ to may now require just a few minutes. This time saving gives the
dispel this "myth." practitioner an incentive to master these various techniques and
It is important to recognize that QA provides a foundation for implement them in routine EMG studies.
the routine needle EMG examination. All the criteria used in We begin by a brief description of the current concept of
signal assessment today are derived from QA studies per- motor unit pathology, the routine EMG examination, and a
formed, in some cases, up to 5 decades ago. In our experience, generic description of quantitative analysis. Next we review the
QA is particularly useful when the abnormalities on routine techniques of EMG quantification using the standard electrodes,
EMG are equivocal. QA techniques are essential when EMG is i.e., CN or MN. This is limited to analysis of MUAPs and the IP
used serially to assess disease progression. The knowledge and signals. Subsequently we review techniques using special elec-
techniques of QA can be easily incorporated into the subjective trodes such as single-fiber and macro EMG. We also review
293
some techniques that estimate the number of MUs in the muscle

based on surface EMG recordings. At each step we propose
strategies to integrate information from QA into the routine
EMG examination.
THE MOTOR UNIT

A motor unit (MU) consists of all muscle fibers innervated
by one motor neuron. The MU architecture refers to the size,
distribution, end-plate area, etc. of the MU. It varies consider-
ably among different normal muscles and is altered by various
disease processes.
The MU size refers to the number and size of muscle fibers
comprising the M U . Small muscles involved in fine motor
6 8 1 6 0
functions such as those of the larynx and face have smaller MUs
compared to large muscles that participate in forceful activity,
e.g., biceps brachii and tibialis anterior. In large muscles, a MU
Figure 8-2. R e i n n e r v a t i o n . A, T w o normal MUs a r e shown
may contain between 50 and 200 muscle fibers, which are dis-
schematically.The endplates are arranged in the middle of each fiber
tributed within a territory of 3 - 1 0 mm in diameter (Fig. 8-
and form an endplate zone.Their territories overlap partially. B, MU
1 ) i7.26.i79.i88 L
A MUs also have a large territory (Fig. 8-1B).
a r g e
# 2 is lost. Some of its fibers are reinnervated by MU # 1 . N o t e the in-
Muscle fibers are classified into three different types based on
creased n u m b e r of muscle fibers in MU #1 with the first signs of
their metabolic properties:, type I, type IIA, and type IIB. On
muscle fiber grouping.The territory has n o t changed.The fibers indi-
muscle biopsy studies, these fibers are identified using appropri-
cated by dotted lines are still denervated and atrophied.
ate stains. It is important to recognize that all fibers in one MU
51
are of the same type. A muscle contains MUs of different types.

The muscle fibers belonging to one MU are dispersed among muscles have MUAP features similar to those recorded from the
fibers of other M U s . Two or more muscle fibers belonging to
57
biceps brachii muscle of patients with a myopathy, i.e., small
the same MU are not usually in close proximity to each other. amplitude, short duration, and possibly a polyphasic character.
In the biceps brachii, the mean muscle fiber has a diameter of The disease processes in myopathy result in loss of muscle
approximately 50-60 pm. The endplates are located roughly at
51
fibers, change in muscle fiber size (i.e., atrophy and hypertro-
the center of the muscle within a 5-10-mm wide zone (Fig. 8- phy), regeneration of muscle fibers from satellite cells, split-
2A). The endplate zone in other muscles may be spread over a ting of fibers, and increased interstitial tissue. Some
much wider area. It is important to recognize that architectural
4
reinnervation can also occur (Fig. 8-1C). An MU may lose,
176
differences in normal muscles generate different characteristics most, perhaps all, of its fibers in severely affected muscles, re-
on EMG recordings. For example, recordings from normal facial sulting in fibrosis. Due to the paucity of fibers in such units, we
may not detect them on needle EMG.
The disease processes in neuropathy result in loss of motor
units and subsequent denervation and reinnervation. In Figure
8-2A two MUs are shown schematically. The fibers within each
MU show no tendency to form groups. In figure 8-2B, MU #2 is
lost due to disease. Some muscle fibers of this MU are close to
the muscle fibers of the surviving MU # 1 , which generates col-
lateral sprouts and provides innervation to the denervated fibers.
The reinnervated muscle fibers change their physiologic proper-
ties to match their new MU. Reinnervation occurs mainly within
the overlapping areas of the two MUs. The surviving MU in-
creases its size by accepting more and more fibers, and its terri-
tory contains clusters of muscle fibers. On muscle biopsy, this is
seen as fiber type grouping. The fibers of the denervated MU
are thus absorbed by the surviving MUs. Some fibers may not
receive reinnervation. They will atrophy and can be seen as small
angulated fibers on the biopsy. The loss of MUs is also com-
pensated by hypertrophy of the surviving muscle fibers. All of
MU in Myopathy M U a f t e r
Reinnervation
these disease processes will affect the MU architecture and
hence the MUAP waveform. These pathophysiologic processes
Figure 8-1. T h e M U a r c h i t e c t u r e . A, A small MU shows a small form the basis for the MUAP changes and their analysis in the
MU territory. B, A large MU has more fibers distributed over a larger routine EMG examination.
area. C, The architecture of the MU in A is changed to illustrate differ-
ent processes that occur in myopathy. N o t e t h e loss of muscle fibers ROUTINE NEEDLE EMG EXAMINATION
in some areas of t h e MU territory. D, T h e size of MU in A increased
due to reinnervation. However, it has a smaller territory compared to The routine EMG examination is performed using a concen-
a normal large MU. tric (CN) or a monopolar needle (MN) electrode. The CN consists
Chapter 8 QUANTITATIVE EMG — 2 9 5
recorded. At present, there is no method to automatically recog-

nize and quantify the potentials elicited during this phase of the
o examination.
In the third step, the patient minimally activates the examined
muscle to recruit a few MUs. The EMG records the electrical
representation of the MU, which is referred to as the motor unit
potential (MUP) or the motor unit action potential (MUAP).
The electrode position is adjusted to record MUAPs that sound
sharp and crisp on the audio monitor. The amplitude, duration,
shape, and discharge rate of the MUAP are assessed.
In the fourth step, the patient is asked to gradually increase
the force of muscle contraction to maximal effort. This recruits
a large number of MUs in the tested muscle. The number and
firing rates of M U s during recruitment are investigated. The
EMG signal at high force levels, called the interference pat-
E 38 tern (IP), is assessed from its amplitude, spike density, fre-
quency content, sound, etc.
Figure 8-3. E M G e l e c t r o d e s . A schematic view of different types The above assessment is performed at different sites within
of EMG electrodes is shown. (A) Single-fiber, (B) Multielectrode, (C) the tested muscle using one or a few different insertion points.
concentric, (D) monopolar, and (E) macro-EMG. The needle electrode is directed along different corridors within
the muscle to explore several locations at each insertion site.
Based on this assessment, the findings may be described as
of a 150-jim metal wire inserted into a hollow metal cylinder "normal," "myopathic," "neuropathic," or interpreted in other
called the cannula (Fig. 8-3C) . The tip of this assembly is
f
ways. The recording is then followed by the assessment of the
ground to a 1 5 angle exposing an elliptical (150 x 580 jim)
c
data collected. In this portion of the evaluation, the need to per-
recording surface on the central wire, called the core. The core form tests of additional muscles is considered as well as what
has a surface area of 0.07 m m and serves as the active record-
2
further studies should be done to formulate a conclusion from
ing surface, whereas the cannula serves as the reference elec- the investigation.
trode in CN EMG recordings. Electrodes with a smaller core
surface area are also commercially available and are frequently ELECTROPHYSIOLOGY OF A MOTOR UNIT
referred to as "facial" needles.
The MN is made from an insulated metal wire (Fig. 8-3D). It A motor unit (MU) consists of all the muscle fibers inner-
is ground to expose a cone-shaped recording surface. This sur- vated by one motor neuron (Fig. 8-4A-B). When the motor
face has an area of 0.2 m m and serves as the active recording
2
neuron discharges, an action potential (AP) propagates along
electrode. A surface electrode is used as the reference. The dif- the nerve fiber and terminal nerve branches, and arrives at the
ference in the size and construction of these electrodes and the neuromuscular junction (also called the endplate). A grouping
reference setup gives them slightly different recording charac-
teristics. These differences, demonstrated by QA, must be rec-
ognized in the routine needle EMG examination. For both
electrodes, a separate ground electrode is positioned near the in-
sertion site.
By way of r e v i e w , -
41
the routine needle EMG examina-
52100
tion may be divided into four stages. First, the electrode is

moved slightly from one position to the other in the tested
muscle. The mechanical irritation provokes a volley of poten-
tials called the insertional activity (IA), which lasts a few hun-
dred milliseconds. The duration of IA depends upon the vigor
behind the needle movement. A gentle slow movement may not
elicit any activity. In pathology, however, needle insertion may
trigger electrical activity in simple or complex configurations
that may persist for several seconds.
In the second stage, the electrode is held fixed in each of a
few (up to ten) places and the patient attempts to completely
relax the muscle. In this condition, spontaneously generated po-
Motor Unit Action Potential (MUAP)
tentials are observed. Examples of spontaneous activity (SA) Extracellular Muscle Fiber Action Potential
are endplate noise, endplate spikes, fibrillation potentials,
positive sharp waves, complex repetitive discharges, etc. Figure 8-4. M U a n d M U A P . A schematic m o t o r unit with five
These potentials are readily recognized by their morphology as muscle fibers shown in longitudinal (A) and cross-sectional (B) view.
well as their sound and firing rates. In normal muscles, SA is
12
N o t e t h e different lengths of the terminal branches and the scatter of
seen rarely. Based on their frequency of occurrence at different the endplates along the muscle length.The individual muscle fiber p o -
sites, the IA and SA are graded subjectively on a numeric scale, tentials in C are summated t o produce t h e MUAP shown in D . T h e
e.g., 1+, 2+, etc. One way to measure them is to record the dark tip of the monopolar electrode is the active recording surface.
number of sites out of 10 in which abnormal IA and SA are The reference electrode placed remotely is n o t shown.
of endplates within a localized region of a muscle is referred to nerve APs from different fibers arrive at the presynaptic termi-
as the endplate zone. At the arrival of a nerve AP, the nerve ternal at different times. The time of neuromuscular transmis-
minals release acetylcholine (ACh), which diffuses across the sion is quite small, but may vary among different fibers and
synaptic cleft and binds to the receptors on the postsynaptic changes from one discharge to the next in the same fiber. (Of
membrane. This depolarizes the membrane, producing the end- note, this variability forms the basis for the jitter measured by
plate potential (EPP). When the E P P exceeds the muscle single-fiber EMG.) Once a muscle fiber AP is generated, its
membrane threshold level, it produces a muscle fiber AP, which propagation time to the electrode will vary due to differences in
propagates in both directions from the endplate towards the ten- the propagation velocity. Large fibers conduct APs faster than
dons. The AP initiates a chain of events that leads to muscle smaller fibers. Finally, the endplates of different muscle fibers
fiber contraction. Thus, every time the motor neuron discharges, are scattered within a region of the muscle. Hence, action po-
the muscle fibers in the MU respond by producing an electrical tentials from different muscle fibers must travel different dis-
and a mechanical output. The MUAP is the summated electrical tances to reach the recording electrode (Fig. 8-4A).
activity of an MU recorded by an EMG electrode (Fig. 8-4D). A needle EMG electrode records the summation of action po-
The generation of the MUAP is shown schematically in Figure tentials from all muscle fibers in the MU, i.e., the MUAP (Fig.
8-4C, which demonstrates three important characteristics of the 8-4D). The MUAP waveform depends not only on the position
muscle fiber potential within the MU. of the electrode in relation to the muscle fibers as described
The shape of the extracellularly recorded muscle fiber AP is above but also upon the MU architecture, i.e., the number,
biphasic or triphasic, which is quite different from the monopha- size, muscle fiber distribution, and endplate distribution.
sic intracellular AP recording. In extracellular recordings (Fig.
2
8-4C), a downward (i.e., positive) deflection is recorded when QUANTITATIVE ANALYSIS

the intracellular AP leaves the endplate and approaches the elec-
trode. When the intracellular AP passes by the electrode, the Quantitative analysis may be divided into three general steps.
main positive-to-negative going spike of the extracellular AP is First the signals are acquired using standardized conditions.
generated. When the intracellular AP propagates away from the These include the size and placement of electrodes, filter set-
electrode, the extracellular potential again deviates in the posi- tings, and the method of signal acquisition and processing. In
tive direction and finally returns to the baseline, thus a triphasic the second step, one makes measurements of the recorded po-
waveform is recorded. Clinical EMG recordings are made in the tential. This may be done manually or by using computer-based
extracellular space. In this chapter the term AP implies an extra- algorithms. Manual measurements may require standardized
cellular recording unless specified otherwise. display settings such as the display gain. The computer-based
When the EMG electrode is in the endplate area, the initial measurements may use criteria that are not always known to the
positive deflection signifying AP propagation from endplate to user. Finally, one performs analysis of measurements by com-
the electrode does not occur. The AP has an initial negative paring them with appropriate reference values.
(upward) deflection. An initial negative deflection is routinely We prefer to use the term reference limits instead of normal
observed in the compound muscle action potential (CMAP) limits for two reasons. First, the subjects recruited for develop-
recordings made when the surface active electrode is placed ing the limits are not tested extensively to establish the lack of
over the endplate area of the tested muscle. any nerve or muscle disease. When one begins recordings, it is
The peak-to-peak amplitudes of APs vary considerably not uncommon to occasionally find "abnormal" EMG charac-
among different fibers of the MU. This occurs due to different teristics by subjective assessments. This presents a dilemma for
distances between the active recording surface of the electrode accepting or rejecting this normal subject for the study. Second,
and the fibers of the MU (Fig. 8-4B). The electrode records a the reference values are often defined as including 9 0 - 9 5 % of
large-amplitude AP from a fiber that is close to the recording all data recorded from the normal population. There is thus
surface (e.g., fiber #4 in Fig. 8-4). These APs are rich in high always a slight chance that a normal subject would have EMG
frequencies and have short rise times. This gives them a high- measurements outside the normal limits. Over the last five
pitched sound, often described as sharp or crisp, on the audio decades, two distinct strategies have been developed to assess
monitor. The AP of a distant muscle fiber has low amplitude QA measurements. In the first, 20 MUAPs are recorded from
(e.g., fiber #1 in Fig. 8-4). It also has a long rise time and mini- different sites. The mean values of these measurements are used
mal high-frequency components, which gives APs from distant to assess abnormalities. This mean value represents the overall
fibers a "dull" sound on the audio monitor. The decline of the sampling of the tested muscle. In serial EMG studies, this
potential amplitude with radial distance depends upon the size allows one to assess changes in the MU due to disease progres-
and shape of the recording electrode. In needle EMG, one may sion. However, it can be time-consuming to acquire the 20 or
use electrodes with small (single-fiber EMG), medium (concen- more potentials required to compute the mean values. Using
tric or monopolar), or large (macro EMG) recording surfaces to fewer measurements may give a false high or low value of the
obtain different rates of amplitude decline, i.e., different selec- mean. 61
tivity. These electrodes give complementary information about Analysis based on mean values is different from subjective
the MU, as described later in this chapter. assessment of EMG signals. In the routine EMG examination,
The peaks of individual muscle fiber APs of the MU occur at one can easily recognize potentials that are "obviously abnor-
different times, indicating that they are not synchronous when mal," e.g., a giant CNEMG MUAP or blocking on single-fiber
they arrive at the recording electrode (Fig. 8-4C). The differ- EMG recordings. This lead Stalberg and coworkers to develop
ence in their arrival time, called the temporal dispersion, re- the outlier approach of assessment. - Instead of defining ref-
190 191
sults from several physiologic factors. erence limits based on mean values, they defined the limits for
The nerve AP propagates through terminal branches of different individual potentials. A potential with measurements outside the
lengths (Fig. 8-4A). Furthermore, the AP propagation velocity reference limits is called an outlier. The limits are defined such
may vary among different terminal branches. As a result, the that all control subjects have fewer than 10% outlier potentials.
When one aims to record up to 20 different potentials, a study The sweep setting is indicated in two different ways: sweep
would be abnormal when the third abnormal potential was duration or sweep speed. The two descriptions are related as
recorded. This could very well occur within the first few record- follows:
ings. At this stage the quantitative procedure has reached diag-
nostic significance and the test may be terminated. In this Sweep speed (ms/div) = Sweep duration (ms)
Number of horizontal divisions
fashion, the time required for the test can be significantly re-
duced. It is important to recognize that the reference limits for In Figure 8-5, the sweep duration is 100 ms. The 100 m s
the outlier approach are much wider than those used for mean sweep is divided into 10 horizontal divisions. This gives a
values. If one intends to follow the disease progression, how- sweep speed of 10 ms/div. The duration of potential Dl in the
ever, 20 or more potentials must be acquired for analysis. top trace is roughly 1 division. Multiplying by the sweep speed
Both methods of analysis are used in quantitative analysis. It gives a value of 10 ms for the MUAP duration. The time inter-
is important to recognize that reference values are affected by val between potentials D2 and D3 is measured as follows: After
technical factors (e.g., type of electrode, filter settings, tempera- potential D2 on the third trace there are 9 divisions to the end of
ture), demographic factors (e.g., age, gender), as well as the trace. The signal continues on the fourth trace and the potential
nerve or muscle tested. One must match all of these factors D3 occurs after 1 division from left. Thus, potentials D2 and D3
before using published reference values. are separated by 10 divisions, i.e., 100 ms. When measured
more precisely using computer-aided cursors the interval was
Manual Measurements 105 ms; however, visual assessment is acceptably close to this
Quantitative analysis does not necessarily require automated value. Similarly, potentials D l and D2 are separated by 12 divi-
measurements. In fact, much of the earlier work on MUAP sions (1 on the first sweep, 10 on the second and 1 on the third)
analysis and single-fiber EMG was performed by making or 120 ms.
manual measurements of EMG signal printouts. Although
modern instruments provide algorithms for automated analysis, MUAP Analysis
the automated measurements may not be accurate when the Measurements. Buchthal and his colleagues pioneered the
signal contains too much noise, interference, and artifacts. technique of MUAP analysis in the early 1950s by measuring
Hence one should be able to make visual assessments of time amplitude, duration, and phases (Fig. S-6). The MUAP ampli-
25
and amplitude measurements. In this chapter, the EMG signals tude is measured between the maximum negative and maxi-
are shown in a raster fashion (Fig. 8-5). The signal display mum positive peaks. The MUAP onset occurs when the signal
begins with the top trace. The bottom trace is the latest-occur- first deviates from the signal baseline as seen at a 100 uV/div
ring EMG signal. The five sweeps represent a 500-ms epoch of display setting. The MUAP end occurs when the waveform re-
EMG for analysis and review. A grid superimposed on the dis- turns and remains at the baseline. A phase represents a change
play divides the trace area in several divisions to facilitate sub- in signal polarity about the baseline. A simple way of measuring
jective measurements. The display gain or sensitivity indicates
the amplitude change per vertical division. In Figure 8-5, the
peak-to-peak deflection of the potential on the top trace is about Amplitude
2 divisions. Multiplying by the gain of 100 uV/div, the MUAP
100 pV
amplitude is estimated to be 200 uV.
5 ms
D1
\<—>\ Spike Duration
0 - 100 ms
H MUAP Duration
100-200 ms
D2
200 - 300 ms Linked Potentials
B
300 - 400 ms
»D4
| MUAP Duration
400 - 500 ms 10 ms 100 MV
h *"| Spike Duration
100 ms
Figure 8-6. MUAP m e a s u r e m e n t s . Different measurements of
MUAP are illustrated using t w o different recorded MUAPs. In A, turns
Figure 8-5. S u b j e c t i v e E M G m e a s u r e m e n t s . Five consecutive are indicated byT. N o t all turns are identified. A questionable phase is
sweeps of the EMG signal are shown.The time of each sweep is indi- indicated by #.An asterisk indicates a peak that does not qualify as a
cated on the left. Four discharges of a single MUAP are labeled D I , D2, turn. In B a polyphasic MUAP is seen, with many late components.The
D3, and D4. See text for details. (slow wave) duration is s h o r t e r that the spike duration.
phases is to count the baseline crossings in the waveform and fiber. A short rise time is therefore an indication that the record-
add one. Usually phase assessment is quite simple, but signals ing electrode is placed inside the MU. In signal analysis, engi-
may deviate minimally from the baseline after a crossing. This neers usually measure the rise time as the time from the baseline
is illustrated by '#' in Figure 8-6. In that MUAP one may count to the peak value of the signal. To exclude noise from these
5 phases by accepting the baseline crossing or ignore the small measurements, rise time may be defined as the time to reach
deflection of the potential and consider it to be triphasic. A 90-95% of the peak amplitude. In MUAP analysis, it is often rec-
polyphasic MUAP has more than four phases. Conversely, a ommended that measured signals should have a short rise time.
simple MUAP has four or fewer phases. This is the most However, the definition of rise time is rather vague. It is usually
common form of description for the MUAP waveform. measured from the maximum positive to maximum negative
In computerized systems, the aforementioned measurements peak. This may not pose a problem for simple triphasic MUAPs;
can be made automatically. Algorithms in these systems detect however, for recordings made with the electrode in the endplate
the MUAP onset and end based on the slope of the signal and/or area, the signal may not have an initial positive peak. In a ser-
its amplitude deviation from the baseline. To exclude small fluc- rated or polyphasic waveform, the maximum positive and maxi-
tuations about the baseline that result from noise, a minimum mum negative peaks may not occur successively. This can give
amplitude criterion should also be used in counting phases. a long rise time value (Fig. 8-6A). To overcome this problem,
,0
Computers allow us to make additional measurements of the automated methods use the rate of change of voltage in a short
waveform. A turn occurs at a peak of the MUAP. To ex-
184185
segment of the MUAP to define this parameter. The rate is high
clude peaks generated by noise inherent to the recording, the when a MUAP has a short rise time and/or high amplitude.
signal must change by some threshold amplitude between suc- Normally the MUAP waveform remains constant during suc-
cessive turns. In Figure 8-6, the small peak in the rising edge of cessive MU discharges. Such MUAPs are called stable (Fig. 8-
the waveform (indicated by the asterisk) does not fulfill this cri- 7A). In contrast, an unstable MUAP changes waveform from
terion, hence it does not qualify as a turn. A MUAP is called one firing to another. This variability is quantified by a feature
serrated when it has more than five turns. A MUAP is called called jiggle, described in the single-fiber EMG section. The
complex when it is polyphasic, serrated, or both. Other mea- 198
measurement of MU firing rate is discussed with the analysis
sures to describe the irregularity of MUAP shape are also de- of interference pattern.
scribed. 221
Techniques. Several techniques are now available to extract
The area of the MUAP is measured between the rectified and quantify the MUAP waveform. We will focus on three dif-
waveform and the signal baseline. The ratio area/amplitude ferent approaches that represent the evolution of these tech-
quantifies the subjective assessment of the MUAP thickness. 131
niques. The first two methods are available on almost all
By combining the amplitude and ratio, a size index can be commercially available modern instruments. Regardless of the
computed. 172
extraction technique, the needle electrode is positioned to
In some recordings, the MUAP waveform may be divided record sharp and crisp-sounding EMG activity. This will also
into two or more portions that are separated by flat baseline increase the MUAP amplitude and reduce its rise time.
(Fig. 8-6B). In general, the longer duration portion (by visual In the manual method of analysis, the EMG is recorded at-
assessment) is used for MUAP duration assessment, whereas minimal force of muscle contraction. Ideally, the signal should
the other portions are called satellite or linked potentials. A contain discharges of a single MUAP. An epoch of this signal is
satellite is separated from the main MUAP by baseline. If satel- frozen on the instrument display for subjective assessment (Fig.
lite potentials are included in measurements of MUAP duration, 8-5). When the signal contains discharges of more than one
these may have very large values, which must be interpreted MUAP, manual analysis may become difficult. In Figure 8-8A,
with caution. The spike duration is often measured between discharges of three MUAPs are identified by visual assessment
the first and last peaks of the MUAP, including any satellite or and appropriately labeled. MUAP #3 may not be used for analy-
linked potentials. A spike duration that is longer than the MUAP sis due to very low amplitude. Activity from distant MUs results
duration indicates the presence of late spike components.
The rise time of the spike component of the MUAP reflects
the distance from the electrode to the nearest component muscle
^ Jl mV
2 ms
Figure 8-8. M a n u a l M U A P analysis. Monopolar needle EMG signals

Figure 8-7. M U A P w a v e f o r m c h a r a c t e r i z a t i o n . T h r e e consecu- were recorded in the biceps brachii muscle of a normal subject. In A, dis-
tive discharges of stable (A) and unstable (B) MUAPs (jiggle and block- charges of MUAPs from 3 MUs are identified and labeled. In the continu-
ing) are shown. ation (B) superimpositions of I and 2 are seen in traces 1,2, and 5.
in small amplitude fluctuations on the baseline, making it diffi- display of the MUAP on the display screen (Fig. 8-9D). One can
cult to assess the MUAP duration. Signals in B are a continua- visually assess these time-locked discharges to assess MUAP
tion of the same recording but contain complex waveforms of features. This approach is often used to assess MUAPs during
long duration that are very different from the MUAPs in A. the routine needle EMG examination.
Unlike the waveforms in A, the complex waveforms in B do not For quantitative analysis, the trigger level is adjusted so that
| repeat. Close scrutiny reveals that the waveforms in B are gen- discharges of a single MUAP will appear time-locked on suc-
erated by superimpositions of the MUAP waveforms seen in A. cessive sweeps (Fig. 8-9D). The other discharging MUAPs
This is a critical observation in manual MUAP analysis. In order appear randomly on the display. In this sense they are similar to
to be identified as a MUAP, a waveform must be seen to occur noise in the recording. When the time-locked sweeps are aver-
at least three times. It is quite unlikely that chance superimpo-
3
aged, the noise is reduced and a clean recording of the MUAP
sition of different MUAPs will result in identical waveforms waveform is obtained (Fig. 8-9E). This is similar to averaging
that are repeated within the analysis epoch. the sensory nerve action potentials to reduce the background
I The manual method of MUAP extraction requires patient co- noise. It is important to realize that the improvement in the
1
operation. Furthermore, duration measurements can be difficult signal-to-noise ratio (i.e., reduction in noise) is proportional to
when the baseline is noisy (Fig. 8-8A). A typical study may re- the square root of the number of averaged sweeps. Most of the
quire over 30 minutes. noise reduction will be seen within the first 50-100 sweeps;
In the free-running mode of signal display (Fig. 8-8), individ- therefore, every effort should be made to minimize the noise,
ual MUAPs appear at random positions on the sweep, making it i.e., the interfering MUAP, by having the patient sufficiently
difficult to assess their configuration (Fig. 8-9A). An amplitude relax the tested muscle. We would like to emphasize that the
trigger and delay line allow one to freeze discharges of a signal processing techniques should not be used to substitute for
MUAP on the display. This electronic device is operated as
39
good recording techniques.
follows: The user defines an amplitude level (indicated by the When a triggered delay line is used, the amplitude level must
arrow in Figure 8-9B) using a control on the instrument panel. be set so that only one MUAP can trigger the sweeps, otherwise
If the EMG amplitude is less than this level, no signals are dis- the sweep will be triggered by several different MUAPs and the
played. The display is "triggered" when the EMG crosses the resulting averaged signal will not be a true MUAP. Needless to
, amplitude level. The signal immediately following the trigger say, this requires that the amplitude level be set high enough to
point is displayed until the sweep ends (Fig. 8-9C). The system be exceeded only by the largest-amplitude MUAP, whereas the
then waits for the next trigger to occur. When the amplitude trig- lower-amplitude MUAPs are excluded from analysis. Other
ger is turned on and the amplitude level is set such that it is forms of triggering mechanisms have been developed to over-
crossed by only one MUAP waveform, the EMG trace will be come this limitation. A window or peak trigger defines two
39
drawn only when that MUAP discharges. The MUAP discharge amplitude levels and accepts only signals with maximum ampli-
will occur at the same position in the display (Fig. 8-9C). This tude between these two levels. Signals that are smaller or larger
scheme does not allow visualization of the portion of the MUAP than both levels do not trigger the sweep.
waveform that precedes the trigger point. To accomplish this, This recording technique also requires considerable patient
the signals are passed through a device called a delay line, which cooperation. Furthermore, one must learn to manipulate the
adds a time delay before passing the signals to the display cir- needle position and simultaneously adjust the trigger levels and
cuitry. A combination of these devices permits the time-locked control signal acquisition. Since the waveforms are displayed in
a time-locked fashion and averaged, the duration can be as-
sessed better than in the manual method described earlier. A
typical study may require over 20 minutes.
With access to fast digital computers, one can automate the
process of MUAP detection, extraction, and measurements
considerably. This automated analysis is often called decompo-
sition 37122
or multi-MUAP analysis ( M M A ) . ' These tech-
6 1 3 7
niques are not available on most commercial systems. The basic

strategies in this approach are as follows. First, a 5-20-second
EMG epoch is acquired and digitized (Fig. 8-1 OA). Next the
signal is high-pass-filtered or differentiated (i.e., the rate of volt-
age change is computed) to enhance the spike component of the
signal (Fig. 8-10B). MUAP activity is detected when the
processed signal exceeds a threshold level. A portion of E M G
activity before and after this detection point is called a tem-
plate. (This is very similar to the amplitude-trigger method de-
scribed earlier.) After the templates in the signal have been
identified, the first template is assigned to the "first class." The
20 ms second template is compared with the first MUAP template. If it
is similar, it is assigned to the first class, otherwise it becomes
Figure 8-9. A m p l i t u d e t r i g g e r a n d d e l a y l i n e . In A, a 300-ms the template of the "second class." The third template is com-
epoch is shown in a free-running m o d e . T h e same signal is shown at pared with the first and second MUAP templates. It is either as-
slower sweep speed in B.The arrow in B indicates the trigger ampli- signed to one of those two classes or becomes the template for
tude level. The triggered sweeps are shown (C) without and (D) with the "third class." The process is repeated until all the MUAPs
a delay line.Three discharges are superimposed in D t o facilitate sub- have been assigned to a class (Fig. 8-10C). Finally, the classes
jective assessment.The averaged MUAP is shown in E. that have more than a minimum number of matching templates
more than 20 MUAPs in less than 3-5 minutes including time

for editing the waveforms. 193
Analysis of Measurements. Buchthal and coworkers

recorded 25 MUAPs from each tested muscle using a manual
analysis method. The muscle was sampled by random
25 1 5 3
needle insertions with no effort to manipulate the electrode to

optimize the MUAP amplitude or rise time. Only MUAPs with
1 2 3 4 amplitude greater than 50 uV were accepted for analysis. The
process of quantification was time-consuming, but nevertheless
their laboratory produced a large set of reference values for
j Class # 1 Class # 2 Class # 3
MUAP features in different muscles and for subjects in different
MUAP # 1 MUAP # 2 age groups (Table 8-1). - Most normal muscles had fewer
25 153
than 12% polyphasic MUAPs. The mean values of amplitude

ID and duration were computed. Muscles with mean durations
1
within 20% of the normal mean were considered normal. Those
investigators excluded polyphasic MUAPs from calculation of
the mean MUAP duration, which gave higher diagnostic sensi-
tivity for patients with myopathy. - - 27 202 203
Figure 8 - / 0 . P r i n c i p l e of M M A . T h e signal in A is filtered (B) t o While the data in Table 8-1 are impressive, it is important to
identify MUAP spikes.The dotted line represents the trigger level.The recognize some of their limitations. Although the values are tab-
templates are indicated by the horizontal bars under the signal and are ulated for more than 20 age groups, data for many muscles were
labeled numerically.The templates form t h r e e classes, two representing obtained from far fewer than 20 subjects. Most likely, some of
MUAP discharges and a third that contains their superimposition (C). the values were obtained from regression analysis. This would
The first t w o classes have sufficient discharges t o extract MUAPs (D). explain why the rate of change in duration values with age is
almost identical for many muscles. Finally, the duration values
for several muscles are extremely similar, occasionally identical
are considered to be comprised of MUAPs valid for analysis. (e.g., biceps = first dorsal interosseous = soleus; gastrocnemius =
Their templates from the recorded signal (Fig. 8-1 OA) are aver- extensor digitorum brevis; infraspinatus = flexor carpi ulnaris).
aged to obtain the MUAP (Fig. 8-10D) and its features are mea- Values within 20% of the mean are considered normal. Other
sured. An example of such an analysis is shown in Figure 8-11. studies show a 30% change from mean as the upper and lower
This approach requires less patient cooperation because it is normal limits (Tables 8-2 and 8-3; Figs. 8-12 and 8-13). The
not necessary to limit the activation to just one or two MUs as in change in MUAP duration with age is also less significant. - - 15 93 198
manual MUAP analysis. Furthermore, no instrument manipula- Stewart and coworkers used an amplitude-triggered delay
198
tion, such as trigger adjustment, is needed. One can easily line and averager to analyze MUAPs in the biceps muscle. The
obtain 3 - 4 different MUAPs from each recording, and with mean values of area/amplitude ratio and turns were also analyzed.
practice and experience it is possible to extract and analyze Only simple MUAPs were included in calculation of duration.
IL BICEPS |
0 4mV11 80ms P 3 T 2 | 0 2mV 12 60ms P 3 T 2
Sweeps Amp
mV
1.1 II 78|| 0.4|| 11.8Q|
\2 021112.601
1.3 1 85|| 0.4l| 5.301
Mean ALL 1 1! 03|| 9.90I
04mV5 30msP2T2 Sd ALL 1 II 0-1II 3.27]
Mean SIMPLE 1 II ' 03|| 9.90|
Sd SIMPLE 1 II 0.11| 3.27|
Total MUPs 1 II II 1
%Polyphasle 1 II II 1
%Simple 1 II II 1
Figure 8-11. E x a m p l e of MMA.Analysis of a 10-second epoch yielded 3 MUAPs. For each MUAP, the discharges are shown as if using a trigger
delay line.These can be seen in light gray color.The averaged MUAP is superimposed on these discharges in black.The superimposed discharges
make it easy t o assess the validity of automated analysis.The tick marks represent the MUAP onset and endpoinLThe MUAP measurements are
shown above the waveform and also in the result table to the right.The firing pattern is indicated in the lower right corner, where each tick mark
represents o n e firing of the MU within t h e 10-second epoch. Gaps in the firing pattern occur when the algorithm cannot resolve superimpositions
of different MUAPs.The user can edit the automated measurements or reject noisy MUAPs.
Table 8-1. Reference Values for MUAP Analysis

Mean MUAP Duration (ms) in Ocular, Facial, Laryngeal, and Masticatory Muscles
Lateral and Orbicularis Aryntenoid, Masseter
Age Medial Oris, Posterior and
(Years) Rectus Frontalis Triangularis Vocalis Cricothyroid Cricoarytenoid Temporalis Lingualis
0 1.6 4.1 4.9 2.1 3.3 3.0 5.6 5.1
3 1.8 4.4 5.1 2.3 3.7 3.3 6.2 5.7
5 2.0 4.5 5.3 2.5 4.0 3.5 6.7 6.2
8 2.2 4.7 5.5 2.8 4.4 3.8 7.3 6.8
10 2.4 4.8 5.7 3.0 4.7 4.0 7.7 7.2
13 2.6 5.0 5.9 3.2 4.9 4.2 8.1 7.6
15 2.7 5.1 6.0 3.3 5.1 4.2 8.4 7.9
18 2.8 5.2 6.1 3.4 5.2 4.4 8.6 8.1
20 2.8 5.3 6.2 3.4 5.2 4.4 8.6 8.1
25 2.8 5.3 6.2 3.4 5.2 4.4 8.6 8.1
30 2.8 5.4 6.3 3.4 5.2 4.5 8.7 8.2
35 2.8 5.4 6.3 3.4 5.2 4.5 8.7 8.2
40 2.8 5.5 6.4 3.4 5.2 4.5 8.7 8.2
45 2.9 5.5 6.4 3.5 5.3 4.5 8.8 8.3
50 2.9 5.6 6.5 3.5 5.3 4.5 8.8 8.3
55 2.9 5.6 6.5 3.5 5.3 4.5 8.8 8.3
60 2.9 5.7 6.6 3.5 5.4 4.6 8.9 8.4
65 2.9 5.7 6.6 3.5 5.4 4.6 8.9 8.4
70 2.9 5.8 6.7 3.5 5.4 4.6 8.9 8.4
75 2.9 5.8 6.7 3.5 5.4 4.6 8.9 8.4
80 2.9 5.9 6.8 3.5 5.4 4.6 8.9 8.4
No. of Subjects 20 49 49 32 32 32 10 10
Mean MUAP Duration (ms) in Neck, Shoulder, and Back Muscles

Age
( Years) Sternocleidomastoid Deltoid Pectoralis Major Infraspinatus Rhomboids Erector Spinae
0 6.5 7.8 6.8 8.6 8.3 8.2
3 6.9 8.3 7.2 9.2 8.9 8.8
5 7.1 8.6 7.5 9.5 9.2 9.1
8 7.5 9.0 7.8 9.9 9.6 9.5
10 7.6 9.3 8.0 10.2 9.8 9.7
13 7.9 9.6 8.3 10.5 10.2 10.1
15 8.1 9.8 8,4 10.6 10.4 10.3
18 8.2 10.0 8.6 11.0 10.6 10.5
20 8.4 10.2 8.8 11.2 10.8 10.7
25 8.7 10.5 9.1 11.5 II.1 11.0
30 8.9 10.7 9.3 1 1.9 11.4 11.4
35 9.2 1 I.I 9.6 12.2 11.8 11.7
40 9.3 11.3 9.8 12.4 12.0 11.9
45 9.4 1 1.4 9.9 12.6 12.1 12.0
50 9.6 11.6 10.0 12.8 12.3 12.2
55 9.8 11.8 10.2 13.0 12.5 12.4
60 10.0 12.1 10.5 13.3 12.9 12.7
65 10.2 12.1 10.7 13.7 13.2 13.1
70 10.2 12.4 10.9 13.9 13.4 13.3
75 10.6 12.8 1 I.I 14.1 13.6 13.5
80 10.8 13.0 11.3 14.3 13.8 13.7
No. of Subjects 10 25 48 24 13 10
(Table continued on next page.)
Table 8-1. Reference Values for MUAP Analysis (cont)

Mean MUAP Duration (ms) in Arm and Hand Muscles
Flexor Extensor Flexor Abductor Abductor
Age Biceps Digitorum Digitorum Brachio- Carpi Opponens Pollicis First Dorsal Digiti
(Years) Brachii Triceps Profundus Communis radialis Ulnaris Pollicis Brevis Interosseous Minimi
0 7.7 9.0 7.4 7.1 7.3 8.6 6.1 6.2 7.7 6.2
3 8.2 9.6 7.9 7.6 7.8 * 9.2 6.7 6.8 8.2 6.8
5 8.5 9.9 8.2 7.8 8.1 9.5 7.2 7.3 8.5 7.3
8 8.9 10.3 8.5 8.2 8.5 10.0 7.8 7.9 8.9 7.9
10 9.1 10.6 8.7 8.4 8.6 10.2 8.2 8.3 9.1 8.3
13 9.4 11.0 9.0 8.7 8.9 10.5 8.6 8.7 9.4 8.7
15 9.6 11.2 9.1 8.8 9.1 10.8 8.9 9.0 9.6 9.0
18 9.8 11.4 9.4 9.0 9.3 11.0 9.1 9.2 9.8 9.2
20 10.0 11.6 9.6 9.2 9.5 11.2 9.1 9.2 10.0 9.2
25 10.3 11.9 9.9 9.5 9.8 11.5 9.1 9.2 10.3 9.2
30 10.6 12.0 10.2 9.8 10.1 11.9 9.2 9.3 10.6 9.3
35 10.9 12.1 10.5 10.0 10.4 12.2 9,2 9.3 10.9 9.3
40 1 I.I 12.2 10.7 10.2 10.5 12.4 9.2 9.3 1 I.I 9.3
45 11.2 12.3 10.8 10.3 10.6 12.5 9.3 9.4 11.2 9.4
50 11.4 12.4 1 1.0 10.5 10.8 12.8 9.3 9.4 11.4 9.4
55 11.6 12.5 1 I.I 10.7 11.0 13.0 9.3 9.4 11.6 9.4
60 11.9 12.6 1 1.4 11.0 11.3 13.3 9.4 9.5 11.9 9.5
65 12.2 12.7 1 1.7 11.2 1 1.6 13.7 9.4 9.5 12.2 9.5
70 12.4 12.8 1 1.9 1 1.4 1 1.8 13.9 9.4 9.5 12.4 9.5
75 12.6 12.8 12.1 11.6 12.0 14.1 9.4 9.5 12.6 9.5
80 12.8 12.8 12.3 11.8 12.2 14.3 9.4 9.5 12.8 9.5
No. of Subjects 238 23 14 31 11 14 15 13 36 36
Mean MUAP Duration (ms) in Leg and Foot Muscles

Extensor
Age Gluteus Biceps Rectus Vastus Vastus Tibialis Peroneus Gastroc Digitorum Abductor
(Years) Maximus Femoris Femoris Medialis Lateralis Anterior Longus nemius Soleus Brevis Hallicis
0 9.2 8.5 8.7 7.9 9.7 9.5 7.6 7.2 7.7 7.2 6.5
3 9.8 9.1 9.2 8.4 10.3 10.1 8.1 7.7 8.2 7.7 7.0
5 10.2 9.4 9.6 8.7 10.7 10.5 8.4 8.0 8.5 8.0 7.2
8 10.7 9.9 10.0 9.1 1 1.2 1 1-0 8.7 8.4 8.9 8.4 7.6
10 10.9 10.1 10.3 9.3 11.5 11.2 8.9 8.6 9.1 8.6 7.7
13 11.3 10.4 10.6 9.6 1 1.8 11.6 9.2 8.8 9.4 8.8 8.0
15 1 1.5 10.6 10.7 9.8 12.1 11.7 9.4 8.9 9.6 8.9 8.2
18 11.8 10.9 II.1 10.0 12.3 12.1 9.6 9.2 9.8 9.2 8.4
20 12.0 II.1 11.3 10.2 12.6 12.3 9.8 9.4 10.0 9.4 8.5
25 12.4 11.4 1 1.6 10.5 13.0 12.7 10.1 9.7 10.3 9.7 8.8
30 12.7 1 1.8 12.0 10.8 13.4 13.1 10.4 10.0 10.6 10.0 9.0
35 13.1 12.1 12.3 II.1 13.7 13.4 10.7 10.2 10.9 10.2 9.3
40 13.3 12.3 12.6 11.3 14.0 13.6 10.9 10.4 II.1 10.4 9.4
45 13.4 12.4 12.7 11.4 14.1 13.8 11.0 10.5 1 1.2 10.5 9.5
50 13.7 12.7 12.9 11.6 14.4 14.0 11.2 10.7 1 1.4 10.7 9.7
55 13.9 12.9 13.1 11.8 14.6 14.3 11.4 10.9 11.6 10.9 9.9
60 14.3 13.2 13.5 12.1 15.0 14.7 1 1.7 1 1.2 1 1.9 11.2 10.2
65 14.6 13.5 13.7 12.4 15.4 15.0 12.0 11.5 12.2 1 1.5 10.4
70 14.9 13.8 14.0 12.6 15.6 15.3 12.2 11.7 12.4 1 1.7 10.6
75 15.1 14.0 14.2 12.8 15.9 15.5 12.4 11.8 12.6 11.8 10.8
80 15.4 14.2 14.4 13.0 16.1 15.7 12.6 12.0 12.8 12.0 1 1.0
No. of Subjiects 19 13 14 12 12 30 12 15 10 24 11
(Table continued on next page.)
Chapter 8 QUANTITATIVE EMG — 303
Table 8-1. Reference Values for MUAP Analysis (cont.)

CN MUAP Amplitude (pV) in Adult Subjects
Muscle Subjects Mean Standard Deviation Range
Deltoid 8 212 147 150-304
Triceps 9 340
Biceps Brachii 57 180 120-390
Extensor Digitorum Communis 34 210 1 15
Abductor Digiti Minimi 23 350
Abductor Pollicis Brevis II 260
Vastus Medialis 10 230 150-360
Vastus Lateralis 9 260 210-370
Rectus Femoris 9 170 130-215
Gastrocnemius 10 160 95
Tibialis Anterior 19 220
Extensor Digitorum Brevis 11 460
Mean values of duration are listed for different age groups in different muscles of normal subjects. Duration values that differ by more than 20% from mean are con-
sidered abnormal. CN, Concentric needle; MUAP, motor unit action potential
From Buchthal F: Electromyography in the evaluation of muscle diseases. In Fuglsang-Frederiksen A (ed): Methods in Clinical Neurophysiology,Vol. 2, Number 2 .
DANTEC Elektronik, Skovlunde (Denmark), 1991, with permission.
The reference values are described graphically in Figure 8-12. of amplitude was used to obtain a Gaussian distribution of the
These investigators found a much higher range of MUAP am- data and to define reference limits. The distribution of the third
plitude (200-800 uV) than the data in Table 8-1. They also largest and third smallest amplitude values in individual deltoid
found up to 20% complex MUAPs in their normal data in con- muscles of normal subjects is shown in Figure 8-14. These
trast to 12% by Buchthal and coworkers. The MUAP duration values are used to define the reference limits using the outlier
in their data increases slightly with age. The minimum to maxi- principle. A study is considered abnormal when more than two
mum range is roughly ± 30% about the mean. Analysis of these outlier MUAPs are found. Some of the reference values for dif-
data is discussed later in greater detail. ferent muscles using this technique are summarized in Tables 8¬
The MMA technique is used extensively by Stalberg and 2-8-4.
coworkers. They recorded 2 0 - 3 0 MUAPs from each tested
193
Comparison of these techniques (Tables 8-2-8-4; Fig. 8-12)
muscle and found a slight increase in MUAP duration with age. demonstrates that different methods arrive at similar values for
The age dependency was different for different muscles (Fig. 8¬ normal MUAP duration. This is partly due to earlier attempts to
13). This study also shows a roughly +30% variation about match the automated measurements with the reference values
mean as the normal range. The MUAP amplitude distribution was described in Table 8-1. The amplitude values are quite different
significantly skewed towards high values. Hence, the logarithm among these methods. This reflects the differences in selecting
Figure B-\2. Results of MUAP analysis. Results in a group of normal subjects are shown in t h e middle row.The mean values of amplitude,
area, area/amplitude ratio, and turns are presented as histograms.The arrows under the histograms indicate the normal limits.To the right, t h e
mean MUAP duration is plotted against age.The t w o solid lines in these plots define the normal limits also. Plots like these are useful in comparing
studies from patients with myopathy (top row) and neuropathy (bottom row). (From Stewart C, Nandedkar SD, Massey JM, e t al: Evaluation of an
automatic method of measuring features of m o t o r unit action potentials. Muscle Nerve 1989;12:141-148, with permission.)
Table 8-2. Reference Values for M M A

Muscle Amplitude (uV) Duration (ms) Thickness (Area/Amplitude)
Mean ± SD Max Min Mean ± SD Max Min Mean ± SD Max Min
Deltoid 550 ± 110 1531 162 10.4 ± 1.3 18.4 4.2 1.56 + 0.22 2.94 0.65
Biceps Brachii 436 ± 115 1414 178 9.9 ± 1.4 16.4 4.2 1.46 ± 0 . 2 2.09 0.56
First Dorsal Interosseous 752 ± 247 2301 188 9.4 ± 1.3 18.0 4.0 1.38 ± 0 . 2 2 2.61 0.49
Vastus Lateralis 687 ± 239 1954 172 I 1.7 ± 1.9 21.6 4.6 1.72 ± 0 . 2 3 3.11 0.6
Tibialis Anterior 666 ± 254 1572 194 I I . 4 ± 1.2 18.4 4.6 1.67 ± 0 . 2 3 2.81 0.58
The statistics of the mean values from individual normal subjects is indicated.The "Max and "Min" refer to the upper and lower limits for individual MUAP measure-
ments used in "outlier" assessment.
From Bischoff C, Stalberg E, Falck B, Eeg-Olofsson K: Reference values of motor unit action potentials obtained with multi-MUAP analysis. Muscle Nerve
1994; 17:842-851, with permisson.
Table 8-3. Reference Values for MMA

Amplitude Area Duration Firing Rate Polyphasic
Muscle # MUAPs (HV) (uV x ms) Area/Amplitude (ms) (Hz) Turns (%)
Biceps Brachii 593 370 ± 151 622 ± 307 1.8 ± 0 . 2 10.4 ± I.I 10.7 ± 1 . 2 2.1 ± 0 . 2 4.7 ± 3.7
Medial Cervical 387 534 ± 91 689 ± 1 9 4 1.3 ± 0 . 2 8.8 ± 1.2 7.9 ± 0.9 2.6 ± 0.3 7.4 ± 4.8
Medial Thoracic 363 588 ± 147 812 ± 1 4 1 1.5 ± 0 . 3 9.7 ± 1.5 8.2 ± 1 . 1 2.7 ± 0 . 2 7.4 ± 6.2
Medial Lumbar 369 563 ± 114 851 ± 3 1 7 1.5 ± 0 . 3 9.3 ± 1.4 7.4 ± 1 . 0 2.6 ± 0 . 3 6.3 ± 6.9
(Multifidus)
Lateral Lumbar 396 462 ± 4 1 795 ± 7 6 1.8 ±0.1 10.8 ± 1.0 7.5 ± 0.8 2.5 ± 0.2 4.6 ± 5.3
(Longissimus)
The statistics for the mean values from individual normal subjects is described.
From Barkhaus PE, Periquet Ml, Nandedkar SD: Quantitative motor unit action potential analysis in paraspinal muscles. Muscle Nerve 1997;20:272-275, with permission.
Table 8-4. Reference Values for MMA

Muscle # MUAPs Amplitude (uV) Area (uV x ms) Area/Amplitude Duration (ms) Firing Rate (Hz)
Cricothyroid 103 179 ± 132 181 ± 153 1.0 ± 0 . 4 5.1 ± I.I 12.8 ± 4 . 1
Vocalis 71 164 ± 141 133 ± 9 3 0.9 ± 0.4 4.9 ± 1.0 13.0 ± 3 . 7
Postcricoryten 102 163 ± 117 152 ± 8 9 1.0 ± 0 . 3 5.2 ± 1.2 13.2 ± 4 . 1
The mean and standard deviation of pooled data from MUAP recordings in the laryngeal muscle of normal subjectsis shown. (Courtesy Barkhaus and Jaradeh.)
MUSCLENR 26 Deltoideus
4.0'
MUSCLENR: 26 Deltoideus 3.5
3.0
2.5
2.0
> 1.5
tude
1.0
amplitude outliers
Q.
E .5 a
3 r dtowe s t
§ 0.0
* 3rd highest
Total Population
AGE
Rsq = 0,1861
Figure 8-14. MUAP amplitude.The outlier limits of MUAP ampli-

AGE tude in t h e deltoid muscle are defined based on the third largest and
third smallest value in normal subjects. N o t e t h e logarithmic scale.The
Figure 8-13. MUAP duration. Multi-MUAP analysis of the deltoid upper line is t h e upper limit of third highest ampliutdes (filled data
muscle of normal subjects shows a very slight increase in MUAP dura- symbols).The lower line is t h e lower limit of third lowest amplitude
tion with age. (open data symbols).
normal A B
0
^ — — "
o • o • o
» » n « ° n 5-
— V — — • ' _U
• •
neuropathy o.
S 25" C D
• .. . : .
'. " " " ^
i
15 65 65
Age (years)
myopathy Figure 8 - / 6 . Diagnostic sensitivity of M U A P features.The

MUAP duration is plotted against age.The solid lines represent t h e
normal limits. See t e x t for details. (From Stewart C, Nandedkar SD,
Massey JM, et al: Evaluation of an automatic method of measuring fea-
tures of motor unit action potentials. Muscle Nerve 1989;12:141-148,
with permission.)
Jsoouv
10ms each study the investigators calculated the mean values of am-
plitude, area, area/amplitude ratio, duration, and number of
Figure 8-15. M U A P in patients. MMA from the biceps muscle of a turns. Polyphasic MUAPs were excluded from calculations of
normal subject (top), a patient with neuropathy (middle), and a patient mean duration.
with myopathy (bottom). The distributions of mean values of amplitude, area, ratio,
and number of turns for normal subjects are shown in the
MUAPs for analysis. The manual method works best when only
one MU is Firing. In contrast, the MMA technique can analyze 25 T
3-4 different MUs in each tested epoch. The amplitude-trig-

gered method is biased towards selecting the largest MUAPs. It
•
is quite obvious that one must use reference data that match the o
analysis technique. Pa
• r-
Clinical Findings. Regardless of the MUAP extraction
method used, many laboratories have demonstrated similar pat- oW a 1
terns of MUAP abnormalities in patients with neuromuscular e

diseases. • • • • ' • • • • We describe one such study in
8 13 28 49 ,03 ,65 ,92 ,93 ,94 ,98
o
greater detail to correlate the MUAP abnormalities with the MU
architecture abnormalities.
Stewart and coworkers performed QA in the biceps brachii
198
muscle of 68 normal subjects, 64 patients with neuropathy, and

58 patients with myopathy. The diagnosis was based on patient
history, laboratory procedures including neurography, routine
needle EMG, and, in some instances, muscle biopsy. The tested
muscle was not necessarily involved at the time of EMG exami-
nation. An amplitude trigger, delay line, and averager were used
to record about 20 MUAPs from each tested muscle. For each Amplitude (mV)
MUAP, the amplitude was measured peak-to-peak. The duration
was measured using an automated marking algorithm. The Figure 8-17. M U A P amplitude and duration in neuropathy.
onset and end points of the MUAP were manually readjusted if The upper limit of normal amplitude and duration are roughly 800 uV
they markedly differed from results of visual assessment. The and 13 ms (Fig. 8-12). In this group of patients, the amplitude shows a
number of phases was measured between the beginning and end much higher increase compared t o duration. See t e x t for details.
of the MUAP. Turns occur at peaks of the MUAP (Fig. 8-6). To (From Stewart C, Nandedkar SD, Massey JM, et al: Evaluation of an au-
exclude peaks generated by noise, a turn is defined only if the tomatic method of measuring features of m o t o r unit action potentials.
amplitude changes by at least 50 jiV from the previous turn. For Muscle Nerve 1989;12:141-148, with permission.)
middle row of Figure 8-12. The study by Stewart and coworkers discussion to include MUAP measurements and the disease
demonstrated increased amplitude and/or duration in patients processes that alter the MU architecture. The core of the CN is
with neuropathy; reduced amplitude and/or duration in patients covered by the cannula on one side (Figs. 8-3 and 8-18).
with myopathy; and increased incidence of polyphasic MUAPs Therefore, the cannula behaves as a shield for the high-fre-
in both patient groups (Fig. 8-12). Similar findings are also re- quency components of potentials arriving from the shielded
ported using other methods of quantitative MUAP analysis. side. The core records sharp, high frequency-rich potentials
These findings provide the basic criteria of MUAP assessment from muscle fibers that are in front it. For this reason, it has a
in the routine needle EMG examination. r6ughly hemispherical uptake area for the spike component of
The QA study by Stewart and coworkers demonstrated 198
the MUAP. It can record low-frequency components from all
some other interesting patterns of abnormality that are not read- fibers around the core including those behind the cannula. 138
ily apparent. Although most patients with myopathy had normal These low-frequency components define the MUAP duration.
or reduced MUAP amplitude, a few studies showed a slight in- The MUAP waveform also depends on the electrode position
crease in this feature (Fig. 8-12). The amplitude increase was relative to the endplates (Fig. 8-19). When recorded from the
65
much less than that seen in patients with neuropathy. When the endplate zone, the MUAP has an initial negative deflection (Fig.
amplitude was increased in myopathy, the MUAP area/ampli- 8-19B). Away from the endplates, the MUAP has an initial posi-
tude ratio was often reduced. In neuropathy, this ratio was tive deflection that signals an approaching potential volley. If the
normal or increased (Fig. 8-12). display gain setting is high, the abrupt onset of the MUAP from
In Figure 8-16, the MUAP duration is plotted against the pa- the endplate area can be seen at any recording position (Fig. 8-
tient's age. Horizontal lines indicate the normal limits. The filled 19B). Towards the end of the MUAP, one can also observe a pos-
symbols indicate that there are abnormalities of MUAP features itive-going dip (Fig. 8-19C). This represents the arrival of the
other than duration. The data are divided into four groups based potential at the tendon. The waveform beyond this point repre-
on diagnosis and muscle strength: (A) mild neuropathy, (C) sents the extinction of the action potentials. If MUAPs are
56
moderate/severe neuropathy, (B) mild myopathy, and (D) mod- recorded with high gain simultaneously at different positions
erate/severe myopathy. It is easily seen that QA has a higher di- along the muscle fibers, the portions of the signal representing
agnostic sensitivity in moderately affected muscles (C, D) than onset in the endplate area and the arrival at the tendon are seen
in mildly affected muscles (A, B). In moderately affected mus- simultaneously at all recording sites (Fig. 8-19B and C).
cles, the duration is more likely to be abnormal. Although the Computer simulations (Fig. 8-18) indicate that the MUAP
132
MUAP duration was normal in many mildly involved biceps amplitude recorded by a CN depends mainly on the number and
muscles, there were abnormalities of amplitude, phases, and size of muscle fibers that lie within 0.5 mm of the recording tip.
turns (filled circles in A and B). In patients with neuropathy, The amplitude is sensitive to the size and distance of the muscle
many mildly affected muscles had MUAPs with high amplitude fibers closest to the active recording surface. Hence, a single
and duration (A). Furthermore, the relative changes in these fea- muscle fiber close to the recording tip of the needle can produce
tures are quite different. The amplitude is increased almost 10¬ a normal or even increased MUAP amplitude in myopathy.
fold, whereas the duration is only doubled (Fig. 8-17). Furthermore, the MUAP amplitude may change significantly
with slight changes in the recording electrode position. This is
MU Architecture and the MUAP Features demonstrated elegantly by the scanning EMG technique de-
We have discussed the relationship between the shape of scribed later.
the MUAP and the MU architecture. We will now expand this MUAP amplitude is higher when the aforementioned uptake
area contains an increased number of muscle fibers from the
MU. This is likely to occur after reinnervation and/or regenera-
Concentric Needle tion of muscle fibers. Conversely, a reduced number of muscle
fibers yields a reduced MUAP amplitude. Of note, muscle hy-
MU Territory pertrophy generates a high AP amplitude. On the other hand, at-
rophic muscle fibers have low amplitude potentials resulting in
smaller MUAPs. If the APs are synchronous, they summate
constructively, generating a high-amplitude MUAP. Increased
temporal dispersion separates the individual APs, yielding
smaller MUAP amplitude.
Spike Shape
(Turns & Phases)
As discussed above, if the MUAP duration is measured at a
high display gain (e.g., 10-20 |j.V/div), one can see the potential
Amplitude onset at the endplate, its arrival at the tendon (the positive dip),
and its extinction at the tendons (Fig. 8-19, B-C). The MUAP
duration measured from its first baseline deviation (i.e., onset at
endplate) to its ultimate baseline return (i.e., potential extinction
at the tendon) is called the physiologic duration. " This dura-
53 56
tion is tens of milliseconds and reflects the time of potential

propagation, temporal dispersion of the potential, and extinction
of the potential, but contains little information about the effect
Figure 8-18. U p t a k e a r e a of C N e l e c t r o d e . T h i s schematic shows of pathology, hence it has minimal clinical value at this time.
the cross-section of an MU and a CN electrode tip.The shaded semicircu- This technique also requires a very high signal-to-noise ratio,
lar areas enclose fibers that affect the MUAP amplitude and spike compo- which cannot be easily achieved in routine clinical recordings.
n e n t The shaded circle encloses the fibers that mainly define the MUAP In clinical analysis, MUAP duration is assessed at a lower
duration.These patterns were derived from computer simulations. display gain (e.g., 100 or 200 |iV/div). This corresponds to
Motor Neuron A
\Axon
Terminal Axon
NMJ Tendon
Branches
t tt 2 3
t
4 5
tt 6
Electrode Position along muscle fibers.
Figure 8-/9. M U A P w a v e f o r m a t d i f f e r e n t e l e c t r o d e p o s i t i o n s . In the biceps muscle of a normal subject, MUAPs w e r e recorded at dif-
ferent positions between t h e t w o tendons (B). For each MUAP the display gain (pV/div) is indicated t o the right of t h e trace.The vertical line
marks the same onset point at all electrode positions. Some MUAPs are shown with a higher display gain in C.The vertical line shows the same lo-
cation of the potential generated when APs arrive at the tendons.The location of the needle within the MU for each MUAP is indicated in t h e
schematic (A).
using an amplitude threshold of about 10-20 p.V to detect the can significantly affect measurements of the total MUAP dura-
onset and end of the MUAP. Measurements made in this tion, which poses a problem in duration measurements. Since
manner, conventionally called the MUAP duration, depend on these potentials are usually polyphasic because of increased
muscle fibers that are up to 2.5 mm from the recording tip of the temporal dispersion, one solution is to exclude polyphasic po-
electrode. This radius represents a substantial portion of the
132
tentials from duration analysis. ' - - - Another solution is
813 27 28 202 209
MU territory (Fig. 8-18); therefore, changes in MUAP duration to exclude the linked potential from duration measurements.
reflect variations in the MU size. The number of MUAP phases and turns may increase due to
The MUAP duration increases if the number of muscle fibers temporal dispersion of APs from MU fibers that are within 1
from the MU that lie within the electrode territory increases due mm of the recording electrode tip. An increase in the number
133
to reinnervation, and decreases if there is a loss of muscle fibers. of fibers within this recording area improves the chance of ob-
Changes in fiber size also affect MUAP duration. Since hy- 133
serving a temporally dispersed MUAP. However, if these poten-
pertrophic fibers generate large potentials, one would expect tials are synchronous, they yield a simple MUAP. Several
that they would produce an increase in the MUAP duration, architectural changes of the MU may increase the temporal dis-
whereas atrophy of fibers should reduce the MUAP duration. persion of the MUAP: longer length of terminal axon branches
When fiber diameters vary, the temporal dispersion is signifi- following reinnervation, slowed conduction in newly formed
cantly increased. The action potential of some muscle fibers axon sprouts, increased width of the endplate zone, slowed con-
close to the recording tip may arrive at the electrode earlier or duction in atrophic fibers, or relatively fast potential propaga-
later than other muscle fibers of the same MU because of fiber tion in hypertrophic fibers. In the case of dispersion due to fiber
size variation. A late-arriving muscle fiber potential may diameter variation, the polyphasicity is much more pronounced
appear as a satellite potential (Fig. 8-6). A single muscle fiber when the recording is performed remotely from the endplate
I—j Monopolar Needle Relative Amplitude versus Radial Distance
MU Territory
Duration
Spike Shape
(Turns & Phases)
Amplitude
Normalized Amplitude versus Radial Distance

Figure 8-20. Uptake area of MN electrode. This schematic
shows t h e cross-section of an MU and an MN electrode.The shaded
circular areas enclose fibers that affect various MUAP features. N o t e
the difference in uptake area for t h e amplitude and spike component
compared t o the concentric needle (Fig. 8-18).
zone. The number of phases is reduced if there is a severe loss

of muscle fibers. In this situation there may be just one muscle
fiber close to the recording tip, which generates a simple bi- or
triphasic MUAP.
Concentric vs. Monopolar N e e d l e Electrode

By virtue of its construction, the MN has a spherical record-
ing area for the high- and low-frequency components of the AP
(Fig. 8-20). Differences in uptake area give slightly different
MUAP characteristics from the CN e l e c t r o d e . - - 34 91101 102136144
Figure 8-2/. Radial decline of AP amplitude.The radial decline

King et al. built large-scale physical models of the concentric
of an extracellular potential computed from physical models. A, The
and monopolar needles and studied their recording characteris-
absolute amplitude of t h e potential is plotted versus t h e radial dis-
tics in a saline bath. A stimulator was used as a current source
tance. B, The amplitude is normalized t o the maximum value recorded
and its potential field at the electrode was recorded. The posi-
by the electrode.The plot shows t h e relative change in amplitude with
tion of the source was changed to simulate different distances
distance. In both plots, t h e inset shows variation at long distances.
between the source and electrode. The measurements were then
(From Dumitru D, King JC, Nandedkar SD: Concentric/monopolar
rescaled to reflect the true electrode dimensions. They observed
needle electrode modelling: Spatial recording territory and physiologic
a smaller absolute amplitude and slower decline of the extracel-
implications. Electroencephalogr Clin Neurophysiol 1997; 105:370-378,
lular AP amplitude for the monopolar needle compared to the
with permission.)
concentric needle (Fig. 8-21 ) . 1 0 1 1 0 2
The slow decline and a spherical uptake area results in a

greater number of muscle fibers contributing to the monopolar recorded by both electrodes to be similar. In clinical record-
MUAP spike component (compare fibers in the uptake area in ings slight differences are seen that can result from a variety of
Figures 8-18 and 8-20). The larger pool of fibers more than off- factors.
sets the lower amplitude of individual muscle fiber potential In the CN recordings, the cannula is used as the reference
recorded by the electrode, hence the monopolar needle elec- electrode. The cannula also records a "MUAP" that is usually
trode records higher-amplitude MUAP compared to the concen- much smaller than the MUAP recorded by the core. The differ-
tric needle. With more fibers contributing to the spike ential amplifier in the EMG equipment subtracts these two sig-
component, we have a better chance of observing their temporal nals to yield the CN MUAP waveform. For the MN, the
dispersion. This also gives a higher incidence of polyphasic reference electrode is on the skin surface, at some distance from
MUAPs in monopolar needle EMG recordings. 94136
the MU, where it records practically no electrical activity that
The MUAP duration is determined by a large number of would cancel the signal recorded by the needle tip. The cancel-
muscle fibers in the MU, most of which are distant from the lation from the cannula potential in the CN electrode should
needle tip. Other than a few muscle fibers that are immediately give a shorter MUAP duration. Indeed, the terminal positive-
behind the bevel of the CN electrode, there is little difference going dip seen on the core and cannula signals is cancelled by
in the number of fibers contributing to this MUAP feature as the differential amplification and thus is not seen in the CN
recorded by both electrodes. Hence we expect the MUAP durations MUAP (Fig. 8-22A).
The amplitude threshold used to detect the onset and end of a

MUAP also affects the MUAP duration differences. In Figure 8-
22B, there is a significant cancellation of the initial positive po- t BICEPS
tential, yet all potentials can be seen to start at the same time in
the endplate area. For a very low threshold value (i.e., a very 1 .
high display gain in subjective assessment), we would measure

the physiologic MUAP duration. As discussed earlier, this is de-
fined by the time of AP propagation, extinction of the AP at the \ j 100 MV
• • •
tendons, etc. This type of duration measurement is independent | 5 ms
of needle types. Similarly, when the amplitude threshold is high,
cancellation by the cannula potential becomes less significant.
This also gives similar MUAP duration measurements for both
needle types. For intermediate amplitude threshold values, we 3 /
expect longer MUAP durations for the MN electrode.

Most of the earlier work on MUAP quantification was per-
formed when the low-frequency setting of the band pass filter Figure 8-22. T h e c o r e , c a n n u l a , a n d C N M U A P . MUAPs
was at 2 - 5 Hz. The monopolar needle EMG recordings are recorded with a C N electrode from t w o different MUs (A, B) in t h e
often performed with the low-frequency setting at 20 Hz. This biceps muscle of a normal subject. For each MU, t h e middle and
higher filter setting attenuates the low-frequency components bottom traces are t h e potentials recorded by t h e core and cannula, r e -
and hence reduces the monopolar MUAP duration. spectively, with respect t o a r e m o t e reference. Subtracting the cannula
C l i n i c a l I n t e r p r e t a t i o n . Figures 8-12 and 8-15-17 demon- potential from t h e c o r e signal gives t h e standard C N MUAP ( t o p
strate the typical patterns that are used in everyday assessment trace). N o t e the tendon potential in the bottom t w o traces on the left
of MUAPs, but they also reveal some MUAP properties that are (indicated by *).
often not appreciated. These can be used to better plan, execute,
and interpret routine EMG recordings.
Diagnostic Sensitivity vs. Specificity. The MUAP duration containing just a few fibers can easily have a duration of 4 - 6
changes in different ways in different disease processes, which milliseconds. In such MUs it would be difficult to demonstrate
makes it suitable to differentiate a neuropathy from myopathy loss of muscle fibers based on duration changes. Therefore, in
(Figs. 8-12, 8-15, and 8-16). We refer to this attribute as speci- myopathy one should select muscles for duration assessment
ficity. Similarly, a very high MUAP amplitude (mean amplitude that normally have long-duration MUAPs (Table 8-1). This also
more than twice the upper limit) is seen only in patients with allows one to assess disease progression as the duration values
neuropathy. This is also a "specific" abnormality that differenti- decrease from low teens to 4 - 5 ms.
ates neuropathy from normal subjects or patients with a myopa- Muscles that are extremely weak should be avoided since
thy. Clearly these are the best parameters to use in distinguishing they may exhibit atypical MUAP waveforms.
among different disease states. Acute vs Chronic. In patients with neuropathy, one may ob-
In the study by Stewart and colleagues the MUAP duration
198
serve very large-amplitude and long-duration MUAPs, even in a
was normal when other features were outside the normal range. minimally affected muscle. This reflects compensatory reinner-
In other words, the duration was not diagnostically sensitive, es- vation after MU loss. When this is an unexpected finding, it re-
pecially in mildly affected muscles. In our experience, an in- flects a remote insult to the M U . ' - m
In a chronic slowly
2 1 2 2 1 4
creased number of polyphasic MUAPs is the earliest abnormality progressing neuropathy, the relative change in amplitude is far
seen in both patient groups. Although sensitive, it is a nonspecific greater than the change in MUAP duration (Fig. 8-17).
finding, and does not aid in the differential diagnosis. Amplitude can be assessed quite easily even in a free-running
A slight increase (mean value less than twice the upper limit display of EMG signals. This is a good MUAP parameter to
of normal as seen in Figure 8-12) in MUAP amplitude may be assess chronic neurogenic disease processes.
seen in both patient groups. This makes it a nonspecific finding. Active vs. Inactive. The instability of the MUAP waveform,
The high-amplitude MUAPs in myopathy often appear "thin" the jiggle, allows one to identify ongoing reinnervation
on visual inspection and have a low MUAP ratio (i.e., area/am- changes. This is discussed in greater details in the single-fiber
plitude), a feature that measures the "thickness" of the MUAP. EMG section.
In contrast, patients with neuropathy have normal or increased Complete Nerve Lesion. Almost all patients studied in
thickness values. The MUAP thickness parameter facilitates dif- Figures 8-12, 8-16, and 8-17 had biceps muscle strength of at
ferential diagnosis when the MUAP amplitude is mildly in- least 3 on the MRC scale. In muscles that are weaker, one may
creased. In a limited number of patients with myopathy, the not observe the above patterns of abnormality. If the nerve suf-
thickness parameter appears to have a higher diagnostic sensi- fers severe trauma, there may not be any functioning MUs in the
tivity than the MUAP duration. 131
muscle. As the nerve regenerates, the newly formed MUs are
Severity of Disease Process. In both patient groups, the very small. Their MUAPs may have low-amplitude, polyphasic
MUAP abnormalities are more pronounced and specific for a waveforms of short duration, similar to those that are usually
disease process when the muscle is moderately affected (Fig. 8¬ seen in patients with myopathy. Describing the MUAP in these
16). Therefore, one should study MUAPs in moderately af- conditions as "myopathic" would be an inaccurate representa-
fected muscles to best differentiate between neuropathy and tion of the underlying pathophysiology, hence we discourage
myopathy the use of such terms.
The MUAP duration has an inherent lower limit that corre- P i t f a l l s . In the manual method of MUAP extraction, one
sponds to the duration of a single muscle fiber AP, i.e., about must select only recurring waveforms. Failure to follow the rule
2-3 milliseconds. Computer simulations indicate that an MU
132
of "three repetitions" will include in the analysis waveforms
(Fig. 8-23, A, C). MUAP duration assessed by visual inspection

A• depends on the display sensitivity. Changing this setting from
'. .30msJ0QyV
200 to 500 uV/div (Fig. 8-23, A, D) gives shorter duration by
visual assessment. This display setting varies among different
B laboratories, but is usually between 100 or 200 uV/div.
In fully automated methods like multi-MUAP analysis, the
' 30ms 20D(A? E'
algorithms can fail and extract discharges of an unstable single
ff 30ms 200MV
MUAP as two or more different MUAPs. Such potentials must
be edited manually.
C
The reference values of MUAP features vary among different
!• .•'
1
- '
30ms 200yV
muscles. They are also seen to vary with the recording tech-
D nique (Tables 8-1-4), choice of electrode, as well as filter set-
tings. Ideally, one should develop his or her own reference data
for quantitative analysis, which is a Herculean task. When using
published reference values, all relevant details of the testing
Figure 8-23. The effect of electronic settings on the M U A P protocols should be followed. Even then it is good practice to
measurements.Ten consecutive discharges of a MUAP in E show test some normal muscles to validate the procedure and refer-
variability of waveform.The filter settings a r e 3-10000 Hz. W h e n aver- ence values. It is obvious that the introduction of quantitative
aged (A), t h e small spike in t h e MUAP is attenuated.The vertical tick analysis in the EMG laboratory requires significant effort and
marks labeled I and 2 are the onset and end of the MUAP determined time. Once established, however, it improves quality, accuracy,
by visual assessment. In B, the high filter frequency is reduced from and in the long run may save time.
10000 Hz t o 2000 Hz. In contrast, in C, t h e low-frequency setting in-
creased from 3 t o 100 Hz.The potential in D is the same as in A, but
the display gain is reduced t o 500 uV/div.
ELECTROPHYSIOLOGY OF THE INTERFERENCE
PATTERN
generated by superimposition of different MUAPs. They are The desired force of muscle contraction is achieved by two
often polyphasic and have a long duration, i.e., as seen in mechanisms: recruitment of MUs and modulation of their firing
pathology (Figs. 8-8B, 8-15). In our opinion, this oversight rate (Fig. 8-24). At a minimal force of contraction, a single MU
often leads to the observation of "increased polyphasic poten- is activated and the EMG contains only discharges of its MUAP.
tials" in an otherwise normal needle EMG examination. When the force of contraction is increased gradually (as in the ,
Whenever possible, polyphasic MUAPs should be directly ob- routine needle EMG examination), the firing rate of the MU in-1
served using an amplitude-triggered delay line. creases and successive twitches of the MU summate to give the
When the MUAP waveform is extracted by signal averaging, necessary incremental force. This process of increasing firing
MUAP variability will be obscured. A serrated waveform may rate is known as rate modulation. With a further increase in
appear smooth and have a smaller amplitude when averaged force, a stage is reached when another MU is activated, which is
(Fig. 8-23, A, E). Stability of the M U A P waveform should be called MU recruitment. The EMG recording now contains dis-
identified by the examiner (Fig. 8-23E). The default filter set- charges of two MUs. With a further increase in force both MUs
tings on the EMG instrument can often be incorrect. If the high- fire faster and a third MU is recruited, e t c . In some muscles
124
frequency setting of the band pass filter is lowered from 10 kHz all of the MUs may be recruited at only moderate force levels,
to 2 kHz, the amplitude is reduced, and peaks in the signal may generating higher force by increasing the MU firing rate. In
be missed (Fig. 8-23, A-B). Conversely, increasing the low fre- other muscles, all MUs may be recruited only with near-maxi-
quency setting from 3 to 100 Hz will reduce the MUAP duration mal effort.
The size of an MU refers to the number and diameter of
muscle fibers contained within it. A muscle contains many MUs
m
l t t t t ttttttttttttt tt t t t t of different size. A normal muscle contains many small MUs
and fewer larger MUs (Fig. 8-25). The MUs are activated based j
2 t t titttt t t t on their size. The small MUs are activated at minimal force,
whereas the large MUs are recruited at higher force levels.
3 t t ttt t t When a patient relaxes, the MU firing rates decrease and each
MU is released in the reversed order of recruitment, i.e., the last
recruited MU will be the first to stop discharging and the first
recruited MU will be the last to stop firing when the muscle is
Force
completely relaxed (Fig. 8-24). This pattern of orderly recruit-
ment and release of MUs is called the size principle. In this
89
context "size" was initially related to the membrane properties

Time
of the anterior horn cell (AHC). A small AHC has high mem-
brane resistance. This was shown in studies of normal animals
Figure 8-24. Force generation.The t w o mechanisms are illus- to be correlated to slow conduction axons and to small number
trated schematically. In all plots, time is shown along t h e X axis.The of muscle fibers.
lower graph shows an increase in the force followed by a decrease in The recruitment and firing rate modulation of MUs during
force as t h e muscle is relaxed. In t h e t o p graph, each vertical a r r o w voluntary force generation is elegantly demonstrated by the
represents one firing of the MU. Small separations between the arrows technique of precision d e c o m p o s i t i o n .
98106171
A special multi-
indicate higher firing rates, and vice versa. See text for details. electrode with four small recording surfaces is used to record
Chapter 8 QUANTITATIVE EMG — 31 I
Normal Muscle Neuropathy

#ofMU
N = 17 N = 11 N =6
MU Size
:::U Loss of MU &
Loss ofMU &
Reinnervation Fractionation
Formation of new
MU after Nerve Reduced MU size ^— Severe fiber loss
Regeneration (Loss/atrophy of fibers) causing loss of MU
N=6 N = 17 N = 13
•
i s .
Severe Trauma to Nerve Myopathy
Figure 8-25. MU size and number in pathology.The distribution of MU size in a normal muscle is illustrated schematically. Observe t h e
change in the number and size of the MU due t o disease processes.
I three channels of bipolar EMG signals. These are used to iden- The results of one such experiment in a normal subject are
tify almost every discharge of every MUAP contained in the shown in Figure 8 - 2 6 . This demonstrates quite vividly the
195
EMG signal. The force of contraction is also monitored. In a change in MU firing rate with the force of contraction. T h e
typical study, the subject gradually increases the force and then technique allows one to observe the high firing rate of MUs
relaxes the tested muscle. The process of decomposition is par- (e.g., 30 Hz for MU 1) even at high force levels. With this tech-
tially automated and does require some operator intervention. A nique we can also see that the last recruited MU stops
plot is constructed to demonstrate the relationship between the discharging first and vice versa. Although the individual M U
firing rate and force versus time. Several seconds of contraction firing rates are different, their firing rates change in parallel.
may require an analysis time of a few minutes to over an hour.
Figure 8-27. Recording for recruitment and firing rate analy-

sis. Monopolar needle EMG recording in t h e biceps brachii muscle of a
normal subject as t h e force of contraction was gradually increased and
SECOND
then decreased.Throughout t h e recording, t h e force of contraction
was minimal. Several different events can b e recognized from this
Figure 8-26. Results of precision decomposition analysis.The epoch. A, Distant MU activity. B, MU # I is recruited. C, MU # 2 is r e -
solid line indicates the force of contraction.The dotted lines show the cruited. D, MU # 3 is recruited. E.The force of contraction is reduced.
mean firing rates of several MUs. N o t e the orderly recruitment and The firing rate of all MUs falls. F, MU # 3 stopped discharging. G, MU #
release of MUs, t h e high firing rates and t h e parallel change in their 2 stopped discharging. H , MU # I stopped discharging. It is difficult t o
firing rates. (From Stashuk D, Deluca CJ: Update on the decomposition recognize discharges of MU # I in the presence of larger MUAPs from
and analysis of EMG signals. In D e s m e d t JE (ed): Computer-Aided o t h e r MUs. MUAP identification by visual assessment of this 20-
Electromyography and Expert Systems. Clinical Neurophysiology second epoch took the first author about 45 minutes. Firing rate analy-
Updates. Basel, Karger, 1989, pp 39-53, with permission.) sis of these events is shown in Figure 8-28.
Manual EMG Decomposition
_JL 4-^-*J *-«-*J— LÂ—oU^-X-J
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Time ( S e c o n d s )
Figure 8-28. Analysis of firing rate and recruitment. The firing

rates of t h e three MUs identified in Figure 8-27. Note the orderly re-
cruitment and release of MUs.The MUs change their firing rate in paral-
lel.Also note the drop in MU firing rate whenever a new MU is recruited.
Increase or decrease in firing rate of one MU usually coincides

with a similar change in firing rate for the other MUs. This
property is quantified as the so-called central drive.
The information observed on the precision decomposition
analysis is present in routine EMG recordings, but is not readily
extracted due to several practical problems (Fig. 8-27). Although
it is quite easy to identify and measure the firing rate of single
MUs activated with minimal force of contraction, when the force
of contraction is increased, and the firing rate of the MU increases
and new MUs are recruited, MUAPs of two or more MUs super-
impose. This makes MUAP identification and the discharge rate
J 1 mV
assessment quite difficult. In a clinical environment one cannot
devote long periods of time to analyzing a single EMG epoch. 100 ms
Hence, the recruitment and firing rates of MUs can be observed
and studied only at minimal force of contraction where less than Figure 8-29. EMG signal at increasing force levels. Concentric
two or three MUs are discharging. At these low force levels, the needle EMG recording in the biceps brachii muscle of a normal subject
firing rates of MUs are much lower than those measured with the as t h e force of contraction was gradually increased from minimal t o
precision decomposition technique (Fig. 8-26). One can also ob- maximal. From this recording six 500-ms epochs at successively higher
serve a slight decline in the firing rate of MUs when a new MU is force are shown ( I - 6 ) . T h e quantitative measurements for these sig-
recruited (Fig. 8-28). This is probably a compensation for the ad- nals are described in Table 8-8.The dotted horizontal lines in 6 indicate
ditional force generated by the newly recruited MU. the estimated envelope amplitude.
At minimal force, the EMG may contain discharges of only
a few MUs (Figs. 8-8, 8-29-1). One may be able to identify In both disease processes the MUAP waveform is altered due to
their individual MUAP waveforms and firing rates. When the changes in the MU architecture. For patients with upper motor
force of contraction is increased, the MU firing rates and the neuron disease, the MU firing pattern (rate) is affected. These
•number of discharging MUs increase (Fig. 8-29-2). The EMG disease processes affect the IP signal, which forms the basis for
signal becomes complex and individual MUAPs cannot be rec- IP assessment.
ognized by visual assessment (Fig. 8-29, 3). As the force in-
creases further, the signal baseline becomes completely Interference Pattern Analysis
obscured by MUAPs. Such a pattern is called "full" (Fig. 8-29, Unlike MUAP analysis, different techniques of IP assessment
4-5). At higher force levels, the signals contain even more measure the signals in distinctly different ways. The IP signal
MUAP spikes. Such a pattern may be called dense (Fig. 8-29, depends primarily on the force of contraction exerted by the
6). The EMG signals recorded in this fashion are called the in- tested muscle. A number of analysis strategies have been devel-
terference pattern (IP). The choice of the term "interference" oped to take into account the degree of muscle activation.
may not be the most appropriate since it usually implies unde-
sired external signals. Nevertheless, the term is now well ac- IP Analysis at Maximal Effort
cepted in the descriptions of EMG signals. The IP signal Technique and Measurements. Buchthal and coworkers
depends on the number of discharging MUs, their firing pat- proposed a simple semiquantitative approach to analyze the IP
tern, and their MUAP waveforms. 128
analysis when the subject exerts maximal force of contraction. 28
In neuropathy, the main disease process is a loss of MUs. In The fullness and envelope amplitude are assessed subjectively.
myopathy, the number of MUs is relatively unaffected (Fig. 8-25). The IP is said to be full when the signal baseline is completely
Table 8-5. Reference Values for IP Analysis

Age Amplitude (mV)
> 15 years 2
10-12 1.5
7-9 1
4-6 0.5
The lower normal limit for the EMG IP envelope amplitude at maximum effort
is indicated. No specific muscles are defined.
From Buchthal F: Electromyography in the evaluation of muscle diseases. In
Fuglsang-Frederiksen A (ed): Methods in Clinical Neurophysiology, Vol. 2,
Number 2.DANTEC Elektronik,Skovlunde (Denmark), 1991,with permisson.
obscured by MUAP spikes. If the baseline can be seen between

MUAP discharges, the IP is said to be reduced or incomplete.
If individual MUAP discharges can be recognized, the IP is
called discrete. The upper border of the E M G envelope is de-
fined by connecting the negative (upward-going) peaks,
whereas the lower envelope border is established by connecting Figure 8-31. Visual IP assessment at different display set-
the positive (downward-going) peaks. The voltage difference tings. Concentric needle EMG IP was recorded in t h e biceps muscle of
between these lines is the envelope amplitude (Fig. 8-29, 6 ) . a normal subject. Signals obtained with moderate effort are shown with
Solitary spikes are excluded from these envelope amplitude de- different display settings.The IP signals in A are shown with typical set-
terminations (Fig. 8-30B). The envelope amplitude limits for a tings used in routine EMG examination.The same signal is shown with a
concentric needle electrode are given in Table 8-5. different sweep duration (B) and display gain (C).This gives a different
Clinical Findings. Buchthal and coworkers described two
28
perception of the baseline and number of spikes in t h e signal.
criteria for IP evaluation at maximal effort (Fig. 8-30): (1) if the
| IP is full in a weak and/or wasted muscle and the envelope am-
plitude is reduced, the pattern is consistent with a myopathy;
and (2) a discrete IP pattern with increased envelope amplitude
is consistent with neurogenic disease processes.
Clinical Interpretation. Assessment of the IP at maximal
effort is an inherent part of the needle EMG examination. 132 2 0 1
200 pV
The aforementioned criteria are used to differentiate myopathy
from neuropathy. However, they may not be a sensitive method
of analysis. As discussed earlier, the spike component of a
MUAP is defined by fibers within 500 um of the needle tip. This
1 mV
50 ms
/a
h
C
ft
• 10 ms
Figure 8-30. IP signal at maximal effort. Concentric needl

EMG recording in the biceps brachii muscle of (A) a patient with my-
opathy, (B) a normal subject, and (C) a patient with neuropathy. In my- Figure 8-32. MU firing rate assessment. In A, a 500-ms epoch is
opathy, the IP is full with reduced amplitude. In neuropathy, the pattern shown as a single sweep. T h e same epoch is shown as 5 rastered
is reduced with fewer spikes and higher amplitude. For t h e normal IP sweeps of 100 ms duration in B. Four MU discharges are recognized,
in B, a high-amplitude solitary spike (*) is easily recognized. which gives an 8-Hz firing rate.
area contains muscle fibers from roughly 20 or more MUs.

Therefore, the DP in a normal muscle is generated by over 20 MUs.
Assuming a 50% loss of MUs, the IP is still generated by 10 MUs.
If there is a discharge rate of 20 Hz and MUAP spike duration of 5
ms, the EMG can still completely obscure the baseline giving a
"full" pattern (10 MU x 20 Hz x 5 msec = 1000 msec of EMG
spikes per second!). As a corollary, a reduced or discrete pattern at
maximal effort implies a significant loss of MUs.
Analysis of the IP at maximal effort requires patient coopera-
tion. If the patient cannot deliver maximal effort due to pain or
other factors, the IP may appear incomplete and have a low en-
velope amplitude. This cannot be considered abnormal. In such
situations, the assessment of MU firing rate and MUAP analysis ]200 JJV
is of great value. 10 ms
Pitfalls. The fullness and envelope amplitude measurements
are inherently subjective and hence the assessment may vary Figure 8-34. Onset and recruitment frequency. Monopolar
among different investigators. The display settings can also needle EMG recording in the biceps brachii muscle of a normal sub-
affect the assessment of IP fullness. If the display gain is low ject. The o n s e t of MU # l is seen in A and recruitment of MU # 2 is
(e.g., > 2 mV/div) one may not be able to see MUAP spikes in seen in B.
the EMG. This can give an impression of a reduced IP (Fig. 8¬
31C). If the sweep duration is long, it will contain more MUAP
spikes compressed within a small portion of the display, which the interval between successive discharges of the same MU is
will make it look more dense (Fig. 8-3IB). Therefore, subjec- 100 ms, i.e., one sweep duration. Hence the MUAP will appear
tive IP assessment should be performed at a standardized dis- at approximately the same place on each sweep (Fig. 8-33B). If
play gain and sweep speed. the discharge rate is less than 10 Hz, the MUAP will appear to
shift to the right on successive sweeps (Fig. 8-33A). If the inter-
IP Analysis at Minimal Effort val between successive discharges, called the interdischarge
Technique and Measurements. At minimal effort one can interval (IDI), is longer than 100 ms, there may be some
recognize and quantify the recruitment of individual MUs and sweeps with no MUAP discharge (Fig. 8-33A, bottom trace).
their firing rates. A simple way of estimating the MU firing
148
Conversely, the MUAP will appear to shift to the left when the
rate is to display a 500-ms epoch on the screen. This may repre- firing rate is more than 10 Hz (Fig. 8-33B). In this case, the IDI
sent one single sweep with a slow sweep speed (Fig. 8-32A) or is shorter than 100 ms and on some sweeps the MUAP may be
several rastered sweeps at a faster sweep speed (Fig. 8-32B). seen twice (Fig. 8-33B, top trace). When the MU firing rate is
Individual MUAP discharges are identified by visual inspection. 20 Hz, one can see two MUAP discharges on each sweep (Fig. •
The number of discharges of the MU in this half-second epoch 8-37B).
is multiplied by 2 to get the firing rate. The firing rate of an MU when it is first recruited is called its
If the sweep duration is set to 100 ms, the MU firing rate can onset frequency (Fig. 8-34A). In contrast, the firing rate of
145
be assessed using some simple rules. At a firing rate of 10 Hz, an MU when another MU is first activated is called the
recruitment frequency. To compute the recruitment fre-
145
quency, the two discharges of the earlier recruited MUAP just
Figure 8-33. MU firing rate assessment. In A, the MUAP is dis-

charging at 8 Hz. T h e dotted line parallels t h e progressive shift in
MUAP position t o t h e right, indicating that t h e firing rate is less than Figure 8-35. Recruitment ratio measurement. In A , monopo-
10 Hz. In B, the position of the smaller MUAP is relatively fixed on the lar needle EMG recordings in the biceps muscle of a normal subject at
screen, paralleled by t h e vertical dotted line, indicating that t h e firing minimal effort are shown. Discharges of three MUs are identified and
rate is 10 Hz.The position of the large amplitude MUAP shifts pro- labeled.A similar recording from a patient with neuropathy is shown in
gressively t o the left, paralleled by the o t h e r dotted line, indicating that B.The recruitment ratio is 4 (i.e., 12 Hz/3 MU) in A, and 9 ( 1 8 Hz/2
the rate is greater than 10 Hz. MU) in B.
250 Table 8-6. Reference Values for Onset

and Recruitment interval
O
UJ
Onset Recruitment
| 200 Muscle Interval (ms) Interval (ms) #MU
Frontalis 102 ± 2 9 46 ± 16 72
150 Orbicularis oris 70 ± 19 34 ± 10 58
All Facial muscles 86 ± 29 40 ± 16 130
100- Deltoid 116 ± 2 3 84 ± 16 53
Biceps brachii 124 ± 2 1 86 ± 14 56
Triceps 132 ± 3 6 84 ± 17 49
50-
Brachioradialis 116 ± 2 2 78 ± 18 36
F = -0.42
R = -0.35 Pronator teres 132 ± 3 8 88 ± 19 35
50 100 150 200 250 First Dorsal Interosseous 142 ± 3 9 98 ± 2 1 51

MSEC Multifidus 152 ± 3 3 102 ± 2 0 39
Vastus lateralis 126 ± 3 0 88 ± 18 80
Gluteus maximus 128 ± 3 0 88 ± 16 48
Tibialis anterior 124 ± 2 6 90 ± 13 85
250
O Biceps femoris 132 ± 2 9 92 ± 16 64
UJ
C/) Medial gastrocnemius 156 ± 2 9 110 ± 2 3 54
200 Extensor digitorum brevis 138 ± 2 9 98 ± 13 43
150 Normal Muscles 132 ± 3 2 9 0 ± |9 693

Neuropathy 48 ± 17 36 ± 12
100 Myopathy 65 ± II 45 ± 8
ALS 64 ± 3 0 42 ± 2 2
The mean and standard deviation of pooled data from normal subjects is indi-
50 cated'for different muscles. At the bottom, the measurements are compared
against patient groups.
F = 0.58 From Petajan JH, Phillips BA: Frequency control of motor unit action potentials.
R = 0.73 Electroencephalogr Clin Neurophysiol 1969;27:66-72, with permission.
IT 100 150 200 250
MSEC
(IDI) vary slightly from one MU discharge to next. This vari-
Figure 8-36. IDI analysis.A scatter plot of the interdischarge inter- ability is quantified by computing the coefficient of variation of
val (IDI) versus t h e previous IDI shows (A) negative correlation and the IDI (i.e., SD/mean). This calculation represents the overall
(B) positive correlation. (From Shahani BT, Wierzbicka M, Parker SW: (or long-term) variation in the firing rate. To study the instanta-
Abnormal single m o t o r unit behavior in the upper m o t o r neuron syn- neous or short-term variation, a plot of IDI versus the immedi-
drome. Muscle Nerve 1991; 14:64-69, with permission.) ately preceding IDI is constructed (Fig. 8-36). The correlation
coefficient and other statistics of this plot may also be com-
p u t ed . - This analysis has not been introduced into routine
71162 216
before the activation of next recruited MUAP are identified assessment of the EMG pattern.
(Fig. 8-34B). The time difference between those discharges is Clinical Findings. The statistics of onset and recruitment in-
the recruitment interval. When the interval is measured in mil- terval in different normal muscles is summarized in Table 8-6.
liseconds, the frequency is obtained by the formula Review of the distribution of these measurements (Fig. 1 in
Petajan, 1974) suggests a normal range of 5-12 Hz for onset
Recruitment frequency (Hz) = 1000/recruitment interval
frequency and 7-16 Hz for recruitment frequency. In a patient
When several MUs are discharging, the recruitment
ratio is calculated by dividing the number of discharging
4143
Table 8-7. Antigravity Testing

MUs into the firing rate of the most rapidly discharging MU
(Fig. 8-35). Subject Group #MUs # Spikes/sec
Petajan proposed the antigravity 45° test to assess the
147148
Control 4.2 ± 1.6 40 ± 2 0

recruitment. The EMG was recorded when the patient main-
Neuropathy 3.2 ± 1.7 29 ± 2 0
tained the forearm at 45° angle and overcame the gravity. He
identified all MUs that had MUAP amplitude of more than 100 Dystrophy 8.3 ± 2.4 122 ± 4 2
uV. The number of MUs, their firing rates, and the total number Polymyositis 6.8 ± 2.6 76 ± 2 9
of MUAP spikes per second were counted by visual inspection. The mean and standard deviation values of measurements in different patient
Even when the force of contraction is maintained constant, groups are compared. Note that only spikes with more than 100 MV amplitude
the MU firing rate is not constant. The interdischarge intervals are used in analysis. 147
. 4 ^ 4 ^ -
Figure 8-37. Firing rate analysis in disease. Concentric needle Figure 8-38. Calculation of recruitment ratio in myopathy. A
EMG recording in t h e biceps muscle of (A) patient with neuropathy 500-ms epoch of CNEMG recording at minimal force is visually exam-
and (B) patient with inflammatory myopathy. Both recordings demon- inated t o identify discharges of different MUAPs. In A, 5 MUAPs are
strate single fast firing units corresponding t o increased recruitment identified, discharging at less than 10 Hz.The recruitment ratio is less
frequency.This finding is atypical for myopathy, but can be seen in se- than 2. In B, even at minimal force, many MUs are recruited. It is rather
verely weak muscles.The asterisk indicates low-amplitude MUAPs that difficult and cumbersome to analyze this epoch for computing the re-
are n o t easily recognized. If we could analyze them t h e recruitment cruitment ratio.
frequency could be normal.
group Petajan found reduced values of onset and recruitment

146
frequency. Let us assume that the second through sixth recruited
intervals. This corresponds to a higher MU firing rate (Table 8¬ MUs are lost. Our measurement of recruitment frequency in this
6). The increased onset and recruitment frequency is readily muscle will correspond to the firing rate of the first recruited
seen in patients with neuropathy (Fig. 8-37A). It is more appar- MU when the seventh MU just begins to discharge. The result-
ent in severely affected muscles of patients with myopathy (Fig. ing higher values of recruitment frequency also mean that single
8-37B). The recruitment ratio is increased in patients with neu- fast firing MUs will appear during the routine needle EMG (Fig. j
ropathy (Fig. 8-35B). In patients with myopathy, the recruit- 8-37A). The reduced number of MUs allows us to observe their
ment ratio may be normal or reduced (Fig. 8-38). discharge at higher rates. Furthermore, the number of MUs con-
The results of antigravity posture EMG recordings are sum- tributing to IP is reduced. This combination leads to calculation*
marized in Table 8-7. The patients with myopathy recruited
147
of an increased recruitment ratio (Fig. 8-35B).
more MUs compared to normal and gave increased number of The greater the loss of MUs, the easier it is to make these
spikes in the recording. A contrasting pattern was seen in pa- measurements. When measuring recruitment frequency, it is not
tients with neuropathy. necessary to demonstrate recruitment of a second MU. A single.
In normal subjects, a scatter plot of successive IDI values MU discharging at more than 20 Hz implies that the recruitment
demonstrates a negative correlation (Fig. 8-36A). In other frequency is even higher than this (Fig. 8-37 A). If two MUs are
words, a short IDI is followed by a long IDI and vice versa. In discharging at more than 15-20 Hz, we know that the recruit-
patients with upper motor neuron disease the MU firing rate ment ratio is increased (Fig. 8-35B).
variability is increased. A positive correlation between succes- In a weak muscle, a patient with myopathy may activate sev-
sive IDIs may be seen in some of these patients (Fig. 8-36B). eral MUs even at minimal effort. This results in a reduced re-
' Clinical Interpretation. In the free-running EMG display, a cruitment ratio. The MUAPs may appear to fire at slightly
MUAP is recognized by its main spike amplitude. As discussed higher than normal rates; however, because the number of dis- j
earlier, the MUAP amplitude is defined by the number and size charging MUs is increased, the recruitment ratio is normal or
of muscle fibers within approximately 0.5 mm of the electrode. reduced. In a myopathy, the primary pathologic process is
This area may contain 1-4 fibers belonging to a normal M U . 132
muscle fiber loss. If the MU has lost 2 or 3 fibers that would
As a result, the electrode can record sharp detectable MUAPs have been close to the needle tip, it will generate a low-ampli-
from over 20 different MUs that lie near the needle tip. The dif- tude MUAP that may not even be detected. If the majority of
ferent MUAPs can superimpose, making it difficult to recognize fibers are lost from a MU, the MUAP may be too small to be de-
their individual waveforms. When more than 3 MUs are re- tected during the recording. This has the same effect as loss of
cruited, assessment of their waveforms and firing rate by visual MUs and increases the calculated recruitment frequency (Fig.
observation may not be possible within the time constraints of 8-37B).
the EMG examination. Therefore, in normal muscles we can re- Submaximal effort gives an incomplete or reduced pattern.
liably assess MU firing rates only at minimal effort, when they When this occurs, the analysis of MUAP waveforms and firing
discharge at low rates. Precision decomposition studies, how- rates can be quite useful. Patients with upper motor neuron dis-
ever, reveal that normal MUs can discharge at much higher fre- ease may not be able to recruit all their MUs and hence the IP
quencies (Fig. 8-26). pattern is incomplete. However, the MUs that are activated tend
In the clinical setting, it is possible to observe MUAPs firing to discharge at slower rates and their discharge pattern is irregu-
at high rates if the number of MUs near the needle tip is re- lar. In contrast, when normal subjects produce an IP that is in-
duced, e.g., in patients with neuropathy. There is a high recruitment complete, MUAPS have normal waveforms and a rhythmic
Tine
1
(nsec)
+
Mean=3JL0Hz Median=190Hz
Figure 8-39. D i f f e r e n t m e t h o d s of IP a n a l y s i s . An epoch of IP signal from a normal biceps brachii muscle is shown at the top.The p o w e r
spectrum of this signal is shown in the middle right plot.The dashed line in this plot indicates the median frequency.The turns and amplitude
(TA) analysis from the same muscle is shown in the left middle plot. More than 90% of the data points fall within the normal cloud.The expert's
quantitative IP (EQUIP) analysis in the same muscle is shown in the lower two plots.The data show a normal distribution, with fewer than 10%
of the data points are outside the clouds. (From Sanders DB, Stalberg EV, Nandedkar SD:Analysis of electromyographic interference pattern.]
Clin Neurophysiol 1996; 13: 385-400, with permission.)
firing pattern. In patients with lower motor neuron disease, the calculation of the recruitment ratio. When several MUs are dis-
MUs appear to discharge faster (i.e., increased recruitment fre- charging, their MUAP waveforms superimpose, making it diffi-
quency and ratio) and the MUAP waveform shows changes due cult to identify individual MUAPs and their firing rate,
to denervation and reinnervation. especially if the MUAP amplitude is low (Fig. 8-38B).
Pitfalls. A MUAP is recognized mainly by its spike compo-
nent, which is generated by muscle fibers closest to the needle Poêr Spectrum Analysis
tip. The EMG signal also contains small-amplitude MUAPs of Technique and Measurements. The sound of the E M G
distant MUs that are often ignored. Unfortunately, there is no signal is very helpful in assessing the IP. Walton used an audio
211
formal criterion to include or exclude MUAPs in the firing rate spectrometer in the early 1950s to demonstrate a shift to high-
analysis. This affects the assessment of the recruitment ratio and frequency components in EMG signals from patients with my-
the recruitment frequency. Signals in Figure 8-35B appear to opathy. The technique of frequency spectrum analysis can now
contain discharges of a single MU firing at a high rate. be carried out quite easily using digital computers. In this
However, if one accepted the smaller amplitude MUAPs in the method, also called Fourier analysis, " the IP signal is rep-
3132
signal, the recruitment would be considered normal. Similarly, resented as the sum of harmonically related sinusoids. A plot of
the inclusion or exclusion of MUAP 3 in Figure 8-8 affects the the power (measured as squared amplitude) for each sinusoid
the values found in normal subjects and plotted against the fre-
Quotients
quency. The four normalized values are close to unity for
normal IP signals, and the line connecting them has a slope
close to zero. When the spectrum shifts to high frequencies, the
normalized power for that band exceeds 1 and the slope is posi-
tive (Fig. 8-40). The slope of this line can also be used to assess
the severity and progression of the disease.
Clinical Findings. In myopathy the mean and median fre-
quency values are increased, indicating a shift towards higher
frequencies (Fig. 8-41). In neuropathy these frequency parame-
ters shift to lower values (Fig. 8-42). In all subjects, fatigue in-
creases the power in the signal and shifts the spectrum towards
low frequencies. - ' - 38 105 109 218
Clinical Interpretation. The shape of the component

MUAPs affects the power spectrum. Polyphasic MUAPs with
short rise times and short durations give rise to high-frequency
components. In contrast, long-rise-time, long-duration MUAPs
give a dull sound to the EMG, i.e., there is a shift to low fre-
quencies. This explains the typical patterns of shift in the spec-
trum observed in pathology. The total power reflects the number
of activated MUs and their MUAP amplitude, thus it increases
with force of contraction. During fatigue, the shape of the
MUAP changes due to alterations in the intracellular action po-
tential. This gives a shift to low frequencies and increased
power in the signal. Although a high-pitched sound of the EMG
50 200 1600 Hz is usually associated with myopathy, it may also be heard in
neuropathy due to increased incidence of polyphasic MUAPs.
Figure 8-40. IP a n a l y s i s u s i n g a b r o a d b a n d filters.The shaded
Pitfalls. The mathematical transformation in the power spec-
area represents the normal range of normalized power values in the
trum analysis makes the measurements less intuitive. For exam-
four frequency bands.The line has a positive slope, indicating a shift t o -
ple, one can look at a signal and estimate its envelope
wards high frequencies. (From Sandstedt P, Henriksson KG, Larsson
amplitude. However, it would be difficult for many to estimate
LE: Quantitative electromyography in polymyositis and dermatomyosi-
the mean or median frequencies. Finally, we are unaware of
tis.Acta Neurol Scand I982;65:l 10-121,with permission.)
well-defined reference values for this technique based on a large
number of control subjects.
against its frequency is the power spectrum (Fig. 8-39). (It is cus-
tomary to use logarithmic scales for these measurements to ac- Turns and Amplitude Measurements at Fixed Force
commodate the wide range of frequency values.) Analysis based Technique and Measurements. Power spectral analysis,
on the spectrum is often called the frequency domain approach. though elegant and fascinating, is rather abstract and too mathe-
The sum of the powers in individual sinusoids, i.e., the area matical for routine EMG assessment. In the early 1960s, it was
under the spectrum curve, is the total power in the signal. The also time-consuming, and time domain measurements were fa-
median frequency divides the spectrum into two halves. The vored. These measurements are made from the actual EMG
mean frequency is computed as: (L Frequency x Power)/Total waveform rather than its mathematical manipulations. A simple
Power. time domain measurement is the EMG amplitude. A signal with
Fourier analysis computes the power in hundreds of narrow- high median frequency should have more peaks. However, peaks
frequency bands. Considerable data are generated by this tech- may also be generated by noise. To exclude noise, Willison de- 215
nique, but most are not analyzed due to practical limitations. fined the so-called turn of the IP signal (Fig. 8-43).
Instead, simple calculations, such as the mean or median fre- • A turn occurs at the peak of the IP signal.
quency, are performed. Sandstedt and coworkers used a 159
• If a turn occurs at a positive-going peak, the immediately
simple approach in which the power was measured in four previous and next turns will occur on a negative going peak,
nonoverlapping bands. These measurements were normalized to and vice versa.
T a b l e 8-8. IP M e a s u r e m e n t s
Epoch # NT (/s) MA (uV) NT/MA Activity (ms) Envelope (uV) NSS (/s) RMS (uV) Median (Hz) Mean (Hz)
1 125 309 0.41 17 519 10 68 140 246
2 182 394 0.46 90 802 40 116 160 222
3 284 460 0.62 159 1152 98 184 140 191
4 332 836 0.40 351 2439 170 395 130 172
5 396 850 0.47 449 2464 240 475 130 156
6 352 1232 0.29 466 3497 174 664 140 162
Different parameters of the IP recordings in Figure 8-29 are tabulated.
A w <»U> Maopatti«
1 I 1 1 1 1 i 1 1 1 h
H 1 1 hr- i H 1 1 h
T ine (usee>
tfaan=446Hz ttodi a n = 2 7 0 H z
Anp/Turn CuU>
ssoo.o i—»— r
Envelope CnU) HSS </s>

JLO.O LOOO.O
O.O
Figure 8-41. IP analysis in a patient with myopathy.An epoch of IP signal from the biceps brachii muscle is shown at t h e top. Note the low
amplitude and increased number of spikes.The power spectrum shows a shift t o high frequencies (middle right plot). In t h e T A analysis from t h e
same muscle (left middle plot), m o r e than 10% of t h e data points fall outside t o t h e lower right side of t h e normal cloud.The EQUIP analysis
(lower two plots) demonstrates a full pattern (activity > 500 ms) with reduced amplitude and increased number of small segments (NSS) (i.e., a
shift t o high frequencies). (From Sanders DB, Stalberg EV, Nandedkar SD: Analysis of electromographic interference pattern. J Clin Neurophysiol
1996; 13:385-400, with permission.)
• Successive turns are separated by at least 100 pV of ampli- used a fixed load for manual resistance, e.g., 2-kg weight for the
tude difference. biceps brachii. In contrast, Fuglsang-Frederiksen and col-
In addition to the number of turns per second (NT) one can leagues - - - used 30% of the force generated by the patient's
72 73 74 75
also quantify the amplitude difference between successive turns. maximal effort as the force level for testing (Table 8-9). T h e
The average value of this measurement is called the mean am- latter group also measured the NT/MA ratio and number of
plitude (MA). short-duration intervals. Both groups recorded the IP signals at
In the biceps muscle, the NT increases linearly with the force different sites within the tested muscle. The mean values of
of contraction until it is roughly 50% of maximum (Fig. 8-44, measurements were compared against reference values derived
Table 8-8). At higher force levels it may increase slightly further from normal subjects.
or may even decrease slightly. The MA usually increases with Clinical Findings. Willison and coworkers observed in-
force through all levels of contraction. creased NT in dystrophy and increased MA in neuropa¬
Since the measurements depend upon the force of contraction thy.87.88.i52.2i5 Fuglsang-Frederiksen and c o w o r k e r s made 72-75
(Table 8-8), recordings forT/A analysis are performed at a standard- similar observations. In addition, they reported an increased in-
ized level of muscle activation. Willison and coworkers - 87 88152 2 , 5
cidence of short-duration intervals in myopathy. They found the
Neuropathy
H r
Tine (HSOC) iOO
Hean=134Hz Median=JLOOHz
flttp/Tum CuU> 'A Power
S9O0.O 14
Envelope ( n U ) NSS <^s>

1 0 . 0 LOOO.O I • • ' » r~
0.0 O.O
O.O lOOO.O lOOO.O
A c t i v i t y (ns> A c t i v i t y (HS>
Figure 8-42. IP analysis in a patient with neuropathy.An epoch of IP signal from the biceps brachii muscle is shown at the top. N o t e the high
amplitude and reduced number of spikes.The power spectrum shows a shift t o low frequencies (middle right plot).TA analysis from t h e same
muscle (left middle plot) shows that more than 10% of the data points fall outside on the upper side of the normal cloud. EQUIP analysis (lower two
plots) demonstrates a full pattern (activity > 500 ms) with increased amplitude and reduced NSS (i.e., a shift t o low frequencies). (From Sanders DB,
Stalberg EV, Nandedkar SD: Analysis of electromographic interference pattern. J Clin Neurophysiol 1996; 13:385—400, with permission.)
NT/MA ratio to be the most sensitive diagnostic feature and

found this IP analysis technique to be complementary to quanti-
tative MUAP analysis with similar diagnostic yields. By com-
bining them, the diagnostic yield was increased.
Amplitude change The Turns and Amplitude (TA) Method

between successive
turns Technique and Measurements. The IP analysis at fixed
effort has three major limitations: (1) requires additional equip-
ment for force monitoring; (2) limited to the muscles in which
the force of contraction can be measured in a relatively simple
fashion; and ( 3 ) requires significant patient cooperation to
maintain the force level.
Figure 8-43. Turns and amplitude measurements.The height Smyth, working with a pediatric population, found it ade-
169
of the gray box indicates the amplitude threshold for detecting turns. quate to analyze the MA/NT ratio. Obviously the force of con-
In the t w o areas indicated by * the small peaks produce an amplitude traction was not maintained constant at a particular level.
change less than the threshold, hence they d o not qualify as turns. Studies were performed in the tibialis anterior and quadriceps
Table 8-9. Reference Value of IP Analysis curved boundaries. Since the regression line continues indefi-
Parameter Biceps Brachii Vastas Medialis Tibialis Anterior nitely, upper limits of NT and MA are also defined to truncate
Female Male Male/Female Male/Female the cloud at the right and the top.
Clinical Findings. In patients with neuropathy (Fig. 8-42),
Subjects 28 33 12 14 some data points fall outside and on the upper side of the
NT 478 ± 67 533 ± 57 471 ± 31 438 ± 56 normal cloud. Data points inside the cloud tend to lie near the
MA 440 1 6 7 542 ± 76 581 ± 103 580 ± 102 upper edge of the cloud. In contrast, in patients with myopathy
(Fig. 8-41), data points tend to lie near the lower border or
NT/MA U0±0.I3 1.0 ± 0 . 1 4 0.82 ± 0 . 1 4 0.77 ± 0 . 1 1
below the normal cloud. Some data points may lie to the right of
The mean and standard deviation of different IP measurements in normal sub- the normal cloud. The normal clouds for some muscles are
jects are summarized.The force of contraction was 30% of maximal effort.
shown in Figure 8-45.
From Buchthal F: Electromyography in the evaluation of muscle diseases. In
Fuglsang-Frederiksen A (ed): Methods in Clinical Neurophysiology, Vol. 2, A study is abnormal when more than two data points out of
Number 2. DANTEC Elektronik,Skovlunde (Denmark), 1991, with permission. 20 are outside the cloud. This could occur within the first two or
three sites, in which case quantitative IP analysis in the tested
muscle can be concluded at this stage. In this respect, the quan-
femoris muscle of children suspected of having neuromuscular titative analysis would not add significantly to the time of the
disease. In uncooperative subjects, the muscle activation was EMG examination.
achieved by tickling the child or having them kick a toy. Clinical Interpretation. The contrasting patterns of data
Patients with no objective evidence of disease were considered point distribution in different disease states make TA useful in
as "apparent normal" and used to define the reference values. differentiating myopathy from neuropathy. Computer simula-
The normal range of mean MA/NT ratio was 0 . 5 2 1 - 1 . 0 7 1 170
tions have been used to study these changes in the NT and
Gilchrist and coworkers also found the MA/NT ratio useful in
80
M A . Initially the NT increases with the number of MU dis-
128
the IP analysis in adult subjects. charges, i.e., the number and firing rates of the MUs. When sev-
Stalberg and coworkers used a similar approach of record-
182
eral MUs discharge at high rates, their MUAPs superimpose.
ing, but a different way of assessing the abnormality of mea- When large- and small-amplitude MUAPs superimpose, the re-
surements. In their technique, now called turns and amplitude sulting signal resembles the larger MUAP (Fig. 8-46). The
(TA), the IP signals are recorded at different force levels rang- small MUAP is masked and we cannot see its turns. In other
ing from minimal to maximal at each tested site. The turns per words, the summated waveform has fewer turns than the sum of
second (i.e., NT) and mean amplitude between turns (i.e., MA) turns in individual MUAPs. Hence, the NT does not change sig-
are measured. A total of 20 or more data points are obtained nificantly at high force levels (Table 8-8). The reduction in N T
from 6-10 different tested sites in the muscle. A plot of MA values also corresponds to a slightly "dull" sound of the IP
versus NT is constructed. An area on this plot is defined, called recorded at high force levels.
the normal cloud, which contains more than 90% of data points The reduced NT in neuropathy reflects a reduced number of
from each normal subject (Fig. 8-39). A study is considered ab- MUs. The MA is increased due to higher MUAP amplitude after
normal if more than 10% of the data points are outside the reinnervation. In contrast, the NT in myopathy is normal or in-
normal cloud (Figs. 8-41 and 8-42). creased (data points to the right of the cloud). This reflects a
The normal cloud (Fig. 8-45) is defined by performing linear normal number of MUs and polyphasic waveforms. When the
regression between the NT and MA (or their logarithms) and se- amplitude of MUAPs is reduced, as seen by a reduced M A ,
lecting a standard deviation band that contains the desired per- there is less masking of the smaller MUAPs. This also gives
centage of data points. When logarithms are used to calculate more turns in the signal.
the regression, the standard deviation bands give the clouds Pitfalls. In the TA method the normal clouds depend on the
type of electrode used, and the patient's age and gender.
Normal clouds have been defined for only a few muscles,
A B which limits the use of this technique. The normal clouds also
appear to depend upon the technique of force measurement.
Mean Amplitude (JJV)
T ' 1200 ~-
Stalberg and c o w o r k e r s used manual resistance against the
182
patient to get the necessary force. This may not be adequate to

get maximum contraction of the tested muscle, especially when
the subject is stronger than the examiner. Nandedkar and
coworkers' had the subject pull against a strain gauge to reach
35
maximal effort. The latter approach gave a much higher values

of amplitude that fell outside the normal cloud of Stalberg et
al. (Fig. 8-47). The difference was seen mainly at high force
182
levels. Due to such differences, it is recommended that one de-

velop his or her own reference values for this technique, which
is a significant undertaking.
Figure 8-44. Turns and amplitude change with force. A,The Other Techniques Based on Turns and Amplitude
peaks in an IP signal epoch w e r e analyzed t o find turns, indicated byT. Many other strategies have been used to increase the diagnos-
The asterisk indicates peaks generated by noise. B.The change in the tic sensitivity of the TA analysis and to make it independent of
number of turns and mean amplitude at t w o different sites is shown. force of contraction - Liguori and c o w o r k e r s
75 79
monitored
107J08
The measurements w e r e made at minimal, mild, moderate, and maxi- the NT/MA ratio as the force of contraction was increased grad-
mal effort.The data points from each site are connected. ually from minimal to maximal. The peak value of the ratio was
NORMAL MATERIAL
AMPL
258o MALES FEMALES
BICEPS
EDC
QUADRICEPS
TIBIALIS ANTERIOR
Figure 8-45. Normal clouds for different muscles in CNEMG study.The cloud with square data points is used for patients over 60 years.
(From Stalberg E, Chu J, Bri! V, et al: Automatic analysis of t h e EMG interference pattern. Electroencephalogr Clin Neurophysiol 1983;56:672-681,
with permission.)
determined for each tested site. This usually occurred at moder- measures the portion of a 1-second epoch that contains spiky
ate effort (Table 8-8). The peak ratio was increased in myopathy MUAP signals. The activity values range from 0 to 1000 ms.
and reduced in neuropathy. The diagnostic sensitivity was The greater the numerical value, the more full/dense the pattern
higher in patients with myopathy. by subjective assessment. As with fullness by subjective assess-
Gilai plotted the total amplitude change (i.e., NT x MA vs.
79
ment, the normal value of the activity parameter differs between
the NT values). The slope of the mathematical curve fit for the muscles. The envelope amplitude was computed by discarding
data distribution was computed. The slope value was increased extreme peaks, which potentially represent superimposed
in neuropathy and decreased in myopathy. MUAP signals. The NSS represents the number of short-du-
134
ration, low-amplitude segments between successive turns. This

Expert's Quantitative IP (EQUIP) Analysis quantifies the high-pitched audio characteristics of the IP. The
Technique and Measurements. The previously discussed data for EQUIP analysis were acquired and analyzed similar to
quantitative measurements of the IP are quite different from the that of TA, i.e., by constructing plots of amplitude and NSS
subjective descriptions of the IP such as "fullness" and quality versus activity (Fig. 8-39).
of the sound. Therefore, Nandedkar and coworkers devel-
128,129
Clinical Findings. This technique of analysis is called
oped three parameters of the IP to measure features of the IP expert's quantitative IP (EQUIP) since it aims to mimic the
that are subjectively assessed: activity, envelope amplitude, subjective assessment performed by an expert electromyogra-
and number of small segments (NSS). The activity parameter pher. In myopathy (Fig. 8-41), the activity values are high (>
A
BICEPS
2500 , ,
— ^ 2
0 I 1 1 1 1 i 1 i t t
0 Turns (/sec) 1250
Figure 8-47. Normal cloud and force measurement.The data

points recorded from normal subjects by N a n d e d k a r and c o w o r k -
e r s ' are superimposed on t h e normal cloud described by Stalberg e t
35
al. (From Nandedkar SD, Sanders DB, Stalberg EV: On the shape of
132
the normal turns-amplitude cloud. Muscle N e r v e 1991;14:8-13, with

permission.)
increase as an indication of MU recruitment based on its size.

The uptake area of the concentric electrode is much smaller
than the MU territory and hence the CN or MN MUAP does not
adequately reflect the MU size. It is important to recognize
62
that the amplitude of the newly recruited MUAP depends on the

number and location of activated muscle fibers close to the
recording surface of the electrode. Hence, a newly recruited M U
may have lower- or higher-amplitude MUAPs compared with
other discharging MUAPs depending on how close its fibers are
to the electrode (Fig. 8-49). When the newly recruited unit lies
5 ma close to the electrode, its recruitment may be recognized quite
easily on visual assessment as an increase in the envelope am-
Figure 8-46. MUAP waveform interaction.A large (I) and a plitude. The small amplitude MUAPs of later recruited M U s
small (2) MUAP were summated (3-10) by computer simulation. N o t e contributes to the NT, NSS, activity, and fullness of the pattern,
that the summated r e s p o n s e resembles the large MUAP when t h e but not to the envelope amplitude.
MUAPs occur close t o each o t h e r (4—8). The summated waveform
contains fewer turns than the turns in the individual MUAPs. N o t e that
the amplitude of the summated MUAP (6) is not significantly greater
than that of t h e large c o m p o n e n t MUAP. (From Nandedkar SD,
Sanders DB, Stalberg EV: Simulation and analysis of the electromyo-
graphic interference pattern. Part I: Turns and amplitude measure-
ments. Muscle Nerve 1986;9:419^426, with permission.)
500 ms), indicating a full pattern. The NSS is increased, indicat-

ing an increase in the high-frequency audible components. The
envelope amplitude is reduced. In contrast, in neuropathy (Fig.
8-42), the activity values may be low or moderate (< 500 ms),
even at maximal effort. This indicates an incomplete or reduced
pattern. The NSS is reduced, indicating reduced high-frequency
components, corresponding to the dull IP sound. Finally, the en-
velope amplitude is increased. Figure 8-48. The change in IP amplitude with force of con-
Clinical Interpretation. The IP amplitude, measured as MA traction. EMG signals w e r e recorded as a normal subject gradually in-
or from the envelope, increases with force. This is recognized creased the force of contraction from minimal t o maximal effort.The
quite easily when a long sweep duration is used for IP record- enevelope amplitude increases with force.The arrow indicates w h e n
ings (Fig. 8-48). One may incorrectly interpret this amplitude the muscle was relaxed.
3 2 4 — PART II BASIC A N D ADVANCED TECHNIQUES
Figure 8-50. Bimodal amplitude. IP recording at maximal effort

Figure 8-49. Amplitude of second recruited MUAP. in the biceps brachii muscle of a patient with myopathy shows mostly
Monopolar needle EMG recordings a t t w o sites (A, B) in a normal low-amplitude MUAPs. O n e MUAP has a thin waveform with high am-
biceps brachii muscle. In A, t h e MUAP of t h e second recruited MU plitude. A u t o m a t e d envelope m e a s u r e m e n t s d o n o t exclude its dis-
(#2) has a smaller amplitude than t h e first recruited MU (#1). In B, the charges as solitary spikes. By visual assessment, one may ignore that
later-recruited MU has a higher-amplitude MUAP. MUAP and t h e resulting envelope is indicated by the dashed line.This
amplitude is reduced compared t o normal.
When different MUAPs superimpose, the resulting signal can

have a higher amplitude than either component MUAP. Pitfalls. For EQUIP analysis, the IP parameters were devel-
Computer simulations (Fig. 8-46) indicate that this occurs rel-
128
oped based on recordings in the biceps brachii muscle and
atively rarely. The peak-to-peak deflection of the MUAP (i.e., its hence may not be suitable for other muscles. This is most likely
rise time) occurs in less than a millisecond. There is a very to be true in muscles in which activity values remain low de-
narrow time window, less than a few hundred microseconds, spite a "full" pattern on visual inspection. As for the TA tech-
during which two MUAPs must occur to summate construc- nique, normal cloud limits must be developed for other muscles
tively. The chance of several MUAPs summating in this fashion in which the technique is to be applied. 30
is low. However, when this does occur, it is recognized as a soli-

tary amplitude spike. As discussed earlier, these spikes are ex- Decomposition
cluded from the envelope assessment. The envelope thus reflects Technique and Measurements. With digital signal process-
the upper limit of amplitude of the MUAPs contained in the IP. ing, IP signals may be decomposed into the constituent
In myopathy, MUAPs may have a bimodal amplitude distrib- M U A P s . The approach is similar to the earlier described
122
ution (Fig. 8-50). The high-amplitude MUAPs have a thin multi-MUAP analysis. However, this technique is more aggres-
waveform, which distinguishes them from neuropathy. Because sive in identifying MU discharges and attempts to resolve
they occur frequently within the analysis epoch, the envelope MUAP superimpositions. Dorfman and coworkers have used 122
amplitude measurement by automated methods may be normal. this approach for signals recorded at mild to moderate (30% of
They could potentially generate a data point that lies on the maximum) force levels. On average they could obtain 7 MUAPs
upper side of the normal "turns-amplitude" cloud. This should from each epoch of analysis.
not be misinterpreted as evidence of neuropathy and myopathy The reference values of features of MUAPs measured by this
in the tested muscle. When subjective assessment is performed, technique at different force levels are given in Table 10. 94
one may measure the envelope amplitude by excluding the dis- MUAP duration values at threshold force are similar to those
charges of that single MUAP (Fig. 8-50). determined by other techniques described earlier (Tables 8-1-8-3).
Table 8-10. ADEMG in Normal Subjects

Muscle Feature Concentric Needle Monopolar Needle
Threshold 10% 30% Threshold 10% 30%
Biceps Amplitude (uV) 422 ± 122 463 ± 139 638 ± 190 689 ± 192 798 ± 166 1172 ± 3 3 1
Brachii Duration (ms) 10.9 ± 1.0 9.4 ± 1.3 7.9 ± 0.8 10.6 ± 1.4 9.5 ± 1.4 7.4 ± 0.8
Turns 1.7 ± 0 . 3 1.8 ± 0 . 4 2.2 ± 0.3 2.1 ± 0.3 2.4 ± 0.3 2.6 ± 0.3
Rate (Hz) 10.6 ± 0 . 9 12.3 ± 1.2 16.0 ± 1.2 10.7 ± 0 . 9 12.1 ± 1.4 16.3 ± 1.4
Tibialis Amplitude (uV) 573 ± 111 586 ± 178 693 ± 2 1 1 1040 ± 3 4 4 II16 ± 3 4 0 1228 ± 3 3 7
Anterior Duration (ms) 12.6 ± 1.3 10.8 ± I.I 9.2 ± 0.7 12.8 ± 1.9 1 1.2 ± 1.7 9.6 ± 1.6
Turns 2.3 ± 0.3 2.5 ± 0.4 2.7 ± 0.6 3.3 ± 0.5 3.6 ± 0.2 3.8 ± 0 . 1
Rate (Hz) 8.4 ± 0.8 9.9 ± 0.9 12.2 ± 1.3 8.5 ± 0.6 9.7 ± 0.7 l l . 9 ± I.I
The statistics of different MUAP measurements is summarized.The MUAP differences at different force levels can be seen in both muscles.The data also show the
difference between the monopolar and concentric needle recordings.
From Howard JE, McGill KC, Dorfman LJ: Properties of motor unit action potentials recorded with concentric and monopolar needle electrodes:ADEMG Analysis.
Muscle Nerve 1988; 11:1051-1055, with permission.
Cumulative Properties for 70 MURPs froB 10 Insertions.

Mean Ualues Standard Deviations
AMP (uU) 506 818 ± 305 387 564 + 271
OUR < R S ) 8.8 8.7 + 1.1 3.5 3.7 ± 0.1
TURNS 2.0 2.3 ± 0.3 1.0 1.1 ± 0.2
FR (Hz) 10.7 1H.5 + 1.1 1.8 2.9+ 0.4
Muscie Force figc Group Gender

NORHs BRflCHIRL BICEPS 305c MUC 60-SO VRS BOTH
Figure 8-51. ADEMG in a patient with primary lateral sclerosis.The data are presented as histograms.The shaded areas represent dis-
tribution in normal subjects. In t h e patient, t h e amplitude, duration and t u r n s are relatively normal, b u t t h e MU firing rate is reduced. (From
Dorfman L; Howard J, McGill K: Clinical studies using automatic decomposition electromyography (ADEMG) in needle and surface EMG. In
D e s m e d t JE (ed): Computer-Aided Electromyography and Expert Systems. Clinical Neurophysiology Updates. Amsterdam, Elsevier, 1989, p p
The MUAP duration appears to be less at higher force levels. and myopathy. In patients with upper motor neuron lesions,
| This is contrary to expectations that larger high-force threshold firing rates were reduced (Fig. 8-51).
l MUs have longer duration. This probably results from the Clinical Interpretation. The analysis strategy is similar to
higher noise in the baseline of averaged MUAPs, which requires the MUAP analysis methods described earlier.
a much higher amplitude threshold to detect the MUAP onset Pitfalls. When several MUs are discharging, one may be
and end. The MU firing rate is increased at higher force levels, able to visually recognize 2 - 3 different MUAPs in the epoch.
as described earlier. The data also demonstrate the differences When the decomposition algorithm presents many more
in the MUAP feature values recorded by a concentric versus MUAPs, it is necessary to validate that individual MUAPs are
monopolar needle as discussed earlier. being measured. Validation of small-amplitude MUAPs can be
Clinical Findings. Although recorded at higher force levels, time-consuming. As with all other quantitative techniques, the
analysis involves measuring individual MUAP waveforms. The reference values have been defined for only a few muscles.
reported abnormalities are similar to those described in the Since this technique also records the higher force threshold
MUAP section. These investigators also analyzed the MU MUs, one may not use reference values from other M U A P
firing rate and found higher than normal rates in neuropathy analysis techniques.
Figure 8-52. Radial decline of the extracellular AP amplitude. Results of computer simulation for a single-fiber EMG (square), concentric
needle (diamond); and macro-EMG (triangle) electrode are shown. See text for details. (From Nandedkar S, Sanders D, Stalberg E: Selectivity of
EMG recoprding electrodes. Med Biol Eng Comput 1985;23:536-540, with permission.)
fiber action potential as recorded by single-fiber, concentric

needle, and macro-EMG electrodes. The extracellular potential
300 Mm pickup radius amplitude was measured as the electrode was positioned at dif-
ferent distances from the muscle fiber. In Figure 8-52A, the
plot of AP amplitude versus distance shows significant differ-
ences among the different electrodes in AP amplitude at short
recording distances. The single-fiber electrode recorded the
25 pm Diameter highest amplitude. Away from the fiber these differences
Wire
become less. In B, the amplitude for each electrode is normal-
m ized to its maximum value recorded near the muscle fiber. The
m relative amplitude decline is greatest for the single-fiber elec-
trode, reaching 10% at only 350 pm. In contrast, the macro-
m
Cannula EMG electrode shows only a gradual decline, reaching 10% at
1450 pm. The concentric needle electrode demonstrates inter-
Figure 8-53. S F E M G e l e c t r o d e s c h e m a t i c . A cross-section of a mediate characteristics.
normal MU is superimposed.The shaded semicircle represents t h e In single-fiber EMG (SFEMG), the recording electrode is a
uptake area of t h e e l e c t r o d e and contains only o n e muscle fiber of 25 pm-diameter wire exposed on a side port of the needle lo-
the MU. cated on the side opposite the bevel, 3 mm back from the tip
(Fig. 8-53). The cannula is used as the reference electrode.
Because of this design, it records very-high-amplitude sharp po-
SPECIAL RECORDING T E C H N I Q U E S tentials, occasionally more than 20 mV, from a single muscle
fiber close to the active recording surface. The potentials of dis-
The amplitude of the extracellular action potential and its de- tant muscle fibers have low-amplitude and mainly low-fre-
cline with distance from the electrode depend on the size of the quency components. By raising the low-frequency setting of the
recording electrode. This principle has been used to design
58
band pass filter to 500 Hz, the APs of distant muscle fiber are
special recording electrodes for the EMG examination. attenuated further. (In the routine needle EMG examination the
Computer simulations have been used to compute the muscle
127
corresponding low-frequency setting is 3-20 Hz.) The amplitude
200 pV
1 ms
1 1 mV
\ 0.5 ms
B
•J
0.5 mV
Figure 8-54. F D a n d M U fiber d i s t r i b u t i o n . Real fiber density signal recordings in B, C.and D are explained using a schematic MU cross-sec-
tion in A. For simplicity, a single MU cross section is used.The recording positions 1,2 and 3 correspond t o signals in B, C, and D, respectivelly.The
semicircle with a 300-pm radius represents t h e uptake area of the electrode.The FD is recorded is I in B, and 2 in C.The small time-locked po-
tentials indicated by the arrow do not fulfil t h e FD criterion and hence only 3 fibers are registered in D
Table 8-11. Reference Values for Fiber Density

Muscle Age (Years)
10 20 30 40 50 60 70 80 90
Frontalis 1.67 1.67 1.68 1.69 1.70 1.73 1.76
Tongue 1.78 1.78 1.78 1.78 1.78 1.79 1.79
SCM 1.89 1.89 1.90 1.92 1.96 2.01 2.08
Deltoid 1.56 1.56 1.57 1.57 1.58 1.59 1.60 1.62 1.65
Biceps Brachii 1.52 1.52 1.53 1.54 1.57 1.60 1.65 1.72 1.80
EDC 1.77 1.78 1.80 1.83 1.90 1.99 2.12 2.29 2.51
ADM 1.99 2.00 2.03 2.08 2.16 2.28 2.46
Quadriceps 1.93 1.94 1.96 1.99 2.05 2.14 2.26 2.43
Tibialis Anterior 1.94 1.94 1.96 1.98 2.02 2.07 2.15 2.26
Soleus 1.56 1.56 1.56 1.57 1.59 1.62 1.66 1.71
The upper normal limit for FD is tabulated for different age groups and in different muscles.The upper limit is higher in older subjects. SCM, Sternocleidomastoid;
ADM, abductor digiti minimi; EDC, Extensor digitorum communis.
From Bromberg M, Scott D: Single fiber EMG reference values: Reformatted in tabular form. Muscle Nerve 1994; 17:820-821, with permission, and from Gilchrist J,
Barkhaus P: Single fiber EMG reference values:A collaborative effort. Muscle Nerve 1992; 15:15 l-l61, with permission.
of the AP of a muscle fiber that is just 300 pm from the elec- it is called a selective recording technique. Single-fiber E M G
trode may be less than 200 p V . When the display is adjusted
178
provides a significant amount of physiologic information about
to accommodate the potential of the closest muscle fibers, sig- just one or a few muscle fibers, but nothing about the rest of the
nals from the distant fibers of the MU are not recognized. The motor unit. Therefore, it is primarily suitable for studying focal
recorded MUAP may represent the AP of just one muscle fiber abnormalities of MU architecture.
of the MU, hence the name of this recording technique. Since The selectivity of the electrode can be increased further by
this electrode records from only a very small portion of the MU, increasing the low-frequency setting of the band pass filter.
mmamammmam Muscle Fiber

• SFEMG Electrode
Figure 8-55. SFEMG jitter recording and analysis. In A, the electrode position is illustrated schematically. In B . t h e EMG signal is shown in
a free-running mode.The discharges of a single MU can be recognized quite easily at this slow sweep setting.The time interval between two MU
firings is the interdischarge interval (IDI). In C, a signal-triggered delay line is used t o freeze t h e signal for observation.The horizontal bar r e p r e -
sents t h e trigger level. In D, t h e triggered discharges are shown in a raster mode.The first potential triggers the sweep, thus appears time-locked
on t h e display.The second AP occurs at a variable interval. Furthermore, it did n o t occur in t h e sweep indicated by *, i.e., it blocked.
Using a bipolar configuration of the recording electrodes also MU microphysiology. Since that time, SFEMG has become a
reduces the uptake area. Such electrodes have been used for
1
major tool in the electrodiagnosis of disorders of the neuromus-
analysis of MU firing rate. cular junction. In conjunction with other recording techniques,
In m a c r o - E M G , the recording surface is the cannula of a it has contributed to a better understanding of MU reorganiza-
modified SFEMG needle. Because of the large recording sur- tion in myopathy and neuropathy.
face, this electrode records with comparatively very low ampli- The SFEMG signals are quantified mainly by measuring the
tude waveforms from single muscle fibers, even when they are fiber density (FD) and jitter. The latter measurement can be
close to the electrode surface (less than a few microvolts). The performed using volitional activity or by using electrical stimu-
potential of distant muscle fibers is less attenuated than with lation. The recording and measurement techniques for FD and
other electrodes (Fig. 8-52). As a result, all muscle fibers, near jitter are quite different and are discussed separately.
and distant, contribute to the macro-EMG MUAP. Since the
electrode records from a large portion of the MU, macro-EMG Fiber Density
is called a nonselective recording technique. This technique is Technique and Measurements. In fiber density measure-
useful to study global abnormalities of the MU, such as change ments (Fig. 8-54), signals are recorded as the patient exerts
in the MU size. minimal force of contraction from the tested muscle. The
Besides the large size of the recording surface, a technique sweep duration is usually 10-20 ms. The display gain is ad-
can become nonselective by virtue of electrode placement. As justed to visualize the entire signal on the screen. The position
seen in Figure 8-52A, the AP amplitude decreases most rapidly of the needle is adjusted to maximize the amplitude and mini-
at a short distance from the fiber. Away from the fiber, the de- mize the rise time of a single-muscle-fiber AP. A signal-trig-
cline is much slower and the differences between AP amplitudes gered delay line is used to "freeze" discharges of this potential
recorded by different electrodes become smaller. All muscle on the display screen. APs of other muscle fibers belonging to
fibers of an MU appear roughly equidistant to a remote record- the same MU appear relatively time-locked on the screen. The
ing electrode. In this case, no single-muscle-fiber AP dominates number of such time-locked potentials that have an amplitude
the MUAP, which makes the technique nonselective. For exam- greater than 200 pV and a rise time less than 300 ps is counted.
ple, the large distance between the M U and the large recording If two APs are partially superimposed, one may not be able to
surface makes surface EMG nonselective. assess their amplitudes and waveform. Nevertheless, it is rec-
Special recording techniques provide complementary infor- ognized by having a distinct notch in the rising or falling edge
mation about the MU architecture. We can obtain valuable infor- of the potential. Such signals are also counted in FD measure-
mation about the pathophysiology and disease progression by ments (Fig. 8-54D). The process is repeated at 20 sites within
combining the results of several special recording techniques. the muscle. The average of these measurements is called the
fiber density (FD).
SINGLE-FIBER EMG Clinical Findings. The FD value is always greater than 1 by
virtue of the recording technique. In normal muscles, record-
Stalberg and Ekstedt developed the SFEMG technique in the ings from most sites contain just the triggering potential or two
early 1960s. Significant subsequent contributions have been time-locked potentials. This gives FD values between 1 and 2.
made by collaborations between Stalberg, Trontelj, and The upper limit of FD varies among different muscles. The limit
others. SFEMG demonstrated that muscle fibers of a normal
191
is also higher for older subjects, especially after the sixth decade
MU are scattered within the muscle. This was in contrast to the of life (Table 8-11).
concept that fibers of an MU are arranged in tightly packed FD values are increased in neuropathy and increase progres-
groups, or subunits. This was a significant step in understanding sively when the disease progresses s l o w l y . - - - -
1810411 IJ50 175186 191 192 200
T a b l e 8-12. Reference Values for Jitter

Muscle Age
10 20 30 40 50 60 70 80 90
Frontalis 34/50 34/50 34/51 36/54 37/58 40/64 44/74
Orbicularis Oculi 40/55 40/55 40/55 40/55 41 / 5 5 42/55 43/56
Orbicularis Oris 35/53 35/53 35/53 35/54 36/56 37/58 38/62 40/67 43/74
Tongue 33/49 33/49 34/50 35/53 37/56 40/62 44/70
SCM 29/45 29/46 30/47 31 / 4 9 33/52 35/58 38/62
Deltoid 33/44 33/45 33/45 33/45 33/45 33/45 33/46 33/46 33/47
Biceps Brachii 30/45 30/45 30/45 30/46 30/46 31 / 47 31/48
EDC 35/50 35/50 35/51 35/51 36/53 37/54 38/57 39/61 41 / 6 7
ADM 44/64 45/64 45/66 46 / 69 48/74 51/83 55/97
Quadriceps 36/48 36/48 37/48 38/49 39/49 41 / 5 0 45/51
Tibialis Anterior 49/80 49/80 49/79 49/78 49/77 48/75 47/71 46/67 44/63
The upper limits are indicated for mean value from 20 pairs followed by upper limit for individual AP pair. The upper limit increases with age. SCM,
Sternocleidomastoid; EDC, Extensor digitorum communis;ADM, abductor digiti minimi.
From Bromberg M, Scott D: Single fiber EMG reference values: Reformatted in tabular form. Muscle Nerve 1994; 17:820-821, with permission, and from Gilchrist J,
Barkhaus P: Single fiber EMG reference values: A collaborative effort. Muscle Nerve 1992;15:151-161, with permission.
In chronic neuropathies, one occasionally finds recording sites MCD = IIPI — IPIi I + IIPI - IPI 1 + . . . + 1 I P I - IPI _, I / N - 1
2 3 2 N N
in which more than 10 muscle fibers are time-locked to the trig-

gering potential. In these situations it may be difficult to count where N represents the number of discharges and IPI is the N
precisely the number of potentials that fulfill the criteria of rise inter-potential interval in the N discharge. At least 50 intervals
th
time and amplitude. The FD values are also increased in some should be analyzed to compute the MCD. The mean value of
types of myopathies. The highest values are usually seen in the IPI (MIPI) is also computed. Jitter measurements are best
Duchenne muscular dystrophy, and other forms of myopathy performed when the MIPI is between 150 and 4,000 ps, for rea-
show much less increase. This increase is explained by abnor- sons discussed later.
mal fiber distribution in the dystrophic muscle (Fig. 8-55B). Sometimes the IPI is affected by the IDI from the previous
discharge (Fig. 8-55B), which makes it vary with the MU firing
Jitter rate. This would add jitter other than that due to neuromuscular
Technique and Measurements. transmission. To reduce the effect of firing rate variability on
Volitional Mode. For jitter measurements, the needle posi- the jitter calculation, one may arrange the IPI values in the as-
tion is adjusted to record APs from two (or more) muscle fibers cending (or descending) order of their corresponding preceding
in one MU (Fig. 8-55, A - B ) . The sweep duration is usually interdischarge intervals (IDIs). The jitter calculated on this or-
5-20 ms. The display gain is adjusted as necessary. One muscle dered set of IPIs is called the mean sorted difference (MSD).
fiber potential is used to trigger the sweep and the signal is de- We use the smaller of the MCD or MSD values to represent the
layed for display (Fig. 8-55C). The triggering potential appears jitter in the pair of potentials.
time-locked and stationary on the screen. The electrode position When jitter is greatly increased, indicating a pronounced neu-
is adjusted to find a position at which each of the individual romuscular disturbance, one fiber may fail to generate a poten-
spike components fulfill the criteria of being generated by a tial with some MU discharges. This is seen when some
single muscle fiber. Unlike measurements of FD, this is not nec- potentials are missing in some MU discharges (Fig. 8-55). This
essarily the position at which either of the APs is maximumal. If phenomenon is called blocking.
another potential appears relatively time-locked on the display, The jitter is measured from 20 different sites in the tested
it is in the same MU as the triggering AP (Fig. 8-55D). Every muscle. Reference values have been defined for the mean jitter
effort should be made to make sure that only one MU triggers from 20 recordings and also for individual values (Table 8-12).
the sweep. When several MUs are active, jitter analysis can be A study is called abnormal when the mean jitter and/or more
quite tedious, time-consuming, and subject to errors. The time than 10% recordings exceed their corresponding upper limits.
interval between the two APs, called the interpotential inter- The 10% limit corresponds to 2 abnormal recordings out of
val (IPI), varies from one discharge to another. This variability 20. It is quite possible that the third abnormal recording will
is the neuromuscular jitter is quantified by the following for- be seen within the first few analyses.The test has reached di-
mula, which calculates the mean difference between consecu- agnostic significance, and may be concluded at this point. If
tive potential intervals (MCD): the mean values are to be used to assess disease progression or
Figure 8-56. Stimulated SFEMG recordings. In A, the technique is indicated schematically.The cathode stimulates 2 different MU axons.The
electrode registers five different single muscle fiber APs. In B, signals recorded from one normal subject contain potentials from five different
muscle fibers.The jitter is normal.These APs belong to different MUs.
Figure 8-57. Automated analysis of jitter.Triggered sweeps recorded by an SFEMG electrode are shown in A.The H-shaped bars indicate
the windows placed on the signals t o detect AP.The signals analyzed based on the window positions and after manual editing are shown in B.The
scatter of sequenctial interpotential intervals is shown in C (sweep number on X axis, IPI on Y axis).The gray rectangle defines a ± 3 standard de-
viation band about t h e MIPI.The numerical data are presented in D. In this recording, the jitter is normal.
the effect of treatment, jitter should be measured in 20 potential portion of the muscle, and its position and the stimulus intensity are
pairs. adjusted to record visually identifiable single-fiber APs. Instead of
The jitter varies slightly among different muscles and tends the intramuscular stimulation, one may use surface stimulation
to increase with age. The reference values used in our laborato- technique to excite the tested nerve. 63
ries for patients aged 20-60 years are summarized in Table 8¬ Jitter recordings are made at a stimulation rate of 10 Hz,
12. Other muscles, such as laryngeal muscles, are now also which approximates the volitional MU discharge rate. It is im-
investigated with SFEMG. 161
portant to recognize that APs time-locked to the stimulus may
One can also measure the time interval between the first and belong to different MUs.
last time-locked AP in the signal. This is the maximum IPI for
that recording. The interspike interval (ISI) is computed as
maximum IPI / (number of potential time locked potentials - 1).
This parameter is often increased in myopathies (Fig. 8-62), but
also in early reinnervation.
Stimulated Mode. Jitter can also be measured following stimu-
lation. 191.204.205,208 p duration is usually 10-20 ms, but one
s w e e
may need to increase this further to view and analyze potentials

with a longer latency. The display gain is adjusted as necessary. A
monopolar needle inserted proximal to the tested muscle stimulates
the intramuscular nerve fibers; a surface electrode is taped to the
skin nearby as an anode (Fig. 8-56). Very short stimulus pulses are
used (50 ps) with an initial rate of 1 - 3 Hz. The stimulus intensity is
gradually increased to produce a small twitch of the tested muscle.
The position of the stimulating electrode is adjusted to get the best
results, which usually occurs at 5-10 mA. The stimulating needle
is then taped in place to maintain this position while the study is
performed. The SFEMG electrode is inserted into the twitching
Table 8 - 1 3 . Reference Values for Jitter in Stimulated Mode

Muscle Mean / Individual Figure 8-58. Jitter and blocking. In A, the SFEMG signals demon-
strate increased jitter and blocking.The blocking (indicated by arrow)
EDC 25/40 is easily recognized by t h e baseline through t h e second AP. In B, the
Orbicularis Oculi 20/30 baseline in the second potential suggests a block. However, the jitter is
The upper limit of mean value is listed, followed by the upper limit for individ- normal. In this recording, the apparent block was caused when a differ-
ual endplates. ent MU triggered the sweep (indicated by arrow).
A B
Split in muscle fiber
1ms
Figure 8-60. Split muscle fiber.The schematic illustrates the split

in muscle fiber between t h e endplate and recording e l e c t r o d e . T h e
Figure 8-59. Jitter and blocking. In these stimulated SFEMG
t w o time-locked action potentials shown below have very low jitter.
recordings, jitter is normal in A. In B.the jitter is increased and a block This recording should be excluded from jitter analysis.
is seen on the fourth trace.
To compute jitter, the IPI is measured from the sweep onset or of these windows is adjusted manually to exclude other noise and
stimulus artifact (Fig. 8-56B). The MCD and MSD are computed interference. The window, indicated by a horizontal bar in Figure 8¬
as described earlier. In a typical study, jitter is measured from 30 57, is usually placed at the steepest rising portion of the AP. When
endplates. The data are analyzed from the individual jitter measure- the potential crosses the bar, an AP is detected. When AP is detected
ments and also from their mean value. The reference values for in both windows, the IPI is measured. The systems allow the user to
stimulated mode are described in Table 8-13. Theoretically, the
205
review individual sweeps to exclude traces that are noisy.
reference values for the stimulated jitter is the voluntary jitter Nl. When the IPI is small, it is difficult to position the windows
Many commercially available systems use two time-amplitude on the rising edge (Fig. 8-54D). One may use the falling edge of
windows to detect a single-fiber potential (Fig. 8-52). The position the potential or its peak to measure IPI.
Figure 8-6/. Jitter and discharge rate. Stimulated SFEMG recordings in a patient with myasthenia gravis when the stimulation rate is (A) 5 Hz,
(B) 10 Hz, and (C, D) 20 Hz.The first spike at the beginning of the sweep is the stimulus artifact This recording demonstrates rate dependent jitter,
velocity recovery function. (See text for details.)
Jitter is increased in diseases of the neuromuscular junction

such as myasthenia gravis (MG) and Lambert-Eaton myas-
thenic syndrome ( L E M S ) . In postsynaptic conditions,
1 5 6 1 5 8 1 7 7
such as MG, the jitter in individual endplates typically increases

as the discharge rate increases from low rates and then de-
creases at rates over 10 Hz. In contrast, in LEMS there typically
is a more consistent decrease in jitter with the firing rate. This
relationship is not seen in all endplates in either condi-
tion. In severely affected muscles of LEMS, it may not
35J5 4 2 0 7
be possible to study some endplates at low activation rates be-

cause of the frequent blocking. This is demonstrated more
easily in the stimulated mode of analysis by using high and low
stimulation rates.
In muscular dystrophy, the recordings may contain several
time-locked potentials with a long maximum IPI (Fig. 8-62). - 20 191
One can record normal, increased, minimal, and frequency-de-

pendent jitter in these muscles. In other myopathies, the maxi-
mum IPI may be just a few milliseconds as in normal subjects,
and the jitter is usually normal or slightly increased. Frequency
dependence may not be prominent.
Jitter may also be increased in neuropathy. In rapidly progres-
sive denervating conditions the jitter tends to be higher. In se- 183
2.5n
verely weak and end-stage muscles the jitter and blocking may
be markedly increased. In these conditions, the jitter and block-
figure 8-62. SFEMG recording in a patient with Duchenne ing may be due to abnormal neuromuscular transmission, or due
dystrophy. N o t e t h e long interval between t h e first and last poten- to inconstant transmission through the nerve fibers. The latter
tial. (From Stalberg EV,Trontelj JV: Single Fiber Electromyography. Old phenomenon is demonstrated when the amount of jitter, block-
Woking, UK, Mirvalle Press, 1979, with permission.) ing, or both, is the same in more than one AP in a MU. This is
called neurogenic jitter and blocking (Fig. 8 - 6 3 ) . 150174
Clinical Interpretation. Fiber Density. Because the ampli-

Clinical Findings. In normal subjects the jitter varies be- tude of signals recorded with the SFEMG electrode decline
tween 5 and 60 psec, depending on the muscle tested and age of rapidly with distance from the electrode, this electrode records
the patient. Broadly speaking, the jitter abnormalities can be de- mainly from muscle fibers within 300 pm of the recording sur-
scribed as (1) increased jitter without blocking (Fig. 8-54D); (2) face. Each normal MU that overlaps this uptake area con-,
178
increased jitter with blocking (Figs. 8-58 and 8-59); (3) very tributes APs from only one or two of its fibers to the signal.
low jitter (< 5 ps) (Fig. 8-60); or (4) frequency-dependent jitter Hence, the SFEMG recordings normally contain just one or two
(Fig. 8-61). In diseases that affect neuromuscular transmission, potentials from any MU (Fig. 8-54). In neurogenic diseases, the
jitter varies considerably among different endplates and from number of muscle fibers belonging to a single MU within the
muscle to muscle.
Intracellular Potential
(mV)
Time
Figure 8-63. Neurogenic jitter and blocking.The SFEMG record-
ings (right) contain 6 different APs from one MU.The middle four poten- Figure 8-64. Genesis of jitter.The relationship between t h e end-
tials shift and block together.A physiologic model (left) suggests a defect in plate potential and jitter is d e m o n s t r a t e d schematically. A, Resting
nerve AP propagation at sites indicated by the arrow.The circle indicates membrane potential, B, Endplate potential, C, Action potential. ( I )
the SFEMG electrode. (From Stalberg EV, Trontelj JV: Single Fiber Normal jitter, (2) Increased jitter, (3) blocking.The gray area r e p r e -
Electromyography. Old Woking, UK, Mirvalle Press, 1979, with permission.) sents the threshold for generating AP.
uptake area of the needle increases due to reinnervation. This

gives higher FD values, which correspond to fiber type group-
ing seen on muscle biopsy. The FD increase is seen before the
type grouping is apparent on biopsy studies. This makes it the
most sensitive technique to study reinnervation. 90
In myopathy, muscle fibers may become hypertrophic or may

split into several small fibers. New fibers may be generated
from satellite cells. Some reinnervation may also occur. All of
these compensatory processes result in focal grouping of fibers
(Fig. 8-1), which yields a mild increase in F D . It is important
176
to recognize that the FD measurements cannot detect the loss of

muscle fibers from an MU. In patients with central core dis-
ease the muscle biopsy may demonstrate fiber type grouping.
14
The FD is normal and thus indicates that the apparent grouping

occurred due to an increased proportion of a fiber type.
Jitter. When the nerve action potential arrives at the end-
plate, the presynaptic terminal releases acetylcholine (ACh).
ACh binds to the postsynaptic receptors and generates an end-
plate potential (EPP) (Fig. 8-64). When the EPP exceeds the
membrane threshold, a muscle fiber AP is generated. The
Figure 8-66. Triggering potential and blocking.The SFEMG
SFEMG electrode records the propagating muscle fiber poten-
recording contains four time-locked APs.The third AP has increased
tial. In normal endplates, a large EPP is generated, which reaches
jitter and blocking. It also has t h e highest amplitude. W h e n used for
the threshold rapidly for each nerve AP. The time to reach thresh-
triggering, as seen in A, t h e remaining three potentials appear t o have
old varies minimally from impulse to impulse. This slight vari-
a high jitter.There is no blocking.The other back ground activity is in-
ability produces the small amount of jitter seen in SFEMG
dicated by *. In B, the smaller amplitude first AP is used for triggering.
recordings in normal muscle (indicated by 1 in Fig. 8-64).
This shows only o n e potential with increased jitter and blocking.
In pathology, the presynaptic terminal may release less ACh
(e.g., LEMS). In MG, the concentration and response of ACh
receptors are reduced due to destruction of the postsynaptic junction disease. During reinnervation, the newly formed end-
membrane. In both diseases, the EPP amplitude is reduced. On plates are immature and have poor neuromuscular transmission,
some occasions it may take much longer for the EPP to reach thereby producing increased jitter and blocking on SFEMG.
the action potential threshold (Fig. 8-64). This produces a Neuromuscular transmission is also uncertain during the process
greater variation in the IPIs measured by SFEMG recordings, of denervation and in some myopathies. Hence, jitter abnormali-
i.e., increased jitter (Figs. 8-54D, 8-58, and 8-59). When the ties, although specific for abnormal neuromuscular transmission,
EPP is very small (indicated by 3 in Fig. 8-64), it may fail to may be seen in a neuropathy or a myopathy, as well as in dis-
generate a propagating muscle fiber AP and will be detected as eases primarily affecting the neuromuscular junction.
a block on SFEMG (Figs. 8-58 and 8-59). 59
Very low jitter (< 5 ps) in a fiber pair during volitional
It is important to recognize that normal muscles also exhibit SFEMG recordings indicates a split muscle fiber or ephaptic
slight jitter (Tables 8-12 and 8-13). Furthermore, an increase in jitter transmission (Fig. 8-60). The recording surface is distal to the
and blocking does not necessarily imply a primary neuromuscular split or ephapse, which results in two potentials. However, both
potentials are initiated by a single endplate, which results in the
very low jitter. If very low jitter is observed in stimulated
Muscle fibre s t i m u l a t i o n recordings (Fig. 8-65), it indicates that the muscle fiber was di-
rectly stimulated by the intramuscular monopolar needle. Since
an endplate is not involved in potential generation, the jitter will
- 1 - be very low, and the results should not be included in calcula-
- f - tions of neuromuscular jitter.
The calculated jitter may be increased by phenomena other
than abnormal neuromuscular transmission. For a brief period
of time following an AP, the muscle fiber AP propagation veloc-
A x o n a l stimulation
ity is increased. If the next AP occurs within this period of faster
propagation speed, the potential will reach the electrode faster
than in the previous MU discharge. Somewhat later, the muscle
fiber propagation velocity may be slower than at baseline, so
that the AP reaches the electrode somewhat later than at base-
line. These changes in AP propagation are described as the ve-
2 ms 0.5 ms
locity recovery function (VRF). This phenomenon is easily
173
demonstrated by stimulated SFEMG. In Figure 8-61, the time

Figure 8-65. Direct muscle stimulation. Stimulated SFEMG interval between the stimulus artifact and the last potential (la-
recordings show very low jitter in A and normal jitter in B. In both beled 1) decreased when the firing rate increased. By visual as-
recordings the potential is shown at two different sweep speeds to ap- sessment one can also note a higher jitter in the same potential
preciate the jitter. In A, the muscle is stimulated directly, whereas in B at higher stimulation rate. The middle potential (labeled 2) does
the axon is stimulated. not demonstrate significant VRF.
During jitter studies with voluntary activation, if the two end

plates lie at different distances from the electrode, or if the
VRF is sufficiently different in the two fibers, changes in the
MU firing rate affect AP propagation differently, which in-
creases the jitter. This effect can be minimized by maintaining
a constant activation rate, or by measuring the MSD. The abil-
ity to control the rate precisely is an advantage of the stimu-
lated mode of acquisition.
The effect of VRF on IPI is more prominent when the IPI is
long. Hence, rate-dependent jitter is seen more prominently in
long-duration complexes, such as seen in muscular dystrophy. To
minimize the effect of VRF on jitter, potential pairs with MIPI
greater than 4 ms should not be included in jitter calculations.
In stimulated mode the jitter is measured from individual end-
plates. In contrast, the IPI variations in the volitional mode result
from the AP generation from two endplates. This gives higher
jitter values for the volitional mode analysis. Jitter measurements
from 20 jitter potential pairs reflect transmission at 40 endplates.
To achieve similar sampling in the stimulated mode, measure-
ments should be made from at least 30 endplates. It should be
recognized that these techniques may assess different MU popu- Figure 8-68. Injury potential on S F E M G recording.
lations, which theoretically could give different diagnostic sensi-
tivity to jitter analysis in mildly affected muscles. 125
Pitfalls: Recognizing Jitter and Block. While the concept When jitter is increased beyond a certain level, we expect to
of jitter and block is straightforward, in practice it can be some- see blocking. However, if the blocking potential is used to trig-
what complicated. Here we present some scenarios that may ger the sweep, and other time-locked potentials do not block,
cause errors and misinterpretations. we may not see the blocking (Fig. 8-66). The blocking is then
For the experienced operator, blocking is most easily recog- best recognized by its sound.
nized by its characteristic sound. On the screen, there are differ- When the SFEMG recording made during voluntary activa-
ent ways in which blocking can be identified in the EMG tion contains acceptable APs from more than two muscle fibers,
signals. It is seen rarely in normal muscles. Blocking is not seen jitter could be calculated among different combinations of po-
when the jitter is normal at the test neuromuscular junction. In tentials. For example, in a recording with three potentials (Fig.
MG, blocking is seen only when jitter at the tested endplate ex- 8-67), jitter could be calculated between APs 1 and 2, 1 and 3,
ceeds 100 ps. If the system indicates that there is blocking when and 2 and 3. However, to accept all three measurements would
jitter is normal, this is almost certainly a technical problem (Fig. create several potential problems. Let us assume that jitter is el-
8-58B). evated at one of the three endplates, for example #3. This will
give two values of increased jitter (2-3 and 1-3). We now have
2/3 abnormal values, although only 1/3 endplates are abnormal.
To reduce this bias, we accept only N - l measurements from N
time-locked APs of one motor unit, no more than one of which
has an elevated value. Other combinations are rejected. This ap-
proach will slightly underestimate the jitter in some complex
MUs, but that is preferable to the alternative.
When the interval between two APs is very short, their wave-
forms may superimpose and appear as a single broad spike. This
makes it difficult to assess the IPI. Therefore, jitter should not
be measured if the leading edge of the second potential "rides"
on the trailing edge of the first (Fig. 8-54D). This is not usually
a problem if the mean interpotential interval (MIPI) is more
than 150 ps.
One occasionally records signals with a waveform resembling
two APs, which in reality comes from a single muscle fiber (Fig.
8-68) The first component has a bi- or triphasic waveform typi-
cal of an AP. The second component has a triangular positive-
going waveform closely following the end of the first potential,
with a variable latency and frequent blocking. The second poten-
Figure 8 - 6 7 . Stable and unstable triplets.The SFEMG signals tial is a so-called injury potential, which results from trauma
contain t h r e e time-locked APs. In A, t h e signals are shown using each from the tip of the electrode needle. These potentials should not
potential t o trigger t h e sweep. Depending on the potential used for be used in jitter analysis. Failure to recognize this artifact can
triggering, o n e gets a different perception of jitter. By visual assess- give abnormal jitter values in normal muscles, and is a problem
ment, jitter appears least in the middle recording. In B, the second and for many during their learning phase of SFEMG.
third potentials show increased jitter and blocking. It may not be pos- In stimulated mode, a low-intensity stimulus may not activate
sible t o measure jitter. the nerve fiber consistently (Fig. 8-69), in which case the time
Figure 8-69. S t i m u l u s i n t e n s i t y a n d j i t t e r . In this stimulated SFEMG study, the intensity was increased gradually (from t o p to b o t t o m ) .
Responses are shown in raster and superimposed mode. See text for details.
to reach the excitation threshold of the axon varies considerably amount of jitter as the recording is being made. Any discrepan-
among successive stimuli. This variability increases the calcu- cies between this subjective assessment and the results of auto-
lated jitter. With some stimuli the axon may not be excited, mated measurements should be resolved before accepting the
which would be seen as blocking. Also, intermittent stimulation data.
failure produces a variable firing rate for the muscle fiber, which Whereas the FD can be measured on any instrument that
introduces artifactual jitter due to the velocity recovery func- offers a trigger, delay line, and a band pass setting of 500-10000
tion, as previously described. To avoid this pitfall, the stimulus Hz, jitter analysis requires special equipment.
should be adjusted to assure that it is optimal for each AP to be Jitter with Routine EMG Electrodes. Attempts have been
measured. 208
made to study the jitter in recordings performed with concentric
In Figure 8-69, the stimulus intensity was increased gradually or a monopolar needle electrodes. This may give values that are
(top to bottom). Initially (top set of superimposed traces), the po- similar to those using an SFEMG electrode. - - However, it is
24 64 213
tential shows a high jitter and long latency. At higher intensity important to recognize that the CN and MN electrodes have a
(second set), the latency and jitter are reduced. In the third set re- much larger uptake area. Thus, the spike component of their
cruitment of more fibers is seen, which have high jitter and MUAP is the summation of several muscle fiber APs. What ap-
blocking. Increasing intensity (fourth set) reduced their jitter and pears as a single spike may not necessarily be a single muscle
there is no block. These potentials have adequate stimulation and fiber potential, but a superimposition of two or more potentials.
they can be used for jitter analysis. Another fiber is seen near the This may lead to an underestimation of the jitter. Similarly, it is
end of sweep. It has high jitter and blocking due to suboptimal not possible to measure FD with these electrodes, since their
intensity and it should not be analyzed. In the last set, the in- recording surfaces are too large to resolve the MUAP into its
crease in intensity stabilizes the last potential and can be used for component APs. Therefore, we prefer to make these measure-
jitter measurements. However, jitter analysis is difficult for po- ments using the SFEMG needle only.
tentials in the center of the sweep where jitter appears increased
due to suboptimal stimulation of the newly recruited fibers. Relevance t o Routine EMG Studies
In stimulated SFEMG, some fibers may be excited when the SFEMG is a very sensitive procedure. It demonstrates in-
stimulation rate is increased (Fig. 8-61, potential 1). Their po- creased FD before fiber type grouping is seen on muscle
tentials may have high jitter and blocking similar to suboptimal biopsy. In disorders of neuromuscular transmission, increased
90
stimulation described earlier. jitter is seen even when there is no weakness or decrement to
The SFEMG electrode is quite expensive and requires care repetitive nerve stimulation in the tested muscle. However, in- 82
and maintenance. Without attention, the electrode impedance creased jitter and/or FD are seen in a number of neuromuscular
may become so high that it may not be possible to get good diseases, making SFEMG diagnostically nonspecific. The re-
recordings from the needle. The tip of the needle must be kept sults from SFEMG testing should be interpreted in light of in-
sharp to reduce discomfort and to avoid trauma to the muscle formation from the patient history, routine EMG, and nerve
fibers during recording. conduction studies. On the other hand, because of its sensitivity,
SFEMG recording is an art that is mastered only after weeks we can say that weakness in a muscle in which jitter is normal
or months of practice. Analysis of traces can be time-consuming cannot be due to abnormal neuromuscular transmission.
when the recording contains discharges of more than one MU. Increased FD is the earliest electrodiagnostic abnormality in
It is quite frustrating to spend an hour with the patient and a reinnervated muscle. Thus, increased FD measurements can
obtain only a few pairs for jitter analysis. There may not be suf- be used to demonstrate neurogenic involvement in muscles in
ficient reimbursement for the amount of time spent on analysis. which reinnervation has prevented development of weakness.
Reliance on automated measurements accounts for many of This may help in the study of motor neuron diseases, where
the errors in SFEMG analysis. The operator should be suffi- neuropathic abnormalities must be demonstrated in several
ciently experienced to identify blocking, and to estimate the muscle groups.
MG Class Mean M C D Fiber Density

no
inactive Ongoing
100 Reinnervation Reinnervation
90
) f-t ~ ~ ~ ~ ~ ~ -<>
«• 80
> 70
60
50
40
/ — M ^ ^ 30
20
-600 -400 -200 200 400 600 800 1000 1200
Days
Jitter
Class MG - * - Jitter (Mean MCD)
Figure 8-7/. The relationship between FD, jitter, and disease
Figure 8-70. Serial jitter studies. Serial measurements of jitter in progression.The shaded area indicates normal findings. See t e x t for
a pregnant woman shows a c o n c o m m i t a n t change in jitter (bottom details. (From Stalberg EV.Trontelj JV: Single Fiber Electromyography.
trace) with disease severity (top trace). (From Massey JM, Sanders DB: Old Woking, UK, Mirvalle Press, 1979, with permission.)
Single fiber electromyography in myasthenia gravis during pregnancy.
Muscle Nerve 1993; 16:458-460, with permission.)
Serial SFEMG studies may be helpful in assessing the response
to treatment, and will usually continue to show abnormal jitter
Serial FD measurements are also useful to demonstrate MU after RNS become normal. The biggest advantage to jitter mea-
remodeling after nerve i n j u r y . " During reinnervation, jitter
4199
surements is in the suspected MG patient with normal RNS test-
and blocking are increased in newly formed endplates. This ing. Because of its great sensitivity, SFEMG demonstrates
electrophysiologic observation indicates that MU remodeling increased jitter in some muscles in virtually all patients with
has not reached its maximum and thus further improvement is M G . Thus, normal jitter values in appropriately tested mus-
157
possible. When reinnervation is complete and the newly formed cles can be very helpful in excluding the disease. If jitter is
endplates are mature, the jitter returns to normal. This indicates normal in a weak muscle, the weakness is not due to abnormal
a status quo that may not be a good prognostic sign for patients neuromuscular transmission.
recovering from nerve injury. The combination of jitter and FD gives information about dis-
In patients with neuromuscular junction disorders, jitter cor- ease progression in patients with neurogenic diseases. In a
relates with the disease activity. When the disease is in remis- slowly progressing neuropathy the FD increases significantly
sion, jitter becomes less, and may even become normal in some while the jitter remains relatively normal (Fig. 8-71, top left).
muscles. Thus, jitter measurements can be used to assess the The slow progression allows the MU to complete reinnervation.
effect of treatment (Fig. SÔ). -" -" - -
95 3
SFEMG is also
5 154 157
This increases the MU size and hence the FD. This combination
useful in assessment of botulism. " 36 2
of findings can also indicate an old MU i n s u l t . - In con-
150155 212
When repetitive nerve stimulation (RNS) demonstrates a trast, a slightly increased FD with significantly increased jitter
decrement, SFEMG may not add much to electrodiagnosis. in a weak muscle indicates an active disease process with more
igure 8-72. Blanket principle. MUAP recording from

(A) normal subject and (C) patient with neuropathy.
Several sweeps are superimposed. Filter settings are
3-10000 Hz.The same MUAPs are shown in B and D, re-
spectively, after increasing low-pass filter frequency t o 500
Hz. N o t e t h e different calibrations for t h e potentials.The
MUAP amplitude and duration are reduced after filtering.
The notch in rising edge of MUAP in A is better resolved in
C.The MUAP variability is seen in D.
rapid progression (Fig. 8-71, bottom right). If both FD and jitter

are increased (Fig. 8-71, top right), this indicates acceleration or
Normal
recent onset of a slow disease p r o c e s s . 186187
As described earlier, raising the low-frequency setting of the

band pass filter to 500 Hz attenuates signals from distant fibers.
Such a filter setting in the routine EMG recordings enhances
any irregularities in the signal. The effect is similar to remov-
ing a blanket over a statue when it is unveiled. This concept is
called the b l a n k e t p r i n c i p l e (Fig. 8-72). Also, this reduces
143
the MUAP amplitude and duration, making them appear "myo-

pathic." A smooth MUAP waveform may be transformed into a
signal with several peaks representing signals from individual
muscle fiber. In normal MUs, the MUAP waveform remains C A D = 0.009
relatively constant during successive discharges. This is called C C C = 0.989
a s t a b l e MUAP. When the waveform varies significantly from
one discharge to the next, the MUAP is called u n s t a b l e (Figs.
8-6 and 8-72). These changes in the MUAP waveform repre- ALS
sent increased jitter and blocking of the muscle fiber potentials
contributing to the MUAP, the same phenomena that can be
seen more precisely with SFEMG. After assessing MUAP sta-
bility, the filter values should be returned to the default set-
tings, otherwise the signals will appear to have reduced
amplitude and duration.
When MUAPs appear unstable by visual assessment,
SFEMG also shows increased jitter and blocking. However,
jitter as seen on SFEMG may be masked when the jittering po-
tential is superimposed on normal fiber potentials. Also, as dis-
cussed above, several fibers may contribute to the APs seen with C A D = 0.770
MN and CN electrodes, and it may not be valid to measure their C C C = 0.677
time variability as in SFEMG. As described previously, the
CNEMG or monopolar EMG, even with filtering, cannot be Figure 8-73. Jiggle. In t h e top recording, the MUAP waveform has
used to quantify the FD values. minimal variability and a jiggle of 0.009. For the recording in a patient
Stalberg and S o n o o have proposed an alternative parame-
189
with ALS (bottom), the variation in waveform is obvious.The jiggle is
ter, called jiggle, to measure this instability of the entire MUAP 0.77. (From Stalberg E, Sonoo M: Assessment of variability in the shape
waveform (Fig. 8-73). A 2.5-msec epoch centered on the peak of t h e m o t o r unit action potential, t h e "jiggle" at consecutive dis-
of the MUAP is selected for analysis. This contains mainly the charges. Muscle Nerve 1994; 17:1 135-1 144, with permission.)
spike component of the MUAP. The amplitude change in suc-
cessive discharges of the MUAP at each MUAP sample point is digitization, as seen in the bottom traces of Figure 8-75. This is
measured. The median value of this measurement is computed. more likely to occur at the most rapidly changing portion of the
The area under the waveform generated by connecting these signals, i.e., at the peak of the waveforms with a low rise time.
median values is divided by the area of the averaged MUAP to
quantify the jiggle, called c o n s e c u t i v e a m p l i t u d e d i f f e r e n c e
(CAD).
In some MUAPs, the instability is recognized by the variabil-
ity of MUAP amplitude. This is seen easily at a slow sweep
speed (Fig. 8-74). This was one of the earliest EMG abnormali-
ties described in patients with MG. However, on some digital
86
100 ms
electromyographs, amplitude variability may also be seen in
very sharp normal MUAPs. This occurs from so-called aliasing,
an artifact of digital s i g n a l p r o c e s s i n g , produced as described
below.
A digital instrument measures the instantaneous voltage of
EMG signals at regular time intervals (Fig. 8-75). Each mea-
surement is called a s a m p l e and the time between successive
samples is the s a m p l i n g i n t e r v a l . The reciprocal value of the
sampling interval is the s a m p l i n g rate. The samples are then
plotted in the same sequence as on acquisition, and are con-
nected with straight lines. The resulting waveform is the digital
representation of the signal. In Figure 8-75, there is a good
match between the digitized signal and the analog waveform in Figure 8-74. Jiggle. In A, one second EMG recording from a patient
the top trace. If the sampling rate is reduced, the system may with neuropathy shows significant variability in its peak t o peak ampli-
fail to acquire adequate samples from the rapidly changing tude. Using a triggered delay line (B) and a faster sweep confirms t h e
components. This gives a distorted view of the EMG signal after variability of the MUAP waveform (C).
acquisition and proper display of these samples due to inherent

hardware or software deficiencies, aliasing will occur, which
can lead to misinterpretation. Unfortunately, the average user
usually does not understand the system architecture well
enough to distinguish this artifact from pathology.
Inadequate sampling can be demonstrated quite easily. In a
normal subject, record a very sharp MUAP with rise time of just
a few hundred microseconds. Change the sweep speed from the
conventional setting of 10 ms/div to 100 ms/div. If the MUAP
amplitude appears to change minimally, there is no aliasing in
the EMG system (Fig. 8-76A). Variable amplitude, as seen in
Figure 8-76B, indicates aliasing. If the EMG system does have
aliasing errors, use faster sweeps (e.g., 2 - 3 ms/div) and an am-
plitude trigger delay line to assess the stability. On many sys-
tems, the indicated change in sweep speed increases the
sampling rate and thus reduces aliasing.
Figure 8-75. Analog to digital converison.This concept is illus- SURFACE EMG AND MU NUMBER ESTIMATION
trated schematically. In solid line indicates t h e analog signal.The circles
are individual sample points.The sampling rate is reduced by 50% in Surface E M G recordings are routinely performed as part of the
the bottom trace.The digital signal indicated by dotted line is quite dif- nerve conduction studies. The compound muscle action potential
ferent from t h e analog signal. (CMAP) amplitude is assessed for abnormalities of conduction, e.g.,
conduction block, and also for muscle wasting and atrophy.
Additionally, surface EMG recordings offer a noninvasive method
The peak of the MUAP appears variably among different dis- of studying muscle fatigue. Multichannel recordings are useful for
charges and looks like a MUAP with varying amplitude. assessing movement disorders and characterizing tremor. Other
This problem can be resolved by increasing the sampling than the nerve conduction studies, surface EMG recordings are not
rate. The Nyquist theorem specifies that the sampling rate used in routine electrodiagnostic medicine evaluation. There is a
should be at least twice the maximum frequency in the signal." great deal of interest in developing techniques to monitor and follow
The maximum frequency in needle E M G recordings is at least disease progression using surface recordings of EMG signals.
10 kHz. Thus the sampling rate should be at least 20 kHz, One such application of surface EMG is the technique of I
preferably more than 40 kHz. If the instrument cannot support motor unit number estimation (MUNE), also known as motor
"Superficial" (<10mm) "Deep" (>20mm)
5 ms
Sweep Duration = 1 second

Figure 8-77. MU depth and SMUAP amplitude. SMUAPs
recorded at minimal force of contraction from the biceps brachii
Figure 8-76. Aliasing or jiggle? In A, a normal EMG signal containing a muscle of a normal subject as described in Figure 8-83. W h e n t h e in-
sharp MUAP was sampled at 50 kHz. In B, the same signal was sampled at tramuscular triggering needle tip was within 10 mm of skin surface,
3 kHz, and shows amplitude variability due t o aliasing. In C, the recording the MU was identified as being superficial. Similarly, when t h e tip was
from a patient shows amplitude variability.This is seen when the sampling more than 20 mm from the skin surface, the MU was considered deep.
rate is 50 kHz and hence it is not a technical artifact. Without knowing the O b s e r v e t h e differences in the SMUAP amplitude for t h e s e MU
sampling rate, we cannot tell if this reflects pathology or aliasing. (From groups. (From Barkhaus PE, Nandedkar SD: Recording characteristics
Barkhaus PE, Nandedkar SD: In Nandedkar SD (ed): EMG on CD. CASA of t h e surface EMG electrodes. Muscle N e r v e 1994; 17; 1317—1323,
Engineering, Hopewell Junction.Volume II, 1999, with permission.) with permission.)
unit counting. Loss of MUs is an important disease process in When the skin and subcutaneous tissue is thick, the distance
patients with neuropathy. Information about the number of MUs between the electrode and the MU is increased. This yields a
in a tested muscle would be very useful in assessing the severity smaller amplitude for the SMUAPs and the corresponding
and progression of the disease process. CMAP (Fig. 8-78A). Changes in skin and subcutaneous tissue
thickness, e.g., after weight loss or gain, also affect surface
Recording Characteristics of Surface EMG Electrodes EMG measurements.
Because of its placement, the surface electrode is roughly In normal subjects there is no direct correlation between the
equidistant from all fibers in the MU. Hence all muscle fibers biceps brachii muscle strength and its CMAP amplitude (Fig. 8-
contribute similarly to the s u r f a c e - r e c o r d e d MUAP (SMUAP), 78B). This is expected since the uptake area of the 10-mm disc is
making this a relatively nonselective recording technique. The not large enough to record from the entire biceps brachii muscle.
SMUAP amplitude and area should reflect the MU size. The electrode records from roughly the same number of genera-
However, if two MUs of the same size are located at different tors among different subjects, regardless of total muscle bulk.
depths from surface, the deeper MU, being farther from the elec- In patients, one can use the CMAP for the tested muscle as a
trode, has a smaller MUAP (Fig. 8-77). Barkhaus and Nandedkar 9
"reference" value. Changes in CMAP size reflect the progres-
estimate that the 10-mm disc electrode used in nerve conduction sion of certain diseases. In myopathy, they reflect mainly the
studies records mainly from muscle fibers that lie within 2 cm of loss of individual muscle fibers within MUs and their atrophy.
the electrode. This recording territory is large enough to contain In neuropathy, the reduced CMAP results mainly from the loss
most muscle fibers in the small distal muscles used in nerve con- of MUs. Hypertrophy of muscle fibers results in a higher
duction studies, e.g., APB, ADM, and EDB. However, this CMAP amplitude.
recording territory contains only a moderate portion of the large
limb muscles, such as the biceps brachii. To record from such MUNE Technique
muscles one should use a recording strip oriented perpendicular MUNE is a three-step process. First, the CMAP is recorded
to muscle fibers, which covers most of the muscle. Such special from the tested muscle using standard nerve conduction tech-
electrodes have been used in MUNE techniques. niques. The CMAP is the sum of the SMUAPs of all MUs in the
muscle. Hence the amplitude (or the area) of the CMAP repre-
(A) C M A P v s Skin Thickness sents the "electrical size" of the whole muscle. In the second
step, the electrical size of individual MUs is estimated by direct
C M A P (mV)
25,0
or indirect assessment of their individual SMUAP amplitude (or
area). Techniques that have been developed for this assessment
will be discussed shortly. Finally, the C M A P amplitude (or
area) is divided by the average SMUAP amplitude (or area) to
estimate the number of MUs in the tested muscle, by the follow-
ing formula:
MUNE = CMAP amplitude (or area) / average SMUAP ampli-
tude (or area)
We review the technique and pitfalls of different methods
used to estimate the SMUAPs and their amplitude.
0.0
0.0 SKIN_THICK (mm) 20.0
(B) CMAP vs Resistance
C M A P (mV)
25.0 | ' ' ' ' ' ' ' ' '
[100 pV
2 ms
RESISTANCE (Kg)
Figure 8-78. Factors affecting CMAP. In 13 normal subjects, 10-

mm disc electrodes w e r e used t o record CMAPs from t h e biceps Figure 8-79. AN-or-none response. Signals recorded from t h e
muscle. The CMAP amplitude was less when t h e skin thickness was ADM muscle as t h e ulnar nerve is stimulated at t h e w r i s t T h e stimulus
greater (A), while it was u n c o r r e c t e d t o the muscle strength assessed intensity was minimal, and was varied slightly during t h e 10 trials dis-
as maximum voluntary force (B). (From Barkhaus PE, Nandedkar SD: played in t h e superimposed m o d e . T h e r e s p o n s e demonstrates t w o
Recording characteristics of t h e surface EMG electrodes. Muscle states, baseline o r a fixed waveform. N o response of intermediate a m -
Nerve 1994; 17; 1317-1323, with permission.) plitude is seen.The waveform represents a single SMUAP.
A B Fig. 8-80). When these criteria are met, the signal is considered
to be an SMUAP. We will call this response "step 1."
! r\
After establishing the first MUAP, the intensity is increased
so
further. The response remains unaffected by the intensity until
m , , f \ , ri ,-rr«r the second axon is stimulated. Once again, slight changes in the
S1
r > p 7- . .V—: . . intensity reveal only three distinct waveforms: baseline (step 0),
® [100'MV
the first SMUAP (step 1), and the third signal, which is consid-
S2
2 ms ered to be the sum of two SMUAPs (step 2) (Fig. 8-80, response
S2). Since no intermediate configurations of the waveform are
S3 recorded, this is a stepwise change. The process can be repeated
^ fs [100 MV to get additional steps in the response, resulting from the pro-
/ \ 2 ms gressive stimulation of additional axons (Fig. 8-80, responses
S4
[2 mV S3 and S4). Dividing the final response amplitude by the
2 ms number of steps (i.e., the number of MUs) gives the average
SMUAP amplitude. This value is used to calculate the MUNE.
c
Each incremental step theoretically represents stimulation of
Figure 8-80. Incremental stimulation technique.The m o t o r one additional axon. Individual SMUAP waveforms can be ob-
nerve conduction program was used t o record evoked responses tained by subtracting the waveforms produced at successive
from t h e APB muscle as the median nerve was stimulated at the wrist. step. For example, subtracting step 0 from step 1 gives the
A stepwise change in t h e response was observed as t h e stimulus in- SMUAP of the first stimulated MU. Subtracting step 1 from
tensity was gradually increased. In A, t w o trials of each step are shown step 2 gives the SMUAP of the second stimulated axon, and so
superimposed.The average of these t w o trials from each step is shown on (Fig. 8-81). It is interesting to note that the amplitude of the
in superimposed fashion in B. N o t e t h e very slight change between summated SMUAPs (Trace S4 in Figure 8-81 A) thus obtained
the responses when t h e intensity was increased.The peak-to-peak am- is smaller than the sum of amplitudes of the individual
plitude a t step 4 was 177 pV, which gives mean SMUAP amplitude of SMUAPs (Traces M1-M4 in Figure 8-81-B). This results from
41.25 uV. In C, t h e CMAP at supramaximal stimulation has peak-to- the phase cancellation among individual SMUAPs as they are
peak amplitude of 13.6 mV.This gives an MU count of 307. summated. Some describe this as "lack of algebraic summation
of SMUAP amplitude in the CMAP." Using the algebraic mean
Incremental Stimulation. McComas and coworkers - 76 l,7 n 9 J 2 0
of SMUAP amplitude rather than the sum will give a smaller es-
pioneered MUNE by describing the technique of incremental timate of the MU count. 47
stimulation. In this method, the surface EMG is recorded in re- Although the incremental stimulation method is conceptually
sponse to nerve stimulation as the stimulus intensity is gradu- simple, it is technically quite demanding. The changes in the re-
ally increased. Initially, the intensity is too low to stimulate any sponse may be so small that they are not easily detected. This is
nerve fibers and the signal contains no MU activity (Fig. 8-79). more likely to occur when MU size is reduced, as in myopathy.
We will refer to this signal as "step 0 " (response SO in Fig. 8¬ Failure to recognize small SMUAPs gives an incorrectly higher
80) and used it as the baseline. The intensity is increased until calculation of SMUAP amplitude, producing a lower MU count. 142
the EMG contains a waveform that is time-locked to the stimu- If axons A and B have very similar stimulation thresholds, we
lus and appears in an all-or-none fashion as the stimulus inten- can observe so-called alternation in responses. Sometimes we
sity is slightly reduced or increased (Fig. 8-79, response SI in will record SMUAP from axon A, sometimes from axon B, and
Figure 8-81. SMUAP extraction from stimulated responses.

A, Stepwise changes in the evoked response from the APB muscle in Figure 8-82. Multi-point stimulation. SMUAPs recorded from a
Figure 8-80A are shown. Subtracting successive responses gives t h e normal subject (A) and patient with m o t o r neuron disease are shown
SMUAP waveform shown in B. Note t h e small SMUAP amplitude.The (B).The amplitude calibration is 100 uV/div in A and 200 pV/div for
sum of amplitudes of these four SMUAPs was 254 pV. In contrast, t h e the patient. N o t e t h e higher SMUAP amplitude for the patient. (From
amplitude of their summated waveform (bottom traces in A) was 177 Lomen-Hoerth C, Olney R: Comparison of multiple point and statisti-
MV.This demonstrates that SMUAP amplitudes do not summate alge- cal motor unit number estimation. Muscle Nerve 2000;23:1525-1533,
braically due t o phase cancellation. with permission.)
Figure 8-83. Surface MUAP recording technique.The free-

running needle EMG (A) and surface EMG (B) are shown at slow
sweep speed.The discharges of o n e MU can be recognized as sharp
Figure 8-84. SMUAP with different triggers. In A, three intra-
spikes in the needle recordings. These are used with t h e triggered
muscular needle MUAP recordings from the biceps muscle of a normal
delay line (C) t o identify MUs for analysis.The time-locked activity on
subject are shown.The filter settings were 500-10000 Hz t o facilitate
the surface EMG is averaged t o produce the SMUAP (D).
triggering. Concurrent SMUAP recordings are shown in B.The active
electrode was over t h e endplate zone. Although t h e needle recorded
MUAPs appear different, t h e SMUAP has the same shape.This indicates
sometimes both axons will be stimulated, giving the sum of the
that all recordings w e r e made from the same MU with different posi-
two SMUAPs. On visual assessment, this will appear as three
tions of the triggering electrode within the MU territory.
stepwise changes in the response, implying stimulation of three
axons instead of two. The SMUAP amplitude will thus be un-
derestimated, yielding a higher MU count. Ballantyne and SMUAP recording using an intramuscular trigger can be per-
Hansen developed an SMUAP identification algorithm in con-
7
formed in most muscles and does not require special hardware
| junction with incremental stimulation to avoid this problem. or software. However, it is an invasive procedure that is some-
The sophisticated software to detect these small SMUAPs and what uncomfortable. It also does not deal with phase cancella-
to identify alternation is not available in most commercially tion among MUAPs. With multi-MUAP algorithms, it may be
available systems. possible to record MUAPs with surface and intramuscular
Multipoint Stimulation. To overcome the technical prob- needle electrodes from many MUs in just a few minutes, which
lems with incremental stimulation method, one may record only would facilitate the MU counting procedure.
the first stimulated MU from the stimulation site. Additional When large muscles are used for MUNE, bias in selecting
SMUAPs are obtained by stimulating the nerve at different sites MUs can affect the estimates. As discussed earlier, superficial
along its length, giving the name multipoint stimulation MUs tend to have higher-amplitude SMUAPs. A bias towards
(MPS) (Fig. 8-82). - This technique permits eliciting individ-
46 69
selecting superficial MUs gives higher SMUAP amplitudes and
ual SMUAPs using hardware available in most commercial smaller MU counts. If MUs are selected mainly from the deep
EMG systems. However, it may be tedious to identify the portion of the muscle, the MU estimate is higher. 9
number of stimulation sites necessary to obtain a large sample Statistical Modeling. When the stimulus intensity is sub-
of SMUAPs. It is more practical to obtain three or four maximal, it may be close to the excitation threshold for some
SMUAPs for each stimulation site. This may be automated axons. As a result, they may or may not be generate a SMUAP
using template matching algorithms that can also detect the al- for each stimulus. This gives a variable shape to the responses
ternation among the excited MUs. - 167 210
elicited by successive stimuli. By modeling the number of ex-
As indicated earlier, the SMUAP amplitudes do not summate cited axons as a Poisson statistical process, the average SMUAP
algebraically. To allow for phase cancellation, the individual
SMUAPs are aligned at their onset point and then summated.
The amplitude of the resulting potential is divided by the
number of SMUAPs to obtain the average SMUAP amplitude. 47
Needle EMG Spike-Triggered Averaging. The surface

recorded EMG signals from individual MUs have low ampli-
tude. This makes it difficult to recognize and quantify SMUAPs
from the free-running EMG. A two-channel recording is neces-
sary, using a needle electrode to identify MU discharges and to
extract the SMUAP (Fig. 8-83). This is called spike-triggered
averaging (STA). - 23 130
The position of the needle is then changed to record from an-

other MU. If the needle selects the same MU, the needle EMG Figure 8-85. MU number estimation by STA. In A, the CMAP
signals may look different but the SMUAP is unchanged. This was recorded from the EDB muscle as the peroneal nerve was stimu-
allows one to discard duplicate recordings from one SMUAP lated supramaximally at t h e ankle (5 mV/div). In B, SMUAP recordings
(Fig. 8-84). Ten or more SMUAPs are thus extracted and their from t h e same muscle are shown (50 pV/div).The peak-to-peak CMAP
mean amplitude is calculated (Fig. 8-85). Reproducibility of the amplitude is 11.2 mV.The average SMUAP amplitude is 67 uV.This
MUNE is reduced if fewer SMUAPs are extracted. 20
gives a count of 167 MUs. (Courtesy Dr. Barkhaus, Milwaukee,WI).
amplitude changes significantly with slight changes in intensity

is selected for stimulation. This yields many different configu-
rations of the response, making it possible to study the statisti-
cal nature of excitation. The step sizes reflect the MU size in the
tested muscle. In patients with ALS, stimulation of a large MU
significantly increases the response area. This can generate very
large steps, sometimes covering more than 20% of the CMAP
area (Fig. 8-87). In general, one chooses intensity levels that
cover 10% of the CMAP area (Fig. 8-87).
The method of estimation is illustrated using the recording in
Figure 8-88. The stimulus intensity was selected to test the
CMAP Amplitude = 1 5.3 mV 5 5 - 7 5 % area range. All testing at this range is called a run.
Within a run, stimuli are delivered in groups of 30. The numer-
ical data from four such groups in the third run is shown in the
Scan R r e a s s
bottom right corner of Figure 8-88. The responses from the
fourth group are indicated to the top left. The distribution of
area from all four groups is indicated in a histogram fashion in
the top right plot. For each group, the average SMUAP ampli-
tude is estimated using the aforementioned formula. In this ex-
ample, the SMUAP amplitude estimate varied from 73 to 95 pV
for the four groups. Dividing the SMUAP amplitude into the
Note t h e rel a t i v e CMAP amplitude yields the MU count, as indicated in the ex-
u n i f o r m i t y of gaps treme right column of the table. A run is concluded when the
coefficient of variation of MU estimate (i.e., standard devia-
between s t e p s - -
tion/mean) from the tested groups (i.e., values in the right
w h i c h correspond t o column of the table) is low, or 10 groups are acquired. The mean
m o t o r unit size. SMUAP amplitude from these groups is divided into the CMAP
amplitude to estimate the number of MU. This is indicated in
the bottom-most line of the numerical result table in Figure 8¬
88 where the number of MUs is 33.
This was the original description of the method when it was
first presented at a scientific meeting. Since then the method
40
MUNE = 200 has been modified significantly. It is argued that the MU size
42
varies when different testing ranges are used. For the recording •
in Figure 8-88, the SMUAP amplitude is representative of the
55-75% range. This is only one fifth of the CMAP and thus ac-
Figure 8-86. Evoked response amplitude versus stimulus in- counts for only 7 motor units in the tested muscle. This is indi-
tensity. Responses t o 30 nerve stimuli of increasing intensity in the cated in the result table in the bottom left corner of Figure 8-88.
EDB muscle of a patient with early ALS are shown at the top.The re- Similar runs gave 4 MUs in the 1 8 - 3 5 % level and 9 MUs in
sponse area curve demonstrates a smooth sigmoid increase with stimu- 40-55% level. The three runs represent 52% of the CMAP area
lus intensity.The MUNE was 200. (Courtesy Dr. B. Smith, Scottsdale.AZ). and it is attributed to 20 MUs (penultimate line of the table).
This left 4 8 % of the sigmoidal curve (Figs. 8-86 and 8-87) for
amplitude can be estimated from the mean and variance of the testing. One could test additional ranges or use extrapolation to
evoked response (ER) amplitude (or area). This technique,
40
predict the number of MUs in the untested range. In Figure 8¬
however, does not yield individual SMUAP waveforms. 88, the lowest SMUAP amplitude (45 pV) is used to estimate
MUs in the untested range. This yields additional 31 units (last
Variance of ER amplitude
Average SMUAP amplitude = line of the table). The total of the tested and untested MUs is 51.
Mean ER amplitude - Minimum ER amplitude
This is the value to be used for assessment of the test.
The stimulus intensity may be significantly above the thresh- Shefner and coworkers have argued that using the lowest
164
old level for some axons. SMUAPs from those axons will con- amplitude value may overestimate the number of MUs. Clearly,
tribute to the evoked response for all stimuli. Their contribution the SMUAP amplitude in other tested ranges is much higher.
is thus a nonvariant part of the measurement and affects the They have proposed a weighted average of SMUAPs from the
mean value but not the variance. To allow for this, the minimum tested runs to estimate the amplitude for the untested range.
SMUAP value of amplitude is subtracted from the mean in the This also reduced the variability of the estimate.
above equation. The statistical method based on the response variability at
In a typical study, the stimulus range for the minimal and submaximal stimulation offers simplicity to the recording tech-
maximal amplitude evoked response is determined. Then the in- nique. However, it is limited only to muscles that are easily
tensity is increased gradually in 30 uniform steps to cover the accessible for multiple stimulation. Abnormalities of neuromus-
stimulus range. A plot of the evoked response amplitude versus cular transmission also contribute to the variability of the re-
stimulus number (i.e., the intensity) has a smooth sigmoid shape sponse to repetitive stimulation. The variability also depends on
with relatively uniform steps (Fig. 8-86). In some studies, there the stimulation level. Reference values must be defined for dif-
may be a sudden stepwise increment in the response followed ferent stimulus ranges (Table 8-14). When a large number of
by a plateau (Fig. 8-87). The intensity level where the response MUs contribute to the response (as in normal muscles), the
Switch: STOP Acquire: Off Rate: 1 Hz Level: 0.0 mfi Dun 0 . 1 ms Trace: 29/30
1 mV 2 - S k Hz 2 ms
St i n :
Start 2 4 . 0 mR
Scan Areas: End : 5 4 . 1 mR
Last Rrfta: 8683 uVma.Rmp! 2 . 9 1 5 frtV

Max Rrea: 8610 u V m s . Rmp: 2 . 8 7 6 mV
TEfP:35.3°C
fill Runs: MUNE(testedMntested) =
Run N Level il SMPuV HUNE
i i i i i i i i i i i i iCurve
20 30
Figure 8-87. Evoked response amplitude versus stimulus intensiry in a patient with ALS.The evoked responses (top left) in the EDB
muscle of a patient with ALS show a stepwise change in t h e area "(right) Four ranges for MUNE are indicated by the dashed rectangles. Note t h e
large step size when the area changed from roughly 60% t o 80% of maximum by a minimal increase in stimulus intensity.The range used for t e s t -
ing falls at the center of such a step. (Courtesy Dr. B. Smith, Scottsdale,AZ).
Table 8-14. Reference Values for MUNE Using Poisson process approaches the Gaussian distribution. In this
Statistical Method case, the MU estimation may require many stimuli to reduce the
variance.
Tibial-
Median- Ulnar- Peroneal- Abductor O t h e r Techniques
Level Thenar Hypothenar EDB Hallucis
Shahani and coworkers recorded surface EMG signals
163
5-10% 210/90 285 / 105 154/52 3 1 0 / 195 during voluntary contraction. Simultaneously the EMG w a s
recorded on a second channel using another surface electrode
15-20% 185/85 223 / 110 137/45 2 5 0 / 167
placed on the tested muscle. The signals from the second chan-
40-50% 153/70 154/70 135/38 1 9 5 / 154 nel were processed to facilitate MU identification as in decom-
position or multi-MU analysis studies. MUs thus identified
70-90% 175/85 2 1 3 / 115 105/35 2 0 2 / 115 were used to trigger extraction of the SMUAP from the first
channel of EMG signals. By changing the position of the sur-
Multipoint 234 / 95 2 5 6 / 115 158/58 2 8 5 / 187 face electrode in the second channel (the triggering electrode),
The mean values in normal muscle are indicated, followed by the lower normal SMUAPs were obtained from different M U s in the tested
limit of normal. Note the difference in values based on the stimulation level.The muscle. The SMUAP waveforms and the CMAP were fed to a
results are also compared with the multipoint stimulation method. Note that
the lower limit is often less than 50% of the mean.
neural network algorithm. The program summated t h e
67
From Daube J: Estimating the number of motor units in a muscle. J Clin SMUAPs in different ways to ultimately synthesize the CMAP.
Neurophysiol 1995; 12:585-594, with permission. When this was achieved the number of SMUAPs required for
Switch: STOP Require: Off Rate: 1 Hz Level: 0.0 nfl Dun 0 . 1 ms Trace: 30/30
St i n :
Run 3 areas
III IIT
SG 55 eo *
Run 3: Area Test Range: 55 - 75 Si
Group flnp mV SMUPuV MUNE

Last Rr«a: 5559 uVms.'flmp: 1 . 6 3 3 mV
Max Area: 8610 u V m s . Rmp: 2 . 8 7 6 mV 1 1.7 73 39
. - „ . , . TEMP:30.0°C 2 1.7 86 33
51 Units 3 1.7 95 30
fill Runs. HUNE(testedtunlested) = 51 4 1.7 92 31
Rur N Level 1 SMUPuV MUNE 5
1 270 18- 35 124t11 4 6
2 120 40- 55 45t 3 9 7

O
3 120 55- 75 87! 5 7 0
4 9
510 52!! tested 85+23 20 10
NONE 48% untested 45 31 Rvg: 1.7 87t 5 33
Figure 8-88. MU e s t i m a t i o n b y t h e s t a t i s t i c a l m e t h o d . Refer t o text for details. (Courtesy Dr. B. Smith, Scottsdale, AZ).
synthesis represented the MU count. This technique has two

major advantages: (1) it is noninvasive, and (2) it attempts to ac-
count for phase cancellation. On the negative side, the analysis
requires special software that is commercially unavailable.
Secondly, the MU estimate may depend upon the underlying
strategy in the neural network.
Stashuk and coworkers stimulated the nerve at submax-
196
imal intensity and recorded the evoked response with sur-

face electrodes to identify repeating F-waves (Fig. 8-89).
These were considered to represent SMUAPs from single
MUs.
In another method described by Stein and coworkers, the 197
mechanical output of the muscle was used to estimate the MU

size. The twitch tension in response to supramaximal stimula-
tion was divided by the average twitch generated by individual
MUs. This approach requires EMG recordings to facilitate
study of individual MUs. Furthermore, force measurement re-
quires special hardware and software.
DeKoning and coworkers used the macro-EMG electrode to
44
record individual MUAPs. This intramuscular electrode was

Figure 8-89. F - w a v e a s S M U A P . F-wave recordings in a normal also used to record the CMAP. The ratio of CMAP to MUAP
subject show variable waveform and latency of t h e F-wave. O c - amplitude was called the MU density. Density measurements
casionally, two F-waves, indicated by t h e arrow, are identical. were made from different sites in the muscle.
Table 8-15. Motor Unit Estimates in Normal Muscles

Investigator Technique Muscle Subjects SMUAP MU Count
Wang et al. (1995) AMPS Thenar All 59 87 ± 28 278 ±113
19-39 24 353 ± 120
41-58 18 258 ± 64
60-87 17 194 ±74
F-wave 54 99 ± 26 253 ±107
Stashuk et al. (1994) MPS Ail 219 ±77
F-wave Thenar 33 ± 11 18 278 ± 103
68 ± 3 15 195 ±34
All 33 245 ± 105
Slawnych et al. (1996) MUESA Thenar <50 26 111 ± 56
Fang et al. (1997) Neural net Thenar 25-37 5 222 ± 98
McComas (1977) Incremental stim Thenar 115 342 ± 89
Stein e t a l . ( l 9 9 0 ) Spike-triggered average Thenar 10 108 ± 38 135 ± 2 7
Stein et al. (1990) Spike-triggered average with force Thenar 10 130 ± 3 9
Brown et al. (1988) Spike-triggered average Biceps Brachii <60 30 16 ± 7 911 ± 2 5 4
Brown et al. (1988) Spike-triggered average >60 10 23 ± 8 479 ± 220
McComas (1977) Incremental stim EDB 151 210 ± 6 5
Nandedkar et al. Spike-triggered average EDB 6 84
Ballantyne and Hansen Incremental stim EDB 39 197 ± 4 9
Slawnych et al (1996) MUESA EDB <50 21 83 ± 3 5
Dekoning Macro-EMG Tibialis anterior 21-40 12 43 ± 17
41-60 15 38 ± 14
61-80 9 21 ± 9
The findings from several different studies in muscles of normal subjects are tabulated. Note the significant variation in MUNE for the same muscle for the different
methods. Also, the lower normal limit (estimated as 2 standard deviations below mean) is less than half the mean value. EDB, extensor digitorum brevis.
•; Findings in Normal Subjects and Patients to monitor changes in the CMAP size. This is certainly useful in
Using the incremental stimulation technique, McComas and the acute phase. If reinnervation is unsuccessful, as in a rapidly
coworkers found a loss of MUs in older subjects. - This
76 H 7 | 2 !
progressive neuropathy, the CMAP should also be useful to
was attributed to the normal aging process. In neuropathy, the monitor MU loss. In a slowly progressing neuropathy, reinnerva-
MU count was reduced as expected. Those investigators also tion may be able to compensate for loss of MUs. This gives a
found a reduced MU count in dystrophy. This formed the basis normal CMAP size despite a reduced MU count. It is in such
for a "neurogenic hypothesis" for dystrophy. It was demon-
118
conditions that the MUNE could be of greater utility than CMAP
strated that the incremental stimulation method might miss amplitude measurements in assessing disease progression.
some small-amplitude SMUAPs. This gives higher amplitudes It is important to remember that a change in the skin and sub-
for individual SMUAPs and thus a reduced MU count in dystro- cutaneous tissue thickness between successive studies also af-
phy, which could be interpreted as neurogenic changes. fects the CMAP measurements. When recording the CMAP,
In the last two decades, the different techniques have been neighboring muscles are also activated. The CMAP contains the
used to study other normal m u s c l e s . The findings of some
119166
volume-conducted activity of these muscles. In contrast, indi-
studies are summarized in Table 8-15. All techniques have vidual MUAPs can be limited to the tested muscle. Thus, MU
demonstrated changes consistent with loss of MUs in normal counts may give overestimates. While all axons will be stimu-
older subjects and in patients with neuropathy. The diagnostic lated in normal subjects, demyelinated axons may require much
sensitivity of these technique has not been compared rigorously higher intensity. If these axons are not stimulated, a smaller
with other EMG parameters such as FD. CMAP and a lower MU count are obtained.
Clinical Interpretation. It is obvious that MU counting tech- MU loss can also be observed, though not quantified, with
niques are still in a developmental stage. The techniques can be other procedures. Changes in the F-wave morphology reflect re-
time-consuming, tedious, susceptible to errors, and require spe- organization of MU architecture due to disease processes. When
cialized hardware and software. Furthermore, the diagnostic sen- MUs are large due to reinnervation, F-waves may become com-
sitivity of this method has not been compared with other routine plex and have high amplitude. This change can also occur if the
test procedures. In many instances, the lower normal limit is less excitability of the motor neurons is increased so that F-waves
than half the mean value (see Table 8-15). This means that up to are generated in more MUs.
50% of MUs could be lost without reducing the MUNE below In nerve conduction studies, the probability that a given MU
normal limits. Despite these limitations, these techniques may be will generate an F-wave is small. However, because a large
useful in assessing disease progression in individual patients. number of MUs participate in the response there is a good
The MU count depends directly on the size of the CMAP. In chance that at least one of them will generate an F-wave at
patients with neuropathy, a simple way to follow the MU loss is supramaximal stimulation. In normal subjects, most stimuli
Table 8-16. Reference Values for Macro-EMG

Age Biceps Brachii Vastus Lateralis Tibialis Anterior
Median Individual Median Individual Median Individual
10-19 65-100 30-350 70-150 20-350 65-200 30-350
20-29 65-140 30-350 70-240 20-525 65-250 30-450
30-39 65-180 30-400 70-240 20-550 65-260 30-450
40-49 65-180 30-500 70-250 20-575 65-330 30-575
50-59 65-180 30-500 70-260 20-575 65-375 40-700
60-69 65-250 30-650 80-370 20-1250 120-375 45-700
70-79 65-250 30-650 90-600 20-1250 120-620 65-500
The range of median macro MUAP amplitude and individual MUAP amplitude is described for normal muscles. Note the increase in amplitude with age.
From Stalberg E, Fawcett PRW: Macro EMG changes in healthy subjects of different ages.J Neurol Neurosurg Psychiatry 1982;45:870-878, with permission.
generate an F-wave. When the number of MUs is reduced, it be- MUNE and Disease Progression
comes more likely that none of them will generate an F-wave. Despite all of the limitations and differences among different
This is seen as a reduced percentage of F-responses. It is neces- methods, the MUNE is a parameter of great value in assessment
sary to deliver a large number of stimuli to demonstrate this re- of progressive degenerative diseases, e.g., amyotrophic lateral
duction. If the excitability of the neurons is reduced, the F-wave sclerosis (ALS). Many studies have demonstrated changes in
22
persistence (i.e., percentage of responses with F-waves when a CMAP, MUNE, grip strength, macro-EMG, and fiber density
large number of stimuli are delivered) also decreases. with disease progression. - - ' - Among different studies, the
5 6 20 70 220
Loss of MU is also reflected by presence of spontaneous ac- MUNE was found to be most suitable in quantifying disease
tivity, abnormal waveform of MUAPs, and incomplete or dis- progression. Yuen and Olney also found fiber density changes
220
crete IP at maximal effort. useful when the disease progression was relatively slow. Yet
there is a lot still to accomplish, e.g., the intrasubject variability
needs to be reduced. Fortunately, all techniques demonstrate
greater reproducibility when the MU count is reduced, i.e., in
pathologic situations. In normal subjects, the estimates have
much higher variability. • ' • ' • As many methods are
5 ,9 69 1,0 ,4S ,68
available only as proprietary software, there may be method-

ologic differences that are not adequately described that may
cause further confusion. Nevertheless, this is an exciting devel-
opment in the electrodiagnostic techniques.
MACRO-EMG
Recognizing the dependence of SMUAP amplitude on the
MU location within the muscle, Stalberg developed the tech- 180
nique of macro-EMG to assess the MU size.

Technique and Measurements. The recording electrode is a
modified SFEMG needle, with a cannula insulated except for the
distal 15 mm. The active recording port of the SFEMG electrode
Figure 8-90. Macro-EMG. A schematic cross section of a MU with

a macro EMG e l e c t r o d e superimposed, and t h e recording montage
(A). N o t e that the cannula passes through t h e MU territory. SFEMG
signals are shown in free-running mode (B).The discharges of a single
MU are easily recognized. In C the macro-EMG signals are shown.The
macro-MUAP (D) can be recognized in C . (From Stalberg E: Macro Figure 8-91. Macro-EMG MUAPs. Recordings from t h e biceps
EMG, a n e w recording technique. J N e u r o l Neurosurg Psychiatry brachii muscle of (A) patient with myopathy, (B) normal subject, and
1980:43:475^82, with permission.) (C) patient with neuropathy are shown.
exits at the center of the bare distal shaft opposite the bevel. Two
channels of EMG activity are recorded (Fig. 8-90A). SFEMG sig-
nals are recorded on the first channel as described earlier. For the
second channel, the cannula is used as the active recording surface
700îv
and a remote surface or monopolar needle is used as reference.
The SFEMG signal is used to identify individual MU activity (Fig.
8-90B). This is used to trigger the sweep and the averager. The sig- 10 roe
300 îv
nals from the second channel (Fig. 8-90C) are delayed and aver-
aged to extract the macro-EMG MUAP (Fig. 8-90D).
Twenty or more MUAPs are recorded from different sites in
the tested muscle. The shape of the Macro MUAP remains rela-
tively constant even when the needle position changes within
the MU territory. Thus, duplicate recordings can be identified
and excluded from analysis. For each MUAP, the amplitude and
area under the rectified waveform are measured. Due to the
skewed distribution of these parameters, the median instead of
mean value is used as the measure of central tendency. Normal
values are defined for the upper and lower limits of the median
as well as individual MUAP measurements.
Findings in Normal Subjects and Patients. In normal sub-
jects the MUAP amplitude increases with age, especially after
the sixth decade. The limits vary among different normal mus-
cles (Table 8-16).
In neuropathy, the macro-MUAP amplitude is increased (Fig.
8-91). Serial studies demonstrate a reduction in amplitude when
the muscle becomes severely weak (Fig. 8-95, 8-96). In myopa-
thy, the macro-MUAP amplitude is normal or reduced (Fig. 8¬
91, 8-98). Serial studies demonstrate progressively reduced
amplitude as the weakness progresses. Figure 8-92. Conmac recordings from patients. A, Patient with
Interpretation. The macro-EMG active recording surface is myopathy; B, patient with neuropathy. For both patients, trace I is conmac
large compared to CN and MN electrode recording surfaces, potential, trace 2 is the routine concentric MUARThe calibrations for t h e
making it nonselective. Thus, when the position of the electrode two patients are indicated separately. (From Gan R,Jabre JF:The spectrum
changes, the MUAP waveform remains relatively constant. This of concentric macro correlations. Part II. Patients with diseases of muscle
has been demonstrated by computer simulations and also by
126
and nerve. Muscle Nerve 1992; 15:1085-1088, with permission.)
experimental recordings. The macro-MUAP of high-force
180
threshold MUs has larger amplitudes than low-force threshold or only slightly reduced macro-MUAP size (Fig. 8-91, 8-98).
MUs. 98181
This also demonstrates that the macro-MUAP reflects These compensatory processes will also be manifest as an in-
the MU size. By virtue of its construction, the electrode passes creased FD on S F E M G . In inclusion body myositis, serial
92
through most of the MU territory (Fig. 8-90). It is close to the studies demonstrate reduced macro-MUAP amplitudes as weak-
muscle fibers of the MU. Therefore, the macro-MUAP stays rel- ness progresses. This reflects progressive loss of muscle fibers
atively constant at different recording positions. In contrast, the and changes in fiber size (Barkhaus, personal communication).
surface MUAP depends on the MU position from the skin sur-
face as well as thickness of the skin. Therefore, the macro-
MUAP is a better estimator of MU size than the surface MUAP.
In neuropathy, the macro-MUAP is larger than in controls, in-
dicating increased MU size due to reinnervation (Fig. 8-91).
The highest values of amplitude are recorded in patients with
slowly progressing disease where reinnervation can compensate
the MU loss. - - - - Serial studies indicate that the MU
85 111 150 186 193
size may increase as much as 20-fold as reinnervation proceeds

to its maximum (Fig. 8-94). Thereafter, the macro-MUAP am-
plitude decreases, which coincides with a new onset of weak-
ness (Fig. 8-95, 8-96). In rapidly progressing degenerative
disease, MUs are lost before they can reach their maximal rein-
nervation capacity. Therefore, the amplitude may show only a
modest increase. Many MUAPs may have normal amplitude
183
although the muscle is moderately or severely weak.

Recall that the primary pathology in myopathies is loss of
muscle fibers. New muscle fibers may be generated from satel-
lite cells or by fiber splitting; additionally, some muscle fibers
become hypertrophic. These changes in the MU may partially
compensate for fiber loss in myopathy. Hence the MU size may Figure 8-93. Conmac position. Two e x t r e m e positions of t h e
be normal or only slightly reduced, which is reflected by normal conmac electrode in an MU are illustrated schematically.
% isokin. strength,knee ext polio, Stalberg and colleagues reassessed the macro-MUAP am-
plitude to account for hypertrophy of muscle fibers. - They con-
85 192
cluded that macro-MUAPs that are 10 times normal size imply

that 20% of the original MUs have survived. The combination of
MU size, percentage of surviving MUs, and the degree of muscle
weakness gives information that is useful in managing patients
with old polio (Fig. 8-94). For example, a weak muscle with
mildly enlarged MUAPs (relative Macro amplitude < 4 {Fig. 8¬
94}, which corresponds to 25-50% surviving MUs, depending on
degree of fiber hypertrophy) indicates a modest pool of surviving
MUs. Thus, strengthening exercises may be of benefit. In contrast,
a weak muscle with very large MUAPs (relative Macro amplitude
> 10 {Fig. 8-94}, corresponding to fewer than 20% surviving
rel Macro amplitude
MUs) indicates few surviving MUs, which are functioning almost
at their peak capacity. These muscles should not be burdened with
Figure 8-94. Macro-EMG in polio.The relative strength of knee
additional tasks. Studies on patients with old polio indicate that
extensors plotted against relative macro-MUAP amplitude in t h e tib-
the muscles begin to weaken when the Macro MUAP size be-
ialis anterior muscle. In muscles with weakness (data points below the
comes 20 times normal. This weakness could result from further
horizontal line), t h e r e is a wide range of macro amplitude measure-
loss or fractionation of MUs, as described later.
ments.This reflects different degrees of MU loss and reinnervation.
The Conmac Electrode. Macro-EMG is the only technique P U T T I N G IT TOGETHER: Q U A N T I T A T I V E

that truly gives information about the MU size. In some condi- ANALYSIS A N D A M O D E L FOR DISEASE
tions, other EMG techniques may demonstrate mixed features. PROCESSES
In such cases, macro-EMG could be useful in demonstrating
changes in MU size. The macro-EMG electrode is rather expen- So far we have looked at the quantitative techniques and fo-
sive, it is not disposable, and requires special care and mainte- cused on information each provides about specific aspects of
nance as well as experience with the technique.
Jabre - developed a two-channel recording technique to
96 97
2000
measure signals similar to the macro-EMG with a CN electrode.
The first channel records the CN EMG signals between the cen-
tral wire and the cannula. The second channel records the activ-
ity from the cannula with reference to a remote reference
electrode. Because the cannulas are similar in size, these record-
ings give similar information to macro-EMG (Fig. 8-92), hence
the name conmac for the technique. In Figure 8-92, concentric
and conmac recordings from two different MUs are shown. 77
The top two traces are from a patient with myopathy, the bottom
traces from a patient with neuropathy. The concentric needle
recordings (A2 and B2) have complex waveform with long du-
ration, satellite components, etc. This is a nonspecific finding
and cannot differentiate between a myopathy and neuropathy.
However, the conmac signal shows very low amplitude for my-
opathy (trace A l ) and increased amplitude in neuropathy (trace
B l ) . Thus, conmac is able to differentiate pathology when the
conventional recordings demonstrate nonspecific findings.
If the needle passes through the MU, the cannula is inside the
MU territory as in macro-EMG, and the macro and conmac
MUAPs should be similar (Fig. 8-90A, 8-93A). However, if the
needle tip is at the edge of the MU territory, the entire cannula is
outside the MU territory (Fig. 8-93B). This gives a low-amplitude
cannula potential. Due to these differences, one expects a slight
difference between the amplitudes of macro-EMG potentials and Mild Moderate Severe
the conmac recordings. A reduced conmac MUAP may not
140
Degree of weakness
necessarily come from a small MU. The probability of such elec-
trode placement (Fig. 8-93B) is expected to be small. On the Figure 8-95. Quantitative EMG in polio.The relative change in
other hand, increased conmac MUAP amplitude would be consis- nFD, and C N MUAP, and macro-MUAP amplitude in muscles with dif-
tent with increased MU size regardless of the electrode position. ferent degrees of weakness from old polio is shown. (From Sanders
Macro-EMG and MU Number Estimation. Since the DB, Massey JM, e t al: Quantitative electromyography after p o -
macro-MUAP amplitude reflects the MU size, it can be used to liomyelitis. In Halstead LS, Wiechers D O (eds): Research and Clinical
assess loss of MUs and reinnervation. If the macro-MUAP am- Aspects of t h e Late Effects of Poliomyelitis. Birth Defects: Original
plitude doubles, it would imply that 5 0 % of MUs have been lost Article Series, March of Dimes. W h i t e Plains, N e w York, 1986, pp
and all muscle fibers have been reinnervated. In patients with 189-200, with permission.)
the MU. When the information obtained from these procedures

is combined, a much better picture of normal and abnormal
MUs is obtained. - -
18 1 185
Sanders and c o w o r k e r s examined F D , CN MUAP, and

155
macro-EMG MUAPs measurements in a group of patients with

old polio. The neighboring fiber density (nFD) is obtained by
subtracting 1 from the FD value. It represents the fibers other
78
than triggering potential that lie within the uptake area of the
electrode. The nFD, mean CN MUAP amplitude, and mean
macro-MUAP amplitude measurements were expressed as per-
centage of their corresponding normal values. The data were
divided into three groups based on strength in the tested
muscle. The mean values of the normalized measurements
from these three groups are shown in Figure 8-95. In muscles Duration (Months )
with mild weakness, the MU size (based on macro-MUAP am-
Figure 8-96. Quantitative EMG in a patient with ALS.The rel-
plitude) is slightly increased. The MU size is maximal in the
ative change in muscle strength, nFD, and C N and macro-MUAP ampli-
moderately weak muscles. In severely weak muscles, the MU
tude in the biceps muscle of a patient with ALS over a 2-year period as
size is slightly increased from normal, but is much less than in
the muscle became weaker. N o t e t h e increase and then decrease in
the moderately weak muscles. The above pattern can also be
macro-MUAP amplitude w h e n t h e muscle b e c a m e weaker. (From
seen in some patients with ALS as their disease progresses
Nandedkar SD: Quantitative electromography in clinical electrodiag-
(Fig. 8-96). The following model is proposed to explain this
nosis. In Goodgold J (ed): Rehabilitation Medicine. St. Louis, Mosby,
pattern of abnormalities. 190
1 9 8 8 , p p 6 8 - 1 1 , with permission.)
A normal MU consists of fibers distributed randomly
within a roughly circular territory. A small portion of the
muscle cross-section contains fibers from several different EMG parameters. Thus, the EMG examination may be d e -
I MUs. We begin with three normal MUs, as shown in Figure ferred for a few weeks, except when the weakness is severe
j 8-97A. In most neuropathies, the main disease process is a and sudden. Evidence of a few active MUs in these cases is a
I loss of MUs (Fig. 8-97B). Acutely after an MU is lost (#2 in good prognostic indicator.
Fig. 8-97B), we would expect to see abnormalities in the form A few days to weeks after the insult, the muscle fibers be-
of a reduced MU count, increased recruitment frequency, re- longing to the lost MUs are recognized from the fibrillation po-
duced recruitment ratio, and a reduced IP at maximal effort. tentials and positive sharp wave recordings on routine needle
As described earlier, these are not necessarily the most sensitive EMG.
Figure 8-97. Concept of MU remodeling. A.Three normal MUs have overlapping territories. B, MU # 2 is IOSL C, MU # I reinnervates and
shows fiber grouping in only a small portion of t h e MU territory. (D) MU # I is significantly enlarged. E, MU # I is lost. Some fibers (indicated by
checkered pattern) are reinnervated. F, MU # I is fractionated. See t e x t for details.
Normal MU Myopathy Neuropathy

fiber density Median macro EMG
MUAP amplitude (pV)
Figure 8-99. Scanning EMG. In A, the cross-section of a normal
MU is shown schematically, The triggering e l e c t r o d e is placed at a
Figure 8-98. Macro MUAP and FD.The FD (A, C) and median
fixed position while t h e scanning electrode is pulled through t h e MU
macro-MUAP amplitude (B, D) in normal subjects (top) and patients
territory in 50-u.m increments.The various positions of the needle tip
with myopathy (below). N o t e the slight increase in FD in myopathy.
are correlated t o t h e scan in Figure 8-IOOA. In B, a cross-section of
T h e r e is a shift t o w a r d s smaller macro-MUAPs, although m o s t sub-
an MU affected by a myopathy is shown. N o t e t h e areas with fiber
jects had values within t h e normal r a n g e . (From Nandedkar SD:
loss and fiber grouping.The dashed line r e p r e s e n t s t h e c o r r i d o r of
Quantitative electromyography in clinical diagnosis. In Goodgold J
the scanning electrode.This MU architecture corresponds t o t h e scan
(ed): Rehabilitation Medicine. St. Louis, Mosby, 1988, pp 68-1 I, with
in Figure 8-I00B. Finally, t h e MU a r c h i t e c t u r e in a n e u r o p a t h y is
permission.)
shown in C. N o t e t h e fiber grouping in t h e b o t t o m p o r t i o n of t h e
t e r r i t o r y similar t o t h e model in Figure 8 - 9 7 C . T h e c o r r e s p o n d i n g
scan is shown in Figure 8-100C.
In a few weeks, the surviving M U s begin to reinnervate the
denervated fibers from the lost M U s (Fig. 8-97C). The newly
formed collateral sprouts are short and thus only those MUs that become 10-40 times their normal size. When this occurs a small
are within the immediate vicinity of the lost MUs participate in cross section of the muscle may contain only a single MU. If the
reinnervation. Therefore, the muscle fiber distribution may be CN or MN electrode is in this area, the "giant" MUAPs may be
normal in a portion of the MU territory (top half of MU #1 in seen. At maximal effort, the IP may contain discharges of just
Fig. 8-97C) while there is fiber grouping in other parts (bottom one MU, i.e., discrete pattern. The recruitment frequency and
portion of MU #1 in Fig. 8-97C). Conventional recording elec- ratio are increased. The FD, or CN, MN or macro-MUAP am-
trodes have a small uptake area compared to the size of the MU plitude will be significantly increased. The MU-counting tech-
territory. As a result, one may record normal and abnormal niques demonstrate a reduced number of MUs. On muscle
MUAPs from the same MU just by changing the electrode posi- biopsy there is marked fiber type grouping.
tion. During reinnervation, the M U A P appears unstable. The If such an enlarged MU is lost (Fig. 8-97E), it represents a
earliest abnormalities during this stage are increased FD on marked decrease in the number of muscle fibers. Some fibers at
SFEMG, and polyphasic and unstable MUAPs on CN or MN the MU periphery may be reinnervated (e.g., MU # 3 or others
EMG. As reinnervation continues, fiber type grouping is seen indicated by checkered pattern), but a vast majority of fibers at
on muscle biopsy. the center are too far away to receive collateral sprouts. This
In a slowly progressing disease, reinnervation may fully com- large group of fibers atrophy and remain denervated. The
pensate for the lost MUs and the surviving MUs are enlarged muscle biopsy shows grouped atrophy of muscle fibers. On
(Fig. 8-97D). It is entirely conceivable that individual MUs may EMG the MUAPs are unstable and their waveform may exhibit
Figure 8-/00. Scanning EMG recordings. (A) Normal

muscle; (B) myopathy; and (C) neuropathy. The anatomic
models for these scans a r e described in Figure 8-99. See
text for details.
Normal tib. a n t . P o l y m y o s i t tib. ant.

significant differences. Significant spontaneous activity (e.g., pulled through the entire MU territory. The sequential plot of
fibrillations, positive sharp waves, etc.) is also observed. Loss MUAPs is called the "scan" (Fig. 8-100).
of large MUs implies reduced macro and concentric MUAP am- When the needle tip is positioned deep within the muscle and
plitude. Since this loss remains uncompensated, the patient may outside the MU territory (position 1, Fig. 8-99), the potential
present with new onset or rapid progression of weakness in the registered by the cannula of the concentric needle is recorded
muscle. (Fig. 8-100A, traces at the top). This is similar to the Macro
It is also postulated that the motor neuron may not be able to EMG recordings, except the waveform appears inverted, since
support nutritionally the enlarged pool of muscle fibers. Thus, the cannula is the reference electrode. As the electrode is pulled
some muscle fibers are lost, which reduces the MU size. This in steps of 50 pm, it enters the MU territory (position 2, Fig. 8¬
phenomenon is called fractionation (Fig. 8-97F). The newly 99). The electrode records a sharp MUAP with low rise time
denervated fibers may not become reinnervated if the remaining (Fig. 8-100A, traces in the middle of scan). Its spike component
MUs have also reached their capacity for reinnervation. This is defined by muscle fiber organization within that portion of
can also result in new onset or rapid progression of weakness. the MU territory. Therefore, the shape of the MUAP continues
Muscle biopsy wilt show atrophic muscle fibers, perhaps in to change as the needle is withdrawn. The last occurrence of
small to moderate groups. Some MUAPs may appear to have spike activity is seen as the electrode exits the MU territory
low amplitude and polyphasic waveform, a characteristic often (Fig. 8-100A, position 3 in Fig. 8-99). When the needle is su-
associated with myopathy. Thus, severely weak muscles should perficial and outside the MU territory (position 4 in Fig. 8-99)
be avoided for EMG testing and muscle biopsy. Similar models no MUAP activity is registered (bottom-most traces in Fig. 8-
have been postulated for MU remodeling by Emeryk-Szajewska 100A). This defines the diameter of the MU territory.
and Kopec. 60
The middle scan in Figure 8-100 was performed in a patient
In many myopathies muscle fibers are lost (Figs. 8-1 and 8-99). with myopathy. The size of the MU territory is not reduced.
This is compensated by regeneration and hypertrophy of fibers. However, we find sections within the scan with no MUAP activ-
As a result, the number of fibers may be reduced only slightly. ity. This represents the area of the MU where muscle fibers have
The MU territory contains some areas that do not have muscle been lost, called silent areas. They cannot be recognized on the
fibers. However, we cannot recognize them on the EMG exami- routine needle EMG examination. Once again a variety of
i nation. Other parts of the MU territory have focal grouping of MUAP waveforms can be seen recorded from the same M U .
muscle fibers due to regeneration and or reinnervation. This com- Initially, the MUAP is simple with a short duration reflecting
bination of disease processes gives increased FD on SFEMG, but loss of fibers from the MU. This is followed by a silent area also
normal or reduced macro-MUAP amplitude (Fig. 8-98). 91
indicating fiber loss. Later, we can see complex waveforms of
long duration due to variability of fiber diameter. In patients
SCANNING EMG with myopathy, the number of silent areas is increased. - 84 91188
The scan in Figure 8-100C was performed in a patient with

The above concepts of MU reorganization are vividly de- neuropathy. Notice that there is no significant change in the M U
scribed by a technique called scanning E M G . In this technique
179
territory. Initially, the MUAP waveform is normal. Later
(Fig. 8-99), an intramuscular needle electrode is positioned to polyphasic and unstable MUAPs are found. A late component is
trigger the sweep by a MUAP from a single MU. A second elec- also seen, representing the portion of MU where reinnervation
trode is inserted through the same MU territory to record activity is taking place.
from the same MU. When the MU discharges, the delayed The scanning EMG demonstrates the variability of MUAP
MUAP is recorded and displayed with a delay. The second elec- waveforms within the MU. It shows that the disease processes
trode is pulled out by 50 pm, and another MUAP is acquired and do not affect the MU in a homogeneous fashion. It further em-
displayed. The process is repeated until the electrode has been phasizes the need to search different corridors within the muscle
Table 8-17. Quantitative Analysis in Electrodiagnosis

Rep Cond
FD Jitter Dur %PP Amp Fib IP Macro Mamp MCV scv F H Stim Block EP
Muscular + X X
Dystrophy
Myositis + X + X
MG X X
MS X X X
ALS + X X +
Polyneuropathy + + X X X
GBS + X X X X
Entrapmenrt + X X X
Root + X X + + +
Central X V X
A summary of suggested tests and their expectedfindingsin different neuromuscular diseases is presented. MG, myasthemia gravis; MS, myasthenic syndrome; ALS, amy-
otrophic lateral sclerosis; GBS, Guillain-Barre syndrome; X, expected abnormalfinding;V, expected normalfinding;+, additional information for differential diagnosis.
From Stalberg EV: Electrodiagnostic assessment and monitoring of motor unit changes in disease. Muscle Nerve 1991; 14:292-303, with permission.
352 — P A R T I I BASIC AND ADVANCED TECHNIQUES
Table 8-18. Quantitative Analysis in Follow-up

Rep Cond
FD Jitter Pur %PP Amp Fib IP Macro Mamp MCV SCV F H Stim Block EP
Muscular • •
Dystrophy
Myositis * • •
MG •
MS
ALS • • •
Polyneuropathy • • • •
GBS
Entrapmenrt • •
Root • •
Central •
Electrophysiologic measurements useful to assess progression of various neuromuscular diseases are indicated. We would also add MUNE for the ALS patients. MG,
myasthemia gravis; MS, myasthenic syndrome; ALS, amyotrophic lateral sclerosis; GBS, Guillain-Barre syndrome; •, suitable for monitoring.
From Stalberg EV: Electrodiagnostic assessment and monitoring of motor unit changes in disease. Muscle Nerve 1991;14:292-303, with permission.
to assess MUAPs. Different sites should be separated by a few onset, progression and prognosis of the disease making EMG an
(5-10) millimeters to avoid duplicate recordings from the same even more powerful test procedure.
MU. While quantitative analysis is a powerful tool, it is also time
consuming. Considering patient discomfort, time, diagnostic
OTHER QUANTITATIVE METHODS sensitivity and specificity, and cost and reimbursement, we
must make a judicious choice of the test procedures to be used
In this chapter we have reviewed the routinely used quantita- for clinical electrodiagnosis. - A technique suitable for di-
187 219
tive test procedures to assess changes in the MU architecture agnosis (Table 8-17) is not necessarily suitable for follow-up
and number. The MU is also affected by other disease and monitoring the disease progression (Table 8-18). This se-
processes. As an example, disuse of a limb after immobilization lection may vary among different investigators, their expertise,
results in atrophy but there is no change in muscle fiber distrib- and resources. Nevertheless, the exercise allows one to gain a
ution or the number of MUs. The change in fiber size reduces better understanding of the diagnostic sensitivity and speci-
its action potential propagation velocity. - When the mechan-
66 173
ficity of the various QA procedures, and their relationship to
ical apparatus of the muscle is affected, the EMG may be the MU.
normal. Nevertheless, we could study the contractile properties
of the motor units using EMG recording procedures. - - - 33 45 50149 217
Surface EMG offers the benefit of noninvasive and painless test ACKNOWLEDGMENTS
procedure. It would be of great interest to develop new tech-
niques for diagnosis, prognosis, and monitoring different neuro- The first author would like to thank Oxford Instruments for
muscular d i s e a s e s . With high-quality amplifiers and
116123131
support in this project. The assistance by Mr. Desh Nandedkar
signal-processing techniques, we can now follow individual in drawing and editing the figures is greatly appreciated. Most
MUs as they are affected by disease processes. - As stated ear-
48 83
are reproduced with permission from the Center for Academic
lier, we may not be able to use these techniques in a routine Scholastic Achievement (www.casaengineering.com).
clinical environment. Yet it is the quantitative analysis that will
increase our understanding of the pathophysiology and progres-
sion of neuromuscular diseases.
REFERENCES
1. Andreassen S, Rosenfalek A: Recording from a single motor unit during strong
effort. IEEE Trans Biomed Engr 1978;6:501-508. "
CONCLUSION 2. Andreassen S. Rosenfalek A: Relationship of intracellular and extracellular
action potentials of skeletal muscle. CRC Crit Rev Bioeng 1981:6:267 306.
Quantitative analysis is useful when the routine EMG studies 3. Andreassen S: Methods for computer-aided measurement of motor unit para-
meters. Electroencephalogr Clin Neurophysiol 1987;Suppl 39:13-20.
are equivocal. The measurements can be used to follow the
4. Aquilonius S-M, Askmark H. Gillberg P-G, et al: Topographic localization of
effect of treatment and disease progression. Coupled with com- motor endplates in cryosections of whole human muscles. Muscle Nerve
puter simulations and muscle biopsy studies, they provide a 1984;7:287-293.
better understanding of the MU architecture in normal subjects 5. Armon C, Brandstater ME, Peterson GW: Motor unit number estimates and
quantitative muscle strength measurements of distal muscles in patients with
and in patients with nerve and muscle diseases. We also have a amyotrophic lateral sclerosis. Muscle Nerve 1997;20:499-501.
better appreciation of the sensitivity and specificity of measur- 6. Armon C, Brandstater ME: Motor unit number estimate based rates of progres-
ing different EMG features. All of this information can be used sion of ALS predict patient survival. Muscle Nerve 1999;22:1571-1575.
to plan and execute the EMG examination in an efficient 7. Ballantyne IP, Hansen S: A new method for estimation of the number of motor
units in a muscle. J Neurol Neurosurg Psychiatry 1974;3:907-915.
manner to obtain even more information about the disease 8. Barkhaus PE, Nandedkar SD. Sanders D: Quantitative EMG in inflammatory
processes. In some instances it may also allow us to define the myopathy. Muscle Nerve 1990;13;247-253.
9. Barkhaus PE. Nandedkar SD: Recording characteristics of the surface EMG 40. Daube J: Statistical estimates of number of motor units in thenar and foot mus-
electrodes. Muscle Nerve 1994; 17; 1317-1323. cles in patients with amyotrophic lateral sclerosis or residual of poliomyelitis.
10. Barkhaus PE, Nandedkar SD: On the selection of concentric needle elec- Muscle Nerve 1988;3:957.
tromyogram motor unit action potentials: Is the rise time criterion too restric- 41. Daube J: Needle examination in clinical electromyography. Muscle Nerve
tive? Muscle Nerve 1996;19:1554-1560. 1991;14:685-700.
11. Barkhaus PE, Periquet MI, Nandedkar SD: Quantitative motor unit action po- 42. Daube J: Estimating the number of motor units in a muscle. J Clin
tential analysis in paraspinal muscles. Muscle Nerve 1997;20:272-275. Neurophysiol 1995;12:585-594.
12. Barkhaus PE, Nandedkar SD: In Nandedkar SD (ed): EMG on CD. CAS A 43. Daube J: Clinical Neurophysiology. Philadelphia, F.A. Davis, 1996.
Engineering, Hopewell Junction, Volume II, 1999. 44. DeKoning P, Wieneke GH, Spijk D, et al: Estimation of the number of motor
13. Barkhaus PE, Periquet MI. Nandedkar SD: Quantitative electrophysiologic units based on macro EMG. J Neurol Neurosurg Psychiatry 1988;51:403¬
studies in sporadic inclusion body myositis. Muscle Nerve 1999:22: 411.
480-^87. 45. Dengler R, Konstanzer A, Kuther G, et al: Amyotrophic lateral sclerosis: Macro
14. Bertorini TE, Stalberg EV, Yuson CP, Engel WK: Single-fiber electromyogra- EMG and twitch forces of single motor units. Muscle Nerve 1990;13:545-550.
phy in neuromuscular disorders: Correlation of muscle histochemistry, single¬ 46. Doherty TJ, Brown WF: The estimated number and relative sizes of thenar
, fiber electromyography and clinical findings. Muscle Nerve 1994; 17:345-353. motor units as selected by multiple point stimulation in young and older adults.
15. Bischoff C, Machetanz J, Conrad B: Is there age-dependent continuous increase Muscle Nerve 1993;16:355-366.
in the duration of the motor unit potential? Electroencephalogr Clin 47. Doherty TJ, Stashuk DW, Brown WF: Determinants of mean motor unit size:
Neurophysiol 199! ;81:304-311. Impact on estimates of motor unit number. Muscle Nerve 1993;16:1326-1331.
16. Bischoff C, Stalberg E, Falck B, Eeg-Olofsson K: Reference values of motor 48. Doherty TJ, Brown W: A method for the longitudinal study of human thenar
unit action potentials obtained with multi-MUAP analysis. Muscle Nerve motor units. Muscle Nerve 1994;17:1029-1036.
1994;17:842-851. 49. Dorfman L, Howard J, McGill K: Clinical studies using automatic decomposi-
17. Bodine-Fowler S, Garfmkel A, Roy RR, Edgerton VR: Spatial distribution of tion electromyography (ADEMG) in needle and surface EMG. In Desmedt JE
muscle fibers within the territory of a motor unit. Muscle Nerve 1990; 13: (ed): Computer-Aided Electromyography and Expert Systems. Clinical
1133-1145. Neurophysiology Updates. Amsterdam, Elsevier, 1989, pp 189-204.
18. Bril V, Werb MR, Greene DA, Sima AA: Single fiber electromyography in dia- 50. Dorfman LJ, Howard JE, McGill KC: Triphasic behavior response of motor
betic peripheral polyneuropathy. Muscle Nerve 1996;19:2-9. units to submaximal fatiguing exercise. Muscle Nerve 1990;13:621-628.
19. Bromberg MB: Motor unit estimation: Reproducibility of spike triggered aver- 51. Dubowitz V, Brooke M: Muscle Biopsy: A Modern Approach. Philadelphia,
aging technique in normal and ALS subjects. Muscle Nerve 1993;16:466-471. W.B.Saunders, 1973.
20. Bromberg M, Forshew D: Motor unit number estimation, isometric strength and 52. Dumitru D: Electrodiagnostic Medicine. Philadelphia, Hanley & Belfus, 1995.
electromyographic measures in amyotrophic lateral sclerosis. Muscle Nerve 53. Dumitru D, King JC, Nandedkar SD: Motor unit potential recorded with con-
1993;16:1213-1219. centric electrodes: Physiologic implications. Electroencephalogr Clin Neuro-
21. Bromberg M, Scott D: Single fiber EMG reference values: Reformatted in tabu- physiol 1997;105:333-339.
lar form. Muscle Nerve 1994; 17:820-821. 54. Dumitru D, King JC, Nandedkar SD: Concentric/monopolar needle electrode
22. Bromberg MB: Electrodiagnostic studies in clinical trials for motor neuron dis- modelling: Spatial recording territory and physiologic implications.
ease. J Clin Neurophysiol 1998; 15:117-128. Electroencephalogr Clin Neurophysiol 1997;105:370-378.
I 23. Brown WF, Strong MJ, Snow R: Methods for estimating numbers of motor 55. Dumitru D, King JC, Nandedkar S: Comparison of single-fiber and macro elec-
units in biceps brachialis muscles and losses of motor units with aging. Muscle trode recordings: Relationship to motor unit action potential duration. Muscle
Nerve 1988;11:423-432. Nerve 1997;20:1381-1388.
24. Buchman A, Garratt M: Determining neuromuscular jitter using a monopolar 56. Dumitru D, King JC Rogers WE: Motor unit action potential components and
electrode. Muscle Nerve 1992;15:615-619. physiologic duration. Muscle Nerve 1999;22:733-741.
25. Buchthal F, Guld C, Rosenfalek P: Action potential parameters in normal 57. Edstrom L, Kubelberg E: Histochemical composition, distribution of fibers and
human muscles and their dependence on physical variables. Acta Physiol Scand fatiguability on single motor units. J Neurol Neurosurg Psychiatry 1968;
1954;32:200-218 31:424-433.
26. Buchthal F, Guld C, Rosenfalek P: Multielectrode study of the territory of a 58. Ekstedt J, Stalberg EV: How the size of needle electrode lead-off surface influ-
motor unit. Acta Physiol Scand 1957;39:83-104. ences the shape of the single muscle fiber action potential in electromyography.
27. Buchthal F: Electrophysiological signs of myopathy as related with muscle Comp Prog Biomed 1973;3:203-212.
biopsy. Acta Neurol (Napoli) 1977;32:1-29. 59. Elmquist D, Hofmann W, Kugelberg J, Quasol DMJ: An electrophysiologic in-
28. Buchthal F, Kamieniecka Z: The diagnostic yield of quantified electromyogra- vestigation of neuromuscular transmission in myasthenia gravis. J Physiol
phy and quantified muscle biopsy in neuromuscular disorders. Muscle Nerve 1964;174:417-434.
1982;2:265-280. 60. Emeryk-Szajewska B, Kopec J, Karwanska A: The reorganization of motor unit
29. Buchthal F: Electromyography in the evaluation of muscle diseases. In in motor neuron disease. Muscle Nerve 1997;20:306-315.
Fuglsang-Frederiksen A (ed): Methods in Clinical Neurophysiology, Vol 2. 61. Engstrom JW. Olney RK: Quantitative motor unit analysis: The effect of sample
Number 2. DANTEC Elektronik, Skovlunde (Denmark). 1991. size. Muscle Nerve 1992;15:277-281.
30. Cao J, Sanders DB: Multivariate descriminant analysis of the electromyo- 62. Ertas M, Stalberg EV, Falck B: Can the size principle be detected in conven-
graphic interference pattern: A statistical approach to discrimination among tional EMG recordings? Muscle Nerve 1995;18:435-439.
controls, myopathies and neuropathies. Med Biol Eng Comput 1996:34: 63. Ertas M, Erdinc O: Surface stimulation single-fiber electromyography. Muscle
369-374. Nerve 1998:21:118-120.
31. Cenkovich FS, Gersten JW: Fourier analysis of the normal human electromyo- 64. Ertas M, Baslo B, Yildiz N, et al: Concentric needle electrode for neuromuscu-
gram. Am J Phys Med 1963;42:192-204. lar jitter analysis. Muscle Nerve 2000;23:715-719.
32. Cenkovich F, Hsu S, Gersten JW: A quantitative electromyographic index that 65. Falck B, Stalberg E, Bischoff C. Influence of recording site within the muscle
is independent of force of contraction. Electroencephalogr Clin Neurophysiol on motor unit potentials. Muscle Nerve 1995;18:1385-1389.
1982:54:79-86. 66. Fang J, Shahani BT: Measurement of muscle fiber conduction velocity by sur-
33. Chan KM. Doherty TJ, Andres LP. et al: Longitudinal study of the contractile face electromyography triggered averaging technique. Muscle Nerve 1996:
and electrical properties of single thenar motor units. Muscle Nerve 19:918-919.
1998;21:839-849. 67. Fang J. Shahani BT, Graupe D: Motor unit number estimation by spatial-tem-
34. Chan RC. Hsu TC: Quantitative comparison of motor unit potential parameters poral summation of single motor unit potentials. Muscle Nerve 1997:
between monopolar and concentric needles. Muscle Nerve 1991; 14:1028-1032. 20:461-468.
35. Chaudhary V. Watson D, Bird S. Cornblath DR: Stimulated single fiber elec- 68. Feinstein B. Lindegard B. Nyman E, Wohlfart: Morphological studies of motor
tromyography in Lambert-Eaton myasthenic syndrome. Muscle Nerve 1991: units in normal human muscles. Acta Anat 1955;23:127-142.
14:1227-1230. 69. Felice KJ: Thenar motor unit number estimates using the multiple point stimu-
36. Chaudhary V, Crawford T: Stimulation single fiber EMG in infant botulism. lation technique: Reproducibility studies in ALS patients and normal subjects.
Muscle Nerve 1999;22:1698-1703. Muscle Nerve 1995;18:1412-1416.
37. Conwit RA, Tracy B. Cowl A, et al: Firing rate analysis using decomposition 70. Felice KJ: A longitudinal study comparing thenar motor unit number estimates
enhanced spike triggered averaging in the quadriceps femoris. Muscle Nerve to other quantitative tests in patients with amyotrophic lateral sclerosis. Muscle
1998;21:1338-13407 Nerve 1997;20:179-185.
38. Christensen H, Fuglsang-Frederiksen A: Power spectrum and.turns analysis of 71. Freund H, Hefter H, Homberg V: Motor unit activity in motor disorders. In
EMG at different voluntary effort in normal subjects. Electroencephalogr Clin Shahani BT (ed): Electromyography in CNS Disorders: Central EMG. Boston,
Neurophysiol 1984:64:528-535. Butterworth Publishers, 1984.
39. Czekajewski J, Ekstedt J, Stalberg E:. Oscilloscopic recording of muscle fiber 72. Fuglsang-Frederiksen A, Mansson A: Analysis of electrical activity of normal
action potentials: The window trigger and the delay unit. Electroencephalogr muscle in man at different degrees of voluntary effort. J Neurol Neurosurg
Clin Neurophysiol 1969;27:536-539. Psychiatry 1975;38:683-694.
73. Fuglsang-Frederiksen, A, Seheel U, Buchthal F: Diagnostic yield of the analy- 107. Liguori R, Dahl K, Fuglsang-Frederiksen A: Turns-amplitude analysis of the
sis of the pattern of electrical activity and of individual motor unit potentials in electromyographic recruitment pattern disregarding force measurement. I:
myopathy. J Neurol Neurosurg Psychiatry 1976;39:742-750. Method and reference values in healthy subjects. Muscle Nerve 1992; 15:
74. Fuglsang-Frederiksen A, Scheel U, Buchthal F: Diagnostic yield of the analysis 1314-1318.
of the pattern of electrical activity and of individual motor unit potentials in 108. Liguori R, Dahl K, Fuglsang-Frederiksen A: Turns-amplitude analysis of the elec-
neurogenic involvement. J Neurol Neurosurg Psychiatry 1977;40:544-554. tromyographic recruitment pattern disregarding force measurement. II: Findings
75. Fuglsang-Frederiksen A: Interference EMG analysis. In Desmedt JE (ed): in patients with neuromuscular disorders. Muscle Nerve 1992;15:1319-1324.
Computer-Aided Electromyography and Expert Systems. Amsterdam, Elsevier, 109. Lindstrom L, Magnusson R, Petersen I: Muscular fatigue and action potential
Clinical Neurophysiology Updates. 1989, pp 161-188. conduction velocity changes studied with frequency analysis of EMG signals.
76. Galea V, Bruin H, Cavasin R, McComas AJ: The number and relative sizes of Electromyography 1970;4:341-356.
motor units estimated by computer. Muscle Nerve 1991; 14:1123-1130. 110. Lomen-Hoerth C, Olney R: Comparison of multiple point and statistical motor
77. Gan R, Jabre JF: The spectrum of concentric macro correlations. Part II. unit number estimation. Muscle Nerve 2000;23:1525-1533.
Patients with diseases of muscle and nerve. Muscle Nerve 1992;15:1085-1088. 111. Luciano CA, Sivakumar S, Spector SA, Dalakas MC: Electrophysiologic and
78. Gath I, Stalberg E: On the measurement of Fiber density in human muscles. histologic studies in clinically unaffected muscles of patients with prior para-
Electroencephalogr Clin Neurophysiol 1982;54:699-706. lytic poliomyelitis. Muscle Nerve 1996;19:1413-1420.
79. Gilai AN: Analysis of turns and amplitude in EMG. In Desmedt JE (ed): 112. Mandler RN: Stimulated single fiber EMG in wound botulism. Muscle Nerve
Computer-Aided Electromyography and Expert Systems. Clinical Neuro- 1996;19:1171-1173.
physiology Updates. Amsterdam, Elsevier, 1989, pp 142-160. 113. Massey JM, Sanders DB, Howard JF Jr. The effect of cholinesterase inhibitors
80. Gilchrist JM, Nandedkar SD, Stewart CS, et al: Automatic analysis of the elec- on SFEMG in myasthenia gravis. Muscle Nerve 1989;12:154-155.
tromyographic interference pattern using turns:amplitude ratio. Electro- 114. Massey JM, Sanders DB: Single-fiber EMG demonstrates reinnervation dy-
encephalogr Clin Neurophysiol 1988;70:534-540. namics after nerve injury. Neurology 1991;41:1150-1151.
81. Gilchrist J, Barkhaus P: Single fiber EMG reference values: A collaborative 115. Massey JM, Sanders DB: Single fiber electromyography in myasthenia gravis
effort. Muscle Nerve 1992;15:151-161. during pregnancy. Muscle Nerve 1993;16:458-460.
82. Gilchrist JM, Massey JM, Sanders DB: Single fiber EMG and repetitive nerve 116. Masuda T, Sadoyama T: Topographical map of innervation zones within single
stimulation of the same muscle in myasthenia gravis. Muscle Nerve motor units measured with a grid surface electrode. IEEETrand Biomed Eng
1994;17:171-175. 1988;35:623-628.
83. Gooch C, Harati Y: Longitudinal tracking of the same single motor unit in amy- 117. McComas AJ, Fawcett PRW, Campbell MJ, Sica REP: Electrophysiological es-
otrophic lateral sclerosis. Muscle Nerve 1997;20:511-513. timation of the number of motor units within a human muscle. J Neurol
84. Gootzen T Vingerhoets D, Stegeman DF: A study of motor unit structure by Neurosurg Psychiatry 1971;34:121-131.
scanning EMG. Muscle Nerve 1992;14:449-454. 118. McComas AJ: Neuromuscular Function and Disorders. London, Butterworths,
85. Grimby G, Stalberg EV: An 8-year longitudinal study of muscle strength, 1977.
muscle fiber size and dynamic electromyogram in individuals with late polio. 119. McComas AJ: Invited review. Motor unit estimation: Methods, results and pre-
Muscle Nerve 1998;21:1428-1437. sent status. Muscle Nerve 1991;14:585-597.
86. Harvey AM, Masland RL: The electromyogram in myasthenia gravis. Bull 120. McComas AJ: Motor unit estimation: Anxieties and achievements. Muscle
Johns Hopkins Hosp 1941;48:1-13. Nerve 1995;18:369-379.
87. Hayward M: Automatic analysis of the electromyogram in healthy subjects of 121. McComas AJ, Quartly C, Griggs: Early and late losses of motor units after po-
different ages. J Neurol Sci 1977;33:397-413. liomyelitis. Brain 1997;120:1415-1421.
88. Hayward M, Willison RG: Automatic analysis of the electromyogram in pa- 122. McGill KC, Dorfman LJ: Automatic decomposition electromyography
tients with chronic partial denervation. J Neurol Sci 1977;33:415-423. (ADEMG): Methodologic and technical considerations. In Desmedt JE (ed):
89. Henneman E, Clamann HP, Gillus JD, Skinner RD: Rank order of motor neu- Computer-Aided Electromyography and Expert Systems. Clinical Neuro-
rons within a pool: Law of combination. J Neurophysiol 1974;37:1338-1347. physiology Updates. Amsterdam, Elsevier, 1989, pp 91-101.
90. Henricksson K, Stalberg EV: Terminal innervation pattern in polymyositis: A 123. Merletti R, DeLuca C: New techniques in surface electromyography. In
histochemical and SFEMG study. Muscle Nerve 1978;1:3-13. Desmedt JE (ed): Computer-Aided Electromyography and Expert Systems.
91. Hilton-Brown P, Stalberg EV: The motor unit in muscular dystrophy, single fiber Clinical Neurophysiology Updates. Amsterdam, Elsevier, 1989, pp 115-124.
EMG and scanning EMG study. J Neurol Neurosurg Psychiatry 1983;46:981-995. 124. Milner-Brown HS, Stein RB, Yemm R: The orderly recruitment of human
92. Hilton-Brown P, Stalberg E: Motor unit size in muscular dystrophy: A Macro motor units during voluntary isometric contractions. J Physiol (London)
EMG and scanning EMG study. J Neurol Neurosurg Psychiatry 1983; 1973;230:359-370.
46:996-1005. 125. Murga L, Sanchez F, Menendez C, Castilla JM: Diagnostic yield of stimulation
93. Howard JE, McGill KC, Dorfman LJ: Age effects on properties of motor unit and voluntary single fiber EMG. Muscle Nerve 1998;21:1081-1083.
action potentials: ADEMG analysis. Ann Neurol 1988;24:207-213. 126. Nandedkar S, Stalberg E: Simulation of macro EMG motor unit potential.
94. Howard JE, McGill KC, Dorfman LJ: Properties of motor unit action potentials Electroencephalogr Clin Neurophysiol 1983;56:52-62.
recorded with concentric and monopolar needle electrodes: ADEMG Analysis. 127. Nandedkar S, Sanders D, Stalberg E: Selectivity of EMG recoprding electrodes.
Muscle Nerve 1988;11:1051-1055. Med Biol Eng Comput 1985;23:536-540.
95. Howard JF, Sanders DB: Serial single-fiber EMG studies in myasthenia gravis 128. Nandedkar SD, Sanders DB, Stalberg EV: Simulation and analysis of the elec-
patients treated with corticosteroids and plasma exchange therapy. Muscle tromyographic interference pattern. Part I: Turns and amplitude measurements.
Nerve 1981;4:254. Muscle Nerve 1986;9:419-426.
96. Jabre JF: Macro-electromyographic studies using a new electrode. J Clin 129. Nandedkar SD, Sanders DB, Stalberg EV: Automatic analysis of the elec-
Neurophysiol 1988;5:204. tromyographic interference pattern. Part II: Findings in control subjects and in
97. Jabre JF: Concentric macro electromyography. Muscle Nerve 1991; 14: some patients with neuromuscular diseases. Muscle Nerve 1986;9:491-500.
820-825. 130. Nandedkar SD, Barkhaus PE: Estimation of the number of motor units in
98. Jabre JF, Spellman NT: The demonstration of size principle in human using human muscle. Proceedings of the 9th Annual Conference of EMBS, Boston,
macro electromyography and precision decomposition. Muscle Nerve 1996; 1987, pp 1999-2000.
19:338-341. 130a. Nandedkar SD: Quantitative electromyography in clinical electrodiagnosis. In
99. Jackson L: Digital Filters and Signal Processing. Boston, Kluwer, 1986. Goodgold J (ed): Rehabilitation Medicine. St. Louis, Mosby, 1988, pp 68-11.
100. Kimura J: Electrodiagnosis of Diseases of Nerve and Muscle:. Principles and 131. Nandedkar S, Barkhaus P, Sanders D, Stalberg E: Analysis of the amplitude and
Practice, 2nd ed. Philadelphia, F.A. Davis, 1989. area of the concentric needle EMG motor unit action potentials. Electro-
101. King JC, Dumitru D, Stegeman D: Monopolar needle electrode spatial record- encephalogr Clin Neurophysiol 1988;69:561-567.
ing characteristics. Muscle Nerve 1996; 19:1310-1319. 132. Nandedkar S, Sanders D, Stalberg E, Andreassen S: Simulation of concentric
102. King JC, Dumitru D, Nandedkar S: Concentric and single fiber electrode spatial needle EMG motor unit action potentials. Muscle Nerve 1988;2:151-159.
recording characteristics. Muscle Nerve 1997;20:1525-1533. 133. Nandedkar SD, Sanders DB: Simulation of myopathic motor unit action poten-
103. Kopec J, Hausmanowa-Petrusewicz I: On-line computer application in clinical tials. Muscle Nerve 1989;12:197-202.
quantitative electromyography. Electroencephalogr Clin Neurophysiol 1976;3!: 134. Nandedkar SD, Sanders DB: Measurement of the amplitude of the EMG enve-
404-406. lope. Muscle Nerve 1990;13:933-938.
104. Lang DJ, Smith T, Smith T, Lovelace R: Postpolio muscular atrophy. Arch 135. Nandedkar SD, Sanders DB, Stalberg EV: On the shape of the normal turns-
Neurol 1989;46:502-506. amplitude cloud. Muscle Nerve 1991;14:8-13.
105. Larsson LE: On the relationship between the EMG frequency spectrum and du- 136. Nandedkar SD, Sanders DB: Recording characteristics of monopolar EMG
ration of symptoms in lesions of the peripheral motor neuron. Electro- electrodes. Muscle Nerve 1991; 14:108-112.
encephalogr Clin Neurophysiol 1975;38:69-78. 137. Nandedkar SD, Barkhaus PE, Charles A: Multi-motor unit action potential
106. Lefever RS, DeLuca CJ: A procedure for decomposing the myoelectric signal analysis (MMA). Muscle Nerve 1995; 18:1155-1166.
into its constituent action potentials. I: Technique, theory and implementation. 138. Nandedkar SD, Dumitru D, King JC: Concentric needle electrode duration
IEEE Trans Biomed Eng 1982;29:149-157. measurement and uptake area. Muscle Nerve 1997;20:1225-1228.
139. Nandedkar S: "Objective EMG": Quantitation and documentation in the routine 174. Stalberg E, Thiele B: Transmission block in terminal nerve twigs: A single fiber
needle electromyographic examination. In Johnson E (ed): Practical Electro- electromyographic finding in man. J Neurol Neurosurg Psychiatry 1972;35:52-59.
myography, 3rd ed. Baltimore, Williams & Wilkins, 1997, pp 41-61. 175. Stalberg E, Schwartz MS, Trontelj JV: Single fiber electromyography in various
140. Nix WA, Scherer A: Single fiber Macro versus concentric trigger Macro EMG: processes affecting the anterior horn cell. J Neurol Sci 1975;24:403-415.
A comparison of methods. Muscle Nerve 1992;15:193-198. 176. Stalberg E: Electrogenesis in human dystrophic muscle. In Rowland LP ( ed):
141. Olney RK, Yuen EC, Engstrom JW: Statistical motor unit number estimation: Pathogenesis of Human Muscular Dystrophies. Amsterdam, Excerpta Medica,
Reproducibility and sources of error in patients with amyotrophic lateral sclero- 1977, pp 570-587.
sis. Muscle Nerve 2000;23:193-197. 177. Stalberg EV, Trontelj JV: Single Fiber Electromyography. Old Woking. UK,
142. Panayiotopoulos C, Scarpalezos S, Papetropolos T: Electrophysiologic estima- Mirvalle Press, 1979.
tion of motor units in Duchenne muscular dystrophy. J Neurol Sci 1974;23:89. 178. Stalberg EV, Gath 1: Measurement of the uptake area of small size electromyo-
143. Payan J: The blanket principle: A technical note. Muscle Nerve 1978; 1:423-426. graphic electrodes. IEEE Trans Biomed Eng 1979;26:374-376.
144. Pease WS, Bowyer BI: Motor unit analysis: Comparison between concentric 179. Stalberg E, Antoni L: Electrophysiological cross section of the motor unit. J
and monopolar electrodes. Am J Phys Med Rehabil 1988;67:2-6. Neurol Neurosurg Psychiatry 1980;43:469-474.
145. Petajan JH, Phillips BA: Frequency control of motor unit action potentials. 180. Stalberg E: Macro EMG: A new recording technique. J Neurol Neurosurg
Electroencephalogr Clin Neurophysiol 1969;27:66-72. Psychiatry 1980;43:475^82.
146. Petajan JH: Clinical electromyographic studies of diseases of the motor unit. 181. Stalberg E, Fawcett PRW: Macro EMG changes in healthy subjects of different
Electroencephalgr Clin Neurophysiol 1974;36:395-401. ages. J Neurol Neurosurg Psychiatry 1982;45:870-878.
147. Petajan J: Antigravity posture for analysis of motor unit recruitment: The 45 182. Stalberg E, Chu J, Bril V, et al: Automatic analysis of the EMG interference pat-
degree test. Muscle Nerve 1990;13:355-359. tern. Electroencephalogr Clin Neurophysiol 1983;56:672-681.
148. Petajan JH: AAEM Minimonograph # 3: Motor Unit Recruitment. Muscle 183. Stalberg E, Sanders D: The motor unit in ALS studies with different electro-
Nerve 1991;14:489-502. physiological techniques. In Rose C (ed): Progress in Motor Neuron Disease.
149. Quinlan JG, Iaizzo PA: Twitch response in a myopathy with impaired relaxation London, Pitman Books, 1983.
but no myotonia. Muscle Nerve 1990;13:326-329. 184. Stalberg E: Single fiber EMG, macro EMG, and scanning EMG: New ways of
150. Ravits J, Hallett M, Baker M, et al: Clinical eletcromyographic studies of post- looking at the motor unit. CRC Crit Rev Clin Neurobiol 1986;2:125-167.
poliomyelitis muscular atrophy. Muscle Nerve 1990; 13:667-674. 185. Stalberg E, Andreassen S, Falck B, et al: Quantitative analysis of individual
151. Roeleveld K, Sandberg A, Stalberg EV, Stegeman DF: Motor unit size estima- motor unit potentials: A proposition for standardized terminology and criteria
tion of enlarged motor units with surface electromyography. Muscle Nerve for measurement. J Clin Neurophysiol 1986;3:313-348.
1998;21:878-886. 186. Stalberg E: Use of single fiber EMG and macro EMG in study of reinnervation.
152. Rose AL, Willison RG: Quantitative electromyography using automatic analy- Muscle Nerve 1990;13:804-813.
sis: Studies in healthy subjects and patients with primary muscle disease. J 187. Stalberg EV: Electrodiagnostic assessment and monitoring of motor unit
Neurol Neurosurg Psychiatry 1967;30:403^H0. changes in disease. Muscle Nerve 1991;14:292-303.
153. Rosenfalek P, Rosenfalek A: Electromyography and sensory/motor conduction: 188. Stalberg EV, Dioszeghy P: Scanning EMG in normal muscle and in neuromus-
Findings in normal subjects. Laboratory of Clinical Neurophysiology, Rigs- cular disorders. Electroencephalogr Clin Neurophysiol 1991 ;81:403-416.
hospitale, Copenhagen, Denmark, 1975. 189. Stalberg E, Sonoo M: Assessment of variability in the shape of the motor unit
154. Sadeh M: Stimulated single fiber EMG in lambert-eaton myasthenic syndrome action potential, the "jiggle" at consecutive discharges. Muscle Nerve 1994; 17:
before and after 3,4-diaminopyridine. Muscle Nerve 1997;20:735-739. 1135-1144.
155. Sanders DB, Massey JM et al: Quantitative electromyography after po- 190. Stalberg EV, Bischoff C, Falck B: Outliers: A way to detect abnormality in
liomyelitis. In Halstead LS, Wiechers DO (eds): Research and Clinical Aspects quantitative EMG. Muscle Nerve 1994;17:392-399.
of the Late Effects of Poliomyelitis. Birth Defects: Original Article Series. J9L Stalberg E, Trontelj JV: Single Fiber Electromyography in Healthy and
White Plains, NY, March of Dimes, 1986, pp 189-200. Diseased Muscle, 2nd ed. New York, Raven Press, 1994.
156. Sanders DB: The effect of firing rate on neuromuscular jitter in Lambert-Eaton 192. Stalberg E, Grimby G: Dynamic electromyography and muscle biopsy changes
myasthenic syndrome. Muscle Nerve 1992;15:256-258. in a 4-year follow up: Study of patients with history of polio. Muscle Nerve
157. Sanders DB, Stalberg EV, Nandedkar SD: Analysis of electromyographic inter- 1995;18:699-707.
ference pattern. J Clin Neurophysiol 1996;13:385-400. 193. Stalberg E, Falck B, Sonoo M, Stalberg S, Astrom M: Multi-MUAP EMG
158. Sanders DB, Stalberg EV: AAEM Minimonograph #25: Single Fiber analysis—a two-year experience with quantitative method in daily routine.
Electromyography. Muscle Nerve 1996;19:1069-1083. Electroencephalogr Clin Neurophysiol 1995;97:145-154.
159. Sandstedt P, Henriksson KG, Larsson LE: Quantitative electromyography in 194. Stalberg E, Nandedkar SD, Sanders DB, Falck B: Quantitative motor unit po-
polymyositis and dermatomyositis. Acta Neurol Scand 1982;65:110-121. tential analysis. J Clin Neurophysiol 1996;13:401-422.
160. Santo Neto H, Carvalho V, Marques MJ: Estimation of the number and size of 195. Stashuk D, Deluca CI: Update on the decomposition and analysis of EMG sig-
human flexor digiti minimi muscle motor units using histological methods. nals. In Desmedt JE (ed): Computer-Aided Electromyography and Expert
Muscle Nerve 1998;21:112-114. Systems. Clinical Neurophysiology Updates. Basel, Karger, 1989, pp 39-53.
161. Schweizer V, Woodson G, Bertorini TE: Single fiber electromyography of the 196. Stashuk DW, Doherty TJ, Kassam A, Brown W: Motor unit number estimates
laryngeal muscles. Muscle Nerve 1999;22:111-114. based on the automated analysis of F-responses. Muscle Nerve 1994; 17:
162. Shahani BT, Wierzbicka M, Parker SW: Abnormal single motor unit behavior in 881-890.
the upper motor neuron syndrome. Muscle Nerve 1991:14:64-69. 197. Stein RB, Yang JF: Methods for estimating the number of motor units in human
163. Shahani BT, Fang J, Shand WR: A new approach to motor unit estimation with muscles. Ann Neurol 1990;28:487Î95.
surface EMG triggered averaging. Muscle Nerve 1995;18:1088-1092. 198. Stewart C. Nandedkar SD, Massey JM, et al: Evaluation of an automatic
164. Shefner JM, Jillapalli D, Bradshaw DY: Reducing intersubject variability in method of measuring features of motor unit action potentials. Muscle Nerve
motor unit number estimation. Muscle Nerve 1999;22:1457-1460. 1989;12:141-148.
165. Sica REP, McComas AJ: Evaluation of an automated method of analysing the 199. Swash M, Schwartz M: A longitudinal study of changes in motor units in motor
electromyogram. Can J Neurol Sci 1978;5:275-281. neuron disease. J Neurol Sci 1982:56:185.
166. Slawnych M, Laszlo C, Hersher C: A review of techniques employed to esti- 200. Thiele B, Stalberg E: Single fiber EMG findings in polyneuropathies of differ-
mate the number of motor units in a muscle. Muscle Nerve 1990; 13: ent aetiology, J Neurol Neurosurg Psychiatry 1975;38:881-887.
1050-1064. 201. Theile B, Bohle A: Anzahl der Spike-Komponenten im Motor Unit Potential.
167. Slawnych M, Laszlo C, Hershler C: Motor unit estimates obtained using the Electroencephalogr-EMG 1978:9:125-130.
new "MUESA" method. Muscle Nerve 1996;19:626-636. 202. Trojaborg W: Motor unit disorder and myopathies. In Halliday AT, Butler SE,
168. Slawnych M, Laszlo C. Hershler C: Motor unit number estimation: Sample size Paul R (eds): A Textbook of Clinical Neurophysiology. New York, John Wiley
consideration. Muscle Nerve 1997:20:22-28. & Sons, 1987, pp 417-437.
169. Smyth D: Quantitative electromyography in babies and young children with 203. Trojaborg W: Quantitative electromyography in polymyositis. A reappraisal.
primary muscle disease and neurogenic lesions. J Neurol Sci 1982;56:199-207. Muscle Nerve 1990; 13:964-971.
170. Smyth DP, Willison RG: Quantitative electromyography in babies and young 204. Trontelj JV, Mihelin M, Fernandez JM, Stalberg E: Axonal stimulation for end-
children with no evidence of neuromuscular disease. J Neurol Sci 1982;56: plate jitter studies. J Neurol Neurosurg Psychiatry 1986;49:677-685.
209-217. 205. Trontelj JV, Khuraibet A, Mihelin M: The jitter in stimulated orbicularis oculi
171. Sogaard K: Motor unit recruitment pattern during low-level static and dynamic muscle: Technique and normal values. J Neurol Neurosurg Psychiatry
contraction. Muscle Nerve 1995;18:292-300. 1988;51:814-819.
172. Sonoo M, Stalberg E: The ability of MUAP parameters to discriminate between 206. Trontelj JV, Zidar J, Denislic M, Vodusek DB: Facioscapulohumeral dystrophy:
normal and neurogenic MUAPs in concentric EMG: Analysis of MUAP "thick- Jitter in facial muscles. J Neurol Neurosurg Psychiatry 1988;51:950-955.
ness" and the proposal of "size index". Electroencephalogr Clin Neurophysiol 207. Trontelj JV, Stalberg EV: Single motor end-plates in myasthenia gravis and
1993;89:291-303. LEMS at different firing rates. Muscle Nerve 1991;14:226-232.
173. Stalberg E: Propagation velocity in single human muscle fibers. Acta Physiol 208. Trontelj JV, Stalberg EV, et al: Jitter of stimulated motor axon. Muscle Nerve
Scand T966;Suppl 287:1-112. 1992;15:449-454.
209. Uncini A, Lange DJ, Lovelace RE, et al: Long duration polyphasic motor unit 216. Yan K, Fang J: Motor unit behavior in stroke patients. Muscle Nerve
potentials in myopathies: A quantitative study with physiologic correlation. 1998;21:1502-1506.
Muscle Nerve 1990;13:263-267. 217. Yang JF, Stein RB, Jhamandas J, Gordon T: Motor unit numbers and contractile
210. Wang F, Delwaide PJ: Number and relative size of thenar motor units estimated properties after spinal cord injury. Ann Neurol 1990;28:496-502.
by an adapted multiple point stimulation method. Muscle Nerve 1995; 18: 218. Yarr I, Niles L: Muscle fiber conduction velocity and mean power spectrum fre-
969-979. quency in neuromuscular disorders and fatigue. Muscle Nerve 1992; 15:
211. Walton JN: The electromyogram in myopathy: Analysis with the audio-fre- 780-787.
quency spectrometer. J Neurol Neurosurg Psychiatry 1952;15:219-226. 219. Yu Y, Murray N: A comparison of concentric needle electromyography, quanti-
212. Weichers D, Hubbell S: Late changes in motor unit after acute poliomyelitis. tative EMG and single fiber EMG in the diagnosis of neuromuscular diseases. J
Muscle Nerve 1981;4:524. ' Neurol Neurosurg Psychiatry 1984;58:220-223.
213. Weichers DO: Single fiber electromyography with a standard monopolar elec- 220. Yuen EC, Olney RK: Longitudinal study of fiber density and motor unit number
trode. Arch Phys Med Rehabil 1985;66:47-48. estimate in patients with amyotrophic lateral sclerosis. Neurology 1997;49:
214. Weichers DO: Single fiber EMG evaluation in denervation and reinnervation. 573-578.
Muscle Nerve 1990;13:829-832. 221. Zalewska E, Hausmanowa-Petrusewicz I: Evaluation of MUAP shape irregu-
215. Willison RG: Analysis of electrical activity in healthy and dystrophic muscle in larity—A new concept of quantification. IEEE Trans Biomed Eng 1995;42:
man. J Neurol Neurosurg Psychiatry 1964;27:386-394. 16-20.
Chapter 9
Somatosensory Evoked
Potentials
Lawrence R. Robinson, M.D.
CHAPTER OUTLINE
Historical Aspects Reference Data and Criteria for Abnormality

Subject Selection • Abnormal Criteria
Anatomical Basis of SEP
Peripheral Nervous System Pathway. • Central Nervous Factors Affecting the SEP
System Pathway • SEP Spinal Pathway Patient Cooperation • Patient Height • Age • Gender
Neural Generators of SEP Waveforms • Temperature • Medication • Sleep • Reproducibility
Recording Electrode Locations • Waveform Generators: Upper
Limb • Waveform Generators: Lower Limb
SEP Techniques
Upper Limb SEPs • Mixed Nerve SEPs • Segmental Sensory
Recommended Standards for Recording Short-Latency Nerve SEPs • Lower Limb SEPs • Lower Limb Segmental
SEPs Somatosensory Evoked Potentials • Dermatomal
Somatosensory Evoked Potentials
Nomenclature for Recording Electrode Sites
Electrode Site Designations Miscellaneous SEP Techniques
SEP Waveforms Pudendal Nerve SEP • Trigeminal Nerve SEP • Medial/Lateral
Polarity • Peak Nomenclature • Peak Latency • Interpeak Plantar and Calcaneal Nerves
Latency • Amplitude Measurement * Waveform Morphology
SEP Applications
SEP Instrumentation Somatosensory Far-Field Potentials
Minimal Standards • Desirable Instrument Options
Prognostication of Central Nervous System Injury
Electrodes
Coma Evaluation
Composition • Recording Electrodes • Comparison of Needle
and Surface Electrodes • Type of Electrode Application SEPs and Prognosis
• Stimulating Electrodes • Stimulation and Ground Electrode
Considerations Conclusion
Somatosensory evoked potentials (SEPs) are waveforms that the supervising physician must directly acquire the technical ex-
can be recorded from the peripheral nerve, spinal cord, and/or pertise necessary to perform all types of SEPs (upper/lower
cerebral cortex following excitation of a peripheral nerve or cu- limb, dermatomal, segmental, etc.) and gain complete familiar-
taneous afferents. The beginning practitioner of electrodiagnos- ity with waveform interpretation and the various technical arti-
tic medicine may be a bit wary of performing SEPs because of facts that may contribute to an erroneous diagnosis. Waveform
inexperience recording these responses. This may simply be a distortions may result from inappropriate filter parameters,
lack of training or, in some institutions, the SEP may be desig- stimulus intensity/frequency effects, suboptimal signal-to-noise
nated as a study obtained by electroencephalographic or elec- ratio, and increased skin-electrode impedance, among other fac-
trodiagnostic technicians for later waveform review by the tors influencing data acquisition. Incomplete understanding of
physician. Prior to allowing a technician to record the information, these important concepts may result in a misdiagnosis, particularly
357
if the SEP waveforms are examined in isolation from the patient This was the first clear demonstration of a cortical electrical
without the opportunity to collect additional information. Thus, event obtained through electronic averaging. - 74 75
as for the needle EMG examination, SEPs should be viewed as The ability to electrically activate a peripheral nerve and av-
a dynamic test, with the plan for a specific patient varying as erage the ensuing depolarization at some other location in the
new information is obtained. neuraxis was not available prior to 1947. Before Dawson's in-
The vast SEP literature can be formidable for the novice prac- strument, all evoked potentials were recorded on a freely run-
titioner because of a lack of standardization regarding recording ning trace. The potentials were difficult to observe because of
techniques and waveform nomenclature. To some extent, the the background noise, which tended to obscure the waveforms
SEP examination can be conceptualized as simply a nerve con- of interest. Unlike the comparatively larger electroencephalo-
duction study over a relatively long portion of both the periph- graphic potentials, which are essentially unrelated to specific
eral and central nervous systems. As most practitioners are sensory input, the few-microvolt SEP has a fixed temporal rela-
rather familiar with obtaining nerve conduction velocities, the tionship to specific sensory stimuli. This temporal relationship
SEP is conceptually no more difficult. There is no reason why to external excitation results in the "evoked" aspect of the SEP.
an individual who routinely obtains nerve conduction velocity The relatively large EEG potentials combined with the environ-
data cannot, with proper instruction and practice, successfully ment's electrical interference require the rather small SEPs to
acquire SEPs. be extracted from this significant background activity.
This chapter approaches the SEP examination from a practi- Dawson's - instrument was capable of minimizing the random
74 75
cal standpoint with the goal of demystifying SEPs and perform- background electrical noise and observing time-locked SEPs
ing routine upper and lower limb studies in an efficient manner. relatively free of noise contamination. Dawson's - electronic 74 75
Although there are some aspects of SEPs that are poorly under- averager summated the regular occurrences, i.e., evoked wave-
stood, it is not necessary for the beginner to become embroiled forms, while the random noise was reduced over time. This
in these complexities. Rather, it is far more productive to master technique improved the signal-to-noise ratio. Development of
the fundamentals of SEP techniques and data acquisition. the average response computer (ARC) by 1958, and further re-
54
Because SEPs are relatively new to many individuals, a good finements in instrumentation, has permitted the pioneering work
beginning point is to briefly consider the historical aspects that of Dawson to be applied routinely in SEP investigations today.
led to the development of the modern SEP evaluation.
A N A T O M I C A L BASIS OF THE SEP

HISTORICAL ASPECTS
PERIPHERAL NERVOUS SYSTEM PATHWAY
Waveforms arising from animal cortical tissue secondary to
peripheral nervous system activation were first detected with two Somatosensory evoked potentials are typically obtained by
electrodes attached to the scalp, or one electrode in direct contact electrically stimulating the peripheral nervous system, which
with the cerebral cortex and the other on the scalp, by Caton in induces a relatively synchronous wave of depolarization to-
1875. Successful localization of the visual cortex was achieved
40
propagate rostrally along the peripheral and central conduction
with the use of candles and magnesium flares for visual stimula- pathways to the cerebral cortex. The type of nerve fibers excited
tion, which represents the first documented observation of the peripherally dictates the central pathways eventually generating
visual evoked potential. Stimulation of a limb yielded cortical re- the cortical SEP. Given that the SEP is evoked by exciting a
sponses resulting in the first S E P s . Caton also first described
4142
mixed, pure sensory peripheral nerve, or cutaneous afferents, it
the spontaneous electrical activity of the brain, i.e., the electroen- is possible to postulate the type of fibers activated.
cephalogram (EEG). Although Caton tried, he could not find Recall that peripheral nerve fibers are classified primarily by
cortical responses to auditory stimuli. As the early investigators their size (diameter) and whether or not they are myelinated,
did not have photographic equipment, sketches had to suffice for using one of two systems (Table 9-1). - Additionally, the di-
120 207
pictorial documentation of the results. 32

ameter of the nerve fiber is correlated to the type of sensory
Independently, Beck substantiated most of Caton's work by input conveyed. - - - - During the performance of a mixed
22 35 134 222 276
1891, particularly in identifying the visual cortex with one elec- nerve SEP, it is necessary to produce a small muscle twitch by
trode on its surface. Beck and Cybulski also described a re-
25 26
optimally adjusting the stimulating current intensity. - 5051 215
sponse to loud shouting when one of the cortical electrodes was SEPs from pure sensory nerves or from stimulating skin in the
attached to the temporal region of Beck's experimental animals' area of a given dermatome are generated using a current inten-
cortices. Additionally, in 1890-1891 Beck confirmed Caton's sity 2 or 3 times the sensory threshold. In mixed peripheral
I07 n i
observation of the EEG. Again, only sketches are available

24 25
nerve SEP studies, both motor and sensory nerves are activated
of these early cortical evoked potentials and spontaneous wave- (see below). Depolarization of motor fibers induce muscle fiber
forms. By 1898, Larinov - successfully localized the corti-
198 200
depolarization as well as an antidromically conducted impulse
cal auditory area by using tuning forks of three different tones. toward the spinal cord. The antidromic impulses are extin-
The first pictorial representations of cortical evoked poten- guished at the level of the anterior horn cell and do not propa-
tials and the EEG were published in 19 1 3 . Human evoked
247 2 4 8
gate beyond this point. The reason motor impulses are not
potentials were initially described by Davis in 1939. - She de- 69 70
conveyed to the central motor pathways is that neurochemical
tected cortical waveforms arising from electric shocks to upper transmission is one way from the suprasegmental axon to the
limb digits. Developments in electronic devices permitted am- anterior horn cell dendrites and not from the motor neuron to
plification and clinical observation of the microvolt potentials at the pyramidal tract, for example. On the other hand, the acti-
the surface of the scalp. Further instrumentation advances al- vated sensory fibers (which may include muscle afferent fibers)
lowed Dawson to stimulate the peripheral nervous system and propagate orthodromically along the peripheral nerve to the
record time-locked cortical events from the scalp in 1947. 72 7 3
spinal cord and following several synapses, terminate in the
Chapter 9 SOMATOSENSORY EVOKED POTENTIALS — 3 5 9
Table 9-1. Peripheral N e r v e Fiber T y p e s
A. Fiber Unmyelinated '

diameter: OJfMm
B. Conduction
velocity: 5 m/sec
C. Fiber class:
(Gasser)
Function:
• skin afferents
£. Motor fibers:
F. Fiber class:
(Lloyd)
G. Function:
muscle afferents
From Shepherd GM: Neurobiology. New York, Oxford University Press, 1983, p 258, with permission.
sensory portion of the cerebral cortex. It is important to

145 , 4 6
travel in the anterolateral aspect of the spinal c o r d . As 22134 3 3 0
distinguish which sensory fibers are activated by the designated noted above, the lemniscal tracts convey data from those so-
impulse parameters and the portions of the spinal cord convey- matic receptors concerned with muscle stretch, joint position,
ing these action potentials to the cerebral cortex. In this discus- vibration, and tactile/pressure stimuli. The cell bodies of these
56
sion, we will use the Lloyd classification system when 207

first order afferent nerve fibers are contained in the dorsal root
referring to different types of sensory nerve fibers. ganglia. Following propagation through the dorsal root entry
In the cat, delivery of 1.5 to 2 times the IA fiber threshold ex- zone, impulses along the lemniscal system enter the ipsilateral
clusively activates the IA fibers. At stimulus intensities above 2
16
dorsal columns and travel rostrally (Fig. 9 - 1 ) . The action
38135
times the IA threshold, group II fibers are excited. Minimal toe potentials within these first-order nerve fibers terminate by
or thumb twitch in humans preferentially activates group I synapsing in the dorsal column nuclei, i.e., the nucleus gracilis
(12-20 pm) and possibly group II (4-12 pm) nerve fibers. 34 3 5 1 6 3
and nucleus cuneatus in the lower region of the medulla (Fig.
The small group III and IV fibers (1 to 4 pm; nociceptive affer- 9_ J )38,135.253,306.315.318
ents) require rather high (possibly painful) current intensities and Lemniscal fibers from the lower limbs are anatomically lo-
are most likely not activated. Table 9-1 demonstrates that group I cated in the most medial aspects of the dorsal columns (Fig. 9¬
and II nerves convey vibration, proprioception, high threshold 1). These first-order fibers synapse with second-order neurons
pressure, and afferent input from the primary and secondary in the nucleus gracilis. - - Nerve fibers conveying dorsal
135 315 318
muscle spindles as well as the Golgi tendon organs and pacinian column data from the upper limbs are located more laterally, but
corpuscles. Once the nerve fibers conveying this information are adjacent to the lower limb nerve fibers. Upper limb afferents to
activated, action potentials travel centripetally to enter the spinal the lemniscal system synapse in the nucleus cuneatus, which is
cord through the medial aspect of the dorsal root entry zone. The positioned lateral to the nucleus gracilis, from which originate
above noted sensory inputs compose the lemniscal s y s t e m . 22134
second-order neurons representing the upper limb. - - 22 134 305 315
The second-order nerve fibers from the dorsal column nuclei

CENTRAL NERVOUS SYSTEM PATHWAY immediately cross to the contralateral side of the brain stem,
forming the decussation of the medial lemniscus, and continue
The lemniscal (dorsal column) system is a term utilized to their rostral ascent as the medial lemniscus itself (Fig. 9¬
distinguish a portion of the afferent somatic fibers that do not 1) 22,i34,3i8 j ^ j g terminates by synapsing in a specific region
t r a c t
Ventral posterior „
nucleus of thalamus
OCT
> Medial Figure 9-1. P r o p r i o c e p t i o n P a t h w a y s . Pathways for

lemniscus
conscious proprioception. (From Barr ML, Kiernan JA:The
cerebellum Human Nervous System, 5th ed. Philadelphia, J.B.
Lippincott, 1988, with permission.)
6
Nucleus «*
cuneatus
Fasciculus -
cuneatus
Dorsal
spinocerebellar -
tract
Nucleus dorsalis-
Thoracic level
(Clarke's column)
Fasciculus .
gracilis
Lumbosacral level
of the thalamus, the ventral posterolateral nucleus or VPL the lateral funiculus may convey a significant portion of the type
(Fig. 9-1). Third-order nerve fibers originating in the VPL, thal- IA and II afferent information to the somatosensory
amocortical fibers, project to the postcentral gyrus of the pari- cortex.55.i97.229.254 -rê proposed pathway is a continuation of pri-
etal lobe, known as the somatosensory cortex (Fig. 9-1). A 123
mary afferents that synapse at the level of Clark's column to
characteristic topographic arrangement of sensory regions is then ascend in the dorsal spinocerebellar tract and subsequently
discretely arranged along the cortex corresponding to lemniscal in the nucleus Z at the level of the medullopontine junction (see
fibers from specific portions of the body, forming the so-called Fig. 9-1). Nerve fibers from nucleus Z then travel to the
214
homunculus (Fig. 9-2). The somatosensory region represent-

134
medial lemniscus and ascend to the rostral region of the VPL.
ing the upper limb is located on the cortex superior to the face The traditional pathway from the VPL to the somatosensory
area. The lower limb somatosensory representation, however, is cortex then serves as the final pathway for the fibers from the
along the medial aspect of the cerebral hemisphere, separated lower limb.
from its contralateral body region in the opposite hemisphere by
the interhemispheric fissure. SEP SPINAL PATHWAY
With respect to the lower limb somatosensory information, a
second fiber route has been proposed, i.e., the spinomedul- In performing SEPs, the peripheral nerve is stimulated and
lothalamic tract. This view suggests that the dorsal aspect of
318
there is no debate as to the initial presumed pathway, i.e., the
Chapter 9 SOMATOSENSORY EVOKED POTENTIALS — 361
peripheral nerve fibers conducting the large afferent impulses. 8
Once the wavefront of depolarization enters the spinal cord,

however, the question arises as to exactly what pathways are uti-
lized by the electrically induced peripheral nerve stimuli. A cor-
relation does not necessarily exist between the artificially
induced electrical depolarization of the peripheral nervous
system and those central pathways traveled by natural stimuli.
With respect to SEPs, the question then becomes, what specific
spinal cord tracts are traveled by the electrically activated pe-
ripheral nerve fibers?
Most clinicians believe that SEP impulses traverse the dorsal
columns and reflect the neural continuity of this anatomic path-
j way. As previously noted, the dorsal columns most probably
convey, from the peripheral nervous system to cortical struc-
tures, a number of modalities such as vibration, proprioception,
light touch, and possibly tactile discrimination. It is usually as-
sumed that an abnormal SEP despite an intact sensory clinical
examination reflects the superiority of objective electrophysio-
logic data compared with a potentially subjective physical ex-
amination. Substantiation of the correlation between SEPs and
the anatomic pathways forming the lemniscal system is lacking
and limiteo to selected patient groups with well-localized
pathology or specific diseases, such as multiple sclerosis.
Although Dawson successfully obtained clinically useful
SEPs in 1947, - the correlation between SEPs and specific dis-
72 73
eases was not performed until I 9 6 0 . A number of patients

132133
| with lesions involving the peripheral nerves, roots, spinal cord,

Figure 9-2. Homunculus. Cortical representation of various a s -
and cerebral hemispheres were investigated. Findings in sub-
pects of the body comprising t h e so-called homunculus. (From Penfield
1 jects with various spinal cord lesions suggested that the SEP
W, Rasmussen T: T h e Cerebral C o r t e x of Man: A Clinical Study of
was primarily mediated by the dorsal columns as opposed to the
.Localization of Function. N e w York,The MacMillan Co., 1950, p 2 1 4 ,
anterolateral pathway. In 1963, a second series of patients with
with permission.)
various lesions producing losses of different ascending spinal
cord pathways were studied. In this investigation, patients
152
with diminished proprioception and vibration demonstrated ab- et al. investigated patients with spinal cord lesions between C 3
normal cortical SEPs, whereas subjects with pain loss did not and L 4 . A correlation was found between the degree of sen-
90
reveal altered SEPs. In patients with multiple sclerosis present- sory loss and magnitude of SEP abnormality. This study, how-
ing with alterations in vibration and proprioception, Namerow ever, did not document the degree of SEP alteration with
found that these individuals had abnormalities of their SEPs in specific sensory modality impairment. A similar study of spinal
proportion to the clinical loss of these particular clinical modal- cord-injured persons in 1984 failed to document a relationship
ities (proprioception and vibration). Comparisons between
228
of proprioceptive loss and abnormal S E P s . In another study
261
patients with thalamic versus brain stem lesions were performed of 50 consecutive patients referred for SEP evaluation, about
in 1975. Investigators noted that in persons with clinical evi-
230
half of the patients with abnormal clinical examinations demon-
dence of thalamic pathology, the SEPs were clearly abnormal, strated alterations in their SEPs. These authors reported that
145
whereas those patients with brain stem lesions not involving the abnormal SEPs were most consistently associated with signs of
lemniscal fibers had normal SEPs. This study strongly suggests pyramidal involvement irrespective of alterations in the sensory
that the dorsal columns do indeed convey impulses generating examination. Additionally, a number of patients symptomatic
the SEP. In 1980, additional studies of a relatively large number for dorsolateral sensory loss had completely normal SEP stud-
of patients with focal lesions producing clinical loss of proprio- ies. These investigators concluded that the best correlation with
ception and vibration were performed. Good correlation was
283
abnormal SEPs is clinical involvement of the pyramidal tract as
obtained between the clinical examination and abnormal SEPs, opposed to the dorsal column pathway.
implying that the SEP directly correlated to both the dorsal Despite a possible correlation of SEPs with pyramidal tract
column pathway and the modalities of vibration and proprio- pathway, it is generally accepted that SEPs are initiated in the
ception. In 1982, van Buggenhout et a l . examined 100 pa-
307
peripheral nerve by excitation of the large myelinated IA affer-
tients with multiple sclerosis and found that in all subjects w4th ent nerve fibers. These nerve fibers concomitantly comprise the
altered proprioception and vibration sensibilities, an abnormal dorsal column and spinocerebellar tracts. It has also been sug-
318
SEP was documented. In 1985, a large group of patients with gested that some extralemniscal afferent pathways may convey
cervical spondylosis were investigated with strict clinical crite- similar information to that noted above. - The majority of this
92 231
ria for sensory loss of all types. The best correlation between
331
work has been performed in the cat, however, and may not be
clinical findings and abnormal SEPs was in patients with altered directly applicable to humans. In addition to the human studies
sensation associated with the dorsal columns. described above, primate preparations in which the dorsal
Despite the above findings, a number of investigators have column was sectioned with a resultant loss in the cortical S E P
failed to confirm the strict association between abnormal SEPs strongly support dorsal columns mediating the SEP. Al- 66
and loss of particular sensory modalities. In 1983, Dimitrijevic though this has not been substantiated in humans following
dorsal cordotomy, intradural needle recordings support the fossa 2-3 cm superior to the insertion of the posterior border of
assertion that the dorsal columns are a primary pathway con- the clavicular head of the sternocleidomastoid muscle. An addi-
veying SEP impulses. 121122
tional set of electrodes is placed over the cervical spine with the
active or E-1 electrode positioned over the second, fifth, or sev-
enth cervical spinous process and referenced to an electrode (E-2)
NEURAL GENERATORS OF SEP located about the forehead region. - The seventh spinous process
6 7
WAVEFORMS is preferred by some investigators because of easily identifiable

bony landmarks. Finally, an E-l electrode is secured to the
326
A neural generator may be defined as nervous tissue such as scalp overlying the contralateral somatosensory cortex referenced
axon, cell bod^.or synapse with the capability of generating or to the previously noted forehead region. The preceding recording
sustaining an -action potential. A series of recording electrodes electrodes are most often utilized for the routine SEP examina-
located on.the peripheral or central nervous systems' pathway tion of most nerves excited in the upper limb.
along which the action potential propagates will document a
series of waveforms associated with the action potential as it Lower Limb
passes beneath the electrodes. The waveforms detected are re- When stimulating the tibial nerve at the medial malleolus, an
ferred to as traveling waves because they represent the propa- E-l recording electrode is placed over the tibial nerve at the
gating action potential. A collection of cell bodies and neural popliteal fossa and referenced to an E-2 electrode positioned at
synapses forming a nucleus can also generate a relatively well- the medial aspect of the knee. - A second E-l recording elec-
6 7
localized region of depolarization. The waveform produced by trode is usually located over the spinous process of the third or
this stationary collection of action potentials is a stationary fourth lumbar vertebra with an E-2 electrode on a spinous
wave. As you might expect, the waveform detected from this process 4 cm rostral, or located on the iliac crest contralateral to
type of neural generator does not propagate, but has a restricted the side of stimulation. A third E-l recording electrode is at LI
spatial distribution. Recording electrodes positioned along pe- or T12 and referenced to an E-2 electrode positioned 4 cm ros-
ripheral and central neural pathways can detect both stationary tral ly on a spinous process or to the iliac crest contralateral to the
and traveling waves, depending upon the electrodes' location stimulated limb. The final E-l electrode is secured to the scalp
(see below). just posterior to the vertex of the skull such that it is superior to
the somatosensory cortex for both lower llimbs and is referenced
RECORDING ELECTRODE LOCATIONS to the forehead region. - Occasionally, cortical recording elec-
6 7
trodes located just lateral to the vertex location may aid in the
The American Electroencephalographic Society (AES) rec- detection of waveforms not optimally observed at the vertex sec-
ommends a number of standard locations for electrode place- ondary to individual patient anatomic variations.
ment when recording SEPs. - In this section, we will refer to the
6 7
relevant active (E-l) and reference (E-2) recording electrodes' WAVEFORM GENERATORS: UPPER LIMB
generic anatomic positions for obtaining a waveform so that we
can discuss SEP generators. A discussion of the preferred elec- Erb's Point
trode and waveform nomenclature is provided in the subsequent Within approximately 9 or 10 ms following median nerve ex-
section. This chapter will only consider short-latency SEP re- citation, a relatively large waveform with a distinct negative peak
sponses. For upper limb studies, a short-latency response is de- is detected by the electrode at Erb's point (Fig. 9-3). This po- 116
fined as a waveform occurring within approximately 25 ms of tential is believed to represent the afferent neural volley passing
the stimulus in normal persons. - Normal lower limb short-la-
6 7
under the electrode at Erb's p o i n t . This waveform is first
48-51105
tency SEPs are usually generated within about 50 ms of the ex- generated at the lateral margin of the clavicle as substantiated by
citing pulse. SEPs resulting from stimulation of the median direct peripheral nerve recordings. Further proof of the potential
nerve at the wrist and tibial nerve at the medial malleolus are arising from the brachial plexus is that it is absent in patients
used as SEP examples to illustrate the major SEP waveforms and with lesions of the brachial plexus but present in patients with
their associated presumed neural generators. Mid-latency
117 3 2 6
cervical root avulsions. 14.15,65.114.142.174.176.196.322 ^ o t avulsion
r 0
and long-latency responses can also be recorded after peripheral spares the dorsal root ganglia, thus preserving the sensory pe-
nerve stimulaton; however, these longer latency responses are ripheral nerves conveying afferent median nerve impulses.
more variable and less clinically useful compared with the short- The specific fibers activated in the median nerve at the wrist
latency responses discussed below. Additional nerves and wave- that yield the Erb's point waveform are muscle afferents present
forms are discussed later. It is important to remember that the at that level. - - Certainly, sensory fibers subserving cuta-
54 246 284
latencies described below for particular waveforms may vary by neous sensation also add to the entire peripheral nerve response.
several milliseconds depending upon the exact site of stimula- The root levels represented in all of these fibers no doubt cover
tion, patient's height, limb temperature, and type of recording levels through most of the brachial plexus such as C-6 and C-7
parameters utilized. for cutaneous sensation, and C-8 and T-l for the muscular affer-
ents. It is also likely that the peripheral location of the Erb's
Upper Limb point electrode permits not only the recording of orthodromic
With respect to upper limb SEPs, one of the most reliable and sensory impulses but also to some extent the antidromic motor
easiest nerves to investigate is the median nerve at the wrist. impulses induced by the electrical stimulation of the median
Usually, separate examination of left and right upper limbs is nerve. Recall that peripheral nerve excitation produces action
required for a complete investigation. The first set of electrodes potentials in both motor and sensory fibers both orthodromi-
is typically placed at the left and right Erb's point. - The Erb's
6 7
cally and antidromically. Essentially, the recording at Erb's
point electrode contralateral to the side of stimulation is used as point represents a mixed nerve action potential containing both
the E-2 electrode. Erb's point is that region of the supraclavicular sensory and motor impulses.
STIM N20
0 10 20 30 40 50 msec
[ I I i i I
Figure 9-3. M e d i a n n e r v e S E P . Median nerve upper limb SEP utilizing a 4-channel recording technique.The b o t t o m t r a c e signifies an Erb's
point recording (EP = N 9 ; EP2-EPI) demonstrating t h e peripheral nerve volley.The spinal recording (C5S-FZ) demonstrating the waveform
recorded at this location.The third trace from the bottom is a far-field recording (noncephalic reference {C3'-EPI}).Top trace shows the typical
negative/positive cortical potential with a cephalic E-2 electrode (C3'-FZ). (From Spehlmann R: Evoked Potential Primer. Boston, Butterworth
Publishers, 1985, with permission.)
Cervical Spinous Process electrode array over the sternocleidomastoid muscle ipsilateral to
A prominent negative potential is usually detected by an elec- the side of stimulation reveals a traveling wave just after the Erb's
trode located at the C2, C5, or C7 spinous process following point potential. This traveling PPV wave is believed to represent
median nerve excitation (Fig. 9-3). Depending upon the instru- the afferent impulses propagating through the proximal portion of
ment's sensitivity and degree of background noise, the main the brachial plexus and the cervical nerve roots (Fig. 9-4, Table 9-2).
peak of this potential may occasionally be preceded and fol- A traveling negative waveform (DCV), first peak of the occa-
lowed by small negative peaks arising from the main negative sionally tri-peaked negative cervical spine potential, can be se-
potential with latencies of approximately 12 ms and 14 ms, re- quentially recorded over the cervical spinous processes at about
spectively. Considerable research has attempted to define the 11 or 12 ms (Fig. 9 - 3 ) . This potential is most likely the N l 1
n6 B 1
generators of these negative peaks.

In attempting to define the neural generator(s) responsible for
the waveforms recorded from the cervical spinous processes
noted above, investigators have focused on five areas of study.
These five areas include the (1) waveform's spatial distribution,
(2) generator dipole spatial orientation, (3) associated action po-
tential's refractory period, (4) effect of focal lesions, and (5)
correlation with animal data. A waveform's spatial distribu-
270
tion allows one to determine if the associated neural generator is

propagating or stationary. The orientation of the generator's
dipole, as determined by its recorded waveform's peak polarity
orientation, suggests if it is pointing in a particular direction
e.g., superior/inferior or anterior/posterior. Propagating action
potentials have refractory periods considerably shorter (3-A ms)
than stationary generators (6-10 ms). Patients sustaining well-
localized lesions of various portions of the neuraxis provide in-
vestigators with valuable information regarding potential
candidates for anatomic neural generator sites. Finally, experi-
mental lesions placed at different locations in animals can also 2ms
contribute to possible generator site identification.
When electrodes are placed over the cervical spine and refer- Figure 9-4. T r a v e l i n g w a v e . Cervical spine t o anterior neck m o n -
enced to the anterior portion of the neck, three negative wave- tage revealing the proximal plexus volley (PPV), dorsal column volley
forms occurring after the Erb's point potential can be recorded (DCV), and the cervical potential (CERV N13). T h e Erb's point p o t e n -
following median nerve stimulation (Fig. 9-4). These three tial is also shown. (From Emerson RG, Seyal M, Pedley TA:
waveforms are the proximal plexus volley (PPV), dorsal column Somatosensory evoked potentials following median nerve stimulation.
volley (DCV), and the cervical p o t e n t i a l . - - ° -
67
A linear
79 8 83n6
I.The cervical components. Brain 1984; 107:169-182, with permission.)
Table 9-2. Short Latency SEP Characteristics

Recording Refractory Voltage Field Presumed
Waveform Site Period (ms) Distribution Generator
MEDIAN NERVE SEP
PPV Ipsilateral neck N o t determined Traveling wave Proximal cervical plexus volley
DCV Cervical spine Less than 3—4 Traveling wave Dorsal column volley
NI3 Cervical spine 8-16 Transverse oriented dipole Dorsal gray of spinal cord at
(spinal); partly axial r o o t entry zone; nucleus
(nucleus cuneatus) cuneatus
TIBIAL SEP
PPV Sacral spine Traveling wave Lumbosacral plexus and roots
DCV Lumbar to cervical Traveling wave Dorsal column volley
spine
N22 T-12 and lumbosacral 6-10 Transverse oriented dipole Dorsal gray of spinal cord at
spine at r o o t entry zone
N29 Upper cervical spine 6-10 Axial oriented dipole Nucleus gracilis
PPV: proximal plexus volley; DCV: dorsal column volley.
Modified from Seyal M Gabor AJ: Generators of human spinal somatosensory evoked potentials. J Clin Neurophysiol 1987;4:177-187.
peak noted in Figure 9-3 preceding the major N13 peak. There Lesion studies demonstrate that the 13a potential is absent fol-
is approximately a 1-ms difference between the caudal and ros- lowing obliteration of the dorsal horn cells, while the 13b potential
tral aspects of the cervical spine with respect to the traveling remains unchanged. Additionally, lesions in the cuneate nucleus
178
wave's peak latency for this potential. The refractory period of resulted in an absence of the 13b waveform but preservation of the
this waveform is short, several milliseconds, suggesting that its 13a potential. The generators of the 13a and 13b potentials, there-
origin is from a conducting neural volley, presynaptic impulse, fore, are believed to be the dorsal gray of the cervical cord and the
and does not involve post-synaptic nuclear generators (Table 9¬ cuneate nucleus, respectively. Combined studies involving stimula-
2). This traveling waveform most likely represents the afferent tion of lower limb nerves and their effect on the two 13-ms peaks
volley propagating rostrally in the dorsal columns of the cervi- further substantiated the two waveforms generators noted
cal spinal cord. - -
62 82
The dorsal root entry zone - "-
83 n 6 , 5 8 2 n 174 2 321
above. 268-271
Additionally, hypothermia reduces the amplitude of
and dorsal horn have also been postulated as sites_of origin
216 2 6 6
this waveform, suggesting that synaptic transmission is involved in
for the waveform. This potential is designated Nl 1/N12 by its production because a traveling wave would have increased in
some investigators. Hypothermia studies demonstrate an in- magnitude similar to that found in the dorsal column studies. The 298
crease of the waveform's amplitude similar to that found in dorsal columns at the superior cervical, - mid-cervical, - - 204 303 324 326 327
peripheral nerves, supporting the dorsal column volley proposi- and foramen magnum ^ levels were also proposed as possible
140 11
tion, i.e., propagating neural volley. - If the generator of this

27 298
generator sites for the N13 potential (Table 9-2). The negative peak
waveform were a post-synaptic potential in a collection of with a 14-ms latency constituting a portion of the cervical spine po-
nuclei, e.g., brain stem nuclei, the amplitude should have de- tential is believed to arise from the afferent neurafyolley propagat-
creased, as this is the anticipated response of these cells to a re- ing in the medial lemniscus. - - - - The N13 and N14 peaks82 83152 153 168 292
duction in temperature. likely represent generators below and above the foramen magnum,
The second negative peak of the major negative spinal poten- respectively. The various subcomponents of the spinal potential
327
tial occurs at about 13 ms and is essentially a stationary potential can often be best observed using a noncephalic referential mon-
demonstrating its greatest amplitude over the cervical root entry tage, e.g., an E-2 electrode located on the ear lobes.
zone (Fig. 9-3). An experimental electrode montage over the
82
anterior cervical region records a positive waveform with the Scalp

same latency as the negative cervical potential. Esophageal 116117
A large negative potential usually presenting a peak latency at 19
electrodes have also documented this positive waveform. - 83 86
or 20 ms after median nerve wrist stimulation is typically recorded
These findings suggest that a dipole is present in the cervical with a scalp electrode over the contralateral somatosensory cortex
region oriented such that the negative pole is directed posteriorly (see Fig. 9-3). The negative peak is then followed by a positive de-
139
while the positive pole faces anteriorly, substantiating a station- flection with a peak latency approximating 22 ms. One proposal sug-
ary neural generator producing the N13 potential. gests that the negative potential originates in the primary relay nuclei
Investigations examining the refractory period of this poten- of the thalamus, whereas the positive potential reflects somatosen-
tial have demonstrated a relatively long refractory period, sug- sory cortical activity induced by the arrival of the afferent im-
gesting the involvement of synaptic transmission (Table 9-2). 166
pulses. - - - - 19.227.234.235 jh\ j
49 51
52 140 2
j considered less accurate
s v e w s n o w
These findings have led some investigators to suggest that the than the suggestion that the large negative and subsequent positive
site of generation of this waveform is within the interneurons of peaks both reflect cortical arrival of the impulses generated at the pe-
the cervical cord's dorsal gray matter. - Further refinements in
82 83
riphery. 3.79.80.83.159.220221,280.321 Legjong involving the postcentral sulcus
recording techniques substantiated that the 13-ms negative peak support the view of the somatosensory cortex producing the large
may also display a bifid character, 13a/13b, implying two dis- negative and positive potentials at 20 ms and 22 ms, respectively.
tinct generators. - - 2
- ' These two generators appear to
82 174196 281 298
Occasionally, a distinct negative peak at 18 ms, 17 ms in some
have a spatial separation resulting in a temporal difference of 1 studies, superimposed on the larger 20-ms negative waveform
ms between the 13a and 13b peak latencies. may be observed. This 18-ms negative potential has a rather
N45
STIM
Figure 9-5. Tibial n e r v e SEP. Four-channel recording of a P37

tibial nerve stimulation. Bottom trace is the popliteal fossa
recording (PF) showing the peripheral nerve volley as it passes
beneath the recording electrode.Adjacent trace demonstrates
the waveform recorded at the lumbar spine montage delineat-
ing the cauda equina potential.Third trace from bottom shows
-TV
2|iV
t h e thoracic spine potential demarcating central nervous
system entry of the potential volley.The top trace shows cor- x +
tical arrival at t h e s o m a t o s e n s o r y c o r t e x (CZ'-FpZ ). (From

1
Spehlmann R: Evoked Potential Primer. Boston, Butterworth

10 20 30 40 50
I I _JL_
STIM
wide distribution over the scalp, suggesting it arises from a (Fig. 9-6) - The conduction velocity of this negative waveform
268 271
rather deep structure capable of projecting its waveform some- is essentially constant along its course. This potential merges with,
what uniformly over the cranium's surface. This potential was and then departs from, the stationary waveform recorded at the
initially suggested to arise from neural activity in the thalamus twelfth thoracic spine (see below), indicating a traveling wave as-
and/or thalamocortical radiations. The thalamus appears to be
83
cending the spinal cord with a conduction velocity approximating
a reasonable structure to generate this waveform given that the 61 m/s. A refractory period of 1-2 ms is noted for this traveling
268
18-ms negative potential precedes that of the cortical waveform. volley when determined over the sacral or thoracic region (Table 9¬
Various lesions in patients with thalamic pathology support the 2). A short refractory period suggests that synaptic relays are not
view that the 18-ms negative peak is associated with the thala- associated with the propagation of this wavefront from the lower
mus or its projections to the cerebral cortex. - - 221 303 325
limb to the thoracic aspects of the spinal cord. The negative wave-
form, therefore, recorded over the third lumbar vertebra most likely
WAVEFORM GENERATORS: LOWER LIMB represents the traveling peripheral nerve volley (cauda equina) as it
progresses through the lumbosacral plexus and nerve roots (Table
Popliteal Fossa 9-2). - ' This same propagating potential, when detected over
64 76 175
Similar to the Erb's point recording in the upper limb, the the thoracic spines, is believed to be the traveling wavefront of elec-
popliteal fossa electrode is positioned directly over a peripheral trical activity ascending the dorsal columns.
nerve pathway. A clearly delineated negative potential that may
be preceded and followed by smaller positive deflections is usu- Twelfth Thoracic Vertebra
ally obtained with this electrode montage (Fig. 9-5). The nega- The T-12 recording electrode documents a well-defined nega-
tive peak latency typically occurs at approximately 7-10 ms, tive potential with a peak latency of approximately 21-22 m s
depending upon the distance between the site of stimulation and (Figs. 9-5 and 9-6). As always, the exact latency of this poten-
recording and nerve conduction velocity. - -' This potential
6 7 79282
tial is dependent upon the length of the individual's leg and po-
arises from the afferent neural volley traveling in the tibial nerve sition of the stimulating electrodes. Unlike the potential
as it passes beneath the recording electrodes. recorded with the L-3 electrode, the T-12 waveform has a rela-
tively restricted spatial distribution. The magnitude of this p o -
268
Third Lumbar Vertebra tential maximizes within 5-15 cm rostral to the L-4 spinous
An electrode located over the third or fourth (some investiga- process. Locations either more superior or inferior result in a
258
tors prefer L4 because it is easy to locate) lumbar vertebra, with diminution of this potential's amplitude.
an E-2 electrode several levels more rostral, records a small Recordings over the anterior abdominal wall and thorax
negative potential that may occasionally be associated with a demonstrate a positive potential with the same latency and tem-
preceding and/or trailing positive deflection (Fig. 9-5). The 62
poral aspects as that of the negative waveform recorded from
negative peak latency normally ranges between 17 and 21 ms, the thoracolumbar spines. Additionally, identical amplitude
267
corresponding to the length of the subjects' lower limb and site characteristics to those observed for the 22-ms negative poten-
of stimulation. Insight is gained into spinal generators by con-
282
tial are found for this positive potential. Similarly, findings of a
sidering the five investigative methodologies utilized for cervi- positive potential with the above aspects are demonstrated with
cal spine generators previously described. an esophageal electrode. The dipole orientation of this potential
84
Sequentially placed electrodes along the spinous processes of appears to be oriented in the transverse direction (Table 9-2).
the entire spinal column demonstrates a negative waveform with a These findings support a well-localized potential originating
linear increase in peak latency from the sacral to the cervical region from the spinal cord about the T-12 region.
AFFERENT The refractory period of the plexus and dorsal column volley is
VOLLEY M22
on the order of 1-2 m s . On the other hand, the refractory
269
CAUDAL* \
period of the negative 22-ms potential approximates 6-10 ms,
strongly suggesting synaptic transmission. This would also ac- 267
count for the somewhat delayed appearance of this potential fol-

lowing the ascending volley, as a temporal delay is necessary to
activate the interneurons because of interposed synapses.
'The combination of the animal and human studies noted
above strongly suggests a site of origin for the negative potential
recorded over the T-12 spinous process. It appears that one can
conclude with some assurance that the negative potential occur-
ring approximately 22 ms following tibial nerve excitation origi-
nates from the dorsal gray's interneuronal population of cells
associated with the root entry zone of the tibial nerve's compart-
ment fibers in the caudal portion of the spinal cord. - - - 75 88 243 269
Second Cervical Potential

Although not routinely performed, a recording electrode
over the second cervical spine records a stationary potential
following bilateral tibial nerve stimulation. - This potential 268 270
demonstrates a latency of about 29 ms following activation of

the tibial nerve at the medial malleolus. The potential is first
noted at the lower cervical region and cannot be detected
above the inion (Fig. 9-7). It is also possible to distinguish the
29-ms negative potential from the ascending dorsal column
volley. Animal studies reveal that the orientation of the nega-
tive potential is a x i a l . A rather long refractory period is
177
noted for this cervical potential. - All of this strongly sug-

268 270
gests that the origin of the 29-ms negative potential is most

likely in the nucleus gracilis.
5 MSEC Scalp
An E-l electrode located 1-2 cm posterior to the skull's
Figure 9-6. T i b i a l n e r v e t r a v e l i n g w a v e . Recordings from a
vertex referenced to the forehead region demonstrates a rela-.
series of electrodes located over the thoracolumbar spine document-
tively large positive/negative complex (see Fig. 9-5). Following
ing the ascent of t h e neural volley following tibial nerve excitation.
tibial nerve activation, this waveform demonstrates mean posi-
N o t e h o w the propagating wave approaches, fuses, and then departs
tive/negative peak latencies of 37 ms and 45 ms, respec-
from t h e stationary potential, which is maximum at L I / T I 2 . (From
tively. - - These values obviously vary with the subject's
86J79 266 281
Seyal M, Gabor AJ:The human posterior tibial somatosensory evoked

age and height (see below). In some instances, the positive
potential: Synapse dependent and synapse independent spinal compo-
potential is preceded by a small negative potential with a peak
nents. Electroencephalogr Clin Neurophysiol 1985;62:323-33 I, with
latency approximating 32 m s . - - Additionally, the posi-
28 106,248 308
permission.)
tive/negative complex is typically followed by large and somewhat
variable potentials (long latency potentials), which will not be
discussed in this chapter. 132
In animals, stimulating the peripheral nerves or dorsal roots The spatial distribution of the positive/negative complex is
produces a triphasic waveform observed over the dorsal aspect of primarily localized to the postcentral gyrus region. - The 138 179
the spinal cord. 13.23.31.36.130.156.157 v e f o r m is thought to repre-

wa amplitude of the cortical waveform is slightly larger over the ip-
sent the afferent neural volley as it travels past the recording elec- silateral scalp. This suggest that the neurons generating the elec-
trode. A somewhat slower negative wave follows this triphasic tric field from the contralateral tibial nerve stimulation is
potential and is believed to originate from the ascending volley's oriented such that the potential gradients are larger over the ip-
excitation of the spinal cord's dorsal gray interneuronal dendrites silateral hemisphere. - - Occasionally, the major positive
65 304 307
via the afferent fibers' collateral branches. This negative wave- waveform of the cortically recorded tibial SEP may be larger
form is of greatest amplitude at the level of the dorsal root entry over the contralateral cortex. - This most likely results from
205 307
zone. A positive potential is noted when an electrode is located the leg area of the cortex being located lateral to the superior lip i
ventral to the dorsal root entry zone in these animal preparations. of the interhemispheric fissure. In this case, the paravertex
116
The similar findings in both animals and humans suggest that the recording electrode locations (C-f and C-2') instead of CZ' may
22-ms negative potential has a similar origin for both groups. 269
better define the cortical SEP waveforms. A number of investi-
Refractory period studies add further support to the above gators have employed both cephalic bipolar and noncephalic
data, suggesting a synaptic involvement in generating the nega- referential electrode montages to define the site of origin for the
tive potential at this anatomic location (Table 9 - 2 ) . As pre-
130298
positive/negative waveform c o m p l e x . - - Although the 84179 210 266
viously discussed, large-diameter myelinated axons demonstrate exact generator of the cortical potential remains controversial,
rather short refractory periods, whereas conduction involving the postcentral gyrus area of the cortex subserving the lower
synaptic transmission has somewhat longer refractory periods. limb is generally believed to be its site.
R E C O M M E N D E D STANDARDS FOR
R E C O R D I N G S H O R T - L A T E N C Y SEPS
Despite a number of well-founded recommendations by the

AES regarding an attempt to standardize electrode site and wave¬
' form morphologic nomenclature, instrumentation parameters,
recording practices, interpretive qualifications, and reference data
collection, - confusion persists in the SEP literature. This confu-
6 7
sion primarily arises from a lack of agreement regarding electrode

placement, filter settings, and nomenclature regarding the manner
in which to designate waveform latencies, i.e., peak latency identi-
! fication and amplitude measurement. In attempting to compare
data from one laboratory to another, obvious difficulties arise with
respect to the occurrence of particular waveforms as they relate to
specific neural generators. This chapter will attempt to follow the
AES guidelines as closely as possible unless otherwise stated. It is
hoped that the reader finds that the authors' justification for devia-
tions from the presented recommendations result in clarity and do
not add to the already confusing literature regarding SEPs.
Figure 9-7. Tibial nerve traveling waves. Recording along t h e
spinal column from sacral region to skull demonstrating various wave-
NOMENCLATURE FOR RECORDING forms along the neuraxis following bilateral tibial nerve stimulation. N o t e
E L E C T R O D E SITES that the traveling neural volley ascends along the spinal axis and inter-
sects the two stationary potentials at the thoracolumbar junction and
The placement of E-l and E-2 electrodes influences the de- cervical region. (From Seyal M, Kraft LW, Gabor AJ:A cervical synapse de-
tection of different waveforms. In general, when the E-l and E- pendent somatosensory evoked potential following posterior tibial n e r v e
2 electrodes are relatively close to each other, a bipolar stimulation. Neurology 1987;37:1417-1421, with permission.) -
montage is created. When these 2 electrodes are located close
to a neural generator, a relatively small potential devoid of elec-
trical events occurring at some distance from the recording site the main cortical waveform. The site of origin for these far-field
is detected. The reason for the absence of events at a distance, potentials remains controversial and is discussed at the end of
i.e., from the far-field, is because this electrical activity is elimi- this chapter, as these potentials are of little clinical utility at the
nated as a common mode signal by the differential amplifier present time with respect to identifying pathology involving the
(see Chapter 3). The net result is the detection of the difference peripheral or somatosensory neural pathways.
between the two rather closely spaced electrodes with elimina-
tion of potentials observed simultaneously by both recording ELECTRODE SITE DESIGNATIONS
electrodes at a distance. Of course, if one could place both the
E-l and E-2 electrodes in the same location, no waveforms Upper Limb Stimulation
would be detected, as all signals would then be common to both Erb's Point. When performing upper limb SEP studies irre-
and subsequently rejected by differential amplification. spective of whether the median or ulnar nerves at the wrist or
An E-2 electrode placed on the scalp and connected to an E-1 cutaneous digital branches of these nerves are excited (or any
electrode on the scalp or cervical spine is referred to as a other upper limb pure sensory or mixed nerve, for that
cephalic reference montage. - In this type of arrangement, cor-
6 7
matter), a standard number of recording electrodes are uti-
281
tical neural generators producing the anticipated waveforms lie lized. As previously noted, the first set of electrodes are located
relatively close to the E-l recording electrode. It is important to over the left and right Erb's point (Table 9-3). An Erb's point
67
keep in mind, however, that the E-2 electrode is most likely recording serves several functions: (1) documents that a periph-
recording a volume-conducted response from the activated corti- eral nerve volley has reached this proximal anatomic location,
cal neural generator and, therefore, is not actually electrically (2) allows one to evaluate the response with respect to slowing
silent. The end result is an SEP waveform with some activity because of limb length or decreased temperature, and (3) per-
arising from the E-2 electrode. This is the type of SEP we will mits the evaluation of peripheral nerve slowing as a cause for a
discuss in the majority of this chapter. As one might anticipate, it central delay. The two electrodes positioned 2-3 cm superior to
is also possible to position an E-l recording electrode on the the clavicle and just lateral to the clavicular head of the stern-
scalp over a portion of the somatosensory cortex of interest (e.g., ocleidomastoid muscle are designated EP1 and EP2 for the left
upper or lower limb location) and reference this electrode to the and right sides, respectively. - In general, electrodes placed on
6 7
contralateral limb with respect to stimulation. For example, one the body are given odd numbers when located left of midline
can place an E-l electrode on the right parietal somatosensory and even numbers when right of m i d l i n e . - - These t w o
46 47169 231
cortex representing the upper limb when activating the left electrodes are typically combined for a left-right or right-left
median nerve at the wrist, and use the right wrist as the location Erb's point montage. The Erb's point electrode defined as E-l is
for the E-2 electrode. This set-up is referred to as a noncephalic ipsilateral to the stimulated limb. Should left upper limb nerves
reference or referential montage, which often produces more be stimulated, EP1 is the E-l electrode and EP2 is the E-2 elec-
"noise" than the bipolar montage. - In this instance, a number of
6 7
trode. In this case, EP1 is referenced to EP2. With right upper
positive potentials called far-field potentials are recorded preceding limb activation, EP2 is the E-l electrode, while EP1 is the E-2
Table 9 - 3 . SEP Recording Electrode Placement cervical vertebra and then counting the appropriate number of
Channel # E-1 Electrode E-2 Electrode spinous processes superiorly. The E-2 electrode is placed on the
scalp (see below for exact position). Irrespective of which spin-
Upper Limb Montage: 4 Channels
ous process is used, essentially the same potentials are observed.
1 C37C4' FpZ' The major negative waveform at 13 ms with superimposed 11-
2 C7S FpZ' or 12-ms, and 14-ms peaks arise from closely spaced generators
3 EP (ipsilateral) EP (contralateral)
(see above) presenting with essentially the same latency from the
4 AF Olecranon
seventh to the second cervical spinous process. The amplitude of
Upper Limb Montage: 2 Channels each may vary depending upon the electrode location, with pos-
1 C37C4* EP (ipsilateral) sibly larger amplitudes recorded at C2S, as one is somewhat
2 C7S FpZ' closer to the neural generators.
266 2 6 9
Lower Limb Montage: 4 Channels Scalp/10-20 International System. The recording of corti-
cal generators require the electrodes to be located on the scalp.
I CZ* FpZ' It is necessary at this juncture to examine the placement of scalp
2 TI2orLI 4 cm superior o r iliac crest electrodes as defined by the International Ten-Twenty (10¬
3 L3 o r L4 4 cm superior o r iliac crest 20) System. Approximately 10 years after the widespread
170 2 3 2
4 PF Medial knee clinical application of electroencephalography, it became obvi-

Lower Limb Montage: 2 Channels ous that a standardized number of scalp recording sites were
i cr FpZ' necessary to promote international understanding when dis-
2 TI2orLI 4 cm superior o r iliac crest cussing EEG pathology and possible neural generators. As a
result, the International Federation of Societies for Elec-
troencephalography and Clinical Neurophysiology proposed the
electrode. Although this is a noncephalic referential montage, 10-20 system of electrode placement in 1957. At the present
170
few far-field potentials are noted because most of the easily de- time, this standard is essentially accepted throughout the world
tectable far-field SEPs are generated at or proximal to Erb's with respect to EEG electrode placements. The underlying prin-
point. The waveform detected by the Erb's point electrode is a ciples governing the development of the 10-20 system are (1)
relatively large typically triphasic waveform with a prominent electrode sites are located on the scalp by using measurements
negative spike of about 10 ms (see Figs. 9-3 and 9-4). based on a percentage of the skull's size with respect to stan-
Cervical Spinous Process. A single spinal electrode is usu- dard landmarks (usually 10 or 20%) and not absolute distances,
ally positioned at one of several locations over the second, fifth, (2) the entire skull is represented in the system, and (3) elec-
or seventh cervical spinous process and designated C2S, CSS, trode sites are labeled according to assumed cortical structures
or C7S, respectively (Table 9-3). - - These positions are easily
6 7 325
over which the electrodes are placed. - The International 10¬
170 232
located by first palpating the spinous process of the seventh 20 System's true value lies in its flexibility and ease of use.
cz
Figure 9-8. 10-20 international system. 10-20 international scalp electrode locations defining the prime designations for preferred SEP elec-
trode locations. (From Nuwer MR: Recording electrode site nomenclature.] Clin Neurophysiol 1987;4:122-133, with permission.)
Four standard landmarks are required to locate all positions M o n t a g e R e c o m m e n d a t i o n s . Should the practitioner have
necessary for SEP purposes. - The n a s i o n (bridge of the
170 232
the good fortune of possessing a 4-channel SEP instrument,
nose) and i n i o n (posterior bony protuberance over the inferior each channel can record different waveforms (Table 9-3).
aspect of the occiput) are the two anatomic landmarks along the Channel 1 may contain an E-l scalp electrode (C3' or C4') refer-
skull's midsagittal plane. That region where the ears attach to enced to the midfrontal region, i.e., FpZ' (Fig. 9-3). Although
the skull, or just anterior to the tragus, form the second pair of the AES recommends FZ as the E-2 location, - as previously
6 7
landmarks in the frontal plane. The 10-20 portion of the mea- noted, FpZ' is preferred by many for ease of localization.
surement system refers to 10% and 20% of the total distance be- Irrespective of location utilized, the SEP latency and to some
tween these two sets of landmarks. There is a slight difference extent amplitude should remain essentially the same. One may
regarding the intent between the recording of EEGs compared also use the earlobe contralateral to the side of stimulation as an
with that of SEPs. Specifically, when obtaining SEPs, one E-2 site. Channel 2 may contain C3' or C4' as the E-l electrode
wishes to be as close as possible to the generators of the SEP, and the contralateral EP site as the E-2 electrode (Table 9-3, Fig.
i.e., the somatosensory cortex of either the upper or lower limb. 9-3). This montage is primarily used to detect far-field poten-
As a result, a minor .modification of the standard EEG electrode tials and will not be discussed at this time. Channel 3 is sug-
locations is required to record SEPs optimally. It is necessary, gested to utilize the cervical CSS or C2S (C7S) referenced to
therefore, for those who wish to perform SEPs to become thor- the scalp's FpZ'. Finally, channel 4 uses EP ipsilateral to the
oughly familiar with the 10-20 system pertinent to the recording side of stimulation as the E-l electrode while the contralateral
ofSEPs(Fig.9-8). EP serves as the E-2 electrode (Fig. 9-3). The mastoid process
Locating the optimal recording positions for upper limb SEPs may also be used as a E-2 site and may be useful for intraopera-
can be performed in only a few minutes by the experienced tive studies.
practitioner. A tape measure is extended between the nasion and If an instrument with two channels is used, it is still possible
inion with the mid-point marked on the scalp. The tape measure to obtain all the pertinent information. Recall that we are pri-
is then positioned and drawn snugly between the regions where marily interested in the waveforms arising from just three
the ears join the scalp. That point at which the line extending recording sites, i.e., Erb's point, cervical spine, and somatosen-
between the two ears crosses the previously defined mid-point sory cortex. Because of just two channels, one of the electrodes
of the sagittal line joining the nasion and inion designates the must perform double duty by not only acting as an E-2 elec-
vertex of the skull, also called the C Z electrode site (Fig. 9-8). trode, but also recording one of the desired waveforms in its
Twenty percent of the total coronal distance between the two "reference" capacity, i.e., producing an inverted response. With
ears extending from CZ on this inter-ear line toward either ear two channels, the guiding principle is to use a E-2 electrode that
marks the electrode sites known as C3 and C4. Recall that odd records a waveform sufficiently displaced in time so as to not
numbers designate left-sided body electrode positions and even overlap with an E-l electrode's near-field potential. The two
numbers right-sided electrode locations; the modifier "z" indi- most-displaced waveforms originate from the most physically
cates midline electrodes. Therefore, C3 approximately over-
170
separated electrode locations. These are Erb's point (EP) and
lies the left and C4 the right somatosensory cortex, subserving the scalp (C3' or C4'). Channel 1, therefore, uses C37C4' as the
the right and left upper extremities, respectively. These two E-l electrode site and EP ipsilateral to the side of stimulation as
electrode sites are standard EEG recording locations. To locate the E-2 electrode. There are two waveforms of interest on chan-
the recording electrode for SEPs optimally, one must move nel 1; the first is an inverted Erb's point potential followed by
these two electrodes 2 cm directly posteriorly from their previ- the typically appearing cortical potential (Fig. 9-9). As the
ously noted positions. A "prime" designation is then given to waveform recorded by EP is clearly defined, inverting its wave-
these newly defined sites, such that standard upper limb scalp form should pose no difficulty in finding its peak latency.
SEP electrodes now become C3* and C4' (Table 9-3, Fig. 9-8). Channel 2 utilizes the standard C2S (C5S or C7S) referenced to
As previously noted, C3' and C4' are the E-l electrode record- FpZ'. With this montage there are two waveforms of interest on
ing positions for the contralaterally excited median nerve. C3' is channel 1, and a single waveform on channel 2 (Fig. 9-9). The
utilized for stimulation of the right upper limb nerves and C4' cortical potential may be somewhat larger than if a cortical E-2
for the left upper limb nerves. is used; however, this should not pose a problem, particularly if
The E-2 electrode is similarly located utilizing a modification side-to-side comparisons are made. A possible alternative mon-
of the 10-20 system. Two midline 10-20 system electrode loca- tage is C37C4' referenced to FpZ' on channel 1 and C7S to FpZ'
tions must first be located. Ten percent of the total distance be- for channel 2. In this situation, if the cervical potential is
tween the nasion and inion, superior to the nasion, constitutes normal, the Erb's point potential may be of limited value unless
an electrode site called F p Z . Twenty percent of the total dis- one is particularly interested in the peripherally recorded re-
tance between the above two sites, nasion and inion, from CZ sponse such as when evaluating possible brachial plexopathy or
toward the nasion designates the recording site F Z (Fig. 9-8). root avulsion.
Halfway between FpZ and FZ is the E-2 location preferred by
some investigators for SEPs and is called F p Z ' . An alternative
6 7 Lower Limb Stimulation
and commonly used E-2 site is FZ. Any one of these preceding Popliteal Fossa. In the performance of lower limb SEP stud-
sites can be used for all cephalic referential recordings as it ap- ies, a peripheral nerve waveform providing a "reference point"
proximates the frontal hairline and is easy to locate. Even in regarding the status of the peripheral nervous system is sug-
persons with frontal balding, the former hairline is still easily gested. - - An Erb's point recording electrode serves this pur-
6 7 232a
noted by having the patient contract the frontalis muscle. That pose in the upper limb SEP recordings ipsilateral to the side of
region of the scalp that does not "wrinkle" and is just superior stimulation. In lower limb tibial nerve SEP investigations, an E-
to the frontalis muscle is about where the former frontal hairline 1 electrode located over the tibial nerve approximately 4 cm
was. In upper limb SEPs, therefore, C3' and C4' are referenced proximal to the popliteal crease midway between the tendons of
to FpZ' in the so-called cephalic r e f e r e n t i a l montage. the semimembranous/semitendinous muscles medially and the
N20 iliac crest electrode as the reference for the same reasons previ-
ously stated. It is important to note that both of these spinal elec-
trodes can be used for both left and right lower limb stimulations.
It is important to recognize that the lumbosacral spine potentials
may normally be unobtainable in older or obese persons, and at
times in healthy young individuals; hence absence of the response
should not be considered abnormal.
Scalp. An E-1 electrode is attached to the scalp to record a
response from the somatosensory cortex associated with the
fibers activated in the peripheral portions of the tibial nerve, i.e.,
that portion of the cortex subserving the lower limb. The best
placement for this electrode is 2 cm posterior to the vertex of
the skull (CZ) utilizing the 10-20 system and is called CZ' (Fig.
9_5) 6,7,52.77,282 j ^ g
s a m£_2 FpZ', as that used for upper limb
e 7
SEP studies can also be utilized for lower limb SEP examina-
tions. Because the E-l electrode (CZ') is midline and can record =
the volume-conducted cortical responses from the inter-hemi- [
Figure 9-9. M e d i a n n e r v e 2 - c h a n n e l SEP.Waveforms recorded spheric somatosensory cortex, the same scalp montage is used
from median nerve stimulation utilizing a 2-channel instrument. for left and right tibial nerve excitation similar to the spine elec-
Channel I (A) employs a noncephalic montage (C37C4'-EP) and dis- trodes (Fig. 9-5).
plays an inverted EP potential and t h e e x p e c t e d cortical potential. Montage Recommendations. In a 4-channel recording in-
Channel 2 (B) detects the cervical spine potential (C7S-FpZ'). N o t e strument, channel 1 is recommended to display the information
that it is possible t o record all potentials of interest successfully with a obtained from CZ' referenced to FpZ' (Table 9-3, Fig. 9-5). - 6 7
2-channel instrument. Channel 2 can record T12S or L I S referenced to the electrode 4

cm superior to this location or the iliac crest. Channel 3 may use
tendon of biceps femoris laterally provides such a peripheral the recording of L3S or L4S referenced to the electrode placed
waveform (Fig. 9-5). - This electrode site is designated PF
76 281
4 cm rostral to this location or the iliac crest. The fourth channel
(Table 9-3). - Should any doubt exist about the proper position-
6 7
records the PF to medial knee waveform. In a 2-channel instru-
ing of the PF electrode, an electrical stimulus delivered to the ment, it is rather difficult to obtain useful recordings from all
proper PF site results in foot plantar flexion and flexion of the four electrode positions despite various montages because of .
toes. The E-2 electrode is usually placed on the medial surface electrode noise. The paraspinal muscle is particularly difficult j
of the knee. to relax, but electrodes relatively close to each other help elimi-
Thoracolumbar Spinous Processes. The third electrode is nate muscle artifact. Attempting to reference T12S(L1S) and
located over the third or fourth lumbar spine and referred to as L3S(L4S) to the scalp or knee may yield technically unsatisfac^
L3S or L4S (Table 9-3, Fig. 9-5 ). - - - - This L3S electrode is
6 7 52 77 282
tory recordings. The practitioner may need to record two sets of
positioned by first locating the fourth lumbar spinous process, electrodes at a time, although this will require twice the number
which is palpated midway between the iliac crests. Proceeding of stimuli. Another possibility may be to forego the PF and one
rostrally 1 spinous process allows one to locate the L3S elec- of the spine sites, e.g., L3S. If the spinal potential is normal, PF
trode site rapidly. The E-2 electrode site for L3S is recom- is of questionable value. Of course, if there is an abnormality in
mended to be situated over the vertebral column 4 cm superior the spinous process electrode chosen, it is then necessary to ex-
to L3S. The waveform recorded from L3S permits one to calcu- plore the possibility of a peripheral nerve versus plexus lesion.
late the interval between PF and L3S, yielding the conduction In this case, the needle electromyographic examination is the
time from the knee to the cauda equina region. As a result, it procedure of choice in delineating the lesion and not the SEP.
may be clinically relevant to obtain this potential in patients sus-
pected of having peripheral nerve lesions. Some investigators
prefer the fourth lumbar spinous process because it is easy to SEP WAVEFORMS
locate, and this site is designated L4S. Essentially the same
waveform is observed at this location as that detected as L3S. Somatosensory evoked potentials consist of waveforms evoked
Although the above are certainly acceptable positions for the typically by an electrical stimulus applied to a peripheral nerve.
two noted electrodes, this recording tends to generate spinal wave- To objectively document the integrity of the neural pathway
forms that may be rather difficult to detect because of their small under study and compare a particular potential with a reference
size. This is expected given that the two electrodes are close to- data base, it is necessary to characterize the evoked potential's
gether and are rejecting common mode signals. It is the authors' waveform. Unfortunately, the manner in which the morphologic
preference to place the E-2 electrode for L3S (L4S) over the iliac characteristics of SEP waveforms are delineated is not universally
crest contralateral to the stimulated l i m b . A somewhat larger
22,a
accepted and a number of systems are currently in use. SEP
potential results without significant waveform distortion but may waveforms may be described by the following: polarity, peak
require more averages to improve the signal-to-noise ratio. 282
nomenclature, peak latency, peak amplitude, peak-to-peak ampli-
Another E-1 electrode in tibial nerve recordings is located over tude, and morphology. - - - - 6
- - " As there is no consen-
7 52 91 104105 107 250 82
the twelfth spinous process and is called T12S. - Some authors 6 7

sus as to how to describe waveforms, confusion with respect to
prefer to use the first lumbar vertebral spinous process (LIS) in- reported waves from one investigator to the next inevitably arises.
stead of T12S. The same potential is detected at either location. It is the responsibility of the practitioner to review the pertinent
Again, the E-2 electrode is placed 4 cm rostral to this location. It literature and decide what is the most advantageous manner to
is certainly possible to also use a second or the same contralateral present the data for one's particular circumstance.
POLARITY connected to the noninverting amplifier such that an upward de-

flection above the isoelectric line results from a net negative po-
Whether a waveform's peak points in the direction above tential difference between the E-l and E-2 electrode. In S E P
(negative), or below (positive), the baseline is dependent upon terminology, a potential with a positive peak is given the desig-
how the electrode leads are connected to the differential ampli- nation of the capital letter "p," i.e., P. When printed, this letter
fier 9 6 . 9 7 . 2 8 2 Polarity refers to the net positivity or negativity of designation may be in plain typeface or bold typeface (e.g., P or
the potential difference between the two recording sites. The P). Similarly, a negative waveform is referred to by N or N. If
SEP waveform may be considered the net result of the potential this concept of polarity poses a particular problem, one may
difference between the two electrodes positioned on the body. wish to recall that in nerve conduction studies the E-l electrode
Simply put, the instrument merely measures the potential differ- is located over the motor point of a muscle and upon stimulating
ence, i.e., the voltage generated at one region of the body as it is that muscle's peripheral nerve, a negative (upward) deflection
recorded by one electrode subtracted from the voltage detected ensues. This is the same electrode connection recommended for
! by the other electrode. The net difference is what is displayed SEP studies.
on the cathode ray tube (CRT). Whether the difference observed
on the CRT proceeds above or below the baseline is a matter of PEAK N O M E N C L A T U R E
how the E-1 and E-2 electrodes are connected to the amplifier.
If a known negative potential difference is connected to the Arguably, perhaps the most confusing aspect of SEP wave-
amplifier through the two recording electrodes such that the form analysis is the manner in which peak latencies are desig-
CRT trace deviates above the baseline, a "negative up'7"positive nated. This confusion primarily arises not because this is a
down" display is created. On the other hand, should the elec- particularly complex issue, but more so because of a lack of uni-
trodes be connected to the amplifier so that this same potential formity/agreement among investigators. There are different
proceeds below the baseline, a "negative down'V'positive up" waveform nomenclatures for the various types of evoked poten-
arrangement exists. Of course, these directional CRT deviations tials, i.e., somatosensory evoked potentials, visual evoked p o -
presuppose a known potential difference directed into the in- tentials, and brain stem (auditory) evoked potentials. - 52 282
strument. In reality, we do not always know the net polarity of a Additionally, for a particular type of evoked potential, identical
neural generator and electrode-instrument connection conven- waveform peaks are given different names depending upon the
| tions become important. laboratory in which these evoked potentials are obtained.
| As there is no universal agreement regarding polarity, the Although a number of peak nomenclature systems exist for
recommendations of the AES - should be followed because
6 7
SEPs, three are reviewed with a recommendation for usage. A
these are the same as those used for needle electromyography simple and at first glance rather appealing peak naming system
and nerve conduction studies. Specifically, the E-l electrode is uses the letters P and N , which represent a positive or negative
Figure 9-/0. Waveform measurement parameters. Example of recording peak latency, peak amplitude, and peak-to-peak amplitude.The
peak latency is measured by noting t h e time interval between the stimulus onset and individual positive o r negative waveform peaks (L|s||, Lpj,
and Lfsj2)- P ^ amplitude is recorded by drawing an imaginary line through a quiet portion of t h e baseline t o the peak of a waveform (e.g.,Ajpj).
e a
Peak-to-peak amplitude is calculated by identifying a particular peak, referencing it t o a preceding o r following peak of opposite polarity. Note t h a t
negative deflections are above and positive deflections below the baseline, i.e., opposite t o AES recommendations. (From Spehlmann R: Evoked
Potential Primer. Boston, Butterworth Publishers, 1985, with permission.)
peak polarity, respectively. - A number follows either the P or

52 282
usually because of superimposed noise from inadequate muscle
N to signify the sequential order of peaks with the same polarity relaxation, filter settings allowing noise to contaminate the
(Fig. 9-10). For example, a waveform with an initial negative waveform's signal, not obtaining a large enough number of
potential followed by a positive potential that in turn has a repe- signal averages, or some other technical problem with the in-
tition of this sequence would be labeled as: N l , P I , N2, P2. strument or electrodes. Occasionally, even after trouble shooting
Similarly, two negative peaks separated by a positive peak is and correcting the above noted problems, peak identification
designated: N l , P I , N2 (Fig. 9-10). The latency for a particular may continue to be difficult. One must initially decide if a
wave, however, is not readily apparent in this system. Also, if a waveform is present and then identify its peak.
potential's waveform for a given individual is missing, one must If there is any question as to whether a noisy tracing contains
then struggle with the sequential numbering of the remaining a true waveform, a number of options should be explored. A
peaks. recommended practice is to always perform two separate trials
A solution to this problem is attempted by designating a of each response to ensure reproducibility. - - - When doubt
6 7 52 282
mean latency to be assigned to the P or N. If the negative corti- remains regarding the presence or absence of a particular wave-
cal waveform to median nerve wrist stimulation normally arises form, more than two averaged runs may be required to ensure a
at 20 ms, it is given the label of N20. The obvious question then waveform's existence. Other trouble shooting measures include
is, what do you call this potential if it occurs at 30 m s ? - 52 282
performing several averaged trials without the delivery of a
Should it be referred to as "N20 occurring at 30 ms," N20/N30, stimulus. An averaged trace obtained in the absence of a stimu-
or simply N30? lus provides a good idea of the background noise. By compar-
The system used by a growing number of investigators and ing averaged traces with and without stimulation, it may be
that recommended by the AES is a combination of all of the easier to appreciate which waveform peaks are noise compared
above proposals. - The capital letters P and N continue to be
6 7
with a true physiologic response. At times it may be helpful to
used for polarity. The characteristic or mean latency of appear- contrast SEP trials recorded with a weak stimulus with those
ance for a specific peak latency is reported with a bar or line obtained utilizing a stronger excitation pulse. A recording with
drawn over the latency value. When this potential is abnormal, a weak stimulus should produce smaller waveforms, thereby en-
the mean peak latency is followed by the observed latency. 91
suring the presence of a possible peak in the stronger stimulus'
This is the so-called observed versus characteristic nomencla- trace. This is a variation on the theme of delivering no stimulus
ture. For example, if the median nerve's cortical SEP is ob- at all.
served at 30 ms, it is designated as: N20 occurring at 30 ms or Should the above methods fail to remove sufficient doubt re-
N20/N30. A simple way in which to reportjhese data in an SEP garding a peak's presence, it is also possible to include a period
report is to designate a data column as N20 and place 30 ms in of time prior to the activation impulse's delivery to assess the
this column thereby assigning the recorded latency to a charac- two regions of the trace with respect to noise versus stimulus-
teristic waveform. related artifact. Another way to attempt to eliminate stimulus-
195
AES Recommendations. For those previously derived related events contaminating the desired waveform is to
upper and lower limb montages noted above for the median and alternate the polarity of the stimulator during data collec-
tibial nerves, the AES - recommends a number of specific des-
6 7
tion. - - This option may result in a slightly different la-
262 263 312
ignations for particular recording locations. In median nerve tency because of the several-centimeter difference between the
SEP studies, channel 1 's C37C4' t o F p Z ' yields a major cortical two stimulus sites. One or a combination of several of the above
waveform with the designation N20. Channel 3's cervical spine techniques should convince the practitioner of the waveform's
(C2, C5, or C7) to FpZ' resultsjn wavdbrms with the peak des- true nature. It is also possible for the stimulus to generate suffi-
ignations N i l , N13 and N14 or N11/N13. The Erb's point cient artifact to interfere with the SEP waveform. Rotating the
recording on channel 4 remains EP and is not given a specific anode about the cathode can at times optimize the initial aspect
letter and number abbreviation. Some authors, however, refer to of the recording that is influenced by the stimulus artifact.
this potential as N9 or N10. - 4 130
A difficulty often encountered in attempting to identify an
In tibial nerve SEP investigations, the popliteal fossa, third SEP peak arises in a particularly noisy tracing in which there
lumbar, and twelfth thoracic recording locations are simply re- are several minor deflections of similar amplitude superimposed
ferred to by their anatomic designations as PF, L3, and T12 po- on a major waveform. There are four recommended methods
tentials, respectively. Letter and number assignments are one may use to assist in the identification of a peak's latency
provided for the cortically detected waveforms. The first se- (Fig. 9-1 1 ) . Let us suppose that a major SEP waveform is
282
quential positive/negative waveform is referred to as P37 and contaminated by sufficient noise to result in two small peaks su-
N45, respectively. perimposed on a major waveform despite all attempts to mini-
mize this noise. The first manner in which to arrive at a peak
PEAK LATENCY latency is simply to identify the peak with the highest amplitude
and assume that this is the primary SEP peak had noise been
Identifying a waveform's peak is obviously important, as this eliminated. A second option is to pick the first peak irrespective
is the manner in which normalcy is decided with respect to la- of the subsequent peaks identified. Thirdly, the time difference
tency. Under optimal conditions, the peak of an SEP waveform of the two peaks can be split such that a point midway between
is noted as the time from the stimulus onset to the region of the the two peaks is chosen and identified as the waveform's la-
maximum positive or negative sequential peak deflections (Fig. tency. Should more than two peaks appear, the difference be-
9-10). In SEP studies, the instrument's latency markers begin tween the two furthest separated peaks may be chosen. The
measuring time from stimulus onset to where the practitioner fourth venue extends the slopes of the first and last minor peaks
locates the time marker on the waveform. Although this is usu- above the major waveform using the point of intersection as the
ally not a problem, it can be rather difficult at times to decide peak latency for the SEP (Fig. 9-11). Of course, none of these
exactly where the waveform's peak occurs. This difficulty is measures are ideal but they are practical solutions to a difficult
Figure 9-11. Waveform parameter identification. Four

possible methods in which t o identify a waveform's peak and its
latency. A,That portion of the waveform with the greatest mag-
nitude can be chosen. B . T h e first of a series of minor peaks
riding on a major peak may be identified. C, It is also possible t o
c h o o s e the mid-point between t w o adjacent peaks superim-
posed on a major waveform. D.The rising and falling phases of a
major waveform can be continued superiorly t o where they in-
tersect. Of course, in this instance t h e amplitude is n o t mea-
sured at the intersection point but only the peak latency. (From
Spehlmann R: Evoked Potential Primer. Boston, Butterworth
problem. Care must be exercised in applying any of these peak pathway, allowing one to quantify these intervals. When assess-
identification methodologies so as to not mistake two separate ing central nervous system propagation by determining the in-
waveforms for one that is contaminated by noise. These four op- terval times between peaks recorded from spinous processes
tions are to be exercised only when the practitioner decides that and the scalp, one can calculate the c e n t r a l c o n d u c t i o n t i m e
the tracing contains noise of such a degree as to result in multi- and thereby minimize peripheral factors that may influence ab-
peaked SEPs not rectified by repeating the trial or averaging solute peak latencies, e.g., peripheral neuropathy. In other 37
more responses. A normal variation can occur in which an opti- words, the peak latency for cortical SEP response may be al-
mal SEP recording reveals a bifid waveform. In this instance, tered by pathology affecting either the peripheral or central ner-
the trace is considered technically acceptable and it is recom- vous system individually or simultaneously. Interpeak latencies
mended that both peak latencies should be recorded. 282
and central conduction times help to distinguish lesions involv-
ing the peripheral and central nervous systems. A central ner-
INTERPEAK LATENCY vous system lesion should not affect the peripherally obtained
impulse (EP or PF) but will alter the interpeak conduction time
In addition to measuring the absolute SEP peak latencies, it is when it is interposed between specific recording points. On the
also possible to record the i n t e r p e a k l a t e n c i e s between the var- other hand, a peripheral nerve lesion delays all peaks, peripheral
ious SEP peaks. - - - The interpeak latency essentially repre-
6 7 52 282
and central, by roughly the same amount. ' * 52 233 282
sents the impulse's time of propagation between recording sites. In upper limb SEPs, for example, one can easily determine
This is a preferred measurement parameter by some investiga- the conduction properties of the peripheral nervous system by
tors because it denotes the conduction time between the struc- evaluating the EP peak latency, assuming technical factors have
tures generating the waveforms along the somatosensory been eliminated. If this latency is prolonged, one can assume a
Table 9-4. Correlation of ARM SEPs to Possible Clinical Interpretations

Abnormal SEP Finding Interpretation
Technical problems
1. Absent SEPs t o arm stimulation at all recording levels Lack of stimulus; lack of synchronization between stimulator
and averager; faulty recording electrodes or equipment
2. Increased latency at SEPs at all recording levels Hypothermia; inaccurate measurement of t h e distance between
stimulating and recording electrodes
Lesions of the nervous system
1. Absent N9 with normal N13 and N20 Normal
2. Absent N9 with absent o r delayed N13 and N20 Peripheral nerve o r plexus lesion; rule o u t technical problems
3. Increased latency of N9 with equally increased latency of Peripheral nerve o r plexus lesion; rule o u t technical problems
N13 and N20: Decreased peripheral conduction velocity
with normal central conduction times
4. Increased N 9 - N 1 3 conduction time with normal N13 Defect above the brachial plexus and below t h e lower medulla
amplitude and shape, normal peripheral conduction velocity,
normal N13-N20 conduction time
5. Absent N l 3 and absent o r delayed N20 Defect above the brachial plexus and below o r at the lower medulla
6. Increased N 1 3 - N 2 0 central conduction time with normal Defect above the lower medulla and at o r below the somatosensory
N 9 - N I 3 conduction time and normal peripheral conduction
velocity
7. Absent N20 and normal N 9 - N 1 3 conduction time and Defect above the lower medulla and at o r below the somatosensory
normal peripheral conduction velocity cortex
8. Decreased peripheral conduction velocity and increased central Combination of peripheral nerve or plexus lesion and central
conduction times defect
From Spehlmann R: Evoked Potential Primer. Boston, Butterworth Publishers, 1985, with permission.
lesion affects the peripheral nerve conveying the electrical im- in accurately measuring the distance along the arm and central
pulse at some point distal to, or at, the Erb's point record- conduction routes as well as latency delays along syn-
ing. - Also, the cervical and cortical potentials could be
52 282
apses. - - -
6 7
Spehlmann has delineated general guidelines
52 104282 282
expected to be delayed a similar amount, as the impulse cannot to help the beginner interpret the multichannel recordings with
make up the time lost peripherally in the central nervous respect to possible lesion location or instrumentation errors that
system. The interpeak latency or central conduction time be- may mimic neural pathology (Table 9-4).
tween the cervical and cortical regions as reflected by their re- A similar line of reasoning to that noted above can be applied
spective interpeak latencies can be anticipated to be normal, to'lower limb SEPs. The PF waveform serves as the marker for
eliminating the possibility of a centraHesion. Should the EP 36
peripheral nerve function. The latency between PF and L3
potential be normal but the cervical N13 potential be delayed or allows one to assess the conduction across the sacral plexus and
absent, a lesion most likely exists between the brachial plexus cauda equina. An interpeak time from L3 to T12 provides insight
and the posterior columns at the cervical spine where N13 is into impulse propagation from the cauda equina to the entry of
generated. Also, the cortical potential would be expected to
282
the lower aspects of the central nervous system. Central conduc-
be abnormal, as the lesion is "down stream" from the scalp elec- tion can be_ evaluated by subtracting the T12 latency from the
trode, which reflects all previous slowing. If the N13 potential is cortical P37 peak latency. If a T12 potential is not recorded, the
delayed but present, the interpeak latency between the cervical L3 potential can be utilized for similar purposes. The presence,
region and the cortex would be normal, although the absolute absence, or delay of the above noted potentials can yield infor-
N20 potential might be delayed. Finally, a normal EP and cervi- mation regarding possible lesions affecting different aspects of
cal spine potential but abnormal scalp potential suggests a the somatosensory conduction pathway (Table 9-5). 282
lesion in the somatosensory pathway rostral to the dorsal

columns in the neck, possibly involving the brain stem, thala- AMPLITUDE MEASUREMENT
mus, thalamocortical radiations, or the cortex itself. - One 52 282
may measure conduction velocities along the somatosensory In addition to peak latency, an SEP's amplitude is also an im-
pathway simply by dividing the distance between recording portant parameter to evaluate with respect to normalcy and com-
sites by the interpeak latencies. However, the accuracy of such parative side-to-side measurements. The amplitude of an SEP
conduction velocities are limited by the difficulty encountered represents the magnitude of the potential difference between the
Table 9-5. Correlation of Leg SEPs to Possible Clinical Interpretations

Abnormal SEP Finding Interpretation
Technical problems
1. Absent SEPS t o leg stimulation at all recording levels Lack of stimulus: lack of synchronization between stimulus and
averager; faulty recording electrodes o r equipment
2. Increased latency of SEPs at all recording levels Hypothermia; inaccurate measurement of the distance between
stimulating and recording electrodes
Lesions of the nervous system
1. PTN stimulation
a. Absent FP potential with absent o r normal L3 potentials Normal
and absent scalp SEPs or normal central conduction velocities
b. Absent PF potential with either absent spinal potentials Lesion between ankle and PF
and absent scalp SEPs or normal central conduction velocities
c. Decreased peripheral conduction velocity to PF and Defect below cauda equina
(1) equally decreased peripheral conduction velocity t o L3 Defect of both distal and proximal peripheral nerve
(2) no decrease of peripheral conduction velocity between Defect between ankle and PF
PF and L3S
d. Decreased peripheral conduction velocity t o L3S with normal Lesion between PF and cauda equina
peripheral conduction velocity t o PF
e. Absent L3 a n d T l 2 potentials, absent o r delayed scalp SEP Probably lesion between PF and cauda equina
with normal peripheral conduction velocity t o PF
2. CPN stimulation
a. Absent L3 potential with present o r a b s e n t T I 2 andT6 Normal
potentials and normal scalp SEP
b. Absent L3,TI2, andT6 potentials, delayed or absent scalp SEPs Defect at o r above the cauda equina, o r both
c. Decreased peripheral conduction velocity to L3 Peripheral defect between PF and cauda equina
3. Either PTN or CPN stimulation
a. Decreased central conduction velocity Defect above the cauda equina and below or at the somatosensory
cortex
b. Absent scalp SEP Suspect defect above the cauda equina and below or at the somato-
sensory cortex
c. Decreased peripheral conduction velocity and decreased Lesions above and below the cauda equina, or a single lesion at the
central conduction velocity cauda equina or lower spinal cord
F r o m Spehlmann R: Evoked Potential Primer. B o s t o n , B u t t e r w o r t h Publishers, 1985, w i t h permission.
two recording electrodes and depends upon the individual mon- It may be necessary to determine a "zero" baseline if the initial
tage utilized. As previously noted, the location of the two peak arises immediately from the preceding baseline (see
recording electrodes directly influences the amplitude of the above). Alternatively, the descending limb of the potential can
recorded potential. For comparison studies in individual pa- be assessed by measuring the vertical displacement of a particu-
tients, the same recording montage must be used to minimize lar peak with a negative to positive descent (Fig. 9-10). Finally,
the technical factors causing amplitude variation. In addition, a third option is to calculate both the ascending and descending
the criteria used (base-to-peak or peak-to-peak) should remain limbs of a peak and determine the mean of both sides of the p o -
consistent. Although not accepted by all investigators, side-to- tential. Of course, there are difficulties with this method if a
side amplitude differences of 50% or more are considered in- large peak is preceded by a smaller peak that does not achieve a
dicative of a possible abnormality. It must be recognized, prior zero baseline level. An additional objection raised to this
however, that normal persons can have considerable side-to-side method of quantifying SEP amplitudes is based upon the theo-
amplitude differences, approaching 80% or m o r e . 102
Unfor- retical consideration of each peak being the result of a neural
j tunately, there is no universally accepted manner in which to de¬ generator. Measuring from one peak to the next may be combin-
I termine the magnitude of various waveforms. Thus, it is ing information from two different generators. The amplitude,
imperative that one standard method of amplitude determina- therefore, is not representative of one group of neurons, but a
tion be chosen and the reference data for that montage used. mixture of neurons possibly from different locations. This type
Adhering to a particular recording methodology also allows the of amplitude measurement may be "mixing apples and or-
practitioner to gain confidence in the method and experience in anges" and not representative of a true loss of neurons from a
52
optimizing the waveforms in either normal or pathologic situa- particular site in pathologic states. Despite the theoretical objec-
tions. Two widely used amplitude measurement schemes are tion, this method is recommended in assessing SEP amplitudes
presented: peak amplitude and peak-to-peak amplitude. - 52 282
because it is rather easy to perform and is useful for subsequent
follow-up studies, particularly when performed by different ex-
Peak Amplitude aminers. It is the authors' observation that novice practitioners
In order to evaluate an individual waveform with the "peak have less difficulty in identifying sequential peaks to calculate
amplitude" method, a reference "zero level" is required. This SEP magnitudes compared with defining a horizontal zero line
relative "zero" potential can be determined in a number of ways. from which to measure an amplitude.
Once the SEP is averaged, the practitioner can subjectively
| assess what appears to be a flat horizontal baseline running WAVEFORM MORPHOLOGY
through the entire recording and designate this line as the "zero"
reference point. One may also select a relatively "quiet" time Many of the SEP waveforms for both upper and lower limbs,
period devoid of any waveforms either preceding or following including spinal potentials, have characteristic shapes. Normal
the SEP component waveforms and declare this the reference variations and subtlety of difference between normal waveforms
baseline. Whatever method is used, the SEP's amplitude is
282
and those found in pathologic situations preclude a simple ref-
measured with respect to the ordained baseline signifying zero erence analysis based solely on shape. Although there has been
potential. The amplitude of an individual peak is then defined as an attempt at classifying SEP waveforms with respect to normal
the maximum vertical displacement above the zero line as mea- and abnormal morphologic p a r a m e t e r s , - ' ° this method-
,05 ,07J,, 25
sured by the instrument's amplitude marker (Fig. 9-10). ology is not widely accepted. - As a result, the shape of S E P
52 282
Specifically, one of the markers is placed on the declared zero waveforms is not heavily weighed in the criteria to decide if a
line, while the other is located on the identified SEP peak. The particular waveform should be described as abnormal.
instrument then displays the vertical displacement above the
zero line in appropriate units, i.e., microvolts. This method
works quite well for SEP recordings with very stable baselines. SEP I N S T R U M E N T A T I O N
Difficulty arises, however, when the trace is rather noisy, sev-
eral sequential waveforms are superimposed on each other, or MINIMAL STANDARDS
the time period preceding or following the waveform is erratic.
It is also somewhat challenging to measure peak amplitudes ac- The fundamental aspects of instrumentation pertinent to the
curately if the responses are particularly small and a "wavy" recording of SEPs are essentially the same as those used for
baseline is present. Fortunately, an alternative method of assess- needle electromyography and nerve conduction studies (see
ing waveform amplitudes is available. Chapter 3), with a few alterations. Minimal criteria regarding
instrumentation parameters as defined by the AES are stated in
Peak-to-Peak Amplitude this section to further clarify the instrument's specific require-
Evaluating SEP's waveform with respect to amplitude by ments to obtain optimal SEP recordings free of distortion. - A 6 7
using the peak-to-peak method is somewhat easier than the peak number of exceptions to the general AES guidelines for all
amplitude method discussed above. - Simply, the peak-to-
52 282
types of evoked potentials are delineated and personal prefer-
peak amplitude is the vertical displacement of the instrument's ences are detailed. Recording parameters should be identical to
amplitude markers when placed between the maxima of consec- those used for collection of reference data whenever possible.
utive SEP peaks of opposite polarity. Depending upon the wave-
form's peak morphology and the practitioner's preference, there Amplifier
are a number of methods by which to calculate the peak-to-peak The minimum differential input impedance must be at least
amplitude. It is possible to place one of the instrument's ampli- 10 MO, and the input signal's magnitude should be capable of
tude markers on the first positive or negative peak and the sub- being amplified between 1000 and 500,000 times. - - - A 6 7 52 282
sequent peak of opposite polarity. This allows one to calculate common mode rejection ratio of least 10,000:1 (80 dB) at the
the amplitude of the ascending limb of the potential (Fig. 9-10). instrument's greatest sensitivity is desirable. Multiple filter settings
capable of creating a wide bandpass of 0.1 Hz to 5000 Hz are intervals the digitizer can sample, just as the horizontal (time)
suggested. Of course, the filter settings for individual SEP stud- resolution is dependent upon the number of horizontal digital
ies will vary with the parameters used by the individual investi- data points. The vertical resolution is defined by the number of
gator. When utilizing the above open bandpass, the instrument's vertical intervals available for the amplitude at each horizontal
inherent noise level should not exceed 3 pV. 320
data point interval. The vertical string of bits is referred to as a
word. Therefore, an 8-bit converter is an A/D converter with a
Averager word length of 8 bits. An 8-bit binary number represents 2 , or 8
The AES recommends that an averager's time (horizontal) 25*6 points of vertical resolution for each horizontal data point.
resolution be at least 80 ps/data point/channel with a minimum An 8-bit digitizer has a total vertical range of 256 independent
of 250 resolution points per channel. The amplitude (vertical) voltage levels. The total number of intervals is one less than the
resolution corresponds to at least an 8-bit converter, and the number of levels, i.e., there is one interval between levels 1 and
maximum number of averages per waveform must meet or 2. The resolution, therefore, is 1 vertical interval out of a total of
exceed 4000 trials should circumstances warrant. - The mini- 6 7
255 intervals, or 0.39% of a waveform's amplitude.
mum number of channels is determined by the type of evoked The practitioner must decide on the number of stimuli to be
potentials performed. averaged for a particular SEP waveform. Although most instru-
Exception is taken to these AES recommendations. Although ments have the capability of performing several thousand con-
such specifications may be appropriate for auditory evoked po- secutive averages, the waveform's signal-to-noise ratio
tentials given the time period of waveform production (about 8 improves only by the square root of the number of averages (see
ms or less for all waveforms), they are inappropriate for SEP Chapter 3). Thus after averaging a few hundred responses, the
studies. For example, consider the minimum requirements marginal improvement in the signal-to-noise ratio is of ques-
noted above of 250 data points combined with a resolution of tionable benefit given the amount of time necessary to collect
80 ps/point/chan'nel. These resolution parameters allow only additional data. For example, to double the signal-to-noise ratio
for a total screen analysis time of 20 ms, which is barely following 100 averages, one would need to acquire 400 sweeps.
enough to perform upper limb SEPs (i.e., 80 ps/point/channel x To again double the signal-to-noise ratio, 1600 sweeps would
250 points = 20,000 ps/channel = 20 ms/channel), and com- be necessary, and so on. Several thousand averages can also be
pletely unacceptable for lower limb SEPs. Given that lower quite time consuming. An SEP waveform is typically optimized
limb SEPs need about 100 ms total analysis time, an instru- by averaging 300-500 times and performing two separate trials
ment's resolution is supposedly inadequate at 400 ps/data point of 300-500 averages. Two trials of 500 averages is considered
(100 ms/250 points/channel = 1 ms/2.5 points/channel = 400 more reliable than one trial of 1000 averages because it verifies
ps/point/channel). the various waveform parameters.
The authors recommend the instrument have a minimum of
250 data points of horizontal resolution for a screen time of 100
ms, yielding a time resolution of 400 ps/data point/channel. A An additional minimal instrument requirement is for some
resolution of 400 ps/data point equates to a resolution or sam- type of ongoing CRT display. The CRT should have amplitude
pling frequency of 2500 Hz (1 data point/400 ps = 2500 data and time scales that can readily be interpreted. It is also neces-
points/1000 ms = 2500 Hz). The highest frequency likely to be sary for a hard-copy of the averaged waveform to be printed out
encountered in most SEP studies is the rise time of the EP or PF and maintained in the patient's records for future reference. - 6 7
potential, which in the authors' opinion is less than 800 Hz. As

the given parameters of 250 data points over 100 ms and 400 DESIRABLE INSTRUMENT OPTIONS
ps/data point produce a resolution at least 3.1 times the EP/PF
rise time, these potentials should be adequately resolved. The The above instrumentation features are those minimally nec-
spinal and cortical waveforms occur over much longer time peri- essary to perform reasonable SEP analysis. There are, however,
ods than the EP/PF potentials and should be resolved without a number of additional instrumentation parameters that may
difficulty. Recall that at least two points of resolution are re- enhance data acquisition and flexibility in difficult recording
quired to minimally resolve a waveform (Nyqvist frequency), 282
conditions.
and in the above worst-case scenario, at least 3.1 points of reso-
lution (2500 Hz/800 Hz = 3.1) are available. A resolution is pre- Amplifier
ferred of 200 ps/data point for lower limb SEPs and 100 ps/data Certainly, the greater the common mode rejection, the quieter
point (10,000 Hz) for upper limb SEPs, i.e.. 100 ms/500 data the recordings, particularly in electrically noisy environments. ^ 6
points/channel (lower limb) and 50 ms/500 data points/channel It is also helpful for the amplifier to have the capability of inter-
(upper limb), or a total of 500 data points per channel. The above nally switching electrode montages and quickly measuring the
explanation deals with the minimum requirements to just resolve electrodes' impedances. From time-to-time, it may be appropri-
any SEP waveform. If your instrument can produce a higher res- ate to verify the calibration of the instrument, and an internal
olution of 80 ps/data point/channel, more than enough resolution device to checkjhe calibration on the CRT screen is helpful.
is present to optimally resolve the waveform in question.
With respect to amplitude, an 8-bit analog-to-digital converter Averager
(A/D converter or digitizer) is recommended. - - In digital in-
6 7 282
Horizontal resolutions in excess of 500 points per channel
struments (all presently available commercial instruments), the and vertical resolutions of 10 bits or higher will improve the
digitizer samples the original analog signal at a fixed frequency. digital reproduction of the original analog signal. Variable
Each sample is converted into a digital representation of ampli- sweep analysis times from 5 ms to 100 ms provide the practi-
tude, i.e., a binary number. This amplitude then consists of a tioner with flexibility regarding the performance of less routine
number of binary digits or bits. The vertical resolution of the SEP studies and allow one the capability of expanding into
signal's amplitude is dependent upon the number of vertical other types of evoked potential work. A nice feature is to be able
to store the previously recorded waveforms for on-screen com- difference between the two recording electrodes, this difference
parisons between multiple traces from the same and other is amplified and can obscure the biologic signal. This is why it
sites. - A preset counter with the capability of ending data col-
6 7
is important, if at all possible, that the two recording electrodes
lection after a designated number of averages are acquired is for any given channel be of the same material. When the two
also a convenience. electrodes are constructed of the same metal, the bias potentials
It is helpful to exclude responses contaminated with large ar- developed at their surfaces are more likely to be similar, thereby
tifacts that may continue to distort the final SEP waveform de- allowing the instrument to eliminate the same bias potentials
spite multiple averages. An automatic artifact reject option through common mode rejection through the process of differ-
accomplishes this by permitting one to set the amplitude voltage ential amplification. When the E-l and E-2 electrode are dis-
criteria beyond which a potential is not included in the averaged similar metals, different bias potentials can develop at each
response. When using an automatic reject option, care must be surface, resulting in a potential difference that is dissimilar and
exercised to not set the voltage criteria too stringently, as arti- subsequently amplified. Additionally, dissimilar electrodes may
fact-free responses may also be excluded. If the instrument is have different impedances, resulting in the recording of slightly
including only a few traces to be averaged, the voltage exclu- different common signals, e.g., 60-Hz interference, thereby am-
sion criteria set on* the automatic reject option should be plifying this difference. This impedance mismatch results in the
checked and the environment investigated for excess noise, and recording, instead of elimination, of 60-Hz noise. The above
all electrode leads examined for discontinuity. One must also statements suggest that it is not a good practice to mix surface
consider a particularly noisy environment or a bad electrode and needle recording electrodes. Metals that are relatively pure
lead introducing excess noise into the system in such cases. and can be kept free of surface contamination are the materials
Particularly noisy traces can be improved somewhat with re- of choice for electrode manufacture. Such metals are silver,
spect to multiple minor irregularities in the baseline by using a gold, tin, platinum, and stainless steel and considered ideal sub-
so called smoothing filter. This is a rather simple technique
282
stances from which to construct electrodes because of the fol-
whereby the data represented by 3 or 5 digital resolution points lowing considerations: cost, availability in pure form, ability to
are averaged to yield one data point. This process is repeated se- resist oxidation, and minimal tissue toxicity. Gold is usually
quentially, thus reducing the small irregularities in the trace. preferred over silver for surface electrodes because it resists en-
The net result is to "smooth" out the final appearance of the vironmental degradation due to oxidation and tarnishes consid-
trace. This option should be used judiciously because it may erably less. With respect to needle electrodes, the preferred
138
alter the SEP shape, amplitude, and inter-peak latency relation- materials are stainless steel or platinum mixed with another
ships. Some instruments provide the practitioner the option of metal so that the needles are good conductors, and strong yet
adding or subtracting waveforms from one channel to that of an- flexible enough to penetrate tough tissues such as the scalp.
other. Finally, the ability to adjust when the collection of data Pure gold is too malleable to be used in needles.
by the averager begins with respect to the stimulus may be of
assistance in some cases, i.e., data collection is delayed until the RECORDING ELECTRODES
stimulus artifact has subsided. Other options are offered by
manufacturers, but the practitioner must clearly explore how Surface Electrodes
such features will be of benefit in particular clinical settings in The most common electrodes used for recording SEPs are
relation to the additional cost to the overall unit. those typically used for EEG recordings. - These surface
52 282
electrodes are a 4 - 1 0 mm metal cup with a hole in the center

and a flat rim. A relatively long (24-48 inches) insulated wire
ELECTRODES serves as the lead that connects the metal cup to the instrument's
amplifier port. After selecting the site from which to record (10¬
Electrodes utilized for the detection of SEPs may be classi- 20 system or other landmarks noted above), one of the most
fied into recording and stimulating electrodes. Both can be challenging aspects to novice practitioners is the optimal secur-
either needle or surface electrodes. Again, there are no stan- ing of electrodes to the patient.
dards regarding the particular material an electrode must con- Prior to attaching surface electrodes to the patient, it is first
tain, and the type of electrode and its method of application are necessary to prepare the skin site adequately. Recall that the
primarily left to the practitioner. Because of this flexibility, one SEP potential is relatively small compared with the background
must be aware of various technical aspects regarding the multi- EEG potentials, muscle activity, and other environmental noise.
ple electrodes available. As a result, the electrode must be secured to the patient such
that it is capable of optimally contacting the patient's electrolyte
COMPOSITION volume conductor conveying the various biologic signals of in-
terest. This can be achieved only if the superficial layers of the
Essentially, an electrode serves the function as an intermedi- skin are adequately abraded, allowing the electrolyte paste
ary between the instrument and the biologic signal of interest placed within the electrode's cupped surface to be in contact
within the body and may be defined as: " . . . a metallic connec- with the patient's electrolytic solution. When this contact is
tion between the complex physiological electrolyte of tissue and achieved by all electrodes, the signal of interest is recorded and
the recording circuitry." Recall from Chapter 3 that the
138
noise is eliminated. It is possible to know when the electrode-
metal/electrolyte interface can result in a transfer of ions be- patient interface is ideal by measuring the impedance through
tween the recording electrode's metal surface and the body's two such prepared electrodes.
ionic solution, giving rise to potentials (bias potentials) at the Recall that resistance is the difficulty a current encounters in
two electrodes possibly resulting in rather large potential differ- an attempt to move through a conducting medium. Resistance is
ences between the two electrodes. If the two electrodes are used
138
the term used when dealing with direct currents but is replaced
For recording minute biologic signals and there is a measurable with the term i m p e d a n c e if alternating currents are encountered
(see Chapter 3). In biologic systems, the hindrance to current stylus and applying collodian around the electrodes' rim. The
flow is discussed utilizing the concept of impedance. When de- drying of collodian is hastened by blowing compressed air over
scribing the electrode-patient contact, the hindrance of flow the site or using a hair dryer. Once the electrode is securely in
52
across this interface is known as impedance. The impedance is place, it is filled with electrolyte cream through the central
clinically measured by passing a weak alternating current of a hole. Following completion of the study, the collodian is re-
specified frequency through the amplifier across both recording moved by dissolving it with acetone. As these materials (ace-
137
electrodes or between each electrode and ground, and calculat- tone and collodian) are volatile, adequate ventilation is
ing the magnitude of impedance to this current. It follows that if mandatory and explosion precautions must be taken. Of course,
there is minimal impedance to an impressed current flow by the all of the collodian is rarely removed and the patient should be
instrument, any currents or differences of voltage in the body instructed that several shampooings may be necessary before
passing by the electrode will also enter the instrument. If the the collodian is no longer present.
impedance is high as determined by the instrument, then most
likely, current will also not flow from the patient into the ampli- N e e d l e Electrodes
fier. Remember that current prefers the path of least resistance. As previously noted, stainless steel and platinum alloy subder-
A low inter-electrode impedance provides an accessible current mal EEG needles are commercially available for SEP studies. The
path into the instrument. High inter-electrode impedance cre- main rationale for using needle electrodes is primarily ease of ap-
ates a situation in which the current continues to flow in the plication. i37.i38.282 Subdermal needle electrodes should be placed at
body's lower impedance, instead of entering a pathway of the appropriate site just under the skin so as to enter the patient's
greater impedance, i.e., across the skin and into the instrument. electrolytic solution conveying the biologic signals. Because the
When applying surface electrodes, an alcohol swab rubbed volume conductor has been entered, the skin site does not need to
across the skin followed by a commercially available pumice is be prepared to reduce the impedance. A local antiseptic wipe is
used to abrade the skin site gently by removing several superfi- used, however, prior to penetrating the skin. The elimination of
cial layers of the skin and skin oils. It is generally accepted that skin preparation considerably reduces the time necessary to place
abrasion is considered sufficient when the impedance measured multiple electrodes onto the patient. Unfortunately, needle elec-
across two such electrode preparation sites is between 1000 and trodes are subject to being pulled out easily. It is recommended
5000 ohms (O). - When the impedance is less than 1000 O,
52 282
that some slack be placed on the electrodes' lead wires prior to
care must be taken to avoid a situation in which the amplifier is taping them to a portion of the body such that an inadvertent pull
short-circuited through aberrant conduction pathways such as on the wire does not remove the needle.
excess perspiration or electrolyte paste between two electrodes.
In this instance, the impedance through the abnormal conduct- COMPARISON OF NEEDLE AND SURFACE ELECTRODES
ing pathway is less than through the electrodes and the current
(biologic signal) would rather travel the path of least imped- The impedance of needle electrodes is measured in the same
ance, thereby bypassing the instrument. If the impedance manner as that for surface electrodes. It is important to realize
through the electrode-patient interface is significantly greater that needle electrodes can have higher impedances than surface
than 5000 Q. (approaching 10,000 Q or more), less biologic cur- electrodes. 52100
This may seem like a contradiction when one
rent enters the instrument and the recorded signal is attenuated. considers that the needle penetrates the skin and does not have
When abrading the patient's skin, it is important to inform the to contend with the impedance of the skin. Actually, the imped-
patient of the mild discomfort about to be experienced and the ance of the surface electrodes are lower than that of the needle
development of a red and tender area about the electrode site. primarily because of the metal from which they are constructed
This area of skin may be uncomfortable for a few days, particu- and the amount of exposed surface area, as well as a dipole
larly when bathing. A scab may also form over the site. The layer that forms about the electrode. That is, the impedance of
patent should be directed to treat this as a superficial wound and gold or silver is lower than stainless steel or platinum. Also, the
keep it clean and dry until healed. surface area of the needle is considerably less than that of the
Surface electrodes can be secured to patients in essentially one surface electrode, which produces an elevation in the imped-
of two ways, depending upon the anticipated duration of the ance. Once the skin is abraded sufficiently, an impedance of
study. For relatively short investigations, 1-2 hours, an electrolyte 2000 Q can easily be achieved with surface electrodes. Needle
paste fills the cup aspect of the electrode and it is firmly pressed electrodes usually have impedances greater than 3000 Q. and
onto the prepared skin. A cotton ball may be placed on top of the can even reach 10,000-13,000 O (personal observation). As
electrode to prevent the paste from drying out and allow a larger long as the input impedance of the amplifier is 10 M Q or more,
contact surface, compared with the electrode, to which tape can the higher needle impedance does not alter the signal apprecia-
apply a counterpressure. Enough tape should be used to permit bly because it is in direct contact with the body's volume con-
mild traction on the electrode lead and yet prevent electrode slip- ductor. This is quite different than surface electrodes. When the
page. It is a good practice to allow a small amount of electrode impedance is much higher than 5000 Q. for the surface elec-
lead slack by taping the electrode's wire to some portion of the trode, the biologic signal is no longer reaching the surface elec-
patient. This prevents the electrode from being dislodged should trode because it cannot cross the skin barrier. The needle
the wire be tugged on accidentally. Although the use of elec- electrode, however, is below the skin and in contact with the bi-
trolyte cream and tape is rather time efficient, this method is po- ologic signal. In this instance, even though the impedance is
tentially subject to more recording and mechanical artifacts. comparatively larger than surface electrodes, the biologic signal
In the event that a period longer than 1-2 hours should be is not lost to the needle electrode because of the high amplifier
necessary, a more satisfactory method of applying electrodes is input impedance (see Chapter 3).
required. The skin site is first prepared as noted above to A number of standard textbooks on SEPs state that caution
reduce impedance. A cup electrode with a hole in the center is needs to be exercised when using needle electrodes because of
secured to the patient by holding it in place with some type of the risk of infection, discomfort, higher electrical noise, and
ease of pulling o u t . - The supposed elevated risk of infec-

52119 282
studies is preferred by these authors. In other regions of the
tion compared with surface electrodes has not been documented body such as Erb's point, popliteal fossa, and spinous processes,
to the authors' knowledge in either short- or long-term studies. the risk of movement or other potential complications must be
On the contrary, with the placement of needle electrodes into the weighed. As discretion at times is the better part of valor, sur-
scalp of normal subjects for approximately 4 hours, no signs of face electrodes may be the most appropriate type to use in these
dermal infection were noted for a follow-up period of 1 week. 100
regions, simply to avoid the possibility of complications. The
Concerns regarding contamination with the human immunodefi- application of a needle electrode about the popliteal fossa may
ciency virus and the hepatitis virus are similar for both needle be appropriate, but one must be aware of the possibility of the
and surface electrodes. Recall that the impedance for surface patient flexing the knee. Should this occur while the needle is
electrodes must be reduced, and this is accomplished by abrad- positioned such that a neural or vascular structure may be punc-
ing the skin sufficiently to expose the patient's volume conduc- tured, obvious repercussions might ensue. Similar concerns can
tor to the metal electrode. It is difficult to reduce the skin's be expressed for inserting a needle electrode in the interspinous
|impedance optimally without also producing some flow of space to record spinal potentials. It is the authors' preference to
serum or possibly blood byproducts, thereby contaminating the use needle electrodes only on the scalp and place surface cup
surface electrodes similar to needle electrodes. Any recommen- electrodes at all other recording locations. The difficulty of se-
dations suggesting that only soap and water are necessary for curing electrodes to the scalp is significantly greater than that of
cleansing surface electrodes while more stringent practices are placement on nonhairy body regions. With practice, applying
somehow necessary for needle electrodes appear contradictory. surface electrodes to nonscalp regions can become rather easy
Once a study is over, the electrode should be carefully cleaned. and avoids possible complications noted above with the use of
A surface electrode should have the paste removed by gently needle electrodes to these same locations.
cleansing it with mild soap and warm water to remove debris
before the electrolyte dries. This electrode should then be prop- STIMULATING ELECTRODES
erly sterilized per the Centers for Disease Control and Prevention
instructions.43,44
Needle electrodes can be handled similarly by Essentially any type of electrode, surface or needle, routinely
placing them in bleach for about 15 minutes and then steam-au- used in nerve conduction studies can also be used to activate the
toclaving them. Any electrode, surface or needle, potentially ex- peripheral nerve in SEP investigations. The major require-
282
posed to Jakob-Creutzfeldt disease or other possible slow virus ment is to accurately locate the electrode over the appropriate
disorders should be safely and immediately discarded. 126
peripheral nerve or dermatome to be excited. Surface electrodes
Commercially available disposable needle electrodes have re- may be placed over a peripheral nerve following a mild cleans-
cently become available similar to disposable electromyographic ing of the skin. The degree of skin abrasion required to reduce
needles, and satisfactory recordings can be achieved. the impedance for recording sites is not necessary for the stimu-
Needle electrodes may actually be more comfortable (per- lus site. If an inordinate amount of current is used to excite a
sonal experience) than surface electrodes, as the skin does not nerve properly and the patient complains of discomfort, one
have to be abraded. - In the only study to compare scalp-
,00 138
should first consider repositioning the electrode to optimize the
recorded SEPs utilizing surface and needle electrodes, no statis- cathode over the nerve. Continued difficulty in exciting the
tically significant difference was detected between waveform nerve may be due to elevated skin impedance hindering current
latency and amplitude. Both recordings were satisfactory re-
100
from reaching the nerve. The skin may need to be abraded
garding artifacts. The conclusion that needle recordings are elec- somewhat to reduce the impedance to current flow in an attempt
trically inferior to surface recordings is simply inaccurate. As to reach the nerve's depth within the tissue. When using needle
noted previously, securing the electrode's lead wire to the patient electrodes to activate a peripheral nerve, the usual caution of
will prevent inadvertent removal during the study. Gloves should near-nerve placement without piercing the nerve should be
be worn at all times when performing any study in which one taken into consideration. A needle electrode requires less cur-
may be exposed to serum, blood, or its constituents. rent duration than surface electrodes, typically 50 ps compared
with 200 ps, because the needle is placed only a short distance
TYPE OF ELECTRODE APPLICATION from the nerve and does not need to "push" current across the
high-impedance barrier of the skin.
The question may arise as to which type of electrode—needle
or surface—is optimal for a specific recording location. As one STIMULATION AND GROUND ELECTRODE
might anticipate, there are no universally agreed upon stan- CONSIDERATIONS
dards. Simply, one may consider this issue with respect to the
potential electrode site. That is, the placement of an electrode Stimulus Isolation/Grounding Considerations. It is recom-
may be scalp or nonscalp. Additionally, if the anticipated elec- mended that all stimulus units be isolated from ground. - - -
96 98 224 265
trode location is covered with hair, one may expect that secur- This is easily accomplished by using a transformer in which the
ing the electrode may be difficult should one wish to not use induced current does not have a physical pathway to enter the in-
collodian. A final consideration might be the amount of poten- strument. Isolating the stimulator assists in minimizing stimulus
tial movement across a particular body segment. artifact from interfering with the SEP waveform and also pro-
As the scalp is usually a hairy region, the use of subdermal tects the patient from aberrant current flow. Fortunately, the
needle electrodes is an ideal consideration for ease of applica- commercial standard for stimulating units utilize the concept of
tion and removal as well as patient comfort. - Appropriate
100 138
isolating the stimulator from ground.
taping ensures the successful completion of the study. Time The use of a ground electrode reduces unwanted current flows
consuming impedance reduction, possibly messy electrolyte and helps produce technically optimal recordings. The skin place-
:reams, and the taping of hair are avoided. The use of needle ment site should be abraded just as for the E-l and E-2 elec-
electrodes for all scalp locations when performing routine SEP trodes. The ground, usually a large metal plate or circumferential
224
band electrode, should be positioned between the E-l electrode and signal with 60-Hz interference and average this noise into the
the stimulating electrodes. ' 9698 224
SEP response. For these reasons, stimulation rates of 5.1 or 2.8,
Stimulus Characteristics. A stimulus pulse of 200-300 ps for example, are frequently used.
duration is preferred for most mixed nerve SEP investiga- Suboptimal stimulus delivery may result in delayed cortical
tions.52,282
The stimulus intensity utilizing the above pulse width potentials of reduced amplitude or altered morphology. - 52 282
should produce a visible twitch of a muscle innervated by the Increasing the stimulus intensity, current or pulse duration, pro-
excited mixed peripheral nerve. A moderately vigorous twitch is duces a normalization of the SEP parameters. Too low a stimulus
desirable, provided it does not cause the patient intolerable dis- may not activate the large and fastest conducting nerve fibers (IA
comfort. - The patient should feel a strong tapping and possibly
6 7
afferents) but instead only excites the slower group II afferents or
a sensation of muscle contraction but not overt pain. Decreasing only a portion of the relatively slower spectrum of group IA affer-
the skin impedance through mild dermal abrasion results in the ents. These slower conducting fibers take longer to reach the
need for less current and, therefore, less pain production. If sig- cortex (delayed arrival), and hence may yield a comparably dif-
nificant adipose tissue overlies the nerve, needle stimulation ferent appearing potential (smaller amplitude and unexpected
may be indicated. When using needle electrodes for stimulation, morphology). Although one might infer that the quality of the
the stimulus duration should be reduced to 50 ps or less to avoid stimulus can be objectively evaluated by observing the peripheral
neural damage, as one no longer needs a long duration to acti- SEP marker potentials at EP and PF for amplitude and/or latency,
vate the nerve optimally. In the event that a pure sensory
238
this unfortunately is not always true. A peripheral neuropathy
nerve or dermatome as opposed to a mixed nerve is studied, the may produce delayed or reduced amplitude potentials that may
stimulus intensity is adjusted to 2.5-3.5 times the sensory appear quite similar to delivering a weak stimulus. A good judge
threshold. - - - Sensory threshold is defined as the current
10 1,1 184 185
of an adequate stimulus for mixed nerve studies is observing an
intensity when the patient first describes a sensation resulting adequate muscle twitch. When performing pure sensory studies
from the stimulating pulse. Raising the current intensity to in patients with altered sensation, one may still attempt to deliver
2.5-3.5 times this value, given patient tolerance, should suffice between 2.5 and 3.5 times the sensory threshold. If this stimulus
in producing a clearly defined sensory SEP. - - 10 184 185
fails because of patient discomfort, increasing the stimulus inten-
As previously stated, the SEP has a rather small amplitude and sity to a mild discomfort level and then reducing the current until
is easily obscured by background noise requiring large numbers a tolerable pulse is delivered can be tried. Of course, caution must
of averages. The process of data collection can take a significant always be exercised in patients with altered sensation.
amount of time. For example, 1000 averages delivered at 2 Hz re- Additionally, higher intensities of stimulation may activate other
quires approximately 8 minutes to collect a response. This time structures, such as nearby mixed nerves, which could produce an
must be doubled if one is to perform two trials. Studying multiple erroneously normal response. Although one can use bilateral
nerves can take quite a long time at the above pace. It is, there- stimulation to investigate the nervous system, unilateral excita-
fore, desirable to increase the delivery rate of the stimulus in an tion of one nerve at a time is usually the best method to elucidate
attempt to decrease the time per examination. Unfortunately, the location of a lesion affecting a peripheral or central portion of
there are limits of how fast an exciting pulse can be delivered the nervous system. Bilateral stimulation may be used to observe
before there is a noticeable deterioration in the waveform. In the for a spinal potential during SEP intraoperative monitoring if uni-
upper limb, the optimal stimulus rate is no more than 5 Hz; in the lateral stimulation does not generate a SEP. 226
lower limb, this limit approaches 2-3 Hz. - Although the phys-
52 282
Stimulator Type. Most commercially available instruments
iologic mechanism underlying these stimulation limits with re- provide the practitioner with the option of using either constant
spect to waveform reproduction is unclear, one can certainly current or constant voltage stimulation. Although either may be
understand how too fast a stimulus adversely affects waveform used, a constant current stimulator is preferable because one is
resolution from purely an instrumentation standpoint. In the assured, within reason, that the same amount of stimulus is de-
upper limb, an analysis time of 50 ms is typically used. livered with each pulse over time even if the impedance under
Specifically, the waveform requires 50 ms to show all of its com- the stimulating electrodes change secondary to electrochemical
ponents on the CRT screen. If one delivers a stimulus prior to the effects between the metal-electrolyte-skin interface. Also, when
complete resolution of the previous potential, there is an overlap measuring side-to-side stimulus thresholds for either mixed or
of waveforms with mutual cancellation of positive and negative pure sensory studies, it is much easier to quantify the amount of
waveforms leading to significant latency and morphology alter- current delivered for the duration of the study.
ation. For an analysis time of 50 ms, the theoretically maximum Typical current intensities vary between 5 mA and 15 mA,
rate of stimulus delivery is approximately 20 Hz (1 potential/50 depending upon the impedance between the electrodes and un-
ms x 1000 ms/1 sec = 20/sec or 20 Hz). In the lower limb, an derlying nervous tissue. Should the patient have a peripheral
282
analysis time of 100 ms is used, yielding a maximum pulse deliv- neuropathy or other pathology affecting the peripheral nervous
ery of 10 Hz (1/100 m s x 1000 ms/1 sec = 10/sec or 10 Hz). The system, the utilization of longer stimulus pulse durations ap-
physiologic limitations are considerably below this instrumenta- proaching 500 ps and higher amperage may be necessary. It is
tion limit. Long latency cortical potentials, greater than 50 ms advisable in such circumstances of altered pain perception to in-
(upper limb), are known to follow peripheral nerve stimulation. crease the current judiciously.
The interaction of these potentials with short latency responses
from a previous stimuli may possibly reduce the theoretical upper
limits of stimulation into or below the above noted ranges. In both REFERENCE DATA A N D CRITERIA
upper and lower limb investigations, the rate of stimulation
FOR ABNORMALITY
should not be an integral of 60 Hz so as to minimize recording
this common environmental noise. Remember that the stimulus Optimal interpretation of SEP information depends directly
rate initiates the CRT sweep and averager. If the stimulus rate is upon a reliable reference data base. The validity of utilizing
divisible into 60, it may be possible to simulate a time-locked reference data is based upon the investigator's skill and technical
30
16 - I .' I „ : \
15 ' » • • -••••"s . — — i , , , . i , \ . , i , , , ,—i—i—• . .
140 150 160 170 180 190 200
A
Height ( o n )
Figure 9-12. LI reference data. Graphic plot of absolute latency t o t h e LI spinous process recording electrode with respect t o patients'
height for posterior tibial nerve stimulation at t h e ankle. (From Chiappa KH: Evoked Potentials in Clinical Medicine. N e w York, Raven Press, 1997,
with permission.)
expertise. All beginning practitioners should develop their ref- which to judge patient results. To produce valid means and stan-
erence values for the commonly performed procedures if fea- dard deviations, the distribution of the recorded reference data
sible. Prior to accumulating this information, it is acceptable base must be determined. If the sample is large enough and sta-
to use another laboratory's published reference values if at tistical analysis reveals a normal gaussian (bell-shaped) curve,
least 20 individuals were used with age ranges comparable the standard deviation and means can be directly calculated. - -
6 7 52
with patients examined by the practitioner and identical instru- On the other hand, if normal values from SEPs, particularly
mentation parameters and technique are employed. small sample sizes, do not yield gaussian distributions, it is nec-
essary to transform these data to a more normal or gaussian dis-
SUBJECT SELECTION tribution through the use of various transformation functions
such as square root, logarithms, or reciprocals before determin-
To develop a reference data base for an individual laboratory, ing means and standard deviations. Two and one-half standard
52
it is critical to select an appropriate population of subjects. A deviations about the mean (98%) of the population tested is
1
careful family and personal history must be taken prior to in- often used to represent "normal data," while a number of labo-
cluding an individual. - - If there is any suspicion of possible
67 52 282
ratories utilize 3.0 standard deviations about the mean (99.5%
nerve pathology, it is best to exclude that person from the of the population tested). - - Using only 2.0 or less standard de-
6 7 52
"normal population." It is also recommended that a complete viations is considered inappropriate by most investigators, since
history for any type of drug use (legal or illicit) be pursued, as an unacceptably high number of false-positive results will be
such use may adversely affect the SEP results. obtained. When applying standard deviations about the mean,
An ideal control population should contain a comparable one may occasionally note that a non-sensical negative number
number of age-matched males and females. In children and can be obtained, e.g., 3.0 ± 1.9. In this example, 2 standard de-
adults, age-specific normal values should be determined by viations below the mean is - 0 . 8 ms, which occurs because of
decade, including the eighth and ninth if possible. For infants, the nongaussian distribution of the data.
reference data are obtained by months, while the perinatal
period requires reference values based on w e e k s . OptimalLy,
58-60
each subgroup should contain at least 20 individuals. One FACTORS AFFECTING THE SEP
author suggests that prior to performing any SEP, one must ex-
amine a minimum of 35 control individuals and use the same PATIENT COOPERATION
parameters on suspected abnormal persons. 52
One of the most important and often forgotten issues in per-

ABNORMAL CRITERIA forming SEPs is the patient's understanding of, and subsequent
ability to participate in, the examination. The SEP study is often
Initially, the pool of normal values obtained must be statisti- a new experience for the patient. He or she may enter the study
cally analyzed to determine a mean and standard deviation from with some level of fear and anxiety. An anxious or uncomfortable
person cannot optimally participate in the study. Frequently, an not necessary. - - - Of course, should one of these agents
52 242 272 273
SEP evaluation may last one or more hours, and the patient's be used, precautions must be given to the patient regarding the
cooperation is perhaps the most crucial portion of the investiga- operation of a motor vehicle or performing any activity requir-
tion. Nervous patients often produce significant conscious and ing mental alertness.
unconscious muscle contractions. The electrical activity from
these contracting muscles can completely, obliterate even rela- PATIENT HEIGHT
tively large normal responses. Relaxation is often the key to a
successful study. One of the best ways to gain an individual's When performing lower limb SEP studies, it is important to
complete cooperation is to explain fully, in terms the patient can record the patient's height. As one might anticipate, the conduc-
understand, what is about to be done and why. It is often possi- tion time for tibial nerve stimulation, for example, from the
ble to visually observe a patient's facial countenance and body medial malleolus to the cortex, is longer for a comparably tall
posture relax once the individual comprehends the SEP test and than short person because of the increased length of both the pe-
experiences its benign nature following the first few stimuli. It ripheral and central conduction pathways. The patient's height
is advisable to then re-zero the averager's stimulus counter at is used as a correction factor for more meaningful latencies
this point and begin data collection without interrupting the when comparing scalp, spinal, or central conduction times.
rhythm of stimulation. Graphic plots (Figs. 9-12, 9-13, and 9-14) and regression
51
To assist in the above goal of patient cooperation and relax- equations (see below) are available to assist in the determina-
ation, the appearance of the room must not be underestimated. tion of reference peripheral and central peak latencies to correct
A relatively quiet, temperature-controlled, and attractively dec- for height. When considering central conduction times for
orated room with neatly aligned electrodes and other equipment median nerve excitation, N13 to N20 inter-peak latency, there
conveys a professional and organized atmosphere. The patient's appears to be little correlation with arm length or height. - 53 159 223
plinth should be comfortable, with clean bed sheets and multi- Apparently, the central conduction pathway from the cervical
ple pillows. Prior to connecting the electrodes, the patient spinal cord to the cortex is_relatively the same in tall and short
should be offered the opportunity to use the lavatory because individuals. The lumbar N20 to cortical P37 (P40 as designated
the test can be quite lengthy. Following the placement of elec- by some authors) inter-peak latency (central conduction time)
trodes, the supine position should not be the only one consid- for tibial nerve stimulation, however, is correlated with an indi-
ered. It is reasonable for the patient to assume whatever is the vidual's height (Fig. 9-14). - 52 53
most comfortable position, provided it does not compromise the

recording of SEP potentials. In the authors' opinion, a com- AGE
pletely relaxed but not asleep patient is desirable. The patient is
asked to keep track of the stimuli in some manner that is not Multiple investigations using different methodologies produce
mentally taxing, but promotes a subdued level of alertness. The conflicting results regarding the influence of age on various
use of mild hypnotics such as chloral hydrate or diphenhy- SEP parameters. The affect of age on SEP cortical amplitude
dramine is recommended by some investigators, but this is usually suggests that there is a "U"-shaped curve in that infants and
Figure 9 - / 3 . S c a l p r e f e r e n c e d a t a : T i b i a l n e r v e . Graphic plot of absolute scalp latencies (P37) following tibial nerve stimulation at the
ankle correlated with patient height. (From Chiappa KH: Evoked Potentials in Clinical Medicine. New York, Raven Press, 1997, with permission.)
25
24
23
22
« 21
6
5 20
a
5 19
d
° 17
* » 16
K 15
i
5 14
13
12
11
10140 150 160 170 180 190 200

Height ( e n )
Figure 9-14. Central conduction time: Tibial nerve. Graphic plot of central conduction times (P37-LI) for tibial nerve stimulation at the
ankle correlated with height. (From Chiappa KH: Evoked Potentials in Clinical Medicine. New York, Raven Press, 1997, with permission.)
children display larger amplitudes than young and middle-aged central nervous system following tibial nerve stimulation has
adults. - - This last group, however, has smaller ampli-
57162 208 274
only once been studied in depth. This investigation examined
181
tudes than persons older than 40 years. Spinal SEP potentials

162
the conduction times along the spinal cord segment and the por-
demonstrate a different pattern. In infants, the lumbar and more tion between the neck and cortex. Both segments demonstrated
rostral spinal potentials were larger than those recorded in chil- slowing of conduction velocity with age. Intra-cortical conduc-
dren and adults. The cervical spinal N13 potential is similar
57
tion time as measured by cortical inter-peak latencies e.g.,
for subjects between the ages of 10 and 39 years, but then dis- N20-P22 or P37-N45, has been shown to increase with age - 80 208
plays a steady amplitude decline for those of 40 years and as well as demonstrate no change. A final consensus regard-
180
greater. The reason for these amplitude findings is unknown

163
ing the effects of aging on central conduction is not yet at hand,
but may be related to the extent of myelination in childhood and but there appears to be a small central conduction velocity d e -
selective fiber loss with aging. 103
cline in persons over 60 years of age.
Unlike age-related amplitude effects, the correlation between
age and conduction time (conduction velocity) is less clear. GENDER
There is agreement that the peripheral nervous system initially
demonstrates slow conduction velocities until about 4 or 5 years A few studies have attempted to compare SEP waveforms be-
of age, at which time adult values are achieved. - - Adult 19 127 300
tween men and women. There is a consistent prolongation of
SEP conduction values are usually achieved between 5 and 8 potential latencies in men. - The increase is most likely a
162 164
years of age. The central conduction time characteristics with result of body stature, as men demonstrate statistically signifi-
respect to age are less straightforward. The majority of studies cant height differences compared with women. - - The central
4 53 223
have focused on the time of conduction between the cervical conduction time appears to bear no relationship to the gender of
N13 to cortical N20 potentials following median nerve stimula- the subject. Several studies suggest that the SEP amplitudes,
1
tion. Several studies investigating a wide range of ages found however, are slightly larger in women, but more work is needed
that although the peripheral nerve conduction velocity slowed to substantiate these findings. At the present time, it seems
53164
with increasing age, the central conduction time remained con- acceptable to pool both latencies and amplitudes from men and
stant. - A larger number of investigations, however, found
223 333
women when developing reference data bases as long as a pop-
that not only did the peripheral nerve conduction velocity slow, ulation representative of various heights is included.
but so did the central conduction velocity. - - - - - One 4 5 93162 ,81 290 292
study comparing central conduction time in younger and older TEMPERATURE

individuals found that in a population with a mean age of 31.6 ±
14.1 years, the central conduction velocity was 55.8 ± 12.1 m/s Temperature is an important factor affecting nerve conduc-
compared with 42.4 ± 1 3 . 1 m/s for persons 74.1 ± 7.5 years. 93
tion velocity. Because the upper and lower limbs consist of sig-
Another group of investigators found the central conduction nificantly less mass than the trunk, they are more prone to
time increased 0.3 ms/yr for persons between the ages of 50 and fluctuations in temperature with respect to the environment
60 years. The central conduction time along the length of the
162
under routine laboratory conditions. As a major portion of the
somatosensory pathway consists of the peripheral nervous Central conduction times are also prolonged. When performing
system, and is, therefore, subject to temperature variations, it is intraoperative SEP monitoring, the avoidance of these agents
important to be cognizant of temperature's role in altering SEPs. has been recommended. These effects are countered by em-
118
A cool limb, particularly digits when performing segmental or ploying a balanced anesthesia technique that uses a strong nar-
dermatomal investigations, may prolong the anticipated laten- cotic in addition to a muscle relaxant plus a weak anesthetic like
cies to all SEP peaks. Although a room temperature of 20-22°C nitrous oxide, or low concentrations of isoflurane. The pre-
226
is recommended, - it is the authors' opinion that this precau-

52 282
ceding approach appears to provide the necessary anesthesia
tion by no means guarantees an acceptably warm limb. The sur- without significantly compromising SEP recordings.
face temperature of the upper limb at or proximal to the site of
nerve stimulation should be maintained at 32°C or more, while SLEEP
the lower limb should be approximately 30°C or higher. This
can be accomplished with hot packs, radiant warmers, or other The ideal SEP should be free of artifact, especially that aris-
means. Of course, if a laboratory has standardized its data at a ing from muscle, to produce clearly defined peaks so that laten-
particular body temperature, then by all means this is the tem- cies can be easily defined. Unfortunately, waveform
perature at which the data should be collected. Use of interpeak contamination from muscle electrical activity is frequently en-
latencies subtracts out peripheral slowing due to cooling of the countered even in apparently relaxed patients. Attempts to cir-
limbs and can be employed when it is not practical to warm the cumvent this interference by encouraging the patient to sleep
limbs. can alter waveform morphology and latency. Stage II sleep may
Central core temperature is subject to less fluctuation with re- result in up to a 2.4-ms prolongation of latency in the wave-
spect to emotional or environmental conditions. Studies investi- forms obtained in tibial nerve SEP studies. In general, sleep
326
gating the effects of central temperature modulation on SEP tends to reduce the amplitude and increase the latencies of
morphology or latency are limited in number. In one upper limb recorded SEPs. Sleep has been shown to reduce the amplitude
SEP study, the investigators elevated the central temperature of of dermatomal SEPs or abolish the responses altogether, and
subjects by 1°C and noted that the_N13 potential's absolute la- hence should be avoided when performing these studies. It is 30
tency decreased by 0.7 ms, the N19 latency decreased by 1.0 preferable to have the patients relax in the supine position but
ms, and the Erb's point potential was reduced by 0.18 m s . On 217
maintain a level of alertness. This can be accomplished by
the other hand, a second investigation considered the effects of asking the patient to clench and relax the mandible several times
temperature on SEPs by measuring latencies in patients with in succession so as to be aware of how to relax these muscles.
fevers (38.0-39.7°C) and again following fever resolution. 187
Additionally, this same maneuver is performed with the
No significant latency changes were noted in these patients. In frontalis and paraspinal muscles. Once the patient is aware of
cancer patients receiving whole body hyperthermia up to core these muscles and what they feel like in the relaxed state, it is
temperatures of 42°C, an absence of cortical SEPs occurred. 95
often possible to reduce muscle artifact dramatically and obtain
Whether this SEP disappearance is a result of peripheral nerve relatively quiet recordings. The patient is also asked to just be
or central pathway conduction failure is unclear. Patients receiv- peripherally aware of the stimulus and make a mental note of
ing hypothermia for intractable seizure control with core tem- the stimulus delivery without actually counting the number, thus
peratures lowered to 29°C demonstrated a cortical SEP latency maintaining some level of alertness. Also, encouraging the pa-
prolongation and amplitude reduction. Again, these may be
285
tient to fix his or her gaze or gently close the eyes can limit as-
central or peripheral effects. sociated external ocular muscle artifact.
MEDICATION REPRODUCIBILITY
Although it is generally believed that medication in ambula- Any technique used to diagnose a possible neural lesion must
tory patients has little effect on SEP latency, amplitude, or mor- be reliable and reproducible. Unfortunately, the reliability and
phology, the relationship between various drugs and SEPs has reproducibility of SEP potentials have not been investigated
not been extensively studied. Phenobarbital in therapeutic systematically or extensively. A few anecdotal reports note that
dosages does not appear to alter the central conduction time in the SEP latencies did not reveal statistically significant changes
comatose p a t i e n t s . In cats, therapeutic levels of pentobar-
84142143
over time within the same patients. - - There is a need for
29 218 310
bital also does not effect SEPs. Utilizing auditory evoked po-
291
large studies examining the variation of latencies and ampli-
tentials, phenytoin has been shown to increase inter-peak tudes within the same subjects and between different individu-
conduction t i m e s . These effects were not noted for primi-
143144
als over time.
done or carbamazepine. Diazepam has been recommended to One study examined spinal responses in both healthy persons
reduce muscle artifact thereby improving SEP recordings, but and in individuals with spinal cord injuries and head injuries. 29
these authors did not report the results of latencies between sub- An average test/retest period of 16.2 months demonstrated that
jects who received the drug and those who did not. Should di-242
the spinal potentials were very stable with respect to amplitude
azepam cause a patient to sleep, one may then anticipate a and latency in all groups examined. Correlation coefficients of
number of SEP changes (see below). 0.84 and 0.78 for controls and patients, respectively, were found.
The largest group of patients who have been studied with re- Additional studies are needed to corroborate these results.
spect to SEP medication effects are those undergoing surgery
and who receive a general anesthetic. Essentially, the volatile
agents in high concentrations such as the halogenated hydrocar- SEP TECHNIQUES
bon inhalation agents (halothane, enflurane, and isoflurane)
produce an increase in cortical potential latencies in addition to This section of the chapter is by no means meant to be an ex-
a reduction in cortical potential amplitudes. - - - - - -
94 236 237 241 259 260 319
haustive litany of the multiple techniques available to elicit
SEPs. The major upper and lower limb mixed and pure sensory and the palmaris longus medially. Although it is convenient to
nerve SEP methods are described and accompanied by available place a bar electrode, cathode proximal (anode at the level of
published reference data. Although there is controversy regard- the distal wrist crease), between these two tendons, separate
ing some SEP methodologies with respect to clinical applicabil- disk electrodes can also be used. A bar electrode is convenient
ity, these concerns are not discussed at this time, but examined because it is easy to firmly secure this electrode combination in
in detail in later sections of the book dealing with the diagnosis place. It is a good idea to abrade the stimulation site gently with
of specific disorders. It is important for the beginner to realize an alcohol swab prior to electrode placement with adequate
that compared with nerve conduction and needle electromyo- conduction paste. The impedance is usually not recorded for the
graphic techniques, diagnostic SEPs are relatively new. As a stimulation site because all that is required is submaximal nerve
result, fundamental questions relating to neural generators, vari- excitation. The stimulus usually consists of a pulse duration of
ability of latencies and amplitudes, reference data, appropriate 200 ps delivered at 5.1 Hz or less at an intensity necessary to
applications, and other important issues remain. These knowl- produce a moderately vigorous thumb twitch. - - - 6 7 32 77
edge gaps result in multiple gray areas directly affecting diag- Ground Electrode. The ground electrode should be located
nostic decisions. Tolerance for the unknown as well as a between the stimulating electrodes and the first recording elec-
willingness to err oh the conservative side in diagnosing pathol- trodes to yield optimal waveforms. This implies that when per-
ogy is mandatory when performing SEPs. forming bilateral studies, it becomes necessary to relocate the
ground electrode. It is often possible to place the ground elec-
UPPER LIMB SEPS trode in the midline, for example, on the sternum or forehead,
and still record clear SEPs. Should artifact be a problem, the
Upper limb SEP techniques may be divided into three pri- ground electrode may be moved to the limb between the stimu-
mary categories. The first major group of studies consists of lus and recording electrodes.
stimulating a mixed peripheral nerve and recording both pe- Recording Electrodes. When the median nerve is investi-
ripheral and central (spinal and scalp) waveforms. Multiple gated, the placement of electrodes is usually standard as d e -
studies have examined the most appropriate manner in which scribed below; however, the manner in which they are
to elicit these potentials and their clinical utility. Additionally, connected to the instrument varies with the number of channels
these waveforms are relatively large and easy to obtain by the available. It is possible to obtain the necessary data with either a
novice. The second category of upper limb SEPs pertains to 4- or 2-channel instrument. The methods for utilizing both types
pure sensory nerves and are occasionally referred to as seg- of instruments and electrode placement are described below.
mental SEPs. Segmental studies primarily involve the excita- Erb's Point. An electrode is secured just superior to the mid-
tion of a major sensory nerve composed of fibers originating portion of the clavicle, 2 - 3 cm superior to the bone where the
from primarily one nerve root innervating the site of stimula- "clavicular head of the sternocleidomastoid muscle inserts. T h e
tion. The third category of upper limb SEPs is dermatomal peripheral nerve volley is recorded as it traverses the brachial
SEPs, which do not involve stimulation of a nerve trunk, but plexus beneath the electrode. - It is a good idea to perform
6 7177
rather an area of skin subserved by a given dermatome. These left/right comparisons. In a 4-channel recording, the fourth
latter two techniques are performed with the goal of recording channel records the Erb's point potential ipsilateral to the side
only from the scalp, as the peripheral and spinal potentials are of stimulation, while the opposite Erb's point electrode is uti-
extremely small and often lost in the surrounding muscle and lized as the E-2 electrode. - - In a 2-channel recording, the
6 7 77
environmental noise. As one would anticipate given this infor- Erb's point electrode is the E-2 electrode for channel 2. The E-l
mation, a check on the peripheral nervous system is needed. electrode placed in channel 1 is noted below.
Peripheral nerve status can be determined through mixed nerve Cervical Spine (C2, C5, C7). The cervical spine location
SEP studies, segmental or dermatomal SEPs, or routine nerve functions as a marker from which to measure the central con-
conduction velocity determinations. duction time. There are three acceptable locations for the cervi-
cal electrode: over the C2, C5, or C7 spinous processes. These
MIXED NERVE SEPS three locations yield essentially the same latencies, although the
C2 region may result in slightly larger amplitudes compared
For the purposes of this discussion, a mixed nerve is defined with the other two locations. In both 2- and 4-channel record-
as a major branch of the peripheral nervous system containing ing, the cervical electrode is connected to the E-l port, while
both motor and sensory fibers. The afferent pool of nerve
330
FpZ' serves as the E-2 electrode. - - 6 7 77
fibers excited convey not only cutaneous sensation but also Scalp. Scalp electrodes record cortical SEP waveforms and
those fibers relaying information from the large muscle spindle utilize the international 10-20 system for the C3' and C4' elec-
afferents. In the upper limb, two mixed nerves are typically ex- trode positions. - - - In the 4-channel system, the first chan-
46 47 170 232
cited to produce SEPs. The median and ulnar nerves are rela- nel contains C3'(left) or C4'(right) referenced to FpZ'. - - - 6 7 52 282
tively easy to stimulate and result in large and reliable SEP This scalp montage should routinely yield relatively large and
waveforms from both the peripheral and central nervous sys- noise-free cortical recordings. If a 2-channel system is used, the
tems. The median nerve usually produces slightly larger poten- C3' or C4' electrode is placed in the E-l port of channel 1 and
tials than the ulnar nerve, - as one might anticipate given its
77 329
the Erb's point electrode serves as the E-2 electrode. Because
fiber content and larger cortical representation. - 22134 330
the Erb's point potential occurs at approximately 10 ms, while
the C37C4' potential is not recorded until approximately 20 ms,
Median Nerve there is no difficulty in recording both of these two potentials on
Stimulation. When performing mixed nerve SEP studies for the same channel. Of course, the Erb's point potential is in-
the median nerve, one typically places the stimulating elec- verted compared with its E-l port location in other montages,
trodes over the median nerve at the wrist. - - The nerve is lo-
52 77 174
but its latency remains essentially unchanged, as does the side-
cated between the tendons of the flexor carpi radialis laterally to-side amplitude comparisons for this technique (Fig. 9-9).
Table 9-6. Mean Central Conduction Time (ms) used to detect far-field potentials. This is accomplished by plac-
Mean C C T SD ing C3' or C4' into the E-l channel and using the contralateral
Erb's point electrode as the E-2 electrode. - - Any other re-
6 7 52
5.6 0.5 159
sponse the investigator is interested in examining can also be

5.6 0.6 293
placed into the fourth channel. Obviously, if a 2-channel instru-
5.5 0.7' 06
ment is utilized, simultaneous optional recordings are limited.
5.7 0.6 130 Instrument Parameters. The most important aspect regard-
ing instrumentation parameters is that the conditions under
5.7 0.5 302
which the reference data were originally acquired must be repro-

5.7 0.5' 61
duced. The performance of upper limb SEPs requires an analysis
5.3 0.5 278
time of about 50 ms, which is accomplished by a sweep speed of
5 ms/div. This total analysis time may need to be increased by 10
5.8 0.8 258
ms or more with severe lesions. Filter setting of 10 Hz to 3000

5.8 . 0.6 4
Hz are adequate to include the major subcomponent frequencies

5.5 0.4 176
contained in the waveforms while simultaneously limiting noise.
5.1 0,9324 A final signal amplification of about 2 pV/div usually results in a
waveform contained on the CRT screen, but occasionally, more
5.4 0.4312
or less sensitivity may be required. Depending upon the size of
5.9 0.5 39
the SEP and the background noise, between 300-500 and 1000
The central conduction time refers to the interpeak latency between N13 and averages are sufficient to resolve most SEPs. At least two trials
Nl9. should be performed for each recording site. - - - 6 7 52 282
Modified from Cant BR, Shaw NA: Central somatosensory conduction time: Reference Data. As previously noted, it is a good idea for
Method and clinical applications. In Cracco RQ, Bodis-Wollnar I (eds): Evoked
Potentials. NewYork.Alan R. Liss Inc., 1986, pp 58-67. each individual to perform sufficient SEPs to obtain a reference
data base. This information should then be compared with pub-
lished values. Multiple studies have been published on the cen-
Alternate Recordings. In the above-noted electrode loca- tral conduction (N13-N20 latency) demonstrating a mean
tions, the 4-channel instrument has one remaining channel that conduction time of 5.6 ms (Table 9-6). " - - Regression
112 3 160 257 293
is not utilized. This open channel can be used to monitor a equations are also available to account for height and a g e . 223
number of other responses. It is possible to detect the median Regarding a regression equation for the EP potential latency,
nerve's peripheral response distal to the brachial plexus by one finds EP = 0.086H + 0.038A - 5.88 for men and EP =
recording at the antecubital fossa. An E-l recording electrode
77
0.054H + 0.03A - 0.59 for women, where H is the subject's
is placed on a point half-way between the medial and lateral height in centimeters and A is the age in_years. An equation de-
humeral epicondyles medial to the biceps brachii tendon. An E- scribing an N13 latency for men is N13 = 0.099H + 0.045A -
2 electrode is then located over the ulnar nerve or the olecranon 4.98, while female latencies can be described as N13 = 0.064H +
process. As previously noted, the second channel can also be 0.035A + 0.78. A cortical N19 latency for men can be arrived at
Table 9-7. Mixed Median Nerve (Wrist) SEP Reference Data

Latency(ms)
To Peak L/R To Peak UR To Peak Max UR
Recording Site
AF (A) 4.3 ± 0.3 0.24-0.99 (B) - — (C) - —
EP 9.9 ± 0.6 0.44-1.74 9.7 ± 0.76 0.2 ± 0.2 9.6 ± 0.7 0.5
C2 ( N i l ) 13.4 ±0.3 0.43-1.52 13.5 ± 0.92 — 13.2 ± 0 . 8 0.6
C37C4* (NI9) 19.2 ± I.I 0.72-3.10 19.0 ± 1.02 — 18.9 ± 1.0 0.9
C37C4' (P22) 25.2 ±2.1 1.08-4.05 22.0 ± 1.29 — — —
Interpeak Latency
AF-EP 4.9 ± 0.2 — — — — —
EP-NI3 3.8 ± 0.3 — 3.8 ± 0.45 0.2 ±0.17 3.5 ± 0.4 0.8
EP-NT9 9.3 ± 0.4 — 9.3 ± 0.53 0.2 ±0.21 — —
EP-P22 __ — 12.3 ± 0 . 8 6 0.3 ± 0.24 — —
NT3-NT9 5.6 ± 0.5 5.5 ± 0.42 0.3 ± 0.25 5.8 ± 0.5 0.5
Amplitude (pV)
AF 3.4 ± 0.7 0.11-2.54 — — — —
EP 2.1 ± 0 . 6 0.17-0.94 3.0 ± 1.86 — 5.4 ± 2.5 49%
NI3 1.9 ± 0 . 3 0.16-1.28 2.3 ± 0.87 — 2.9 ± 1.3 46%
Nl9 0.6 ± 0.2 0.19—1.81 1.0 ± 0 . 5 6 41.7% 2.8 ± 1.6 50%
NL9-P22 2.1 ± 0 . 9 0.21-2.31 2.2 ± I.I 25.7% — —
In A, filter setting for scalp recorded SEPs are 0-2,000 Hz while all other responses utilize a bandwidth of 10-3,000 Hz. Bandwidth of B is not specified but likely cor-
respond to 10-3,000 Hz. Filter settings for C are 20-2000 Hz. Amplitudes of peripheral and cervical responses are measured from baseline to peak while cortical
potentials are measured from baseline to peak unless otherwise specified e.g., peak-to-peak for NT9-P22. L/R signifies the left/right difference. AF: antecubital fossa;
L/R: left/right difference. From DeLisa et al. (A), Chiappa (B),andYiannikas (C).
77 52 329
HZO Ulnar Nerve

S t i m u l a t i o n . The ulnar nerve at the wrist can be readily
activated either just medial or lateral to the tendon of the
A v / \ flexor carpi ulnaris. Similar to the median nerve, the anode is
placed at the level of the distal wrist crease, while the cathode
is located several centimeters proximal to this site. A bar elec-
trode is convenient. Gentle skin abrasion with an alcohol
swab may be of assistance in reducing the current intensity
needed to excite the nerve and reduce patient discomfort. A
stimulus pulse with a duration of 200 ps, rate of 5 Hz or less,
and an intensity capable of producing a moderately vigorous
but nonpainful twitch of the abductor digiti minimi is usually
employed. - 52 77282
G r o u n d . The general rule of placing the ground electrode

|2.5/*V between the stimulating electrodes and first set of recording
electrodes should be followed. One may also wish to locate the
5ms ground electrode in the midline on the sternum or forehead if
the level of background noise or stimulus artifact permits.
Figure 9 - / 5 . U l n a r n e r v e S E P . Ulnar nerve SEP utilizing a 2-chan- Recording Electrodes. Similar practices as those utilized
nel technique. N o t e t h e inverted EP potential preceding t h e cortical for the median nerve also apply to the mixed ulnar nerve stud-
response on channel I (A).The second channel (B) records the N I 3 ies. As always, the more channels an instrument has, the more
potential with a C7S-FpZ' montage. flexible one can be regarding how the channels are connected to
the recording sites. Identical recording sites are used for the
ulnar nerve as those for the median nerve (see above).
by using N19 = 0.095H + 0.049A + 1.19, and for women the Erb's Point. Essentially the same type of preparation and lo-
equation is N19 = 0.085H + 0.043A + 2.72. The equations de- cation site as that used for the median nerve are recommended.
j scribe a gender difference for SEP latencies, but not all authors Because the ulnar nerve tends to produce a waveform with a
agree that gender results in a significant difference for SEP arrival slightly smaller amplitude than that obtained for the median
' times. These regression equations provide expected mean values; nerve, meticulous technique is suggested. In a 4-channel instru-
limits of normal depend upon the standard deviation for each ment, the E-l Erb's point electrode is referenced to Erb's point
variable. A number of investigators' reference data are included •contralateral to the side of stimulation. - For a 2-channel instru-
6 7
for the reader to compare reproducibility between reputable labo- ment, the Erb's point electrode is used as the E-2 electrode for
ratories and for technique verification purposes (Table 9-7). This channel 1 (see median nerve) (Fig. 9-15).
table consists of data from individuals with varied heights and Cervical Spine (C2, C5, C7). Either C2S, CSS, or C7S can
i ages, although the specific parameters of these persons are not be used for recording the spinal potential. With 4 channels, the
always provided. Thus, the regression equations may be of value. E-l cervical spine electrode is referenced to the forehead site,
T a b l e 9-8. Mixed Ulnar N e r v e ( W r i s t ) SEP R e f e r e n c e D a t a

Latency(ms)
To Peak L/R To Peak UR To Peak MaxUR
Recording Site
AF (A) 4.3 ± 0.5 0.31-1.18 — (C)- —
EP 9.9 ± 0.8 0.39-1.80 10.7 ± 1.3 — 10.0 ± 0 . 9 0.4
C2 ( N i l ) 14.0 ± I.I 0.38-1.44 14.1 ± I.I — 13.9 ± I.I 0.5
C37C4' (NI9) 19.5 ± I.I 0.45-1.77 19.6 ± 1.4 — 19.3 ± 1.2 0.6
C37C4' (P22) 24.1 ± 2.6 0.95-4.22 — — — —
Interpeak Latency
AF-EP 5.7 ± 0.5 — — — — —
EP-NI3 4.3 ± 0.8 — 3.5 ± 0.8 — 4.0 ± 0.4 0.5
EP-NI9 9.6 ± 1.4
_ — — — —
EP-P22 — — — — — —
Nl3-Nl9 6.0 ± 0.8 — 5.6 ± 0.8 — 5.3 ± 0.4 0.6
A m p l i t u d e (pV)
AF 2.1 ±0.6 0.01 —1.51 — — — —
EP 1.5 ±0.4 0.17-1.83 3.2 ± 1.4 — 2.9 ± 1.6 48%
NI3 0.9 ±0.31 0.13-0.92 1.6 ± 0 . 7 — 1.7 ± 0 . 8 56%
Nl9 I.I ±0.6 0.19-1.62 1.4 ± 0 . 6 — 1.8 ± I.I 55%
NI9-P22 1.9 ±0.8 0.1-2.88 — — __
In A, comparable filter settings to median nerve used. For data in B, filter settings are 16 Hz-1.6 kHz. For C filter settings are 20-2000 Hz (peripheral potentials) and
2-2000 Hz cortical potentials. L/R signifies the left/right difference. AF: antecubital fossa; L/R: left/right difference. From Delisa et al. (A), Ganes (B), and Yiannikas
77 129
et al. (C).
329
i.e. FpZ'. In a 2-channel setup, the second channel contains the intensity is slowly increased until the patient first notes the tin-
E-l cervical spine electrode, also referenced to FpZ'. gling sensation produced by the stimulating electrodes, which
Scalp. Again, the international 10-20 system is utilized to usually occurs at approximately 3-A mA. - - This defines 184 185 282
place the cortical electrodes in proximity to the somatosensory the patient's sensory threshold. The stimulus intensity is then
cortex by using the C3' or C4' recording sites as for the median slowly increased to between 2 and 3 times the individual's sen-
nerve. - With 4 channels, either C3\ov C4* is referenced to
170 232
sory threshold, which approximates 6-12 mA. If the patient
FpZ'. A 2-channel instrument requires the appropriate cortical cannot tolerate this excitation, it is decreased until a strong but
electrode to be placed into the E-1 amplifier port of channel 1 tolerable pulse is delivered. Of course, if the patient is hyper-
using the contralateral Erb's point electrode as the E-2 (Fig. 9¬ sensitive and the current is insufficient to generate maximal
15). In the 2-channel instrument, the rationale for which the SEPs, the study becomes invalid because insufficient current in-
electrodes are connected to the amplifier is the same as that for tensities result in low amplitude and delayed potentials. - As 52 282
the median nerve. previously noted, a constant current stimulator is preferred be-
Alternate Recordings. If one has the luxury of 4 channels, cause this enables the practitioner to easily quantify the amount
all channels can be used. The mixed nerve action potential from of current delivered to ensure proper neural excitation with each
the ulnar nerve can be recorded by placing an E-1 electrode over pulse during the study.
the ulnar nerve in the ulnar groove with an E-2 electrode located Three digits can be analyzed when performing median nerve
anteriorly between the humeral epicondyles. This is the reverse segmental studies. - - - Ring electrodes are located on any
77 129 264 297
of the electrode orientation noted above for median nerve of the first three digits of the hand. Although some authors rec-
recordings. Of course, the fourth channel can simply be turned ommended a 4-cm separation between cathode (proximal) and
off as well. anode, this really is not necessary for stimulation. The ring
77 297
Instrument Parameters. The same parameters used for the electrodes used are the same when obtaining antidromic median
median nerve are employed. Because of the comparatively sensory nerve action potentials. The electrodes should be placed
smaller ulnar nerve amplitudes, the practitioner should be pre- such that the opening of the ring's noose is located posteriorly,
pared to increase the amplifier's sensitivity. thereby maximizing contact of the cathode and anode with the
Reference Data. The ulnar nerve has not been as extensively median nerve on the volar aspect of the digit. Recall that the
studied as the median nerve. A number of investigators' pub- radial nerve innervates the dorsum of the finger and is not the
lished normal values are provided to assist the reader in per- nerve to be excited at this time.
forming these studies as well as to provide reference values to Ground. Because of segmental SEPs' small amplitude,
compare one's own reference data (Table 9-8). - - 52 77 129
meticulous electrode placement is critical. It should be placed
proximal to the recording electrode. This position will vary de-
SEGMENTAL SENSORY NERVE SEPS pending upon the placement of the recording electrodes. As de-
scribed below, only cortical electrodes are performed and,
Somatosensory evoked potentials arising from the excitation therefore, the ground is positioned on the mid-frontal region of
of sensory nerves as opposed to mixed nerves can also be the skull or on the sternum.
recorded. One can preferentially activate a peripheral sensory Recording Electrodes. The number of recording electrodes"
nerve, e.g., superficial radial (segmental stimulation), or a por-
77
used and their location are dependent upon the particular tech-
tion of the skin innervated by a nerve at a more distal site, e.g., nique. The relatively small amplitude of the segmental SEP can
lateral aspect of the foot (S-l dermatomal stimulation). - 184 185
render the Erb's point and cervical spine sites somewhat difficult
Generally, segmental stimulation results in more easily obtain- to record. Although it is by no means impossible to obtain these
able and somewhat larger responses than dermatomal SEPs, as responses, there is little margin for error and meticulous tech-
297
more nerve fibers are excited. Compared with mixed nerves, nique combined with a relaxed patient is needed. Cortical poten-
few studies have documented in a controlled manner either tials, on the other hand, are significantly less difficult to record j
normal values or the utility of pure sensory SEPs in diagnosing because the central amplification effect produces relatively |
pathology. easily identifiable segmental cortical potentials in most pa-
tients.107109
The central amplification effect merely amplifies
Median N e r v e the magnitude of the cortical response with respect to the periph- j
Median nerve segmental stimulation can be performed by eral impulse. For example, if a peripheral sensory response I
stimulating any of the median nerve's digital branches innervat- cannot be recorded, chances are a cortical SEP will continue to
ing the first three digits of the hand. The main differences be-
297
be detected because of the so-called central amplification effect.
tween median nerve wrist and digit SEP waveforms are twofold. It is the authors' experience, however, that some normal individ-
First, the latency of segmental responses are slightly prolonged uals can have rather small cortical potentials.
compared with mixed nerve stimulation. - - This latency
77 129 264
Erb's Point. This is the same location as that for mixed
prolongation is a result of the distal placement of the stimula- median or ulnar nerve studies. - The absence of a response at
77 264
tion site plus the small-diameter fibers in the digits conducting this location may simply be due to the small nature of the re-
impulses more slowly than the larger caliber wrist fibers. The sponse combined with incomplete patient relaxation and does
second waveform difference is the comparably smaller ampli- not necessarily reflect pathology.
tudes for segmental stimulation at all recording sites. Cervical Spine. An identical location as that for mixed
Stimulation. When attempting to generate segmental SEPs, median and ulnar nerve studies is used. Because of the compa-.
the sensory portion of the peripheral nervous system must be rably smaller segmental SEP spine amplitudes, one may wish to
preferentially activated. As a result, one does not look for a consider the C2S location if spinal recordings are attempted.
muscle twitch. Instead, a stimulus of 2.5-3 times sensory The somewhat larger amplitude at this location may assist in
threshold is applied. - Again, a stimulus pulse duration of
10111 282
documenting the desired waveforms more readily than at C5S
200 ps is preferred with a rate of 5 Hz or less. The current (voltage) or C7S. - 264 297
Scalp. Scalp-recorded potentials are obtained by placing the 0.0093)A + (0.155 + 0.026)AL ± 0 . 9 1 . Note that these equa-
264
recording electrodes over C3' or C4' just_as for the mixed tions utilize arm length and not height. Also, note the C6 spin-
median nerve studies. A similar N19 or N20 waveform mor- ous process is used rather than C2, C5, or C7.
phology to those for mixed median SEPs is obtained, but the la-
tencies are somewhat prolonged as noted above. 297
Ulnar Nerve
Alternate Recordings. The number of channels used can be Ulnar nerve segmental SEPs are comparable in technical
less than for mixed nerve studies. Because of the smaller ampli- demand with median nerve segmental studies. Segmental
tude for Erb's point and cervical spine locations, these sites may studies in general, and ulnar sensory SEPs in particular, have
be eliminated. Only 1 channel is then required to record seg- been touted for a number of diagnostic uses and are d i s -
mental SEPs. As there is still 1 channel in a 2-channel instru- cussed later. Only techniques are described in this section.
ment, a peripheral nerve recording can be obtained by locating a Once again, meticulous technique is crucial to obtaining
set of recording electrodes over the median nerve just proximal these responses.
to the medial epicondyle over the neurovascular bundle about Stimulation. One convenient way to activate ulnar sensory
the medial aspect of the arm. This enables one to calculate pe-
77
fibers properly is to securely place two ring electrodes on the
ripheral nerve conduction velocities. Certainly, if a 4-channel fifth digit, with the cathode proximal. A pulse width of 200
297
instrument is available, it is possible, although technically diffi- ps is applied at an intensity approximating 2 - 3 times the sen-
cult, to record not only the cortical potentials but also the pe- sory threshold at a rate of 5 Hz or less as previously described
ripheral (arm and Erb's point) as well as spinal potentials. for segmental median nerve s t u d i e s . - - Again, the gap in
77 ,07 m
Instrument Parameters. The same instrument settings used the noose portion of the electrode should be positioned posteri-
for mixed nerve studies can also be employed for segmental orly. Sufficient electrode paste is placed on the electrodes to
studies. One investigator who has extensively studied segmental provide adequate electrical contact yet not ooze toward its
SEPs prefers a bandwidth of 0.5-200 H z because of introduc-
107
neighboring electrode to avoid forming a conductive bridge
ing less high frequency noise yet not distorting the potential across the skin. The ring electrodes are ensured good contact
compared with the 10-3000 Hz and 0-2000 Hz bandwidths pre- by rotating them several times on the finger whereby the skin
viously noted for peripheral and cortical responses, respec- impedance is reduced.
tively. - Slightly higher sensitivities and more averages may
77 297
Ground. Similar concepts apply regarding ground electrode
be required, particularly in less than completely relaxed patients. placement as previously noted for the median nerve.
Reference Data. There is less reference data (Table 9-9) Recording Electrodes. The same electrode sites are utilized
available for segmental studies, as they have not been exten- as previously described for both mixed and sensory median
sively documented in either control subjects or patients with nerve studies. An attempt at recording potentials from Erb's
various lesions. Although the technique is somewhat more de- .point or the cervical region is not always performed because of
manding than for mixed nerve examinations, with a little prac- potential amplitude variability. 297
tice rewarding recordings can be obtained. Optimal reference Erb 's Point, Cervical Spine, Scalp. Identical electrode sites
values would account for various patient ages and heights, but and preparation are used for ulnar nerve sensory SEPs as those
available data often do not. Regression equations accounting for for median nerve sensory SEPs.
age and arm length (metacarpophalangeal joint to C-6 spinous Alternate Recordings. The number of recordings, as
process with arm pronated andabducted 90°) are available. For always, depends upon the number of channels available.
the first digit an equation is: N19 = 7.17 + (0.0219 ± 0.0109)A + Possibly the most fruitful additional recording site is that of a
(0.168 ± 0.036)AL, where A is age (years) and AL is the pa- peripheral nerve. - This response should be large enough to
52 77
tient's arm length (cm) using a standard deviation of+ 1.1 m s . 264
produce reliable amplitude and latency measuring points. The
The third digit has an equation of N19 = 7.86 + (0.0299 + peripheral potential can serve as a marker for peripheral nerve
Table 9-9. Median Nerve (Segmental) SEP Reference Data

Latency(ms)
To Peak UR To Peak UR To Peak MaxUR
First Digit (Thumb) Second Digit Third Digit
Recording Site
AF (A) - — (B)- — (C) 7.9 + 0.9 0.4-1.3
EP 13.2 ± 1.0 0.06 13.0 ± 1.2 — 13.1 ±0.8 0.3-1.4
C2 (NI3) 17.1 ± 1.3 0.08 16.6 ± 1.5 — 16.9 ± 1.3 0.2-1.8
C37C4' (NI9) 23.0 ± 1.4 0.07 22.2 ± 1 . 5 — 22.6 ± 1.3 0.6-2.5
C37C4' (P22) — — — 28.8 ±4.1 0.9-3.4
Interpeak Latency
AF-EP — — — — 5.4 ± 0.4 —
EP-NI3 3.9 ± 0.7 — 3.6 ± I.I — 3.6 ± 0.8 —
EP-NT9 9.8 ± 0.9 — — — 9.9 ± 1.2 —
EP-P22 — — — — — —
NT3-NT9 5.9 ± 0.8 — 5.4 ± 1.4 — 5.9 ± 0.7 —
Filter settings are as follows: A: 5-1500 Hz, B: 16-1600 Hz, and C: 10-3000 Hz peripheral and 0-2000 Hz central recordings. L/R signifies the left/right difference.
Refer to original references for details regarding instrumentation parameters. AF: antecubital fossa; UR: left/right difference. From Schimsheimer et al. (A), 264
Gaines (B),and DeLisa et al. (C).

129 77
Table 9-10. Ulnar Nerve (Segmental) SEP Reference Data

Latency(ms)
To Peak L/R To Peak To Peak L/R
Fifth Digit Fifth Digit Fifth Digit
Recording Site
AF (A) 8.0 + 0.6 0.38-1.46 (B)~ (C)- —
EP — — — ^ 12.8 ± 0 . 8 0.2 ± 0.2
C2(NI3) — — __ 16.4 ± 1.0 0.5 ± 0.5
C37C4' (Nl9) 24.0 ± 1.8 0.71-2.34 22.0 ± 1.4 21.4 ± I.I 0.5 ± 0.4
C37C4' (P22) 31.0 ± 3 . 4 0.95-3.67 28.7 ± 1.5 — —
Filter parameters are A&C 10-3000 Hz (peripheral recordings & Central for C) and 0-2000 Hz (central responses: A only) and B: 0.5-200 Hz. UR signifies the
left/right difference. Refer to original references for details regarding instrumentation parameters. AF: antecubital fossa; UR: left/right difference. From DeLisa et al. 77
(A) and Eisen (B) and Synek (C).

107 297
function particularly in light of the fact that the other two needle electrodes can also be utilized to produce sufficient
recording sites, Erb's point and cervical spine, are often not ob- stimuli at a lower current and may be tolerated better by some
tainable even in normals. patients. 147
Instrument Parameters. Similar instrument settings uti- Ground. Similar placement is used as previously recom-
lized for the median nerve sensory SEPs also apply for the ulnar mended for median and ulnar nerve studies.
nerve. The most important concept to remember is that what- Recording Electrodes. Erb's Point, Cervical Spine, Scalp.
ever reference data base is used, the same filter and amplifier These are the same sites used for the above-noted nerves. Again,
parameters must be used. the small nature of the responses may preclude recording of
Reference Data. If the median nerve cortical sensory SEP waveforms from these sites.
can be obtained, one can expect to detect the ulnar nerve SEP. A Alternate Recordings. If a peripheral nerve response is de-
few normal studies utilizing a limited number of patients, espe- sired, one can place recording electrodes halfway between the
cially of relatively younger age groups, have been published biceps tendon and the lateral epicondyle along the antecubital
(Table 9-10). - -" 77 107 1
crease. 77
Instrument Parameters. See ulnar and median nerve sen-

Superficial Sensory Radial N e r v e sory SEP recommendations.
SEPs obtained by stimulating the superficial radial nerve at Reference Data. As with all segmental studies, few investi-
the wrist yield relatively reliable and reproducible cortical SEP gations have documented normal values for the various SEP
responses. -' 77
Once proficiency is gained with digital nerve
07147
waveforms (Table 9-11). The interested reader is referred to the
excitation, little trouble should be experienced in performing original investigations for discussions regarding the utility of
superficial radial nerve stimulation. these procedures in various pathologic c a s e s . - 77107 329
Stimulation. The superficial radial nerve can be excited at

the wrist about 2 cm proximal to the radial styloid. - - 77 147 329
Lateral Antebrachial Cutaneous
Readers are encouraged to develop their own reference data (Musculocutaneous Sensory) N e r v e
base for this location. A 200-ps stimulating pulse delivered at 5 SEP evaluation of this nerve has been recommended to assist in
Hz or less at an intensity of 2-3 times sensory threshold results the diagnosis of peripheral nerve injuries to the musculocutaneous
in well-defined superficial radial nerve SEPs. A surface bar nerve or lesions involving the C5 nerve root. Obtaining repro-
111
electrode, anode distal, is preferred. However, subcutaneous ducible SEPs of this nerve may take a little practice and patience.
Table 9-11. Superficial Radial Nerve (Segmental) SEP Reference Data

Latency(ms)
To Peak UR To Peak To Peak UR
Recording Site
AF (A) 3.7 ± 0.5 0.23-0.86 (B)- (C) - —
EP 9.5 ± 0.6 0.24-0.97 — 9.5 ± 0.8 0.5
C2(NI3)
C37C4' (NT9)
13.3
18.8
±
±
I.I
1.2
0.47-1.86
0.49-1.84
_18.6 ± 0 . 4 5 13.5 ± I.I
18.8 ± 1.0
0.6
0.6
C37C4* (P22) 25.4 ± 4.3 1.03-3.49 23.7 ± 0.48 — —
Interpeak Latency
AF-EP 5.6 ± 0.7 — — — —
EP-NI3 — — — 3.9 ± 0.5 0.5
EP-NI9 9.1 ± 1.3 — — — 0.6
NI3-NI9 — — — 5.3 ± 0.5 —
Filter settings are: A: see ulnar nerve; B: 10-3000 Hz; and C: 20-2000 Hz for peripheral data and 2-2000 Hz for cortical potentials. UR signifies the left/right differ-
ence but in C it is the maximum UR difference. Refer to original references for details regarding instrumentation parameters.AF:Antecubital fossa; UR: left/right dif-
ference. From DeLisa JA et al. (A), Grisolia et al. (B), and Yiannikas et al. (C).
77 N7 329
Table 9-12. Antebrachial Cutaneous Nerve (Segmental) SEP Reference Data

Latency (ms)
To Peak L/R To Peak L/R To Peak
Recording Site
Elbow (A) - — (B)- — (C) 3.010.33
AS 3.1 ±0.4 0.22-0.82 — — —
EP 4.9 ± 0.6 0.28-1.02 — — 9.7 ± 0.8
C2(NT3) 7.8 ± I.I 0.37-1.91 — — 13.55 ± 0 . 9
C37C4' (NI9) 14.1 ±0.7 0.51-2.10 17.4 ± 1.2 0.5 + 0.4 18.8 ± I.I
C37C4' (P22) — — 25.4 ± 1.4 0.8 + 0.8 —
I Interpeak Latency
I AF-EP_ 1.6 ±0.4
EP-NT1 • 2.7 ± 0.3
EP-NT9_ 9.1 ± 1.3
NI3-NI9 6.5 ±0.7
Filter settings for A: 10-3000 Hz peripheral and 0-2000 Hz cortical responses, B: 0.5-200 Hz, and C: 10-3000 Hz. AS: Anterior shoulder; L/R signifies the left/right
difference. From DeLisa et al. (A), Eisen et al. (B),and Synek (C).
77 107 294
Stimulation. Placing the cathode 2 cm lateral to the biceps sensory evoked potential examinations involve either segmental
brachii tendon and approximately two fingerbreadths distal to or dermatomal excitation. Segmental stimulation is similar to
the antecubital crease optimally excites the lateral antebrachial that noted for upper limb studies where an accessible branch of
cutaneous n e r v e . The anode is positioned distal to the cath-
77107
a peripheral nerve is excited. Dermatomal studies are performed
ode. As previously noted, a bar electrode, surface disks, or sub- by activating the terminal branches of peripheral nerves as they
cutaneous needles may be used. An alternative method of innervate the patch of skin representing the signature area for a
stimulating this nerve is accomplished by locating the cathode particular dermatomal distribution. It is the authors' experience
5.5 cm proximal to the radiocarpal joint, between the tendons of that lower limb dermatomal and segmental studies are some-
the flexor carpi radialis and the radial artery. The anode is 294
what easier to perform than upper limb segmental SEPs because
placed 2.5 cm distal to the cathode. A 200-ps pulse duration is lower limb cortical waveforms are usually larger.
delivered at 2-3 times sensory threshold at a rate of 5 Hz or less.
One should avoid stimulating the underlying muscles (as visible Mixed N e r v e S o m a t o s e n s o r y Evoked Potentials
from a twitch) because this can generate evoked potentials due The tibial and common peroneal nerves are both relatively
to muscle spindle afferent activation that may arise from differ- large, subcutaneous, and readily excited by routine peripheral
ent segmental levels than those desired. nerve stimulation techniques at easily demarcated anatomic
Ground. The ground electrode can be placed on the mid- sites. The tibial nerve is found just medial to the medial malleo-
portion of the biceps brachii or on the sternum as previously lus and proximally in the mid-popliteal fossa, while the
described. common peroneal nerve lies just posterior to the fibular head.
Recording Electrodes. A slight variation in the recording Stimulating electrodes placed at these sites yield large and char-
electrodes is necessary to observe the peripheral nerve response acteristic cortical responses. Additionally, only one E-l scalp
(see below). The remainder of the peripheral and central elec- recording site is required for left and right stimulation because
trode locations are the same as those for median and ulnar nerve. this electrode is placed just posterior to the cranium's vertex
Erb's Point, Cervical Spine, Scalp. See previous electrode (CZ'), which overlies that region of the volume conductor be-
suggestions for the median sensory SEP response. tween the two cerebral hemispheres. The somatosensory cortex
Alternate Recordings. To record the peripheral nerve re- representing the lower limb is located on the medial aspect of
sponse of the lateral antebrachial cutaneous nerve successfully, the brain facing the contralateral side (Fig. 9-2). Occasionally
a recording electrode should be placed medially in the deltopec- one may wish to use paravertex recordings (C17C2') for those
toral groove at the level of the humerus' greater tubercle. An 77
cases when cortical responses are not optimally obtained or
E-2 electrode can be positioned over the ipsilateral acromion. when doubt exists as to the presence of pathology. The spinal
Instrument Parameters. See ulnar and median nerve sen- potentials are somewhat more difficult to obtain than the cervi-
sory SEP recommendations. cal potential in the upper limb.
Reference Data. Very few investigations have attempted to
document reference data for this nerve (Table 9-12). - * 77 107 294
Tibial Nerve
Practitioners are encouraged to develop their own reference Stimulation. The tibial nerve can be excited at either of two
values. easily located regions, i.e., the ankle or popliteal fossa. - Both
249 309
sites yield relatively large responses, and it is recommended that

LOWER LIMB SEPs the practitioner become familiar with both techniques. Although
neither location holds a distinct advantage over the other, the
Just as for upper limb SEPs, there are several methods for ob- practitioner may prefer one of the two techniques or individual
taining lower limb SEPs: mixed nerve techniques and two types patient circumstance may necessitate a specific stimulus site, e.g.,
of sensory SEP studies. The mixed nerve investigations usually a below the knee amputation requires popliteal fossa stimulation.
involve either the tibial or peroneal nerve. The tibial nerve is by Medial Malleolus. The tibial nerve can be easily located
far the most common lower limb nerve examined. Lower limb posterior to the medial malleolus by observing or palpating for
the arterial pulsations in this region. Either surface or needle spinous processes. The L4 spinous process is easily identified
electrodes can be used for stimulation. Should an individual by palpating the superior margins of the iliac crests with the fin-
have significant adipose tissue overlying the tibial nerve, needle gers and bisecting a line between them with the thumbs palpat-
electrodes may be required to deliver an adequate stimulus. In ing the intervening spinous process, i.e., L4. The next most
most cases, however, a surface electrode suffices. A bar elec- rostral spinous process is L3. An E-l electrode is placed over
trode (inter-electrode separation not critical) is preferred be- one of these spinous processes after the skin has been appropri-
cause significant pressure can be applied to the electrodes by ately abraded to reduce the impedance to less than 5000 Q. An
placing the tape over the intervening plastic bar and thereby ap- E-2 electrode is then secured about 3 cm proximal to the E-l
proach the nerve. Plastic tape with a slight amount of elasticity electrode, which approximates the level of the L2 spinous
is optimal, as it develops a counterpressure when firmly secured process. Alternatively, the E-2 electrode can also be situated on
and stays in place throughout the procedure. An alternate the iliac crest contralateral to the stimulated l i m b . - The 244 303
method of stimulating the tibial nerve is to place the stimulating lumbar potentials are relatively small compared with the corti-
cathode 8 cm proximal to a point located 1 cm posterior and 1 cal SEP waveforms. Preferentially increasing the gain on the
cm inferior to the navicular tubercle on the medial side of the third channel may aid in recording the lumbar potentials. It is
foot. This may be a bit too proximal in some patients to excite
77
crucial for the patient to be completely relaxed, especially with
the nerve adequately because of subcutaneous tissue. Similar respect to the paraspinal muscles, to detect the lumbar potential.
stimulation parameters as those described for the median nerve Some authors recommended giving the patient medication to
can be used. - A stimulus rate of between 2 and 3 Hz rather
52 282
maximize relaxation. - - If sedation is administered, it is
242 272 273
than 5 Hz is preferred to record lower limb SEPs optimally. The incumbent upon the physician to ensure that the patient is ade-
goal of the stimulating pulse is to induce a moderately vigorous quately warned regarding operating motor vehicles or perform-
twitch of the great toe or abductor hallucis muscle. Should the ing any other task requiring one to be fully alert. Although
patient complain of excess stimulus discomfort, the skin should investigators report the lumbar potential being present in all
be mildly abraded to reduce the impedance and, therefore, the normal individuals, this has not been the authors' personal expe-
amount of current required. If discomfort continues, the current rience, particularly in elderly subjects. - - - Thus, the ab-
175 201 239 244
simply may be too intense for the depth of the nerve and slight sence of a potential is not diagnostic of pathology.
repositioning of the stimulus site should be tried or the use of Thoracic Spine (T12, LI). Following placement of the
stimulating needle electrodes employed. lumbar electrodes, the next set of electrodes are located over the
Popliteal Fossa. The tibial nerve is excited at the popliteal thoracic spinous processes and connected to the second chan-
fossa by asking the patient to first flex the leg slightly. This nel. The T12 spinous process is found by locating the L4 spine
allows the tendons of the semitendinosus and semimembra- and placing the E-l recording electrode on the fourth spinous
nosus medially and the biceps femoris laterally to become process cephalad. This T12S electrode serves as the central con-
prominent. Half-way between these tendons and 2 cm proximal duction marker for the most caudal portion of the spinal cord.
to the popliteal crease is the recommended location for the cath- An E-2 electrode is usually positioned about 3 cm rostral to T12
ode. The anode is located several centimeters distal to this site.
77
or over the T10 spinous process. The same remarks regarding
Similar stimulation parameters as those noted for the ankle are skin preparation and ease of recording for the lumbar potential
used, but more current may be required. apply to the thoracic potential. An iliac crest E-2 electrode may
Ground. As with previously delineated SEP techniques, the help record particularly small spine potentials because of the
ground should be placed between the stimulating and first set of referential nature of the recording (see above). Some authors
recording electrodes, preferably closer to the recording elec- prefer the LI as opposed to T12 spinous process. In either case,
trodes. When recording from the lumbosacral region (see essentially the same potentials and latencies are recorded.
below), the ground can be conveniently located on the low back Scalp. An E-l electrode is located at CZ', while the E-2 elec-
region in the midline, which suffices for both left and right sided trode is placed at FpZ' for the first channel recording. This i
excitations. recording montage serves both left and right tibial nerve stimu- I
Recording Electrodes. A 4-channel technique is recom- lation, as do the spinous process electrode sites.
mended for recording lower limb SEP studies. In order to
67
Alternate Recordings. When the tibial nerve is stimulated
obtain the maximum amount of data from a lower limb SEP at the popliteal fossa, it is obviously impossible to record from
evaluation, one must use four recording sites. However, one this same location simultaneously. In this instance, the fourth
should not conclude that practitioners with 2-channel instru- channel records a response from the sciatic notch. An E-l 77
ments are precluded from doing acceptable lower limb SEP electrode is positioned over the sciatic notch (as described
studies (see below). below for the peroneal nerve), while the E-2 electrode may be
Popliteal Fossa (PF). When stimulating the tibial nerve located over the greater trochanter. Reducing the impedance is
about the ankle region, the first recommended recording site is crucial for successful recordings because of the amount of
the popliteal fossa. - - - The potential obtained from this po-
6 7 77 2 0 1 281
tissue overlying the sciatic notch. Like the thoracolumbar po-
sition is the propagating mixed nerve action potential initiated tentials, the sciatic notch potential may not be present in all in-
posterior to the medial malleolus. The popliteal fossa potential dividuals. The use of a needle recording electrode is not
serves as the marker for peripheral nerve function. An E-l recommended. When studying the tibial nerve, some investiga-
recording electrode is positioned exactly as that noted for the tors have utilized a C7-FpZ' recording montage to delineate the
stimulating cathode at the popliteal fossa as described above. 77
spinal cord volley in the cervical region in order to calculate the
The E-2 electrode is located on the medial joint line of the knee. spinal cord conduction t i m e . Of note, the cervical potential
269-271
This is the electrode montage usually recorded on channel 4. is extremely small in some individuals and bilateral tibial nerve
Lumbar Spinous Process (L3, LA). The third channel is used stimulation may be required to observe this potential.
to record the nerve volley traveling in the cauda equina as Instrument Parameters. As the distance for the neural
recorded by an electrode placed over the L3 (L3S) or L4 (L4S) volley is greater for lower compared with upper limb studies, an
Table 9-13. Mixed Tibial Nerve (Ankle) SEP Reference Data

Latency(ms)
To Peak L/R To Peak UR To Peak L/R
Recording Site
PF (A) 8.4 ± 0.9 0.18-0.88 (B) 8.2 ± 0 . 6 (C) 8.37 ± 1.04 0.02 ± 0.4
L3 17.9 ± 1.4 0.14-1.48 19.4 ±2.1 — 19.1 ± 1.9 0.06 ± 1.2
TI2 21.6 ± 1.6 0.12-1.29 22.1 ±2.2 0.4 ± 0.4 21.9 ± 2 . 3 0.2 ± 0 7
CZ' (P37) 38.5 ± 2.8 0.45-3.05 38.3 ± 3.3 I.I ± 0.9 37.3 ± 3.0 0.5 ± I.I
CZ' (N45) 48.1 ± 4.1 0.67-5.92 46.4 ± 3.2 1.6 ± I.I 45.3 ± 2.9 0.4 ± 1.3
Interpeak Latency
PF-L3 9.3 ± 0.8
PF-TI2 — — — — 13.6 ± 1.5 0.01 ± 0 . 7
L3-TI2 3.5 ± 0.2 — — — —
L3-P37 — — — — _15.2 ± —
TI2-P37 15.5 ± 1.7 — 16.0 ± 1.6 0.41 ± 1.2 1.3 0.29 ± I.I
Amplitude (pV)
PF 2.3 ± 0.6 0.12-1.92
L3 0.7 ± 0.2 0.16-0.65 — — _0.96 —
TI2 0.8 ± 0.3 0.14-0.72 I.I ± 0 . 6 0.2 ± 0.2 ±0.6 0.01 ± 0 . 3
P37 I.I ±0.3 0.13-0.97 2.4 ± 1.5 0.7 ± 1.0 1.9 ± I.I 0.01 ± 0 . 8
P37-N45 1.4 ±0.5 0.19-1.42 2.3 ± 1.3 0.5 ± 0.6 2.5 ± 1.4 0.3 ± 0.9
Filter setting for A: scalp: 0-2000 Hz and peripheral/spinal: 10-3000 Hz; B: 10-3000 Hz; and C: 30-1500 Hz. In the above table L3 is equivalent to L4 recordings while
TI2 is equivalent to LI recordings as some investigators prefer one or the other level. Magnitude of cortical potentials are baseline to peak for P37 and peak-to-
peak for P37-N45. PF: popliteal fossa; L/R signifies the left/right difference. From DeLisa et al. (A), Baran et al. (B),and Lastimosa et al. (C).
77 21 20l
analysis time of 100 ms (sweep 10 ms/div) is required. Because Figs. 9-12, 9-13, and 9-14). As previously noted, it is a good
cortical potentials are relatively large, a sensitivity of 2.0 pV/div practice to develop one's own reference data base.
usually suffices, although one should be prepared to increase
the sensitivity. Lumbar potentials may require a final signal am- • Peroneal Nerve
plification of 1.0 pV/div or possibly 0.5 pV/div. About 500 aver- SEPs of the peroneal nerve can be performed as easily as
ages result in distinct cortical potentials, but 1000 or more tibial SEPs, although cortical responses are often smaller. T h e
averages may be necessary to resolve the spine potentials, espe- authors know no advantage regarding neural information in per-
cially in less than completely relaxed patients. The usually rec- forming peroneal compared with tibial SEPs unless there is a
ommended filter settings of 10 Hz and 3000 Hz produce specific clinical indication to examine the peroneal nerve, e.g.,
adequate responses, although they can be a bit noisy. absent tibial nerve, lesion affecting the peroneal nerve, or a re-
Decreasing the high-frequency filter to 1000 Hz will reduce the search protocol. There is comparably less SEP data available for
noise without significantly affecting the responses' latency or the peroneal than tibial nerve. - ' - - -
20 65 255 304 307 329
amplitude. Stimulation. The peroneal nerve can be easily identified in

Reference Data. The tibial nerve has been extensively in- its location posterior to the fibular head about the knee. - A 77 154
vestigated with respect to both reference data and alterations in cathode is positioned just inferior to the fibular head, while the
pathologic states. -"- - - The large amplitudes and ease of de-
9 12 52 77
anode is placed several centimeters distal to the cathode. T h e
tection are major reasons this nerve has been explored in such same stimulus parameters as used for the tibial nerve are uti-
detail. The beginning practitioner is strongly urged to consider lized. One should note the clinical manifestation of peroneal
beginning one's SEP experience with the tibial nerve. The refer- nerve activation, i.e., primarily ankle dorsiflexion and eversion.
ence data presented are compiled from several sources and Ground. The ground electrode is again located near the first
demonstrate the consistency of the tibial response (Tables 9-13 set of electrodes at the sciatic notch, but one may wish to place
and 9-14). Regression equations considering height (cm) and the ground electrode on the back, which obviates the need of
age (years) have been formulated to assist in the development of repositioning this electrode.
reference parameters injarge populations. In men, the spinal 53
Recording Electrodes. The first E-l recording electrode is
potential LI latency (N22) is found to ha v e a n equation of N22 slightly different than that for the tibial nerve, as stimulation is
= 0.174(H) + 0.076A - 9.2525, while P37 (40) = 0.199H + now in the popliteal fossa. A slightly more proximal site is sug-
0.0852A + 3.8025, where H and A equate to height and age^je- gested for peroneal nerve studies. Although 4 channels are rec-
spectively. The same respective equationsjn women are: N22 = ommended, - - it is the authors' opinion that adequate studies
6 7 77
0.1619H + 0.0694A - 7.5235 and P37 (40) = 0.2222H + can be carried out using 2 channels. The most important data to
0.5995A + 1.1210. Equations j o r the central conduction times beôllected are the spinal arrival time (T12) and cortical arrival
for menjmdwomen are N22/P37 = 0.944H + 0.0233A - 0.2730 (P37/N45). If more data are required, the sciatic notch and
329
and N22/P37 = 0.0943H + 0.0425A - 0.2076, respectively. 53

lumbar potentials can be collected with a second set of stimuli,
Slight differences are believed to occur secondary to gender by inconvenient though this may be.
the investigators reporting the above regression equations. Not Sciatic Notch. Because the popliteal fossa is not available for
all authors agree with these findings (see above). Graphic plots recording, the next anatomic site of possible interest is the sci-
can also be used from which the equations can be derived (see atic notch, which is recorded on channel 4 . The E-1 electrode 77
Table 9-14. Mixed Tibial Nerve (Popliteal Fossa) SEP Reference Data
Latency(ms)
To Peak L/R To Peak To Peak
Recording Site
SN (A) 5.4 ± 0.3 0.21-0.82 (B)- (C)-
L4 10.9 ±0.8 0.19-1.51 8.75 ± 1.06 11.4 ± 0 . 9
TI2 12.8 ± I.I 0.15-1.32 10.33 ± 1.7 14.4 ± 1.3
CZ' (P37) 31.2 ±2.6 0.52-3.24 — —
CZ' (N45) 39.4 ± 4.6 0.74-6.14 — —
Interpeak Latency
SN-L4 5.3 ± 0.6 — — —
L3-TI2 1.7 ± 0 . 3 — — —
TI2-P37 18.4 ± 2 . 1 * — — —
SN I.I ±0.5 0.14-1.28
L4 0.8 ± 0.3 0.19-0.84
TI2 0.9 ± 0.3 0.14-0.88
P37 1.6 ±0.4 0.12-1.42
P37-N45 1.9 ±0.7 0.18—1.51
In the above table L3 is-equivalent to L4 recordings whileTH is equivalent to LI recordings as some investigators prefer one or the other level. In B the spinal po-
tejntials_are measured to L5 andTI2 with filter settings of 2-3000 Hz. In C the bandwidth is 10-3000 Hz.Amplitudes for P37 are from baseline to peak of P37 while
P37-N45 amplitude is a peak-to-peak magnitude. SN: sciatic notch; UR signifies the left/right difference. From DeLisa et al. (A), Delbeke J, et al. (B), and
77 76
Dimitrijevic (C).
89
site is located by palpating the greater trochanter and ischial morphology, latency, and amplitude as well as their anatomic
tuberosity. An imaginary line is drawn between these two areas significance. An E-2 electrode is located 3 cm proximal to the
and bisected. This region should place the E-l electrode in the E-l electrode or placed on the iliac crest contralateral to the side
vicinity of the sciatic nerve as it is about to enter the pelvis. The of excitation.
greater trochanter serves as the E-2 electrode site for the sciatic Thoracic Spine (T12). A third set of electrodes records the
notch. action potential volley traversing the spinal cord's caudal seg-
Lumbar Spinous Process (L3, LA). On the third of the 4 ment from the region of T12. The E-2 electrode again is placed
channels, the L3S or L4S spinal potential is recorded. Some in- 3 cm proximal to the spinous process of the twelfth thoracic
vestigators prefer L3, while others record from the L4 spinous vertebra (about T10) or on the iliac crest. When using the iliac
process. The potentials are essentially the same in terms of crest, one should be prepared for a relatively more noisy trace,
Table 9-15. Peroneal Nerve SEP Reference Data

Latency(ms)
To Peak L/R To Peak L/R
Recording Site
SN (A) 5.7 ± 0.4 0.29-O.9I (B) — —
L3 10.4 ± 0.6 0.08-0.71 10.8 ± 0.9 0.5
TI2 1 I.I ± 0.9 0.07-0.82 — —
CZ' (P37) 32.5 ± 2.4 0.59-5.87 27.3 ± 1.5 2.2
CZ' (N45) 41.3 ± 3.9 1.92-11.72 — —
Interpeak Latency
SN-L3 4.4 ± 0.7 — — —
L3-TI2 1.5 ± 0 . 3 — — —
L3-P37 — — 16.5 ± 0 . 9 5 2.3
TI2-P37 15.4 ± 0.8 — — —
PF 0.8 ± 0.3 0.18-0.83 — —
L3 0.7 ± 0.3 0.36-1.29 — —
TI2 0.8 ± 0.4 0.41-1.32 — —
P37 — — 0.6 ± 0.47
_
P37-N45 1.7 ± 0 . 7 0.17-1.72 — —
For the above table L3 is equivalent to L4 recordings whileTI2 is equivalent to LI recordings as some investigators prefer one or the other level. Bandwidths for the
above values are:A) see_tibial nerve; and B) 2-2000 Hz. E-2 electrode for spinal potentials are 3 cm proximal to E-l electrode. Magnitude of cortical potentials are
baseline to peak for P37 and peak-to-peak for P37—N45. UR signifies the left/right difference except in B the numbers signify a maximum value; SN: sciatic notch.
From DeLisa et al. (A).andYiannikis et al. (B).
77 329
as there is less common mode signal between these 2 electrodes particularly as it travels posterior to the lateral malleolus. S-l
compared with the situation when both electrodes are close to nerve root fibers constitute the primary but not exclusive root
each other on the thoracic spines. level contributing to the sural nerve. - ' - 107 m 154 184
Scalp. The same locations used for tibial nerve SEPs elec- Stimulation. The sural nerve can be rather easily excited by
trodes, CZ' and FpZ', are utilized for peroneal nerve cortical placing a cathode in the depression between the lateral malleo-
recordings. lus and the Achilles tendon. The nerve lies in this anatomic
Alternate Recordings. As noted above, if only the cortical region and generates relatively large cortical SEPs. An anode
and T12 spine potentials can be recorded, sufficient data have should be placed several centimeters distal to the cathode. The
been obtained for most routine purposes. An inability to obtain a fact that the sural nerve is purely sensory requires that stimula-
sciatic notch potential or L4 potential does not necessarily imply tion be effected in a particular manner quite distinct from that
that an inadequate study has been performed or that the patient is used for the mixed tibial and peroneal nerves, i.e., determining
abnormal. The peripheral nervous system can certainly be exam- the sensory threshold and exceeding it by 2.5-3.0 times.
ined with peripheral nerve conduction techniques. Ground. The ground electrode can be located just distal to
Instrument Parameters. See tibial nerve recommendations. the popliteal fossa or on the sternum. Good skin preparation
Reference Data. As previously noted, there is not an abun- should be performed at the chosen site.
dance of normal values for peroneal nerve studies (Table 9¬ Recording Electrodes. The recording of segmental SEPs is
15) 77.329 Although this is not a technically difficult study to somewhat different than mixed nerve studies with respect to
perform, it is a good idea to practice this study on normal sub- the number of channels and recording sites. The extremely
jects prior to investigating potential abnormalities of the per- small amplitude of segmental spinal potentials precludes their
oneal nerve. routine detection. As a result, only 1 channel is routinely uti-
lized by the authors to observe the cortical potential. O c -
LOWER LIMB SEGMENTAL SOMATOSENSORY casionally, a popliteal fossa recording is attempted to assess
EVOKED POTENTIALS peripheral nerve function. It is also possible to record spinal
potentials in patients capable of maintaining adequate
As previously defined, segmental SEP studies examine the paraspinal muscle relaxation. - 239 240
peripheral activation of a pure sensory nerve such as the superfi- Popliteal Fossa (PF). The E-l and E-2 electrode sites rec-
cial radial nerve. In the lower limb, a number of segmental SEPs ommended for the sural nerve are the same as those for the
i can be easily performed by the beginning practitioner. These tibial nerve. Although good skin preparation is required for
77
studies have primarily been utilized to evaluate lumbosacral any SEP, segmental studies are particularly demanding, as the
r a d i c u l o p a t h i e s . - The sensitivity and specificity of seg-
I07m 240
pure sensory nature of responses can at times result in a rela-
mental studies with respect to lumbosacral radiculopathies tively small potential.
remain controversial. This controversy is not presently explored Scalp. Sural segmental SEPs utilize the same cortical recording
but will be discussed in the chapter considering radiculopathies. positions used for tibial nerve SEPs, i.e., CZ' and FpZ'. Again, the
At this time, only the technical aspects of eliciting segmental same recording electrodes are employed for both left and right
studies and reference data are presented. limbs. The morphology of segmental cortical responses is similar
to mixed nerve SEPs, i.e., there is usually an initial large positive
Sural Nerve deflection followed by a large negative deflection. If the peripheral
The sural nerve is a relatively large sensory nerve comprised nerve potential is also recorded, channel 2 is used to document this
of fibers originating from branches of both the tibial and peroneal response; otherwise, channel 1 is used. Thus, a 2-channel instru-
nerves. - Anatomically, the sural nerve is rather superficial,
154 286
ment is quite adequate for segmental SEP studies.
Table 9-16. Sural Nerve SEP Reference Data

Latency(ms)
To Peak L/R To Peak L/R To Peak L/R
Recording Site
PF (A) 8.7 ± 0.5 0.17-0.84 (B)-
TI2 24.3 ± 1.4
L3 20.2 ± 1.6 0.6 ± 0.3
C Z ' (P37) 44.3 ± 3.7 0.59-5.71 38.7 ± 2.9 0.7 ± 0.4 42.0 ± 2.4 1.04 ± 0 . 7 2
CZ' (N45) 51.7 ± 5 . 9 1.42-10.48
Interpeak Latency
TI2-P37 — — 24.3 ± 1.4 —
L3-P37 18.5 ± 2 . 0 0.9 ± 0.6
Amplitude (pV)
PF 1.8 ± 0 . 3 0.12-1.92
L3 0.16-0.65 — — —
P37-N45 I.I ± 0 . 5 0.19-1.42
In the above table L3 is equivalent to L4 recordings whileTI2 is equivalent to LI as some investigators prefer one or the other level. Filter setting for the above are:
A) 0-2000 Hz peripheral responses and 10-3000 Hz cortical potentials, B) 10-3000 Hz, and C) 3-3000 Hz. E-2 electrode for spinal potentials are 3 cm proximal to
E-l electrode. Magnitude of cortical potentials are peak-to-peak for P37-N45. L/R signifies the left/right difference. PF: popliteal fossa. From DeLisa et al. (A), and
77
Chiappa (B) and Perlik (C).

52 240
Alternate Recordings. The small amplitude of the segmen- Scalp. The montage previously noted for the sural nerve, CZ'
tal response usually precludes spinal recordings. and FpZ', is also used for this nerve. The morphology of the cor-
Instrument Parameters. A sweep of 10 ms/div and sensi- tical response is similar to that noted for other lower limb SEP
tivity of 1-2 pV/div with filter settings of 10-3000 Hz are typi- responses.
cal for sural nerve segmental studies, which are the same as Alternate Recordings. Because of the small nature of the
those used for the tibial nerve. Averaging approximately 500 re- spinal potentials, attempts to record from the lumbar or thoracic
sponses typically yields good waveforms. spine are quite challenging and preclude routine observation of
Reference Data. Because of the ease of performing the sural these responses. A 2-channel instrument, therefore, is quite suf-
SEP, a number of investigators have provided data for this nerve ficient for the superficial sensory peroneal nerve.
(Table 9-16). - -
52 77 240
Instrument Parameters. The same instrument parameters
noted for the sural nerve suffice. As always, two trials of each
Superficial Sensory Peroneal N e r v e response should be performed to ensure reproducibility.
The superficial sensory peroneal nerve is the cutaneous branch of Reference Data. Compared with the sural nerve, fewer stud-
the superficial peroneal nerve. - Approximately 12 cm proxi-
554 1 6 7 286
ies have been performed on the superficial sensory peroneal
mal to the ankle, the superficial sensory peroneal nerve pierces the nerve (Table 9 - 1 7 ) . - Although the technique is relatively
77107 240
superficial fascia to become a cutaneous nerve. This nerve pro- 169

easy, the clinical utility of SEPs from this nerve is questionable
ceeds distally to form the medial and intermediate dorsal cutaneous and hence the lack of various reference data bases.
nerves, which pass anterior to the extensor retinaculum to innervate
the foot's dorsal aspect. The cutaneous fibers of this nerve are be- Saphenous N e r v e
lieved to consist primarily of fibers from the L5 nerve root. It is, The saphenous nerve is the cutaneous termination of the
therefore, possible to perform an SEP from the superficial sensory femoral nerve. This nerve enters the femoral triangle along with
peroneal nerve and investigate the sensory portion of the L5 root. the femoral artery and passes through Hunter's canal. Distal 154
Stimulation. The superficial sensory peroneal nerve can be to this region, the saphenous nerve provides a cutaneous branch
anatomically located by drawing an imaginary line connecting to the medial portion of the knee known as the infrapatellar
the medial and lateral malleoli. A cathode is located at the junc- branch of the saphenous nerve. - The main saphenous trunk
154 254
ture between the middle and lateral thirds of this imaginary continues distally into the leg in the groove formed by the
line. 77168
The anode is located several centimeters distally. medial aspect of the tibia and the bulk of the medial gastrocne-
Electrodes imbedded in a plastic bar are convenient to use as the mius muscle. Branches of the nerve terminate on the medial
cathode and anode. The same parameters described above for aspect of the foot by passing anterior to the medial malleolus.
the sural nerve are utilized for the superficial sensory peroneal The cutaneous distribution of the descending saphenous nerve
nerve. It is also possible to stimulate the superficial peroneal is the medial aspect of the leg, ankle, and foot, sometimes
nerve 12 cm proximal to the lateral malleolus. -" - 107 1 240
reaching the great toe. The saphenous nerve can be excited
Ground. The ground can be on the patella or more proximally. along its course at the infrapatellar region, medial aspect of the
Recording Electrodes. The same principles described for leg, or the ankle.
the sural nerve also apply to the superficial sensory peroneal Stimulation. One may wish to excite different portions of
nerve. Only 1 channel is used to perform this study, although a the saphenous nerve along its course depending upon the possi-
logical argument can be made for a 2-channel recording. ble lesion site. - The most distal aspect of the saphenous
97 296
Popliteal Fossa (PF). Channel 1 may record potentials at the nerve can be stimulated at the ankle as it accompanies the
popliteal fossa, as the neural impulses traverse this region. To saphenous v e i n . A bar electrode is convenient, but separate
107111
detect this potential properly, it is necessary to use a rather high surface electrodes are equally effective. The patient is requested
sensitivity (1.0 pV/div) because of the response's small ampli- to gently dorsiflex the ankle on the side of stimulation, and the
tude. Alternatively, the peripheral nerve can be examined with tendon of the tibialis anterior is palpated. If the patient is inca-
routine nerve conduction techniques. pable of dorsiflexing the foot, this same site can be located
Table 9-17. Superficial Sensory Peroneal Nerve SEP Reference Data

Latency(ms)
To Peak UR To Peak To Peak L/R
Recording Site
PF (A) 11.6 ± 0 . 8 0.16-0.73 (B)- (C)-
TI2 — 23.4 ± 1.35
C Z (P37) 45.2 ± 2.4 0.39-2.74 39.9 ± 1.8 41.1 ± 2.0 1.02 ± 0 . 6 9
CZ* (N45) 58.5 ± 4.2 0.52-5.84
Interpeak Latency
PF-P37 31.4 ± 1.7 — — —
TI2-P37 17.1 ± 1.7
Amplitude (pV)
PF 0.43 ± 0 . 1 9 0.07 ±0.41 — — —
P37-N45 I.I ± 0 . 3 5 0.15 ± 1.19
Magnitude of cortical potentials are peak-to-peak for P37-N45. UR signifies the left/right difference. Bandwidths for above values are: A) 10-3000 Hz for peripheral
response and 0-2000 Hz for central potentials, B) 0.5-200 Hz, and 3-3000 Hz. In A the nerve is excited at the ankle while in B and C the site of stimulation is 12 cm
proximal to the lateral malleolus. From DeLisa et al. (A), Eisen (B),and Perlik (C).
77 107 240
approximately one fingerbreadth anterior to the medial malleo- Knee. When stimulating the saphenous nerve at either the
lus. The cathode is located at this site, i.e., just medial to the tib- ankle or mid-leg level, the investigator can attempt to record
ialis anterior tendon in the soft tissue hollow formed between from the peripheral portion of the saphenous nerve as it crosses
the tibialis anterior tendon and the medial malleolus. An anode the knee joint. In this instance, an E-l electrode is positioned
is located a few centimeters distal to the cathode. medially between the tendons of the sartorius and gracilis mus-
A second site of stimulation is in the groove formed by the cles 1 cm superior to the inferior border of the patella. A con- 77
medial aspect of the tibia and the medial gastrocnemius. 77,3,1

venient location for the E-2 electrode is the anterior aspect of
The cathode is located approximately midway between the the patella. This recording montage is connected to channel 2.
medial malleolus and medial tibial plateau. The anode is again This potential may be difficult to detect despite a successful cor-
placed several centimeters distal to this site. Another author rec- tical potential.
ommends that the cathode be positioned 15 cm distal to a point Scalp. The scalp location for the recording electrodes is the
between the sartorius and gracilis tendons where these muscles same as those used for the sural nerve, i.e., CZ' and FpZ'.
are 1 cm proximal to the inferior border of the patella. 77
Cortical potential morphology is similar to the sural nerve in
The infrapatellar branch of the saphenous nerve can be stud- that an initial positive deflection followed by a negative deflec-
ied by locating a cathode in the soft tissue depression formed by tion is typical.
the inferior border of the medial aspect of the patella and the su- Alternate Recordings. The small amplitude of peripheral
peromedial aspect of the medial tibial plateau. The anode is
97 2 9 6
and spinal potentials obviates their routine detection.
again positioned several centimeters distal to this site. Essentially, the potential typically recorded for the saphenous
The same stimulus parameters for the sural nerve are also uti- nerve is the cortical potential.
lized for all portions of the saphenous nerve excited. A sensory Instrument Parameters. Saphenous nerve SEP studies re-
threshold is first determined, and then a current intensity ap- quire similar instrument parameters as those utilized for the
proximating 2 - 3 times this sensory threshold is delivered. A sural nerve. In some patients, 500 or more responses may need
small muscle twitch of the medial gastrocnemius may occasion- to be averaged in order to obtain clearly recognizable responses.
ally be observed when attempting to activate the saphenous The filter setting used by individual authors may vary somewhat
nerve at mid-leg level. This is most likely direct muscle stimula- and should be consulted when reproducing a technique for par-
tion not involving a significant number of IA afferents and, ticular patients.
| therefore, should not alter the results. Of course, should this Reference Data. There are only a limited number of normal
j muscle activation be painful, the electrode should be reposi- values available for the saphenous nerve at each of the three
tioned either more proximal/distal or medial/lateral until the stimulation sites (Table 9-18). - * - Again, it is highly rec-
77 107 240 296
discomfort is minimized. ommended that the individual practitioner develop laboratory-

Ground. The ground electrode can be located on the ipsilat- specific normal values.
eral patella or preferably closer to the recording electrodes. As
peripheral recording sites are not very successful in picking up a Lateral Femoral C u t a n e o u s Nerve
I response, the ground can often be placed on the sternum or some A direct continuation of the lumbar plexus, the lateral
other convenient location without distorting the cortical SEP. femoral cutaneous nerve consists of sensory fibers from L2 and
Recording Electrodes. There is little opportunity to per- L3 nerve roots. - The nerve courses along the rim of the
154 254
form successful peripheral recordings when stimulating the pelvis to emerge through a tunnel formed by the anterior supe-
saphenous nerve because of the small nature of this response. rior iliac spine and the lateral portion of the inguinal ligament.
When exciting the saphenous nerve at the knee region, the pe- The nerve then descends subcutaneously along the thigh to
ripheral and spinal volley is simply too small to record over the divide, approximately 12 cm distal to the anterior superior iliac
thoracolumbar region. spine, into two branches. One branch provides cutaneous sensation
Table 9-18. Saphenous Nerve SEP Reference Data

Latency (ms)
To Peak L/R To Peak To Peak
MID-LEG STIMULATION ANKLE STIMULATION KNEE STIMULATION
Recording Site
Knee (A) 2.6 ± 0.3 0.15-0.79 (B)- (C)- (D)—
TI2 — — — 25.3 ± 1.4
C Z ' (P37) 37.1 ± 2.8 0.32-2.81 43.4 ± 2.2 43.4 ± 2 . 6 37.6 ± 2.0
C Z ' (N45) 47.3 ± 4.6 0.47-5.62 — — —
Interpeak Latency
Knee-P37 32.5 ± 1.9 — — — —
TI2-P37 — — — 18.3 ± 2.1 —
Amplitude (pV)
Knee 0.64 ± 0 . 1 8 0.06-0.39 — — —
P37 0.43 ± 0 . 1 2 0.15-0.36 — — —
P37-N45 0.94 ± 0.23 0.21-0.76 — —
Magnitude of cortical potentials are baseline to peak for P37 and peak-to-peak for P37-N45. Filter setting for the above techniques are: A) 10-3000 Hz for periph-
eral and 0-2000 Hz for cortical responses, B & D) 0.5-200 Hz, and C) 20-2000 Hz. L/R signifies the left/right difference. From DeLisa et al. (A), Eisen (B & D), and
77 107
Perlik (C).
240
to the anterior aspect of the thigh, while the second is distrib- nerve recording at the ASIS is desired, increasing the sensitivity
uted over the lateral portion of the thigh as far distal as the patel- to 0.5-1.0 pV/div may be desirable, provided the noise level is
lar region. sufficiently controlled.
Stimulation. The lateral femoral cutaneous nerve can be ex- Reference Data. Only a few authors have investigated the
cited by locating a cathode 12 cm distal to the anterior superior lateral femoral cutaneous nerve and provided normal values
iliac spine and placing the anode several centimeters distal to (Table 9-19). - As with the saphenous nerve, it is especially
77 107
the c a t h o d e .
77107
Sural nerve stimulation parameters are again important for practitioners to develop their own reference values
used to activate the lateral femoral cutaneous nerve. to ensure proper recording techniques because of the technical
Ground. In the event of using a peripheral recording loca- difficulty in recording this SEP.
tion (see below), the ground should be placed just distal to this
site between the stimulating and recording electrodes. If only a DERMATOMAL SOMATOSENSORY
cortical recording montage is used, the ground can be situated EVOKED POTENTIALS
on the sternum or other convenient location.
Recording Electrodes. In the authors' opinion, the lateral SEPs obtained as a result of dermatomal compared with seg-
femoral cutaneous nerve SEP waveform is somewhat more diffi- mental stimulation are morphologically similar, but the two stim-
cult to detect than that of all of the previously noted nerves. The ulation techniques are different. When attempting to record a
102
reason for this is the small nature of the response at both the pe- dermatomal SEP, the stimulus is applied to a region of skin that
ripheral and cortical recording locations. Complete relaxation of represents the autonomous zone of a particular nerve root. The
the patient is particularly important, as any muscle artifact will number of nerve fibers, cutaneous afferents, directly under the
impede detection of the waveform. Meticulous technique is re- cathode is considerably less than when a pure sensory nerve
quired at the stimulating and recording sites with respect to ac- branch believed to be composed of fibers primarily from one
curate placement of electrodes and skin preparation. nerve root is activated. The clinical utility of both segmental and
Anterior Superior Iliac Spine (ASIS). Should one wish to dermatomal studies is unclear, and the relative value of either
pursue a peripheral recording during the SEP examination, an compared with the other is also unknown. In this section, no at-
E-1 recording electrode can be located over the emergence of tempt is made to critically analyze the diagnostic importance of
the lateral femoral cutaneous nerve from the inguinal tunnel. 77
either technique; only the methodology of obtaining reproducible
This electrode is positioned about 1 cm medial to the ASIS, results is delineated. For an appraisal of the current thinking re-
while the E-2 electrode is placed over the ipsilateral greater garding the diagnostic applicability of dermatomal studies, con-
trochanter. Adequate skin preparation is mandatory to observe sult the section of this text discussing radiculopathies.
this potential. Do not be surprised if this response is not ob- Although a number of studies have been published expound-
tained, even on the asymptomatic side. ing the merits of dermatomal SEPs with respect to correctly di-
Scalp. CZ' and FpZ' are again used. Particular attention agnosing pathology, minimal work has been done on
10J8 4 1 8 5 2 5 2
must be directed at getting patients to relax cervical and mas- attempting to acquire reference data and develop criteria to es-
seter muscles to minimize muscle artifact contaminating this tablish normality. The techniques to record dermatomal SEPs
small waveform. The first channel can be utilized for the corti- are relatively straightforward. While there are some data on
cal potentials. normal values of various dermatomal SEPs, multiple reference
Alternate Recordings. Very small lateral femoral cutaneous data bases, particularly regarding the reproducibility and normal
nerve responses preclude spinal recordings. A scalp montage of variance of the dermatomal SEPs, are noticeably absent.
C3/C4-OZ is preferred by some. An initial negative deflection
295
Considerable latency (up to 8-9 ms) and amplitude (80%) side-
is the first cortical response in this instance. to-side differences can occur in normal p e r s o n s . 101102
Instrument Parameters. The same instrumentation settings

used for the sural nerve are replicated for this nerve. The only L5 D e r m a t o m a l SEP
change required may be an increase in the number averaged to The L5 dermatome's signature area is believed to be located
resolve the response, i.e., 1000. Additionally, if a peripheral about the medial aspect of the first metatarsophalangeal joint in
Table 9-19. Lateral Femoral Cutaneous Nerve SEP Reference Data

Latency (ms)
To Peak L/R To Peak UR To Peak
Recording Site
ASIS (A) 2.9 ± 0.5 0.31-0.75 (B)- —
CZ" (P37) 30.2 ± 2.4 0.41-2.89 29.8 ± 1.6 1.5 31.8 ± 1.8
CZ' (N45) 42.5 ± 4.5 0.61-6.72 — — —
Interpeak Latency
ASIS-P37 25.2 ± 2.8 — — — —
Amplitude (pV)
ASIS 0.42 ±0.21 0.08-0.41 — —
_— —
P37 0.32 ± 0 . 1 2 0.09-0.28
P37-N45 0.41 ± 0 . 1 9 0.15-0.41 —
Bandwidths for above techniques are: A) 10-3000 Hz peripheral and 0-2000 Hz cortical responses, B) not provided, C) 0.5-200 Hz. Magnitude of cortical potentials
are baseline to peak for P37 and peak-to-peak for P37-N45. UR signifies the left/right difference. ASIS: anterior superior iliac spine. In B, a cortical montage of
C3/C4-OZ is used. As a result, an initial negative potential is detected at approximately N29. From DeLisa et al. (A), Synek (B) and Eisen (C).
77 295 107
the f o o t .
125154
A more exclusive L5 dermatomal region is be- at the fifth metatarsophalangeal joint. - This portion of the
125 154
lieved to be on the dorsum of the foot between the first and foot is supplied by the sural nerve, which is thought to consist
second digit by some investigators. In any event, the region
10
exclusively of fibers originating from the SI nerve root. Aside
most likely to represent L5 is on the dorsum of the foot sur- from developing a dermatomal SEP technique, there is little dis-
rounding the first metatarsophalangeal joint. cussion regarding the clinical utility of S1 dermatomal stimula-
Stimulation. The stimulus parameters are similar to those tion compared with sural nerve segmental stimulation.
for segmental studies in that a pulse duration of 200 ps is deliv- Stimulation. Identical stimulus parameters are used for the
ered at a rate less than 5 Hz at between 2 and 3 times the sen- SI dermatomal SEP as were utilized for the L5 dermatomal
sory threshold. In one study, the mean sensory threshold in SEP. A cathode is placed at the level of the fifth metatarsopha-
normal individuals was noted to be 4.1 ± 1.1 mA for the L5 der- langeal joint on the lateral aspect of the foot. An anode is lo-
matome with a final stimulus delivery of 9.2 ± 2.6 m A . The 184185
cated about 3 cm distal to this location. A convenient set of
cathode should be located at either the level of the first metatar- stimulating electrodes is the bar electrode. Mildly abrading the
sophalangeal joint along its medial a s p e c t - or in the web
184185 245
lateral margin of the foot beneath the electrodes helps to reduce
space between the first and second digit in the foot. An anode 10
skin impedance secondary to callus formation. A stimulus of
is located about 2 or 3 cm distal to the placement of the cathode. 2-3 times sensory threshold (4.2 ±1.1 mA) is used, and this has
Mild abrasion of the skin beneath the stimulating electrodes been found to be about 9.2 ± 2.6 m A . 184
may be necessary in some individuals, as the foot can be heav- Ground. The same location and parameters as those for the
ily calloused. L5 dermatome are used.
Ground. The ground electrode can be located on the ster- Recording Electrodes. As with the L5 dermatomal re-
num, cranium, or some other convenient location such as the ip- sponses, the small number of fibers excited precludes reliable
silateral patella. It is a good idea to mildly abrade the skin under peripheral or spinal recordings. The absence of peripheral
the ground site to ensure good contact to reduce interference recording sites, however, is somewhat of a limitation in that an
and artifact because the dermatomal responses can be rather abnormal cortical response may result from abnormalities any-
small, particularly in various pathologic conditions. where along the nervous system from just proximal to the site
Recording Electrodes. Stimulating a dermatome involves a of stimulation to the somatosensory cortex itself. This is one ob-
small number of cutaneous afferents. This fact most likely previous limitation of the dermatomal response.
cludes recording reproducible peripheral or spinal responses. As Scalp. Because the somatosensory cortical representation of
a result, all investigations to date have utilized only cortical the entire foot is in the same general area, there is no difference
recording electrodes. As with other SEPs, the scalp electrodes in location for the SI scalp recording electrode compared with
can be either surface or subdermal needle electrodes. the L5 electrode. As a result, the E-l electrode is placed at CZ'
Scalp. The scalp recording electrodes are positioned in the while the E-2 electrode is at FpZ'. This location should result in
same location as that used for all lower limb SEP studies. An E- clearly identifiable waveforms that are similar in appearance to
1 electrode is placed at CZ', while an E-2 electrode is located at those obtained for tibial nerve stimulation.
FpZ'. This cortical montage should result in well-defined L5 Alternate Recordings. The same as those for the L5 der-
dermatomal responses. Because only the cortical potentials are matomal SEP.
recorded, a single channel is all that is required. Instrument Parameters. Same as the L5 dermatomal SEP.
Alternate Recordings. There are no additional responses nor- The details in the original studies need to be reviewed to fully
mally recorded for dermatomal responses. However, should more appreciate the nuances in successfully obtaining responses that
channels be available, various cortical montages can be used to conform to the waveforms as initially reported.
help identify the vertex responses, e.g., C l ' or C2' referenced to Reference Data. With respect to reference data, the same
FZ or C Z referenced to the contralateral C3VC4' locations. 10
comments apply to S1 dermatomal responses as those already
Instrument Parameters. An analysis time of 100 ms noted above for L5 waveforms (Table 9 - 2 1 ) . The regres-10184 2 4 4
(sweep of 10 ms/div) combined with 500 to 1000 averages sion equation for the_S 1 dermatome relating age (years) and
should be sufficient to resolve most L5 dermatomal responses. height (meters) is: P(40) = 8.6 + 24.0 (Height) + 0.038 (Age) ±
Amplifier sensitivity of 1 or 2 pV/div is adequate. Two studies 2.9 m s . The same general rules of prolonged peak latency for
184
have presented reference data each using different filter settings.

One of the investigations examined the responses at various
bandwidths and concluded that there was no statistically signifi- T a b l e 9-20. L5 D e r m a t o m a l S E P R e f e r e n c e D a t a
cant difference in waveform parameters between filter setting of
Latency(ms)
5-250 Hz and 1-1500 H z . - 184
A second study utilized
185
30-3,000 Hz. As always, until more data are available, it is a

10
To Peak L/R To Peak UR
good idea to reproduce exactly the conditions under which par- Recording Site
ticular data are obtained if these data are to be utilized. CZ' (P37) (A)48.4 ± 3.9 -0.6 ± 1.7 (B)51.0 -0.55 ± 2.1
Reference Data. Very few studies have systematically and CZ (N45)
f
58.8 ± 4 . 3 — — —
clearly attempted to formulate a complete data base (Table 9¬
A m p l i t u d e (pY)
20). io.184,245 Dermatomal latencies are directly correlated to age
Max Ratio
(years) and height (meters) just asjire other SEPs. A regression
CZ' (P37) _ 0.09-2.36 4:1 0.6 ± 0.4 —
equation for the L5 dermatomal P37(or P40) latency responses
CZ' (P37-N45) 0.28-4.17 — —
is given as: P40 = 8.3 + 22.4 (Height) + 0.086 (Age) ± 2.7 m s . 184 =
L5 was stimulated at the medial aspect of the first digit for A while the first web
SI D e r m a t o m a l SEP spacewas utilized for B. Magnitude of cortical potentials are baseline-to-peak
for P37 and peak-to-peak for P37-N45. Filter setting for techniques above are:
An exclusive region of dermal cutaneous innervation by the A & C) 5-250 Hz, and B) 30-3000 Hz. L/R signifies the left/right difference.
S1 nerve root is thought to occur at the lateral margin of the foot From Katifi et al. (A).Aminoff et al. (B).
184 10
Table 9-21. SI Dermatomal SEP Reference Data to between 2 and 3 times the sensory threshold delivered at less
Latency(ms) than 5 Hz is o p t i m a l . An alternate technique is to place a
148149
bar electrode on either side of the penis consecutively to com-

To Peak UR To Peak UR
pare left and right responses.
Recording Site In the female, the same stimulation parameters are employed
CZ' (P37) (A) 49.9 ± 3 . 9 -0.15 ± 2.3 (B) 52.3 -1.17 ± 1.7 but obviously the electrode locations must differ. Typically a bar
C Z ' (N45) 60.2 ± 4 . 4 — — — electrode is placed on the labia majora for left-right comparison
Amplitude (pV) studies. In the authors' experience, the pudendal SEP tech-
151
Max Ratio nique is somewhat more demanding in the female than in the
CZ' (P37) _ 0.26-2.84 3:1 0.5 ± 0.4 — male patient. In normal females, it is often difficult to record
C Z (P37-N45) 0.30-5.85 — — — either the left or right response. The main reason for this diffi-
culty may be cutaneous moisture that short-circuits the current
Filter settings for the above techniques are: A) 5-250 Hz and B) 30-3000 Hz.
Magnitude of cortical potentials are baseline-to-peak for P37 and peak-to-peak across the skin's surface as opposed to activating the cutaneous
for P37-N45. UR signifies the left/right difference. From Katifi et a l . (A), ,M skin afferents, i.e., in the female, it is a dermatomal response
Aminoff etal.'° (B). but segmental stimulation in the male. Proper drying of the labia
is critical to obtaining pudendal SEPs. In both the male and
an increase in age and height apply to the S1 dermatome SEP as female patient, stimulation is only mildly uncomfortable and
they do for all other SEPs. patients often require reassurance because of the stimulator's
location. A chaperone is appropriate when performing this pro-
cedure in patients of the opposite sex to the practitioner.
MISCELLANEOUS SEP TECHNIQUES Ground. A mid-sternal or abdominal ground electrode can
be utilized, and either location works equally well.
Although an SEP can be performed for just about any acces- Recording Electrodes. The cortical SEPs are easily ob-
sible nerve, only a few nerves have been examined with respect tained and are of sufficient amplitude to be readily recognized.
to reference data. The clinical utility of these procedures is dis- The investigator - who popularized this technique even per-
148 149
cussed in detail when appropriate in other sections of this text forms spinal recordings, although it is the authors' experience
relating to specific lesions involving the nervous system. The that spinal potentials can often be difficult to document.
technique, instrumentation parameters, and reference values are Scalp. The pudendal SEP is best recorded with an E-1 elec-
presently described so as to allow the successful recording of trode located at CZ' referenced to FpZ'. As for all other scalp
these waveforms. SEP recordings, subdermal needle recording electrodes are pre-
ferred, but surface electrodes provide equally successful re-
PUDENDAL NERVE SEP sponses with proper skin preparation.
Spine (LJ). An LI electrode can occasionally record a re-
The pudendal nerve is a continuation of the sacral plexus and sponse in some m e n , but this response is rarely observed in
148
is formed by nerve roots s 2 - 4 . The cutaneous afferent

1 3 6 1 5 4 2 8 6
women. The reason for this difference is most likely that seg-
151
component of the pudendal nerve of interest to SEP investiga- mental responses are obtained in men and hence more nerve
tions is the dorsal nerve of the penis in males and dorsal nerve fibers are activated compared with dermatomal (less nerve
of the clitoris in females. Pudendal SEPs are of interest because fiber) stimulation in women. When spinal potentials are at-
they investigate the afferent component of the lower sacral tempted, they may be recorded on channel 2.
nerve roots, which are difficult to examine by other electrodiag- Alternate Recordings. No other recording/stimulating
nostic techniques. Although the anal sphincter (motor compo- montages have been reported.
nent S2-4) muscle can be explored with a needle electrode, 314
Instrument Parameters. Because of the segmental or der-
continuous tonus of this muscle often renders the study of activ- matomal nature of these responses, similar instrument settings
ity at rest less than optimal. The clinical utility of pudendal as those used for dermatomal studies are recommended, specifi-
SEPs is in the area of sexual dysfunction and compromise of the cally an analysis time of 100 ms (sweep of 10 ms/div) with a
lower sacral nerve r o o t s . I 3 6 J 4 8 1 4 9 1 5 0 1 5 1
sensitivity of 1 or 2 pV/div and filter settings of 0-2000 H z or
76
Stimulation. In males, the cutaneous portion of the puden- 5-250 H z . Approximately 500 averages should result in rec-
148
dal nerve can be easily activated by placing ring electrodes ognizable and reproducible cortical waveforms. Spinal poten-
around the shaft of the penis with the cathode p r o x i m a l . 148149
tials may require a sensitivity of 0.5 pV/div. This high
There should be several centimeters of separation between the sensitivity necessitates complete relaxation on the part of the
cathode and anode. A stimulus pulse duration of 200 ps raised patient to minimize muscle artifact.
Reference Data. Very little reference data have been pub-
lished (Table 9-22) regarding pudendal SEPs even though this
Table 9-22. Pudendal Nerve SEP Reference Data technique is frequently used with respect to urologic investiga-
Latency (ms) tion.
148-151
As with all other studies, it is recommended that prac-
To Peak To Peak titioners develop their own data base from which to judge
abnormality.
Recording Site
_ MEN WOMEN
TRIGEMINAL NERVE SEP
CZ' (P37) 42.3 ± 1.9 39.8 ± 1.3
C Z (N45) 52.6 ± 2.6 49.1 ± 2 . 3 In performing trigeminal SEPs, the most common nerve
LJ 9.9 ± 1.37 fibers activated are those supplying the cutaneous aspects of the
Filter setting are 5Hz to 250 Hz. From Haldeman et al. 148151
face, specifically the sensory divisions of the trigeminal nerve.
Stimulation of various aspects of the face most likely activates

tactile sensory receptors that convey information to the princi-
pal sensory nucleus and rostral portion of the spinal nucleus of
the spinal tract after passing through but not synapsing in the
trigeminal (semilunar, gasserian) g a n g l i o n . From these re-
21134
gions, second-order neurons form the dorsal trigeminothalamic

tract that project both ipsilaterally and contralateral^ to the
ventral posteromedial (VPM) thalamic nucleus. Third-order
neurons subsequently project to the postcentral gyrus. The
trigeminal cutaneous afferents conveying pain and possibly
other afferent information have their cell bodies located in the
trigeminal ganglion. These fibers then project to the nucleus of
the spinal tract, which in turn provides second-order neurons
contralaterally to form the ventral trigeminothalamic tract ter-
minating in the VPM. Third-order neurons of these fibers also
project to the postcentral gyrus. That aspect of the parietal lobe
representing these afferents is rather inferior and lies just supe-
rior to the temporal l o b e . Compared with the cortical repre-
134
sentation of the hand for upper limb SEPs, the facial cortical
area is more inferior on the homunculus and therefore cortical
recording electrodes are placed more inferior (toward the ears
as described below). Trigeminal SEPs are rather difficult to
obtain primarily because of overlap between the stimulus arti-
fact and desired response. There are a number of techniques de-
scribed to record trigeminal SEPs, but only the most reliable is
presented.
Stimulation. The stimulation technique presented should
result in successful trigeminal SEPs when used on most pa-
tients. The patient is placed supine and instructed to relax with
the lips slightly parted. - - A hand-held stimulator is gently
108 287 288
placed into the patient's mouth such that the cathode is in con-
tact with the angle of the mouth, i.e., where the lips join. The
anode is rested on the lower lip. The patient is then asked to
close the mouth gently such that the upper and lower lips con- Figure 9-16. T r i g e m i n a l SEP.Trigeminal SEP response to stimula-
tact the cathode and anode projections. Care must be taken not tion of the upper and lower lips. Note the initial negative deflection of
to touch the teeth, as this may result in noticeable patient dis- the waveform. (From Stohr M, Petruch F, Scheglmann K: Somatosensory
comfort. It is important to note that the practitioner must hold evoked potentials following trigeminal nerve stimulation in trigeminal
the stimulator very steady, because slight movements can result neuralgia.Ann Neurol 1981 ;9:63-66, with permission.)
in contacting the teeth.
A stimulus pulse of 200 ps at 2-3 times sensory threshold de-
livered about 2-3 Hz is suggested. - Some patients may not
108 287
electrode is placed at C5' for left trigeminal nerve activation and
be able to tolerate this intensity, in which case the current at C6' for right trigeminal nerve stimulation. C5' and C6' are 2
should be adjusted to tolerance level. The lips should be dried cm posterior to the line bisecting the ears and 10% of the total
with gauze because if excess saliva is present, the current will coronal distance superior to the tragus region. - - An FpZ'
46 47 170232
preferentially travel along the surface of the lips as opposed to E-2 electrode is used for both left and right cortical recordings.
activating the subcutaneous nerve fibers. If insufficient current Alternate Recordings. None.
is delivered, it may not be possible to detect an SEP. A slight Instrument Parameters. A total analysis time of 50 ms
twitch of the lip can occur but should not affect the SEP wave- (sweep of 5 ms/div) with a sensitivity of 1-2 pV/div and a band-
form. The stimulus artifact is rather noticeable in trigeminal width of 2-2000 Hz is suggested. - About 500 averages
287 288
SEPs because of the short distance and hence time interval be- should result in clearly definable responses.
tween stimulation and the first cortical waveform. Reference Data. There are a number of techniques in which
Ground. The ground electrode may be placed on the inion latencies for cortical trigeminal waveforms can be recorded
or the side of the face ipsilateral to the stimulus above the from the angle of the mouth, - mental nerve, and infraor-
108 287 279
zygoma. The skin should be prepared with mild abrasion, as bital nerve. The above described technique, angle of the
202
stimulus artifact is rather significant in this SEP technique. If mouth, is most often quoted in the literature. The original refer-
the inion is utilized for the ground location, subdermal needle ence data are provided (Table 9- 23). - If this nerve is to be
108 287
electrodes are convenient because they are easy to apply. frequently examined, it is good practice to develop one's own
Recording Electrodes. Only one channel is required for reference values.
trigeminal SEPs. Cortical electrodes are the only ones required. Cautionary Note. There is debate as to the efficacy of using
The morphology of the trigeminal SEP is triphasic with an ini- any form of stimulation other than direct neural activation in
tial negative deflection (Fig. 9-16). one of the bony foramina on the face through which a branch of
Scalp. The location for scalp electrodes necessary to record the trigeminal nerve exits. The above-noted surface techniques
trigeminal SEPs has not been described previously. The E-l have been criticized with respect to primarily generating poorly
Table 9-23. Trigeminal Nerve SEP Reference Data muscle through which the nerves pass forms the so-called tarsal
Latency(ms) tunnel. Techniques were developed to evaluate the above-noted
peripheral branches of the tibial nerve to possibly assist in the
To Peak UR To Peak UR
diagnosis of plantar neuropathies.
Recording Site Stimulation. A constant-current stimulator utilizing a pulse
C57C6' (A) 12.5 ± 0.87 — (B) 12.8 ± 0.9 0.6 ± 0.5 duration of 200 ps delivered at 2 - 3 Hz with an intensity to pro-
(NI3) duce a moderately vigorous toe twitch of either the first digit
C57C6' 18.5 ± 1.51 0.55 ± 0.55 19.3 ± 1.4 0.6 ± 0.4 (medial plantar nerve) or fifth digit (lateral plantar nerve) is
(PI 9) used (Fig. 9-17)." A convenient manner of exciting these nerves
C57C6' 26.9 ± 2.2 — 28.6 ± 1.7 1.2 ± 0.8 is with a bar electrode. To activate the medial plantar nerve, a
(N30) line connecting the interdigital region of the first and second toe
Amplitude (pV) with the heel is bisected." The midpoint of this line is where the
Nil/ 2.6 ± 1 . 1 0.51 ± 0 . 5 4 — — cathode is located, with the anode placed several centimeters
PT9 distal. Activation of the lateral plantar nerve is achieved by
forming a similar line between the fourth and fifth digits, and
Filter settings are 2-2000 Hz (A) and 0.5-200 Hz (B). From Stohr et al. (A) 287
and Eisen et al. (B).

108
the lateral portion of the heel. Again, the half-way point demar-
cates the location where the cathode excites the lateral plantar
nerve. The calcaneal nerve is stimulated by placing a bar elec-
defined and possibly far-field potentials not representative of trode on the heel a few centimeters plantar to the posterior
trigeminal neural pathway conduction with respect to specific margin of the h e e l . " The current intensity used for calcaneal
neural structures. stimulation is 3 times the sensory threshold at this region of the
foot. Abrading the skin is recommended for all stimulating sites
MEDIAL/LATERAL PLANTAR AND CALCANEAL NERVES because the plantar surface of most feet is calloused.
Additionally, the plantar aspect of the foot contains thick con-
The medial/lateral plantar and calcaneal nerves are the three nective tissue, which may impede the passage of current from
terminal branches of the tibial nerve. After entering the ankle the stimulator. Although the medial and lateral plantar re-
beneath the laciniate ligament (flexor retinaculum), the tibial sponses should be obtained in all normal persons, the calcaneal
nerve divides into its three terminal branches. The cal-
154 2 8 6
nerve response may be absent because of too much impedance
caneal nerve is a pure sensory nerve that may branch from the on the heel.
tibial nerve just proximal to, or under, the flexor retinaculum to Ground. A ground electrode is located in a convenient loca-
provide cutaneous sensation to the heel of the foot. The tibial 78
tion such as the sternum or ipsilateral patella.
nerve then terminates in two mixed nerves, the medial and lat- Recording Electrodes. The cortical recording electrodes
eral plantar nerves, which supply all of the foot intrinsic mus- utilize the same placement as that for the tibial nerve. The
cles except for the extensor digitorum brevis and possibly the medial and lateral plantar, as well as calcaneal SEP waveforms,
first dorsal interosseous (deep peroneal nerve). Each nerve are similar in morphology to all other lower limb potentials.
enters the foot through its own opening in the abductor hallucis Scalp. An E-l recording electrode is placed at CZ', while a
muscle. The medial plantar nerve provides cutaneous sensa-
141
E-2 electrode is located at FpZ'. The same montage is used for
tion to the medial aspect of the foot's plantar surface including recording SEPs from all three nerves. Although surface elec-
the first three and one-half toes. The lateral plantar nerve sup- trodes were used in the original study," subdermal recording
plies sensation to the remainder of the foot. The laciniate liga- needle electrodes may also be used without a need to alter the
ment in combination with that region of the abductor hallucis recording/stimulating parameters.
Figure 9-17. Medial/lateral plantar and calcaneal

SEPs. Location of electrodes for stimulation of the medial and
lateral plantar and calcaneal nerves.The midpoint of a line be-
tween t h e interdigital region and posterior aspect of t h e heel
(note arrows) is used t o stimulate t h e plantar nerve (left
figure).Tibial nerve excitation is performed 12 cm proximal t o
the medial plantar nerve stimulation point (arrows in right
figure). (From Dumitru D, Kalantri A, Dierschke B:
Somatosensory evoked potentials of t h e medial and lateral
plantar and calcaneal nerves. Muscle Nerve 1991;14:665-671,
with permission.)
Table 9-24. Medial/Plantar/Calcaneal SEP Reference Data

To Peak UR To Peak UR To Peak UR
Medial Plantar Lateral Plantar Calcaneal
Recording Site
CZ' (P37) (ms) 42.3 ± 3.0 0.04-0.9 43.5 ± 3 . 0 0.01 ± I. I 46.9 ± 3.2 - 0 . 4 ± 1.2
Amplitude (pV)
CZ" (P37/N45) 3.3 ± 1.9 0.4 ± 0.8 2.9 ± 1.7 0.1 ± 0.9 1.4 ± 0.9 -0.01 ± 0.6
Max L/R % 41.2 48.6 50.0
Magnitude of cortical potentials are peak-to-peak for P37-N45. L/R signifies the left/right difference while Max UR % refers to the maximum percentage difference
expressed as one side compared to the other. Filter settings are 10-3000 Hz. From Dumitru et al."
Alternate Recordings. Although the original study did not de- If two electrodes are located relatively close to each other
scribe spinal potential recordings, it is anticipated that responses and placed in the near-field, one creates a near-field bipolar
should be obtainable. Of course, one would have to establish refer- recording montage. Should the E-2 electrode be displaced to
ence data for this location prior to utilizing it clinically. some region of the far-field, a referential near-field montage
Instrument Parameters. An analysis time of 100 ms exists. *
189 190
This is one way to conceptualize the differ-
192-194
(sweep of 10 ms/div) and filter settings of 10-3000 Hz are the ence between bipolar and referential recording montages. If
parameters initially used." A sensitivity of 1-2 pV/div should both electrodes are relocated to a neural generator's far-field,
suffice for both medial and lateral plantar nerve responses; how- interesting results ensue depending upon the two electrodes'
ever, 1 pV/div may be necessary for the calcaneal response, as it location with respect to each other. Two electrodes relatively
is comparatively smaller. A total of 500 averages are adequate close to each other in the far-field region most likely will not
to resolve most responses, but more may be required for partic- detect a waveform because the difference in potential b e -
ularly small calcaneal responses. tween these two locations is negligible. Another way of
Reference Data. The medial and lateral plantar nerves as saying this is to note that if the same potential exists at both
well as the calcaneal response result in typical cortical lower electrodes and differential amplification records only what is
| limb SEP potentials." Latencies to the initial positive and sub- different between the electrodes, the absence of a difference
sequent negative peaks are measured, while the amplitude noted results in no net recorded potential. By displacing the t w o
is from the first positive to following negative peaks (Table 9¬ electrodes rather distant from each other, but remaining in the
2 4 ) . " As only one study has been published to date regarding far-field, a small "far-field" potential may be recorded, partic-
these nerves, it is a good idea for the practitioner to replicate ularly if multiple averages are summated to improve the
these data prior to diagnosing pathology. signal-to-noise ratio. Two electrodes in the far-field but dis-
tant from each other will detect the small potential difference
between them; however, this difference is small and hence the
necessity of averaging multiple trials. It is possible to perform
SEP APPLICATIONS
far-field SEP recordings.
SEPs have been utilized to investigate many disorders. - - 39191 203
The usual technique of clinically recording cortical SEPs is
j^g j j
212,225,278.302.313.317 w a y of possible diagnostic applica-
c e a r r to place both an E-l and E-two electrode on the scalp and
tions can be simplified by considering two major categories in record the potential difference between these two scalp loca-
which SEPs may be useful: central nervous system disorders tions, e.g., C37C4' and FpZ' for upper limb studies. - - The 6 7 282
and peripheral nervous system lesions. Potential uses of SEPs in result is an easily definable N19/P22 potential. By relocating
both diagnostic and prognostic situations regarding central dis- the E-2 electrode from FpZ' to the_wrist contralateral to the one
orders are explored in detail when both central nervous system being stimulated, the simple N19/P22 waveform becomes a
and anterior horn cell disorders are considered (see chapters re- rather complex sequential_array of positive and negative poten-
lated to specific clinical disorders). The diverse applicability of tials preceding the N19/P22 complex, which also is somewhat
SEPs with respect to peripheral nervous system pathology is altered.
considered in the remainder of this book when particular dis- Specifically, following median nerve wrist stimulation with
ease entities affect various portions of the peripheral nerves. the above noted scalp-contralateral wrist montage, 4 monopha-
sic positive potentials__are noted (Fig. 9-1S). - - - ' - -
61 62 63 80 81 82 83
SOMATOSENSORY FAR-FIELD POTENTIALS Additionally, the N19/P22 complex increases in_amplitude and
duration. The larger and longer duration N19/P22 waveform is
The electrical field generated by neural tissue may be consid- easily understood if one considers the principles of differential
ered to consist of a near-field and f a r - f i e l d . The near-
165171-173
amplification (see Chapter 3). - Two electrodes located rela-
96 98
field region is that portion of the current distribution in which tively close to a neural generator will, to some extent, share a
the associated isopotential lines change in magnitude with rela- certain amount of the data, i.e., these two electrodes will detect
tively small alterations in position within the field. At some some of the same electrical potential. That portion of the elec-
point, as one moves away from the neural generator's source tric field that is common to both electrodes is cancelled through
and sink currents, the isopotential field lines change little with differential amplification. The result is amplification of the re-
different positions. This region of minimal potential change maining electrical activity. When one of the electrodes is re-
may be referred to as the far-field aspect of that neural genera- moved to a further location, there is now little electrical activity
tor. A more complete discussion of this topic may be found in in common and the entire potential as observed by the near-by
Chapter 2. electrode is amplified, producing a relatively larger potential
that is also longer in duration. This is the most likely explana- for the production of the four early far-field potentials preced-
tion forthejarger amplitude and longer duration cortical poten- ing the cortical response to upper limb median nerve excitation.
tial N19/P22). The observation of 4 positive potentials is The first far-field potential demonstrates a rather consistent
somewhat more difficult to explain. peak at approximately 9 ms and is referred to as the P9 potential
Based upon the above discussion, one can conclude that when (Fig. 9-18). Electrodes located at the axilla and Erb's point,
both electrodes are located on the scalp, the four positive poten- median nerve near-field recordings, demonstrated that the P9
tials are not observed because they are common to both record- far-field potential occurred in time after the corresponding axil-
ing electrodes and hence eliminated as a common mode signal, lary near-field potential, but before the Erb's point near-field
i.e., they have the same potential at each of the recording elec- potential. - The P9 neural generator is, therefore, postulated to
63 79
trodes. In other words, the two scalp electrodes are in the far- occur in the proximal axilla. Because there are no known
field of four generators that produce the same potential in the synapses or nuclei in this area, the median nerve afferent volley
cranium. By relocating the E-2 electrode to the wrist (any loca- itself is the most likely generator. Close inspection of the P9 po-
tion will do, as long as it is not on the skull), there is now enough tential revealed that it consisted ofjwo separate peaks in over
of a potential difference between the two electrodes to result in a 70% of individuals, referred to as P9a and P9b. According to 328
recorded waveform even though both electrodes are in the far- far-field theory, an alteration in direction of the median nerve
field (see above discussion). By placing the E-2 electrode on fibers as they turned to enter the thorax from the arm results in a
some body part other than the skull, a so-called referential mon- dipolar moment imbalance of current associated with the action
tage is c r e a t e d . The question then becomes, what are the
192194
potential. This imbalance in dipole moments is de-
8 8 1 7 2 1 7 3 1 8 2
neural generators producing the four positive far-field potentials? tected by the differential amplifier as a difference in potential
This rather straightforward question unfortunately does not between the two referentially located electrodes producing an
have a simple answer. The full understanding of the mechanism observable waveform. The change in median nerve direction
underlying far-field potential production continues to elude in- yields two far-field potentials, one within the axillary_region
vestigators, but a few working models do exist (see Chapter 2). (P9a) and a second potential just distal to Erb's point (P9b).
A multitude of neural generators have been proposed to account A second far-field potential is observed at about 11 ms fol-
lowing median nerve stimulation utilizing a referential montage
and is known as PI 1 (Fig. 9-18). A number of neural generators
have been postulated to produce this potential, including synap-
tic and post-synaptic activity in the nucleus cuneatus and medial
lemniscus. - - Considering the conduction velocity of the
1314 63 79182
traveling neural volley, it has been suggested that this far-field

potential arose where the nerve root_entered the spinal
cord. - - The true generator of the PI 1 potential remains
189 289 322
controversial.
The third positive far-field potential following median nerve.
wrist stimulation has a latency of 13 ms; hence the designation
PI 3 (Fig. 9-18). Multiple neural generators have been postu-
lated such as the synaptic and post-synaptic activity in the thala-
mus, brain stem, medial lemniscus, cerebellum, and ventral
posterior thalamus. - - - - - Calculating the distance trav-
17 63 79 166 275 316
eled by a neural impulse based upon averagedconduction ve-

locities suggests that the site of origin for PI 3 is within the
cervical_spinal cord. Agreement does not exist as to the most
189
likely PI 3 neural generator.

An inconsistently observed fourth monophasic positive far-
field potential, PI 4, has also had a number of possible genera-
tors (Fig. 9-18). Various sites of origin for this potential include
medial lemniscus, cerebellar connections, thalamocortical radi-
ations, and the cerebeHum. - ^ - - As with the two previous 13 1 9 165 316
far-field potentials, PI 1 and PI 3, the region of nervous tissue re-

sponsible for PI4 is unknown.
Tibial nerve stimulation also revealed 4 positive far-field po-
tentials when recorded with a referential scalp montage. The po-
tentials have been designated PI 7, P24, P27, and P 3 l (Fig.
9-18). - - - - - Postulated but by no mean accepted sites of
79 85 87 209 210 323
origin for these four respective potentials are: just distal to the
Figure 9-18. Far-field p o t e n t i a l : M e d i a n n e r v e S E P s . (A) Far- sacral plexus, impulse entry into the conus medullaris, rostral
field SEP recording following median n e r v e wrist stimulation. Four spinal cord, and brain stem. Additional work is required re- 323
positive far-field potentials are recorded: P9, PI I, PI3, and PI4. (B-E) garding both upper and lower limb far-field potential production
Stimulation of the tibial nerve with a cortical referential montage t o before its nature and site of origin can be fully understood.
the contralateral knee (K) demonstrating far-field potentials and their
development from t h e gluteal fold region along the spinal column t o Clinical Utility of SEP Far-Field Potentials
the scalp: Pl7, P24, P27, and PIT. (Modified from Desmedt et al., and 79
A number of investigations have attempted to clinically eval-
Yamada et al. ) 323
uate patients' lesions based upon SEP far-field findings. - - 13 65 71
The most current thinking on the origin of far-field potentials is coma. - - - Specifically, bilateral absence of cortical re-
33 124 161 332
that an action potential undergoes a dipolar imbalance when an sponses (BACR) to median nerve stimulation is usually associ-
alteration in the surrounding volume conductor's conductivity ated with a very poor prognosis. Studies have shown that, in
occurs (see Chapter 2). - - A dipolar imbalance may also be
88 172 173
patients with nontraumatic encephalopathy, essentially com-
produced when an action potential crosses the boundary be- plete cortical necrosis is required to obliterate the cortical re-
tween two body segments of different size or a nerve impulse sponse. Accordingly, the complete absence of SEP responses
256
changes direction. Finally, if an action potential encounters the indicates a poor prognosis for recovery from coma.
termination of active tissue such as the end of a nerve, a far-field
potential can be seen. Because far-field potentials do not have Nedian Nerve SEP Methodology
specific anatomic neural generators but depend upon joint posi- The literature documents the use of median nerve SEPs for
tion or other v a r i a b l e s , their latency and amplitude may
84192193
predicting prognosis in coma. Although other nerves may pro-
vary in less than predictable ways. As a result, it is recom- vide similar information, the median nerve-generated cortical
mended that far-field potentials not be used for clinical diagnos- SEP response is usually the most robust of all mixed-nerve
tic purposes until such time that more is understood about SEPs.
far-field generation. - - -
114 192 193 328
Because of this reasonable recommendation, the utilization Stimulation

of additional channels for referential montage recordings must Since patients are comatose, a motor twitch threshold is uti-
be questioned. 110115
It is certainly acceptable to apply referential lized for setting stimulus intensity because sensory thresholds
recording techniques to experimental situations in the hope of are not useful. Typically, 1.5 times the motor threshold is used.
obtaining more information about far-field potential generators. As previously noted, stimulus rate influences the amplitude of
Utilizing noncephalic referential SEP montages effectively in- the cortical response, where faster rates reduce the amplitude.
creases the noise of the trace, as common mode rejection is re- For example, in normal subjects, the cortical amplitude with 6
duced and additional potentials are included from poorly Hz stimulation is about 8 5 % of that with 3 Hz stimulation.
understood far-field generating sources. It is recommended that Thus, to compare with published studies, a consistent stimulus
the beginner initially refrain from performing noncephalic ref- rate is required. For short-latency studies, most authors use a 2
erential montages and use primarily bipolar (scalp-scalp: FpZ' Hz stimulation rate, while for long-latency studies, approxi-
as the E-2 for C37C4' and CZ') electrodes placements. Spinal mately a 1 Hz stimulation rate is used.
referential montages (e.g., T12-iliac spine), however, may be of
some assistance because the rather small bipolarly recorded Recording
spinal potentials (T12-T10) are increased in magnitude and When studying patients with coma to ascertain prognosis, it
hence easier to detect with a referential montage. is extremely important to avoid a false-positive prediction of
nonawakening, i.e., saying someone will not awaken when they
might. False-negative predictions (i.e., saying someone might
PROGNOSTICATION OF CENTRAL awaken when they do not) are more acceptable, since patients
NERVOUS SYSTEM INJURY are not erroneously removed from life support in this instance.
In order to rely on cortical potentials, a number of recording
COMA EVALUATION sites must be used to ensure that the ascending volley actually
reaches the brain. In the trauma setting, multiple injuries to the
Clinicians caring for comatose patients are continually seek- neuraxis frequently coexist. Failure to record the ascending
ing better ways to predict outcome. Methods used for this pur- volley could mean that an absent cortical response may be due
pose should be sensitive for predicting nonawakening. More to an undetected spinal cord lesion, and hence, result in an erro-
importantly, however, such methods need to be exquisitely spe- neous false-positive prediction for nonawakening. Thus, one
cific, i.e., they should not erroneously predict nonawakening typically records from the following sites:
when the patient would actually awaken if supportive treatment 1. Peripheral nerve (axilla or Erb's point)
were provided. 2. Cervical spine (C7 spine-Fz)
A number of methods have been studied for their ability to 3. Brain stem (Fz-mastoid)
predict the prognosis of coma recovery. Clinical assessments 4. Contralateral cortex (C3' or C4'-Fz)
like the Glasgow Coma Scale and the evaluation of brain stem Responses from the first three recording sites above indicate
reflexes offer clear predictive value, as patients with low scores an intact neuraxis up to the brain stem. Unrecognized spinal
generally do worse than those with higher scores. However, cord injury and atlanto-occiptal dislocation in patients with
some patients with very low scores ultimately awaken, making traumatic brain injury referred for coma evaluation have been
the information not reliable enough to use for decisions regard- detected by using these methods.
ing withdrawal of life support. Recently, creatine kinase BB Sufficient amplification, averaging, and analysis time are all
band levels in the cerebrospinal fluid (CSF-CK) have been necessary to avoid missing an otherwise present response.
shown to indicate a uniformly poor prognosis when sufficiently Typically, at least 2 5 0 - 5 0 0 responses for each trial should be
elevated in patients with anoxic encephalopathy. - This tech-
183 301
averaged, depending upon the level of background noise. A dis-
nique, however, is not particularly sensitive for detecting non- play sensitivity of 0.2-1.0 pV/div is used, and analysis times
awakening and is not applicable in patients with coma due to vary from 50 to 200 ms, depending upon whether short- or long-
trauma, since even focal trauma will release CSF-CK. latency responses are evaluated.
A number of electrophysiologic methods have been studied
For their ability to predict outcomes in comatose patients. Of Assessment
these methods, median SEPs have some of the strongest evi- Although there are a variety of parameters that can be measured
dence to suggest their utility in predicting outcome after from cortical responses, including absolute latencies, interpeak
Figure 9 - / 9 . S E P e x a m p l e s in p a t i e n t s . A shows a normal left median nerve somatosensory evoked potential (SEP) recorded from a co-
matose patient w h o did not awaken.The first tracing is (active-reference) the C4'-Fz; the second tracing is the mastoid-Fz; the third tracing is the
C7 spine-Fz; and t h e fourth tracing is t h e axilla-shoulder. B shows a left median nerve SEP recorded from a c o m a t o s e patient w h o did not
awaken.The cortical response is absent. N o t e the inverted mastoid response in the t o p tracing: this should not be confused with a cortical re-
sponse, since the latency is t o o early.The first tracing is (active-reference) the C4'-Fz; the second tracing is the mastoid-Fz; the third tracing is the
C7 spine-Fz; the fourth tracing is the axilla-shoulder. C shows a left median nerve SEP recorded from a comatose patient who did awaken.The
cortical response is present and robust, and the N3 (N70) is present. N o t e the slower sweep speed (20 ms/div) compared with the earlier figures
at 5 ms/div.The first tracing is (active-reference) the C4'-Fz.
latencies, and amplitudes, only a few have been shown to have to be sure that the stimulus actually reaches the brain. As indi-
clinical relevance. The strongest indicator of a poor prognosis is cated above, recordings should be made from the peripheral
a bilateral absence of the short-latency cortical response with nerve, cervical spine, and brain stem. In the presence of a sig-
median nerve stimulation. To the authors' knowledge, no adult nificant abnormality at any of these sites, the prognosis for
patients with nontraumatic coma, and very few adult patients awakening cannot be stated confidently. Specific lesions that
with coma of traumatic onset (about 6%), ever awaken in this may interfere with recording at these subcortical sites include
setting; of those who do awaken, most have been reported to peripheral nerve lesions or polyneuropathy, brachial plexus
have severe disability. In some cases, baseline noise or arti-
332
injury, spinal cord injury, and brain stem lesions.
fact might obscure a very small response. Thus, for practical Most medications do not obliterate the short-latency cortical
purposes, absence of a response is often defined as potentials response with median nerve stimulation. These responses are
less 0.5 pV in magnitude. usually easily recordable even under general anesthesia. There
Absolute peak latency, interpeak latency, and amplitude of the is less known, however, about the effect of medications on long-
short-latency cortical responses have not been shown to have a latency responses.
strong prognostic significance. Although patients with normal Volume conduction of far-field potentials may also impair ac-
SEPs do better, on average, than those with present but abnormal curate recording. At times, the brain stem potential recorded
responses (Fig. 9-19), the level of certainty is not usually be- from the mastoid electrode may be volume conducted to the
lieved sufficient to base clinical decision-making on the results. cortical electrodes (Fig. 9-19B). This should not be confused
Recent data suggest that long-latency responses (i.e., greater with a cortical response since the latency is too short (13-14
that 75 ms in the upper limbs) may be of use in predicting out- ms) for a cortical response (19-20 ms).
come as well. Two studies have demonstrated that patients with
bilaterally absent N3 or N70 do not awaken. - One study 213 277
Specificity and Sensitivity
found a cutoff of 118 m s and another 176 m s for predicting
213 277
When detecting nonawakening, the specificity (avoiding
nonawakening; i.e., those with latencies exceeding these values falsely predicting nonawakening) is of greater importance that
did not awaken. sensitivity (correctly predicting nonawakening). Thus, emphasis
is placed on correctly identifying nonawakening. A literature
Pitfalls search was conducted recently to review the medical literature
There are a number of potential pitfalls that may lead to a for all studies reporting the use of SEPs in prediction of out-
false-positive prediction of nonawakening. First, it is important come after coma back to 1966. 251
Articles were included if they specified the etiology of coma have a very poor prognosis for awakening. Further life-sustain-
(traumatic, nontraumatic, or mixed), specified whether adults or ing measures on these patients are futile. Patients in coma due
children or both were studied, and specified outcomes sufficient to traumatic injury with BACRs to median-nerve stimulation
to allow designation of either persistent vegetative state have a poor prognosis, with only 5% awakening but a 95% CI
(PVS)/death, or awakening. Methodology for SEPs also needed of up to 12%. These patients should be considered unlikely to
to be sufficiently specified to allow for assurance that standard awaken, and if they do awaken, they are likely to be severely
technical procedures were followed. Only English language ar- disabled.
ticles were included. Excluded were single case reports, case
series with less than four patients, and articles not found in peer- Traumatic Myelopathy Evaluation
reviewed journals. The authors' search recovered 51 peer-re- Clinicians caring for spinal cord-injured patients are inter-
viewed articles. ested in seeking better ways to predict outcome. As prediction
Hypoxic-Ischemic Coma (Adults). Of 962 adult patients methods are refined, patients and families could be given more
with hypoxic-ischemic coma who had SEPs performed, 29% information about expected functional outcome. Methods used
awakened. Of these 962 patients, 285 (29%) had bilateral ab- for this purpose should be better than current clinical methods,
sence of cortical responses (BACR) to median-nerve stimula- i.e., physical examination including strength, sensation, and
tion. All 285 (100%) either died or remained in a PVS 0-1.0%. reflex testing.
Of the 677 adults with hypoxic-ischemic coma in whom median As discussed above, SEPs are thought to travel through the
nerve SEPs were present, 4 1 % awoke (95% confidence interval dorsal column pathways of the spinal cord. Unfortunately, this
{CI} 37-45%). represents a disadvantage for predicting functional outcome
Coma Due to Intracranial Bleed (Adults). Of 157 adult after spinal cord injury (SCI). The dorsal columns are not only
patients with coma due to intracranial bleed who had SEPs per- topographically distinct from the more functionally significant
formed, 2 1 % awakened. Of these 157 individuals, 72 (46%) had motor pathways, but also have a different blood supply. Thus, it
bilateral absence of cortical responses to median nerve stimula- is possible to have marked differences in the level of impair-
tion; 71 of 72 (99%) either died or remained in PVS. One ment for descending motor pathways and ascending somatosen-
person (1%) awakened (95% CI 0 ^ . 0 % ) . Median nerve SEPs sory pathways. Clinically, this can be apparent when joint
were present in 85 of the 157 (54%) adults with coma due to in- position sense is preserved in the setting of complete paralysis.
tracranial bleed. Of these 85 patients, 38% awakened (95% CI
27-48%).
Traumatic Coma (Adults). Of 711 adult patients with trau- SEPS A N D PROGNOSIS
matic causes of coma who had SEPs reported, 58% awakened
overall. Of these 711 patients, 226 (32%) had bilaterally absent There is good evidence of a correlation between light touch,
cortical responses to median nerve stimulation. Of the 226, vibration, joint position sense (JPS), and S E P s . There is also 30
twelve (5%) awakened (95% CI 2-8%). Of the 12 who awak- good evidence that those with complete or incomplete SCI have
ened, 10 (83% of those who awakened) were reported to have smaller tibial SEP amplitudes than control subjects. However, 45
severe disability on the Glasgow outcome scale, and, 2 of the 12 there have not been significant differences between those with
(17%) had outcomes better than severe disability. clinically complete versus incomplete SCI when examining
There are a number of caveats and limitations to this review. S E P s . Consequently, even in persons with chronic SCI, it is
45
First, there is always the possibility that reports of people awak- difficult to tell the difference between poor outcomes and rela-
ening after having BACR were either missed by the literature tively good outcomes using SEPs.
review or that such reports were not published. Studies attempting to predict outcome in patients with acute
Second, there is always the possibility in many of these stud- SCI, using SEPs as a predictor, have produced variable results.
ies that a self-fulfilling prophecy existed. Perhaps, once median Li and colleagues, studying 36 patients with acute SCI, found
206
SEPs were noted to be absent, a full effort was not made to sus- that the amplitude of the SEP at 2 weeks post-injury (taking the
tain the patient who died before being given an adequate mean ulnar and tibial responses) contributed to the prediction of
chance to awaken. Most authors indicate that SEPs did not in- outcome (as measured by Barthel scores at 6 months), adding
fluence their clinical decisions, but subtle influences cannot be more than the clinical examination (Table 9-25) alone.
excluded. In contrast, Katz and colleagues have found that using
186
Third, the follow-up period varies widely among studies and SEPs offered no added prognostic value when compared with
there could be some late awakeners unknown to the authors. the clinical examination in 57 patients with acute SCI. No pa-
While most patients who eventually awaken do so within days tient with a complete SCI on initial physical evaluation ever
to weeks, there are some patients, particularly after traumatic
brain injury, who do not awaken until months after injury. T a b l e 9-25. P r e d i c t o r s of O u t c o m e in P a t i e n t s
Finally, the SEP records a long segment of the nervous w i t h Spinal C o r d Injury
system, form peripheral nerve to brain. It is possible that in some
patients (particularly with traumatic injuries), a second lesion at % Outcome
the spinal cord or brain stem level could have produced absence Predictor rValue Explained PValue
of cortical responses. While most authors typically record from SEP amplitude (mean ulnar .75 56% .0001
subcortical sites to exclude these possibilities, not all do. It is and tibial)
strongly recommended that recordings from peripheral nerve, Joint position sense .64 41% .0001
cervical spine, and mastoid (reflecting brain stem function) be
included in all SEPs performed for prognostic purposes. Motor strength, joint position .87 76% .0001
sense, and pinprick plus SEP
In conclusion, the data indicate that adults in coma due to
amplitude
nontraumatic causes with BACRs to median-nerve stimulation
developed motor return. An initial examination demonstrating 19. Baer RD, Johnson EW: Motor nerve conduction velocity in premature infants.
Arch Phys Med Rehabil 1965;46:698-704.
incomplete SCI heralded a result of walking or better in 56.4%
20. Bas GA, Cracco JB, Cracco RQ: Scalp-recorded short latency cortical and sub-
of incomplete patients with SCI. Both the initial physical exam- cortical somatosensory evoked potentials to peroneal nerve stimulation.
ination and evoked potentials were reasonable predictors of fur- Electroencephalogr Clin Neurophysiol 1981;52:1-8.
ther motor improvement. Evoked potentials, however, added 21. Baran EM, Daube JR: Lower extremity somatosensory e%'oked potentials: An
AAEM workshop. Rochester, MN, American Association of Electromyography
little or no useful prognostic information to the initial physical and Electrodiagnosis, 1984.
examination in either complete or incomplete SCI patient 22. Barr ML, Kiernan JA: The Human Nervous System (5th ed). Philadelphia, JB
groups. Thus, there is no strong evidence to suggest that SEPs Lippincott Co., 1988.
can play a significant role in the prognosis of recovery follow- 23. Beall JE, Appelbaum AE, Foreman R, et al: Spinal cord potentials evoked by
cutaneous afferents in the monkey. J Neurophysiol 1977;40:199-211.
ing SCI. 24. Beck A: The determination of localization of brain and spinal cord function by
means of electrical phenomena. Cbl Physiol 1890;4:473-476.
25. Beck A: Determination of localization in the brain and spinal cord by means of
electrical phenomena. Rozpr Wydz mat-przyr Polsk Akadam Ser II, 1891:1:
CONCLUSION 186-232.
26. Beck A, Cybulski N: Further research on the electrical phenomena of the cere-
Somatosensory evoked potentials involve a significant bral cortex in monkeys and dogs. Rozpr Wydz mat-przyr 1891;32:369-375.
27. Benita M, Caude H: Effects of local cooling upon conduction time and synaptic
amount of complex neural theory but are relatively easy to per- transmission. Brain Res 1972;36:135-151.
form. It is paramount, however, that the practitioner take the 28. Beric' A, Prevec TS: The early negative potential evoked by stimulation of tibial
necessary time to understand the theoretical foundation upon nerve in man. J Neurol Sci 1981 ;50:299-306.
which the field of SEPs is based to obtain optimal, error-free 29. Beric' A: Stability of lumbosacral somatosensory evoked potentials in a long-
term follow-up. Muscle Nerve 1988; 11:621 -626
recordings. Significant work remains to be accomplished in this 30. Beric' A: Cortical somatosensory evoked potentials in spinal cord injury pa-
area. Applying the principles outlined in this chapter, and read- tients. J Neurol Sci 1992;107:50-59.
ing the appropriate references for each technique, should allow 31. Bernhard CG: The spinal cord potentials in leads from the cord dorsum in rela-
tion to peripheral source of afferent stimulation. Acta Physiol Scand 1953;29
one to apply SEPs in the diagnosis of pathology.
(SuppI 106): 1-29.
32. Brazier MAB: Pioneers in the discovery of evoked potentials. Electro-
encephalogr Clin Neurophysiol 1984;59:2-8.
33. Brunko E, Zegers de Beyl D: Prognostic value of early cortical somatosensory
REFERENCES evoked potentials after resuscitation from cardiac arrest. Electroencephalogr
Clin Neurophysiol 1987;66:15-24.
1. Abbruzzese G, Abbruzzese M, Cocito L, et al: Conduction time of the lemnis- 34. Burke D, Skuse NF, Lethlean AK: Cutaneous and muscle afferent components
cal pathway in males and females. Acta Neurol Scand 1980;62:132-136. of the cerebral potentials evoked by electrical stimulation of human peripheral
2. Allison T, Hume AL: A comparative analysis of short latency somatosensory nerves. Electroencephalogr Clin Neurophysiol 1981;51:579-588.
evoked potentials in man, monkey, cat, and rat. Exp Neurol 1981;72:592-611. 35. Burke D, Gandevia SC, McKeon B, et al: Interactions between cutaneous and
3. Allison T: Scalp and cortical recordings of initial somatosensory cortex activity muscle afferent projections to cerebral cortex in man. Electroencephalogr Clin
to median nerve stimulation in man. Ann N Y Acad Sci 1982;388:671-678. Neurophysiol 1982;53:349-360.
4. Allison T, Wood CC, Goff WR: Brain stem auditory, pattern-reversal visual, 36. Campbell B: The distribution of potential fields within the spinal cord. Anat
and short-latency somatosensory evoked potentials: Latencies in relation to age, Rec 1945;91:77-88.
sex, and brain and body size. Electroencephalogr Clin Neurophysiol 37. Cant BR, Shaw NA: Central somatosensory conduction time: Method and clin-
1983;55:619-636. ical applications. In Cracco RQ, Bodis-Wollnar I (eds): Evoked Potentials. New
5. Allison T, Hume AL, Wood CC, et al: Developmental aging changes in so- York, Alan R Liss Inc., 1986, pp 58-67.
matosensory, auditory and visual evoked potentials. Electroencephalogr Clin 38. Carpenter MB, Stein BM, Shriver JE: Central projections of dorsal spinal roots
Neurophysiol 1984;58:14-24. in the monkey II. Lower thoracic, lumbosacral, and coccygeal dorsal roots. Am
6. American Electroencephalographic Society: Recommended standards for the J Anat 1968;123:75-118.
clinical practice of evoked potentials. J Clin Neurophysiol 1984; 1:6-14. 39. Carter LP, Raudzens PA, Gaines O, et al: Somatosensory evoked potentials and
7. American Electroencephalographic Society: Recommended standards for short- cortical blood flow during craniotomy for vascular disease. Neurosurgery
latency somatosensory evoked potentials. J Clin Neurophysiol 1984;1:41-53. 1984;15:22-28.
8. Amassian VE, Berlin L: Early cortical projection of Group I afferents in the 40. Caton R: The electric currents of the brain. Br Med J 1875;2:278.
forelimb muscle nerves of cat. J Physiol 1958; 143:61 R 41. Caton R: Interim report on investigation of the electric currents of the brain. Br
9. Aminoff MJ, Davis SL, Panitek HS: Serial evoked potential studies in patients Med J 1877(Suppl 1 ):62-62.
with definite multiple sclerosis: clinical relevance. Arch Neurol 1984;41: 42. Caton R: Research on electrical phenomena of cerebral grey matter. Ninth Inter
1197-1202. Med Congr 1887,3:246-249.
10. Aminoff MJ, Goodin DS, Barbara NM, et al: Dermatomal somatosensory 43. Center for Disease Control: Recommendations for prevention of HIV transmis-
evoked potentials in unilateral lumbosacral radiculopathy. Ann Neurol sion in health-care settings. MMWR 1987;36(Suppl 2s).
1985;17:171-176. 44. Center for Disease Control: Universal precautions for prevention of transmis-
11. Aminoff MJ: Use of somatosensory evoked potentials to evaluate the peripheral sion of human immunodeficiency virus, hepatitis B, and other blood bourne
nervous system. J Clin Neurophysiol 1987:4:135-144. pathogens in health care setting. MMWR 1988;37:377-382, 387-388.
12. Aminoff MJ: Evoked potential studies in neurological diagnosis and manage- 45. Chabot R, York DH, Watts C, Waugh WA: Somatosensory evoked potentials
ment. Ann Neurol 1990;28:706-710. evaluated in normal subjects and spinal cord injured patients. J Neurosurg
13. Anziska B, Cracco RQ, Cook AW, et al: Somatosensory far-field potentials: 1985;63:544-551.
Studies in normal subjects and patients with multiple sclerosis. Electro- 46. Chatrian G, Lettich E, Nelson PL: Ten percent electrode system for topographic
encephalogr Clin Neurophysiol 1978;45:602-610. studies of spontaneous and evoked EEG activities. Am J EEG Technol 1985;25:
14. Anziska B, Cracco RQ: Short latency somatosensory evoked potentials: Studies 83-92.
in patients with focal neurological disease. Electroencephalogr Clin 47. Chatrian G, Lettich E, Nelson PL: Modified nomenclature for the 10% elec-
Neurophysiol 1980;49:227-239. trode system. J Clin Neurophysiol 1985;5:183-186.
15. Anziska BJ, Cracco RQ: Short latency SEPs to median nerve stimulation: Com- 48. Chiappa KH: Pattern shift visual, brainstem auditory, and short-latency so-
parison of recording methods and origin of components. Electroencephalogr matosensory evoked potentials in multiple sclerosis. Neurology 1980:30:
Clin Neurophysiol 1981 ;52:531 -539. 110-123.
16. Appelberg B, Hulliger M, Johansson H, et al: Actions on a-motoneurones 49. Chiappa KH, Choi SK, Young RR: Short-latency somatosensory evoked poten-
elicited by electrical stimulation of group I muscle afferent fibers in the hind tials following median nerve stimulation in patients with neurological lesions.
limb of the cat. J Physiol 1983;335:237-253. Clinical uses of cerebral, brainstem and spinal somatosensory evoked poten-
17. Arezzo J, Legatt AD, Vaughan HG: Topography and intracranial sources of so- tials. Prog Clin Neurophysiol 1980;7:264-281.
matosensory evoked potentials in the monkey. Electroencephalogr Clin 50. Chiappa KH, Ropper AH: Evoked potentials in clinical medicine (Part I). N
Neurophysiol 1979;46:155-172. Engl J Med 1982;306:1140-1150.
18. Austin GM, McCouch GP: Presynaptic component of intermediary cord poten- 51. Chiappa KH, Ropper AH: Evoked potentials in clinical medicine (Part II). N
tial. J Neurophysiol 1955;18:441-451. Engl J Med 1982;306:1205-1211.
52. Chiappa KH: Evoked Potentials in Clinical Medicine. New York, Raven Press, 85. Desmedt JE, Cheron G: Spinal and far-field components of the human so-
1997. matosensory evoked potentials to posterior tibial nerve stimulation analysis
53. Chu N: Somatosensory evoked potentials: Correlations with height. with esophageal derivation and noncephalic reference recordings. Electro-
Electroencephalogr Clin Neurophysiol 1986;65:169-176. encephalogr Clin Neurophysiol 1983;56:635-651.
54. Clark WA: Average response computer (ARC). Quart Progr Rep, Research 86. Desmedt JE, Tran Huy N: Bit-mapped colour imaging of the potential fields
Laboratory of Electronics MIT, Cambridge MA, 1958, pp 114-117. of propagated and segmental subcortical components of somatosensory
55. Cohen L, Pratt H, Starr A: Cerebral somatosensory potentials evoked by muscle evoked potentials in man. Electroencephalogr Clin Neurophysiol 1984;58:
stretch, cutaneous taps, and electrical stimulation of peripheral nerves in lower 481-497.
extremities of humans, Brain 1985;108:103-122. 87. Desmedt JE, Bourguet M: Color imaging of parietal and frontal somatosensory
56. Cook AW, Browder EJ: Function of the posterior columns in man. Arch Neurol potential fields evoked by stimulation of median or posterior tibial nerve in
1965;12:72-79. man. Electroencephalogr Clin Neurophysiol 1985;62:1-17.
57. Cracco JB, Cracco RQ, Graziani L: The spinal evoked response in infants and 88. Deupree DL, Jewett DL: Far-field potentials due to action potentials traversing
children. Neurology 1975;25:31-36. curved nerves, reaching cut nerve ends, and crossing boundaries between cylin-
58. Cracco JB, Cracco RQ, Stolov WR: Spinal evoked potential in man: a matura- drical volumes. Electroencephalogr Clin Neurophysiol 1988;70:355-362.
tional study. Electroencephalogr Clin Neurophysiol 1979;46:58-64. 89. Dimitrijevic MR, Larsson LE, Lehmkuhl D et al: Evoked spinal cord and nerve
59. Cracco JB, Cracco RQ: Somatosensory evoked potentials in the evaluation of root potentials in humans using a non-invasive recording technique. Electro-
infants and children with spinal dysraphism. Electroencephalogr Clin Neuro- encephalogr Clin Neurophysiol 1978;45:331-340.
physiol 1981;52:590. * 90. Dimitrijevic MR, Prevac TS, Sherwood AM: Somatosensory perception and
60. Cracco JB. Cracco RQ: Spinal somatosensory evoked potentials: Maturational cortical evoked potentials in established paraplegia. J Neurol Sci 1983;60:
and clinical studies. Ann NY Acad Sci 1982;388:526-537. 253-265.
61. Cracco RQ: The initial positive potential of the human scalp-recorded so- 91. Donchin E, Callaway E, Cooper R, et al: Publication criteria for studies of
matosensory evoked response. Electroencephalogr Clin Neurophysiol 1972;32: evoked potentials (EP) in man: Report of a committee. Prog Clin Neurophysiol
623-629. 1977;1:1-11.
62. Cracco RQ: Spinal evoked response: Peripheral nerve stimulation in man. 92. Dong WK, Harkins SW, Ashleman BT: Origins of cat somatosensory far-field
Electroencephalogr Clin Neurophysiol 1973;35:379-386. and early near-field evoked potentials. Electroencephalogr Clin Neurophysiol
63. Cracco RQ, Cracco JB: Somatosensory evoked potential in man: Far-field po- 1982;53:143-165.
tentials. Electroencephalogr Clin Neurophysiol 1976;41:460-466. 93. Dorfman LJ, Bosley TM: Age-related changes in peripheral and central nerve
64. Cracco RQ, Cracco JB, Sarnowski R, et al: Spinal evoked potentials. Prog Clin conduction in man. Neurology 1979;29:38-44.
Neurophysiol 1980;7:87-104. 94. Drummond JC, Todd MM, U HS: The effect of high dose sodium thiopental on
65. Cracco RQ, Anziska BJ, Cracco JB, et al: Short-latency somatosensory evoked brain stem auditory and median nerve somatosensory evoked responses in man.
potentials to median and peroneal nerve stimulation: Studies in normal sub- Anesthesiol 1985;63:249-254.
jects and patients with neurologic disease. Ann N Y Acad Sci 1982;388: 95. Dubois M, Cappola R, Buchsbaum M, et al: Somatosensory evoked potentials
412-425. during whole body hyperthermia in humans. Electroencephalogr Clin
; 66. Cruse R, Klem G, Lesser RP, et al: Paradoxical lateralization of cortical poten- Neurophysiol 1981;52:157-162.
j tials evoked by posterior tibial nerve. Arch Neurol 1982;39:222-225. 96. Dumitru D, Walsh NE: Practical instrumentation and common sources of error.
J 67. Cusick JF, Myklebust J, Larson SJ, et al: Spinal evoked potentials in the pri- Am J Phys Med Rehabil 1988;67:55-65.
mate: neural substrate. J Neurosurg 1978;49:551-557. 97. Dumitru D, Windsor RE: Subsartorial entrapment of the saphenous nerve of a
68. Cusick JF, Mykelbust JB, Larson SJ, et al: Spinal cord evaluation by cortical competitive body builder. Physician Sports Med 1989,17:116-125.
evoked responses. Arch Neurol 1979;36:140-143. 98. Dumitru D, Walsh NE: Electrophysiologic instrumentation. In Dumitru D (ed):
69. Davis PA: The electrical response of the human brain to auditory stimuli. Amer Clinical Electrophysiology; PM&R State of the Art Reviews. Philadelphia,
J Physiol 1939;126:475-476. Hanley & Belfus, 1989, pp 683-699.
70. Davis PA: Effects of acoustic stimuli on the waking human brain. J 99. Dumitru D, Kalantri A, Dierschke B: Somatosensory evoked potentials of the
Neurophysiol 1939;2:494-499. medial and lateral plantar and calcaneal nerves. Muscle Nerve 1991; 14:
71. Davis SL, Aminoff MJ, Panitch HS: Clinical correlations of serial somatosen- 665-671.
sory evoked potentials in multiple sclerosis. Neurology 1985;35:359-365. 100. Dumitru D, J Lester: Comparison of needle and surface somatosensory evoked
72. Dawson GD: Cerebral responses to electrical stimulation of peripheral nerve in potential normative data. Arch Phys Med Rehabil 1991;72:989-992.
man. J Neurol Neurosurg Psychiatry 1947;10:137-140. 101. Dumitru D, Powell GD, King JC: The effect of different needle recording elec-
73. Dawson GD: Investigations on a patient subject to myoclonic seizures after sen- trodes on somatosensory evoked potentials and internal waveform variation.
sory stimulation. J Neurol Neurosurg Psychiatry 1947;10:141-162. Am J Phys Med Rehabil 1992;71:164-169.
74. Dawson GD: Cerebral responses to nerve stimulation in man. Brit Med Bulletin 102. Dumitru D, Newton BY, Drey fuss P: Segmental vs dermatomal somatosensory
1950:6:326-329. evoked potentials: Normal intertrial variation and side-to-side comparison. Am
75. Dawson GD: A summation technique for the detection of small evoked poten- J Phys Med Rehabil 1993;72:75-83.
tials. Electroencephalogr Clin Neurophysiol 1954;6:65-84. 103. Eggermont JJ: On the rate of maturation of sensory evoked potentials.
76. Delbeke J, McComas AJ, Kopec SJ: Analysis of evoked lumbosacral potentials Electroencephalogr Clin Neurophysiol 1988;70:293-305.
in man. J Neurol Neurosurg Psychiatry 1978;41:293-302. 104. Eisen A, Nudleman K: F-wave and cervical somatosensory response conduc-
77. DeLisa JA, Mackenzie K, Baran EM: Manual of Nerve Conduction Velocity tion from the seventh cervical spinous process to cortex in multiple sclerosis.
and Somatosensory Evoked Potentials (2nd ed). New York, Raven Press, 1987. Can J Neurol Sci 1978;5:289-295.
78. Dellon AL, Mackinnon SE: Tibial nerve branching in the tarsal tunnel. Arch 105. Eisen A, Stewart J, Nudleman K, et al: Short-latency somatosensory responses
Neurol 1984;41:645-646. in multiple sclerosis. Neurology 1979;29:827-834.
79. Desmedt JE, Cheron G: Central somatosensory conduction in man: Neural gen- 106. Eisen A, Odusote K: Central and peripheral conduction times in multiple scle-
erators and interpeak latencies of the far-field components recorded from neck rosis. Electroencephalogr Clin Neurophysiol 1980;48:253-265.
and right or left scalp and earlobes. Electroencephalogr Clin Neurophysiol 107. Eisen A, Elleker G: Sensory nerve stimulation and evoked cerebral potentials.
1980:50:382^103. Neurology 1980;30:1097-1105.
80. Desmedt JE, Cheron G: Somatosensory evoked potentials to finger stimulation 108. Eisen A, Paty D, Purves S, et al: Occult fifth nerve dysfunction in multiple scle-
in healthy octogenarians and in young adults: Waveforms, scalp topography, rosis. J Can Neurol Soc 1981;8:221-225.
and transit times of parietal and frontal components. Electroencephalogr Clin 109. Eisen A, Purves S Hoirch M: Central nervous system amplification: Its poten-
Neurophysiol 1980;50:404-425. tial in the diagnosis of early multiple sclerosis. Neurology 1982;32:359-364.
81. Desmedt JE Cheron G: Somatosensory pathway and evoked potentials in 110. Eisen A , Cracco RQ: Overuse of evoked potentials: Caution. Neurology
normal human aging. Clinical uses of cerebral, brainstem and spinal somatosen- * 1983;33:618-621.
sory evoked potentials. Prog Clin Neurophysiol 1980;7:162-169. 111. Eisen A, Hoirch M, Moll A: Evaluation of radiculopathies by segmental stimu-
82. Desmedt JE, Cheron G: Prevertebral (oesophageal) recording of subcortical so- lation and somatosensory evoked potentials. Can J Neurol Sci 1983; 10:
matosensory evoked potentials in man: The spinal PI3 component and dual 178-182.
nature of spinal generators. Electroencephalogr Clin Neurophysiol 1981;52: 112. Eisen A, Burton K, Larsen A, et al: A new indirect method for measuring spinal
257-275. conduction velocity in man. Electroencephalogr Clin Neurophysiol 1984;59:
83. Desmedt JE, Cheron G: Non-cephalic reference recording of early somatosen- 204-213.
sory potentials to finger stimulation in adult or aging normal man: Differentia- 113. Eisen A, Hoirch M, Fink M, et al: Noninvasive measurement of central sensory
tion of widespread N18 and contralateral N20 from the prerolandic P22 and and motor conduction. Neurology 1985;35:503-509.
N30 components. Electroencephalogr Clin Neurophysiol 1981;52:553-570. 114. Eisen A, Odusote K, Bozek C, et al: Far-field potentials from peripheral nerve:
84. Desmedt JE, Nguyen TH, Carmeliet J: Unexpected latency shifts of the station- Generators at sites of muscle mass change. Neurology 1986;36:815-818.
ary P9 somatosensory evoked potential far-field with changes in shoulder posi- 115. Eisen A, Odusote K, Li D, et al: Comparison of magnetic resonance imaging
tion. Electroencephalogr Clin Neurophysiol 1983;56:628-634. with somatosensory testing in ms suspects. Muscle Nerve 1987;10:385-390.
116. Emerson RG, Seyal M, Pedley TA: Somatosensory evoked potentials following 150. Haldeman S, Bradley WE, Bhatia NN, et al: Pudendal evoked responses in neu-
median nerve stimulation. 1. The cervical components. Brain 1984; 107: rologic diseases. Neurology 1982;89:A67.
169-182. 151. Haldeman S, Bradley WE, Bhatia NN, et al: Cortical evoked potentials on stim-
117. Emerson RG: Anatomic and physiologic bases of posterior tibial nerve so- ulation of pudendal nerve in women. Urology 1983;6:590-593.
matosensory evoked potentials. Neurol Clin 1988;6:735-749. 152. Halliday AM, Wakefield GS: Cerebral evoked potentials in patients with disso-
118. Engler GL, Spielholz Nl, Bernhard WN, et al: Somatosensory evoked potentials ciated sensory loss. J Neurol Neurosurg Psychiatry 1953;26:211-219.
during Harrington instrumentation for scoliosis. J Bone Joint Surg 1978;60A: 153. Hashimoto I: Somatosensory evoked potentials from human brainstem: origin
528-532. of short latency potentials. Electroencephalogr Clin Neurophysiol 1984;57:
119. Epstein M, Andriola M: Introduction to EEG and Evoked Potentials. 221-227.
Philadelphia, JB Lippincott, 1986, p 6. 154. Haymaker W, Woodhall B: Peripheral Nerve Injuries. Philadelphia, WB
120. Erlanger J, Gasser HS: Electrical Signs of Nervous Activity. Philadelphia, Saunders, 1953.
Philadelphia Univ. Press, 1937. 155. Head H, Campbell AW: The pathology of herpes zoster and its bearing on sen-
121. Ertekin C: Studies on the human evoked electrospinogram. I. The origin of the sory localization. Brain 1900;23:353-523.
segmental evoked potentials. Acta Neurol Scand 1976;53:1-19. 156. Hore J, Preston JB, Durkovic RG, et al: Responses of cortical neurons (areas 3a
122. Ertekin C: Studies on the human evoked electrospinogram. II. The conduction and 4) to ramp stretch of hindlimb muscles in the baboon. J Neurophysiol
velocity along the dorsal funiculus. Acta Neurol Scand 1976;53:21-38. 1976;39:484-500.
123. Evarts EV: Motor control. In Brooks VB (ed): Handbook of Physiology. Section 157. Howland B, Lettvin JY, McCulloch WS, et al: Reflex inhibition by dorsal root
1, Vol II. Philadelphia, Williams & Wilkins, 1981. interaction. J Neurophysiol 1955;18:1-17.
124. Facco E, Munari M, Baratto F, et al: Multimodality evoked potentials (auditory, 158. Hughes J, Gasser HS: Some properties of the cord potentials produced by a
somatosensory, and motor) in coma. Neurophysiol Clin 1993;23(2-3):237-258. single afferent volley. Am J Physiol 1934;108:295-306.
125. Foerster O: The dermatomes in man. Brain 1933;56:1-39. 159. Hume AL, Cant BR: Conduction time in central somatosensory pathways in
126. Gajdusek DC, Gibbs CJ, Asher DM, et al: Precautions in medical care of han- man. Electroenceph Clin Neurophysiol 1978;45:361-375.
dling materials from patients with transmissible virus dementia (Creutzfeldt- 160. Hume AL, Cant BR, Shaw NA: Central somatosensory conduction time in co-
Jakob disease). N Engl J Med 1977;297:1256-1258. matose patients. Ann Neurol 1979;5:379-384.
127. Gamstorp I: Normal conduction velocity of ulnar, median, and peroneal nerves 161. Hume AL, Cant BR: Central somatosensory conduction after head injury. Ann
in infancy, childhood, and adolescence. Acta Paediat. 1963;(Suppl 146):68-76. Neurol 1981;19:411^119.
128. Gandevia SC, Burke D, McKeon B: The projection of muscle afferents from the 162. Hume AL, Cant BR, Shaw NA, et al: Central somatosensory conduction time
hand to cerebral cortex in man. Brain 1984;107:1-13. from 10 to 79 years. Electroencephalogr Clin Neurophysiol 1982;54:49-54.
129. GanesT: A study of peripheral cervical and cortical evoked potentials and affer- 163. Hunt CC: Relation of function to diameter in afferent fibers of muscle nerves. J
ent conduction times in the somatosensory pathway. Electroencephalogr Clin Gen Physiol 1954;38:117-131.
Neurophysiol 1980;49:446-451. 164. Ikuta T, Furuta N: Sex differences in the human group mean SEP. Electro-
130. GanesT: Somatosensory conduction times and peripheral, cervical and cortical encephalogr Clin Neurophysiol 1982;54:449^57.
evoked potentials in patients with cervical spondylosis. J Neurol Neurosurg 165. Iragui-Madoz VJ, Wiederholt WC: Far-field somatosensory evoked potentials
Psychiatry 1980;43:683-689. in the cat. Electroencephalogr Clin Neurophysiol 1977;43:646-657.
131. Gasser HS, Graham HT: Potentials produced in the spinal cord by stimulation 166. Iragui VJ: The cervical somatosensory evoked potential in man: far-field con-
of the dorsal roots. Am J Physiol 1933;103:303-320. ducted and segmental components. Electroencephalogr Clin Neurophysiol
132. Giblin DR: The effect of lesions of the nervous system on cerebral responses to 1984;57:228-235.
peripheral nerve stimulation. Electroencephalogr Clin Neurophysiol 1960; 12: 167. Izzo KA, Sridhara CR, Rosenholtz H, et al: Sensory conduction studies of the
262. branches of the superficial peroneal nerve. Arch Phys Med Rehabil 1981;62:24-27.
133. Giblin DR: Somatosensory evoked potentials in healthy subjects and in patients 168. Jacobson GP, Tew JM: The origin of scalp recorded P14 following electrical
with lesions of the nervous system. Ann N Y Acad Sci 1964; 112:93-142. stimulation of the median nerve. Intraoperative observation. Electroencephalogr
134. Gilman S, Winans SS: Manter and Gatz's Essentials of Clinical Neuroanatomy Clin Neurophysiol 1988;71:73-76.
and Neurophysiology. Philadelphia, FA Davis, 1982, pp 15-18. 169. Jabre JF: The superficial peroneal sensory nerve revisited. Arch Neurol
135. Glees P, Soler J: Fiber content of posterior column and synaptic connections of 1981;38:666-667.
the nucleus gracilis. Z Zellforsch Mikrosk Anat 1951;36:381-400. 170. Jasper HH: The ten-twenty electrode system of the International Federation.
136. Glick M, Haldeman S: The electrodiagnostic and neurovisceral evaluation of Electroencephalogr Clin Neurophysiol 1958;10:371-375.
patients with spinal cord injuries. Neurology 1984;34:208. 171. Jewett DL, Williston JS: Auditory-evoked far-fields averaged from the scalp of
137. Goff WR, Rosner BS, Allison T: Distribution of cerebral somatosensory evoked humans. Brain 1971;94:681-696.
responses in normal man. Electroencephalogr Clin Neurophysiol 1962; 14: 172. Jewett DL, Deupree DL: Far-field potentials recorded from action potentials
697-713. and from a tripole in a hemicylindrical volume. Electroencephalogr Clin
138. Goff WR: Human average evoked potentials: Procedures for stimulating and Neurophysiol 1989:72:439-449.
recording. Bioelectric Recording Techniques, Part B, Electroencephalography 173. Jewett DL, Deupree DL, Bommannan D: Far-field potentials generated by
and Human Brain Potentials, Thompson RF, Patterson MM (eds). New York, action potentials of isolated frog sciatic nerves in a spherical volume.
Academic Press, 1974, pp 101-131. Electroencephalogr Clin Neurophysiol 1990;75:105-117. j
139. Goff WR, Allison T, Vaughn HG: The functional neuroanatomy of event related 174. Jones SJ: Short latency potentials recorded from the neck and scalp following |
potentials. In Callaway E, Tueting P, Koslow SH (eds): Event Related Potentials median nerve stimulation in man. Electroencephalogr Clin Neurophysiol
in Man. New York, Academic Press. 1978, pp 1-79. 1977;43:853-863.
140. Goldie WD, Chiappa KG, Young RR, et al: Brainstem auditory and short-la- 175. Jones SJ, Small DG: Spinal and subcortical evoked potentials following stimu-
tency somatosensory evoked responses in brain death. Neurology 1981;31: lation of the posterior tibial nerve in man. Electroencephalogr Clin ;
248-256. Neurophysiol 1978;44:290-306.
141. Goodgold J, Kopell HP, Spielholz Nl: The tarsal tunnel syndrome. N Engl J 176. Jones SJ, Carrol WM, Halliday AM: Peripheral and central somatosensory
Med 1965;273:742-745. nerve conduction in Charcot-Marie-Tooth disease and comparison with
142. Green JB, McLeod S: Short latency somatosensory evoked potentials in pa- Friedreich's ataxia. J Neurol Sci 1983;61:135-149.
tients with neurologic lesions. Arch Neurol 1979;36:846-851. 177. Jones SJ, Wynn Parry CB, Landi A: Diagnosis of brachial plexus traction le-
143. Green JB, Walcoff MR, Lucke JF: Comparison of phenytoin and phenobarbital sions by sensory nerve potentials and somatosensory evoked potentials. Injury
effects on far-field auditory and somatosensory evoked potential interpeak la- 1981;12:376-382.
tencies. Epilepsia 1982;23:417—421. 178. Kaji R, Ritsuho T, Kawaguchi S, et al: Origin of short latency somatosensory
144. Green JB, Walcoff M, Lucke JF: Phenytoin prolongs far-field somatosensory evoked potentials to median nerve stimulation in the cat. Comparison of the
and auditory evoked potential interpeak latencies. Neurology 1982;32:85-88. recording montages and effect of laminectomy. Brain 1986;109:443-468.
145. Greenberg JA. Erwin CW: Somatosensory evoked potentials and the neurologic 179. Kakigi R, Shibasaki H, Hashizume A, et al: Short latency somatosensory
exam: A study of clinical correlates. Electroencephalogr Clin Neurophysiol evoked spinal and scalp-recorded potentials following posterior tibial nerve
1986;64:31P-32P. stimulation in man. Electroencephalogr Clin Neurophysiol 1982;53:602-611.
146. Greenberg JA, Kaplan PW, Erwin CW: Somatosensory evoked potentials and 180. Kakigi R, Shibasaki H: Scalp topography of the short latency somatosensory
the dorsal column myth. J Clin Neurophysiol 1987;4:189-196. evoked potentials following posterior tibial nerve stimulation in man.
147. Grisolia JS, Wiederholt WC: Short latency somatosensory evoked potentials Electroencephalogr Clin Neurophysiol 1983;56:430-437.
from radial, median, and ulnar nerve stimulation in man. Electroencephalogr 181. Kakigi R: The effect of aging on somatosensory evoked potentials following
Clin Neurophysiol 1980;50:375-382. stimulation of the posterior tibial nerve in man. Electroencephalogr Clin
148. Haldeman S, Bradley WE, Bhatia NN, et al: Pudendal evoked responses. Arch NeurophysiolT987;68:277-286.
Neurol 1982;89:280-283. 182. Kameyama S, Yamada T, Matsuoka H, et al: Stationary potentials after median
149. Haldeman S, Bradley WE, Bhatia NN: Evoked responses from the pudendal nerve stimulation: changes with arm position. Electroencephalogr Clin Neuro-
nerve. J Urol 1982;128:974-980. physiol 1988;71:348-356.
183. Karkela J, Bock E, Kaukinen S: CSF and serum brain-specific creatine kina- 214. Magherini P, Pompeiano O, Sequin JJ: Responses of nucleus Z neurons to vi-
seisoenzyme (CK-BB), neuron-specific enolase (NSE) and neural cell adhesion bration of hindlimb extensor muscles in the decerebrate cat. Archs Itial Biol
molecule (NCAM) as prognostic markers for hypoxic brain injury after cardiac 1975;113:150-189.
arrest in man. J Neurol Sci 1993;! 16:100-109. 215. Magladery JW, Porter WE, Park AM, et al: Electrophysiological studies of
184. Katifi HA, Sedgwick EM: Somatosensory evoked potentials from posterior nerve and reflex activity in normal man. Nerve and reflex activity in man. IV.
tibial nerve and lumbosacral dermatomes. Electroencephalogr Clin Neuro- Bull Johns Hopkins Hosp 1951;88:499-519.
physiol 1986;65:249-259. 216. Matthews WB, Beauchamp M, Small DC: Cervical somatosensory evoked re-
185. Katifi HA, Sedgwick EM: Evaluation of the dermatomal somatosensory evoked sponse in man. Nature 1974;252:230-232.
potential in the diagnosis of lumbo-sacral root compression. J Neurol 217. Matthews WB, Read DJ, Pountney E: Effect of raising body temperature on
Neurosurg Psychiatry 1987;50:1204-1210. visual and somatosensory evoked potentials in patients with multiple sclerosis.
186. Katz RT, Toleikis RJ, Knuth AE: Somatosensory-evoked and dermatomal- J Neurol Neurosurg Psychiatry 1979;42:250-255.
evoked potentials are not clinically useful in the prognostication of acute spinal 218. Matthews WB, Small DG: Serial recording of visual and somatosensory evoked
cord injury. Spine 1991;16:730-735. potentials in multiple sclerosis. J Neurol Sci 197940:11-21.
187. Kazis HA, Vlaikidis N, Xafenies D, et al: Fever and evoked potentials in multi- 219. Mauguiere F, Courjon J: The origins of short-latency somatosensory evoked po-
ple sclerosis. J Neurol 1982;227:1-10. tentials in humans. Ann Neurol 1981;9:607-611.
188. Kazis A, Vlaikidis N, Pappa P, et al: Somatosensory and visual evoked poten- 220. Mauguiere F, Desmedt JE, Courjon J: Asteriognosis and dissociated loss of
tials in human aging. Electromyogr Clin Neurophysiol 1983;23:49-59. frontal or parietal components of somatosensory evoked potentials in hemi-
189. Kimura J, Yamada T, Shivapour E, et al: Neural pathways of somatosensory spheric lesions. Brain 1983;106:271-311.
evoked potentials: clinical implications. Electroencephalogr Clin Neurophysiol 221. Mauguiere F, Desmedt JE, Courjon J: Neural generators of N18 and P14 far-
1982;(Suppl 36):328-335. field somatosensory evoked potentials studied in patients with lesions of the
190. Kimura J, Mitsudome A, Yamada T, et al: Stationary peaks from a moving thalamus or thalamocortical radiations. Electroencephalogr Clin Neurophysiol
source in far-field recording. Electroencephalogr Clin Neurophysiol 1984;58: 1983;56:283-292.
351-361. 221a. Mauguiere F, Allison T, Babiloni C, et al: Somatosensory evoked potentials.
191. Kimura J: Abuse and misuse of evoked potentials as a diagnostic test. Arch Electroencephalogr Clin Neurophysiol. 1999(Suppl 52):79-90.
Neurol 1985;42:78-80. 222. McCloskey Dl: Kinesthetic sensibility. Physiol Rev 1978;58:763-820.
192. Kimura J, Kimura A, Ishida T, et al: What determines the latency and amplitude 223. Mervaala E, Paakkonen A, Partanen JV: The influence of height, age and gender
of stationary peaks in far-field recordings? Ann Neurol 1986; 19:479^486. on the interpretation of median nerve SEPs. Electroencephalogr Clin Neuro-
193. Kimura J, Ishida T, Suzuki S, et al: Far-field recording of the junctional poten- physiol 1988;71:109-113.
tial generated by median nerve volleys at the wrist. Neurology 1986;36: 224. Misulis KE: Basic electronics for clinical neurophysiology. J Clin Neurophysiol
1451-1457. 1989;6:41-74.
194. Kimura J, Kimura A, Masafumi M, et al: Model for far-field recordings of SEP. 225. Molaie M: Scalp-recorded short and middle latency peroneal somatosensory
In Cracco RQ, Bodis-Wollner I (eds): Evoked Potentials. New York, Alan R evoked potentials in normals: Comparison with peroneal and median nerve
Liss, Inc., 1986, pp 246-261. SEPs in patients with unilateral hemispheric lesions. Electroencephalogr Clin
195. Kobayashi H, Yaguchi K: A statistical method of component identification of Neurophysiol 1987;68:107-118.
average evoked potentials. Electroencephalogr Clin Neurophysiol 1981 ;51: 226. Mailer AR: Evoked Potentials in Intraoperative Monitoring. Baltimore,
213-214. Williams & Wilkins, 1988, pp 78-79, 155-158.
196. Kritchevsky M, Wiederholt WC: Short-latency somatosensory evoked poten- 227. Nakanishi T, Tamaki M, Ozaki Y, et al Origins of short latency somatosensory
tials. Arch Neurol 1978;35:706-711. evoked potentials to median nerve stimulation. Electroencephalogr Clin Neuro-
197. Landgren S, Silfvenius H: Nucleus Z, the medullary relay in the projection path physiol 1983;56:74-85.
to the cerebral cortex of group I muscle afferents from the cat's hind limb. J 228. Namerow NS: Somatosensory evoked responses in multiple sclerosis patients
Physiol 1971;218:551-571. with varying sensory loss. Neurology 1968; 18:1197-1204.
198. Larinov VE: On the cortical centers of hearing in dogs. Obozr Psychiat Neurol 229. Netsky MG: Syringomyelia, a clinicopathologic study. AMA Arch Neurol
1897;2:419-424. Psychiatry 1953;70:741-777.
I 199. Larinov VE: The auditory tracts. Neurolog Vestr 1898;3:162-203. 230. Noel P, Desmedt JE: Somatosensory cerebral evoked potentials after vascular
200. Larinov VE: Galvanometric determination of cortical currents in the area of the lesions of the brain-stem and diencephalon. Brain 1975;98.T13^128.
tonal centers under stimulation of peripheral acoustic organs. Neurolog Vestr 231. Norrsell U, Voorhoeve P: Tactile pathways from the hindlimb to the cerebral
1899;7:44-64. cortex in cat. Acta Physiol Scand 1962;54:9-17.
201. Lastimosa ACB, Bass NH, Stanback K, et al: Lumbar spinal cord and early cor- 232. Nuwer MR: Recording electrode site nomenclature. J Clin Neurophysiol
tical evoked potentials after tibial nerve stimulation: effects of stature on nor- 1987;4:122-133.
mative data. Electroencephalogr Clin Neurophysiol 1982;54:499-507. 232a. Nuwer MR, Aminoff M, Desmedt J, et al: IFCN recommended standards for
202. Leadri M, Parodi CI, Zattoni J, et al: Subcortical and cortical responses follow- short latency somatosensory evoked potentials. Electroencephalogr Clin
ing infraorbital nerve stimulation in man. Electroencephalogr Clin Neurophysiol 1994;91:6-11.
Neurophysiol 1987;66:253-262. 233. Oken BS, Chiappa KH: Somatosensory evoked potentials in neurological diag-
203. Leblhuber F, Reisecker F, Boehm-Jurkovic H, et al: Diagnostic value of differ- nosis. In Cracco RQ, Bodis-Wollner I (eds): Evoked Potentials. New York, Alan
ent electrophysiologic tests in cervical disk prolapse. Neurology 1988;38: R Liss inc., 1986, pp 379-389.
1879-1881. 234. Oscarsson O, Rosen I: Projection to cerebral cortex of large muscle-spindle af-
204. Lesser RP, Leuders H, Hahn J, et al: Early somatosensory potentials evoked by ferents in forelimb nerves of the cat. J. Physiol 1963;169:924-945.
median nerve stimulation: Intraoperative monitoring. Neurology 1983;31: 235. Oscarsson O, Rosen I: Short-latency projections to the cat's cerebral cortex
1519-1523. from skin and muscle afferents in the contralateral forelimb. J Physiol 1966;
205. Lesser RP, Leuders H, Dinner DS, et al: The source of "paradoxical localiza- 182:164-184.
tions" of cortical evoked potentials to posterior tibial stimulation. Neurology 236. Pathak KS, Ammadio M, Kalamehi A, et al: Effects of halothane, enflurane and
1987;37:82-88. isoflurane on somatosensory evoked potentials during nitrous oxide anesthesia.
206. Li C, Houlden DA, Rowed DW: Somatosensory evoked potentials and neuro- Anesthesiol 1987;66:753-757.
logical grades as predictors of outcome in acute spinal cord injury. J Neurosurg 237. Pathak KS, Amaddio MD, Scoles PV, et al: Effects of halothane, enflurane, and
1990;72:600-609. isoflurane in nitrous oxide on multilevel somatosensory evoked potentials.
207. Lloyd DPC: Neuron pattern controlling transmission of ipsilateral hindlimb re- Anesthesiol 1989;70:207-212.
flexes in cat. J Neurophysiol 1943;6:293-315. 238. Pease WS, Fatehi MT, Johnson EW: Monopolar needle stimulation: Safety con-
208. Leuders H: The effects of aging on the wave form of the somatosensory cortical siderations. Arch Phys Med Rehabil 1989;70:412-^114.
evoked potential. Electroencephalogr Clin Neurophysiol 1970;29:450-460. 239. Pelosi L, Cracco JB, Cracco RQ: Conduction characteristics of somatosensory
209. Leuders H, Andrish J, Gurd A, Weiker Garry, Klem G: Origin of far-field sub- evoked potentials to peroneal, tibial, and sural nerve stimulation in man.
cortical potentials evoked by stimulation of the posterior tibial nerve. Electroencephalogr Clin Neurophysiol 1987;68:287-294.
Electroencephalogr Clin Neurophysiol 1981;52:336-344. 240. Perlik S, Fisher M, Patel DV, et al: On the usefulness of somatosensory evoked
210. Leuders H, Dinner DS, Lesser RP, et al: Origin of far-field subcortical evoked responses for the evaluation of lower back pain. Arch Neurol 1986;43:907-913.
potentials to posterior tibial and median nerve stimulation. A comparative study. 241. Peterson DO, Drummond JC, Todd MM: Effects of halothane, enflurane, isoflu-
Arch Neurol 1983;40:93-97. rane and nitrous oxide on somatosensory evoked potentials in humans.
211. Lueders H, Lesser R, Hahn J, et al: Subcortical somatosensory evoked poten- Anesthesiol 1986;65:35-40.
tials to median nerve stimulation. Brain 1983;106:341-372. 242. Pfeiffer FE, Peterson L, Daube JR: Diazepam improves recording of lumbar
212. Lueders H, Dinner DS, Lesser RP, et al: Evoked potentials in cortical localiza- and neck somatosensory evoked potentials. Muscle Nerve 1989;12:473-475.
tion. Clin Neurophysiol 1986;31:75-84. 243. Phillips CG, Powell TPS, Wiesendanger M: Projection from low-threshold
213. Madl C, Grimm G, Kramer L, et al: Early prediction of individual outcome after muscle afferents of hand and forearm to area 3a of baboon's cortex. J Physiol
cardiopulmonary resuscitation. Lancet 1993;341:855-858. 1971 ;217:419-^46.
244. Phillips LH, Daube JR: Lumbosacral spinal evoked potentials in humans. 275. Shaw NA: Far-field correlates of thalamic activity in cortical and cervical so-
Neurology 1980;30:1175-1183. matosensory evoked potentials. Clin Electroencephalogr 1988;19:219-224.
245. Pratt H, Starr A, Amlie RN, et al: Mechanically and electrically evoked so- 276. Shepherd GM: Neurobiology. New York, Oxford University Press, 1983, p 258.
matosensory evoked potentials in humans. Neurology 1979;29:1236-1244. 277. Sherman AL, Tirschwell DL, Micklesen PJ, et al: Somatosensory potentials, CSF
246. Pratt H, Politoske D, Starr A: Mechanically and electrically evoked somatosen- creatine kinase BB activity, and awakening after cardiac arrest. Neurology (in press).
sory potentials in humans: Effects of stimulus presentation. Electroencephalogr 278. Shibasaki H, Kakigi R, Ohnishi A, et al: Peripheral and central nerve conduc-
Clin Neurophysiol 1980;49:240-249. tion in subacute myelo-optico-neuropathy. Neurology 1982;32:1186-1189.
247. Pravdich-Neminsky WW: Einige electrische erscheinungen im zentralnerven- 279. Singh N, Sachdev KK, Brisman R: Trigeminal nerve stimulation: Short latency
system bei rana temporaria. Arch Anat Physiol 1913;321—324. somatosensory evoked potentials. Neurology 1982;32:97-101.
248. Pravdich-Neminsky VV: Experiments on the registration of the electrical phe- 280. Slimp JC, Tamas LB, Stolov WC, et al: Somatosensory evoked potentials after
nomena of the mammalian brain. Zbl Physiol 1913;27:951-960. removal of somatosensory cortex in man. Electroencephalogr Clin Neuro-
249. Riffel B, Stohr M: Spinal and subcortical somatosensory evoked potentials after physiol 1986;65:111-117.
stimulation of tibial nerve. Arch Psychiatr Nervenkr 1982;232:251-263. 281. Sonoo M, Shimpo T, Genba K, et al: Posterior cervical N13 in median nerve SEP
250. Roberts KB, Lawrence Pd, Eisen A: Dispersion of the somatosensory evoked po- has two components. Electroencephalogr Clin Neurophysiol 1990;77:28-38.
tential (SEP) in multiple sclerosis. IEEE Trans Biomed Eng 1983;30:360-364. 282. Spehlmann R: Evoked Potential Primer. Boston, Butterworth Publishers, 1985.
251. Robinson LR, Micklesen PJ, Lew H, et al: Predictive value of somatosensory 283. Spudis EV, Fullerton W, Fernandez H, et al: Somatosensory central latencies
evoked potentials (SEPs) for awakening from coma—a review of the literature, and disc discrimination in multiple sclerosis. Clin Electroencephalogr 1980;
(in preparation). 11:48-56.
252. Rodriquez AA, Kanis L, Rodriquez AA, et al: Somatosensory evoked potentials 284. Starr A, McKeon B, Skuse N, et al: Cerebral potentials evoked by muscle
from dermatomal stimulation as an indicator of L5 and SI radiculopathy. Arch stretch in man. Brain 1981;104:149-166.
Phys Med Rehabil 1987;68:366-368. 285. Stejskal L, Travnecek V, Sourek K, et al: Somatosensory evoked potentials in
253. Rosen I: Afferent connections to group I activated cells in the main cuneate nu- deep hypothermia. Appl Neurophysiol 1980;43:1-7.
cleus of the cat. J Physiol 1969;205:209-236. 286. Stewert JD: Focal Peripheral Neuropathies. New York, Elsevier, 1987.
254. Ross ED, Kirkpatrick JB, Latimosa CB: Position and vibration sensations: 287. Stohr M, Petruch F: Somatosensory evoked potentials following stimulation of
Functions of the dorsal spinocerebellar tracts. Ann Neurol 1979;5:171-176. the trigeminal nerve in man. J Neurol 1979;220:95-98.
255. Rossini PM, Cracco RQ, Cracco JB, et al: Short latency somatosensory evoked 288. Stohr M, Petruch F, Scheglmann K: Somatosensory evoked potentials following
potentials to peroneal nerve stimulation: Scalp topography and the effect of dif- trigeminal nerve stimulation in trigeminal neuralgia. Ann Neurol 1981;9:63-66.
ferent frequency filters. Electroencephalogr Clin Neurophysiol 1981;52:540-552. 289. Stohr M, Riffel B: Short-latency somatosensory evoked potentials to median
256. Rothstein TL, Thomas EM, Sumi SM: Predicting outcome in hypoxic-ischemic nerve stimulation: components N13-P13, N14-P14, PIS, P16, and P18 with dif-
coma. A prospective clinical and electrophysiologic study. Electroencephalogr ferent recording methods. J Neurol 1982;228:39^17.
Clin Neurophysiol 1991;79:101-107. 290. Strenge H, Hedderich J: Age-dependent changes in central somatosensory con-
257. Rumpl E, Prugger M, Gerstenbrand F, et al: Central somatosensory conduction duction time. Eur Neurol 1982;21:270-276.
time and short latency somatosensory evoked potentials in post-traumatic coma. 291. Sutton UN, Frewen JT, Marsh R, et al: The effects of deep barbiruate coma on
Electroencephalogr Clin Neurophysiol 1983;56:583-596. multimodality evoked potentials. J Neurosurg 1982;57:178-185.
258. Rumpl E, Prugger M, Battista HJ, et al: Short latency somatosensory evoked 292. Suzuki I, Mayanagi Y: Intracranial recording of short latency somatosensory
potentials and brain-stem auditory evoked potentials in coma due to CNS de- evoked potentials in man: Identification of origin of each component.
pressant drug poisoning. Preliminary observations. Electroencephalogr Clin Electroencephalogr Clin Neurophysiol 1984;59:286-296.
Neurophysiol 1988;70:482^189. 293. Symon L, Hargadin J, Zawirski M, et al: Central conduction time as an index of
259. Salzman SK, Beckman AL, Marks HG, et al: Effects of halothane on intraoper- ischemia in subarachnoid hemorrhage. J Neurol Sci 1979;44:95-103.
ative scalp-recorded somatosensory evoked potentials to posterior tibial nerve 294. Synek VM: Somatosensory evoked potentials from musculocutaneous nerve in
stimulation in man. Electroencephalogr Clin Neurophysiol 1986;65:36-45. diagnosis of brachial plexus injuries. J Neurol Sci 1983;61:443-452.
260. Samra SK, Vanderzant CW, Domer PA, et al: Differential effects of Isoflurane 295. Synek VM, Cowan JC: Somatosensory evoked potentials from stimulation of
on human median nerve somatosensory evoked potentials. Anesthesiol cutaneous femoris lateralis nerve and their application in meralgia parasthetica.
1987;66:29-35. Clin Electroencephalogr 1983;14:161-163.
261. Schiff JA, Cracco RQ, Rossini PM, et al: Spine and scalp somatosensory 296. Synek VM, Cowan JC: Saphenous nerve evoked potentials and the assessment
evoked potentials in normal subjects and patients with spinal cord disease: eval- of intraabdominal lesions of the femoral nerve. Muscle Nerve 1983;6:453-456.
uation of afferent transmission. Electroencephalogr Clin Neurophysiol 297. Synek VM: Somatosensory evoked potentials after stimulation of digital nerves
1984;59:374-387. in upper limbs: normative data. Electroencephalogr Clin Neurophysiol
262. Schimmel H: The (±) reference: Accuracy of estimated mean components in av- 1986;65:460-463.
erage response studies. Science 1967;157:92-94. 298. Taylor MJ, Borrets DS, Coles JC: The effects of profound hypothermia on the
263. Schimmel H, Rapin I, Cohen MM: Improving evoked response audiometry. cervical SEP in humans: Evidence of dual generators. Electroencephalogr Clin
Results of normative studies for machine scoring. Audiology 1975;14:466-479. Neurophysiol 1985;62:184-192.
264. Schimsheimer RJ, Ongerboer de Visser BW, Bour LJ, et al: Digital nerve so- 299. Therman PO: Transmission of impulses through the Burdach nucleus. J
matosensory evoked potentials and flexor carpi radialis H reflexes in cervical Neurophysiol 1941;4:153-166.
disc protrusion and involvement of the sixth or seventh cervical root: Relations 300. Thomas JE, Lambert EH: Ulnar nerve conduction velocity and H-reflex in in-
to clinical and myelographic findings. Electroencephalogr Clin Neurophysiol fants and children. J Appl Physiol 1960;15:1-9.
1988;70:313-324. 301. Tirschwell DL, Longstreth WT Jr, Rausch-Matthews ME, et al: Crebrospinal
265. Seaba P: Electrical safety. Arn J EEG Tech 1980;20:1-13. fluid creatine kinase BB isoenzyme activity and neurologic prognosis after car-
266. Sedgwich EM, El-Negamy E, Frankel H: Spinal cord potentials in traumatic diac arrest. Neurology 1997;48:352-357.
paraplegia and quadriplegia. J Neurol Neurosurg Psychiatry 1980;43:823-830. 302. Trojaborg W, Bottcher J, Santrup O: Evoked potentials and immunoglobulin ab-
267. Seyal M, Emerson RG, Pedley TA: Spinal and early scalp recorded components normalities in multiple sclerosis. Neurology 1981;36:866-871.
of the somatosensory evoked potential following stimulation of posterior tibial 303. Tsuji S, Shibasaki H, Kato M, et al: Subcortical, thalamic and cortical so-
nerve. Electroencephalogr Clin Neurophysiol 1983;55:320-330. matosensory evoked potentials to median nerve stimulation. Electro-
268. Seyal M, Gabor AJ: The human posterior tibial somatosensory evoked poten- encephalogr Clin Neurophysiol 1984;59:465-476.
tial: synapse dependent and synapse independent spinal components. Electro- 304. Tsumoto T, Hirose N, Nonaka S, et al: Analysis of somatosensory evoked po-
encephalogr Clin Neurophysiol 1985;62:323-331. tentials to lateral popliteal nerve stimulation in man. Electroencephalogr Ciin
269. Seyal M, Kraft LW: Synapse-dependent somatosensory evoked potential (SEP) Neurophysiol 1972;33:379-388.
to posterior tibial nerve (PTN) stimulation recorded from the cervical spine. 305. Tsumoto T, Hirose N, Nonaka S, et al: Cerebrovascular disease: Changes in so-
Neurology 1986;36(Suppl 1):82. matosensory evoked potentials associated with unilateral lesions. Electro-
270. Seyal M Gabor AJ: Generators of human spinal somatosensory evoked poten- encephalogr Clin Neurophysiol 1973;35:463-473.
tials. Journal of Clin Neurophysiol 1987;4:177-187. 306. Udenberg N: Differential localization in dorsal funiculus of fibers originating
271. Seyal M, Kraft LW, Gabor AJ: A cervical synapse dependent somatosensory from different receptors. Exp Brain Res 1968;4:367-376.
evoked potential following posterior tibial nerve stimulation. Neurology 307. van Buggenhout E., Ketelaer P, Carton H: Success and failure of evoked poten-
1987;37:1417-1421. tials in detecting clinical and subclinical lesions in multiple sclerosis patients.
272. Seyal M, Palma GA, Sandhu LS, et al: Spinal somatosensory evoked potentials Clin Neurol Neurosurg 1982;84:3-14.
following segmental sensory stimulation. A direct measure of dorsal root func- 308. Vas GA, Cracco JB, Cracco RQ: Scalp recorded short latency cortical and sub-
tion. Electroencephalogr Clin Neurophysiol 1988;69:390-393. cortical somatosensory evoked potentials to peroneal nerve stimulation.
273. Seyal M, Sandhu LS, Mack YP: Spinal segmental somatosensory evoked poten- Electroencephalogr Clin Neurophysiol 1981;32:1-8.
tials in lumbosacral radiculopathies. Neurology 1989;39:801-805. 309. Vera CL, Perot PL, Fountain EL: Scalp recorded somatosensory evoked poten-
274. Shagass C, Schwartz M: Age, personality, and somatosensory cerebral evoked tials to posterior tibial nerve stimulation in humans. Electroencephalogr Clin
responses. Science 1965;148:1359-1361. Neurophysiol 1983;56:159-168.
310. Vogel P, Vogel H: Somatosensory cortical potentials evoked by stimulation of 322. Yamada T, Kimura J, Nitz DM: Short latency somatosensory evoked potentials
leg nerves: analysis of normal values and variability; diagnostic significance. J following median nerve stimulation. Electroencephalogr Clin Neurophysiol
Neurol 1982;228:97-111. 1980;48:367-376.
311. Wainapel SF, Kim DJ, Ebel A: Conduction studies of the saphenous nerve in 323. Yamada T, Machida M, Kimura J: Far-field somatosensory evoked potentials
healthy subjects. Arch Phys Med Rehabil 1978;59:316-319. after stimulation of the tibial nerve. Neurology 1982;32:1151-1158.
312. Wang AD, Cone J, Symon L, et al: Somatosensory evoked potential monitor- 324. Yamada T, Kimura J, Wilkinson JT, et al: Short- and long-latency median so-
ing during the management of aneurysmal SAH. J Neurosurg 1984;60: matosensory evoked potentials. Findings in patients with localized neurological
264-268. lesions. Arch Neurol 1983;40:215-220.
313. Waxman SG: The structural basis for axonal conduction abnormalities in de- 325. Yamada T, Graff-Radford NR, Kimura J, et al: Topographical analysis of so-
myelinating diseases. Clinical uses of cerebral, brainstem and spinal somatosensory evoked potentials in patients with well localized thalamic infarc-
matosensory evoked potentials, Prog Clin Neurophysiol 1980;7:179-189. tions. J Neurol Sci 1985;68:31-46.
314. Waylonis GW, Kruger KC: Anal sphincter electromyography in adults. Arch 326. Yamada T, Kamayama S, Fuchigami Y, et al: Changes of short latency so-
Phys Med Rehabil 1970;51:409^12. matosensory evoked potential in sleep. Electroencephalogr Clin Neurophysiol
315. Whitsel BL, Petrucelli LM, Sapiro G: Modality representation in the lumbar 1988;70:126-136.
and cervical fasciculus gracilis of squirrel monkeys. Brain Res 1969; 15: 327. Yamada T: The anatomic and physiologic bases of median nerve somatosensory
67-78. evoked potentials. Neurol Clin 1988;4:705-733.
316. Wiederholt WC, Iragui-Madoz VJ: Far-field somatosensory potentials in the rat. 328. Yasuhara A, Yamada T, Seki Y, et al: Presence of two subcomponents in P9 far-
Electroencephalogr Clin Neurophysiol 1977;42:456-465. field potential following stimulation of the median nerve. Electroencephalogr
317. Wiederholt WC, Meyer-Hardting E, Budnick B, et al: Stimulating and record- Clin Neurophysiol 1990;77:93-100.
ing methods used in obtaining short-latency somatosensory evoked potentials 329. Yiannikas C, Shahani BT, Young RR: Short-latency somatosensory evoked po-
(SEPs) in patients with central and peripheral neurologic disorders. Ann N Y tentials from radial, median, ulnar and peroneal nerve stimulation in the assess-
Acad Sci 1982;388:349-358. ment of cervical spondylosis. Arch Neurol 1986;43:1264-1271.
318. Willis WD, Coggeshall RE: Sensory Mechanisms of the Spinal Cord. New 330. York DH: Somatosensory evoked potentials in man: differentiation of spinal
York, Plenum Press, 1978. pathways responsible for conduction from the forelimb vs hindlimb. Prog
319. Wolfe DE, Drummond JC: Differential effects of isoflurane/nitrous oxide on Neurobiol 1985;25:1-25.
posterior tibial somatosensory evoked responses of cortical and subcortical 331. Yu YL, Jones SJ: Somatosensory evoked potentials in cervical spondylosis.
origin. Ane.sth Analg 1988;67:852-859. Brain 1985;108:273-300.
320. Wong PKH, Bickford RG: Brainstem auditory evoked potentials: The use of 332. Zandbergen EG, deHaan RJ, Stoutenbeek CP, et al: Systematic review of early pre-
noise estimate. Electroencephalogr Clin Neurophysiol 1980;50:25-34. diction of poor outcome in anoxic ischemic coma. Lancet 1998;352:1808-1812.
321. Wood CC, Cohen D, Cuffin BN, et al: Electrical sources in human somatosen- 333. Zegers de Beyl D, Celberghe X, Herbaut AG, et al: The somatosensory central
sory cortex: Identification by combined magnetic and potential recordings. conduction time: physiological considerations and normative data. Electro-
Science 1985;227:1051-1053. encephalogr Clin Neurophysiol 1988;71:17-26.
!
Chapter 10
Magnetic Stimulation o f
t h e C e n t r a l and Peripheral
Nervous Systems
Lawrence R. Robinson, M.D.
CHAPTER OUTLINE
Magnetic Stimulation Magnetic Cortical Stimulation t o Assess Descending Motor

Historical Aspects • Physical Principles • Generation of the Pathways in Disease States • Intraoperative Monitoring
Magnetic Field • Advantages of Magnetic Stimulation • Coil • Mapping and Assessing Cortical Representation • Inhibitory
Geometries • Equipment Safety Phenomena
Cortical Stimulation Stimulation of the Peripheral Nervous System

Physiology • Facilitation • Descending Pathways • Recording Stimulation over the Spine and Nerve Roots
and Measuring MEPs after Magnetic Cortical Stimulation
• Calculation of Central Motor Conduction Times • Use of Safety Considerations
HISTORICAL ASPECTS OF was capable of supramaximal depolarization of the median nerve

at the wrist. Given the advantages of electrical stimulation for
MAGNETIC STIMULATION these types of studies, which was easier to use and quite satisfac-
Interest in using magnetic fields for examining patients and tory in its standard applications, magnetic stimulation using this
treating illness is at least several hundred years old. Most early early device did not develop any immediate clinical applicability.
attempts, however, used only relatively constant magnetic fields, The greatest technical improvement in magnetic stimulation
which do not effectively depolarize underlying tissue. The phys- came in the early 1980s when a unit that stimulated via a hand-
iologic effect of these constant fields is still poorly understood. held coil was developed. A separate large unit charged the ca-
8
Some of the earliest reports of using a time-varying magnetic pacitors and discharged a brief, large current (via a cable) to a 9
field (which can depolarize underlying nervous structures) cm diameter round coil; this could be held over virtually any part
come from Magnusson and Stevens who, in 1914, developed a of the body. Although it was first used to stimulate areas in the
large coil through which they fed alternating current. These96a
peripheral nervous system and muscle, the investigators soon
enthusiastic investigators placed their heads into the coils and learned how to effectively and relatively painlessly stimulate
reportedly saw phosphenes, the appearance of greenish lights. areas of the brain. Soon commercially available units were pro-
Although, initially interpreted as stimulation of the visual duced and subsequently approved by the U.S. Food and Drug
cortex, these phenomena were probably more related to dis- Administration (FDA) for use in the peripheral nervous system.
charges within the retina rather than excitation of the visual
cortex. Although these experiments were of some interest, there
were no real research or clinical applications at the time. PHYSICAL PRINCIPLES OF MAGNETIC
In the mid-1960s there were a number of experiments using a STIMULATION
15
small coil to depolarize nerves in the distal upper limb. Since this
coil was an integral part of the main charging unit, which was sub- Magnetic stimulation of the nervous system can only really
stantial in size and weight, it was not well-suited to stimulate areas occur in the setting of a rapidly changing magnetic field.
of the brain or proximal nervous system. Nevertheless, this unit Subjects exposed to a constant field strength (e.g., MRI units)
415
do not experience stimulation of nervous tissue. As described system compared to conventional methods of electrical stimula-
below, the intensity of the secondarily produced electrical field tion. What then is the advantage of magnetic stimulation? The
in nervous tissue (and hence the intensity of nervous stimula- principal advantage of using a magnetic stimulator lies in the
tion) is related to the rapidity of the change in magnetic field depth of penetration and its ability to penetrate intervening tis-
strength. sues regardless of electrical resistance.
Figure 10-1 demonstrates the theoretical drop in field inten-
GENERATION OF THE MAGNETIC FIELD sity with stimulation at the scalp, comparing electrical with
magnetic stimulation applied at the surface; field strength at
11
Formation of a brief magnetic pulse used for nervous system each depth is plotted as a percentage of surface intensity. The
stimulation starts within the main unit of the stimulator. A large electrical field drops off very rapidly with depth from the sur-
bank of heavy-duty capacitors (which incidentally contribute face of the scalp, in large part due to the electrical resistance of
most of the weight to the unit) are electrically charged. When the scalp, skull, and CSF; most current is shunted along the in-
triggered, these capacitors rapidly discharge through a cable tervening tissues before it ever reaches the brain. In contrast,
into a hand-held coil, producing a brief burst of very high cur- while the magnetic field also drops off* with distance, the rate is
rent (usually several thousand amperes). The current moving slower than for electrical stimulation. Moreover, for all practical
through the hand-held coil then produces a large (in the range of purposes, the type of tissue between the coil and the site of in-
1-3 tesla) time-varying magnetic field. One tesla (T), equivalent tended stimulation is irrelevant; the drop-off is essentially the
to 10,000 G (gauss), is much larger than the Earth's magnetic same for air, bone, fat, muscle, saline, etc. Thus, to stimulate
field, which is roughly 5 x 10~ T. The duration of the current
5
areas of the brain or other deep nervous tissues, it takes a com-
flow, and hence of the magnetic field, is very brief; it typically paratively lower surface field strength with magnetic stimula-
lasts for only about 50 microseconds, although some stimulation than for electrical stimulation, i.e., the penetration is far
tors have a decaying sinusoid wave with subsequent smaller better.
peaks. 98108
As a result of this brief magnetic field, a secondary As mentioned above, it should be kept in mind that the ab-
electric field is produced in nearby tissues or objects that allow solute field strength (in T) is not the most important feature for
current to pass; these in effect act as secondary coils. The stimulation of nervous tissue, but it is rather the rate of change
strength of the electric field produced by this stimulation is re- of the magnetic field. Thus, a 2-T stimulator is not necessarily
lated in part to the first derivative of the magnetic flux over time better than a 1-T unit and could be worse if it discharges its ca-
(dB/dt); the more rapidly changing the magnetic field, the pacitors more slowly. Pulse duration and shape as well as mea-
stronger the intensity of the generated secondary electric field surement of induced currents in secondary coils may provide
and the consequent nervous stimulation. The strength of the in- more useful information about the intensity of nervous stimula-
duced current is also proportional to the conductivity of the tion provided by various units.
medium in which it is generated. Thus, in cerebrospinal fluid a
relatively high current is generated, whereas in osseous struc- COIL GEOMETRIES
tures the current is negligible. The induced current has the re-
verse direction of the originating current in the coil. A number of different coil geometries have been used for
magnetic stimulation. Generally the tradeoffs in choice of coils
ADVANTAGES OF MAGNETIC STIMULATION involve intensity or depth of penetration vs. focality of stimula-
tion. Large diameter (e.g., 9 cm) coils produce greater intensity
It is considerably more complicated and expensive to produce fields that penetrate more deeply but stimulate over a wider area
a large rapidly changing magnetic field to stimulate the nervous and are less focal than small-diameter coils. Small diameter
(e.g., 2-cm) coils, in contrast, can stimulate in a more focal fash-
ion but do not penetrate very deeply or stimulate very intensely.
induced Electric Fields (% of Surface)
For some applications, the focality of stimulation is not criti-
% Surface intensity
60 j cal since the selectivity of the testing procedure is achieved on
the recording end of the system. When measuring conduction
time of the descending motor pathways supplying the intrinsic
muscles of the hand, for example, one can place recording elec-
trodes over the thenar or hypothenar muscles and stimulate over
a large area of motor cortex; although one might stimulate a
large area of cortex, only the events directed toward the hand
muscles are recorded. Additional selectivity also can be
achieved by directing facilitation (see below) to isolated groups
of muscles. In the above example, one can limit facilitation (a
small voluntary contraction) to just the thenar muscles and thus
limit the muscles activated.
20 40 60 80 100 There are, however, situations in which limitation of the area
Depth in mm of stimulation is critical. Mapping studies of the motor cortex
S H 8 Magnetic Stim 2I~J Electric Stim
require activating focal areas of the brain. Similarly, isolated pe-
ripheral nerve stimulation requires focal magnetic fields to
Figure 10-1. Theoretical drop off in magnetic and electric avoid activating the nerve over a wide area or to avoid activating
field strength at varying depths from the surface of the scalp. adjacent nerves.
Field strength is plotted as a percentage of that applied at the scalp Several different strategies have been directed toward limiting
surface. spread of current in these cases. One may change the orientation
Chapter 10 MAGNETIC STIMULATION OF THE CENTRAL A N D PERIPHERAL NERVOUS SYSTEMS — 417
of the coil, such as holding it up on one edge; although this may CM

reduce the field strength entering the tissues, it also limits the
spread of excitation. Another modification may take place in the i 10 12 14
coil shape. Butterfly or "figure-of-eight" coils usually have two
small coils that intersect in the middle of the hand held device
(Fig. 10-2). This point of intersection has a much higher current
density and magnetic flux than the surrounding areas and the
stimulation is strongest under this intersecting point. This modi-
fication, however, also drops the total stimulation intensity com-
pared with a single large round coil.
The site of stimulation on the coil has been extensively stud-
ied by several authors. 90-92
The highest intensity induced electric
fields are measured at the edges of the coil, with lower or no in-
tensities in the center*; intensity decreases with increasing dis-
tance from the edge (Fig. 10-2). Butterfly coils have the highest
intensities at the intersection of the two smaller coils. B
o
0
10 12 14
EQUIPMENT SAFETY 'JL.' >
As mentioned above, these devices release a very large cur-

rent with each stimulation; this is probably several times that re- SI *
quired to be lethal if one were exposed to the current directly. f t f •
tell t
Thus, in addition to safety concerns with respect to patients/sub- l&l
t t
1
jects (see below, cortical stimulation), one should avoid the pos-
sibility of injury from exposure to the currents normally
handled within the stimulator and coil. Since the capacitors
»t i i \ % •
store and release a very large charge for each stimulation, man-
ually handling the internal components of the device can be po-
tentially deadly. Most devices have locked cabinets that can be
opened only by the manufacturer. This is one situation in which I 10 12 14
it is truly wise to leave servicing to qualified personnel only.
Likewise, in the event of coil or cable breakage, it is unwise to
use the device before repairs are made, because of the possibil-
ity of releasing large currents.
CORTICAL STIMULATION
The most significant advance that has resulted from develop-
ment of the magnetic stimulator is the ability to noninvasively
stimulate the cerebral cortex in a relatively pain-free manner.
Figure 10-2. Various coil geometries and associated current
PHYSIOLOGY flow. Current Induced in a tank of isotonic saline by a tangentially ori-
ented round coil (A), an orthogonally oriented round coil (B), and a
Direct electrical stimulation of the exposed cerebral cortex tangentially oriented figure-of-eight coil (C). Length of arrow is p r o -
has been extensively studied in animal models for a number of portional t o the c u r r e n t strength. (From Maccabee PJ.Amassian VE,
years. Motor cortical physiology has been extensively studied in Eberle LP, et al: Measurement of the electric field induced into inho-
the baboon in the 1950s and 1960s and the interested reader is
114
mogenous volume c o n d u c t o r s by magnetic coils: Application t o

directed to appropriate reviews. Pertinent to the discussion of
1,5
human spinal neurogeometry. Electroencephalogr Clin Neurophysiol

magnetic stimulation are a number of physiologic findings from 1991 ;81:224-237, with permission.)
these and other studies.
Stimulation of the cortex with a single pulse has been shown
to produce more than just a single volley of descending corti- Although the first D wave may not bring the alpha motoneuron to
cospinal activation. Using a killed end recording from the pyra-
4
threshold, summation of subsequent I waves may then result in
midal tracts in baboons and electrically stimulating in the firing of the alpha motoneuron. While this summation usually re-
superficial layers of the cortex, a series of descending impulses sults in a single discharge, it has also been shown that the lower
has been observed following each stimulation. There is an ini- motor neuron may fire repeatedly after a sufficiently intense
tial D (direct) wave, which is followed by a series of I (indirect single cortical stimulation. As a consequence, amplitude of the
129
waves) coming at periodic intervals (usually separated by inter- CMAP after cortical stimulation can be larger than that produced
vals on the order of 1 ms) (Fig. 10-3). The later I waves are by corresponding supramaximal peripheral nerve stimulation.
likely generated by events occurring within the motor cortex. Several lines of evidence suggest that magnetic cortical stim-
As these multiple volleys descend down the corticospinal ulation does not activate the corticospinal neurons directly but it
tracts, they summate at the anterior horn cells in the spinal cord. may instead activate descending pathways via corticocortical
contractions, but does not appear to increase past about 10% of

maximal voluntary contraction (MVC).
Recent work suggests that there may be both specific facilita-
tion from target muscle contraction and nonspecific types of
facilitation from distant muscles, such as facial muscle contrac-
tion facilitating intrinsic hand muscle recording. It also has
6
been shown that activation of the contralateral muscle can be

used for facilitation. The obvious advantage is that the base-
150
line is not contaminated with EMG activity and that in case of

severe paresis or paralysis, facilitation is still possible. The
107
specific facilitation may have contributions from both spinal

(motor neuron) and supraspinal influences, whereas nonspecific
facilitation probably depends only upon supraspinal influences.
I MSEC *
DESCENDING PATHWAYS
I I i I I I I
Figure 10-3. Stimulation of the motor cortex with killed end The pathways these descending impulses take from the cortex
recording from pyramidal tracts in the baboon.A single cortical are not definitively known. There is, however, evidence suggest-
stimulation produces multiple discharges from the pyramidal tracts; ing that they travel via the corticospinal pathways. In cats, it has
the first direct wave (D) is followed by a series of periodic indirect (I) been found that the signal was strongest in the area of the spinal
waves. In the middle.trace, I waves are absent as a result of cortical de- cord near the corticospinal tracts and in the anterior horn cell
pression from anesthesia, but recover after anesthesia. (From Amassian area. - Lesioning studies have confirmed that most of the
81 82
VE, Steward M, Quirk GJ, et al: Physiological basis of motor effects of a signal travels in the area of the corticospinal tract. The conduc-
transient stimulus t o cerebral cor tex. Neurosurgery 1987;20:74-93, tion velocity measured in several studies (60-70 m/s) agrees
with permission.) with propagation through fast corticospinal tract axons.
RECORDING AND MEASURING MOTOR

connections. Studies recording from single motor units, measur- EVOKED POTENTIALS AFTER MAGNETIC
ing jitter between successive stimulations, have suggested that CORTICAL STIMULATION
there is at least one extra synapse at the cortical level in addition
to those at the anterior horn cell and the neuromuscular junction. Motor evoked potentials (MEPs), which typically involve
Comparing the technique with electrical cortical stimulation stimulating the motor cortex and recording caudally or periph-
(which is thought to activate neurons at the axon hillock), a erally, can be performed many different ways with respect to
higher jitter has been noted in single-fiber EMG (SFEMG) methods of stimulation and recording. Recording after motor-
recording during magnetic cortical stimulation (about 900 ps) cortex stimulation is usually performed over the target muscle
than during electrical stimulation (about 300 ps), consistent with rather than proximally over peripheral nerve or spinal cord; this
multiple s y n a p s e s . Findings of 1-2 ms longer latencies with
irM28
is in large part due to the relative size of the responses. While
magnetic stimulation compared to corresponding electrical stim- peripheral nerve responses can be recorded, as can spinal cord
ulation are also consistent with at least one extra synapse in the responses with epidural electrodes, they are much smaller than
pathway. Recent studies have indicated that the direction of the muscle responses and often require averaging multiple stimuli.
coil also may influence the site of activation. In conscious There are a number of different parameters that can be mea-
humans at threshold intensities, electric stimulation evokes D sured in MEPs recorded from muscle after motor cortex stimu- I
waves and magnetic stimulation (with a posteroanterior-induced lation. Most investigators agree that latency of the motor '
current) evokes I waves, whereas magnetic stimulation (with a response is the most useful and reliable parameter to measure.
lateromedial-induced current) evokes both activities. 44
Latency does need to be controlled for the level of facilitation
since a shortening of latency by several ms is seen when stimu-
FACILITATION lation is superimposed upon a small voluntary contraction.
Recording the response during both relaxation and facilitation
Facilitation is a well-described phenomenon with both mag- is thus often helpful.
netic and electrical cortical stimulation. When stimulation is su- Some authors report that amplitude measurements can be
perimposed upon a small voluntary contraction of the target useful - when measured as a percentage of the CMAP obtained
47 150
muscle (i.e., the muscle being recorded), there is both a shorten- after supramaximal peripheral nerve stimulation. The MEP/
ing of latency and increase in amplitude of the muscle (EMG) CMAP ratio can be regarded as the "efficacy" of the cortical stim-
response. This phenomenon is probably related to differing ulation. Since left and right normalized MEP amplitudes were
thresholds within the population of cortical neurons. Single shown to be strongly correlated, this can be used as a criterion of
motor unit recordings have demonstrated that the first units to abnormality in case of a unilateral low respones. Others, how-
150
be activated with magnetic cortical stimulation are also the first ever, report that amplitude measurements are too variable to be
ones to fire under voluntary control. These are slowly conduct-
68
clinically useful; they clearly depend greatly upon the level of fa-
ing fibers with small motor unit territories. During slight volun- cilitation and the site and intensity of stimulation.'29
tary contraction, it is postulated that the higher-threshold,

faster-conducting motor fibers (which usually have larger motor Calculation of Central Motor Conduction T i m e s
unit territories) are brought closer to firing threshold and thus fire Simply measuring the latency from cortical stimulation to the
during the stimulation. This effect is noted with small voluntary muscle response is useful, but if slowing is present, it does not
allow resolution of peripheral vs. central slowing. Thus, meth- Table 10-1. Mean Values of Magnetic Cortical Stimulation
ods have been derived to subtract peripheral conduction time with Ipsilateral and Contralateral Facilitation of the Target
from the total scalp-to-muscle latency. Muscle (n = 36)
One method is to stimulate with a second magnetic or elec-
Ipsilateral Contralateral
trical impulse over the cervical or lumbar spine. This probably
Activation (SD) Activation (SD)
activates roots distal to the neural foramen (see root stimula-
tion below). Needle stimulation is usually required to activate Latency (ms)
the roots when using electrical impulses for stimulation. Cortex-APB 18.9 (1.23) 20.6(1.20)
Another method uses F-waves elicited from electrical stimula- CMCT-APB 5.7 (0.90) 7.4 (0.87)
tion at the wrist or ankle to calculate peripheral conduction Cortex-TA 25.7 (2.0) 27.8 (2.1)
t i m e . While F-waves are probably better at considering the
120
CMCT-TA 14.7 (0.87) 14.9 (0.61)
entire length of the lower motor neuron, they have the disad- Normalized amplitude (MEP/M-wave)
vantage of being slowed in some central processes (e.g., sy- APB 0.39(0.13) 0.36 (0.19)
ringomyelia); thus they could lead to an underestimation of TA 0.56 (0.28) 0.30 (0.20)
central conduction times. On the other hand, using magnetic or
CMCT, central motor conduction time; SD, standard deviation; APB, abductor
electrical root stimulation in the calculation adds a small pe-
pollicis brevis;TA, tibialis anterior.
ripheral (proximal) conduction time to the central conduction. Adapted from Zwarts MJ: Central motor conduction in relation to contra- and
Subjects find magnetic stimulation of nerve roots or F-wave iipsilateral activation. Electroencephalogr Clin Neurophysiol 1992;85:425-429.
measurements more tolerable than needle stimulation of nerve
roots, although intraindividual trial-to-trial variability of cen-
tral motor conduction time is similar for all methods, with co- Sites of stimulation have been well described based on map-
efficients of variation of 13% for the F-wave latency, 15% for ping s t u d i e s - using surface stimulation. Typically, distal
3233 36
cervical magnetic stimulation, and 1 1 % for cervical needle upper limb stimulation is best at sites about 7 cm lateral to C z
stimulation.130
(the intersection of the nasion-inion and tragus-tragus lines), the
Independent of how the calculation is performed, subtraction proximal upper limb at 5 cm lateral to C and the lower limb at
z
of peripheral conduction time from the total scalp to muscle la- the midline (C ). Optimal orientation of the coil depends on the
z
tency yields the central motor conduction time (CMCT). This type of stimulator used and the current waveform going through
| parameter is most commonly used for reporting results of motor the coil; flat orientations typically produce more robust stimula-
cortex stimulation in healthy subjects and controls. Tables 10-1 tion, while tangential orientations produce more focal stimula-
and 10-2 provide some reference values from the literature for tion. It is known that posterior-to-anterior induced currents in
different target muscles and different activation procedures. It the hemisphere preferentially induce motor responses. Thus, it 68
should be realized that the CMCT to the leg muscles is signifi- is important with monophasic pulses to be aware of the position
cantly dependent on body height. of the side of the coil when using a flat orientation. Because of
the influence of facilitation on latency and amplitude, the level
Practical Issues Involved in Performing of facilitation should be kept constant; this can also be mea-
Magnetic Cortical Stimulation and sured with small hand-held dynamometers.
Recording Motor Evoked Potentials As mentioned above, it is usually most efficient to record di-
There are a number of practical issues involved in performing rectly from muscle with surface electrodes. Standard recording
motor cortex stimulation. First, one must obtain the necessary settings, similar to those used for motor conduction studies, are
approvals prior to performing the procedure. As of this writing, employed; however, when stimulus artifact is a problem, one
use of magnetic stimulation for cortical activation is considered
experimental. Prior to starting the procedure, one must obtain
approval from the applicable local human subjects review board Table 10-2. Reference MEP Data of 150 Control Subjects
or institutional review board. Aged 20-83 Years
There are a number of exclusion criteria that should be evalu- Muscle Latency (ms) MEP/Mx 100 (%) CMCT (ms)
ated before proceeding with magnetic stimulation. Any im-
planted or nearby electronic devices can be damaged by the Biceps l l . 8 ± 1.2 41.0 ± 18.7 6.1 ± 1.3
magnetic field produced. Thus subjects (or examiners) with (N = 49) (9.1-14.7) (21.3-108) (4.3-8.4)
pacemakers, cochlear implants, or other implanted electronic de- EDC 15.2 ± 1.5 3 7 . 2 1 22.1 6.4 ± 1.2
vices should not be near the magnetic coil. Metallic devices, (N = 42) (12.2-18.4) (24.5-110) (4.1-8.6)
such as aneurysm clips, are usually considered a contraindica-
tion because of the possibility of movement. There are reports of Thenar 20.4 ± 1.5 46.1 ± 2 3 . 5 6.7 ± 1.2
magnetic cortical stimulation producing seizures in seizure- (N = 95) (16.8-23.8) (27.8-113.4) (4.9-8.8)
prone individuals; those with a risk or history of seizures should TA 27.7 ± 2.4 34.9 ± 19.7 13.1 ±3,8*
not have cortical stimulation (see section on safety below). (N = 83) (20.2-32.5) (19.3-87.6) (10.1-16.3)
While kindling of a new seizure focus is a concern with high
Amplitudes measured peak to peak. CMCT, central motor conduction time;TA,
rates of stimulation, kindling has not been induced in animals tibialis anterior; EDC, extensor digitorum communis.Values are mean ± SD
with stimulation frequencies below 10 Hz, despite prolonged du- (range). MEP: motor evoked potentials. M: compound muscle action potential;
rations. Although one study demonstrated a high frequency of CMCT: central motor conduction time.
hearing loss in rabbits exposed to repeated stimuli, audiologic
37
'Measured using F response. Mean ages (ranges) of subjects, biceps: 49.5 ± 18.5
testing in humans has failed to corroborate these findings. 112 (20-83); EDC: 54.5 ± 16.9 (20-83); thenar: 44.6 ± 16.8 (20-83);TA: 37.9 ± 15.0
(20-76).
Nevertheless, many investigators use hearing protectors during Adapted from Eisen AA, Shtybel W: Clinical experience with trancranial mag-
stimulation. netic stimulation.Muscle Nerve 1990;13:995-101 I.
may need to adjust the low-frequency filter accordingly and/or Spinal Cord injury
consider the use of shielded recording cables. After complete spinal cord injury (SCI), usually no re-
sponse can be obtained from muscles below the level of the
USE OF MAGNETIC CORTICAL STIMULATION lesion, while normal responses can be obtained above. One
TO ASSESS DESCENDING MOTOR PATHWAYS group has reported very prolonged motor responses from the
58
IN DISEASE STATES abductor pollicis brevis in a few clinically complete C5 or C6

lesions, suggesting that subclinical sparing of motor tracts
Multiple Sclerosis may occur in some patients. This finding, however, is not com-
monly reported.
There are a number of disease states in which magnetic corti- MEPs and SEPs have been studied after inducing spinal cord
cal stimulation has been applied. Multiple sclerosis (MS), a de- injury in cats. The MEP recorded below the lesion was the
82
myelinating disease of the central nervous system, is a condition most sensitive indicator of injury. On follow-up, the MEP re-
in which slowing of central motor conduction time (CMCT) is sponse always returned before the animals started ambulating,
commonly reported. A number of-studies have documented pro- although this was often only 1-2 days before. No cats without a
longed CMCTs in patients with MS, consistent with demyelina- response started to ambulate again.
tion of motor tracts. - - - - - Some evidence suggests that
13 38 66 67 100 134
Perhaps the most useful application of MEPs in patients with
MEPs are more sensitive than SEPs, with an overall incidence SCI is in detecting new neurologic compromise. There are data
of abnormality of slightly more than 70%. In a study of 83 pa- suggesting that MEPs reflect the degree of descending motor
tients with definite or probable MS recording the response from tract impairment in posttraumatic syringomyelia; patients with
86
abductor digiti minimi (ADM), responses were of very pro- weakness or progressive neurologic deficit usually show a pro-
longed latency; amplitudes were often reduced as well. e7
longation of CMCTs. We have also noted improvement in
Abnormalities in MEPs correlated well with brisk finger flexor CMCTs after successful decompression of posttraumatic
jerks. The incidence of electrophysiologic abnormalities com- syrinx, - - although this finding has not been universally re-
86 87 121
paring the various evoked potentials was: MEPs 7 2 % , VEPs ported. Mapping studies (see below) also have been of interest
97
67%, SSEPs 59%, and BAERs 39%. In comparison, MRI of the in demonstrating changes in the motor cortex after spinal cord
brain has sensitivities of 80% or more. injury.
There is also some evidence that the degree of abnormality
on central conduction times is associated with the degree of dis- Stroke
ability. In a study of 45 patients with multiple sclerosis, the pro- A number of studies have looked at motor evoked potentials
longation in C M C T (from head to cervical region) correlated in patients after stroke, primarily to evaluate prognosis. In a
with the level of disability. 9
study of 20 patients with chronic hemiparesis or hemiplegia, on
Others have also found a good correlation between disabil-
74
the affected side, 15 patients had no response in either the APB
ity and CMCTs. Moreover, they found that patients who im- or biceps brachii, 2 patients had an absent response in one
proved on steroids had a concomitant shortening in CMCT, muscle, and the remaining 3 patients had delayed responses. 14
while those who had no improvement had stable CMCTs. There was no clear correlation between electrophysiologic ab-
normalities and clinical deficits; no responses were recorded in
Cervical Spondylosis some patients who had voluntary movement.
Cervical spondylosis is a very common cause of myelopathy In a study of 19 patients with recently completed stroke,
in the elderly. - - - While imaging modalities (CT and MRI)
I9 20 77 79
delays in CMCT were found predominantly in subcortical le-
can detect cervical spondylosis and compression of the spinal sions, whereas those with severe cortical strokes were more
cord, these changes may be seen only in the presence of a severe likely to have absent MEP responses. It was also found that
93
neurologic deficit. - 78 79
patients with preserved MEPs had smaller infarcts on CT and |
MEPs recording from the thenar and the tibialis anterior that MEPs had a slightly better predictive value for functional |
muscle appear especially sensitive for detecting myelopathy prognosis than SEPs. Cortical MEPs were present in 9 of 10 pa-
secondary to cervical spondylosis, whereas patients with 1
tients who made some degree of functional recovery, whereas
radiculopathy alone are generally reported to have normal they were absent in 8 of 9 patients who made no recovery or
MEPs. Other muscles also can be useful in determining which died.
spinal cord levels are most affected. Some evidence suggests 27
Another study has evaluated 33 patients within 3 days after
that MEPs may be more sensitive than SEPs, possibly because stroke. Two months later, those with present or prolonged
45
cervical spondylosis (often with prominent bony spurs project- MEPs had improved motor function compared to those without
ing from the vertebral bodies) predominantly involves the an- responses. MEPs may be somewhat predictive of recovery from
terolateral quadrants of the spinal cord. This could potentially dysphagia, in that those with larger increased pharyngeal repre-
affect descending motor tracts in the corticospinal tracts, leav- sentation in the unaffected hemisphere have a better recovery,
ing dorsal column pathways relatively unaffected. suggesting a role for the intact hemisphere reorganization in re-
In one report of 67 patients with cervical spondylosis, cover}'. The largest study on 118 first-ever stroke patients stud-
62
CMCTs were prolonged in 84% of patients with radiologic evi- ied the natural history of MEP changes following stroke and the
dence of cord compression and in 22% of those without. These 96
predictive value of MEP responses. - It was shown that the
64 65
MEP abnormalities correlated well with upper motor neuron presence (delayed or normal) or absence of a response follow-
signs. Postoperatively, however, no improvement in MEPs oc- ing magnetic cortex stimulation on day 1 had a strong correla-
curred. In another study of subjects with asymptomatic cervical tion with outcome after 12 months. Thus, clear prognostic
spondylosis, the majority of such patients (92%) had normal information can be obtained from MEPs in patients with stroke,
MEPs, suggesting that in the absence of clinical manifestations even in a very early stage. It is not yet clear how this compares
MEPs are likely to be normal. 138
with information from clinical assessment alone.
C h a p t e r 10 MAGNETIC STIMULATION OF THE CENTRAL A N D PERIPHERAL NERVOUS SYSTEMS — 421
Motor Neuron Disease Nevertheless, there are some reports of magnetically acti-
As would be expected, MEPs are abnormal in motor neuron vated MEPs being successfully used for monitoring of invasive
disease. Abnormalities include absence of a response, pro-
47103
procedures. There is one report, for example, of MEPs recorded
longation of latency, and, most commonly, low-amplitude motor from the distal limb muscles in response to transcranial magne-
responses. Slowing is thought to reflect dropout of faster-con- toelectrical stimulation in 10 patients during embolization of ar-
ducting axons, rather than demyelination, whereas amplitude teriovenous malformations. Stable potentials in 8 of 10
123
changes may reflect either loss of anterior horn cells, cortico- patients coincided with an uneventful neurologic outcome,
motorneurons or both. while prolongation of the CMCT in 2 patients coincided with
MEPs may be relatively sensitive for detecting upper motor transient hemiparesis, probably caused by ischemia.
neuron abnormalities in patients with motor neuron disease. For At this point it remains too early to know if magnetic stimula-
instance, one group found that MEPs revealed evidence of tion of the motor cortex will be as reliable as other techniques
upper motor neuron dysfunction in 84 of 121 (69%) patients for intraoperative assessment and whether technical difficulties,
with motor neuron disease, including unsuspected upper motor particularly if effects of anesthesia, can be overcome.
neuron involvement in 6 of 22 (27%) patients who had purely
lower motor neuron syndromes clinically. Increased MEP MAPPING AND ASSESSMENT OF
140
threshold was the abnormality observed most frequently. CORTICAL REPRESENTATION

Similar findings have been reported by others. 104
Measuring the areas representing various muscles or move-

INTRAOPERATIVE MONITORING ments can be of interest both for the study of cortical physiol-
ogy in normal subjects and for examining plasticity and other
The idea of using motor cortex stimulation during spinal processes after neurologic perturbations.
surgery is conceptually appealing for several reasons. First, the In animals the areas of motor cortex corresponding to various
most functionally significant impairments after intraoperative movements or muscle contractions have been studied in a vari-
iatrogenic myelopathies are probably related to the resulting ety of species. Typically, electrical pulses have been applied to
34
motor deficits; thus it would be useful to monitor these descend- the exposed cerebral cortex and the resulting movements either
ing motor tracts directly. Second, due to the anatomy of the directly visually observed or recorded via EMG. While E M G
spinal cord blood supply, it is at least theoretically possible to recording is often used to record single muscle activity after
infarct the anterior two thirds of the spinal cord (via compro- cortical stimulation, it should be noted that, at the cortical level,
' mise of the anterior spinal artery) and preserve SEPs that ascend the organization is more "movement-specific" rather than
through the posterior columns. muscle-specific; thus there are probably multiple overlapping
Levy et al. were among the first to report on using MEPs to areas of the cortex representing each muscle. Nevertheless, be-
71
monitor the descending motor tracts during surgical procedures. cause of ease of recording and quantification, EMG recording
They reported on 98 cases of MEP monitoring with electrical of individual muscles is still used.
stimulation during neurosurgical procedures. - They found 80 83
Despite the fact that mapping studies had been performed in
! that the peripheral nerve response during monitoring was much multiple animal models over the last 100 years, mapping of the
more sensitive to injury than the spinal cord response. Rever- human cortex had been generally limited to stimulation of the
sible loss of the peripheral nerve signal was not associated with exposed (usually abnormal) brain during surgical procedures. It
a postoperative deficit, suggesting that MEPs could be of value has only been through the advent of surface stimulation, at first
for intraoperative monitoring. electrical and now magnetic, that investigators could noninva-
One technical problem with using MEPs in the operating room, s i v e ^ map the motor cortex both in healthy individuals and in
however, has been the effects of anesthesia. Recent studies 46
those with diseases or lesions. Surface electrical stimulation
demonstrated a 60% reduction in MEP amplitude to the abductor was used initially and potentially offers the advantage of more
digiti minimi with the administration of midazolam and fentanyl. focal stimulation than magnetic coils. However, with the devel-
Another group has reported an average reduction of MEP am-
148
opment of more focal coils (e.g., the butterfly coil), the rela-
plitudes to about 10% of preoperative values during anesthesia tively painless technique of magnetic stimulation has become
with nitrous oxide and fentanyl. Normal volunteers breathing ni- widespread.
trous oxide alone had a similar marked reduction. Intravenous nar- There are two alternative approaches to mapping the cortex
cotics (fentanyl, flunitrazepam, and thiopental) have had more with magnetic coil stimulation. - One technique maps out
32 61
moderate reductions in amplitude (30-50% reduction), but it is not stimulation thresholds, i.e., the center of the "map' for a given
1
practical to use these agents alone for the majority of neurosurgi- muscle has the lowest threshold for activation (usually mea-
cal or orthopedic cases; doing so carries the risk of requiring pro- sured in percent of maximum output of the stimulator), with in-
longed ventilatory support. In monkeys, use of etomidate has creasing thresholds as one moves out away from the center.
permitted MEPs to be recorded under anesthesia, but latencies are Because it can be difficult and time-consuming to measure
significantly prolonged (up to 10%) and amplitudes reduced (up-to thresholds, this has not been the method of choice. The alterna-
60%), making it of questionable value for sensitive monitoring. 56
tive method uses a standard stimulus intensity somewhat above
Enflurane or isoflurane alone also markedly reduces MEP ampli- threshold and maps the amplitude of the EMG muscle response
tudes, even in subanesthetic concentrations. 26135
with stimulation over various sites; thus the center of the "map"
These problems largely do not exist with electrical, as opposed has the largest-amplitude EMG response. The second method is
to magnetic, stimulation of the brain or spinal cord. Some au- 141
quicker and more commonly used, but it is not yet clear how the
thors have reported using electrically activated motor evoked po- two methods compare.
tentials intraoperatively; one group, has reported recording
149
Results of mapping the motor cortex in normal subjects have
from spinal cord and cauda equina. Of 40 patients, there were re- been reported - and are similar to those obtained with stimula-
32 33
portedly no false negatives, 3 true positives and 5 false positives. tion of the exposed cortex. For the upper limb, distal muscles
116145
are represented more laterally than proximal muscles, e.g., the tourniquet are mapped, they demonstrate an increase in MEP
site for activating intrinsic hand muscles is about 7 cm lateral to amplitude within minutes. The time frame of these findings
the midline, while the biceps brachii is more medial, about 5 cm again raises the possibility of unmasking of established path-
lateral to midline. Lower limb muscles are represented closer to ways. The duration of these changes in amputees has been re-
the midline (near C ) .z ported to last more than 20 years post-amputation. 124
In general, responses are recorded only contralaterally; however, Even in normal subjects, some changes in motor maps can be
occasionally for proximal upper limb muscles we have recorded ip- induced. Focal transcranial magnetic stimulation has been used
silateral responses, possibly from uncrossed corticospinal path- to examine the effects of 120 synchronized thumb and foot
ways (unpublished observations with Dr. Greg Malloy). movements on the motor output map of the right abductor polli-
Perhaps the most interesting capability provided by mapping cis brevis (APB) muscle. The center of gravity (CoG) of the
84
techniques is the opportunity to noninvasively examine changes output map moved medially in the direction of the foot repre-
in cortical representation after a perturbation, i.e., neuroplastic- sentation area (mean movement 7 mm ) and returned to its orig-
ity. This has been examined in a number of circumstances. inal location within 1 hour after the experiment.
After spinal cord injury (SCI), muscles immediately rostral to
the level of the lesion have been examined and mapped. * INHIBITORY PHENOMENA
36 136
These investigations have demonstrated plastic changes in the

motor cortex after SCI; muscles rostral to the lesion (e.g., biceps While most of the preceding discussion has focused on exci-
brachii in C6 quadriplegia) have enlarged territories compared tation of muscles resulting from stimulation of the motor cortex,
to control subjects. These changes may be seen within days there are a number of interesting inhibitory phenomena reported
after onset of SCI, raising the possibility that they represent un- as well. These events have been produced from stimulation of
masking of established pathways rather than development of the motor areas of the cortex as well as other remote areas (e.g.,
new anatomic connections. It is unclear what implications these occipital cortex).
changes have for treatment or spontaneous recovery of function. Silent periods induced by stimulation of the motor cortex prob-
Patients with peripheral nervous system lesions are reported ably represent one type of inhibitory phenomena. Silent periods
to have changes in cortical representation as well. Patients with are easily observed when a subject exerts a maximal voluntary
prior polio, for example, have smaller cortical areas in muscles contraction in a target muscle (e.g., APB) and a magnetic stimulus
affected by polio compared with the analagous muscles in the is applied over the contralateral motor cortex; one will record a
contralateral limb. 110
silent period during which the subject cannot continue or restart
Similar changes are reported in patients with amputations. In the contraction for up to a few hundred ms (Fig. 10-4). The silent
subjects with traumatic, surgical, or congenital amputations, period varies from one muscle to another but can last for up to 300
muscles just proximal to the amputation have enlarged maps. 35
ms in the thenar muscles; the duration is proportional to the size
122
Some of these changes probably occur very early. Brasil-Neto of cortical representation (e.g., longer in hand muscles than more
and colleagues have cleverly mimicked the early changes
21
proximal muscles), and to the intensity of magnetic stimulation.
after amputation by using an inflatable double tourniquet, which Silent periods may be altered in some disease states. They
produces anesthesia distally. When muscles just proximal to the have been reported to be prolonged in patients with hemiparesis
and shortened in those with Parkinson's disease. Shortened
63
ECS silent periods have been found after stroke, ALS, and cervical
Stlm.CMAP SlUnt P e r i o d myelopathy. 143
Inhibitory effects produced from stimulating areas of the

cortex corresponding to language have also been observed. Using
special prototypical stimulators, application of rapid (25 Hz)
magnetic stimulation to the dominant side produced temporary i
speech arrest. These findings are of special interest because of '
111
the potential ability to quickly determine in a noninvasive way

which side of the brain is dominant for language.
Magnetic stimuli applied over the occipital cortex can cause a
temporary loss of information coming in through the visual
system, although it is unclear whether this represents activation
of inhibitory pathways or simply a scrambling of stored infor-
mation. When healthy subjects are presented with three letters
(a trigram) for 14 ms on a computer screen, they can remember
and repeat these three letters with essentially perfect accuracy.
However, when a magnetic stimulation is applied over the oc-
cipital cortex 40 to 120 ms after the trigram is presented, there
Figure 10-4. Silent periods recorded from the abductor pol- is significant impairment in the identification of the letters,
12
licis brevis.The subject is asked t o sustain a maximal voluntary con- i.e., the letters are "forgotten."
traction during which a magnetic stimulation is applied t o the
contralateral m o t o r cortex. Despite repeated trials, silent periods of
up t o 300 ms a r e reliably produced. ECS, electromagnetic cortical STIMULATION OF THE PERIPHERAL
stimulation; CMAP, compound muscle action potential. (From NERVOUS SYSTEM
Robinson LR, Goldstein BS, Little JW: Silent periods after electromag-
netic stimulation of t h e m o t o r c o r t e x . Am J Phys Med Rehabil When first developed, magnetic stimulators were thought to
1993;72:23-28, with permission.) hold promise for use in the peripheral nervous system (PNS) as
well as the brain. The potential advantages were thought similar have been no successful reports of activating the cervical spinal
to those applicable to cortical stimulation: deeper penetration of cord with magnetic stimulation, i.e., subjects do not demonstrate
stimulation and less painful stimulation. However, in contrast to lower limb movement or report upper limb paresthesias.
cortical stimulation, there are already very well established, When applying magnetic stimulation over the cervical spine,
broadly accepted methods for stimulating peripheral nerves, depolarization probably occurs at the root distal to the interver-
namely electrical stimulation. Despite initial hopes, magnetic tebral foramen rather than within the spinal canal itself. - 24 52109
stimulation has demonstrated a number of disadvantages that There are several lines of evidence supporting this hypothesis:
have made it far less useful in the peripheral nervous system (1) comparison with direct electrical root stimulation suggests
than initially thought. that the point of stimulation (in the cervical area) is about 7 cm
Perhaps the biggest problem with the magnetic coil is the distal to the intervertebral foramen; (2) comparison with F- 115
lack of focality of the stimulation. As opposed to many applica- wave latencies suggests the depolarization point may be about 7
tions of cortical stimulation, peripheral nerve conduction stud- cm from the anterior horn cell; and (3) modeling of current 24
ies require knowing the precise site of stimulation as it is critical flows from magnetic stimulation over the spine have shown that
to accurate measurement of latency and conduction velocities. the greatest current densities are in the area of the intervertebral
A number of studies have documented that several centimeters foramen (Fig. 10-5). - - The latencies of electrical and mag-
49 89 91
of nerve length near the coil are usually depolarized and that the netic root stimulation were not significantly different, suggest-
site of stimulation is variable from one stimulation to the ing an identical excitation site. Since there is no clear shift in
next. 108109
Moreover, the site of stimulation for different fiber latency with higher magnetic stimulus strength, a preferential
populations may differ with magnetic stimulation; larger fibers site of excitation is postulated, probably at the neurofora-
are probably stimulated at a greater distance from the coil than m e5o.53.i42 j
n j b
tn a sshown that even with low magnetic
a s o e e n
smaller fibers, producing a greater degree of temporal disper- stimulus strengths, it is possible to excite sensory root fibers and
sion than with corresponding electrical stimulation. While 40
record antidromic sensory potentials with ring electrodes at the
changing the orientation, shape, or size of the coil may reduce fingers. 151
the area of stimulation or its variability, the precision obtained For magnetic stimulation of the roots to become well ac-
with this technique is still not nearly as good as for its electrical cepted, it will need to be shown to be as good as, or superior to,
counterpart. 109
electrical root stimulation; the latter has been well described
Not only are latencies and velocities difficult to accurately and is clinically useful in the evaluation of plexopathies and
determine with magnetic stimulation, it is also difficult or im- proximal nerve lesions. While electrical stimulation requires
95
possible to supramaximally stimulate a single peripheral nerve needle insertion and sometimes high (uncomfortable) voltages,
without simultaneously activating nearby nerves. - Thus, am-
28 51
plitudes of peripheral nerve responses have been less than those

with electrical stimulation and volume conduction to other
nerves has been frequently reported.
Nevertheless, there remain some potentially useful applica-
tions of magnetic stimulation of peripheral nervous system that
are still being investigated. Phrenic nerve conduction studies
usually require high levels of electrical stimulation and may be
more easily performed with magnetic stimulation. Facial 88
nerve studies also may be performed with magnetic stimula-

tion. With surprisingly low stimulus strength it is possible to
146
stimulate the peripheral part of the facial nerve within the skull
by positioning the coil over the parieto-occipital area. - * 57 59125 126133
It is also possible to stimulate the motor cortex and measure the

central conduction to the facial muscles. Recording from the
nasalis muscle usually gives a clear, biphasic C M A P . With 125
magnetic stimulation the facial nerve is probably excited at the

transition of cerebrospinal fluid to bone in the facial canal (the
labyrinthine segment). Comparing the latency with electrical
stimulation at the stylomastoid foramen results in a transosseal
conduction time (average 1.3, SD ± 0.15 ms). This technique
125
provides the opportunity to asses the localization of an acute

conduction block of the facial nerve such as in Bell's palsy. It
has been shown that an absent response at the first day predicts
-40 4 1 H « 1—~+—-—i . t—
a poor recovery. " It is also possible to differentiate an in-
118 9
mm 0 4 8 12 16 20 24 28 32
franuclear from a supranuclear facial nerve lesion. Magnetic144
stimulation to other cranial nerves and muscles is also feasible, Figure 10-5. Voltages measured across neural foramen at
for example the lingual muscles. 105
the T2-T3 level during magnetic coil stimulation.The highest
voltages are measured within the foramen (point H).The lower trace
STIMULATION OVER THE SPINE AND NERVE ROOTS d e m o n s t r a t e s the first spatial derivative of t h e electric field. (From
Macabee PJ.Amassian VE, Eberle LP, et al: Measurement of the electric
A number of studies have examined the possibility of stimulat- field induced into inhomogenous volume c o n d u c t o r s by magnetic
ing over the spine using magnetic stimulation in order to evaluate coils: Application t o human spinal neurogeometry. Electroencephalogr
the roots and proximal P N S . To this author's knowledge, there
109
Clin Neurophysiol 1991 ;81:224~237, with permission.)
it has the advantages of producing supramaximal stimulation. before and within 30 minutes after magnetic cortical stimulation
The advantage of magnetic stimulation is that it is relatively (mean of 35 stimuli), there was no change in EEG, verbal
painless. When the amplitude of the respons is not an impor-
131
memory, language fluency, visual memory, attention, or grip
tant item, magnetic stimulation seems preferable and can show strength. * While one group has reported that maximal mag-
22 23
focal conduction slowing in the proximal parts of the peripheral netic stimulation can produce a mild change (10%) in short-term
nervous system. It is already an acknowledged technique for memory, it is the same with stimulation over the cortex or cervi-
measuring the peripheral conduction time in order to calculate cal spine, suggesting this might not be a cortical effect. Rapid 54
the conduction in central motor pathways. Several reports 13

rate transcranial magnetic stimulation (stimulation frequency > 1
have shown delayed responses in lumbar root disease and plex- Hz) can induce significant side effects, such as temporary hear-
opathy. 7.29.30.94.144 j j j
t suggested that the latencies of mag-
s a s o ing threshold changes and seizure provocation; guidelines for
netic root stimulation could replace F response determinations preventing unwanted side effects have been suggested. 113
because it is easier and faster to perform. Further, if F re-

24
sponses are absent due to disease, it is often still possible to

obtain magnetic evoked root responses. One study used the la-
CONCLUSION
tencies of cervical and lumbar root stimulation to assess the
homogeneity of peripheral conduction in generalized (polyneu- Magnetic stimulation holds promise primarily with respect to
ropathies), focal (lumbar stenosis), and multifocal (inflamma- its ability to noninvasively activate areas of the brain. Potential
tory demyelinating polyneuropathy) disease. However, more 152
clinical applications include study of multiple sclerosis, spinal
studies are needed to investigate the place of this promising cord injury or other types of myelopathy, and stroke.
technique in the work-up of peripheral nerve disorders. Intraoperative monitoring may be an application if effects of
anesthesia can be adequately controlled. Interesting research ap-
plications include cortical mapping, study of neuroplasticity, as
SAFETY CONSIDERATIONS well as study of inhibitory phenomena. Applications regarding
the peripheral nervous system concern the easy stimulation of
Although magnetic stimulators have been approved for use in several cranial nerves at their intracranial course and the rela-
the peripheral nervous system, they are not yet FDA-approved tively painless root stimulation, providing the opportunity to
for cortical stimulation. In addition to some of the concerns assess proximal conduction.
about the electrical safety to the user mentioned earlier, there
are still a number of concerns about the safety of providing this
type of stimulation to the brain. REFERENCES
When applying electrical or magnetic stimulation to the brain
there is a concern about either inducing a single seizure or kin- 1. Abbruzzese G, Dall'Agata D, et al: Electrical stimulation of the motor tracts in
cervical spondylosis. J Neurol Neurosurg Psychiatry 1988;51:796-802.
dling a seizure focus. There are reports of inducing seizures
2. Abbruzzese G, Morena M, Dall'Agata D, et al: Motor evoked potentials
with magnetic cortical stimulation in patients susceptible to (MEPs) in lacunar syndromes. Electroencephalogr Clin Neurophysiol 1991 ;81:
seizures. One patient with a large stroke has had a new seizure 202-208.
induced after magnetic stimulation, although no subsequent 3. Agnew WF, McCreery DB: Considerations for safety in the use of extracranial
stimulation for motor evoked potentials. Neurosurgery 1987;20:143-147.
spontaneous seizures followed (i.e., no seizure focus was likely 4. Amassian VE, Cracco RQ: Human cerebral cortical responses to contralateral
induced). Attempts to intentionally produce seizures in seizure
69
transcranial stimulation. Neurosurgery 1987;20:148-155.
prone individuals have been successful, although multiple
43137
5. Amassian VE, Steward M, Quirk GJ, Rosenthal JL: Physiological basis of
stimulations have often been required to produce a seizure. motor effects of a transient stimulus to cerebral cortex. Neurosurgery
1987;20:74-93.
Kindling is a much less likely event than producing a single 6. Andersen B; Rosier KM; Lauritzen M: Nonspecific facilitation of responses to
seizure. While application of prolonged, high rates of electrical transcranial magnetic stimulation. Muscle Nerve 1999;22:857-863. j
stimulation has produced seizure foci in animals, investigators 7. Banerjee TK, Mostofi MS, Us O, et al: Magnetic stimulation in the determina- I
tion of lumbosacral motor radiculopathy. Electroencephalogr Clin Neuro-
have not been able to induce kindling in animals with stimula- physiol 1993;89:221-226.
tion frequencies below 10 Hz, despite prolonged durations. 60
8. Barker AT, Jalinous R, Freeston IL: Non-invasive magnetic stimulation of the
There are a number of other potential adverse effects that could human motor cortex. Lancet 1985; 1:1106-1107.
9. Barker AT, Freeston IL, Jalinous R, Jarratt JA: Clinical evaluation of conduction
be theoretically problematic with magnetic stimulators, but these
time measurements in central motor pathways using magnetic stimulation of
have not been shown to be reason for particular concern within the human brain. Lancet 1986; 1:1325-1326.
the parameters used in commercially available units. The mag- 11
10. Barker AT, Freeston IL, Jalinous R, Jarratt JA: Magnetic stimulation of the
netic field strength itself is within the range of, or lower than, human brain and peripheral nervous system: An introduction and the results of
an initial clinical evaluation. Neurosurgery 1987:20:100-109.
many magnetic resonance imaging units that have not been 11. Barker AT: Basic principles of magnetic stimulation of the central and periph-
shown to be harmful (except for movement of metallic objects). eral nervous system. In AAEE/AEEGS Joint Symposium on Somatosensory
With regard to current intensities, magnetic stimulation is esti- Evoked Potentials and Magnetic Stimulation. Rochester, MN. American
Association of Electrodiagnostic Medicine, 1988.
mated to induce 1.9 microcoulombs per square centimeter per
12. Beckers G, Homberg V: Impairment of visual perception and visual short-term
phase ( C/cm /ph) at the scalp and 0.8 C/cm /ph at the brain's
u
2
u
2
memory scanning by transcranial magnetic stimulation of occipital cortex. Exp

surface." Data from animal experiments suggest that induced Brain Res 1991;87:421-432.
fields of < 10 C/cm /ph are safe, even with continuous 50-Hz
u
2 3
13. Benecke R: Magnetic stimulation in the assessment of peripheral nerve disor-
ders. Baillieres Clin Neurol 1996;5:115-128.
stimulation over many hours. The amount of thermal energy dis-
14. Berardelli A, Inghilleri M, Manfredi M, et al: Cortical and cervical stimulation
sipated is below the threshold for concern; with 0.3 Hz stimula- after hemispheric infarction. J Neurol Neurosurg Psychiatry 1987;50:861-865.
tion the maximum average power dissipation is about 53 pW, 5 15. Berardelli A, Inghilleri M. Formisano R, et al: Stimulation of motor tracts i n
orders of magnitude less than the adult basal metabolic rate. motor neuron disease. J Neurol Neurosurg Psychiatry 1987;50:732-737.
16. Berardelli A, inghilleri M, Cruccu G, Manfredi M: Corticospinal potentials
There are probably no significant cognitive changes after after electrical and magnetic stimulation in man. Electroencephalogr Clin
magnetic cortical stimulation. In studies of normal subjects Neurophysiol 1991(Suppl 43); 147-154.
C h a p t e r 10 MAGNETIC STIMULATION OF THE CENTRAL A N D PERIPHERAL NERVOUS SYSTEMS — 4 2 5
17. Bickford RG, Guidi M, Fortesque P, Swenson M; Magnetic stimulation of 48. Eisen AA, Shtybel W: Clinical experience with trancranial magnetic stimula-
human peripheral nerve and brain: Response enhancement by combined mag- tion. Muscle Nerve 1990;13:995-1011.
netoelectrical technique. Neurosurgery 1987;20:110-116. 49. Epstein CM, Fernandez-Beer E, Weissman JD, Matsuura S: Cervical magneitc
18. Boyd SG, Rothwell JC, Cowan JMA, et al: A method of monitoring function in stimulation: The role of the neural foramen. Neurology 1991;41:677-680.
corticospinal pathways during scoliosis surgery with a note on motor conduc- 50. Epstein CM: Magnetic mapping of human cervical nerve roots: Variation in
tion velocities. J Neurol Neurosurg Psychiatry 1986;49:251-257. normal subjects. Electroenceph Clin Neurophysiol 1993:89;145-153.
19. Brain L: Some unsolved problems of cervical spondylosis. Br Med J 1963; 51. Evans BA, Litchy WJ, Daube JR: The utility of magnetic stimulation for routine
1:771-777. peripheral nerve conduction studies. Muscle Nerve 1988; 11:1074-1078.
20. Brain WR, Northfield D, Wilkinson M: The neurological manifestations of cer- 52. Evans BA, Daube MD, Litchy WJ: A comparison of magnetic and electric stim-
vical spondylosis. Brain 1952;75:187-225. ulation of spinal nerves. Muscle Nerve 1990;13:414-420.
21. Brasil-Neto JP, Cohen LG, Pascual-Leone A, et al: Rapid reversible modulation 53. Evans BA: Magnetic stimulation of the peripheral nervous system. J Clin
of human motor outputs after transient deafferentation of the forearm: A study Neurophysiol 1991:8;77-84.
with transcranial magnetic stimulation. Neurology 1992;42:1302-1306. 54. Ferbert A, Mussmann N, Menne A, et al: Short-term memory performance with
22. Bridgers SL, Delaney RC: Assessment of short term effects of magnetic cortical magnetic stimulation of the motor cortex. Eur Arch Psychiatry Clin Neurosci
stimulation. Abstract, AAEE/AEEGS Joint Symposium, San Diego, 1988. 1991;241:135-138.
23. Bridgers SL: The safety of transcranial magnetic stimulation reconsidered: Evi- 55. Gandevia SC, Rothwell JC: Activation of the human diaphragm from the motor
dence regarding cognitive and other cerebral effects. Electroencephalogr Clin cortex. J Physiol 1987;384:109-118.
Neurophysiol 1991 fSuppl 43): 170-179. 56. Ghaly RF, Stone JL, Levy WJ,et al: The effect of etomidate on motor evoked
24. Britton TC, Meyer BU, Herdman J, Benecke R: Clinical use of the magnetic potentials induced by transcranial magnetic stimulation in the monkey.
stimulator in the investigation of peripheral conduction time. Muscle Nerve Neurosurgery 1990;27:936-942.
1990;13:396-406. 57. Ghezzi A, Callea L, Zaffaroni M, Zibetti A: Motor potentials of inferior orbicu-
25. Calancie B, Nordin M, Wallin U, Hagbarth KE: Motor-unit responses in human laris oculi muscle to transcranial magnetic stimulation: Comparison with re-
wrist flexor and extensor muscles to transcranial cortical stimuli. J sponses to electrical peripheral stimulation of facial nerve. Electroenceph Clin
Neurophysiol 1987;58:1168-1185. Neurophysiol 1992:85;248-252.
26. Calancie B, Klose KJ, Baier S, Green BA: Isoflurane-induced attenuation of 58. Gianutsos J, Eberstein A, Ma D, et al: A noninvasive technique to assess com-
motor evoked potentials caused by electrical motor cortex stimulation during pleteness of spinal cord lesions in humans. Exp Neurol 1987;98:34-40.
surgery. J Neurosurg 1991;74:897-904. 59. Glocker FX, Magistris MR, Rosier KM, Hess CW: Magnetic transcranial and
27. Chan KM; Nasathurai S; Chavin JM; Brown WF: The usefulness of central electrical stylomastoidal stimulation of the facial motor pathways in Bell's
motor conduction studies in the localization of cord involvement in cervical palsy: Time course and relevance of electrophysiological parameters. Electro-
spondylytic myelopathy. Muscle Nerve 1998;21:1220-1223. encephalogr Clin Neurophysiol 1994;93:113-20.
28. Chokroverty S: Magnetic stimulation of the human peripheral nerves. 60. Goddard GV, Mclntyre DC, Leech CK: A permanent change in brain function
Electromyogr Clin Neurophysiol 1989;29:409^H6. resulting from daily electrical stimulation. Exp Neurol 1969;25:295-330.
29. Chokroverty S, Sachdeo R, Dillulo J, Duvoisin RC: Magnetic stimulation in the 61. Hallet M, Wassermann EM, Cohen LG: Mapping the motor cortex using mag-
diagnosis of lumbosacral radiculopathy. J Neurol, 1989:52;767-772. netic stimulation. In 1993 AAEM Course F: The Role of Magnetic Stimulation
30. Chokroverty S, Flynn D, Picone MA, et al: Magnetic coil stimulation of the in Clinical Neurophysiology. Rochester, MN, American Association of
human lumbosacral vertebral column: Site of stimulation and clinical applica- Electrodiagnostic Medicine, 1993.
tion. Electroenceph Clin Neurophysiol 1993:89:54-60. 62. Hamdy S, Aziz Q, Rothwell JC, et al: Recovery of swallowing after dysphagic
31. Cohen LG, Hallet M: Cortical stimulation does not cause short-term changes in stroke relates to functional reorganization in the intact motor cortex. Gastro-
the electroencephalogram. Ann Neurol 1987;21:512-513. enterology 1998;115:1104-1112.
32. Cohen LG, Hallet M: Methodology for non-invasive mapping of human motor 63. Haug BA, Schonle PW, Knobloch C, Kohne M: Silent period measurement re-
cortex with electrical stimulation. Electroencephalogr Clin Neurophysiol vives as a valuable diagnostic tool with transcranial magnetic stimulation.
1988;69:403-411. Electroencephalogr Clin Neurophysiol 1992;85:158-160.
33. Cohen LG, Hallett M: Noninvasive mapping of human motor cortex. Neurology 64. Heald A, Bates D, Cartlidge NEF, et al: Longitudinal study of central motor
1988;38:904-909. conduction time following stroke. 1. Natural history of central motor conduc-
34. Cohen LG, Roth BJ, Nilsson J, et al: Effects of coil design on delivery of focal tion. Brain 1993;116:1355-1370.
magnetic stimulation: Technical considerations. Electroencephalogr Clin 65. Heald A, Bates D, Cartlidge NEF, et al: Longitudinal study of central motor
Neurophysiol 1990;75:350-357. conduction time following stroke. 2. Central motor conduction measured within
35. Cohen LG, Bandinelli S, Findley TW, Hallett M: Motor reorganization after 72h after stroke as a predictor of functional outcome at 12 months. Brain
upper limb amputation in man: A study with focal magnetic stimulation. Brain 1993;116:1371-1385.
1991;114:615-627. 66. Hess CW, Mills KR, Murray NMF: Measurement of central motor tract con-
36. Cohen LG, Bandinelli S, Topka HR, et al: Topographic maps of human motor duction in multiple sclerosis by magnetic brain stimulation. Lancet 1986;2:
cortex in normal and pathological condition: Mirror movements, amputations 355-358.
and spinal cord injuries. Electroencephalogr Clin Neurophysiol 1991(Suppl 67. Hess CW, Mills KR, Murray NMF, SchrieferTS: Magnetic brain stimulation:
43):36-50. Central motor conduction studies in multiple sclerosis. Ann Neurol 1987;22:
37. Counter SA, Borg E, Lofqvist L, Brismar T: Hearing loss from the acoustic arti- 744-752.
fact of the coil used in extracranial magnetic stimulation. Neurology 1990;40: 68. Hess CW, Mills KR, Murray NMF: Responses in small hand muscles from
1159-1162. magnetic stimulation of the human brain. J Physiol 1987;388:397-419.
38. Cowan JMA. Dick JPR, Day BL. et al: Abnormalities in central motor pathway 69. Homberg V, Netz J: Generalised seizures induced by transcranial magnetic
conduction in multiple sclerosis. Lancet 1984;2:304-307. stimulation of motor cortex [letter]. Lancet 1989;2:1223.
39. Cracco RQ: Evaluation of conduction in central motor pathways: Tech- 70. Hugon J, Lubeau M, Tabaraud F, et al: Central motor conduction in motor
niques, pathophysiology, and clinical interpretation. Neurosurgery 1987;20: neuron disease. Ann Neurol 1987;22:544-546.
199-203. 71. Humphrey DR: Representation of movements and muscles within the primate
40. Cros D, Day TJ, Shahani B: Spatial dispersion of magnetic stimulation in pe- precentral motor cortex, historical and current perspectives. Fed Proc 1986;
ripheral nerves. Muscle Nerve 1990;13:1076-1082. 45:2687-2699.
41. Day BL, Dick JPR, Marsden CD, Thompson PD: Differences between electri- 72. Ingram DA, Swash M: Central motor conduction is abnormal in motor neuron
cal and magnetic stimulation of the human brain. J Physiol 1986;378:36P. disease. J Neurol Neurosurg Psychiatry 1987;50:159-166.
42. Day BL. Rothwell JC. Thompson PD, et al: Motor cortex stimulation in intact 73. Kandler R: Safety of transcranial magnetic stimulation [comment]. Lancet
man: 2. Multiple descending volleys. Brain 1987; 110:1191-1209. 1990;335:469-470.
43. Dhuna A, Gates J, Pascual-Leone A: Transcranial magnetic stimulation inpa- ,74. Kandler RH, Jarratt JA, Davies-Jones GA, Gumpert EJ: The role of magnetic
tients with epilepsy. Neurology 1991 ;41:1067-1071. stimulation as a quantifier of motor disability in patients with multiple sclerosis.
44. Di Lazzaro V, Oliviero A, Profice P et al: Comparison of descending volleys J Neurol Sci 1991;106:31-34.
evoked by transcranial magnetic and electric stimulation in conscious humans. 75. Koh TH, Eyre JA: Maturation of corticospinal tracts assessed by electromag-
Electroencephalogr Clin Neurophysiol 1998:109:397-401. netic stimulation of the motor cortex. Arch Dis Child 1988;63:1347-1352.
45. Dominkus-M, Grisold-W, Jelinek-V: Transcranial electrical motor evoked po- 76. Konrad PE, Tacker WA, Levy WJ, et al: Motor evoked potentials in the dog: Ef-
tentials as a prognostic indicator for motor recovery in stroke patients. J Neurol fects of global ischemia on spinal cord and peripheral nerve signals.
Neurosurg Psychiatry 1990;53:745-748. Neurosurgery 1987;20:117-124.
46. Drummond JC, Tung HC, Bickford RG: The effects of sedative agents on mag- 77. LaRocca H: Cervical spondylotic myelopathy: Natural history. Spine 1988; 13:
netic motor evoked potentials. Abstract, AAEE/AEEGS Joint Symposium, San 854-855.
Diego, 1988. 78. Larsson EM, Holt S, Cronqvist S, Brandt L: Comparison of myelography, CT
47. Eisen A, Shytbel W, Murphy K, Hoirch M: Cortical magnetic stimulation in myelography and magnetic resonance imaging in cervical spondylosis and disk
amyotrophic lateral sclerosis. Muscle Nerve 1990;13:146-151. herniation. Acta Radiol 1989;30:233-239.
79. Lestini WF, Wiesel SW: The pathogenesis of cervical spondylosis. Clin Orthop 110. Oliveri M, Brighina F, La Bua V, et al: Reorganization of cortical motor area in
1989;239:69-93. prior polio patients. Clin Neurophysiol 1999.110:806-812.
80. Levy WJ, York D, McGaffrey M: Motor evoked potentials from transcranial 111. Pascual-Leone A, Brasil-Neto JP, Valls-Sole J, et al: Simple reaction time to
stimulation of the motor cortex in humans. Neurosurgery 1984;15:287-302. focal transcranial magnetic stimulation. Comparison with reaction time to
81. Levy WJ, McCaffrey M, York DH, Tanzer F: Motor evoked potentials from acoustic, visual and somatosensory stimuli. Brain 1992; 115:109-122.
transcranial stimulation of the motor cortex in cats. Neurosurgery 112. Pascual-Leone A, Cohen LG, Shotland LI, et al: No evidence of hearing loss in
1984;15:214-226. humans due to transcranial magnetic stimulation. Neurology 1992:42;647-65I.
82. Levy WJ, McCaffrey M, York D: Motor evoked potentials in cats with acute 113. Pascual-Leone A, Houser CM, Reese K, et al: Safety of rapid-rate transcranial
spinal cord injury. Neurosurgery 1986;19:9-19. magnetic stimulation in normal volunteers. Electroencephalogr Clin
83. Levy WJ: Clinical experience with motor and cerebellar evoked potential moni- * Neurophysiol 1993;89:120-130.
toring. Neurosurgery 1987;20:169-182. 114. Patton HD, Amassian VE: Single and multiple unit analysis of cortical stage of
84. Liepert J, Terborg C, Weiller C: Motor plasticity induced by synchronized pyramidal activation. J Neurophysiol 1954;17:345-363.
thumb and foot movements. Exp Brain Res 1999; 125:435^39. 115. Patton HD, Amassian VE: The pyramidal tract: Its excitation and functions. In
85. Little JW, Goldstein BS, Robinson LR: Prolonged cortical inhibition after elec- Field J, Magoun HW, Hall VE (eds): Handbook of Physiology: Neurophysiology,
tromagnetic stimulation of the motor cortex. Muscle Nerve 1991; 14:894. II. Washngton, DC, American Physiology Society, 1960, pp 837-861.
86. Little JW, Robinson LR, Goldstein BS, et al: Electrophysiologic findings in 116. Penfield W, Boldrey E: Somatic motor and sensory representation in the cere-
post-traumatic syringomyelia: Implications for clinical management. J Am bral cortex of man as studied by electrical stimulation. Brain 1937;60:389-443.
Paraplegia Soc 1992; 15:44-52. 117. Ravnborg-M, Dahl-K: Examination of central and peripheral motor pathways
87. Little JW, Robinson LR: Electrodiagnosis in post-traumatic syringomyelia: by standardized magnetic stimulation. Acta Neurol Scand 1991;84:491-497.
Case report and review of the literature. Muscle Nerve 1992;15:755-760. 118. Rimpilainen 1, Karma P, Laranne J, et al: Magnetic facial nerve stimulation in
88. Luo YM, Johnson LC, Polkey MI, et al: Diaphragm electromyogram measured Bell's palsy. Acta Otolaryngol 1992; 112:311-316.
with unilateral magnetic stimulation. Eur Respir J 1999;13:385-390. 119. Rimpilainen I, Eskola H, Laippala P, et al: Prognostication of Bell's palsy using
89. Maccabee PJ, Amassian VE, Cracco RQ, Cadwell JA: An analysis of peripheral transcranial magnetic stimulation. Acta Otolaryngol Suppl 1997;529:111-115.
nerve mor stimulation in humans using the magnetic coil. Electroencephalogr 120. Robinson LR, Jantra P, MacLean IC: Central motor conduction times using
Clin Neurophysiol 1988;70:524-533. transcranial stimulation and F-wave latencies. Muscle Nerve 1988; 11:174-180.
90. Maccabee PJ, Amassian VE, Eberle LP, et al: Measurement of the electric field 121. Robinson LR, Little JW: Motor evoked potentials reflect spinal cord function in
induced into inhomogenous volume conductors by magnetic coils: Application post-traumatic syringomyelia. Am J Phys Med Rehabil 1990;69:307-310.
to human spinal neurogeometry. Electroencephalogr Clin Neurophysiol 1990; 122. Robinson LR, Goldstein, BS, Little JW: Silent periods after electromagnetic
76:131-141. stimulation of the motor cortex. Am J Phys Med Rehabil 1993;72:23-28.
91. Maccabee PJ, Eberle L, Amassian VE, et al: Spatial distribution of the electrical 123. Rohde V, Will BE, Hahn G, et al: Motor evoked potentials during embolization
field induced in volume by round and figure 8 magnetic coils: Relevance to ac- of arteriovenous malformations for the detection of ischemic complications.
tivation of sensory nerve fibers. Electroencephalogr Clin Neurophysiol 1991; Zentralbl Neurochir 1999;60:74-80.
81:224-237. 124. Roricht S, Meyer BU, Niehaus L, Brandt SA: Long-term reorganization of
92. Maccabee PJ, Amassian VE, Cracco RQ, et al: Stimulation of the human ner- motor cortex outputs after arm amputation. Neurology 1999;53:106-111.
vous system using the magnetic coil. J Clin Neurophysiol 1991;8:38-55. 125. Rosier KM, Hess CW, Schmid UD: Investigation of facial motor pathways by
93. Macdonell RA, Donnan GA, Bladin PF: A comparison of somatosensory electrical and magnetic stimulation: Sites and mechanisms of excitation. J
evoked and motor evoked potentials in stroke. Ann Neurol 1989;25:68-73. Neurol Neurosurg Psychiatry 1989:52:1149-1156.
94. Maegaki Y, Inagaki M, Takeshita K: Cervical magnetic stimulation in children 126. Rosier KM, Schmid UD, Hess CW: Transcranial magnetic stimulation of the
and adolescents: Normal values and evaluation of the proximal lesion of the pe- facial nerve: Where is the actual excitation site? Electroencephalogr Clin
ripheral motor nerve in cases with polyradiculoneuropathy. Electroencephalogr Neurophysiol Suppl 1991;43:362-368.
Clin Neurophysiol 1994;93:318-323. 127. Rossini PM, Caramia M, Zarola F: Mechanisms of nervous propagation along
95. MacLean IC: Spinal nerve stimulation. Phys Med Rehabil Clin North Am central motor pathways: Noninvasive evaluation in healthy subjects and in pa-
1994;5:509-529. tients with neurological disease. Neurosurgery 1987;20:183-191.
96. Maertens de Noordhout A, Remade JM, Pepin JL, et al: Magnetic stimulation 128. Rossini PM: The anatomic and physiologic bases of motor-evoked potentials. *
of the motor cortex in cervical spondylosis. Neurology 1991;41:75-80. Neurol Clin 1989;6:751-769.
96a. Magnusson CE, Stevens HC: Visual sensations created by a magnetic field. 129. Rothwell JC, Thompson PD, Day BL, et al: Motor cortex stimulation in intact
Phil Mag 1914;24:188-207. man: 1. General characteristics of EMG responses in different muscles. Brain
97. McDonald CM, Robinson LR: Transcranial cortical stimulation in sy- 1997;110:1173-1190.
ringomyelia. Muscle Nerve 1994; 17:114-115. 130. Samii A, Luciano CA, Dambrosia JM, Hallett M: Central motor conduction
98. Merton PA, Morton HB: Stimulation of the cerebral cortex in the intact human time: reproducibility and discomfort of different methods. Muscle Nerve
subject. Nature 1980;285:227. 1998;21:1445-1450.
99. Merton PA, Morton HB, Hill DK, Marsden CD: Scope of a technique for electri- 131. Schmid UD, Walker G, Hess CW, Schmid J. Magnetic and electrical stimula-
cal stimulation of human brain, spinal cord, and muscle. Lancet 1982;2:597-600. tion of cervical motor roots: Technique, site and mechanisms of excitation. J
100. Mills KR, Murray NMF: Corticospinal tract conduction times in multiple scle- Neurol Neurosurg Psychiatry 1990;53:770-777.
rosis. Ann Neurol 1985;18:601-605. 132. Schmid UD, Walder G, Schmid-Sigron J, Hess CW: Transcutaneous magnetic
101. Mills KR, Murray NMF: Electrical stimulation over the human vertebral and electrical stimulation over the cervical spine: Excitation of plexus roots
column: Which neural elements are excited. Electroencephalogr Clin Neuro- rather than spinal roots. Electroencephalogr Clin Neurophysiol 1991;(Suppl
physiol 1986;63:582-589. 43):369-384.
102. Mills KR, Murray NMF, Hess CW: Magnet and electrical transcranial brain 133. SchrieferTN, Mills KR, Murray NMF, Hess CW. Evaluation of proximal facial
stimulation: Physiological mechanisms and clinical applications. Neurosurgery nerve conduction by transcranial magnetic stimulation. J Neurol Neurosurg
1987;20: 164-168. Psychiatry 1988;51:60-66.
103. Mills KR; Nithi KA: Peripheral and central motor conduction in amyotrophic 134. Snooks SJ, Swash M: Motor conduction velocity in the human spinal cord:
lateral sclerosis. J Neurol Sci 1998; 15; 159:82-87. Slowed conduction in multiple sclerosis and radiation myelopathy. J Neurol
104. Miscio G, Pisano F, Mora G, Mazzini L: Motor neuron disease: Usefulness of Neurosurg Psychiatry I95;48:l 135-1139.
transcranial magnetic stimulation in improving the diagnosis. Clin 135. Stone JL, Ghaly RF, Levy WJ: A comparative analysis of enflurane anesthesia
Neurophysiol 1999; 110:975-981. on primate motor and somatosensory evoked potentials. Electroencephalogr
105. Muellbacher W, Mathis J, Hess CW: Electrophysiological assessment of central Clin Neurophysiol 1992;84:180-187.
and peripheral motor routes to the lingual muscles. J Neurol Neurosurg 136. Streletz LJ, Jones SM, Belevich JKS, et al: Rapid reorganization of motor path-
Psychiatry 1994;57:309-315. ways in acute spinal cord injury revealed by transcranial magnetic stimulation.
106. Murray NMF: Physiology and pathophysiology of motor mechanisms. In Electroencephalogr Clin Neurophysiol 1992:83:85.
AAEE/AEEGS Joint Symposium on Somatosensory Evoked Potentials and 137. Tassinari CA, Michelucci R, Forti A, et al: Transcranial magnetic stimulation in
Magnetic Stimulation. Rochester, MN, AAEM, 1988. epileptic patients: Usefulness and safety. Neurology 1990:40:1132-1133.
107. Murray NMF Magnetic stimulation of cortex: clinical applications. J Clin 138. Tavy DL. Franssen H, Keunen RW, et al: Motor and somatosensory evoked po-
Neurophysiol 1991;81:66-67. tentials in asymptomatic spondylotic cord compression. Muscle Nerve I999;22:
108. Nilsson J, Panizza M, Roth BJ, et al: Determining the site of stimulation during 628-634
magnetic stimulation of peripheral nerve. Electroencephalogr Clin Neuro- 139. Thompson PD. Dick JPR, Asselman GB, et al: Examination of motor function
physiol 1986:63:582-589. in lesions of the spinal cord by examination of the motor cortex. Ann Neurol
109. Odderson I: The role of magnetic stimulation in evaluation of the peripheral 1987;21:389-396.
nervous system. In 1993 AAEM Course F: The role of magnetic stimulation in 140. Triggs WJ, Menkes D, Onorato J, et al: Transcranial magnetic stimulation iden-
clinical neurophysiology. Rochester. MN, American Association of Electro- tifies upper motor neuron involvement in motor neuron disease. Neurology
diagnostic Medicine, 1993. 1999;53:605-611.
141. Ubags LH, Kalkman CJ, Been HD, et al: A comparison of myogenic motor 146. Yamakawa T, Yoshikawa H, Arai A, et al: A clinical study on the magnetic stim-
evoked responses to electrical and magnetic transcranial stimulation during ni- ulation of the facial nerve. Laryngoscope 1999;109:492-497.
trous oxide/opioid anesthesia. Anesth Analg 1999;88:568-572. 147. Zentner J, Schumacher M, Bien S: Motor evoked potentials during interven-
142. Ugawa Y, Rothwell JC, Day BL, et al: Magnetic stimulation over the spinal en- tional neuroradiology. Neuroradiology 1988;30:252-255.
largements. J Neurol Neurosurg Psychiatry 1989;52:1025-1032. 148. Zentner J, Kiss I, Ebner A: Influence of anesthetics—nitrous oxide in particu-
143. Uozumi T, Tsuji S, Murai Y: Motor potentials evoked by magnetic stimulation lar—on electromyographic response evoked by transcranial electrical stimula-
of the motor cortex in normal subjects and patients with motor disorders. tion of the cortex. Neurosurgery 1989;24:253-256.
Electroencephalogr Clin Neurophysiol 1991;81:251-256. 149. Zentner J: Motor evoked potential monitoring during neurosurgical operations
144. Westerink M, van Woerkom TC, Tavy DL: Transcranial magnetic stimulation in on the spinal cord. Neurosurg Rev 1991;14:29-36.
peripheral facial nerve palsy of central origin. Clin Neurol Neurosurg 150. Zwarts MJ: Central motor conduction in relation to contra- and ipsilateral acti-
1991;93:45-49. vation. Electroenceph Clin Neurophysiol 1992;85:425-429.
145. Woolsey CN, Erickson TC, Gilson WE: Localisation in somatic sensory 151. Zwarts MJ. Sensory potentials evoked by magnetic stimulation of the cervical
and motor areas of human cerebral cortex as determined by direct record- spine. Muscle Nerve 1993;16:289-293.
ing of evoked potentials and electrical stimulation. J Neurosurg 1979;51: 152. Zwarts MJ: Magnetic stimulation of the peripheral nervous system: Local versus
476-506. generalized disorders. Electromyogr Clin Neurophysiol 1998;38:309-316.
Chapter I I
Q u a n t i t a t i v e Sensory T e s t i n g :
Basic Principles a n d C l i n i c a l
Applications
Gil I.Wolfe, M.D.
CHAPTER OUTLINE
Quantitative Sensory Testing Clinical Applications of QST
Historical Background • Physical Properties • The Rationale Diabetic Nephropathy • Uremic Neuropathy • Vitamin B i 2
for Performing QST • QST Units and Equipment • Test Deficiency States • Alcoholic Neuropathy * HIV-Related
Algorithms for QST • Reference Data and O t h e r Testing Neuropathy • Cryptogenic Sensory Polyneuropathy (CSPN) •
Issues O t h e r Applications • Illustrative Case
HISTORICAL BACKGROUND used, but these reports account for only a small fraction of the
literature. For instance, cutaneous pressure can be quantified to
Over 25 years have passed since quantitative sensory test- some extent with simple instruments, such as the graded
ing (QST) was introduced in human subjects. Since this first
30
Semmes-Weinstein monofilaments popularized in the evalua-
report, a variety of QST instruments have been developed, and tion of patients with Hansen's disease. Electrical current per-
investigators have applied QST to a wide range of disorders af- ception, where sensory receptors are bypassed and axons are
fecting sensory function with varying degrees of success. In 92
directly depolarized by an electrical current, has also been stud-
most disorders, including peripheral neuropathies with sensory ied,1361
but questions regarding the reliability of this approach
involvement, investigative work is predominantly descriptive, remain. 53
providing the rate of abnormalities to a particular sensory The advantages of thermal and vibratory stimuli include the
modality. In diabetic neuropathy, however, extensive work has evaluation of the entire somatosensory axis, from receptor to
defined how QST can be used in the diagnosis, staging, and cortex, and the assessment of a wide spectrum of sensory fibers
follow-up of patients, with a focus on the application of QST in (Table 11-1). Vibratory stimuli are first detected by Pacinian
clinical trials. - In several sensory disorders, QST has been
17 27
corpuscles located in the skin, which in turn activate large
compared with other laboratory measurements of nerve func- myelinated fibers (A beta) that carry the impulses to the dorsal
tion, including nerve conduction studies and autonomic testing, column system of the spinal cord. Thermal and painful stimuli
often showing a higher yield of abnormality. are detected by bare nerve endings and are transmitted to the
Most work with QST has concentrated on sensory impair- spinoreticulothalamic tracts of the spinal cord by small myeli-
ment related to peripheral neuropathies. This bias is a likely out- nated (A delta) and/or unmyelinated (C) fibers, depending on
growth of the advances made in neuroradiology, particularly the exact nature of the stimulus. It is important to realize that
magnetic resonance imaging. To date, these modern imaging QST does not effectively localize the site of injury along the so-
techniques have shown greater utility and popularity in the as- matosensory axis.
sessment of central nervous system disorders. 92
Threshold measurements are the most common parameter
used in QST. Although QST instruments are designed to deliver
quantitative data* test results are generated from subjective re-
PHYSICAL PROPERTIES OF Q S T sponses, as the test subject must signal when he or she perceives
the appropriate stimulus. Therefore, methodologic differences
I Temperature and vibration are the sensory modalities most or poor adherence to protocols can result in contradictory re-
commonly assessed in QST. Other sensory stimuli have been sults. Subject age and height, testing site, the size and pressure
429
TABLE 11-I. Sensory Receptors and Afferent Pathways

QST UNITS A N D EQUIPMENT
Afferent
Stimulus Receptor Fiber Type Central Pathways Stimulus intensity can be expressed in two types of units. An
absolute value such as a temperature or micrometer can be used.
Vibration Pacinian corpuscle A-beta Dorsal columns A second system uses "just noticeable differences," (JNDs),
Cold Naked nerve endings A-delta, C Spinothalamic tracts the minimal difference between two stimuli that permits them to
Warm Naked nerve endings C Spinothalamic tracts be distinguished, assuming normal sensory function. The full
range of sensation, from threshold to tolerance, can be divided
Cold-pain Naked nerve endings A-delta, C Spinothalamic tracts
in JNDs. Expressing QST data using JNDs has a physiologic
Heat-pain Naked nerve endings A-delta, C Spinothalamic tracts advantage in that JNDs are small near the sensory threshold and
grow progressively larger at higher stimulus intensities. 89
Applying JNDs obtained from normal controls to patients with

applied by the stimulator, and subject training all can affect sensory impairment is not a simple process, but, fortunately, has
QST results. Although QST is based on subjective responses, become available with computerized commercial systems.
data collected from cooperative patients have proven to be reli- QST instruments can be divided into those that generate spe-
able over time. A variety of techniques are in practice to identify cific vibratory or thermal stimuli, and those that deliver electri-
poorly cooperative patients including null stimuli and thresh-
23
cal impulses at a variety of frequencies to stimulate the different
old variance. 88
classes of sensory fibers. Vibratory stimuli may be produced in
a variety of ways, including vibrating electromagnets and move-
ment of a surface mounted to a motorized shaft that undergoes
THE RATIONALE FOR PERFORMING QST vertical displacement proportionate to the current. Thermal
stimuli are uniformly produced using the Peltier principle, first
There are several reasons why Q S T is of potential value in applied to QST by Kenshalo and Bergen. By alternating the
56
both research and routine clinical settings. First, routine nerve direction of current through a metal thermocouple (Peltier ele-
conduction studies (NCS) are a poor test of small-fiber sensory ment), one can alternatively warm or cool the metal surface. The
function, while small fiber modalities (e.g., thermal and pain amount of current passing through the thermocouple will deter-
thresholds) can be quantified with QST. In addition, QST also mine the actual temperature, which is transmitted to a high con-
provides a measure of sensory function, something that may or ductance material that serves as the "active surface." A heat sink
may not correlate with the physiologic properties of the nerve of circulating water sits adjacent to the thermocouple in order to
determined by NCS. As will be described later, there are many improve the efficiency and accuracy of thermal testing. The
instances where Q S T is more sensitive than NCS in detecting water temperature is maintained by a second thermocouple, al-
sensory nerve impairment. Since most peripheral nerve disor- lowing for tight control of the active surface's temperature with
ders will eventually impact both small and large sensory fibers, only small adjustments of current. A thermister on the active
QST and NCS can be viewed as complementary evaluations. surface provides continuous feedback to the current generator to
Other advantages are that QST is relatively simple to perform maintain the temperature at the desired level. System software
both in the clinic and in the field and is largely painless. can generate temperatures throughout the physiologic range
Therefore, QST has the potential to serve as a useful screening (0-50°C), alter the duration of the stimulus, and change the
and follow-up procedure in a variety of clinical settings without temperature at rates as high as 6°C per second (Fig. 11-1).
being a burden to test subjects.
TEST ALGORITHMS FOR QST

There are two basic test algorithms used by the various QST
instruments. The first is referred to as the method of limits or
ramps and requires patients to depress a switch or button when
an increasingly strong stimulus (ascending ramp) is first per- !
ceived. Likewise, the patient may be asked to depress the indi-
cator when a decreasing stimulus (descending ramp) is no
longer felt. The rate of change in the ascending or descending
ramp can be adjusted. The second algorithm is the method of
levels or forced choice. In this methodology, test subjects
simply respond whether or not they perceive a single sensory
stimulus during a given time period or, alternatively, select
which of two devices is delivering a stimulus. Sensory thresh-
olds are generally higher for method of limits testing than for
method of levels or forced choice testing. - - - In method of
22 23 45 87
limits testing, the stimulus intensity continues to increase

during the reaction time, the interval required for the peripheral
FIGURE I /-/. Thermal threshold testing using a computer- stimulus to travel along afferent pathways to affect a motor re-
ized QST system.The active surface is strapped t o the dorsal sur- sponse allowing the hand to depress the indicator switch or
face of t h e right hand. Using a switch, t h e subject indicates whether or button. Therefore, method of limits testing is reaction time-in-
not the stimulus is perceived with his free left hand.The computerized clusive, while forced-choice methods are reaction time-exclu-
components are in the foreground. sive. Although reaction time-inclusive methods may produce
Chapter 11 QUANTITATIVE SENSORY TESTING — 431
higher sensory thresholds, they can take less time to complete, 25 | . . . r-

and theoretically will minimize errors resulting from patient s t i m u 1 i —•—
boredom. Producing accurate and reliable data in a reasonable measured threshold
n u 1 1 s t i mu1i +
time frame poses a challenge in QST. Although reaction time-

inclusive methodology is more efficient from a time standpoint, 20 -
it poses special challenges for some stimuli such as heat-pain,
since a ramp method could potentially cause skin burns.
o
For response time-inclusive methods, the vibrating or thermal z
stimulus intensity increases at a linear or exponential rate from 15 -
a neutral starting point. The ramp is then stopped when the sub-
ject perceives the specified sensation. Usually several trains are
delivered, and a mean is calculated. Some investigators have
found that a reliable mean vibration threshold can be calculated
after four trains. 19
In response-time exclusive or forced choice testing, the sub-

ject is asked whether or not a stimulus of predetermined inten-
sity is felt. In general, if the subject perceives the first stimulus, 5
the next one will be of smaller intensity. A larger stimulus
would follow a negative response. The difference between the
two consecutive stimuli is termed a "step." The sophisticated
equipment and calibration systems required to generate discrete f f f s f
0 - -» ..+ . . . + . . . + +
steps are now commercially available. There are several differ-
22
ent approaches to response time-exclusive testing. 0 5 10 15 20

Test number
In the method of levels, each step is increased or decreased
by a factor of 2 depending on the subject's prior response. For FIGURE 11-2. Graphic depiction of the 4-2-1 stepping algo-
instance, if the response to the first stimulus is negative, the rithm. In this case, cooling perception was being assessed in the hand.
stimulus is increased two-fold. If the first response is positive, Testing begins on t h e left side of the figure at J N D level 13. Since t h e
the stimulus is reduced by 50%. The test is completed once the patient perceives t h e stimulus (s = success, f = failure t o perceive), t h e
step size has fallen to a certain level. The threshold is the mean intensity of each successive stimulus is reduced 4 JNDs, until a turning
between the last positive and negative response. This type of point occurs at J N D level 5. At that point, steps of 2 JNDs are intro-
testing can be completed rapidly and, therefore, is suitable for duced, followed by I J N D after the next turning point. Five null stimuli
thermal pain testing. However, it is prone to error if the sub-
85 8 6
were inserted at random in t h e testing sequence and are depicted by

ject is not fully attentive to each stimulus. the zero line on t h e y-axis.The patient responded inappropriately t o
In the staircase method, three steps are predetermined: gross, only t h e fourth null stimulus. Results were normal in this instance.
medium, and fine. Gross steps are used until the first turn point,
followed by medium steps, and then fine steps. Turn points refer
to the stimulus intensity where the subject either no longer per- be performed with two stimulators, one active and one inactive,
ceives or begins to perceive the stimulus. After four negative re- instead of using a single stimulator. - 3 6
sponses using fine steps, the testing is terminated and a mean is

calculated from the "yes" and "no" responses. Two or more
29
staircase sequences can be combined randomly in the same train REFERENCE DATA AND OTHER
to eliminate the educated guesswork that may influence subject TESTING ISSUES
responses once they recognize the basic principle of staircase
testing. While control data are available for a variety of testing meth-
The 4-2-1 stepping algorithm applies the concept of JNDs. ods, sensory modalities, body sites, and patient age, - - - - - - - 4 9 14 38 46 55 87 88
Testing begins at the J N D level of 13, the midpoint of the 25 it is crucial that these values only be used if the testing condi-
JNDs that divide the full range of sensation for the given modal- tions are carefully replicated in the QST laboratory. The test- 89
ity (Fig. 11-2). Depending on the subject's response to JND ing site, rate of stimulus change, and stimulator size can all
level 13, the next step will increase or decrease four JNDs. After influence the results. In general, thresholds increase slightly
42 2
the first turn point, the step is reduced to two JNDs, and after with age, there are no significant gender or side-to-side differ-
the next turn point, to one JND. The test ends after 20 stimuli ences, and response-time inclusive methods produce higher
are delivered, five of which are null. The threshold is calcu-
21
threshold values. Data are presented with an upper limit of
89
lated as a mean of the turn points at the one-JND steps. This normal representing 2 SDs or the 95th percentile. Thresholds 18
method also can be used for heat-pain testing, with the stimulus beyond this level are considered abnormal. For thermal pain
sequentially increasing to the specified point followed by a fall testing, reference data are presented with lower and upper limits
back to adaptation. To prevent skin injury, the ceiling tempera- of normal to represent hyperalgesia and hypoesthesia, respec-
ture is set at 50°C. tively. QST for thermal thresholds is feasible and reproducible
Forced-choice testing can also be done with dual stimuli, in children as young as age 3 years. - 42 46
asking the subject to choose which of two time epochs con- Overall, the sensitivity of the various QST methodologies has
tained a stimulus. Since by chance alone, the subject will be not differed significantly in studies. Results from method of
correct 50% of the time, a minimum of three of four correct re- limits and forced choice testing are similar in diabetic pa-
sponses is generally required before the stimulus intensity is tients. - There is good agreement between the 4-2-1 stepping
10 59
lowered. 15 5 0
Alternatively, the dual stimulus methodology can protocol and other types of forced-choice testing for vibratory
25 j . . . r- sessions whenever possible. In most testing conditions, reaction

stimul i -*— time-exclusive methods that use single stimuli are preferred as
measured threshold they are less prone to threshold overestimation and still can be
nul1 stimulI +
5
completed relatively quickly. The one exception is thermal pain
measurements where a method of limits algorithm may be pre-
ferred to reduce the number of painful stimuli delivered and
minimize the influence of prior painful stimuli on pain percep-
o
z: tion and subsequent test results. Dyck et al. have described a
89
—>
heat-pain algorithm that uses nonrepeating ascending ramps
with null stimuli. The algorithm demonstrated good repro-
24
ducibility and is able to document altered pain thresholds in a

variety of situations including diabetic polyneuropathy.
c
CO
CLINICAL APPLICATIONS OF QST

Q)
DIABETIC NEUROPATHY
Since it is so common and is frequently the subject of clinical
trials, diabetic neuropathy has been extensively studied with
QST. The application of QST in the evaluation of diabetic neu-
ropathy was recommended in an international conference. 1
0 5 10 15 20 Because of the insensitivity of routine nerve conduction studies

T e s t number to small-fiber impairment, QST of thermal thresholds can be par-
ticularly useful. Thermal threshold abnormalities may be found
94
FIGURE 11-3. Example of an unreliable test using the 4-2-1 prior to the development of neuropathic symptoms. - - QST 30 52 93
stepping algorithm to determine vibratory thresholds in the remains more sensitive than electrophysiologic studies even in
foot.The patient responded appropriately (f = failure to perceive) to
later stages of diabetic neuropathy when there is significant large
the first two null stimuli depicted by the zero line on the y-axis.
fiber involvement. - Thermal sensitivity was abnormal in 86%
37 66
However, the patient claimed to perceive (s = success) the last two

of 142 type-1 diabetics, including approximately 50% of patients
null stimuli, negating the reliability of the results. In this setting, the
who had normal N C S . In another study of 280 diabetics, 7 8 %
65
testing procedure was again reviewed with the subject and the algo-
demonstrated abnormalities on warm-cold testing and 39% on
rithm was restarted.
heat-pain. Of the 46 diabetics with normal physical examina-
66
tions, 5 7 % had abnormal thermal testing and only 2 0 % had

and thermal thresholds in healthy subjects and neuropathy pa- nerve conduction study abnormalities. In a study of 81 diabetics,
tients. In addition, the 4-2-1 stepping method takes only 25%
21
abnormalities on vibratory testing were seen in 88% of patients,
of the time required to perform other forced-choice methods. warm sensation in 78%, and cold in 7 7 % . Combining thermal 81
Experience has shown that method of limits testing is the fastest and vibratory testing, the sensitivity for detecting neuropathy
of the algorithms. While test durations are similar for the differ- was over 90% with a specificity of 77-86%. Typically, both ther-
ent methodologies in normal subjects, forced choice may take
87
mal and vibratory testing are abnormal; however, up to one-third
six times longer than method of limits testing in patients. 10
of patients have isolated thermal abnormalities. 37
Since QST is based on subjective responses, assessing the test Some investigators have proposed that QST serve as one of five
subject's cooperativeness is an important issue. The most evaluations in the diagnosis of diabetic neuropathy, along with
straightforward approach is to include random null stimuli in the scored symptoms, neurologic signs, nerve conduction studies, and
test sequence, something that is automatically done with comput- autonomic testing. Abnormalities in two or more of these evalua-
20
erized systems. If subjects respond to null stimuli, the test instructions can establish the diagnosis, with NCS or autonomic studies
tions should be reviewed with them again. QST data will not be
22
accounting for at least one of the abnormalities. Given the subjec-
reliable in those subjects who repeatedly respond to null stimuli tive nature of QST, experts have cautioned against its use as the
(Fig. 11-3). Yarnitsky and colleagues have proposed the use of
1 52 9
sole diagnostic criterion for diabetic neuropathy. From a staging 25
variance measurements of consecutive thermal stimuli in deter- standpoint, QST has been found to correlate well with the degree
mining the level of subject cooperation with the testing. 87 8 8
of small-fiber impairment on examination and the neuropathy 39
When considering the reliability or reproducibility of QST disability score. In several studies, vibratory thresholds have
91
results, little data have been published. Reaction time-exclusive demonstrated a close relationship to a variety of measures on
methods, however, are generally favored. Reaction time-inclu- motor and sensory NCS - - In individual patients, QST thresh-
48 57 65
sive methods have demonstrated large inter-session differences olds appear to increase over time, especially in the setting of poor
and problems with repeatability. - Using method of limits test-
28 87
glucose control, - but do not correlate well with the duration of
41 75
ing, intersession differences were 150%, compared to only 5%

28
diabetes in general. In a study of 405 diabetics followed for 10
65
using a forced-choice algorithm. Still, it is important to realize

50
years or more, toe vibration thresholds were a better predictor of
that no matter what algorithm is employed, consistency of the future diabetic foot complications than standard sensory and reflex
examiner and the testing environment are crucial in generating findings on neurologic examination. 11
reproducible results. The examination room should be quiet and Although most studies have supported a role for QST in the di-
free of distraction, instructions should be read at a modest pace, agnosis and staging of diabetic neuropathy, the application of QST
and the same examiner should administer the test in subsequent in the longitudinal assessment of patients enrolled in clinical trials
1.R, 2. Ipsilateral R 2
Amplitude »9«
Amplitude
I
'Latency - Duration -
- Latency -
— Average
Amplitude
Imulus
artifact
3. Contralateral R 0
Figure 6-7. Blink reflex p a t h w a y . T h e afferent impulse traverses the supraorbital nerve and then enters the pons t o divide into a rostral and
caudal pathway.The rostral fibers synapse in t h e principal sensory nucleus and then descend t o synapse with the facial nucleus.The fibers not con-
necting with the principal sensory nucleus descend in the lateral aspect of the medulla to then send a contralateral and ipsilateral group of fibers
rostrally t o synapse with both the left and right facial nucleus.The facial nerve then conveys t h e initial ipsilateral and s h o r t e r pathway t o generate
the Rj while the longer bilateral pathway produces t h e t w o R waveforms. (From Kimura J: Electrodiagnosis in Diseases of N e r v e and Muscle:
2
Principles and Practice. Philadelphia, F.A. Davis, 1989, with permission.)
E-l. Two E-l electrodes are located bilaterally on the pa- peripheral nerves are recommended. The current intensity of the
tient. Each is positioned as if one is performing a facial nerve stimulator is slowly increased until stable, reproducible, and
study to the orbicularis oculi (Fig. 6-8). maximal RI and R2 responses are obtained. Because the blink
E-2. There are a number of positions one may choose for E-2. reflex involves a multisynaptic pathway, there is some variabil-
It is possible to locate E-2 on the temporal region bilaterally or 125
ity between successive activations of the supraorbital nerve (es-
just superior to the nasalis muscle (Fig. 6-8). One can also use
141
pecially with respect to the R2) and at least 8-10 responses
a single E-l electrode placed on the tip of the nose and using a should be elicited with the shortest used for measurement.
"jumper" cable connect it to both E-2 ports on the instrument's Following completion of the study on one side, the contralateral
amplifier, i.e., a common reference for both channels. side is stimulated and responses recorded. Care should be exer-
G r o u n d . The ground electrode can be placed on the chin, cised at all times as it is easy to concentrate on the CRT screen
forehead, or cheek. and allow the cathode to slip inferiorly into the patient's eye.
S t i m u l a t i o n . The cathode is positioned directly over the A particularly annoying problem during blink reflex studies
supraorbital notch, i.e., the supraorbital nerve (Fig. 6-8). With is that of stimulus artifact that can obscure the RI response. To
the cathode in this location, the anode is directed superiorly and minimize stimulus artifact production in the face it is crucial to
laterally. It is important not to rotate the anode too far medially remove all makeup, facial oils, and perspiration. This needs to
as the contralateral supraorbital nerve will become activated be accomplished for the entire face and not just around the stim-
through anodal break excitation, thus producing bilateral RI re- ulus site as current from the stimulator will follow the path of
sponses and confusing the diagnostic utility of the blink reflex. least resistance and may still arrive at the electrodes prior to the
It may be necessary to rotate the anode about the cathode to op- response, resulting in possible RI contamination. Attention to
timize the effects of stimulus artifact, which can be a problem detail is especially important in attempting to generate optimal
because of the close association between the cathode and blink reflex responses.
recording electrode. As long as the above caution is kept in I n s t r u m e n t a t i o n P a r a m e t e r s . The R1 and R2 response is
mind regarding anodal break excitation, there should be no dif- relatively small and requires a sensitivity of 50-200 p.V/div. The
ficulty with anode rotation. The stimulation site may be some- delayed R2 necessitates a sweep speed of 10 ms/div. Filter set-
what uncomfortable for patients and a slow stimulus rate of 1 tings are those used for routine motor studies.
Hz is preferable. Additionally, the stimulator prongs should rest Reference V a l u e s . Reference values are provided for both
lightly on the supraorbital nerve as this is a rather sensitive the ipsilateral RI and R2, as well as the contralateral R2 (Table
structure. Stimulator parameters similar to those used for other 6-4). Because of the variability of the responses, three standard
S present. On the other hand, should the R/D ratio fall below
125
the expected reference values D is prolonged, implying an ab-

normality of the facial nerve.
The same recording and instrumentation parameters can be
used for recording the cornea reflex. - The stimulation is
99 132
easiest done electrically with a cotton thread soaked in saline

and positioned on the sclera with anode as a surface electrode
near the eye. The main difference from the blink reflex is that
the afferent arc consists of thin (A-delta) fibers with a slow con-
duction. The RI is normally not present and a bilateral response
of 35-50 ms is expected. It is also possible to measure reflexes
limited to the sensory and motor part of the trigeminal nerve
itself. This can be done by eliciting the masseter tendon reflex
by tapping on the chin with a reflex hammer and recording the
responses of the masseter muscle bilaterally with surface elec-
trodes. The advantage of this technique in comparison to the
clinical examination is that unilateral absence or delay of the
reflex can be s h o w n . The last reflex is the masseter in-
166
hibitory reflex. The mentalis nerve is stimulated on the left and

right with surface electrodes and recording is done bilaterally
on the masseter with surface electrodes while maximally
clenching the jaws. After unilateral stimulation, two phases of
EMG interruption (silent period) on both sides occur at 10-15
Gd and 40-50 ms latencies, respectively.
Gd
Cranial N e r v e XI (Spinal Accessory Nerve)
Figure 6-8. Blink reflex stimulation. Stimulation (S) of the right The cervical portion of the eleventh cranial or spinal acces-
supraorbital nerve t h a t can usually b e palpated in t h e supraorbital sory nerve originates from cervical levels C 1 - C 5 . Individual
90
notch. Bilateral recording from t h e orbicularis oculi with E-1 (R ) and a

nerve rootlets from these cervical segments proceed superiorly,
E-2 (R ) electrodes positioned for optimal recording of the blink reflex
r
fusing with each sequentially rostral segment until the spinal
responses. (From Ma DM, Liveson JA: N e r v e Conduction Handbook. accessory nerve trunk is formed. It continues to course rostral- i
Philadelphia, F.A. Davis, 1983, with permission.) ward entering the cranium through the foramen magnum. While
intracranial, this nerve joins with nerve fibers arising from the
tenth cranial nerve to exit the skull by way of the jugular fora- .
deviations are used to compute reference values. Temperature is men. Once extracranial, the spinal accessory nerve separates
not routinely measured as the face is usually quite warm. from the tenth-nerve fibers to descend into the neck to innervate
Distance is inconsequential in this study and therefore the the sternocleidomastoid and trapezius muscles. The spinal ac-
above-noted anatomic landmarks are used. One can also com- cessory nerve is joined by additional nerve fibers from cervical
bine information from the facial and trigeminal nerves to arrive segments C1-C4 via a communication with the cervical plexus
at an optimal ratio of latencies to the orbicularis oculi muscle. while in the neck region. These fibers preferentially innervate
Specifically, the indirect facial response through supraorbital the trapezius muscle after joining the spinal accessory nerve.
nerve excitation RI (R) is divided by the facial nerve latency to After innervating the sternocleidomastoid muscle, the spinal ac-
direct activation of the facial nerve (D) to arrive at R/D (Table cessory nerve is superficial just posterior to the posterior border
6-4). If R/D exceeds the normal limits because RI is prolonged of this muscle at approximately the muscle's mid-point. The
but D is normal, a lesion involving the trigeminal nerve is likely nerve then continues distally to innervate the trapezius muscle. ;
The superficial location of the spinal accessory nerve posterior I
to the sternocleidomastoid muscle allows easy access to stimu-
Table 6-4. Blink Reflex' 25
lation. As for previous NCSs, a technique using surface stimula-

Latency (ms) Amplitude (mV) tion and recording is preferred.
Ipsilateral RI 10.6 ± 0.8; < 13.1 0.38 ± 0.23 Recording Electrodes. E-L A surface E-l electrode is lo-
Ipsilateral R2 31.3 ± 3 . 3 3 ; < 41.0 0.53 ± 0.24 cated on the trapezius muscle approximately 5 cm lateral to the
C7 spinous process on a line between this structure and the
Contralateral R2 31.6 ± 3.78; < 43.0 0.49 ± 0.24
acromion.
L/R;RI 1.2 E-2. This electrode is located over a lower thoracic spinous
L/R:S; R2 5.0 process. One may also position this electrode on the acromion.
L/LR/R; R2 8.0 Ground. Although one investigator recommends that
ground be located on the acromion, positioning it between the
141
R/D 2.6-4.6 stimulus and E-l is preferred.

L/R, left-right difference for shortest RI latencies; L/R:S, left-right difference for Stimulation. The cathode is located approximately 1-2 cm
R2 responses simultaneously obtained for a particular stimulus; L/LR/R, R2 dif- posterior to the posterior margin of the sternocleidomastoid
ferences for the same side obtained with opposite-side stimuli, e.g., R2 latency
on the right obtained with right-sided stimulation subtracted from R2 latency muscle mid-way between the mastoid process and the supraster-
on the right obtained with left-sided stimulation; R/D, RI divided by direct facial nal notch. This location approximates the superior margin of the
nerve response. thyroid cartilage. The anode is directed superior to the cathode.
Chapter 11 QUANTITATIVE SENSORY TESTING — 4 3 3
is likely to be of greatest value. Sensory thresholds have decreased present with distal paresthesias, symmetric stocking-glove sen-
in patients who received continuous subcutaneous insulin infu- sory loss, and gait ataxia. Vibratory thresholds were checked 40
sions to maintain tight control of blood glucose levels. In a 6

in 42 post-gastrectomy patients with vitamin B levels less than 1 2
placebo-controlled trial, the Nerve Growth Factor Study Group 200 pg/ml. Some patients had deficits that could be explained
71
found that six months of treatment with recombinant human nerve by myelopathy alone whereas others had evidence of combined
growth factor resulted in a trend toward improvement in various central and peripheral degeneration. Vibratory thresholds
neuropathy measures, including cooling and heat-pain assess- recorded from the foot decreased significantly in the 25 patients
ments on QST. However, of the QST measures, only the cooling
2
who were compliant with vitamin B injections. Likewise, neu- 1 2
detection threshold was significantly different when compared to ropathy symptoms improved in this group. Neither vibratory
placebo. In the Rochester Diabetic Neuropathy Study, QST was of thresholds or neuropathic symptoms improved in the remaining
greatest value in measuring neuropathy progression when com- patients.
bined with other assessments including examination findings,
NCS, and autonomic testing. QST has also been employed in di-
25
ALCOHOLIC NEUROPATHY
abetic neuropathy treatment trials of gamma-linoleic acid and 54
aldose reductase inhibitors. Unfortunately, vibratory and thermal

31
The neuropathy associated with longstanding alcohol abuse
testing was not found to be a sensitive indicator of improvement in involves both large and small sensory fibers, supporting a role
26 insulin-dependent diabetes patients who received combined for both vibratory and thermal QST in this clinical setting. In 92
pancreas and kidney transplantation for end-stage nephropathy. 100 men with 11-13 years of heavy alcohol consumption, foot
QST did not mirror the improvement seen in clinical and electro- vibratory thresholds were significantly higher than in 52 control
physiological parameters in six patients with functioning trans- subjects. Higher thresholds were seen in the alcoholic group,
74
plants at a mean postoperative interval of 41 months. 79

whether or not they had neuropathic symptoms, suggesting that
QST could be helpful in detecting subclinical neuropathy.
UREMIC NEUROPATHY Thermal testing was performed on a group of 50 patients with
chronic alcohol abuse of at least 7 years. Cold thresholds were
Uremic neuropathy is an axonal sensorimotor polyneuropa- increased in 62%, warm thresholds in 24%, and heat-pain in
thy that predominantly affects large myelinated fibers. As a 2 2 % . Complete hypoesthesia to heat-pain or cold-pain, and a
43
result, vibratory thresholds have been the focus of QST in this paradoxical sensation of warmth to cold stimuli were seen in
j form of neuropathy. Vibratory thresholds were higher in 97 pa- about 10% of patients each.
I tients with chronic renal failure compared to a large control
group. Vibratory abnormalities correlated with the degree of HIV-RELATED NEUROPATHY
67
renal impairment in men but not in women. Overall, the vibra-

tory testing results reflected the clinical severity of the neuropa- QST has been applied in HIV-infected patients, primarily to
thy. NCS were actually more sensitive in detecting neuropathy, characterize the distal, symmetric, and painful neuropathy that
with 82% of patients showing abnormalities compared to only develops in as many as one-third of patients with advanced dis-
32% on QST In a study of 64 nondiabetic patients with chronic 62.72.82
e a s e j e s s neurotoxicity from antiviral agents.
a n ( t 0 a s S
renal failure, clinical signs of neuropathy were present in Thermal testing, especially for warmth perception, was more
65%. - NCS were again more sensitive than QST in detecting
60 78
sensitive than NCS in detecting sensory nerve impairment in
neuropathy, with 90% of patients demonstrating electrophysio- HIV patients. Thermal thresholds were abnormal in 50% of
82
logic abnormalities compared to 36% on vibratory testing and symptomatic patients with normal NCS and in 30% of patients
30% on thermal testing. However, vibratory thresholds showed without symptoms. In another study of 179 HIV-seropositive
the closest correlation with clinical findings. In another study,
60
patients, vibratory thresholds had a similar rate of abnormality
vibratory testing was found to be more sensitive than using a as N C S . As expected, QST abnormalities were more common
36
tuning fork in detecting large-fiber sensory impairment in both in the 28 subjects who had progressed to AIDS-related complex
uremic and alcoholic patients. Several studies have docu-
44
(ARC) or AIDS, with elevated vibratory thresholds found in
mented stabilization or improvement of vibration thresholds in 36% of these patients. Vibratory thresholds were more sensitive
uremic patients receiving hemodialysis. - 12 77
than NCS in screening for subclinical neuropathy related to the
While thermal testing has largely been found to have less antiviral agent 2',3'-dideoxycytidine (ddC), but once patients
value than vibratory testing in uremic neuropathy, in one study were symptomatic, both NCS and vibratory thresholds were
42% of patients with end-stage renal failure developed a sensa- abnormal in the large majority of patients. Vibratory threshold 5
tion of warmth or heat to cold stimuli delivered to the feet. 90

abnormalities were a rare finding in patients receiving zidovu-
This paradoxical response is seen in less than 10% of normals. dine. However, thermal thresholds, which may be a more ap-
8
The paradoxical response was the only QST abnormality in propriate test for the painful, hyperesthetic symptoms caused by
11 % of patients and correlated with serum creatinine levels. The this drug, were not performed.
authors suggested this response could serve as an indicator of
worsening renal failure or insufficient hemodialysis. Heat-pain CRYPTOGENIC SENSORY POLYNEUROPATHY (CSPN)
thresholds are not particularly sensitive in the detection of
uremic neuropathy, being abnormal in less than 10% of patients Chronic sensory or sensorimotor polyneuropathy of unknown
with clinically-evident neuropathy. 90
cause is a common disorder that tends to present in the sixth to
seventh decade of life. These patients represent anywhere 83
VITAMIN B, DEFICIENCY STATES

2
from 10% to one-third of all polyneuropathies seen in referral
centers. - - - - Approximately 70-80% of these neuropathies
16 58 63 68 70
Vitamin B deficiency may lead to both central and periph-

I 2
are painful - - - and when small-fiber involvement predomi-
33 34 68 84
eral nerve dysfunction. Patients with this neuropathy typically nates, electrophysiologic testing may be normal in nearly 5 0 %
of patients. Therefore, QST, especially thermal testing, could

33
mediate warmth—are less affected by compression than small
serve as a valuable laboratory assessment in these patients. unmyelinated axons that subserve cold stimuli.
In our experience, QST for cold and vibration detection
thresholds has a slightly higher yield than NCS in the detection Illustrative Case
of abnormalities in CSPN. Sensory NCS were abnormal in
84
History. A 46-year-old man presented with a 20-month his-
77% of our patients, compared to QST abnormalities in 8 5 % . 84
tory of uncomfortable sensations in his feet that began during a
This figure is similar to the 88% frequency of QST abnormali- hospitalization for acute pancreatitis. He received total par-
ties in a report of patients with idiopathic painful sensory neu- enteral nutrition for three months with resolution of the foot
ropathy. In our CSPN population, both cold and vibration
69
symptoms. However, several months later, the burning, stinging
thresholds were abnormal in two-thirds of patients. Abnormal 84
and shock-like sensations returned, spreading from his toes up
cold but normal vibration thresholds were seen in 9 patients, to the heels. Similar dysesthesias and numbness also developed
but abnormal vibration and normal cold thresholds in only two, in the fingertips of both hands. He denied weakness, only re-
perhaps reflecting the greater degree of small-fiber impairment porting that his feet felt "tired" by the end of the day. His bowel
in these patients. In approximately 10% of patients with abnor- and bladder function were unchanged. Medical history was no-
mal QST, there was no evidence of neuropathy on N C S . table for a single episode of pancreatitis 2 years earlier that re-
Studies of idiopathic small-fiber neuropathies in the setting of mained unexplained. He underwent a hernia repair 12 years
normal electrophysiologic testing have demonstrated thermal earlier and had pneumonia as a child. There was no history of
threshold abnormalities in anywhere from 5 7 % to 100% of47
diabetes or ethanol abuse. He was not on any medications.
patients. In another study describing 30 patients with painful,
51
P h y s i c a l E x a m i n a t i o n . On neurologic examination, cra-
burning feet, only one had abnormal N C S . On the other hand,
73
nial nerves were intact. Muscle bulk and tone were normal,
either warming or cooling thresholds were elevated in 12 and strength was full in both the upper and lower limbs, in-
(40%) of these patients. cluding toe extensors and flexors. Deep tendon reflexes were
grade 2 except for absent responses at both ankles. Plantar re-
OTHER APPLICATIONS sponses were flexor bilaterally. On sensory testing, light
touch, temperature, and pinprick were reduced in both feet up
QST has been evaluated in a variety of other neuropathic to the ankle level. Timed vibration was slightly reduced in the
states, including carpal tunnel syndrome and lumbosacral toes. Position sense was intact. His gait was normal without a
radiculopathies. Since these disorders are relatively common, Romberg sign.
QST has potential attractiveness as a mass screening tool in L a b o r a t o r y D a t a . Laboratory testing was unremarkable, in-
these clinical scenarios. Unfortunately, the sensitivity - -
92 32 35 64
cluding 2-hour glucose tolerance testing, thyroid function, vita-
and specificity of QST - in carpal tunnel syndrome has been
7 49
min B , serum protein and immunofixation electrophoresis,
l 2
poor in several studies. For instance, thresholds in the fifth ESR, ANA, RF, SSA, SSB, HIV, and syphilis serologies. Anti-
finger are often abnormal in patients with carpal tunnel syn- Hu antibody testing was negative.
drome, raising serious questions about the utility of QST in this N e r v e C o n d u c t i o n S t u d i e s . Nerve conduction studies were-
mononeuropathy. - Overall, there is little evidence to support
7 49
performed in the right upper and lower limbs and were normal.
the use of QST in the diagnosis of carpal tunnel syndrome. It
DSL S Amp DML M Amp NCV Fwave
has a lower sensitivity than standard electrophysiologic testing
and often fails to demonstrate a pattern consistent with Nerve (ms) (rlV) (ms) (mV) (m/s) (ms)
mononeuropathy. R median 3.5 44.3 3.9 6.9 56.4 28.3
R ulnar 3.0 31.3 3.1 8.4 58.5 30.8
The role of QST in evaluation of lumbosacral radiculopathy
R peroneal 3.8 8.6 44.4 50.5
has not been established. Thermal thresholds for warmth were
R tibial 4.4 9.9 43.6 52.7
increased in dermatomes ipsilateral to the root compression
R sural 3.8 9.1
while heat-pain thresholds were normal ipsilaterally but surpris-
ingly decreased in contralateral dermatomes. The authors pro-
76
DSL, distal sensory latency; S Amp, sensory amplitude, DML,
posed that this discrepancy is related to different responses of distal motor latency; M Amp, motor amplitude; NCV, nerve
fiber types to chronic root compression. Compression of fibers conduction velocity; ms, milliseconds; pV, microvolts; m/s,
mediating warmth would manifest only ipsilateral findings. meters per second. Motor and sensory amplitudes are measured
Nociceptive fiber compression, however, would produce bilat- baseline-to-peak. Sensory latencies are measured to peak, and
eral changes. The decreased heat-pain threshold on the con- motor latencies are measured to initial negative onset.
tralateral side could be related to a loss of inhibition resulting
from large fiber compression. On the ipsilateral side, compres- N e e d l e E l e c t r o m y o g r a p h y . A needle electromyographic in-
sion of both large fibers and nociceptive small fibers would ef- vestigation of distal muscles in the right upper and lower limbs
fectively produce a sensory equilibrium, with no significant was normal, except for rare fasciculation potentials in the ab-
change in the heat-pain threshold. In another study, warm ductor hallucis.
thresholds were significantly higher on the symptomatic side
compared to the asymptomatic limb in 40 patients with L5 or Rest MUAP Recruit-
SI radiculopathies. Cold thresholds were also significantly
95 Muscle activity Morphology ment
higher on the symptomatic side in those patients with surgically R gastrocnemius Silent Normal Normal
confirmed disk herniation. Since warm thresholds were affected R tibialis anterior Silent Normal Normal
to a greater degree than cold thresholds, the authors hypothe- R abductor hallucis Silent Normal Normal
sized that inflammation plays a greater role than compression R first doral interosseous Silent Normal Normal
in the generation of sciatic pain. This conclusion is based on Q u a n t i t a t i v e S e n s o r y T e s t i n g . Q S T was performed on both
prior observations that unmyelinated axons—such as those that hands and both feet using the 4-2-1 stepping algorithm (Fig. 11-4).
Chapter I ! QUANTITATIVE SENSORY TESTING — 4 3 5
over the next year. No further diagnostic work-up is recom-

s t imul i - 0 — mended at this time. Treatment is largely symptomatic (e.g,
measured threshold
antiepileptic medications, tricyclic antidepressants, tramadol).
null stimuli 4¬
20 - Comments
This case illustrates the clinical utility of QST. The patient's
history and examination are suggestive of a predominantly small-
fiber sensory neuropathy. The motor and sensory nerve conduc-
tion studies as well as electromyography were completely
normal. As noted previously, QST is more sensitive in evaluating
individuals with small fiber neuropathies than routine nerve con-
10 -
duction studies, which mainly assess large myelinated fibers. In
this case, the abnormal QST was helpful in confirming the clini-
cal impression that the patient had a small-fiber neuropathy.
5 - Despite an extensive laboratory work-up for causes of small-fiber
neuropathy, no etiology was determined and he was diagnosed
with cryptogenic or idiopathic sensory polyneuropathy (see
f f f f f Chapter 23). The patient's symptoms have remained relatively
L '. I !_ '. I ! I stable and have not interfered with his daily activities. Therefore,
0 5 10 15 20 he opted not to start pharmacotherapy for symptom relief.
T e s t number
FIGURE / / - 4 . Abnormal cooling perception in the right

hand using the 4-2-1 stepping algorithm.Testing begins on t h e CONCLUSIONS
left side of the figure at JND level 13.The testing results in a computed
J N D of 13.17, corresponding t o an age-matched percentile of > 99 While QST equipment and applications continue to evolve,
(i.e., > 99% of an age-matched population have better cooling percep- this relatively new technology has already made valuable contri-
tion in the hand than this patient). butions to screening, natural history, and drug therapy studies in
several sensory disorders. Despite the introduction of computer-
ized equipment and efforts to standardize testing algorithms and
The QST data were collected using CASE IV, WR Medical generate site and age-related reference data, careful attention by
Electronics, Stillwater, MN. The QST thresholds considered are the examiner to the methodology and testing environment re-
abnormal if they are greater than the 95th percentile (i.e., > 95% mains of utmost importance in generating reproducible data. In
of the population have better sensory perception). Abnormal addition, results from QST are based on psychophysical re-
values are shown in bold. sponses and are subject to issues of patient attention and cooper-
ation. For these and other reasons, clinicians should not rely on
R hand L hand R foot L foot
QST findings alone to diagnose peripheral or central neurologic
Cooling
disease. When QST is used in multicenter investigations, the
26
JND 13.1 7 15.2 16.1

same equipment and methodology should be used at each center,
Percentile 99 75 97 96
and inter-rater reliability between the clinical evaluators should
be demonstrated on control and possibly disease subjects prior to
Vibration
the study. These issues are equally, if not more, relevant to rou-
JND 6.8 3 21 15.7 tine application of QST in clinical practice. The generation of
Percentile 30 1 95 50 reference values and confirmation of reproducibility are likely to
be even greater challenges to community practices than to acade-
Summary of Findings
mic centers. While QST can provide useful measurements of
1. Normal right median, ulnar, and sural SNAPs. sensory function in individual patients, the interpretation of an
2. Normal right median, ulnar, peroneal, and tibial CMAPs. abnormal result should always be made with caution. Given the
3. Normal EMG of selected distal muscles in the right leg test's subjective nature, the potential for results to be influenced
and arm. by poor attention, reduced cooperation, or psychogenic factors
4. QST revealed abnormal cooling perception in the right needs to be considered. Finally, analagous to electrodiagnostic
80
hand and bilateral feet. Cooling perception was normal in the studies, QST is best considered an extension to the neurologic
left hand. examination and should always be interpreted in the context of
5. QST revealed borderline vibratory perception only in the the patient's clinical presentation.
right foot. Vibratory perception was normal in the left foot and
both hands.
Impression REFERENCES
On the basis of normal nerve conduction studies but abnormal 1. Anonymous: Report and recommendations of the San Antonio Conference on
cooling perception threshold on QST, the patient has a small- Diabetic Neuropathy. Diabetes 1988;37:1000-1004.
fiber sensory neuropathy. He already underwent extensive work- 2. Apfel SC, Kessler JA, Adornato BT, et al: Recombinant human nerve growth
factor in the treatment of diabetic polyneuropathy. Neurology 1998;51:695-702.
up to look for identifiable causes of painful sensory neuropathy
3. Arezzo JC, Schaumberg HH, Laudadio C: Thermal sensitivity tester: Device for
and none were found. He falls into the category of cryptogenic quantitative assessment of thermal sense in diabetic neuropathy. Diabetes
or idiopathic sensory neuropathy. He remained stable on follow-up 1986;35:590-592.
4. Bartlett G, Stewart JD, Tamblyn R, Abrahamowicz M: Normal distributions of 38. Hagander LG, Midani HA, Kuskowski MA, Parry GJ: Quantitative sensory test-
thermal and vibration sensory thresholds. Muscle Nerve 1998;21:367-374. ing: Effect of the site and skin temperature on thermal thresholds. Clin
5. Berger AR, Arezzo JC, Schaumburg HH, et al: 2',3'-dideoxycytidine (ddC) toxic Neurophysiol 2000;! 11:17-22.
neuropathy: A study of 52 patients. Neurology 1993;43:358-362. 39. Hansson P, Lindblom U, Lindstrom P: Graded assessment and classification of
6. Bertelsmann FW, Heimans JJ, Van Rooy JC, et al: Peripheral nerve function in impaired temperature sensibility in patients with diabetic polyneuropathy. J
patients with painful diabetic neuropathy treated with continuous subcutaneous Neurol Neurosurg Psychiatry 1991;54:527-530.
insulin infusion. J Neurol Neurosurg Psychiatry 1987;50:1337-1441. 40. Healton EB, Savage DG, Brust JCM, et al: Neurologic aspects of cobalamin de-
7. Borg K, Lindblom U: Diagnostic value of quantitative sensor}' testing (QST) in ficiency. Medicine 1991;70:229-245.
carpal tunnel syndrome. Acta Neurol Scand 1988;78:537-541. 41. Hillson RM, Hockaday TDR, Newton DJ: Hyperglycemia is one correlate of de-
8. Bozzette SA, Santangelo J, Villasana D, et al: Peripheral nerve function in per- %
terioration in vibration sense during the 5 years after diagnosis of type 2 (non-in-
sons with asymptomatic or minimally symptomatic HIV disease: Absence of zi- sulin-dependent) diabetes. Diabetologia 1984;26:122-126.
dovudine neurotoxicity. J Acquired Immune Defic Syndr 1991;4:851-855. 42. Hilz M, Glorius SE, Schweibold G, et al: Quantitative thermal perception testing
9. Claus D, Hilz MJ, Hummer 1, Neundorfer B: Methods of measurement of ther- in preschool children. Muscle Nerve 1996;19:381-383.
mal thresholds. Acta Neurol Scand 1987;76:288-296. 43. Hilz MJ, Claus D, Neundorfer B, et al: Is heat hypoalgesia a useful parameter in
10. Claus D, Hilz MJ, Neundorfer B: Thermal discrimination thresholds: A compar- quantitative thermal testing of alcoholic polyneuropathy? Muscle Nerve 1994;
ison of different methods. Acta Neurol Scand 1990;81:533-540. 17:1456-1460.
11. Coppini DV, Young PJ, Weng C, Macleod AF, Sonkesen PH: Outcome on dia- 44. Hilz MJ, Zimmermann P, Rosl G, et al: Vibrameter testing facilitates the diagno-
betic foot complications in relation to clinical examination and quantitative sen- sis of uremic and alcoholic polyneuropathy. Acta Neurol Scand I995;92:
sory testing: A case-control study. Diabet Med 1998;15:765-771. 486-490.
12. Daniel CR, Bower JD, Pearson JE: Vibrometry and uremic peripheral neuropa- 45. Hilz MJ, Glorius SE, Beric A: Thermal perception thresholds: Influence of deter-
thy. South Med J I977;70:1311-1313. mination paradigm and reference temperature. J Neurol Sci 1995;129:135-140.
13. Dinh S, Marroquin E, Raj PP: Neuroselective quantification of allodynia by cur- 46. Hilz MJ, Stemper B, Schweibold G, et al: Quantitative thermal perception testing
rent perception threshold evaluation in CRPS (RSD) patients (abstract). Regional in 225 children and juveniles. J Clin Neurophysiol 1998;15:529-534.
Anesthesia 1997,22:44. 47. Holland NR, Crawford TO, Hauer P, et al: Small-fiber sensory neuropathies:
14. Doeland HJ, Nauta JJP, van Zandbergen JB, et al: The relationship of cold and Clinical course and neuropathology of idiopathic cases. Ann Neurol 1998;44:
warmth cutaneous sensation to age and gender. Muscle Nerve 1989;12:712-715. 47-59.
15. Dyck PJ, Zimmerman IR, O'Brien PC, et al: Introduction of automated systems 48. Horowitz SH: Correlation of near-nerve sural conduction and quantified sensory
to evaluate touch, pressure, vibration, and thermal cutaneous sensation in man. testing in patients with diabetic neuropathy. Muscle Nerve 1995;18:1202-1204.
Ann Neurol 1978;4:502-510. 49. Imai T, Matsumoto H, Minami R: Asymptomatic ulnar neuropathy in carpal
16. Dyck PJ, Oviatt KF, Lambert EH: Intensive evaluation of referred unclassified tunnel syndrome. Arch Phys Med Rehabil 1990;71:992-994.
neuropathies yields improved diagnosis. Ann Neurol 1981;10:222-226. 50. Jamal GA, Hansen S, Weir Al, Ballantyne JP: An improved automated method
17. Dyck PJ, Bushek W, Spring EM, et al: Vibratory and cooling detection thresh- for the measurement of thermal thresholds: 1. Normal subjects. J Neurol
olds compared with other tests in diagnosing and staging diabetic neuropathy. Neurosurg Psychiatry 1985;48:354-360.
Diabetes Care 1987;10:432-440. 51. Jamal GA, Hansen S, Weir Al, Ballantyne JP: The neurophysiologic investiga-
18. Dyck PJ: Detection, characterization and staging of polyneuropathy: Assessed tion of small fiber neuropathies. Muscle Nerve 1987;10:537-545.
in diabetics. Muscle Nerve 1988;11:21-32. 52. Jensen TS, Bach FW, Kastrup J, et al: Vibratory and thermal thresholds in diabet-
19. Dyck PJ, Karnes JL, Gillen DA: Comparison of algorithms of testing for use in ics with and without clinical neuropathy. Acta Neurol Scand 1991;84:326-333.
automated evaluation of sensation. Neurology 1990;40:1607-1613. 53. Kahn R: Proceedings of a Concensus Development Conference on Standardized
20. Dyck PJ, Kratz KM, Lehman KA, et al: The Rochester Diabetic Neuropathy Measures in Diabetic Neuropathy. Quantitative sensory testing. Diabetes Care
Study: Design, criteria for types of neuropathy, selection bias, and reproducibil- 1992;15:1092-1094.
ity of neuropathic tests. Neurology 1991;41:799-807. 54. Keen H, Payan J, Allawi J, et al: Treatment of diabetic neuropathy with gamma-
21. Dyck PJ, O'Brien PC, Kosanke JL, et al: A 4, 2, and 1 stepping algorithm for linoleic acid. Diabetes Care 1993;16:8-15.
quick and accurate estimation of cutaneous sensation threshold. Neurology 55. Kenshalo DR, Holmes CE, Wood PB: Warm and cool thresholds as a function of
1993;43:1508-1512. rate of stimulus temperature change. Percept Psychophys 1968;3:81-84.
22. Dyck PJ, Zimmerman I, Gillen DA, et al: Cool, warm, and heat-pain detection 56. Kenshalo DR, Bergen DC: A device to measure cutaneous temperature sensitiv-
thresholds: Testing methods and inferences about anatomic distribution of recep- ity in humans and subhuman species. J Appl Physiol 1975;39:1038-1040.
tors. Neurology 1993;43:1500-1508. 57. Klima RR, Weigand AH, DeLisa JA: Nerve conduction studies and vibration per-
23. Dyck PJ, Karnes J, O'Brien PC: Detection thresholds of cutaneous sensation in ception thresholds in diabetic and uremic neuropathy. Am J Phys Med Rehabil
humans. In Dyck PJ, Thomas PK, Griffin JW, et al (eds): Peripheral Neuropathy, 1991;70:86-90.
3rd ed. Philadelphia, W.B. Saunders, 1993, pp 706-728. 58. Konig F, Neundorfer B, Kompf D: Polyneuropathien im hoheren Lebensalter.
24. Dyck PJ, Zimmerman IR, Johnson DM, et al: A standard test of heat-pain re- Deutsch Med Wschr 1984;109:735-737.
sponses using CASE IV. J Neurol Sci 1996;136:54-63. 59. Levy D, Abraham R, Reid G: A comparison of two methods for measuring ther-
25. Dyck PJ, Davies JL, Litchy WJ, O'Brien PC: Longitudinal assessment of dia- mal thresholds in diabetic neuropathy. J Neurol Neurosurg Psychiatry 1989;52:
betic polyneuropathy using a composite score in the Rochester Diabetic 1072-1077.
Neuropathy Study cohort. Neurology 1997;49:229-239. 60. Lindblom B, Tegner R: Thermal sensitivity in uremic neuropathy. Acta Neurol
26. Dyck PJ, Dyck PJB, Kennedy WR, et al: Limitations of quantitative sensory test- Scand 1985;71:290-294.
ing when patients are biased toward a bad outcome. Neurology 1998;50:1213. 61. Liu S, Kopacz DJ, Carpenter RL: Quantitative assessment of differential sensory
27. Dyck PJ, O' Brien PC: Quantitative sensation testing in epidemiological and nerve block after lidocaine spinal anesthesia. Anesthesiology 1995;82:60-63.
therapeutic studies of peripheral neuropathy. Muscle Nerve 1999;22:659-662. 62. Marra CM, Boutin P, Collier AC: Screening for distal sensory peripheral neu-
28. Fagius J, Wahren LK: Variability of sensory threshold determination in clinical ropathy in HIV-infected persons in research and clinical settings. Neurology
use. J Neurol Sci 1981;51:11-27. 1998;51:1678-1681.
29. Fowler CJ, Carroll MB, Burns D, et al: A portable system for measuring cuta- 63. McLeod JG, Tuck RR, Pollard JD: Chronic polyneuropathy of undetermined
neous thresholds for warming and cooling. J Neurol Neurosurg Psychiatry cause. J Neurol Neurosurg Psychiatry 1984;47:530-535.
1987;50:1211-1215. 64. Merchut MP, Kelly MA, Toleikis SC: Quantitative sensory thresholds in carpal
30. Fruhstorfer H, Lindblom U, Schmidt WG: Method for quantitative estimation of tunnel syndrome. Electromyogr Clin Neurophysiol 199G;30:119-124.
thermal thresholds in patients. J Neurol Neurosurg Psychiatry 1976;39:1071-1075. 65. Navarro X, Kennedy WR, Fries TJ: Small nerve fiber dysfunction in diabetic
31. Gill JS, Williams G, Ghatei MA: Effect of the aldose reductase inhibitor, ponal- neuropathy. Muscle Nerve 1989;12:498-507.
restat, on diabetic neuropathy. Diabete Metabolisme 1990;16:296-302. 66. Navarro X, Kennedy WR: Evaluation of thermal and pain sensitivity in type-1 di-
32. Goadsby PJ, Burke D: Deficits in the function of small and large afferent fibres abetic patients. J Neurol Neurosurg Psychiatry 1991;54:60-64.
in confirmed cases of carpal tunnel syndrome. Muscle Nerve 1994;17:614-622. 67. Nielsen VK: The peripheral nerve function in chronic renal failure. IV. An analy-
33. Gorson KC. Ropper AH: Idiopathic distal small fiber neuropathy. Acta Neurol sis of the vibratory perception threshold. Acta Med Scand 1972;191:287-296.
Scand 1995;92:376-382. 68. Notermans NC. Wokke JHJ, Franssen H, et al: Chronic idiopathic polyneuropa-
34. Grahmann F. Wimerholler M, Neundorfer B: Cryptogenic polyneuropathies: An thy presenting in middle or old age: a clinical and electrophysiological study of
outpatient follow-up study. Acta Neurol Scand 1991;84:221-225. 75 patients. J Neurol Neurosurg Psychiatry 1993;56:1066-1071.
35. Grant KA, Congleton JJ. Koppa RJ: Use of motor nerve conduction testing and 69. Periquet MI, Novak V, Collins MR et al: Painful sensory neuropathy: Prospective
vibration sensitivity testing as screening tools for carpal tunnel syndrome in in- evaluation using skin biopsy. Neurology 1999;53:1641-1647.
dustry. J Hand Surg 1992; 17A:71-76. 70. Prineas J: Polyneuropathies of undetermined cause. Acta Neurol Scand 1970;
36. Gulevich SJ, Kalmijn JA, Thai LJ, et al: Sensory testing in human immunodefi- 44(suppl):l-72.
ciency virus type 1-infected men. Arch Neurol 1992;49:1281-1284. 71. Roos D: The vibration perception threshold in gastrectomized patients with low
37. Guy RJ, Clark CA, Malcolm PN, Watkins PJ: Evaluation of thermal and vibra- serum B : A clinical and biothesiometric follow-up after intensive B therapy.
i2 P
tion sensation in diabetic neuropathy. Diabetologia 1985;28:131-137. Acta Neurol Scand 1977;56:551-562.
Chapter I I QUANTITATIVE SENSORY TESTING — 4 3 7
72. Simpson DM, Tagliati M: Neurologic manifestations of HIV infection. Ann 84. Wolfe Gl, Baker NS, Amato AA, et al: Chronic cryptogenic sensory polyneu-
Intern Med 1994;121:769-785. ropathy: Clinical and laboratory characteristics. Arch Neurol 1999;56:
73. Smith SJM, Ali Z, Fowler CJ: Cutaneous thermal thresholds in patients with 540-547.
painful burning feet. J Neurol Neurosurg Psychiatry 1991;54:877-881. 85. Yarnitsky D, Ochoa JL: Release of cold-induced burning pain by block of cold-
74. Sosenko JM, Soto R, Aronson J, et al: The prevalence and extent of vibration specific afferent input. Brain 1990;! 13:893-902.
sensitivity impairment in men with chronic ethanol abuse. J Stud Alcohol 86. Yarnitsky D, Ochoa JL: Studies of heat pain sensation in man: Perception thresh-
1991;52:374-376. olds, rate of stimulus rise and reaction time. Pain 1990;40:85-91.
75. Sosenko JM, Kato M, Soto R, Bild DE: A prospective study of sensory function 87. Yarnitsky D, Sprecher E: Thermal testing: Normative data and repeatability for
in patients with type 2 diabetes. Diabet Med 1993;10:110—114. various test algorithms. J Neurol Sci 1994; 125:39^*5.
76. Strian F, Lautenbacher S, Karlbauer G: Disturbances of C-fibre-mediated sensi- 88. Yarnitsky D, Sprecher E, Tamir A: Variance of sensory threshold measurements:
bility in lumbosacral spine disease. J Neurol Neurosurg Psychiatry 1991 ;54: Discrimination of feigners from trustworthy performers. J Neurol Sci 1994; 125:
1013-1014. 186-189.
77. Tegner R, Lindholm B: Uremic polyneuropathy: Different effects of hemodialysis 89. Yarnitsky D: Quantitative sensory testing. Muscle Nerve 1997;20:198-204.
and continuous ambulatory peritoneal dialysis. Acta Med Scand 1985;218:409-416. 90. Yosipovitch G, Yarnitsky D, Mermerlstein V, et al: Paradoxical heat sensation in
78. Tegner R, Lindholm B: Vibratory perception threshold compared with nerve con- uremic polyneuropathy. Muscle Nerve 1995;18:768-771.
duction velocity in the evaluation of uremic neuropathy. Acta Neurol Scand 91. Young MJ, Boulton AJM, Macleod AF, et al: A multicenter study of the preva-
1985;71:284-289. lence of diabetic peripheral neuropathy in the United Kingdom hospital clinic
79. Trojaborg W, Smith T, Jakobsen J, Rasmussen K: Effect of pancreas and kidney population. Diabetologia 1993;36:150-154.
transplantation on the neuropathic profile in insulin-dependent diabetics with 92. Zaslansky R, Yarnitsky D: Clinical applications of quantitative sensory testing
end-stage nephropathy. Acta Neurol Scand 1994;90:5-9. (QST). J Neurol Sci 1998;153:215-238.
80. Verdugo RJ, Ochoa JL: Use and misuse of conventional electrodiagnosis, quanti- 93. Ziegler D, Mayer P, Gries FA: Evaluation of thermal, pain, and vibration sensa-
tative sensory testing, thermography, and nerve blocks in the evaluation of neuro- tion thresholds in newly diagnosed type-1 diabetic patients. J Neurol Neurosurg
pathic syndromes. Muscle Nerve 1993;16:1056-1062. Psychiatry 1988;51:1420-1424.
81. Vinik Al, Suwanwalaikorn S, Stansberry KB, et al: Quantitative measurement of 94. Ziegler D, Mayer P, Wiefels K, Gries FA: Assessment of small and large fiber
cutaneous perception in diabetic neuropathy. Muscle Nerve 1995;18:574-584. function in long-term type 1 (insulin-dependent) diabetic patients with and with-
82. Winer JB, Bang B, Clarke JR, et al: A study of neuropathy in HIV infection. out painful neuropathy. Pain 1988;34:1-10.
Quar J Med. 1992;83:473-488. 95. Zwart J-A, Sand T, Unsgaard G: Warm and cold sensory thresholds in patients
83. Wolfe Gl, Barohn RJ: Cryptogenic sensory and sensorimotor polyneuropathies. with unilateral sciatica: C fibers are more severely affected than A-delta fibers.
Semin Neurol 1998;18:105-111. Acta Neurol Scand 1998;97:41^5.
Chapter 12
Intraoperative
Neurophysiologic Monitoring
John C. King, M.D.
Jaime R. Lopez, M.D.
Tod B. Sloan, M.D., Ph.D.
CHAPTER OUTLINE
Introduction Intraoperative Neurophysiologic Monitoring: Specific

Neurophysiologic Systems Monitored Intraoperatively Applications
• Somatosensory Evoked Potentials • Brain Stem Auditory Ischemia and Electrophysiologic Studies • Carotid
Evoked Potentials • Electromyography • Nerve Conduction Endarterectomy • Intracranial Neurovascular Surgery
Studies * Motor Evoked Potentials • Electroencephalography • Aneurysm • Arteriovenous Malformations • Brain Tumor
• Visual Evoked Potentials • Equipment Requirements Surgery • Posterior Fossa and Cranial Nerve Surgery
• General Results and Efficacy • Interventional Neuroradiology Procedures • Spinal Surgery
• Peripheral Nerve Surgeries
Anesthesia Considerations
Inhalational Agents • Intravenous Analgesic Agents Technical Considerations
SEP and BAEP Recording Techniques • Operating Room
Other Intraoperative Physiologic Factors Recording • Nerve Conduction Techniques • Intraoperative
Blood Flow • Intracranial Pressure • Hypoxemia • Blood EMG Techniques
Rheology • Ventilation • Temperature • O t h e r Physiologic
Variables Conclusion
INTRODUCTION any decline in the neurophysiologic system's function. Such de-

tection may alert the surgeon to modify the operative approach
The intraoperative environment can be hazardous, and surg- with resulting preservation of function. The duration of many
eries can place various neuromusculoskeletal systems at risk. deficits, such as neurovascular compromise, nerve traction, or
Anesthesia is in part applied for the express purpose of directly direct pressure, are time-dependent for creating permanent neu-
suppressing a patient's motor and sensory function during rologic deficits. - - Therefore, if reversal of the offending
102 105 199
surgery. Though highly desirable for surgery, this suppression procedure can be rapidly accomplished, permanent deficits may
leaves the surgeon without timely clinical information to warn be avoided. - - Of course, not all loss of function is readily
102 105 199
of impending harm. Alternative methods of monitoring and reversible, such as when a nerve has been completely lysed or
safeguarding a patient's neurologic function while the patient stretched beyond axonal limits, though such stretch does lead to
remains completely anesthetized is the goal of intraoperative detectable loss of function prior to irreversible lysis.
neurophysiologic monitoring. Ideally, neurophysiologic moni- When functional loss occurs, patient suffering and the eco-
toring procedures should not add to the operative risk, but rather nomic impact on society can be enormous. Prevention is much
38
have positive demonstrable effects in reducing the incidence of more desirable than any subsequent treatment, which is often
harm to the neuromuscular system. One goal of intraoperative vastly more costly. Because of the cost-effectiveness of preven-
neurophysiologic monitoring is to permit real-time detection of tion and the current medicolegal environment, intraoperative
439
Table 12-1. Types of Neurophysiologic Monitoring and System Evaluated

Technique Pathway
1. SEP: Somatosensory evoked potentials: Posterior spinal columns and radiating
also called SSEP pathways t o the cerebral cortex
2. SCEP: Spinal cord evoked potentials Tracts along the spinal cord
3. BAEP: Brain stem auditory evoked potential: Cochlear t o auditory c o r t e x pathways
also called auditory brain stem responses (ABR)
or auditory evoked potentials (AEP)
or brain stem auditory evoked response (BAER)
4. EMG: Electromyography Anterior horn cells through peripheral nerve o r cranial
nerve m o t o r axons t o muscles
5. NCS: Nerve conduction studies Peripheral nerves and central conduction time also
recordable as well as stimulation intensity studies
6. MEP: Motor evoked potentials Central and peripheral m o t o r pathways from cortex t o muscles
7. EEG: Electroencephalography C o r t e x functional integrity
8. VEP: Visual evoked potentials Visual pathways from retina t o visual cortex
neurophysiologic monitoring for high-risk procedures has The function of those nervous system pathways listed in Table
become, or is emerging as, the standard of care. In 1995, 88% 12-1 should ideally be monitored continuously during the surgi-
of surgeons performing scoliosis procedures routinely used cal procedure, or at least during that portion of the procedure that
intraoperative electrophysiologic monitoring. Owing to in- 288
places the nervous system at greatest risk. This may involve real-
creasing demand, it is important for competent clinical neuro- time monitoring, such as over a loudspeaker or on a continuous
physiologists to be able to offer their expertise to help establish video display, or often by means of a periodically sampled, aver-
or directly provide such services. N o one is more specifically aged, and video-displayed response. Periodicity is necessary to
trained, capable of understanding the procedures, able to inter- average many dozens to hundreds of stimulations to minimize
pret electrophysiologic monitoring results, and appreciative of the random environmental electrical noise that contaminates and
their limitations and potential pitfalls than electrodiagnostic obscures the electrophysiologic signal of interest. Sources of the '
medicine consultants. electrical noise commonly encountered in the operative suite and
steps to reduce these effects are listed in Tables 12-3 to 12-6. In
NEUROPHYSIOLOGIC SYSTEMS order to permit more rapid deficit onset detection, the epochs of
MONITORED INTRAOPERATIVELY
sufficient numbers of averages to produce a reliable response are
continuously repeated during the entire or key portions of the
Many types of neurophysiologic monitoring systems and surgery. Detailed reviews exist for intraoperative monitoring
techniques are available (Table 12-1). In addition to direct elec- techniques, frequency of use, and their clinical utility, which are
trical stimulation of these neural pathways, other stimuli can be recommended for additional reading. - - - •^• - - - -
8 9 11 21 223 281 288 336 344
used to provoke recordable responses (Table 12-2). Those neuro-

physiologic systems commonly monitored have been published
with at least case series results demonstrating effectiveness or Table 12-3. Increase in Amount of Artifact
usefulness - - 3.24.27,54.80,94,102.106.114.132.116.162.16.1. r65.1so.201.221.227.228.233.264.
1 5 1
Problem/Cause Detection Action

265.288.312.330.337.395-397,423,425
High-impedance Check electrode impedance Switch t o duplicate
electrode electrodes
Increased EMG View input signal, check Patient may be
Table 12-2. Some Stimuli That Can be Used to anesthetic level "light"; request in-
Elicit Evoked Potentials creased anesthesia
Visual Auditory or use NM block;
Diffuse flash Clicks increase number
Checkerboard reversal Brief tones of stimuli
Moving gratings or bars Syllables Electrocautery High-amplitude noise caus- Wait, resume when
Flickering Words ing amplifier block cautery stops and
Partial field stimuli Musical chords (saturation) amplifiers recover
Other Rhythmic, high-amplitude Turn off o n e source
Somatosensory Other
equipment noise.View input t o of noise at a time
Median nerve Brief joint movements
determine frequency while looking at
Digital nerve Respirator air bursts
input
Posterior tibial nerve Olfaction
Trigeminal nerve branches Temperature EMG = electromyography; NM = neuromuscular.
Skin dermatomes Complex tasks From Erwin CW, Erwin AQThe use of brain stem auditory evoked potentials in
intraoperative monitoring. In Russell GB, Rodichok LD (eds): Primer of
From Nuwer MR (ed): Evoked Potential Monitoring i the Operating Room. Intraoperative Monitoring. Boston, Butterworth-Heinemann, 1995, pp 135-158,
New York, Raven Press, 1986, p 6, with permission. with permission.
C h a p t e r 12 INTRAOPERATIVE NEUROPHYSIOLOGIC MONITORING — 4 4 1
Table 12-4. Most Common Sources of Artifacts in the (SEPs). The distal nerve (usually a compound nerve) is stimu-
Operating Room lated and the cortical response monitored. This was one of the
1. The Bovie coagulator first and most common techniques used in an attempt to monitor
2. Heating devices (blood warmer, heating blanket) spinal cord function during scoliosis surgery. ' ' 37
-6 9 J 2 2 a 2 8 8 2 9 0 3 4 4 3 7 5
Intraoperative SEP monitoring results for scoliosis surgery have

3. Operating microscopes
been studied by the Scoliosis Research Society in a large series
4. X-ray view boxes
of 51,263 patients, - with its sensitivity, specificity and posi-
288 290
5. X-ray devices, such as a C-arm tive as well as negative predictive values demonstrated (Table
6. The operating table itself 12-7). It is also the most common intraoperative neurophysio-
7. Metal placed on o r near the patient, such as spinal logic monitoring technique performed, especially during spine
instrumentation o r metal head holders surgeries or direct spinal cord neurosurgical procedures that
From Rodichok LD, Schwentker MC: Special problems in the operating room. may threaten the spinal cord's function. 6-18.22,49.66.68.73.89.95.96.123.138.
1
In Russell GB, Rodichok LD (eds): Primer of Intraoperative Monitoring. Boston, i 95.196,198.199.232.243.247.251.257,264.266,274,282.285.289,290,321.326.344,356.374.375,403.413,

Butterworth-Heinemann, I995,pp 241-244, with permission. 416,430.437,439
For some spine surgeries, monitoring multiple threatened

SOMATOSENSORY EVOKED POTENTIALS nerve roots is also desirable, for which dermatomal somato-
sensory evoked potentials (DSEPs) have been attempted. - - 58 406 408
Peripheral nerve through cortical sensory pathways can be A 2 % incidence of nerve root deficits has been found after
monitored by means of somatosensory evoked potentials scoliosis surgery, which had not been predicted by S E P
Table 12-5. Artifactual Shift of Latency or Loss of Response During SEP/BAEP Monitoring
Problem/Cause Detection Action
Stimulus Input Defects
Cable/electrodes faulty Sudden loss of response, including first Reseal connections, verify correct placement, or
peaks and peripheral responses; change t o backup cables/electrodes; Replace
verify connections/positioning, use defective cables/electrodes for next time
back up cabling/electrodes
Stimulator/click Suspect during surgery, connect new Increase stimulus intensity; if n o improvement,
transducer faulty stimulator/transducer t o confirm connect new stimulator after testing with backup
cables/electrodes
Stimulation electrodes o r Suspect during surgery, confirm by Increase stimulus intensity, improve placement and
transducer dislodged inspecting placement and securing techniques next time
Conductive loss (BAEP) Suspect during surgery, confirm by Increase stimulus intensity, improve sealing of ear
(fluid in ear canal) inspecting placement canal (adhesive dressing, etc.)
Recording Pickup Defects
Cable/electrode breakage Sudden loss of EP response with pre- Check electrodes placement, use backup electrodes/
o r dislodgement servation of peripheral responses cables if functional, while checking cable/electrode
(that verify input) connections. If no EP response with back ups,
notify surgeon. If during high risk maneuvers,
notify surgeon before checking back up systems.
Excessive noise Averaged response much more ragged, Increase stimulus intensity, reposition dislodged
peaks less discernible, reproducibility electrodes, verify cabling connections are secure
diminished and initial peripheral input detectable o r see
above
Excessive high-frequency Check filter settings Return filter t o baseline settings o r decrease high
noise frequency cutoff
Check filter settings Return filter t o baseline settings o r increase low
Excessive baseline drift frequency cutoff
Sudden high amplitude noise apparent Increase stimulus intensity, pause averager during
Noise masking in input and averaged response drilling, and/or clear averaged response and
(drilling, cautery, etc.) restart
Slow drift over minutes Compare with previous and baseline Check with anesthesiologist for concurrent
in amplitudes or latency averages anesthetic o r body temperature changes, if none
sufficient t o account for changes, and at alert
thresholds, notify surgeon (possible vascular or
mild traction insult)
From Erwin CW, Erwin AC: The use of brain stem auditory evoked potentials in intraoperative monitoring. In Russell GB, Rodichok LD (eds): Primer of
Intraoperative Monitoring. Boston, Butterworth-Heinemann, 1995, pp 135-158, with permission.
Table 12-6. Procedures to Help Decrease reliable technique than the usually employed peripheral mixed
Intraoperative Artifacts nerve stimulation SEP in detecting root i n j u r i e s . - 151154169 298
Remove grease and abrade skin before applying disc scalp electrodes. However, if the selective root is directly stimulated, the result-
Glue electrodes down with collodion. ing SEP may reliably indicate that particular root's compro-
If electrodes are on overnight, re-gel and abrade scalp again in the
m i s e . - Peripheral nerves at risk can also be monitored by
228 360
means of SEP, such as may occur in various orthopedic proce-

operating room.
dures, e.g., total hip arthroplasty and fracture fixa-
Keep electrode impedances at approximately 2000 ohms.
tions. i 3 . i * 2 . 2 3 7 . 3 8 i . 3 8 4 j addition to direct nerve compromise by
1
n
Use the shortest electrode wires that can be securely fastened out of
physical means, SEP has been used to monitor vascular proce-
the way. dures that may hemodynamically compromise spinal cord func-
Use short intereiectrode distances between pairs of recording tion, such as spinal cord arteriovenous malformations,
electrodes. cardiopulmonary bypass procedures, or thoracic aortic
Braid the recording electrode wires together. aneurysm resections. - - - - - -^ 23 60 65-67 72 83 86
Have back-up stimulus and recording, electrodes available and already 234.244.253.280.296.340.347.353.383,426.433
in place on the patient. Owing to the anesthetic blunting effects on the cortical re-
Keep recording and stimulating wires and cords as far as practical sponse, others have additionally monitored the spinal cord tracts
away from each other. directly, usually by means of peripheral nerve stimulation and
D o not cross cables o r wires over o t h e r cables, especially power recording a near-field spinal cord evoked potential (SCEP)
cables. through interspinous ligament, epidural, or subdural elec-
Do not kick, jar, or sway wires (secure o r tape out of the surgeons trodes. 10.108.130.163,228.272.328.421-423.439 Advantages of SCEP record-
and anesthetists' way). ings include less deterioration from anesthetics, less
Keep the low filter above I Hz whenever possible. deterioration at higher stimulation rates, permitting up to 30-Hz
Unplug unused equipment (not just off). stimulation as opposed to 5 Hz or less usually required to opti-
mize cortical SEP responses. - - - - - - - Higher stimu-
96 283 285 355 371 395 396 409
Avoid appliances with 2-pronged power plugs (ungrounded).

lation frequencies permit more rapid averaging of the evoked
Stop averaging whenever amplifiers are saturated (e.g., during
potential, obtaining results in epochs as short as 3-4 seconds,
electrocautery).
compared with 1-2 minute epochs required for cortical SEPs. 285
Adjust sensitivity so that some normal trials result in artifact A disadvantage of direct spinal cord recordings is the use of
rejection activation. more invasive electrodes, usually being required in the opera-
Use enough neuromuscular junction blocking agents (if only SEP and tive field or epidural space (Table 12-8). Direct spinal cord stim-
not EMG o r MEPs recorded). ulation is often possible intraoperatively and has been used with
Delay recording until several ms after the stimulus (to avoid inter- cortical (SEP) recordings or recordings made directly from the
ference by the stimulus artifact). spinal cord (SCEP). - The recording site for SCEP is usually
262 272
Verify adequate stimulus input by a peripheral pickup t o verify stimuli more cephalad to the surgical area with cord stimulation dis-
are being applied. tally. Also, recordings from distal muscles or nerves can be used
(From Nuwer MR (ed): Evoked Potential Monitoring in the Operating Room. with direct spinal cord stimulation. - - - - - - - - 2 106 152 196 231 271 297 299
New York, Raven Press, 1986, p 39, with permission.) 3i2.397.395.396.4oo intraoperative SEP continues to be useful and
one of the most frequently used techniques for monitoring
many types of surgical procedures. SEP intraoperative mon-
monitoring. DSEP studies result in smaller cortical evoked
136
itoring provides early warning of impending harm to the
responses and have not had the reproducibility and reliability
found with peripheral mixed nerve SEPs. - Fortunately, con-
406 408
tinuous selective electromyography has been found to be a more Table 12-8. Comparison of Cortical SEP Versus SCEP
Recording Techniques
Spinal Cord
Table 12-7. SEP Monitoring in Scoliosis Cortical SEP Evoked Potential
Sensitivity 92% Minimum time to 1-2 minutes 10-30 seconds, epidural
417745r produce each EP 1-2 minutes, ligamental
Specificity 98.8% Forane 0.5% Usually gone Usually preserved
50,207t/50,78|ttt
Can monitor caudal Yes Yes
Positive predictive value 42% to surgery t o o
4177991* Begin monitoring At or before After wound opening
Negative predictive value 99.3% induction
50,207^/50,2411+ Discontinue After awakening Upon wound closing
Key: * = Patients with new permanent neurologic deficits (NPND) as pre- monitoring
dicted by monitoring
m
= Total number of patients with NPND Wires Hidden In surgical field
* = Patients without NPND with none predicted by monitoring Risk of damage Negligible Epidural-small
= Total number of patients with no monitoring prediction of deficits Ligamental-negligible
ttt = Total number of patients without NPND
* = Total number of patients with monitoring prediction of deficits From Nuwer MR: Monitoring spinal cord surgery with cortical somatosensory
From Nuwer MR: Spinal cord monitoring. Muscle Nerve 1999;22:1620-1630, evoked potentials. In Desmedt JE (ed): Neuromonitoring in Surgery.
with permission. Amsterdam, Elsevier, 1989, pp 158-164, with permission.
Chapter 12 INTRAOPERATIVE NEUROPHYSIOLOGIC MONITORING — 443
Table 12-9. Somatosensory Evoked Potentials Field Distributions

Basis ofVoltage
Waveform Recording Site Field Distribution Putative Generators
Posterior tibial SEPs

N22 Lumbosacral spine Transverse oriented dipole Dorsal gray of spinal cord at root entry
zone
W3 Sacral spine Traveling wave Reflexly evoked ventral root discharge
after PTN stimulation at the knee
PV Sacral spine Traveling wave Lumbosacral plexus and roots
DCV Lumbar t o cervical spine Traveling wave Dorsal column volley
N29 Upper cervical spine Axially oriented dipole Nucleus gracilis
P3I Scalp (widespread) Far-field potential Brain stem
N34 Scalp (widespread) Far-field potential Brain stem
P37 Ipsilateral central region Mesiolateral tangential dipole Primary sensory c o r t e x of the foot
in longitudinal fissure
Median SEPs
Erb's point Erb's point Traveling wave Brachial plexus
NTT (DCV) Cervical spine Traveling wave Dorsal column volley
NT3 Cervical spine Transverse and axially Dorsal gray of spinal cord at root entry
oriented dipole zone; nucleus cuneatus
Pl4 Scalp (widespread) Far-field potential Medial lemniscus at cervicomedullary
junction o r caudal medulla
NT8 Scalp (widespread) Far-field potential Subcortical
N20 Contralateral central region Tangential dipole Primary sensory cortex of the hand
P22 Contralateral frontal region Unknown Primary motor c o r t e x of the hand
SEPs = somatosensory evoked potentials; DCV = dorsal column volley; PTN = posterior tibial nerve.
From Lee EK, Seyal M: Generators of short latency human somatosensory-evoked potentials recorded over the spine and scalp. J Clin Neurophysiol
1998; 15:227-234, with permission.
sensory function of the spinal cord and helps to guide surgi- BRAIN STEM AUDITORY EVOKED POTENTIALS
cal progress. - - - - - - - - -^ The SEP components that
8 17 18 31 94 264 273 274 290
can be monitored and their field distributions are shown in The special sensory tracts involved in hearing can be moni-
Table 12-9. The cortical JSEP_waveform peaks mo.re com-
211
tored by brainstem auditory evoked potentials (BAEPs). - - - 53 74 103
monly monitored are the_P37/N45 potentials for tibial nerve 124,139.140.239.245,295.399.414,430 X n " e s e» beCaUSe
a r e c a l l e d b m m s t e m
stimulation, and the N20/P22 potentials for median nerve much of the brain stem is involved in the transmission of audi-
stimulation. tory input before it is relayed to the auditory cortex. The BAEP
consists of many peaks that represent various transmission and
synaptic brain stem centers. The most important waveform
Table 12-10. Brain Stem Auditory Evoked Potentials peaks to monitor are the first (representing the auditory nerve
Field Distributions portion of the VIII cranial nerve) and the fifth (representing the
10-20 final pathway through the brain stem) (Table 12-10). 97
Recording
Montage Cz-Ai Cz-Ac ELECTROMYOGRAPHY
Wave I Upgoing, later than 1.3 ms Virtually absent
If present, similar amplitude
Electromyography can evaluate irritation to peripheral motor
Wave II Usually present but absent
to Cz-Ai -0.1 ms later
nerves in real time and can alert the surgeon to inadvertent trac-
in some normal individuals
tion on a nerve. - -
45 n 6
- - -
, 3 ,
By use of multi-trace
, 4 0 !5,170 185a 298 328a
(not an obligate wave) real-time recordings, many different cranial or peripheral nerves
Wave I Present in normals Lower amplitude (sometimes or nerve roots can be monitored simultaneously. However, there
virtually absent), —0.01 ms are multiple intraoperative sources of motor nerve irritation,
earlier than Ai which can include such benign events as saline nerve lavage,
Wave IV N o t obligate If present, —0.01 ms later N
which nevertheless create the spontaneous motor discharges
than Ai called neurotonic discharges (Fig. 12-1). Neurotonic dis-
WaveV Obligate, largest of waves, Larger, more separated from charges are irregular rapid motor unit firings that can easily be
followed by major down- wave IV than in Ai, auditorily monitored by the surgeon and the clinical neurophys-
going negativity usually -0.01 ms later iologist. Frequently, multiple motor nerves may be involved
,85a
than Ai in the operative field, and thus multiple muscles must be moni-
tored simultaneously. Unfortunately, complete lysis of a nerve
From Erwin CW, Erwin AOThe use of brain stem auditory evoked potentials in can occur without necessarily generating neurotonic discharges.
intraoperative monitoring. In Russell GB, Rodichok LD (eds): Primer of
Intraoperative Monitoring. Boston, Butterworth-Heinemann, 1995, pp 135-158, Additionally, detection of such discharges is obscured during
with permission. the electrically noisy event of electrocautery. 141
tissue from other tissue in the operative field. - - - - - - - - 84 90 91 175 276 278 291 312
4oi.405.4ii j
n presence of an intraoperatively detectable nerve
e
action potential across a suspected area of compromise is asso-

ciated with a 90% chance of recovery to a useful motor state. 405
The stimulation intensity required to activate a nerve has

been used to assess intervening bony integrity such as during
pedicle screw p l a c e m e n t . - - - A decrease in stimula-
45 46154169 235 419
tion needed to activate a spinal nerve suggests a loss of bony

cortical (higher electrical resistance) integrity, alerting the sur-
geon that possible misplacement of the screw has occurred.
Stimulation intensity can also help distinguish between neural
or connective tissue, thus helping with the progress of dissec-
tion. A study of tethered cord surgery found that a response
312
recorded with a 1-10 ma stimulation meant the stimulator was

directly on the nerve, whereas 11-25 ma implied near the nerve
but with intervening tissue, and greater than 25 ma indicated the
nerve was distant to the stimulation site. 312
Figure 12-1. Neurotonic discharges. A 52-year-old female under- MOTOR EVOKED POTENTIALS
going a T 12-L2 intraspinal tumor resection.The neurotonic discharges
Motor evoked potentials (MEPs) assess the motor systems
in traces 2, 3,6, and 7 are seen during t h e time of resection.
Recordings w e r e d o n e using intramuscular hook-wire electrodes
from the motor cortex to the anterior horn cell and then by way
placed in the following muscles: of the peripheral nerves to the muscle proper. - - - - - - 43 47 70 92 127 144
160.166.177.203,204.2)7,243.263.267.271.297.308.324.341,440.441 Avoiding paralysis is

Trace I: Left vastus lateralis Trace 5: Right vastus lateralis
Trace 2: Left tibialis anterior Trace 6: Right tibialis anterior a major intraoperative concern, and monitoring these pathways
Trace 3: Left medial Trace 7: Right medial should theoretically permit better detection of intraoperative
gastrocnemius gastrocnemius motor function loss. Animal studies suggest greater sensitiv-
290
Trace 4: Left anal sphincter Trace 8: Right anal sphincter ity to spinal cord traction may occur with MEPs than SEPs, with
a change in the MEP at 5-10% distraction, but no change in the
SEP until 15% distraction. The advantages and disadvantages
335
NERVE CONDUCTION STUDIES

of MEP versus SEP monitoring are further discussed in Table
Nerve conduction studies (NCSs) help to quantify nerve in- 12-11. Unfortunately, electrical transdermal cortical motor
tegrity. Whereas real-time EMG may detect even small irrita- stimulation is painful, though not a significant obstacle in the
tions, a decline in the nerve conduction response is somewhat anesthetized patient. - - - - - - ' - Less uncomfortable
42 43 47 146 172 177 335 379 410
more proportional to the severity of compromise of the nerve magnetic stimulation systems have been developed. 127.144,172.217.
being tested. Traction results in rapid declines within 1-2 335.341,398 However, both these techniques remain experimental in
113a
minutes, whereas partial vascular compromise results in slowly the United States. Both types of transcranial stimulation place
335
progressive potential declines over longer times, approaching the patient at increased risk of a seizure; however, during anesthe-
10-20 minutes, although if the vasculature is completely sia this is rarely a problem. Several papers addressing the safety
oblated, immediate decline in neural function will result. If such and theoretical concerns about transcranial cortical stimulation
neural compromise persists for greater than 30 minutes, irre- have been favorable. - - Monitoring of motor pathways by means
3 26 59
versible neurologic insults may ensue. - - - If a tourniquet is

63 66 128 285
of MEPs, NCSs, or EMG requires that the motor system is not
used to exsanguinate an operative field, an additional 20 min- subject to total motor blockade, and the anesthetic requirements to
utes for nerve function recovery following tourniquet release accomplish this are discussed below in Anesthesia Considerations.
may be required, making this a relative contraindication to mon-
itoring. NCSs can be used to evaluate improvements across
25
ELECTROENCEPHALOGRAPHY
decompressed sites, which can help guide a surgeon as to the
adequacy of nerve decompression, as well as to monitor for Electroencephalography (EEG) is used to monitor the vascu-
worsening of function. The NCS can also help identify neural lar supply to the cortex during neurosurgical procedures such as
Table 12-11. Comparison of Intraoperative SEP versus HEP Spinal Cord Monitoring Techniques
Modality Advantages Disadvantages
Somatosensory-evoked Readily available sites for stimulating and recording Requires experienced personnel and dedicated equipment
potentials (SEPs) Subcortical recordings allow use of potent inhaled agents Cortical recordings attenuated by potent inhaled agents
Generally detects spinal cord injuries May n o t detect isolated m o t o r injuries
Motor-evoked Depending on stimulating and recording sites, does n o t May n o t be readily o r universally available. Experienced
potentials (MEPs) require significantly more equipment than SEPs alone monitoring team requested
May be more sensitive t o impending injury than SEPs Potent agents and muscle relaxants restricted in certain
Allows monitoring in selected cases with absent SEPs circumstances
Combined with SEPs, provides "whole cord" monitoring Movement during stimulation may interfere with surgery
From Adams DC, Emerson RG: Intraoperative spinal cord monitoring. Curr Opin Anaesthesiol 1996;9:372, with permission.
carotid endarterectomy or aneurysm clipping. - - - - - - - ' - 4 55 56 200 250 268 304 3 0

Surface electrodes may be used such as gold cup EEG elec-
319.320.322.332.388.407.415.438
A w j n d ortical blOOd flOW is
w e x i s t s a s CO trodes, bar electrodes, flat plate electrodes or standard NCS ring
compromised in which cortical electroencephalographic activ- electrodes, subdermal EEG electrodes, or highly specialized
ity is diminished and then abolished prior to reaching the criti- custom-made probe electrodes. However, many of these custom
cal level of blood flow and oxygen supply that leads to electrodes are becoming more generally available through elec-
irreversible cell death. - - The loss of EEG activity
352 386 387 428
trode companies and are autoclavable for re-use. This includes
allows the surgeon to take corrective actions to improve cortical electrode forceps that can touch or hold a structure. These for-
blood flow and minimize permanent neurologic ischemic post- ceps are designed for stimulation, with each prong being the
operative deficits. pole of a stimulator. They can also be configured to serve as
pickup electrodes. When not being used as a stimulator or as
VISUAL EVOKED POTENTIALS sensing electrodes, these forceps can be used to dissect as with
any forceps. Often, custom hooks are used as stimulation and
Visual evoked potentials (VEPs) monitor the visual pathways pickup electrodes. Though usually spaced closer than the ideal
from the retina to the occipital visual c o r t e x . " - - - - 270 323 382 431435
4 cm for recording, in order to better accommodate intraopera-
VEPs are used infrequently, owing to both technical difficulties tive field limitations, these active and reference electrodes can
and risks. Usually, other methods can cover the same general reliably detect a response and a decrement in that response. 405
operative areas. Reports of intraoperative VEP monitoring

284
Monopolar recordings with a distant reference have also been
have generally been unfavorable. - 99 382
used effectively when operative fields are restricted. When 68
used as stimulating electrodes, the shock artifact can be rather

EQUIPMENT REQUIREMENTS large if both the stimulation and pickup are occurring in the
same relatively small operative field. To minimize the shock ar-
Multiple neurologic systems can be monitored by the many tifact, sometimes a monopolar cathodal stimulating electrode (a
techniques described (see Table 12-1). Certain surgical proce- single hook or probe) is used with a distant anode, usually a sur-
dures will require concurrent monitoring of several different face electrode. This also permits a more focal stimulation with
systems. Some techniques require electrical or other (see Table less likelihood of cross-stimulation to adjacent nerves. The spe-
12-2) stimulation of a nervous structure with recording of sub- cific placement and connection of these electrodes, called the
sequent time-locked responses. These responses are often of montage, are discussed with each technique.
j small amplitude compared with the operating room's electri- The operating environment is both electrically and mechani-
cally noisy environment, requiring the averaging of many stim- cally hostile. Electrodes initially secured for continuous moni-
ulations to obtain a reliable response. Sources of electrical noise toring use may become dislodged during a procedure. Backup
in the operating room include ventilators, blood warmers, elec- electrodes and equipment are essential to minimize technical
trocautery, and other devices used to monitor the depth of anes- failure. Constant vigilance for technical failures is essential to
thesia and motor blockade (see Table 12-4). Other techniques 327
prevent unnecessary alarm or delays in the surgical procedure.
require real-time display by auditory loudspeaker or continuous Protocols to minimize electrical interference and decrease intra-
video trace display, and some channels may require different operative artifacts are listed in Tables 12-3, 12-5 and 12-6. - 284 286
sweep speeds as well as different sensitivity settings. Fre-

quently, multimodal monitoring, combining multiple displays GENERAL RESULTS AND EFFICACY
of real-time and averaged signals, is desired in order to evaluate
all or as many systems as possible that are at risk. Since some Many of the above-described techniques are sensitive but not
techniques require averaging of the response to stimulation and particularly specific. False-positive results (loss of the desired
others are real-time displays, sophisticated equipment may be response without actual pathology to the monitored system)
required that allows differing time bases (sweep speeds) and occur much more commonly (approximately 5% rate, with
triggering for each trace to be displayed concurrently. Also, the about half of these accountable to transient technical failures)
same or similar cortical structures may need to be stimulated al- than true-positives (which for scoliosis surgery SEP monitoring
ternately and each side's response averaged to check both sides occurs at about a 0.4% r a t e ) . -
44 96102
- Even a transient loss of
290
of the neuraxis such as when performing bilateral alternating signals places patients at greater risk of permanent deficits,
tibial or peroneal nerve stimulation. Such ability to interlace 290
though the highest risks are for those with initially easily ob-
stimulations, averages, and real-time response recordings is tainable responses that gradually disappear, having about a 5 0 %
available in the high-end electrophysiologic instrumentation risk of permanent s e q u e l a e .
102105
Some of the "false-positive"
packages. These modern electrophysiologic instruments typi- responses, especially with SEP, do correlate with transient post-
cally have excellent noise reducing preamplification that needs operative paresthesias suggesting a partial injury, but given their
to be verified, as this is essential in the electrically noisy intra- resolution, the study is considered a false-positive in terms of
operative environment. However, if just one modality is to be predicting permanent deficits. False-negative studies (maintain-
assessed, such as SEPs, often the simplest of electrodiagnostic ing intact electrophysiologic responses, but the patients awaken
instruments can be used, though adequate performance again with new permanent neurologic deficits) are quite rare but do
should be verified in the operative suite. Such instruments may occur. - -ni.136.139.162.214.255.261.290.396.425
20 52
Many of these reported
not be as effective in rejecting noise and thus require longer av- false-negative studies are not actually an electrophysiologic
eraging times, which results in less continuous monitoring. If false-negative because the monitored neurologic system was not
the attempts at intraoperative monitoring are going to be opti- the system in which the deficit occurred. This emphasizes the
mized, the more sophisticated equipment becomes a necessity importance of the electrodiagnostic consultant's expertise in
in order to have adequate versatility, performance, and flexibil- planning for effective intraoperative monitoring given the pro-
ity to meet all demands. Specific equipment needs and proto- cedure to be performed and potential structures at risk. An ex-
cols are discussed further with each technique. ample is the use of SEPs, which specifically monitor sensory
function and not motor function. - Another is the perfor- 214 267
entirely for approximately 1 of every 200 patients moni-
mance of bilateral simultaneous peripheral nerve stimulation, tored. - - Given the enormous economic impact of paraple-
69 288 290
which can lead to missing a rare hemicord dysfunction. Other 255

gia, the additional cost of intraoperative electrophysiologic
false-negatives are the late appearance of deficits that may re- monitoring for even the fairly low risk procedure of scoliosis
flect the onset of edema or vascular insufficiency occurring surgery is cost-effective. One series of cervical spine surgeries
postoperatively, which could not be detected intraoperatively as compared 100 consecutive monitored cases with 0% subsequent
such impairments had not yet occurred. 162
adverse outcomes to the previous 218 unmonitored patients who
Often technical problems will lead to the loss of a signal intra- had a 3.7% tetraplegia and 0.5% mortality adverse outcomes. 95
operatively. The search for such a technical problem must be Other disorders such as surgery for aortic coarctation (0.5%
made expediently. The location and integrity of the recording paraplegia incidence), thoracoabdominal aneurysms (up to 15%
montage and technical system must be verified prior to alerting incidence of paraplegia), and surgical decompression for spinal
the surgeon as to a technical problem or an apparent electrophys- cord tumors or trauma carry even higher risks (up to 20%) and,
iologic change. "Alert criteria" for clinically significant changes therefore, may benefit from intraoperative neurophysiologic
in the monitored responses often include a sustained loss, usu- monitoring; however, such cost-effectiveness benefits have not
ally greater than 10 minutes, in part to allow investigation as to been studied as extensively. - - - 9 50 108 118
anesthetic, or technical problems than may have contributed to One assessment of cost versus liability applied the "learned
the observed changes. - The clinical neurophysiologist must
376 436
hand rule," which states that legal negligence occurs whenever it
verify the monitoring system's integrity and check with the anes- would cost less to prevent a mishap than to pay for the damages
thesiologist for other factors, prior to concluding that a com-
36
predicted to result from it. When restated in mathematical
123
promise in neurological function has been detected. Access to all symbols, whenever the cost (C) is less than the probability (P)
electrodes to verify correct positioning is necessary, though this multiplied by damages or the loss (L): C < P x L; then negli-
is usually somewhat awkward and therefore is a contingency for gence has occurred if the cost, in this case of providing intraop-
which one must plan in advance. erative neurophysiologic monitoring to such an at risk patient, is
Despite these challenges, significant reductions in morbidity not expended. Using one of the most reliable statistics we have,
have occurred from the increasing use of intraoperative electro- that of the scoliosis group, if the risk is chosen as a very conserv-
physiologic monitoring. - 287
Hearing loss has been pre-
290
ative value of 0.7% (P, without monitoring) and approximately
vented. 2 4 5 . 2 9 5 . 3 9 9 , 4 1 4 paraplegia from scoliosis surgery has three quarters of a million dollars of loss occurs from paraplegia
declined from as much as 4-6.9% to 0-0.7% with the advent of (L) onset age 15, then any cost of less than $5240 per case to
75
intraoperative SEP monitoring. - The Multi-Center Study of

95 251
provide intraoperative neurophysiologic monitoring should be
Spinal Cord Monitoring in Scoliosis Surgery found at least a expended in order to avoid this definition of negligence. 115123
60% decline in morbidity with intraoperative monitoring during

scoliosis surgery, with serious neurologic deficits prevented
ANESTHESIA CONSIDERATIONS
Establishing intraoperative neurophysiologic protocols must
SEIZURES
involve close cooperation with the anesthesiologist. Anesthesia
optimized to facilitate neurophysiologic monitoring will, never-
theless, produce drug effects that alter sensory and motor
MYOCLONUS evoked responses that must be appreciated and appropriately in-
/CATALEPTIC
terpreted by the monitoring neurophysiologist. Although some
H ANESTHESIA generalizations exist about anesthesia drug effects, the relative
B«*C potency and specific location of drug actions differ between
agents, so that some discussion of each agent is necessary to
better understand the alternatives and their implications.
Some of the differences between anesthetic agents relate to
the anatomic site of the anesthetic effect. Other differences
329
relate to the neurophysiologic site of drug action. The major

AWAKE target for anesthetic action appears to be at the gamma
aminobutyric acid (GABA) and /V-methyl-D-aspartate (NMDA)
receptors mediating electrolyte channels (Na , Cl~, C A ) at + 2+
SLEEP central nervous system synapses.

In the sensory system, the response generated by synapses in
— DEPRESSANT cortical structures will be the most affected, with less effect oc-
ANESTHESIA
m
curring at more peripheral structures, where fewer synapses are

MEDULLARY involved. Because the most prominent anesthetic effects for
371
"PARALYSIS the sensory system is on the synaptic-rich cortically generated

responses, it is not surprising that anesthetic effects on cortical
\
I100/AV
evoked potentials parallel the drug effects described for the
1 sec EEG, which is also a cortical synaptically mediated response
0EATH
(Fig. 12-2). A schema is proposed for anesthesia effects on
428
Figure 12-2. C o r t i c a l EEG s t a g e s t y p i c a l of a n e s t h e s i a . (From cortical sensory evoked potentials (Fig. 12-3). Unfortunately,427
W i n t e r s W D : Effects of drugs on the electrical activity of the brain: most commonly used anesthetic drugs today produce a dose-re-
Anesthetics.Ann Rev Pharm Toxicol 16:413-426, 1976, with permission.) lated depression of the recordable EEG, as well as decreased
EEG
GENERALIZED
SEIZURE
EFFECTS OF VARIOUS ANESTHETIC AGENTS

ON AUDITORY EVOKED RESPONSE Awake
UNRESPONSIVE
1 MAC
1.5 MAC
UNRESPONSIVE
/ / ^
AWAKE
_ INAPPROPRIATE
HYPERACTIVE MOVEMENTS
B-S
Suppr.
1 MAC
Figure 12-5. Cortical SEP and EEG recorded at various

doses of isoflurane. (From Porkkala T, Jantti V, Kaukinen S, Hakkinen
Figure 12-3. Cortical evoked potentials stages typical of V: Somatosensory evoked potentials during isoflurane anaesthesia.
anesthesia. (From W i n t e r s W D , Mori K, S p o o n e r CE, Bauer RO: Acta Anaesthesiol Scand 1994;38:206-210, with permission.)
The neurophysiology of anesthesia. Anesthesiology 1967;28:65, with
permission.)
site may have one of two effects. First, partial synaptic blockade
may compound a loss of I waves, making it more difficult to
amplitude and increased latency of cortical SEP and myogenic bring the anterior horn cell to threshold. At higher doses, synap-
motor evoked potential (MEP), making the anesthetic choice for tic blockade may inhibit synaptic transmission at this site re-
monitoring with these electrophysiologic modalities particu- gardless of the composition of the descending spinal cord volley
! larly challenging. 372
of activity. The third major synaptic location for anesthetic ef-
Based on the major effect of anesthetic drugs occurring at the fects in the motor pathway is at the neuromuscular junction.
synapses, three locations in the motor pathways will be the most Fortunately, with the exception of neuromuscular blocking
susceptible to anesthesia. The first location is within the motor agents and drugs, which alter acetylcholine transmission, anes-
cortex where internuncial neurons and synapses participate in thetic drugs have little effect at the neuromuscular junction.
activation of the motor cortex by transcranial stimuli. When Similarly, neuromuscular blocking agents have little effect on
electrical or magnetic pulses activate pyramidal cells, they pro- central nervous system synaptic transmission and axonal con-
duce a direct activation of the cells producing a " D " wave and duction in motor pathways other than at the neuromuscular
activation via the internuncial pathways (dependent upon junction.
synapses) producing a series of I waves (Fig. 12-4). Weaker Finally, it should be noted that anesthetic drugs may have an
magnetic impulses appear to depend on synaptic activation for effect on evoked responses indirectly by altering other physio-
production of a response. The implication of anesthetic effects logic factors that influence the provision of nutrient supply to
on these internuncial synapses is that the production of D waves the neural tracts. This is discussed in the sections that follow.
will be relatively immune to anesthetic effects whereas the pro-
duction of I waves will be reduced with anesthetic agents that INHALATIONAL AGENTS
depress synaptic function. Of further consideration is that
synaptic function may be a delicate balance of inhibitory and Halogenated Inhalational Agents
excitatory influences. The second major sites of anesthetic Perhaps the most common anesthetics in use today are the
action in the motor system are the synapses at the anterior horn halogenated inhalational agents (desflurane, enflurane,
cell. At this location, the summated D and I waves bring the an- halothane, isoflurane, sevoflurane). Paralleling their effects on
terior horn cell to threshold with a resulting peripheral nerve the EEG, all halogenated inhalational agents produce a dose-re-
action potential leading to a muscle response. Anesthetics at this lated increase in latency and reduction in the amplitude of the
cortically recorded evoked potential responses (SEP, VEP,
BAEP). Although the effects of halogenated inhalational agents
appear to be dose related, the changes observed in some studies
appear to plateau at low concentrations ( 0 . 5 - 1 % inspired con-
centration). Figure 12-5 shows the effects of isoflurane on
339
EEG and on the cortical SEP, demonstrating a parallel effect. 315
Studies support differences in the potency of the halogenated

inhalational agents on the cortical SEP. The relative order seen
is isoflurane (most potent), enflurane, and halothane (least
potent). Studies with sevoflurane and desflurane suggest that
371
they are similar to isoflurane at steady state, but owing to their

more rapid onset and offset of effect (because of their relative
insolubility), they may appear to be more potent during periods
Figure 12-4. D wave and I waves with MEP stimulation. when concentrations are increasing.
E F F E C T O F I S O F L U R A N E ON B A E P
I E UL W-Y
I L_J I U U I 1
0 2.0 4.0 6.0 8.0 10.0 I I I I I I I » 1 L
20 40 60 80
msec
TIME (ms)
figure 12-6. Influence of Isoflurane alone on BAEP. Latency of
Figure 12-7. Motor evoked responses t o transcranial electri-
peaks III and IV-V increased at 1.0% b u t plateaued with increasing
cal stimulation during nitrous oxide/sufentanil anesthesia before,
anesthetic depth. (From Manninen PH, Lam AM, Nicholas JF:The ef-
during, and after administration of isoflurane (0.3% end-tidal). (From
fects of isoflurane and isoflurane-nitrous oxide anesthesia on brain-
Kalkman CJ, D r u m m o n d JC, Ribberink AA: Low concentrations of
stem auditory evoked potentials in humans. Anesth Analg 1985; 64:43,
isoflurane abolish m o t o r evoked responses t o transcranial electrical
with permission.)
stimulation during nitrous oxide/opioid anesthesia in humans. Anesth
Analg I99l;73:4l0,with permission.)
The most prominent effect of halogenated inhalational agents

is on cortical responses, with markedly less effect on subcortical suggested that the most prominent anesthetic effect on the MEP
structures. Studies of recordings at Erb's point (over the brachial is at the anterior horn cell level. However, the loss of I waves
plexus) and over the cervical spine show minimal changes from a cortical effect may be sufficient to block myogenic re-
(0-9%), which are not dose related. As a subcortical response, sponses, even without significant anesthetic effects at the ante-
the BAEP is minimally affected by halogenated inhalational rior horn cell. This is because a series of I waves appear to be
agents. The more prominent latency changes occur in wave V,
with III being less affected and wave I being little affected,
having amplitude changes that are minimal (Fig. 12-6). 238
MEPs recorded in muscle (myogenic) are the most easily

abolished by halogenated inhalational agents. Single pulse stim-
ulation transcranial motor evoked myogenic potentials
(tcMEP) appear to be so easily abolished by inhalational agents
that they are often unrecordable in the presence of these agents.
When recordable, the major effect may occur at low concentra-
tions (e.g., less than 0.2-0.5% isoflurane) (Fig. 1 2 - 7 ) . - ,29 171442
This effect is likely a result of the combination of the anterior

horn cell synapse depression as well as loss of I waves due to
anesthetic effects on the internuncial synapses. Changes in
147
the H-reflex confirm an effect of halogenated inhalational 1 1 1 1—

agents at the spinal level. 246
0 5 10 15
In contrast to myogenic responses, the D response seen in the Figure 12-8. Effect of isoflurane on epidural recordings fol-
epidural space is highly resistant to the effects of these agents lowing transcranial electrical stimulation of t h e m o t o r pathways.The
and is easily recordable at high volatile anesthetic concentra- first wave (D wave) remains intact as t h e concentration of isoflurane is
tions and can be used for monitoring (Fig. 12-8). It has been
127
increased, but there is a progressive loss of I waves.
Chapter 12 INTRAOPERATIVE NEUROPHYSIOLOGIC MONITORING — 4 4 9
necessary for producing myogenic responses in the unanes-

thetized state.
Studies comparing transcranial magnetic motor evoked
potentials (tcMMEP), using an externally applied magnetic
field, and transcranial electrical motor evoked potentials
(tcEMEP), using an electrical voltage applied across the cra-
nium, suggest that the magnetic technique can be more sensitive
to the inhalational agents, probably because magnetic stimu-
367
lation relies more on transsynaptic activation. High magnetic

strength tcMMEP (which can produce D waves) appears to min-
imize this cortical difference.
Because the D wave is resistant to anesthetic depression, the
anesthetic effect at the anterior horn cell can be partially over-
come at low concentrations by high-frequency (multiple-pulse)
transcranial stimulation. In this circumstance, the multiple D
370
waves formed (and I waves, if produced) summate at the ante- Figure /2-9. Effect of multiple pulse transcranial stimulation
rior horn cell resulting in a peripheral nerve activation and sub- on compound muscle action potentials (CMAPs) during 0.9%
sequent motor response (Fig. 12-9). Low concentrations of isoflurane.The amplitude of the CMAP increases as a second stimula-
inhalational agents appear acceptable when high-frequency tion pulse is added, with a maximum effect in this study when the in-
transcranial stimulation is used (trains of stimuli with interstim- terstimulus interval (ISI) is 3 milliseconds.
ulus interval [ISI] of 2 - 5 milliseconds). - ° As predicted,
1 7 7 m 3 7
higher concentrations of these agents eliminate myogenic re- inhalational anesthetic agent. - - Like halogenated agents,
161 359 371
sponses from this stimulation. Clinical experience suggests that effects on subcortical and peripheral sensory responses and on
anesthetic plans avoiding the inhalational agents may still be epidurally recorded MEP are minimal.
desirable for optimal MEP monitoring, even with the high-fre-
quency stimulation technique. 309
Studies with direct spinal or epidural stimulation show minimal

effects of anesthesia on neurogenic or myogenic responses. 300 N 0 on
9
However, the above described effects at the anterior horn cell sug-
gest that depression may change the mixture of orthodromic motor
and antidromic sensory contributions to the recorded responses. A
study of the responses in the peripheral nerve and muscle follow-
ing epidural stimulation in the cat revealed that single-pulse stimu-
lation produced a response that was eliminated by pentobarbital,
low-dose isoflurane, or by posterior column transection (but not
lateral column transection). This suggests the response recorded
254
from the peripheral nerve was largely mediated by sensory path-

ways, especially those of the posterior column. When a pair of
stimuli were used (interstimulus interval 1-5 milliseconds), a new
complex in the peripheral nerve response was seen. This complex N 0 off —
2
and the compound muscle action potential (CMAP) were elimi-

nated only by high-dose isoflurane or lateral spinal cord transec-
tion (lysing the descending motor pathways). Therefore, the type
of spinal cord stimulation and the anesthetic agents used may alter
the balance of sensory and motor contributions to the peripheral
nerve and muscle response from direct spinal stimulation. Recent
studies suggest that with isoflurane anesthesia, the motor compo-
nent is preferentially blocked, perhaps by interaction at the
synapses in the anterior horn cell or by differential effects on con-
duction in the spinal tracts in humans. These studies do not, how-
71
ever, clearly allow a recommendation of anesthesia that will

preferentially promote monitoring of motor pathways with direct
spinal, epidural, or paraspinal stimulation.
Nitrous Oxide
"100 ms
Nitrous oxide produces SEP cortical amplitude reductions
and latency increases when used alone or when combined with Figure 12-10. Effect of nitrous oxide on cortical recordings
halogenated inhalational agents or opioid agents (Fig. 12-10). of posterior tibial nerve somatosensory evoked potentials.The ampli-
Studies of nitrous oxide in a hyperbaric chamber confirm the tude of the response is markedly reduced over the 10—15 minutes fol-
depressant nature of its effect at higher doses as well. When 334
lowing the introduction of nitrous oxide, and a response returns after
compared at equipotent anesthetic concentrations, nitrous oxide agent is removed. (From Sloan TB, Koht A: Depression of cortical s o -
produces more profound changes in cortical SEP and muscle matosensory evoked potentials by nitrous oxide. Br J Anaesth 1985;
recordings from transcranial motor stimulation than any other 57:850, with permission.)
Despite the depressant effect of nitrous oxide, it has been INTRAVENOUS ANALGESIC AGENTS
used with recording of responses, particularly when combined
with opioids ("nitrous-narcotic" anesthetic technique). When Most anesthesia techniques utilize a mixture of different
combined with other agents, nitrous oxide may be "context-sen- anesthetic agents such as supplementation with inhalational
sitive" in its effects, similar to its effects on the EEG (i.e., the agents (halogenated agent or nitrous oxide) with opioids or in-
actual effect may vary depending on the other anesthetics al- travenous sedatives (e.g., benzodiazepines, etomidate, droperi-
ready p r e s e n t ) .
249369
dol, or propofol). If the inhalational agents need to be
As with sevoflurane and desflurane, nitrous oxide is rela- completely avoided, intravenous agents can be combined to pro-
tively insoluble. Therefore, anesthetic effects can change duce a total intravenous anesthetic (TIVA).
rapidly when concentrations are varied intraoperatively. Since a
decrease in concentration will be associated with a rapid in- Opioid A g e n t s
crease in amplitude and decrease in latency, it may "mask" am- The effects of opioid analgesics (fentanyl, sufentanil, alfen-
plitude and latency changes that may be occurring from tanil, remifentanil) on sensory and motor evoked responses are
concurrent neural compromise. Therefore, such changes should less adverse than inhalational agents, making them important
be avoided during critical portions of the surgery when the mon- components of anesthetic planning for monitoring evoked re-
itored structures may be at higher risk. sponses. Effects are similar for most evoked sensory modali-
306
Also, nitrous oxide can increase middle ear pressure and ties. Minimal changes in spinal or subcortical recordings are
hearing threshold, thereby presenting the possibility for dispro- noted with some amplitude depression and latency increases in
portionate effects on BAEP and cortical auditory evoked poten- cortical responses, especially loss of late cortical peaks (over
tial (AEP) responses when eustachian tube dysfunction occurs. 100 ms) at doses sufficient to produce sedation (Fig. 12-11). 184
This could result in false-positive monitoring deficits occurring As with systemic opioids, the spinal application of morphine or
in the BAEP. Therefore, avoidance of an increase in nitrous fentanyl for postoperative pain management produces minimal
oxide during critical portions of surgery requiring BAEP moni- changes in the SEP and fails to alter the H-reflex. 345
toring is also important. Changes in such anesthetic agent con- Opioid-based anesthesia is frequently used when cortical
centrations should be relayed from the anesthesiologist to the SEP responses and transcranial motor evoked potentials are
clinical neurophysiologist because they are required to help monitored. Studies with myogenic responses from tcMEP
305
with correlating electrophysiologic monitoring changes to the with electrical and magnetic methods show only mild amplitude
operative environment. decreases and latency increases that permit good recording. 1,2 :2,2
With respect to the latter, fentanyl has been suggested to be

useful in reducing background spontaneous muscle contractions
Cervical Cortical
and associated motor unit potentials, which may further im-
prove muscle recordings. Since the opioids do not guarantee se-
dation or amnesia, opioid-based anesthesia must include an
additional sedative agent to produce TIVA.
Ketamine
The effects of ketamine on the evoked responses also differ
from those of inhalational agents. Ketamine can produce central
nervous system excitement with associated enhancement of cor-
tical sensory and myogenic responses. Thus, an increase in
369
cortical SEP amplitude and an increase in amplitude of mus-

345
cle and spinal recorded responses following spinal stimulation
0 25 0 60 J 1 I L _ J 1 1
Time(ms) Time(ms) LATENCY (Sms/tfiv)
Figure 12-11. Changes in median nerve cervical and cortical Figure 12-12. Example of SCEP waveforms before and after
SEP recording with time in one patient after sufentanil 5 g/kg. induction with ketamine at times 2, 5, 10, 15, 20, and 30 minutes.
Two baseline recordings at time zero are shown. (From Kimovec MA, (From Schubert A, Licina MG, Lineberry PJ:The effect of ketamine on
Koht A, Sloan TB: Effects of sufentanil on median nerve somatosensory human somatosensory evoked potentials and its modification by ni-
evoked potentials. Br J Anaesth 1990;65:169, with permission.) trous oxide.Anesthesiology I990;72:33, with permission.)
Spinal cord Cortex

.SC
Figure 12-13. SEP responses recorded from the cervi-

cal and cortical electrodes before (0) and a t several times
up t o 12 minutes following t h e injection of t h i o p e n t o n e (4
mg/kg). (From Sloan TB, Kimovec MA, Serpico LC: Effects of
t h i o p e n t o n e on median nerve somatosensory evoked poten-
tials. Br J Anaesth 1989;63:51, with permission.)
IluV
4 0 ms 50 ms
has been seen. This latter effect on muscle responses may be

174
a silent EEG. For this reason, sensory evoked responses have
mediated by the same mechanisms that potentiate the H- been used successfully to monitor neurologic function during
reflex. However, effects on subcortical and peripheral sensory
354
barbiturate-induced coma (Fig. 12-13). 362
responses are minimal (Fig. 12-12). Minimal effects are also ob-
served in myogenic tcMEP with ketamine. 112
Cortical
Because of these effects, ketamine is a desirable agent for
monitoring responses that are usually difficult to record under
anesthesia (e.g., dermatomal evoked responses and transcra-
n i a l ^ elicited muscle motor evoked responses). However, its
hallucinatory potential and known increase in intracranial pres-
sure with intracranial pathology have led to a reluctance to uti-
lize this agent routinely.
Sedative-Hypnotic Drugs
Intravenous sedative agents are frequently used to induce or
supplement general anesthesia. If inhalational agents must be
avoided, sedative-hypnotic agents are routinely combined with
opioids or ketamine to ensure adequate sedation, anxiolysis, and
amnesia. Although ketamine doses produce some dissociative
effects in addition to analgesia, supplementation can reduce the
risk of excitatory events including hallucinations.
Droperidol
When combined with fentanyl ("neurolept anesthesia"),
droperidol appears to have minimal effects on SEP, VEP, and
M E P . However, since its effect is quite long-lasting, many
173
anesthesiologists would prefer to utilize a more rapidly metabo-

lized sedative hypnotic for TIVA.
5 pV
Barbiturates
Thiopental remains a popular drug for anesthesia induction,
though transient decreases in amplitude and increases in latency
of cortical sensory responses occur. Longer latency cortical Time ( m s ) Time (ms)
waves are most affected, while minimal effects are seen on the
subcortical and peripheral responses. Studies with another bar- Figure 12-14. SEP responses recorded from the cervical and
biturate, phenobarbital, demonstrate that the BAEP is virtually cortical electrodes are shown before and at I -minute intervals after
unaffected at doses that produce coma; changes are not seen the injection of midazolam (.2 mg/kg). (From Sloan TB, Fugina ML,
until doses sufficient to produce cardiovascular collapse are Toleikis JR: Effects of midazolam on median nerve s o m a t o s e n s o r y
reached. Similarly, the SEP is unaffected at doses that produce evoked potentials. Br J Anaesth 1990;64:590, with permission.)
242
Barbiturates are not commonly used during recording of Etomidate

tcMEP because the CMAP responses are particularly sensitive Like ketamine, etomidate has the unusual effect of response
to barbiturates. Further, the effect appears quite prolonged; in enhancement. Hence it has become a desirable agent for some
one study, the induction bolus eliminated the CMAP from monitoring uses. Etomidate increases the amplitude of cortical
tcEMEP for a period of 45-60 minutes, suggesting that barbi-
417
SEP components following injection with no changes in sub- 194
turates are a poor induction choice when monitoring with this cortical and peripheral sensory responses (Fig. 12-15). This am-
modality is needed. For this reason, most anesthetic protocols plitude increase appears coincident with the myoclonus seen
do not use induction with thiopental. One exception, methohex- with the drug, suggesting a heightened cortical excitability (no
ital, has been used in one TIVA protocol with opioids and keta- evidence of seizure activity, however, was seen). A sustained
mine. Fortunately, this drug is more rapidly metabolized and
417
amplitude increase with constant drug infusion has been used to
appears to have excitatory properties (low doses can be used to enhance SEP cortical recordings that were otherwise not moni-
identify seizure foci during cortical mapping of epilepsy). torable. This effect has also been used to enhance amplitude
361
in motor evoked responses. - Fortunately, the enhancing ac-

194 361
Benzodiazepines tivity occurs at doses that are consistent with the desired degree
Midazolam has desirable properties of amnesia and has been of sedation and amnesia needed for TIVA.
used for monitoring cortical SEPs. However, because of its sig- Studies with tcMEP have suggested that etomidate is an ex-
nificant effects on myogenic MEP and its slow metabolism, it cellent agent for induction and monitoring of these modali-
has been replaced by other agents. Midazolam, in doses consis- ties. Of the several intravenous agents studied, etomidate had
368
tent with induction of anesthesia (0.2 mg/kg) and in the ab- the least degree of amplitude depression after induction doses or
sence of other agents, produces a mild depression of cortical continual intravenous infusion. Thus, etomidate has been 112
S E P and minimal effects on subcortical and peripheral sen-

363
used for induction of anesthesia and as a component of TIVA,
sory evoked responses (Fig. 12-14). As with thiopental, mida- combined with opioids. - - - 144 203 204 434
zolam produces prolonged marked depression of tcMMEP,

suggesting that it also may be a poor induction agent for MEP Propofol
recording. This effect has been interpreted as due to inhibi-
172
Propofol has a very rapid metabolism, making it an excellent
tion of cortical pyramidal cell neurons. In addition to possible component of TIVA. Although propofol is a depressant agent,
cortical locations for the benzodiazepine effect, an effect at the adjustment of infusions often will allow adequate monitoring.
spinal cord dorsal root has been suggested by a study of poste- Propofol induction produces amplitude depression in cortical
rior tibial stimulation, which revealed a marked decrease in
170
AEP (Fig. 12-16) and cortical SEP with rapid recovery after ter-
the amplitude of the H-reflex with no effect on the stimulated mination of infusion. When the SEP is recorded in the epidural
51
CMAP (M-wave). space, propofol has no significant effect. This is consistent with
the postulated site of anesthetic action of propofol on the cere-
bral cortex. Studies with transcranial electrically or magneti-
12
cally elicited motor evoked potentials have demonstrated a

depressant effect on response amplitude, also consistent with a
cortical effect. - - Propofol does not appear to enhance cor-
160 172 181
tical responses, but its rapid metabolism allows the depth of anes-
thesia and effects on evoked responses to be adjusted rapidly.
Post auricular peak

V I Pa Nb
i.v
2pV
0 Time (ms) 50
S0MAT0SENS0RISCH EVOZIERTE POTENTIALE (SEP)
NACH MEDIANUSSTIMULATION
Figure 12-16. Cortical AEP before anesthesia and at differ-
Figure 12-15. Cortical SEP from median nerve stimulation ent concentrations of propofol. Arrows indicate t h e position of
before and following 10 mg etomidate. (From Russ W.Thiel A.Schwandt waves V, Pa, and N b . (From Chassard D, Joubaud A, Colson A, e t al:
HJ, Hempelmann G: Somatosensorisch evozierte Potentiale unter Auditory evoked potentials during propofol anaesthesia in man. Br J
thiopental und etomidat.Anaesthesist 1986; 35:679, with permission.) Anaesth 1989;62:522, with permission.)
However, as a component of TIVA, infusions of propofol have

been combined with opioids to produce acceptable conditions
for myogenic MEP monitoring. - - -
47 I66 308 309
Regional Anesthesia
Major conduction anesthesia (e.g., epidural or spinal) does
not appear to be an acceptable alternative to general anesthesia
for monitoring evoked responses that depend on the tracts anes-
thetized. This effect has also been seen with intravenous re-
371
gional block and specific nerve blocks. However, local

Fraction EMG Remaining
anesthesia placed away from the neural pathway mediating the
monitored evoked response (such as scalp anesthesia for awake Figure 12-17. Plot of tcMMEP amplitude as recorded from
craniotomy) is satisfactory for monitoring unless systemic ab- t h e n a r muscles as plotted versus t h e fraction of a single twitch r e -
sorption is substantial. maining following median nerve stimulation. (From Sloan TB, Erian R:
Effect of atracurium induced neuromuscular block on cortical m o t o r
Muscle Relaxants evoked potentials. Anesth Analg 1993;76:979, with permission.)
Since muscle relaxants have their major site of action at the
neuromuscular junction, they have little effect on electrophysio-
logic recordings such as the SEP that do not derive from muscle OTHER INTRAOPERATIVE
activity. In fact, they may improve, or be essential for, some PHYSIOLOGIC FACTORS
types of recordings where the muscle activity near the recording
electrode may create unwanted noise. This is true for epidural BLOOD FLOW
or peripheral nerve recordings where the activity of overlying
muscle obscures the response from transcranial stimulation. For As measures of neural function, the evoked responses are re-
recording of these responses, complete or near-complete neuro- sponsive to a variety of physiologic factors that alter neuronal
muscular blockade is highly desirable. function and viability. Numerous studies - - - - - 14
have
28 29 33 392 394
Certainly, complete neuromuscular blockade will prevent demonstrated a threshold relationship between regional cerebral
recording of CMAP responses during MEP. However, partial blood flow and cortical evoked responses. The cortical SEP re-
neuromuscular blockade has the benefit of reducing a substan- mains normal until blood flow is reduced to about 2 0
tial portion of the movement that accompanies the testing. Some ml/min/100 g m . At more restricted blood flows of between 15
360
degree of neuromuscular blockade may facilitate some surgical and 18 ml/min/100 gm of tissue, the S E P is altered and lost
procedures where muscle relaxation is needed for adequate (Fig. 12-18). "' - * 19.i34.i93,2i3 Subcortical responses appear less
3 32 64 1
tissue retraction. sensitive, though global hypotension, affecting both subcortical

Two methods are customarily utilized to assess the degree of and cortical blood flow, is associated with cortical SEP loss at
neuromuscular blockade. The method that best quantitates the higher rates of cerebral blood flow than with middle cerebral
blockade involves measuring the amplitude of the CMAP ( T l ) artery occlusion, which more selectively impairs cortical than
or M-wave produced by supramaximal peripheral motor nerve subcortical blood flow. - The difference in sensitivity to is-
119 213
stimulation. When neuromuscular monitoring is conducted this chemia between cortical and subcortical structures may explain
way, most monitoring protocols use neuromuscular blockade of why the central conduction time (CCT) of the SEP bears a
Tl that is 10-20% of baseline. Clinically, anesthesiologists parametric relationship to cerebral blood flow (Fig. 12-18). - 103 134
often assess neuromuscular blockade by counting the number of Regional factors may produce focal ischemia not predicted
twitches remaining when four motor nerve stimuli are delivered by systemic blood pressure. For example, during spinal surgery,
at a rate of 2 Hz. Measured this way, acceptable CMAP moni- the effects of hypotension may be aggravated by spinal distrac-
toring has been conducted with only 2 of 4 responses remain- tion, such that an acceptable limit of systemic hypotension
ing (this corresponds to a T l response in the range of
47
cannot be determined without m o n i t o r i n g . - Other 35 82120122
10-20%). When intense neuromuscular blockade is required examples include peripheral nerve ischemia from positioning,
(e.g., recording of epidural or neurogenic responses), T l re-
sponse of less than 10%, or no more than 1 of 4 twitches, is gen-
erally recommended.
When using neuromuscular blockade, tight control of the
blockade is necessary so that excessive blockade does not elim-
inate the ability to record or mimic the loss of the response with
neural injury (Fig. 12-17), creating a false-positive result. 366
Because of varying muscle sensitivity to muscle relaxants, the

neuromuscular blockade may need to be evaluated continuously
EES
in the specific muscle groups used for monitoring. It is impor-
ec/min/IOOgm^
tant to note that the use of neuromuscular blockade is controver-
O IO 20 30 50
sial during monitoring of muscle responses from mechanical REGIONAL CEREBRAL BLOOD FLOW
stimulation of nerves, as partial paralysis may reduce the ability
to detect these responses (e.g., facial nerve monitoring or moni- Figure 12-18. Relationship between the SEP and EEG elec-
toring for pedicle screw placement). Often muscle relaxants are trical response and regional cerebral blood flow. (From Sloan T:
avoided when monitoring these latter mechanically evoked American Society of Anesthesiologists: Refresher Courses, O c t o b e r
muscle EMG neurotonic responses. 1985, Lecture 211, with permission.)
tourniquets, or vascular interruption, - - spinal cord is- BLOOD RHEOLOGY

98 280 432
chemia from aortic blood flow compromise, carotid artery inter-

ruption, vertebrobasilar insufficiency aggravated by head
331
Since changes in hematocrit can alter both oxygen carrying
extension, cerebral artery constriction by vasospasm, and cere- capacity and blood viscosity, the maximum oxygen delivery is
bral ischemia due to retractor pressure. 394
often thought to occur in a mid-range hematocrit (30-32%).
MEPs and SEPs are sensitive to spinal cord events produced Evoked response changes with hematocrit are consistent with
by vascular ischemia (aortic cross-clamping) or mechanical this optimum range. In a study of VEPs and upper limb SEPs in
compression (epidural balloon). However, because these tracts the' baboon, N a g a o observed an increase in amplitude with
269
are topographically removed from one another, the MEP and mild anemia, an increase in latency at hematocrits of 10-15%,
SEP may show differential sensitivity to an ischemic event. 149
and further latency changes and amplitude reductions at hemat-
MEP studies using transcranial^ generated MEPs and ocrits below 10%. These changes were partially restored by an
recorded epidurally are sensitive to ischemia but not to anterior increase in the hematocrit.
horn cell injury. This is postulated to be due to persistent con-
duction in the corticospinal tracts.. This is in contrast to the VENTILATION
92
recording of peripheral nerve or muscle response with MEPs, in

which anterior horn cell injury can destroy the anterior horn cell In addition to changes in oxygenation, alterations in ventila-
function that is required to translate the descending neural tion can alter blood carbon dioxide, thus altering spinal cord and
signal into a peripheral nerve or muscle response. As with the 389
cortical blood flow (hypocapnia producing vasoconstriction and
SEP, myogenic MEP is sensitive to spinal cord ischemia associ- hypercapnia producing vasodilation). The most significant
ated with thoracic aortic clamping, - - and a decrease in re-
149 208 389
changes in cortical SEP occur when the carbon dioxide is ex-
sponse has been shown to correlate with reduced spinal cord tremely low, suggesting excessive vasoconstriction may pro-
blood f l o w . 92149
duce ischemia (carbon dioxide tensions below 20 mm Hg).
Hypocapnia may aggravate hypotension as a result of arterial
INTRACRANIAL PRESSURE vasoconstriction. This effect has been suggested to contribute to
alterations in SEP during spinal surgery or in BAEP during 124
Another factor leading to regional (cortical) ischemia is ele- posterior fossa surgery in the sitting position. 124
vated intracranial pressure (ICP). Elevated ICP is associated

with reductions in amplitude and increases in latency of corti- TEMPERATURE
cally generated visual, somatosensory, and brain stem auditory
evoked responses. The BAEP is altered as uncal herniation Hypothermia can also alter evoked responses by changing
occurs. The relationship of the VEP to ICP has suggested the
270
nerve depolarization (increased action potential duration, re- 187
VEP as a means of noninvasive ICP testing (Fig. 12-19). 435

duced conduction velocity, - and decreased synaptic func-
73 197
Increased ICP, probably by virtue of its effect on cortical tion ), resulting in latency increases and decreases in evoked
418
structures, produces a gradual increase in onset of the tcMMEP response amplitude. These changes have been observed in
83
until a response can no longer be detected (i.e., threshold ex- visual, - brain stem auditory, - cortical auditory, and
323 431 85168 380 183
ceeds the capacity of the stimulator). T h e increase in latency somatosensory evoked potentials. - - - Contrarily, induced
41 83 155 316
suggests that the central component of the motor pathway has hyperthermia can reduce SEP latencies, but this is less often an
87
slowed conduction velocity. 365

intraoperative concern. Hypothermia appears to affect central
nervous system synaptic function more than conduction, prob- 41
HYPOXEMIA ably by interference in the postsynaptic membrane. Thus, 399
changes are more prominent at the cephalic end of long neural

Hypoxemia can also cause evoked potential deterioration. tracts (such as the SEP) or in components of responses associ-
This has been recorded in one case when the P 0 reached 41 a 2
ated with multiple synaptic elements (Fig. 12-20). Hence, re-
mmHg before other clinical parameters had changed.
121
sponses recorded from peripheral nerves are minimally affected,
whereas those produced by cortical structures are markedly af-
fected for the same degree of c o o l i n g . Core temperatures83155
IOOO
o 800
X
E
£ 600
CL
U
400
200-
70 80 90 100 110 120 Milliseconds 70

N2 LATENCY (msec)
Figure 12-20. Changes in the cortical S E P (median nerve)
Figure 12-19. Relationship of latency of V E P with intracra- with whole body hypothermia as esophageal temperature is low-
nial pressure. (From York DH, Pulliam MW, Rosenfeld JG, Watts C: ered from 37 t o 28 degrees Centigrade. (From Sloan TB: Evoked po-
Relationship between visual evoked potentials and intracranial pres- tentials. In Albin MS (ed): Neuroanesthesia. N e w York, McGraw-Hill,
sure. J Neurosurg 1981,55:909, with permission.) 1997, pp 221-276, with permission.)
frequently drop by greater than 1°C, but peripheral nerves gradually came into use. The intraoperative application and
lifted from the body for focal stimulation may be as cool as clinical utility of EEG, BAEPs, SEPs, and EMG in a variety of
20°-30°C. 38
different surgical procedures are discussed below. The clinical
Whole body hypothermia, either inadvertent or intentional, is utility of VEPs is very limited; therefore, VEPs will be only
the most obvious temperature change that occurs during briefly discussed.
surgery. Changes in regional temperature can also occur, result- The cortical EEG is produced by the spontaneously gener-
ing in evoked response alterations that would not be otherwise ated electroencephalographic brain wave activity and is easily
predicted based on unchanged body (core) temperature. For ex- recorded, requiring no stimulus. However, evoked potential
ample, cold irrigation solutions applied to the spinal cord, 61
(EP) recordings can only be generated by applying an external
brain stem, or cortex routinely cause evoked response changes. stimulus. In general, stimulation is provided peripherally using
These cold irrigation solutions may also irritate the nerve, caus- peripheral nerve electrical stimulation for SEPs, auditory clicks
ing increased muscular activity if the nerve has motor compo- for BAEPs, and light flashes for VEPs. In addition to recording
nents. Similarly, limb cooling (as from cold intravenous
256
EPs, spontaneous or stimulus triggered EMG can be recorded
solutions) can alter the SEP originating from stimulation to a from muscle using either intramuscular or surface electrodes.
nerve from that limb. Electrical or magnetic stimulation of peripheral nerves, nerve
With hypothermia, the tcMMEP demonstrates a gradual in- roots, spinal cord, cranial nerves, or cortical structures can gen-
crease in onset as temperature decreases from 38°C to 32°C erate a motor evoked response that also can be recorded from a
esophageal. An increase in stimulation threshold has also been muscle or nerve.
observed at lower temperatures. This is consistent with both
cortical initiation and peripheral conduction being affected by ISCHEMIA AND ELECTROPHYSIOLOGIC STUDIES
the temperature drop. 364
Different techniques have been used in order to identify and

OTHER PHYSIOLOGIC VARIABLES attempt to prevent damage caused by cerebral ischemia. The
most extensively studied and commonly used techniques are
Changes in a variety of other physiologic variables may EEG and SEPs. The rationale for employing electrophysiologic
produce alterations in the evoked responses during surgical techniques as a marker for ischemia is the good correlation be-
monitoring. Significant reduction in blood volume can alter tween these techniques and regional cerebral blood flow
j evoked responses as a result of changes in blood flow distribu- (rCBF). As previously noted, studies demonstrated that in pa-
I tion, despite absence of significant blood pressure changes tients undergoing carotid endarterectomies (CEA), major EEG
(e.g., limb ischemia altering the SEP as blood flow to central changes occurred with rCBF < 10 ml/100 gm/min, and less
organs is spared). An increase in superior vena caval pressure severe EEG changes were seen with rCBF between 10 and 18
during cardiopulmonary bypass has been associated with SEP ml/100 gm/min, with a critical level defined as 15 ml/100
changes. 148
gm/min. - - In contrast, primate studies show that SEPs are
352 386 387
Other physiologic events may occur too slowly to be noted as maintained at levels of rCBF >16 ml/100 gm/min but absent at
changes in the evoked response. For example, changes in glu- levels below 12 ml/100 gm/min. At rCBF levels between 14 and
cose, sodium, potassium, and other electrolytes important in
74
16 ml/100 gm/min, there is a sharp decline in the evoked re-
the neurochemical environment are likely to also result in sponse amplitude, with a 50% amplitude reduction correspond-
evoked response changes over time. ing to a rCBF of 16 ml/100 gm/min. - - In addition to
28 30 32 391
altering the SEP amplitude, ischemia also appears to prolong

the CCT, with a threshold rCBF (<15 ml/100 gm/min) similar to
INTRAOPERATIVE NEUROPHYSIOLOGIC those previously reported for EP amplitude r e d u c t i o n . 134213
Interestingly, there is experimental evidence suggesting a dif-

MONITORING: SPECIFIC APPLICATIONS ferential susceptibility to local ischemia as one descends the
As previously noted, intraoperative monitoring (IOM) can neuraxis, demonstrated by increasing resistance of electrophysi-
be described as the application of neurophysiologic, usually ologic function to systemic hypotension. 32
electrophysiologic, techniques to detect changes in the func- There clearly seems to be a threshold relationship between
tional state of the nervous system consistent with ischemia or cerebral blood flow and alteration of the EEG and SEP. There is
injury. Electrophysiologic techniques may also assist in localiz- also a good deal of experimental evidence for a rCBF threshold
ing neural structures, identifying specific cortical functional and cellular membrane failure. - An investigation has found
15 391
areas, and delineating an epileptogenic cortex. The information that at rCBF levels between 12 and 16 ml/100 gm/min, the SEP
obtained may also help determine the mechanism of injury, cor- was abolished and small, self-limiting increases in extracellular
related with the surgical proceedings, and serves to prevent potassium activity were detected. This study further demon-
30
damage by detecting cellular dysfunction prior to its reaching strated that the rCBF threshold range for a massive irreversible
an irreversible cell death stage. Outcome studies have in general rise in extracellular potassium, associated with structural
been promising in that intraoperative electrophysiologic studies changes of infarction, occurred between 7.6 and 11.4 ml/100
do make a positive difference in patient care. - - - - - -
69 245 287 290 295 399 414
gm/min. In baboon chronic stroke models, following middle
Intraoperative neurophysiologic monitoring for surgical pro- cerebral artery (MCA) occlusion, areas of infarction corre-
cedures where the central and peripheral nervous systems are at sponded to blood flow levels of 10 ml/100 gm/min or less. - 390 391
risk has become increasingly popular over the past several In acute stroke primate models, infarction occurred only in
years. IOM of facial nerve activity, utilizing EMG, was first per- areas where rCBF measured < 12 ml/100 gm/min (Table 12¬
formed in the late 19th century, and EEG was subsequently used 12).167.186.260 Thus, these findings suggest that significant (50%)
in 1965. Other electrophysiologic techniques such as SEPs,
310
reduction in amplitude of the SEP, which corresponds to a rCBF
BAEPs, VEPs, and peripheral nerve compound action potentials of 14-16 ml/100 gm/min, is indicative of ischemia and is a
Table 12-12. Clinical Correlates with Cerebral Perfusion investigators had "never had a patient emerge from anesthesia
Pressure (CPP) and Cerebral Blood Flow (CBF) with a new deficit that was not predicted by the E E G . " The 387
CPP CBF use of EEG monitoring has also resulted in the reduction in fre-
(mmHg) (ml/IOOg/min) Clinical Finding quency of intraoperative carotid shunt use as well as improved
neurological outcome. - Assessment of the EEG data from
56 369
100 50 Normal
the different studies suggests that EEG is very sensitive in de-
50 > 20-25 Slowing EEG; cerebral impairment tecting cerebral ischemia and useful in predicting outcome.
25-40 15-20 EEG may be flat
Intraoperative EEG Changes
<25 < 10 Irreversible brain damage
EEG changes may be due to several different factors in the
From King FG: Functional neurophysiology. In Russell GB, Rodichok LD (eds): operating room.
Primer of Intraoperative Monitoring. Boston, Butterworth-Heinemann, 1995, p
30, with permission. Ischemia. The two major changes are generalized slowing
and decreased amplitude from the ischemic hemisphere - 310 407
and attenuation of the anesthetic-induced fast rhythms. As ex-

55
warning of possible progression from reversible to irreversible pected, ipsilateral hemispheric EEG changes are the most
cell damage. common findings due to ischemia after unilateral carotid cross-
clamping. However, bilateral EEG changes have been reported
CAROTID ENDARTERECTOMY in patients with compromised collateral circulation. It has also
53
been noted that patients with, an abnormal preoperative EEG

Possible brain ischemia as a direct result of CEA is an impor- have a higher frequency of intraoperative EEG changes. 387
tant factor that requires special attention, since cross-clamping Anesthesia. Anesthetics have a significant effect on the
of the common carotid artery is an unavoidable component of EEG, typically producing rhythmic beta activity. Initially,
CEA. If collateral circulation is inadequate, ipsilateral hemi- during induction, 12-18 Hz beta waves appear over the anterior
spheric ischemia is the result of carotid cross-clamping. With hemispheres. At lighter levels of steady-state anesthesia, the
this in mind, a variety of IOM techniques have been used to amplitude increases, becomes widespread, and slows to 8-14
identify the presence and magnitude of such cerebral ischemia. Hz. Intermittent delta waves may also be seen anteriorly. Burst
In general, techniques such as stump pressure measurements, suppression of these patterns can be seen with high doses of dif-
radionuclide cerebral blood flow measurements, and transcra- ferent anesthetics, including isoflurane, halothane, and enflu-
nial Doppler (TCD) assess cerebrovascular integrity; whereas rane as well as barbiturates. EEG changes due to anesthesia
250
EEG and cortical SEPs indirectly measure cerebral ischemia by are bilateral and symmetrical.
assessing cerebral function. Based on the changes seen during Mean Arterial Pressure (MAP). In some cases, bilateral or
IOM, some surgeons use induced systemic hypertension and/or unilateral (ipsilateral to diseased carotid) EEG changes consis-
intraluminal shunting as strategies to reverse detected cerebral tent with ischemia can be seen when the MAP drops below a
ischemia. Unfortunately, there is no clear consensus regarding certain critical level, which is dependent on the patient's base-,
the necessity for an indwelling arterial shunt during CEA. line MAP (JL, personal observation).
Surgeons usually belong to one of three groups: those who Interventions. Ipsilateral EEG changes after carotid artery
always shunt, those who never shunt, and those who selectively cross-clamping are interpreted as due to reduced unilateral
shunt based on the results of IOM. - The routine use of shunts
99 224
cerebral perfusion. Thus, in order to reverse cerebral ischemia,
in CEA is controversial because shunt placement is not without surgeons who use IOM will usually shunt once the EEG shows
complications and may result in embolization and possible major changes. There is clinical evidence to suggest that
stroke or arterial damage. Additionally, shunt malfunc-
100133 222
shunting those cases with major EEG changes consistent with
tions from various causes have been reported. 218,420
cerebral ischemia leads to improved overall neurologic out-
come. - - - A study comparing neurologic outcome in two
56 250 387 388
Continuous EEG Monitoring groups of CEAs performed either with or without EEG IOM
EEG monitoring during CEA has been in use for over 30 found that the use of IOM and selective shunting reduced the
years. Initial studies compared EEG changes with postoperative frequency of carotid shunt use from 49% to 12% of cases and
neurologic deficits. Subsequently, EEG monitoring grew in
310
decreased major neurologic morbidity and mortality from
popularity because of readily available equipment, familiarity 2.3% to 1.1%. 52
with technique, ease of set-up, and reliability in determining

carotid artery cross-clamp-dependent ischemia. The ability to Computer-Processed EEG
determine ischemia and predict postoperative neurologic Conventional intraoperative EEG generates a large amount of
deficits reliably with EEG was reported by several data and requires the use of large, complex equipment, precise
groups.56.352.387.4i5 Patients with EEG changes during CEA were time-consuming electrode placement, and trained personnel to
also found to have a higher incidence of stroke as opposed to continuously interpret the EEG data. Therefore, in an attempt to
patients who did not develop changes. In a series of 293 CEAs, facilitate IOM using EEG, digitally processed EEG methods
the postoperative neurologic deficit risk was much higher in the have been developed. - Spectral analysis of EEG data has
35 268
subgroup with major EEG changes, 18%, as opposed to those been performed by digitizing analog EEG signals and, using a
without changes, 2 % . In a separate study, 59 of 369 patients
322
computer to perform a Fourier analysis, producing a power
(431 CEAs) developed EEG changes and 17% of these (10 of versus frequency display. Once the EEG has been digitized
268
59) developed postoperative complications. EEG changes 438

and computer processed, it can be displayed in different formats
were more predicative than rCBF, with EEG changes predicting including compressed spectral arrays (CSA), which are dis-
50% of the new deficits in those with false negative r C B F . 438
played in a linear fashion, and density-modulated spectral
Another study reported that in 1 152 consecutive CEAs the arrays (DSA), which are shown by areas of density (intensity). 216
•Vy. y^^K-v^w n ^ ^ A v - v ~*-"
— .-.^
Figure 12-21. Intraoperative EEG and SEP recordings were o b -

tained from a 67-year-old female with a history of severe left carotid
stenosis undergoing a left carotid endarterectomy.
Traces 1-3: Left-sided EEG (F , C , a n d P ; referenced t o linked ears)
3 3 3
Traces 4-6: Right-sided EEG (F , C , and P ; referenced t o linked

4 4 4
ears)
Trace 7: Right cortical (C '-Fz) SEP after left median nerve stimu-
4
lation
Trace 8: Left cortical (C3'-Fz) SEP after right median nerve stim-
ulation
A, Precarotid cross-clamp baseline EEG shows burst-suppression pat-
t e r n as a result of thiopental administration for neuroprotection. B,
Five minutes after cross-clamping of t h e left internal carotid artery.
N o t e attenuation of the left hemispheric EEG and a > 50% amplitude
reduction of t h e left cortical SEP. C,Two minutes after shunt place-
ment; t h e SEP is back t o baseline, and t h e r e is improvement of t h e
EEG. Patient developed no new neurologic deficits.
These types of displays allow for easier interpretation of the cortex, which is supplied by the M C A and thus is at risk
EEG, showing ischemic changes as loss of amplitude in the during carotid cross-clamping. There are also several technical
power spectrum for all frequencies and a shift of the EEG spec- advantages of SEPs over EEG for IOM, such as relative resis-
trum to the lower frequency range. Several studies using com- tance to general anesthesia, fewer electrode sites, easier
puterized EEG analyses have found the technique useful in recording and interpretation, and serial comparison. - - 76 200 241
identifying cerebral ischemia and predicting neurologic out- Moorthy et al. were the first to report the use of median nerve
come following CEA. In fact, some investigators have
4 5 5 3 1 9 3 2 0
SEPs during CEA and to correlate S E P changes with neuro-
claimed that computer-processed EEG proved more useful than logic deterioration. Subsequently, Markand reported on 38
259
analog EEG. - However, a prospective study of 103 patients

4 55
CEAs monitored with S E P . One patient had a sudden loss of
240
undergoing CEA monitored with analog EEG and DSA con- consciousness and loss of the cortical SEP following carotid
cluded that DSA was not sufficiently sensitive in detecting lim- cross-clamping. The SEP changes were reversed, and the pa-
ited cerebral ischemia associated with mild analog EEG tient regained consciousness after shunt placement. In three
changes. Thus, it appears that there is a role for computerized
179
other cases, shunting reversed SEP changes. Other investiga-
EEG analysis during CEAs, with the caveat being that it may tors have reported on the reliability and accuracy of SEPs in
lack sensitivity in detecting mild EEG real-time pattern changes CEAs. - "
7 76
- - - However, at least one study has con-
0138178 333 347 404
consistent with cerebral ischemia. cluded that SEPs are not sensitive enough markers of cerebral
ischemia because neither 50% amplitude reduction nor in-
S o m a t o s e n s o r y Evoked Potentials crease in CCT has been established as a physiologic marker of
As previously mentioned, there is a strong correlation be- intraoperative ischemia in h u m a n s . However, in this study
200
tween SEPs and cerebral ischemia; therefore, SEPs have SEP criteria were measured against the EEG "gold standard"
gained popularity as a useful IOM technique to assess cerebral instead of using neurologic outcome as an endpoint. The in-
function during cerebrovascular surgery. Median nerve
158
vestigators found that 10 out of 23 patients with EEG evidence
SEPs reflect the cerebral functional status of the primary sensory of ischemia after cross-clamping had prolongation of the C C T
Figure 12-22. 28-year-old female admitted for clipping of a

giant right M C A trifurcation aneurysm w h o presented with left
face and arm numbness.
Trace I: Right cortical SEP (C '-Fz) after left median nerve stimula-
4
tion
Trace 2: Left Erb's EP (left Erb's-Fz) after left median nerve stimula-
tion
Trace 3: Left cortical SEP (C '-Fz) after right median nerve stimula-
3
tion
Trace 4: Right Erb's EP (right Erb's-Fz) after right median nerve stim-
ulation
A, Baseline SEPs. B, Loss of right cortical SEP 2 minutes after t e m p o -
rary clipping of the MI branch of the right MCA. C, Recovery of corti-
cal SEP 2 minutes after t e m p o r a r y clip was removed. Patient
experienced no new postoperative neurologic deficits.
but only one had amplitude reduction of 50% or more. In con- Intraoperative SEP Changes:
trast, another group stated that there is a higher incidence of Ischemia. As noted above, severe cerebral ischemia causes a
false-positives and the use of shunts with EEG, since there is reduction of the cortical SEP amplitude to < 50% of the base-
less agreement in identifying EEG evidence of ischemic risk line value. In addition, the peak latency of the cortical SEP can
as compared with SEPs. Although SEP criteria for cerebral
7
be delayed and the CCT prolonged.
ischemia are not uniform and differ among various investiga- Temperature. Hypothermia can prolong the latencies of all
tors, with some using amplitude reduction, latency delay, or peaks and needs to be taken into consideration because surgery
specific prolongation of CCT or a combination of the three to frequently leads to a 1-3°C drop in the core body temperature.
signify cerebral ischemia, most studies have found that a 50% Hypotension. Can lead to amplitude reduction and/or delay
amplitude reduction of the N19-P24 or a 1-millisecond delay in the cortical SEP latencies. Critical level depends on the pa-
of the C C T correlates best with postoperative neurologic tient's preoperative blood pressure.
de Fic i ts. - - °-
34 77 11
- -
142,178 179 344
Interventions. The interventions performed when SEP
A review of seven studies comprising 3028 patients re- changes occur are similar to those described for EEG changes.
ported that 170 patients, or 5.6%, had a significant decline in Installation of a shunt is the most common strategy. However,
intraoperative SEPs that correlated with surgical manipula- head repositioning and restoration of blood pressure may also
tion and that 34, or 20%, of the 170 patients developed signif- be used. - '
126 3 8
icant postoperative deficits. Additionally, the reported

99
sensitivity and specificity in six studies, between 1985 and INTRACRANIAL NEUROVASCULAR SURGERY
1991, were 83-100% and 83-99%, respectively. It was con- 200
cluded that conventional EEG and SEP monitoring modalities The type of electrophysiologic technique to be used de-
during CEAs have similar sensitivity and specificity. A sepa- pends on several factors, such as type of surgery, anatomic lo-
rate study found that intraoperative SEP amplitude decre- cation, intracranial vascular supply at risk, and preoperative
ments of 5 0 % or more were associated with worsening of neurologic deficits. EEG monitoring has had limited use in
neurophysiologic function, determined by comparing pre- aneurysm surgery because craniotomy and brain relaxation
and postoperative neurophysiologic testing. These findings 34
produce air spaces between the dura and arachnoid that inter-
suggest that SEP IOM is useful in improving neurologic out- fere with recording of the E E G . However, a combination
93
come during CEA (Fig. 12-21). 50

of EEG and SEPs can be used to monitor MCA aneurysm
C4'/C3'-FZ RIGHT LEFT C4'/C3'-FZ RIGHT
l-»t PP f*np L»t PP Pm ip

20.0 30 .7 1 .40 20 .2 20 ST . T o
J
OT/S'-Erb's
34 4 20 . B 1 OO23 .7 28 .0'"7.32
is a
CZ-FPZ
PP Fkip PP ftnp
42.« 47 . O0 .07 44 O40.S
SO 6 02 *
CS'/C^'-FZ
l-Mt
Fknp PP FWpi
SO Qtn "i.o 49 . * D S4
9 «J oS O
LEFT C4'/C3'-FZ RIGHT
Lit PP P«ip PP Fknp

2S .7 31 O1 .30 25 320 .3"?.oo
• XtH "
C4'/3'-Erb's
PP f*np PP »np
24 .4 ao .4 1 SO23 .7 28 .4"7 47
13 . 1
CZ-FPZ
P> P«p PP P
BJO
43 .7 40 Q O SB4<.1 4D .a 1 O
.O
oa a
C3*/C4'-FZ
'-*'
pp p-p t.»t PP pjnp
30 a 45 SQ . 72 412 40 .2 'o.03
«o 1 'o* S
Figure 12-23. 48-year-old male presented with a 2.2-cm anterior communicating artery aneurysm after a subarachnoid hemor-
rhage.
Trace I: Right cortical (C '-Fz) SEP after left median nerve stimulation
4
Trace 2: Left Erb's (C '-ipsilateral Erb's) SEP after left median nerve stimulation
4
Trace 3: Left cortical (C '-Fz) SEP after right median nerve stimulation
3
Trace 4: Right Erb's (C '-ipsilateral Erb's) SEP after right median nerve stimulation
3
Trace 5: Right cortical (Cz-Fz) SEP after left posterior tibial nerve stimulation
Trace 6: Right cortical (Cz'-Fz) SEP after left posterior tibial nerve stimulation
Trace 7: Left cortical (Cz-Fz) SEP after right posterior tibial nerve stimulation
Trace 8: Left cortical (Cz'-Fz) SEP after right posterior tibial nerve stimulation
A, Baseline intraoperative SEPs. B, Five minutes after temporary clips were applied t o both AI arteries, there was a 5 0 - 6 5 % amplitude reduction
in the cortical SEPs after right leg stimulation. C, Removal of the temporary clips resulted in recovery of the SEPs over 21 minutes.The patient did
not develop any postoperative new neurologic deficits.
surgery. The EEG was used to detect burst suppression pat- ANEURYSM
304
terns, thus allowing for the minimal dosage of thiopental re-

quired to achieve neuroprotection, while SEPs monitored Middle Cerebral Artery/Internal Carotid Artery
cerebral ischemia. A review of the reported clinical outcomes in 10 studies in-
MCA and internal carotid artery (ICA) vascular territories volving MCA aneurysm surgery demonstrated the reliability of
can be monitored with median nerve SEPs, since the so- median nerve SEPs in predicting outcome, which ranged from
matosensory cortex representing the hand is supplied by the 7 8 - 1 0 0 % . An investigation of intraoperative monitoring in 50
303
MCA. In addition, median nerve SEPs also provide informa- aneurysm surgeries (12 MCA) reported that a CCT greater than
tion on the posterior cerebral artery (PCA) vascular territory 9 milliseconds (ms) correlated with postoperative neurologic
through the monitoring of thalamic activity. Several studies
39
deficits. In addition to prolongation of the CCT, another study
77
have demonstrated the utility of median nerve SEPs in pre- found that a decrease in cortical SEP amplitude or disappear-
dicting postoperative neurologic outcome. ' - - ° - - - -
39 40 8IJ04 , 5 248 258 343
ance of the cortical EP accurately predicted postoperative
392.393 deficits. A study of 53 MCA aneurysm procedures found new
104
T irwr 1«-M=<V. IS 38 3S
c
- ^-^^-^
B
Figure 12-24. Intraoperative monitoring of a 33-year-old

female undergoing resection of a left sylvian and insular region 3-
cm diameter angiographically occult vascular malformation after pre-
senting with a first-onset seizure.
Traces I and 2: Right cortical SEPs ( C ' - F z and 0 4 - 0 3 ' ) after left
4
median nerve stimulation

Traces 3 and 4: Left cortical SEPs ( C ' - F z and C3-C4') after right
3
median nerve stimulation

Traces 5 and 6: Right cortical SEPs (Cz-Fz and Cz'-Fz) after left
posterior tibial nerve stimulation
Traces 7 and 8: Left cortical SEPs (Cz-Fz and Cz'-Fz) after right
posterior tibial nerve stimulation
A, Baseline SEPs. B, Surgeon alerted of a > 50% amplitude reduction of
the left cortical SEPs (following right median nerve stimulation) during
AVM resection. C , N o improvement was seen in t h e SEPs in spite of
raising the patient's mean arterial pressure. Postoperatively, the patient
was noted t o have new right arm weakness and numbness.
postoperative deficits in the following: four out of four patients identifying significant ischemia due to accidental clipping of
with a significant change in the SEP and incomplete return to blood vessels, manipulation of brain structures, and excessive re-
baseline or loss of the EP; and one out of five patients with signif- traction. In a study of 134 aneurysms, there was reported
81343
icant SEP changes that returned to baseline. In contrast, only

105
loss or alteration of SEPs in 12 cases of temporary occlusion,
one out of 37 patients without SEP changes developed a new post- two cases of accidental clipping, one case of permanent occlu-
operative deficit. An investigation monitoring 37 patients undergo- sion, and one case each of retraction of the MCA and cerebellar
ing aneurysm surgery (17 MCA) concluded that preservation of retraction. In 15 out of 17 instances of SEP changes, the surgi-
343
conduction, even if CCT increased to 10 ms, was associated with a cal course was altered. Changes included repositioning of vascu-
good outcome. However, prompt disappearance of the N20
393
lar clips and repositioning of retractors. In rare instances, cortical
(within 1 minute), as well as slow recovery of the SEP (> 20 min- SEP changes may identify a peripheral ischemic process such as
utes), was associated with new postoperative deficits. a malfunctioning blood pressure cuff causing arm ischemia. 105
In an attempt to determine the permissible temporary occlu-

sion time in aneurysm surgery, a retrospective analysis of the re- Anterior Cerebral Artery (ACA)
sults of IOM with median nerve SEPs in 97 patients undergoing A study found median nerve SEPs to be poor indicators of is-
MCA and ICA surgery found the loss of the SEP in 42 patients chemia in anterior communicating artery (ACA) vascular terri-
during temporary occlusion, with all but three patients recover- tory. The most likely reason for lack of change is because the
258
ing the SEP back to baseline levels after recirculation. The 253
ACA vascular distribution does not supply the median nerve so-
three patients who did not recover their SEP experienced a rapid matosensory cortex, and the cortical generators of the P40-N45
loss of the SEP (between 1 and 5 minutes) and developed post- posterior tibial SEPs lie within the vascular territory of the
operative deficits. One patient without SEP changes developed ACA. Some investigators recommended SEP monitoring using
postoperative hemiplegia. The authors concluded that tempo- posterior tibial nerve stimulation combined with median nerve
rary vascular occlusion is relatively safe for a period of approx- stimulation, since posterior tibial nerve SEPs are inadequate in
imately 10 minutes after gradual loss of the SEP (Fig. 12-22). detecting ischemia in the recurrent artery of Heubner. More 343
In addition to providing information about the development of recently, in a small series of 15 patients, posterior tibial nerve
cerebral ischemia during temporary clipping, SEPs are useful in SEPs were used during ACA surgery requiring temporary arterial
C h a p t e r 12 INTRAOPERATIVE NEUROPHYSIOLOGIC MONITORING — 461
Figure 12-25. An example of sensory and motor lo-

calization using the "phase-reversal" technique in a
43-year-old female undergoing resection of a left parietal
glioma. An eight-contact cortical strip electrode was placed
on t h e left parietal c o r t e x . Averaged recordings w e r e ob-
tained after right median nerve stimulation. C o n t a c t
number I corresponds to trace I and contact number 8 t o
trace 8. In this case, the orientation of the electrode strip is
such that contact number I is anterior and number 8 is
posterior. The evoked potential phase reversal can be seen
between channels 2 and 4. T h e response from channel 2
corresponds t o t h e m o t o r c o r t e x , channel 4 corresponds
t o the somatosensory cortex, and channel 3 localizes to the
sylvian fissure.
occlusion to identify the onset of cerebral ischemia. Sig- 338

had BAEP changes associated with a new neurologic change. 225
nificant SEP changes were seen in 11 of 15 patients. Unilateral However, 1 out of 17 patients had a new neurologic deficit de-
A l (first branch of the ACA) occlusion led to SEP changes in spite no change in BAEP (false-negative). The investigators
four out of seven patients, while bilateral Al occlusion caused
changes in six of eight patients (Fig. 12-23).
Posterior Circulation
Several investigators have found IOM of vertebral-basilar
aneurysms with SEPs and BAEPs to be unreliable. - - A 105 219 258
possible explanation is that ischemia, owing to basilar perforator

occlusion, may not affect the brainstem auditory or somatosen-
sory p a t h w a y s . However, one team of investigators found
93105
that dual-modality monitoring (SEPs and BAEPs) was useful

during posterior fossa aneurysm surgery. They examined a
239
total of 70 aneurysms of the vertebral basilar circulation and

found that 10 patients had a change in their BAEP with six de-
veloping a postoperative neurologic deficit. Fourteen patients
had SEP changes, and eight of these had new postoperative
changes. All patients who had permanent EP changes developed
postoperative neurologic deficits. They also reported that the in-
cidence of false-negative and false-positive results for both
modalities was 20% and 30%, respectively. These findings sug-
gest that identification of brain stem ischemia with IOM is en- Figure 12-26. 25-year-old male with pontine glioma resec-
hanced by using a dual-modality approach. A recent study using tion. Intraoperative monitoring of this case involved multimodality
multimodality (SEP and BAEP) monitoring for both anterior and recordings of the following:
posterior circulation aneurysms showed that posterior circulation Trace I: BAEPs after left ear stimulation
ischemic changes as well as anterior circulation ischemia were Trace 2: Right cortical SEP (C '-Fz) following left median nerve stim-
4
accurately identified. Based on EP changes, postoperative

226
ulation
neurologic deficits were equally predicted in both groups. Trace 3: BAEPs after right ear stimulation
Trace 4: Left cortical SEP (C '-Fz) after right median nerve stimula-
3
tion
ARTERIOVENOUS MALFORMATIONS Continuous EMG activity was recorded using intramuscular hookwire
electrodes from:
The first to report successful use of posterior tibial SEP mon-
Trace 5: Left masseter
itoring in ACA territory arteriovenous malformation (AVM)
Trace 6: Left orbicularis oris
surgery was reported in 1982. However, there are no large
125
Trace 7: Right masseter

series reporting the outcome of AVM surgery and IOM, as most Trace 8: Right orbicularis oris
of the reports are combined with aneurysm surgical series. In a N o t e the presence of neurotonic discharges from bilateral orbicularis
series of 54 patients undergoing AVM surgery with the aid of oris muscles, which o c c u r r e d during t u m o r resection. Baseline right
IOM (56 procedures), investigators reported that 5 out of 54 pa- cortical SEP was unregisterable, a result of a previously attempted r e -
tients had SEP changes (four transient, one permanent) coincid- section. Postoperatively, the patient experienced worsening of his right
ing with clinical neurologic changes, and 1 out of 17 patients hemiparesis and bilateral facial weakness.
Table 12-13. Muscles Accessible for Cranial Nerve concluded that SEP and BAEP changes were useful in predict-
EMG Monitoring ing postoperative neurologic function (Fig. 12-24).
Cranial Nerve Muscles
III Inferior rectus BRAIN TUMOR SURGERY
IV Superior oblique The techniques used to monitor brain tumor surgery depend
V Masseter on the location and extent of the tumor. If the tumor is located in
the cerebrum, IOM usually involves identifying the functional
VI Lateral rectus sensory and motor cortex. This can be achieved by directly
VII Frontalis, orbicularis oculi, stimulating the cortex and observing or recording motor activity
and orbicularis oris in the face, arms, or legs of an anesthetized patient. Cortical
Posterior pharyngeal muscles
stimulation in an awake patient can also be used to identify the
IX
cortical areas corresponding to language. Another method is to
X Cricothyroid o r vocalis electrically stimulate a peripheral nerve (usually the median
XI Sternocleidomastoid or trapezius and/or posterior tibial) in a fashion similar to that used for SEPs
and record a "phase-reversal" directly from the cortex (Fig. 12¬
XII Tongue
25). SEPs recorded from the postcentral gyrus generate a biphasic
3 m s T r i g
lOO u V Rmp 1
10CJ u V Rmp 3
3 ms T r i g
TOO u V R m p «* 1QO u V Rmp 4
3 m s T r i g 3 m s T r i g
lOO uV Rmp 5 1QO u V Rmp 5
D i s k S p a c « : CJ| 5Q3MB 10 JftN D i s k S p a c e : 0 | 5Q3MB 10 JftN
Figure 12-27. A 55-year-old female undergoing resection of a pontine AVM was monitored for cranial nuclei activity.The facial nerve
nucleus was localized using monopolar electrical stimulation prior t o incising the pons. Evoked CMAPs were recorded using intramuscular hook-
wire electrodes placed in the following muscles:
Trace I: Left orbicularis oculi
Trace 2: Left orbicularis oris
Trace 3: Left masseter
Trace 4: Right orbicularis oculi
Trace 5: Right orbicularis oris
A, Brain stem stimulation (1.0 mA/0.05 msec) showing a small response from the left orbicularis oris. B, Slight caudal movement of the stimulation
electrode reveals larger CMAP indicating t h e location of the nucleus. Stimulation parameters remained unchanged from the previous trace.The
AVM was not resected because it did n o t c o m e t o the surface and it was located directly beneath t h e nucleus.
negative-positive electrical potential, while the EP, recorded defining them as "distinctive discharges of a motor unit poten-
from the precentral gyrus appear as a mirror image biphasic po- tial in rapid, irregular bursts" (Fig. 1 2 - 2 7 ) . In contrast,
68292
tential (positive-negative). Using the phase-reversal tech-

325
CMAPs monitor integrity and continuity of the nerve from the
nique, a cortical strip or grid is placed on the cortex, in an area point of stimulation (proximal to the site of surgery) to the
thought to correspond to the somatosensory and motor cortex. muscle. The onset latency of the CMAP indicates the conduc-
Peripheral nerve stimulation will produce a sensory EP easily tion time of fastest fibers, while its amplitude is approximately
recorded from the exposed cortex. If the cortical grid is lying proportional to the number of available a x o n s . 113a185
across the sylvian fissure, a corresponding electrical potential of CN VIII. The need to preserve hearing in surgeries of the
opposite polarity can be recorded from the motor cortex. This cerebellopontine angle and in acoustic neuroma resections has
technique is relatively fast, reliable, and simple to perform. led to a variety of auditory evoked potential techniques. - 215 351
Tumors in the posterior fossa and in areas adjacent to cranial The most commonly used recording technique is the BAEP;
nerves will usually require multimodality monitoring using a however, other less commonly used techniques such as electro-
combination of SEPs. BAEPs, and spontaneous and/or triggered cochleography (ECoG) and direct recording of nerve action
EMG (Fig. 12-26).* potentials (NAP) are useful in assessing eighth nerve func-
tion. The advantage of the BAEP is that it reflects the electri-
157
POSTERIOR FOSSA AND CRANIAL NERVE SURGERY cal activity in the auditory nerve and auditory pathways of the
brain stem in response to cochlear stimulation. Five distinct
Posterior fossa surgery presents difficult and unique chal- vertex positive waveforms can be recorded, with waves I, III,
lenges to the IOM team because various types of surgeries are and V being the most significant for IOM purposes. Waves I and
performed, each requiring a different monitoring protocol. II originate from the auditory nerve at the level of the cochlea,
Thus, the IOM protocol is designed according to the neural wave III within the lower pons, and wave V in the midbrain near
structures at risk. The most commonly used modalities are the inferior colliculus. Direct intracranial recordings NAPs
53
BAEPs, spontaneous and triggered EMG, and S E P s . It is 53

can be obtained using a small cotton-wick electrode applied to
common to use these modalities simultaneously in order to the eighth nerve close to its entrance to the brain stem. T h e
monitor a larger group of neural structures. ECoG is the least commonly used modality and can be recorded
The most common use of IOM in posterior fossa surgery is using a needle electrode placed through the tympanic mem-
for acoustic schwannoma resections, but monitoring is also uti- brane and resting on the promontory of the middle ear. ECoG is
lized in surgeries for other types of tumors, hemifacial spasm, used to document integrity of the cochlea.
I trigeminal neuralgia, and vascular structures. For monitoring CN VIII function, the BAEP seems to be the
Monitoring Cranial Nerve (CN) I. Monitoring of the most comprehensive modality because it assesses peripheral
visual pathways has been attempted using VEPs. However, and central auditory pathways, has few technical limitations,
VEPs are very sensitive to technique, anesthesia, temperature, can be used during the entire procedure, and has good correla-
blood pressure, and other surgical factors and are affected in an tion with postoperative hearing status (Table 1 2 - 1 4 ) . In a 1392,5
unpredictable manner. - One investigator stated that flash

99 382
study of 90 BAEP-monitored patients compared with 90 un-
VEPs are not useful for IOM during removal of visual pathway monitored historically matched controls undergoing acoustic
lesions, owing to a combination of uneventful monitoring, false- neuroma surgery, it was found that BAEP monitoring reduced
positive, and false-negative cases. 382
the risk of hearing loss in patients with acoustic neuromas less
CNs HI-VII and IX-XII. IOM of the motor component of than 2 cm and was statistically significant for tumors less than
any cranial (or peripheral) nerve is performed in essentially the 1.1 cm in diameter. In addition, BAEP monitoring improved
139
same fashion. The basic principle involves EMG monitoring of the chance of preserving useful hearing.
spontaneous and triggered muscle activity. Obviously, the cra- Owing to the close proximity to CN VII to acoustic neuro-
nial nerve(s) at risk will determine which muscles from which mas, techniques to also monitor this cranial nerve during
to record (Table 12-13). Recordings can be obtained by placing acoustic neuroma resections have been described. - 62 63135
two intramuscular wire electrodes within the muscle or using

surface electrodes. Simultaneous spontaneous EMG and INTERVENTIONAL NEURORADIOLOGY PROCEDURES
CMAP recordings can be obtained by using intramuscular wire
and surface electrodes. Using intramuscular electrodes in- There are few reports of IOM monitoring in neuroradiologic
creases the sensitivity for detecting spontaneous EMG activity, procedures. - - *
57 79128
However, EEG and EPs have been
302 307
while surface electrodes allow for more reliable monitoring of

C M A P amplitude and m o r p h o l o g y . Additionally, a sen-
681133
Table 12-14. Hearing Outcome As a Function of Presence

sory nerve conduction study has been applied for intraoperative of Evoked Potentials at End of 164 Cerebellopontine Angle
CN V monitoring. 156
Tumor Operations
Also, brain stem cranial nerve nuclei can be identified and
N l Present N1 N o t Present
mapped by electrically stimulating discrete areas of the brain
stem. For example, direct stimulation of areas within the floor WaveV present Hearing (31) ** (none)
of the fourth ventricle can activate the trigeminal and facial WaveV not present Indeterminant (66): N o hearing (67)
nuclei. Recording is performed in a manner similar to that de- "Useful" hearing (38)
scribed above (Fig. 12-27). N o t "useful" hearing (12)
The rational for using spontaneous EMG activity monitoring N o hearing (16)
is based on the property that thermal, mechanical, or metabolic
From Levine RA: Monitoring auditory evoked potentials during cerebellopon-
irritation of the intracranial portion of the motor CNs will lead tine angle tumor surgery. In Schramm J, Moller AR (eds): Intraoperative Neuro-
to characteristic activity in the innervated m u s c l e s . -
54J07 328a
physiologic Monitoring in Neurosurgery. New York, Springer, 1991, pp 193-204,
Some investigators termed this activity "neurotonic discharges," with permission.
13 SO 1
Star
1 U A A Xa'«i
HBR Latencies & ftnplitudes flBR Latencies & ftnplitudes
L.mt
L.Jit
•***
j"„ i" 7 00
-"t-I
il- e"„
T
a'ei i " 30
ae
' o Vra MP L«1
V'BI 17 V'BS
\s j »\jSD (B^ ^3 t -« «-«
—iTiT i — n V I»F» i -«
*a"ai *a" 5a V 3- ' a "*«

LEFT KEOIflN SEP RIGHT RIGHT
^ ^ ^ ^
LEFT HEDIRN SEP
U»t
SO 7 a? * *ao o*a"' Q ao° e30 *
a
i "?*, a aT. i
l.ll
*M
Vr*.37
j...
30.0
* a os ""io Vt Q'aoi
\o a 'io' - l
"°
Figure 12-28. An 80-year-old female with a giant basilar aneurysm and recent subarachnoid hemorrhage treated with en-
dovascular occlusion.The patient was believed to be a poor surgical candidate, and endovascular placement of coils was not possible owing t o
the morphology of t h e aneurysm. Basilar o r vertebral occlusion, as a means of reducing arterial flow and t o promote thrombosis of the aneurysm,
was determined to be the best option.
Trace I: BAEP (A,-Cz) after left ear stimulation
Traces 2 and 3: Right cortical SEPs (C '-Fz and C ^ - ^ ' ) after left median nerve stimulation
4
Trace 4: Left Erb's (left Erb's-Fz) following left median nerve stimulation
Trace 5: BAEP (A -Fz) after right ear stimulation
2
Traces 6 and 7: Left cortical SEPs (C '-Fz and C3'-C ') after right median nerve stimulation
3 4
Trace 8: Right Erb's (right Erb's-Fz) following right median nerve stimulation
A, Baseline BAEPs and SEPs. B,Three minutes after balloon occlusion of the basilar artery. N o t e loss of left cortical SEP and delay and reduced
amplitude of right cortical SEPs. N o significant changes were seen in the BAEPs. C, Return of SEPs t o baseline levels 5 minutes after the balloon
was deflated. Occlusion of the right vertebral artery produced no significant changes in the evoked potentials, and, therefore, it was permanently
occluded.The patient experienced no new neurologic deficits.
successfully used in the monitoring of selected interventional Some investigators reported the use of SEPs in 2 3 embolization
procedures. The most common neuroradiologic procedures sessions performed in 17 patients. In their study, two patients
307
monitored are carotid artery angioplasty, deliberate arterial oc- were unable to undergo embolization because of positive results
clusion of a variety of extracranial and intracranial vessels, coil- of the Amytal testing. In one case, SEPs demonstrated an ampli-
ing of intracerebral aneurysms, and in the glue embolization of tude reduction of greater than 50% and the results of the physi-
cerebral AVMs. The modalities chosen to monitor such cases cal examination showed a new neurologic deficit after the
depend on the cerebrovascular territory at risk and whether the administration of Amytal; in the other case, only the SEPs were
patient will be awake or under general anesthesia during the used because the embolization was performed under general
procedure. The recordings are performed in a similar fashion to anesthesia. The investigators believed that the combined use of
those described in previous sections. One important exception SEPs and neurologic examination was valuable in the treatment
is in the awake patient, where neurologic examination and su- of rolandic AVMs. In a series of 17 patients undergoing balloon
perselective Amytal testing can augment IOM. Amytal testing occlusion test of the ICA due to tumor infiltrate, a combination
consists of injecting a small dose of Amytal directly in the pedi- of clinical, electrophysiologic (median nerve SEPs and trans-
cle that is to be embolized or the vessel that is to be occluded. cortical motor evoked potentials), and Doppler sonographic
1~
B
Figure 12-29. 1 4 - y e a r - o l d f e m a l e w i t h scoliosis and rapid curve progression following spinal ganglioblastoma resection and radiation t h e r -
apy I year prior. IOM of anterior and posterior spinal fusion (T3-L3) with instrumentation was performed.
A, Baseline posterior tibial SEPs.
Traces 1-3: Right cortical SEPs (Cz-Fz, Cz'-Fz,and 0 3 - 0 4 ' ) after left side stimulation
Trace 4: C - 7 (C7-Fz) after left side stimulation
Traces 5-7: Left cortical SEPs (Cz-Fz, Cz'-Fz, and C '-C3') after right side stimulation
4
Trace 8: C - 7 (C7-Fz) after right side stimulation

B, Loss of cortical (traces 2 - 3 and 6-7) and cervical spine-generated SEPs (traces 4 and 8) during the placement of the second rod but prior t o
distraction. Peripheral (popliteal fossa) recorded EPs are present (traces I and 5), indicating intact peripheral conduction. Scalp-recorded median
nerve SEPs (not shown) were unchanged from baseline.These findings are consistent with acute cord injury.Ali spinal hardware was removed as a
result of the SEP changes and wake-up test was equivocal. Unfortunately, the patient developed postoperative severe paraparesis and lower ex-
tremity sensory loss.After several weeks, she began to regain sensation and leg strength.
monitoring was used to detect severe cerebral complications. 79

been used. The first technique employed was the intraoperative
These investigators concluded that neurophysiologic monitor- wake-up test. - The so-called wake-up test is simple; anes-
21 412
ing played an important role in predicting cerebral complica- thesia is discontinued at a critical stage during surgery (usually
tions after permanent occlusion of the ICA (Fig. 12-28). after distraction/derotation), and patients are asked to wiggle
In addition to EPs, EEG has also been used in cerebral and their toes. Unfortunately, the wake-up test has several limita-
carotid balloon test occlusion. In one study, continuous poly- tions. It requires a cooperative patient who can comprehend,
graph and quantitative EEG were used along with repeated hear, and follow simple commands; requires presurgical re-
detailed clinical examinations in 17 cases. Four of 17 con-
57
hearsal; interrupts the surgical procedure; can take up to 15 min-
secutive patients showed changes in either their EEG or their utes or more to complete, which means it may occur up to 20
clinical examinations during carotid occlusion. The authors minutes after the spinal distraction; and may fail to detect spinal
concluded that continuous EEG monitoring and repeated clin- cord injury. - It is not without hazards, which may include ac-
78 282
ical examinations provide useful ways of evaluating cerebral cidental extubation, vascular line interruptions, air embolisms,
circulation during carotid test occlusions. In a separate study, and dislodging the spinal instrumentation just placed, and some
the EEG was used to monitor 19 patients during cyanoacry- patients may not be able to cooperate adequately. - 94 180 282
late AVM embolization. Focal or diffuse EEG abnormali-

302
Furthermore, the wake-up test provides only a one-time neuro-
ties in the lower frequency range could be followed by logic assessment and does not allow continuous monitoring.
clinical hazards (3 out of 11 cases). Theoretically, it is possible that by the time the wake-up test is
performed, damage to the spinal cord has already occurred and
SPINAL SURGERY may be irreversible. This subsequently has led to the use of elec-
trophysiologic techniques that would allow continuous monitor-
The main use of IOM where the spinal cord is at risk involves ing of spinal cord function. Different strategies in stimulation
1222
scoliosis surgery. However, some authors believe IOM should and recordings have been used; typically peripheral nerve or
be used in any patient who is at risk for injury to the spinal cord spinal epidural stimulation is performed, and recordings are ob-
during any surgical procedure. 101275
tained from epidural, interspinous ligaments, spinous process,
or scalp sites. - The most commonly used methods are the
229 377
Scoliosis Surgery spinal evoked potential (recorded from the skin, interspinous
The risk of postoperative paraplegia and paraparesis as a ligament, vertebrae, or epidural space) and the scalp-recorded
result of instrumentation procedures for scoliosis correction has SEP. Generally, tibial or peroneal nerve SEPs are used to moni-
been reported to be 0.72%. Spinal cord injury may result from
230 tor below the C8 level, and median or ulnar nerve SEPs are per-
several factors, including compression or traction of the cord, formed to monitor above the C8 level. The criteria for SEP
159
intraoperative vascular insufficiency, and trauma." With this in compromise is a 50% amplitude reduction or >5 ms latency
mind, several methods of monitoring spinal cord function have shift sustained for greater than 10 minutes. Although SEPs do
376
not monitor the central motor pathways, animal studies suggest PERIPHERAL NERVE SURGERIES
that SEPs are sensitive to acute spinal cord damage from either
ischemic or mechanical insults to the spinal cord (Fig. 12-29). 324
Peripheral nerve surgery can be monitored with intraoperative
Several studies have demonstrated the validity of SEP moni- neurophysiologic techniques, especially NCS, EMG, and SEP
toring in scoliosis surgery. - - A large multicenter survey as-
94 282 288
techniques. - - - - ' This monitoring has been successfully
I91 192 291 314 357 358
sessed surgical outcome from centers that use SEP monitoring implemented with the benefits of helping to guide dissection,
in scoliosis surgery, obtaining data on 51,263 procedures, and identify neural structures, and evaluate for traction or vascular
found that experienced SEP monitoring teams had fewer neuro- compromise. - - - - - - - - - - Typical procedures
175 188 190 202 291 293 317 357 384 405 429
logic deficits and that false-negatives occurred in only 0.063% that call for such monitoring are neurolysis, nerve grafting after
of patients. The investigators recognized that intraoperative
288
traumatic nerve injuries, excision of tumors that encroach upon or
intervention (secondary to SEP changes) may have prevented or encase nerves, resection of cysts, and nerve releases or transposi-
reduced the severity of any postoperative deficit, thereby influ- tions. Also surgical reexploration of scarred areas where anatomy
encing the outcome and artificially raising the number of false- can be difficult to identify and the scar itself can be contributing
positive cases and undercounting true-positive cases. With this to nerve compromise and any procedure that may lead to acute
assumption, sensitivity and specificity of SEP monitoring were nerve traction or stretch such as total joint arthroplasties or joint
calculated as 92% and 98.9%, respectively. They also found a contracture releases warrant monitoring.
negative predictive value of 99.33% but only a mediocre posi- Spontaneous EMG monitoring of more distally innervated
tive predictive value of 42%. They concluded that SEP monitor- muscles can help to detect nerve irritation or stimulation, alert-
ing detects more than 90% of the intraoperative neurologic ing a physician as to the proximity to those nerves. This is very
deficits and that SEP monitoring for scoliosis surgery is a clini- similar to the techniques described above for posterior fossa
cally useful, valid procedure. tumor explorations. Additionally, triggered EMG can be used
wherein the surgeon stimulates suspected tissue and evaluates
Other Spinal Surgeries which monitored muscles become activated by that stimulation.
The use of SEPs as a measure of spinal cord function is not This is a variant of nerve conduction studies in which the mus-
limited solely to scoliosis surgery. It has also been found useful cles may be studied by surface electrodes, though more often by
in other spinal surgical procedures such as spinal AVMs and intramuscular electrodes to optimize isolation of each specific
tumors, unstable fractures, resection of syringomyelia, spinal muscle. With intramuscular electrodes, free-running sponta-
decompression, embolization of spinal AVMs, and in children neous EMG activity, triggered EMG data, or both can be de-
undergoing selective dorsal rhizotomy. - - 99
Nash and
311 377
rived and monitored simultaneously during a procedure.
B r o w n stated that "it is standard practice to conduct some
275
Examples of some complex setups that can be used include
form of monitoring when performing any spinal operation that (1) brachial plexus: serratus anterior, rhomboid, supra-
is associated with a high risk of neurological injury. Generally, spinatus/infraspinatus, deltoid, biceps, brachii, triceps, wrist ex-
operations are considered to carry such a risk when corrective tensors, wrist flexors, thenar eminence, hypothenar eminence,
forces are being applied to the spine, the patient has pre-existing first dorsal interosseous; or from the lower limb: (2) sciatic
neurological damage, the cord is being invaded, or an os- nerve: biceps femoris, peroneal component—tibialis anterior,
teotomy or other procedure is being carried out in immediate peroneus longus, extensor digitorum brevis—and tibial compo-
juxtaposition to the spinal cord." In spite of the impressive track nent—gastrocnemius. 358
record of SEPs in monitoring spinal cord function, false-nega- Nerve conduction studies can be used with intraoperative
tive cases have been reported and are of concern. - The criti- 99 288
field stimulation of the nerves being explored or from stimula-
cism of mixed nerve SEP spinal cord monitoring is that it tion of the nerve outside the operative field. This requires pre-
measures conduction in the ascending sensory pathways located operative setup and baseline testing. If muscle relaxants can be
in the posterior columns while the anterior, motor pathways avoided, by adequate alternative anesthesia regimens, intraoper-
remain unmonitored. Thus, several techniques have evolved in ative CMAP monitoring can be used. CMAPs have significant
an attempt to monitor the spinal cord motor pathways. intraoperative advantages, owing to the CMAP amplitude being
Motor evoked potentials (MEPs) can be generated by apply- at least two orders of magnitude (100 times) greater than nerve
ing an electrical or magnetic stimulation transcranial^ to the action potentials' (NAPs) amplitudes. This feature makes
motor cortex or to the spinal cord and recording distal to the CMAP recordings easier to obtain and more resistant to the dis-
spinal level of interest from the spinal cord, peripheral nerve, or torting effects of the high levels of electrical noise common to
muscle. - Unfortunately, the clinical utility of MEPs is
42160181 217
the operating room environment than are the much lower ampli-
limited owing to the effect of anesthetic agents on the motor tude NAPs. However, if muscle relaxants are used, averaged
pathways, causing severe reduction of MEP amplitude. NAP amplitudes may be more reliable than the more easily
However, different anesthetic techniques are currently being in- recorded CMAPs, which change amplitude in response to the
vestigated that may facilitate recording of MEPs in the operat- muscle blocking agent, and thereby, compound interpretation of
ing room. Transcranial electrical and magnetic stimulation is any changes in amplitude. Peripheral nerves, however, are sen-
still considered experimental in the United States, and the de- sitive to core body temperature changes, which can significantly
vices are not approved by the Food and Drug Administration. alter the peripheral nerve's conduction latencies and amplitudes.
Certain centers use direct spinal cord stimulation with record- Because of this, any changes detected in latency or amplitude
ings obtained from the lower extremities using either surface or measured along peripheral nerve paths require obtaining the
intramuscular EMG recordings to assess the motor pathways. concurrent core body temperature to assist in interpreting
However, this technique can be technically challenging to the whether such changes are physiologic and expected, or herald a
neurophysiologist and anesthesiologist because the patient potential nerve compromise. Surface CMAPs or NAPs correlate
cannot be completely paralyzed and the level of neuromuscular well to the number of functioning axons, making these useful
blockade must be kept constant throughout the procedure. for quantifying the degree of compromise. When the peripheral
I
nerves to be monitored can be accessed outside the operative field,

the bilateral comparisons can be continuously performed by bilat-
erally symmetric setups. This can be very helpful in determining
whether the patient's physiology as modified by anesthesia is
leading to the observed changes, in which case, one expects the
same changes on both sides. If an asymmetrical change occurs
that is consistent with a peripheral nerve compromise on the op-
erative side, a warning can be more confidently and quickly
conveyed to the surgeon. In general, if the CMAP or NAP am-
plitude remains greater than 2 5 % of baseline testing, no clini-
cally significant postoperative peripheral nerve deficits are
expected. Another sign of intraoperative peripheral nerve com-
promise is a significantly increasing CMAP or NAP latency,
usually by 1-5 ms, where increase is contemporarily correlated
with an operative maneuver.
SEP recordings can be obtained even with small peripheral
nerve inputs, such as DSEPs. Therefore, the presence of any
physiologic continuity of the peripheral nerve segment be-
tween its distal stimulation site to the centrally recorded SEP
site can often be best established by SEP testing. Recordings
can be the typical scalp recordings or epidural recording of po-
tentials from the spinal cord. SEPs are less helpful to quantify
the degree of peripheral nerve compromise, owing to central
amplification effects, but are very sensitive to any information
transmitted through the spinal cord. SEPs can be very helpful
in determining the intactness of the preganglionic sensory
pathways, which are important for intraoperative decisions
while exploring the brachial plexus and deciding which trau- Figure 12-30. Stimulating (S) and recording somatosensory
matic trunks or roots to cable-graft. If there is a preganglionic (RI) and nerve action potentials ( R 2 ) during brachial plexus
lesion that precludes sensory input from reaching the spinal repair.The reference electrode for the SEP is n o t shown. (From Landi
cord, cable grafting will not result in sensory recovery. Total A, Copeland SA,Wynn Parry CB, Jones SJ:The role of somatosensory
loss of peripheral nerve function between the site of stimula- evoked potentials and nerve conduction studies in the surgical man-
tion and the central nervous system will also result in loss of agement of brachial plexus injuries. J Bone Joint Surg I980;62B:
the SEP response, and this can be used for evaluating exces- 492-496, with permission.)
sive peripheral nerve traction. A peripheral nerve will develop
a reversible conduction block after a mild degree of stretch. 350
This occurs before axonotmesis, or permanent injury, occurs The brachial plexus has also been monitored prophylactically
at typically 6-30% elongation, depending on time course, with during sternotomies as a preventative measure as a result of an inci-
shorter periods tolerating lesser stretch before axon ruptures dence of post-sternotomy brachial plexus injuries of approximately
occur. If this window of initial nerve conduction failure can
385
be identified rapidly, procedures that lead to this stretch can be C 5 Root Recording
reversed or eliminated to prevent permanent or prolonged
nerve injuries.
The more common upper limb nerves that have been moni-
tored during surgical procedures include brachial plexus, spinal
accessory, supraspinatus, axillary, radial, posterior interosseous,
musculocutaneous, median, anterior interosseous, and ulnar
nerves. The more commonly monitored lower limb nerves in-
clude: sciatic, peroneal, tibial, femoral, and ilioinguinal.
Brachial Plexus Intraoperative Monitoring. Exploration C6 Root Recording
of traumatic brachial plexus injuries can be greatly enhanced by

intraoperative neurophysiologic techniques. - - - The find-
82a 202 293 384
ing of intact SEP responses from stimulating the proximal por- MEDIAN N E R V E STIMULATION
tion of residual root implies an intact central nervous system 5 ms&c/div
pathway, which helps to determine which nerve stumps are suit- Upper Arm Recording
able for cable grafting (Figs. 12-30 and i2-31). - - If a co-
82a 202 358
morbid complete preganglionic lesion is present, sensory Figure 12-31. Location of intact distal nerve segment in a
recovery could not be expected from peripheral cable grafting. severe preganglionic brachial plexus injury by recording in-
The presence of a comorbid preganglionic lesion is manifested trafield nerve compound action potentials (NCAPs).The lower right
by the absence of an SEP response to that nerve stump's stimu- panel demonstrates that the distal segment is intact, producing a clear
lation. The brachial plexus can also be involved in tumor resec- upper arm recording.The larger amplitude intrafield recordings indicate
tions, for which intraoperative neurophysiologic techniques can the position of the C6 nerve r o o t (From Slimp JC: Intraoperative mon-
help with dissection and preservation of function. 317
itoring of nerve repairs. Hand Clinics 2000; 16:25-36, with permission.)
the median, axillary, radial, and musculocutaneous nerves

during shoulder procedures including arthroscopy. The ulnar 314
nerve can be intraoperatively assessed by the so-called inching

techniques to further evaluate the site of ulnar nerve entrapment
about the elbow, as well as to assess the adequacy of decom-
pression (Figs. 12-32 and 12- 3 3 ) . Incidental peripheral
4 8 1 3 7
ulnar nerve compromise has been detected intraoperatively by

SfiP techniques incorporated for other purposes. Surgical ex- 19
ploration of the mid-forearm median nerve has been described,

using SEP, and NCS stimulating within the operative field while
recording both outside and within the field. 358
Hip Surgery Sciatic Nerve Monitoring. Total hip arthroplasty

can result in compromise of the sciatic nerve, especially the per-
oneal portion, reported in 0.6-3.5% of cases, with almost twice that
(msec) incidence after arthroplasty revision surgeries." * Table 12-15 3 381
lists some of the many potential causes of neural compromise that

Figure 12-32. Example of CMAPs In a control subject. They have been reported specifically with hip arthroplasty; however, this
were recorded from the abductor digiti quinti in response t o sequen- list includes common causes to be considered in many surgical pro-
tial stimulation over I -cm segments of t h e ulnar nerve.The medial epi- cedures. In an attempt to decrease this, both NCS - and SEP
113 175 182
condyle (ME) is considered the 0 point, and stimulation progresses techniques have been used to warn of impending compromise
2793 8 1
from proximal (-) t o distal (+). N o t e t h e rather uniform latency so that corrective action can be taken. Multiple muscle spontaneous
changes (0.05-0.25 msec over each segment shown to the left of each EMG monitoring has also been advocated. The criteria for SEP 358
trace) and the constancy of the evoked potential's waveform. (From compromise is a 50% amplitude reduction or greater than 5 ms la-
Campbell W W , Sahni SK, Pridgeon RM, Leshner RT: Intraoperative tency shift sustained for greater than 10 minutes. For significant 376
electroneurography management of ulnar neuropathy at the elbow. NCS criteria to warrant operative changes, an evoked amplitude
Muscle Nerve 1988:1 1:75-81,with permission.) decrement of 7 5 % was found to better predict those at increased
risk of compromise. No hip arthroplasty patients having decre-
ments less than 75% of their baseline evoked potentials developed
5%, most often to the C8/T1 roots, lower trunk, or medial cord, postoperative causalgic symptoms, yet 50% of those with greater
presumably from stretch or retractor pressure. ^5,153.210.316.349 than 7 5 % decrements complained of such symptoms postopera-
Some advocate the routine monitoring of the brachial plexus in tively (Figs. 12-34 and 12-35). 182
other non-brachial plexus related surgeries to avoid deficits Lower Limb Peripheral Nerve Monitoring. During in-
from prolonged intraoperative positioning of the arm. - - 10 236 346
guinal hernia repair the ilioinguinal nerve can be placed at risk,
Upper Limb Peripheral Nerve Monitoring. Arthroplasty and monitoring techniques can help with its identification and.
and even arthroscopy has also been associated with peripheral preservation. Monitoring of the peroneal nerve during knee
358
nerve injuries. - This has lead to the successful monitoring of

294 373
surgery has also been described. 429
Table 12-15. Total Hip Arthroplasty Causes

of Nerve Compromise
Retraction
Compression
Hemorrhage
Hematoma
Excessive leg lengthening
Release of contractures
Penetrating injuries
Ischemia
Patient positioning
A Lotency Surgical dissection
(msec)
Fixation of components
Figure 12-33. Marked increase in latency (1.4 ms) over the
Cauterization
segment just proximal to the medial epicondyle. In addition,
the shape of the CMAPs recorded from proximal stimulation is slightly Prosthetic placement
more dispersed than that of response to distal stimulation. Patient un- Reduction maneuvers
derwent epineurolysis and microscopically guided limited internal neu-
Removal of old components (revision)
rolysis and then anterior transposition of t h e nerve. Epineural biopsy
revealed neuroma in continuity. (From Campbell W W , Sahni SK, Preparation of bone-to-place component
Pridgeon RM, Leshner RT: Intraoperative electroneurography manage- Postoperative dislocation
ment of ulnar neuropathy at the elbow. Muscle Nerve 1988:11:75—81, From Goldberg G, Goldstein H: AAEM case report 32: Nerve injury associated
with permission.) with hip arthroplasty. Muscle Nerve 1998:21:519-527, with permission.
Stimulus type = clicks

Polarity = alternating or rarefaction
Rate is between 10.1 and 11.7 Hz.
1000 stimulations are averaged
Recording
SEPs
Standard disk or needle EEG electrodes can be used. If disk
electrodes are utilized, it is recommended that they be applied
Figure 12-34. Location of the four electrodes from sciatic with collodion. Electrode impedance should be maintained
nerve monitoring during total hip arthroplasty. (From Kennedy below 5 ohms for surface electrodes but not needle electrodes.
WF, Byrne TF, Majid HA, Pavlak LL Sciatic nerve monitoring during revision Electrodes are placed bilaterally on the parietal scalp at C3' and
total hip arthroplasty. Clin O r t h o p 1991 ;264:223-227, with permission.) C4' (2 cm behind the C3/C4 positions of the 10-20 interna-
tional electrode placement system (Table 12-16 and Fig. 12-
TECHNICAL CONSIDERATIONS 35)348 j
a n c o parasagittal scalp at CZ' (2 cm behind C Z )
n t n e
overlying the primary somatosensory hand and leg area, respec-

SEP AND BAEP RECORDING TECHNIQUES tively. A reference electrode is placed at the midfrontal (FZ) lo-
cation. Median nerve cortical SEP recordings should be obtained
Baseline recordings should be obtained prior to any surgical from both cerebral hemispheres using C 3 - F Z and C 4 - F Z .
intervention. Other available channels can be used to monitor subcortical and
Stimulation
SEPs are recorded after bilateral independent median nerve Table 12-16. A Method of Placing Electroencephalograph
stimulation using standard bipolar bar surface or subdermal Electrodes by the International Ten-Twenty System
needie electrodes at the wrist. Stimulation is at a rate of 3-5 Hz 1. Measure the distance from the nasion t o t h e inion and make a
with a 0.1-0.3 msec pulse duration and a constant current inten- mark at 50% of this distance.
sity, usually less than 25 mA, but sufficient to produce a visible
2. Measure the distance from the tragus t o tragus and make a mark
i muscle twitch. The ground electrode is placed on the arm proxi-
at 50% of this distance.
| mal to the stimulating electrode. A minimum of 250 stimula-
The intersection of these t w o marks is Cz.
tions are averaged. Stimulating at faster rates than 3 Hz (lower
extremity) to 5 Hz (upper extremity), though desirable for ob- 3. With the tape passing through Cz, measure and mark 10% of t h e
taining more rapid warning of deterioration, is not advisable distance from nasion t o inion on t h e forehead for Fp (also called
owing to a decline in SEP amplitudes (an attenuation of at least Fpz), and 10% of the distance from inion t o nasion at the occiput
80% when rates of 15 Hz or higher are used). - 96 283
for Oz.
BAEPs are obtained after bilateral independent ear stimula- 4. With the tape passing through Cz, measure and mark 10% of t h e
tion, using ear inserts. Stimulation parameters are the following: distance from tragus t o tragus.These points a r e T 3 on the left
and T4 on t h e right.
5. Measure the circumference of the head passing through the Fp,
T4,Oz,andT3.
6. Mark 10% of this distance on either side of Fp.These points a r e
Fpl on the left and Fp2 on the right. Also mark 10% of this dis-
tance on either side of O Z T h e s e points are O I on the left and
0 2 on the right.
7. Mark half the distance from T3 t o Fp I for F7. Repeat on the right
for the location of F8.
8. Mark half the distance fromT3 t o O I f o r T 5 . Repeat on the right
for the location of T6.
9. Mark half the distance from Cz t o T 3 for C 3 . Repeat on the right
for the location of C4.
10. Mark half the distance from Cz t o Fp for Fz. Mark half the dis-
tance from Cz t o O z for Pz.
11. Mark half t h e distance from C3 t o Fp I for F3. Mark half the dis-
tance from C3 t o O I for P3. Repeat for t h e right side of the head
for F4 and P4.
12. Primes indicate 2 cm posterior (toward t h e occiput) t o the above
points, e.g., C 3 ' and C 4 ' are 2 cm horizontally more occipital than
C3 and C4 derived as above.
Figure 12-35. Recording hook electrode around exposed sci-
From Schwentker MC, Forney DJ, Gieski R, Winters JliTechnical standards and
atic nerve. (From Kennedy WF, Byrne TF, Majid HA, Pavlak LL: Sciatic techniques for basic electroencephalography. In Russell GB, Rodichok LD (eds):
nerve monitoring during revision total hip arthroplasty. Clin O r t h o p Primer of Intraoperative Monitoring. Boston, Butterworth-Heinemann, 1995, p
1991 ;264:223-227, with permission.) 56, with permission.
of EEG activity over both cerebral hemispheres. However, a

minimum of 8 channels covering the parasagittal regions may be
used if a 16-channel EEG is not available. The high-frequency
filter is set at least at 30 Hz and preferably at 70 Hz; the low-fre-
quency filter is set at 1.0 Hz. In some cases, it may be necessary
to use a 60-Hz filter, but it should be avoided if possible.
OPERATING ROOM RECORDING

If possible, initial intraoperative recordings are obtained prior
to anesthesia induction and intubation. At a minimum, record-
ings are acquired after induction but prior to first incision.
Recordings are continued at various intervals during the surgi-
cal procedure, while simultaneously documenting the anesthetic
level, minimum anesthetic concentration (MAC) times a factor,
and temperature.
During the period when the nervous system is at risk, EPs and
other recording modalities should be continuously monitored
and their morphology, latency, and amplitude serially com-
pared. Non-continuous monitoring is continued until the patient
is awake.
Figure 12-36. 10-20 S y s t e m e l e c t r o d e p o s i t i o n s . (From Nuwer
Critical Changes
MR (ed): Evoked Potential Monitoring in t h e Operating Room. N e w
York, Raven Press, 1986. Appendix, with permission.) 1. 50% or more reduction in amplitude of N 1 9 - P 2 4 /
P40-N45
2. Increase in CCT of 1 msec or more
brachial plexus (Erb's point) evoked potentials. The low and 3. Loss of waves III-V of the BAEP
high filters are set at 30 Hz and 3 kHz, respectively.
Recommended Montages: Median nerve SEPs Moderate Changes
C3-FZ 1. Cortical SEP latency increase of 5%
C4-FZ 2. Latency delay of 1 msec or more in the wave IV/V com-
• These channels provide recordings of near-field cortical plex of the BAEP
SEP components (N19, P24, N30).
Cl (7th cervical spine)-FZ or contralateral shoulder-FZ. NERVE CONDUCTION TECHNIQUES
• This set-up would allow_for identification of subcortical far-
field potentials (PI4, and permit monitoring of CCT Stimulation
(time interval between P I 4 and N19). CCT reflects the in- Most often, a 0.05 to 0.3 ms constant-current stimulation
tracranial conduction time between foramen magnum and pulse is used. This is usually adjusted for supramaximal stimu-
somatosensory cortex. lation, usually 10-90 ma, with intraoperative field stimulation
Ipsilateral brachial plexus-contralateral brachial plexus requiring currents at the lower end and surface stimulation re-
(EP1/EP2). quiring the higher currents. The amount of current, as discussed
• This would allow recording of peripheral nerve EPs. above, can be used to help determine within the operative field
Recommended Montages: Posterior tibial SEPs how close the surgeon is to neural tissue.
CZ-FZ
CZ'-FZ Recording
C7-FZ The pickup electrodes can be placed outside the operative
field, in which case usual surface electrodes are most often
Ipsilateral popliteal fossa-knee reference used, well secured to prevent being dislodged during surgeon
and staff incidental contact and pressure. If identification of ex-
BAEPs
actly which branch is required, occasionally the pickup elec-
EEG scalp electrodes are prepared in a similar manner as trode will be an intramuscular wire (see below). These wires
listed above and electrode impedance maintained below 5 kohm. allow very precise localization of which nerve portion was stim-
Stimulation is by ipsilateral alternating compression and rarefac- ulated but do not permit the quantification of nerve or muscle
tion clicks at 90 dB HL of 100 p.s duration at 20 Hz rate with 70 evoked response decrement that surface electrodes can perform.
bB contralateral wideband pseudorandom masking noise. Intraoperative field electrodes, typically sterile hooks or special
Electrodes are placed in the vertex region at CZ and CZ' and on forceps, can be used to either stimulate or pick up a nerve or
both ears (Al and A2). Low filter is set at 150 Hz, and the high muscle response within the field of dissection. The nearer the
filter at 3 kHz. Typically at least 1000 averages are required. 343
electrode is to the stimulated tissue, the larger the response ex-

EEG pected. Averaging may not be required for near-nerve record-
Usually all scalp electrodes are applied and held securely with ings but usually will be required for surface nerve evoked
collodion. Electrode impedance is maintained below 5 kohm. A potential recordings. Triggered EMG and motor evoked poten-
bipolar anteroposterior 16-channel montage covering the tial recordings often do not require averaging, but multiple trials
parasagittal and temporal regions provides adequate evaluation are necessary to confirm results.
C h a p t e r 12 INTRAOPERATIVE NEUROPHYSIOLOGIC MONITORING — 4 7 !
INTRAOPERATIVE EMG TECHNIQUES 2. Adams DC, Emerson RG: Intraoperative spinal cord monitoring. Curr Opin
Anaesthesiol 1996;9:371-375.
3. Agnew WF, McCreery DB: Considerations for safety in the use of extracranial
Recording intraoperative isolated muscle EMG signals requires stimulation for motor evoked potentials. Neurosurgery 1987;20:143-147.
initially placing intramuscular fine wires (0.03-mm insulated 4. Ahn S, Jordan SE, Nuwer MR, et al: Computed electroencephalographic topo-
nickel-chromium wires with hooked 3-mm bared tips) into target graphic brain mapping: a new and accurate monitor of cerebral circulation and
function for patients having carotid endarterectomy. J Vase Surg 1988;8:247-254.
muscles through a 26-gauge needle preoperatively. These leads 378
5. Albanese SA, Spadaro JA, Lubicky JP, Henderson NA: Somatosensory cortical
are then secured so that incidental pressure does not pull out the evoked potential changes after deformity correction. Spine 1991;16(suppl):
electrodes, which can be easily dislodged or removed by traction S371-S374.
6. Allen GC: Assessing and responding to detected abnormalities during intraop-
on the percutaneous leads. The muscle sites examined are usually
erative neurophysiologic monitoring. In: Russell GB, Rodichok LD (eds):
outside the surgical area and thus not within the dissection field. Primer of Intraoperative Monitoring. Boston, Butterworth-Heinemann, 1995, pp
With these electrodes, spontaneous free running EMG signals can 245-259.
be monitored for "neurotonic discharges," which may be bursting 7. Amantini A, Bartelli M, de Scisciolo D, et al: Monitoring of somatosensory
evoked potentials during carotid endarterectomy. J Neurol 1992;239:241-247.
or in trains, both suggesting motor nerve irritation. They can also 8. American Academy of Neurology: Assessment: Intraoperative neurophysiol-
be used to record stimulated responses in triggered EMG mode, to ogy. Neurology 1990;40:1644-1646.
assist with dissection or nerve branch identification. 358
9. American Electroencephalographic Society: Guideline Eleven: Guidelines for
intraoperative monitoring of sensory evoked potentials. J Clin Neurophysiol
1994;11:77-87.
10. Anderson SK, Loughnan BA, Hetreed MA: A technique for monitoring evoked
CONCLUSION potentials during scoliosis and brachial plexus surgery. Ann Royal Coll Surg
Engl 1990;72:321-323.
11. Andrews R (ed): Intraoperative Neuroprotection. Baltimore, Williams &
The field of IOM is relatively new and continues in develop- Wilkins, 1996.
ment with increasing acceptance and demand, which requires the 12. Angel A, LeBeau F: A comparison of the effects of propofol with other anaes-
understanding of clinical neurophysiologic techniques across mul- thetic agents on the centripetal transmission of sensory information. Gen
Pharmac 1992;23:945-963.
tiple disciplines. Not all new intraoperative neurophysiologic 13. Apel DM, Marrero G, King J, et al: Avoiding paraplegia during anterior spinal
monitoring techniques prove beneficial. - Although some of
143 279
surgery: the role of somatosensory evoked potential monitoring with temporary
these techniques were initially investigational, others are now occlusion of segmental spinal arteries. Spine 1991;16(suppl):S365-S370.
standard protocol during certain surgical procedures. In 1992, the 14. Astrup J, Symon L, Branston NM, Lassen NA: Cortical evoked potential and
extracellular K+ and H+ at critical levels of brain ischemia. Stroke 1977;
Scoliosis Research Society issued a position statement that intra- 8:51-57.
operative neurophysiologic monitoring during spinal cord surgery 15. Astrup J: Energy-requiring cell functions in the ischemic brain: their critical supply
involving instrumentation is not investigational and is considered and possible inhibition in protective therapy. J Neurosurg 1982;56:482-497.
16. Baba H, Tomita K, Umeda S, et al: Clinical study of spinal cord evoked poten-
a viable alternative as well as an adjunct to the wake-up test. 117 3 0 1
tials. In DuckerTL, Brown RH (eds): Neurophysiology and Standards of Spinal

After an extensive review of the literature, the position state- Cord Monitoring. New York, Springer, 1988, pp 216-221.
ment of the Therapeutics and Technology Subcommittee of the 17. Baba H, Kawahara N, Nagata S, et ai: Spinal cord evoked potentials in cervical
and thoracic myelopathy. In Jones SJ, Boyd S, Hetreed M, Smith NJ (eds):
American Academy of Neurology - concluded that the fol-
99 402
Handbook of Spinal Cord Monitoring. London, Kluwer, 1994, pp 99-103

lowing intraoperative neurophysiological monitoring tech- 18. Baba H, Kawahara K, Tomita K, et al: Spinal cord evoked potentials in spinal
niques are useful and non-investigational: cord intermittent claudication. In Jones SJ, Boyd S, Hetreed M, Smith NJ (eds):
1. EEG, CSA, and SEP in CEA and brain surgeries that po- Handbook of Spinal Cord Monitoring. London, Kluwer, 1994, pp 104-109.
19. Baumann SB, Welch WC, Bloom MJ: Intraoperative SSEP detection of ulnar
tentially compromise cerebral blood flow nerve compression or ischemia in an obese patient: a unique complication asso-
2. BAEP and CN monitoring in surgeries performed in the ciated with a specialized spinal retraction system. Arch Phys Med Rehabil
region of the brain stem or inner ear 2000;81:130-132.
20. Ben-David B, Haller G, Taylor P: Anterior spinal fusion complicated by para-
3. SEP monitoring performed for surgical procedures poten-
plegia: a case report of a false-negative somatosensory evoked potential. Spine
tially involving ischemia or mechanical trauma of the spinal cord. 1987;12:536-539.
The subcommittee also came to the conclusion that, although 21. Ben-David B: Spinal cord monitoring. Scoliosis 1988; 19:427^48.
promising, MEPs and VEPs are still investigational. 22. Bennett HL, Benson DR: Somatosensory evoked potentials for orthopaedic
spine trauma. J Orthop Trauma 1989;3:11-18.
Future developments include enhancing the automation of in- 23. Berenstein A. Young W, Ransohoff J, et al: Somatosensory evoked potentials
traoperative monitoring, monitoring additional physiologic para- during spinal angiography and therapeutic transvascular embolization. J
meters not yet investigated (see Table 12-2), and further studies to Neurosurg 1984;60:777-785.
verify cost-effectiveness for the many procedures for which vari- 24. Bieber E, Tolo V, Uematsu S: Spinal cord monitoring during posterior spinal in-
strumentation and fusion. Clin Orthop 1988;229:121-124.
ous intraoperative electrophysiologic techniques have become 25. Bolton CF, McFarlane RM: Human pneumatic tourniquet paralysis. Neurology
customarily employed, including establishing MEPs as a routine 1978;28:787-793.
and effective clinical tool. Automation, with the assistance of arti- 26. Boyd SG, Silva LD: EEG and serum prolactin studies in relation to transcuta-
neous stimulation of central motor pathways. J Neurol Neurosurg Psychiatry
ficial intelligence, offers economic promise but must survive the 1986;49:954-956.
medicolegal challenges that still propel much of the interest in the 27. Bradshaw K, Webb JK, Fraser AM: Clinical evaluation of spinal cord monitor-
currently labor-intensive effort of intraoperative electrophysio- ing in scoliosis surgery. Spine 1984;9:636-643.
logic monitoring. - "Above all, do no harm," the physician's
252 424
28. Branston NM, Symon L, Crockard HA, Pasztor E: Relationship between the
cortical evoked potential and local cortical blood flow following acute middle
creed, is clearly the best incentive for continued development
150
cerebral artery occlusion in the baboon. Exp Neurol 1974;45:195-208.

of the field of intraoperative electrophysiologic monitoring. 29. Branston NM, Symon L, Crockard HA: Recovery of the cortical evoked re-
sponse following temporary middle cerebral artery occlusion in baboons: rela-
tion to local blood flow and P 0 . Stroke 1976;7:151-157.
2
30. Branston NM, Strong AJ, Symon L: Extracellular potassium activity, evoked
REFERENCES
potential and tissue blood flow: relationships during progressive ischaemia in
baboon cerebral cortex. J Neurol Sci 1977;32:305-321.
31. Branston NM, Ladds A, Symon L, et al: Somatosensory evoked potentials in
1. Abel MF, Mubarak SJ, Wenger DR, et al: Brainstem evoked potentials for scol- experimental brain ischemia. In Pfurtscheller G, Jonkman EG, Lopes da Silva
iosis surgery: a reliable method allowing use of halogenated anesthetic agents. J FH (eds): Brain Ischemia: Quantitative EEG and Imaging Techniques, Progress
Pediatr Orthop 1990;10:208-213. in Brain Research, Vol 62. Amsterdam, Elsevier, 1984, pp 185-199.
32. Branston NM, Ladds A, Symon L, Wang AD: Comparison of the effects of is- 62. Colletti V, Fiorino F, Policante Z, Bruni L: Intraoperative monitoring of facial
chaemia on early components of the somatosensory evoked potential in brainstem, nerve antidromic potentials during acoustic neuroma surgery. Acta Otolaryngol
thalamus, and cerebral cortex. J Cerebral Blood Row Metabol 1984;4:68-81. (Stockh) 1997;117:663-669.
33. Brierley JB, Prior PF, Caverley, et al: The patogenesis of ischemic neuronal 63. Colletti V, Fiorino FG: Advances in monitoring of seventh and eighth cranial
damage along the cerebral arterial boundary zones in papio anubis. Brain nerve function during posterior fossa surgery. Am J Otol 1998;19:503-512.
1979;103:929-965. 64. Crockard HA, Brown FD, Trimble J, Mullan JF: Somatosensory evoked poten-
34. Brinkman SD, Braun P, Ganji S, et al: Neuropsychological performance one tials, cerebral blood flow and metabolism following cerebral missile trauma in
week after carotid endarterectomy reflects intra-operative ischemia. Stroke monkeys. Surg Neurol 1977;7:281-287.
1984;15:497-503. 65. Cunningham JN Jr, Laschinger JC, Merkin HA, et al: Measurement of spinal
35. Brodkey JS, Richards DE, Blasingame JP, Nulsen FE: Reversible spinal cord * cord ischemia during operations upon the thoracic aorta: Initial clinical experi-
trauma in cats: additive effects of direct pressure and ischemia. J Neurosurg ence. Ann Surg 1982;196:285-296.
1972;37:591-593. 66. Cunningham JN Jr, Laschinger JC, Spencer FC: Monitoring of somatosensory
36. Brodsky JB: What intraoperative monitoring makes sense? Chest 1999; 115: evoked potentials during surgical procedures on the thoracoabdominal aorta: IV.
101S-105S. Clinical observations and results. J Thorac Cardiovasc Surg 1987;94:275-285.
37. Brown RH, Nash CL Jr, Berilla JA, Amaddio MD: Cortical evoked potential 67. Dasmahapatra HK, Coles JG, Taylor MJ, et al: Identification of risk factors for
monitoring: a system for intraoperative monitoring of spinal cord function. spinal cord ischemia by the use of monitoring of somatosensory evoked poten-
Spine 1984;9:256-261. tials during coarctation repair. Circulation 1987;76(suppl III): 14—18.
38. Brown WF, Veitch J: AAEM Minimonograph #42: Intraoperative monitoring of 68. Daube JR: Monitoring of spine surgery with evoked potentials. In Schramm J,
peripheral and cranial nerves. Muscle Nerve 1994;17:371-377. M0ller AE (eds): Intraoperative Neurophysiologic Monitoring in Neurosurgery.
39. Buchtal A, Belopavlovic M, Mooij JJA: Evoked potential monitoring and tem- New York, Springer; 1991, pp 127-137.
porary clipping in cerebral aneurysm surgery. Acta Neurochir 1988;93:28-36. 69. Dawson EG, Sherman JE, Kanim LEA, Nuwer MR: Spinal cord monitoring:
40. Buchtal A, Belopavlovic M: Somatosensory evoked potentials in cerebral Results of the Scoliosis Research Society and the European Spinal Deformity
aneurysm surgery. Eur J Anaesth 1992;9:493-497. Society survey. Spine 1991; 16(suppl):S361-S364.
41. Budnick B, McKeown KL, Wiederholt K: Hypothermia-induced changes in rat 70. de Haan P, Kalkman CJ, de Mol BA, et al: Efficacy of transcranial motor-
short latency somatosensory evoked potentials. Electroenceph Clin Neuro- evoked myogenic potentials to detect spinal cord ischemia during operations for
physiol 1981;51:19-31. thoracoabdominal aneurysms. J Thorac Cardiovasc Surg 1997;113:87-101.
42. Burke D, Hicks R, Stephen J, et al: Assessment of corticospinal and somatosen- 71. Deletis V: Intraoperative monitoring of the functional integrity of the motor
sory conduction simultaneously during scoliosis surgery. Electroencephalogr pathways. Adv Neurol 1993;63:201-14.
Clin Neurophysiol 1992;85:388-396. 72. de Mol BAJM, Boezeman EHJF, Hamerlijnck RPHM, de Geest R: Experimental
43. Burke D, Hicks RG: Surgical monitoring of motor pathways. J Clin Neuro- and clinical use of somatosensory evoked potentials in surgery of aneurysms of
physiol 1998;15:194-205. the descending thoracic aorta. Thorac Cardiovasc Surg 1990;38:146-150.
44. Burke D, Nuwer MR, Daube J, et al: Intraoperative monitoring. In Deuschl G, 73. Desmedt JE (ed): Neuromonitoring in Surgery. Amsterdam, Elsevier, 1989.
Eisen A (eds): Recommendations for the Practice of Clinical Neurophysiology. 74. Deutsch E, Sohmer H, Weidenfeld J, et al: Auditory nerve brain-stem evoked
Amsterdam, Elsevier, 1999, pp 133-148. potentials and EEG during severe hypoglycemia. Electroencephalogr Clin
45. Calancie B, Lebwohl N, Madsen P, Klose KJ: Intraoperative evoked EMG mon- Neurophysiol 1983;55:714-716.
itoring in an animal model: a new technique for evaluating pedicle screw place- 75. DeVivo MJ, Whiteneck GG, Charles ED: The economic impact of spinal cord
ment. Spine 1992;17:1229-1235. injury. In Stover SL, DeLisa JA, Whiteneck GG (eds): Spinal Cord Injury: Clinical
46. Calancie B, Madsen P, Lebwohl N: Stimulus-evoked EMG monitoring during Outcomes from the Model Systems. Gaitherburg, MD, Aspen, 1995, pp 234-271.
transpedicular lumbosacral spine instrumentation. Spine 1994;19:2780-2786. 76. De Vleeschauwer P, Horsch S, Haupt WF, Huber P: The use of somatosensory
47. Calancie B, Harris W, Broton JG, et al: "Threshold-level" multipulse transcra- evoked responses in carotid surgery for monitoring brain function. Acta Chir
nial electrical stimulation of motor cortex for intraoperative monitoring of Belg 1985;85:293-298.
spinal motor tracts: description of method and comparison to somatosensory 77. De Vleeschauwer P, Horsch S, Matamoros R: Monitoring of somatosensory
evoked potential monitoring. J Neurosurg 1998;88:457^470. evoked potentials in carotid surgery: results, usefulness and limitations of the
48. Campbell WW, Sahni SK, Pridgeon RM, et al: Intraoperative electroneurogra- method. Ann Vase Surg 1988;2:63-68.
phy: management of ulnar neuropathy at the elbow. Muscle Nerve 1988; 11: 78. Diaz JH, Lockhart CH: Postoperative quadriplegia after spinal fusion for scol-
75-81. iosis with intraoperative awakening. Anesth Analg 1987;66:1039-1042.
49. Chabot RJ, John ER, Princhep LS: Real-time intraoperative monitoring during 79. Dietz A, von Kummer R, Adams HP, et al: Balloon occlusion test of the internal
neurosurgical and neuroradiological procedures. In Ducker TB, Brown RH carotid artery for evaluating resectability of blood vessel infiltrating cervical
(eds): Neurophysiology and Standards of Spinal Cord Monitoring. New York, metastasis of advanced head and neck cancers—Heidelberg experience.
Springer, 1986, pp 207-215. Laryngo-Rhinoot-Ologie 1993;72:558-567.
50. Chang SD, Lopez JR, Steinberg GK: The use of electrophysiologic monitoring 80. Dinner DS, Luders H, Lesser RP, et al: Intraoperative spinal somatosensory
during intracranial cerebrovascular surgery: A cost-benefit analysis. Stroke evoked potential monitoring. J Neurosurg 1986;65:807-814.
1999; 30:257. 81. Djuric S, Milenkovic Z, Klopcic-Spevak M, Spasic M: Somatosensory evoked
51. Chassard D, Joubaud A, Colson A, et al: Auditory evoked potentials during potential monitoring during intracranial surgery. Acta Neurochir 1992; 119:
propofol anaesthesia in man. Br J Anaesth 1989;62:522-526. 85-90.
52. Chatrian G-E, Berger MS, Wirch AL: Discrepancy between intraoperative 82. Dolan EJ, Transfeldt EE, Tator CH, et al: The effect of spinal distraction on re-
SSEPs and postoperative function. J Neurosurgery 1988;69:450-454. gional blood flow in cats. J Neurosurg 1980;53:756-764.
53. Cheek JC: Posterior fossa intraoperative monitoring. J Clin Neurophysiol 82a. Dolence VV: Contemporary treatment of peripheral nerve and brachial plexus
1993;10:412-424. lesions. Neurosurg Rev 1986;9:149-156.
54. Cheliout-Heraut F, Vital JM, Pouliquen JC, et al: Surveillance per-operatoire 83. Dolman J, Silvay G, Zappulla R, et al: The effect of temperature, mean arterial
des scolioses chez Tadulte jeune par les potentiels evoques somesthesiques: a pressure, and cardiopulmonary bypass flows on somatosensory evoked poten-
propos de 33 cas. Neurophysiol Clin 1989;19:297-310. tial latency in man. Thorac Cardiovasc Surg 1986;34:217-222.
55. Chiappa KH, Burke SR, Young RR: Results of electroencephalogram monitor- 84. Donner TR, Kline DG: Extracranial spinal accessory nerve injury. Neuro-
ing during 367 carotid endarterectomies: use of a dedicated minicomputer. surgery 1993;32:907-910.
Stroke 1979;10:381-388. 85. Doyle WJ, Fria TJ: The effects of hypothermia on the latencies of the auditory
56. Cho I, Smullens SN, Streletz LJ, Fariello RG: The value of intraoperative EEG brainstem response (ABR) in the rhesus monkey. Electroenceph Clin
monitoring during carotid endarterectomy. Ann Neurol 1986;20:508-551. Neurophysiol 1985;60:258-266.
57. Cloughesy TF, Nuwer MR. Hoch D, et al: Monitoring carotid test occlusions with 86. Drenger B, Parker SD, McPherson RW, et al: Spinal cord stimulation evoked
continuous EEG and clinical examination. J Clin Neurophysiol 1993;10:363-369. potentials during thoracoabdominal aortic aneurysm surgery. Anesthesiology
58. Cohen BA, Huizenga BA: Dermatomal monitoring for surgical correction of 1992;76:689-695.
spondylolisthesis: a case report. Spine 1988; 13:1125-1128. 87. Dubois M, Coppola R, Buchsbaum MS, Lees DE: Somatosensory evoked po-
59. Cohen LG, Hallett M: Non-invasive electrical stimulation of the brain does not tentials during whole body hyperthermia in humans. Electroenceph Clin
cause short-term changes in the electroencephalogram. In Rossini PM, Marsden Neurophysiol 1981:52:157-162.
CD (eds): Non-Invasive Stimulation of Brain and Spinal Cord: Fundamentals 88. Ducati A, Landi A, Cenzato M, et al: Monitoring of brain function by means of
and Clinical Applications. New York, Liss, 1988, pp 159-161. evoked potentials in cerebral aneurysm surgery. Acta Neurochir Suppl
60. Coles JG, Wilson GJ, Sima AF, et al: Intraoperative detection of spinal cord is- 1988;42:8-13.
chemia using somatosensory cortical evoked potentials during thoracic aortic 89. Ducker TL, Brown RH (eds): Neurophysiology and Standards of Spinal Cord
occlusion. AnnThorac Surg 1982;34:299-306. Monitoring. New York, Springer. 1988, pp 261-267.
61. Coles JG, Wilson GJ, Sima AF, et al: Intraoperative management of thoracic 90. Echeverri A, Flexon PB: Electrophysiologic nerve stimulation for identifying
aortic aneurysm. Experimental evaluation of perfusion cooling of the spinal the recurrent laryngeal nerve in thyroid surgery: Review of 70 consecutive thy-
cord. J Thorac Cardiovasc Surg 1983;85:292-299. roid surgeries. Am Surgeon 1998;64:328-333.
91. Eisele DW: Intraoperative electrophysiologic monitoring of the recurrent laryn- 120. Griffiths IR, Trench JG, Crawford RA: Spinal cord blood flow and conduction
geal nerve. Laryngoscope 1996;106:443-449. during experimental cord compression in normotensive and hypotensive dogs. J
92. Elmore JR, Gloviczki P, Harper CM, et al: Failure of motor evoked potentials to Neurosurg 1979;50:353-360.
predict neurologic outcome in experimental thoracic aortic occlusion. J Vase 121. Grundy BL, Heros RC, Tung AS, Doyle E: Intraoperative hypoxia detected by
Surg 1991;14:131-139. evoked potential monitoring. Anesth Analg 1981 ;60:437-439.
93. Emerson RG, Turner CA: Monitoring during supratentorial surgery. J Clin 122. Grundy BL, Nash CL, Brown RH: Arterial pressure manipulation alters spinal
Neurophysiol 1993;10:404-411. cord function during corrections of scoliosis. Anesthesiology 1981;54:249-253.
94. Engler GL, Spielholz Nl, Bernhard WN, et al: Somatosensory evoked potentials 122a. Grundy BL, Nash GL, Brown RH: Deliberate hypotension for spinal fusion:
during Harrington instrumentation for scoliosis. J Bone Joint Surg [Am] Prospective randomized study with evoked potential monitoring. Can Anesth
1978;60:528-532. Soc J 1982;29:452-462.
95. Epstein NE, Danto J, Nardi D: Evaluation of intraoperative somatosensory- 123. Grundy BL: Monitoring of sensory evoked potentials during neurosurgical op-
evoked potential monitoring during 100 cervical operations. Spine 1993; 18: erations: methods and applications. Neurosurgery 1982; 11:556-575
737-747. 124. Grundy BL, Jannetta PJ, Procopio PT, et al: Intraoperative monitoring of brain-
96. Erwin CW, Erwin AC: Up and down the spinal cord: intraoperative monitoring stem auditory evoked potentials. J Neurosurg 1982;57:674-681.
of sensory and motor spinal cord pathways. J Clin Neurophysiol 1993; 10: 125. Grundy BL, Nelson PB, Lina A, Heros RC: Monitoring of cortical somatosen-
425-436. sory evoked potentials to determine the safety of sacrificing the anterior cere-
97. Erwin CW, Erwin AC: The use of brain stem auditory evoked potentials in bral artery. Neurosurgery 1982;11:64-67.
intraoperative monhoring. In Russell GB, Rodichok LD (eds): Primer of 126. Guerit JM, Witdoeckt C, de Tourtchaninoff M, et al: Somatosensory evoked po-
Intraoperative Monitoring. Boston, Butterworth-Heinemann, 1995, pp 135— tential monitoring in carotid surgery. I. Relationships between qualitative SEP
158. alterations and intraoperative events. Electroencephalogr Clin Neurophysiol
98. Fava E, Bortolani EM, Ducati A, Ruberti U: Evaluation of spinal cord function 1997;104:459-469.
by means of lower limb somatosensory evoked potentials in reparative aortic 127. Gugino LD, Aglio LS, Segal ME, et al: Use of transcranial magnetic stimulation
surgery. J Cardiovasc Surg 1988;29:421-427. for monitoring spinal cord motor paths. Semin Spine Surg 1997;9:315-336.
99. Fisher RS, Raudzens P, Nunemacher M: Efficacy of intraoperative neurophysi- 128. Hacke W: Evoked potentials monitoring in interventional neuroradiology. In Des-
ological monitoring. J Clin Neurophysiol 1995;12:97-109. medt JE (ed): Neuromonitoring in Surgery. Amsterdam, Elsevier, 1989, pp 334-340.
100. Fode NC, Sundt TM, Robertson JT, et al: Multicenter retrospective review of 129. Haghighi SS, Green KD, Oro JJ, et al: Depressive effect of isoflurane anesthe-
results and complications of carotid endarterectomy in 1981. Stroke 1986; 17: sia on motor evoked potentials. Neurosurgery 1990;26:993-997.
370-376.' 130. Hahn JF, Lesser R, Klem G, Lueders H: Simple technique for monitoring intra-
101. Follett KA: Intraoperative electrophysiological spinal cord monitoring. In operative spinal cord function. Neurosurgery 1981;9:692-695.
Loftus CM, Traynelis VC (eds): Intraoperative Monitoring Techniques in 131. Haines SJ, Torres F: Intraoperative monitoring of the facial nerve during de-
Neurosurgery- New York, McGraw-Hill, 1994, pp 231-238. compressive surgery for hemifacial spasm. J Neurosurg 1991;74:254-257.
102. Forbes HJ, Allen PW, Waller CS, et al: Spinal cord monitoring in scoliosis 132. Halonen JP, Jones SJ, Edgar MA, Ransford AO: Multi-level epidural recordings
surgery: experience with 1168 cases. J Bone Joint Surg [Br] 1991;73:487-491. of spinal SEPs during scoliosis surgery. New Trends Adv Tech Clin Neuro-
103. Friedman WA, Kaplan BJ, Gravenstein D, Rhoton L Jr: Intraoperative brain- physiol 1990;41:342-347.
stem auditory evoked potentials during posterior fossa microvascular decom- 133. Halsey JH: Risks and benefits of shunting in carotid endarterectomy. Stroke
pression. J Neurosurg 1985;62:552-557. 1992;23:1583-1587.
104. Friedman WA, Kaplan BL, Day AL, et al: Evoked potential monitoring during 134. Hargadine JR, Branston NM, Symon L: Central conduction time in primate
aneurysm operation: observations after fifty cases. Neurosurgery 1987;20: brain ischemia: A study in baboons. Stroke 1980; 11:637-642.
678-687. 135. Harner SG, Daube JR, Ebersold MJ, Beatty CW: Improved preservation of
105. Friedman WA, Chadwick MA, Verhoeven FJ, et al: Monitoring of somatosen- facial nerve function with use of electrical monitoring during removal of
sory evoked potentials during surgery for middle cerebral artery aneurysms. acoustic neuromas. Mayo Clin Proc 1987;62:92-102.
Neurosurgery 1991;29:83-88. 136. Harper CM, Daube JR, Litchy WJ, Klassen RA: Lumbar radiculopathy after
106. Fujioka H, Shimoji K, Tomita S, et al: Spinal cord potential recordings from the spinal fusion for scoliosis. Muscle Nerve 1988; 11:386-391.
extradural space during scoliosis surgery. Br J Anesthesiol 1994;73:350-356. 137. Harper CM: Intraoperative monitoring of peripheral nerves. 1989 AAEE
107. Fukaya C, Katayama Y, Kasai M, et al: Intraoperative electrooculographic mon- Course B: Intraoperative Monitoring; AAEE Twelfth Annual continuing
itoring of oculomotor nerve function during skull base surgery. J Neurosurg Education Courses. Washington, DC, September 14, 1989, pp 19-26.
1999;91:157-159. 138. Harper CM, Daube JR: Surgical monitoring with evoked potentials: The Mayo
108. Fukui Y, Satomi K, Okuma T, et al: Intraoperative spinal cord monitoring using Clinic experience. In Desmedt JE (ed): Neuromonitoring in Surgery.
evoked spinal cord potentials on spinal cord tumors. In Shimoji K, Kurokawa T, Amsterdam, Elsevier, 1989, pp 275-301.
Tamaki T, WillisW (eds): Spinal Cord Monitoring and Electrodiagnosis. New 139. Harper CM, Harner SG, Slavit DH, et al: Effect of BAEP monitoring on hearing
York, Springer, 1988, pp 53-57. preservation during acoustic neuroma resection. Neurology 1992:42:1551-1553.
109. Galla JD, Ergin A, Sadeghi AM, et al: A new technique using somatosensory 140. Harper CM: Brain stem and cranial nerve monitoring. In Daube JR (ed):
evoked potential guidance during descending and thoracoabdominal aortic re- Clinical Neurophysiology. Philadelphia, FA Davis, 1996, pp 451-456.
pairs. J Cardiac Surg 1994;9:662-672. 141. Harper CM, Daube JR: Facial nerve electromyography and other cranial nerve
110. Gigli GL, Caramia M, Marciani MG, et al: Monitoring of subcortical and cortical monitoring. J Clin Neurophysiol 1998;15:206-216.
somatosensory evoked potentials during carotid endarterectomy: Comparison with 142. Haupt WF, Horsch S: Evoked potential monitoring in carotid surgery: a review
stump pressure levels. Electroencephalogr Clin Neurophysiol 1987;68:424-432. of 994 cases. Neurology 1992;42:835-838.
111. Ginsburg HH, Shelter AG, Raudzens PA: Postoperative paraplegia with pre- 143. Hays RM, McLaughlin JF, Bjornson KF, et al: Electrophysiological monitoring
served intraoperative somatosensory evoked potentials. J Neurosurg 1985;63: during selective dorsal rhizotomy, and spasticity and GMFM performance.
296-300. Develop Med Child Neurol 1998;40:233-238.
112. Glassman SD, Shields CB, Linden RD: Anesthetic effects on motor evoked po- 144. Herdmann J, Lumenta CB, Huse KO: Magnetic stimulation for monitoring of
tentials in dogs. Spine 1993;18:1083-1089. motor pathways in spinal procedures. Spine 1993;18:551-559.
113. Goldberg G, Goldstein H: AAEM case report 32: Nerve injury associated with 145. Hickey C, Gugino LD, Aglio LS, et al: Intraoperative somatosensory evoked
hip arthroplasty. Muscle Nerve 1998:21:519-527. potential monitoring predicts peripheral nerve injury during cardiac surgery.
113a. Goldbrunner RH, Schlake H-P, Mileuski C, et al: Quantitative parameters of Anesthesiology 1993;78:29-35.
intraoperative electromography predict facial nerve outcome for vestibular 146. Hicks RG, Burke DJ, Stephen JPH: Monitoring spinal cord function during sco-
schwannoma surgery. Neurosurg 2000:46:1140-1146. liosis surgery with Cotrel-Dubousset instrumentation. Med J Aust 1991;154:
114. Gonzalez EG, Hajdu M. Keim H, Brand L: Quantification of intraoperative so- 82-86.
matosensory evoked potential. Arch Phys Med Rehabil 1984;65:721-725. .. 147. Hicks RG, Woodforth IJ, Crawford MR: Some effects of isoflurane on I waves
115. Gonzalez EG, Hajdu M, Keim H, et al: Intraoperative somatosensory evoked of the motor evoked potential. Br J Anesth 1992;69:130-136.
potential monitoring. Orthop Rev 1984;13:573-578. 148. Hill R, Sebel PS, de Bruijn N, Neville W: Alterations in somatosensory evoked
116. Grabb PA, Albright AL, Sclabassi RJ, Pollack IF: Continuous intraoperative potentials associated with inadequate venous return during cardiopulmonary
electromyographic monitoring of cranial nerves during resection of fourth ven- bypass. J Cardiothoracic Anesth 1987;1:48-50.
tricular tumors in children. J Neurosurg 1997;86:1-4. 149. Hitchon PW, Dyste GN, Osenbach RK, Jensen AE: Response of spinal cord
117. Graybeal JM: Intraoperative neurophysiological monitoring: A historical per- blood flow and motor and sensory evoked potentials to aortic ligation. Surg
spective. In Russell GB, Rodichok LD (eds): Primer of Intraoperative Neurol 1990;34:279-285.
Neurophysiologic Monitoring. Boston, Butterworth-Heinemann, 1995, pp 9-14. 150. Hippocrates: Epidemics, Book I, Chapter 11. In Bartletts' Famous Quotations,
118. Greenwood J: Surgical removal of intramedullary tumors. J Neurosurg 1967; 16th ed. New York, Little Brown, 1992.
26:276-282. 151. Holland NR, Kosluik JP: Continuous electromyographic monitoring to detect
119. Gregory PC, McGeorge AP, Fitch W, et al: Effects of hemorrhagic hypotension nerve root injury during thoracolumbar scoliosis surgery. Spine 1997;22:
on the cerebral circulation. II. Electrocortical function. Stroke 1979;10:719-723. 2547-2550.
152. Homma S, Tamaki T (eds): Fundamental and Clinical Applications of Spinal 181. Keller BP, Haghighi SS, Oro JJ, Eggers GWN: The effects of propofol anesthe-
Cord Monitoring. Tokyo, Saikon Press, 1984. sia on transcortical electric evoked potentials in the rat. Neurosurgery 1992;
153. Honet JC, Raikes JA, Kantrowitz A, et al: Neuropathy in the upper extremity 30:557-560.
after open-heart surgery. Arch Phys Med Rehabil 1976;57:264-267. 182. Kennedy WF, Byrne TF, Majid HA, Pavlak LL: Sciatic nerve monitoring during
154. Hormes JT, Chappuis JL: Monitoring of lumbosacral nerve roots during spinal revision total hip arthroplasty. Clin Orthop 1991;264:223-227.
instrumentation. Spine 1993;18:2059-2062. 183. Kileny P, Dobson D, Gelfand ET: Middle-latency auditory evoked responses
155. Hume AL, Durkin MA: Central and spinal somatosensory conduction times during open-heart surgery with hypothermia. Electroencephalogr Clin
during hypothermic cardiopulmonary bypass and some observations on the ef- Neurophysiol 1983;55:268-276.
fects of fentanyl and isoflurane anesthesia. Electroencephalogr Clin Neuro- 184,. Kimovec MA, Koht A, Sloan TB: Effects of sufentanil on median nerve so-
physiol 1986;65:46-58. matosensory evoked potentials. Br J Anaesth 1990;65:169-172.
156. Jaaskelainen SK, Teerijoki-Oksa T, Forssell K, et al: Intraoperative monitoring 185. Kimura J: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and
of the inferior alveolar nerve during mandibular sagittal-split osteotomy. Practice. Philadelphia, FA Davis, 1989, p 81.
Muscle Nerve 2000;23:368-375. 185a. Kinney SE, Prass R: Facial nerve dissection by use of acoustic (loudspeakers)
157. Jackson LE, Roberson JB Jr: Acoustic neuroma surgery: Use of cochlear nerve facial EMG monitoring. Otolaryngol Head Neck Surg 1986;95:458-463.
action potential monitoring for hearing preservation. Am J Otol 2000:21: 186. King FG: Functional neurophysiology. In Russell GB, Rodichok LD (eds): Primer
249-259. of Intraoperative Monitoring. Boston, Butterworth-Heinemann, 1995, p 30.
158. Jacobs LA, Brinkman SD, Morrell RM, et al: Long-latency somatosensory 187. Klee MR, Pierau FK, Faber DS: Temperature effects on resting potential and
evoked potentials during carotid endarterectomy. Am Surg 1983;49:338-344. spike parameters of cat motoneurons. Exp Brain Res 1974; 19:789.
159. Jacobson GP, Tew JM: Intraoperative evoked potential monitoring. J Clin 188. Kline DG: Surgical repair of peripheral nerve injury. Muscle Nerve 1990; 13:
Neurophysiol 1987;4:152-176. 843-852.
160. Jellinek D, Jewkes D, Symon L: Noninvasive intraoperative monitoring of 189. Kline DG, Hudson AR: Acute injuries of peripheral nerves. In Youmans JR (ed):
motor evoked potentials under propofol anesthesia: Effects of spinal surgery on Neurological Surgery: A Comprehensive Reference Guide to the Diagnosis and
the amplitude and latency of motor evoked potentials. Neurosurgery 1991 ;29: Management of Neurosurgical Problems, 3rd ed. Philadelphia, WB Saunders,
551-557. 1990, pp 2423-2510.
161. Jellinek D, Piatt M, Jewkes D, Symon L: Effects of nitrous oxide on motor 190. Kline DG, Happel LT: Penfield lecture. A quarter century's experience with
evoked potentials recorded from skeletal muscle in patients under total anesthe- intraoperative nerve action potential recording. Can J Neurol Sci 1993;20:3-10.
sia with intravenously administered propofol. Neurosurgery 1991;29:558-562. 191. Kline D, Kim D, Midha R, et al: Management and results of sciatic nerve in-
162. Johnston CE II, Happel LT Jr, Norris R, et al: Delayed paraplegia complicating juries: A 24-year experience. J Neurosurg 1998;89:13-23.
sublaminar segmental spinal instrumentation. J Bone Joint Surg [Am] 1986;68: 192. Kliot M, Slimp JC: Techniques for assessment of peripheral nerve function at
556-563. surgery. In Loftus CM, Traynelis VC (eds): Intraoperative Monitoring
163. Jones SJ: Clinical applications of short-latency somatosensory evoked poten- Techniques in Neurosurgery. New York, McGraw-Hill, 1994, pp 275-285.
tials. Ann NY Acad Sci 1982;388:369-387. 193. Kobrine AI, Evans DE, Rizzoli HV: Relative vulnerability of the brain and
164. Jones SJ, Edgar MA, Ransford AO: Sensory nerve conduction in the human spinal cord to ischemia. J Neurol Sci 1980;45:65-72.
spinal cord: epidural recordings made during scoliosis surgery. J Neurol 194. Kochs E, Treede RD, Schulte AM, Esch J: Increase in somatosensorially evoked
Neurosurg Psychiatry 1982;45:446-451. potentials during induction of anaesthesia with etomidate. Anaesthetist 1986;
165. Jones SJ, Edgar MA, Ransford AO, Thomas NP: A system for the electrophysi- 35:359-364.
ological monitoring of the spinal cord during operations for scoliosis. J Bone 195. Kopf GS, Hume AL, Durkin MA, et al: Measurement of central somatosensory
Joint Surg [Br] 1983;65:134-139. conduction time in patients undergoing cardiopulmonary bypass: An index of
166. Jones SJ, Harrison R, Koh KF, et al: Motor evoked potential monitoring during neurologic function. Am J Surg 1985;149:445-448.
spinal surgery: Responses of distal limb muscles to transcranial cortical stimu- 196. Koyanagi I, Iwasaki Y, Isu T, et al: Spinal cord evoked potential monitoring
lation with pulse trains. EEG Clin Neurophysiol 1996;100:375-383. after spinal cord stimulation during surgery of spinal cord tumors. Neuro-
167. Jones TH, Morawetz RB, Crowell RM, et al: Thresholds of focal cerebral is- surgery 1993;33:451-460.
chemia in awake monkeys. J Neurosurg 1981;54:773-782. 197. Kraft H: Effects of temperature and age on nerve conduction velocity in the
168. Kaga K, Takiguchi T, Myokai K, Shiode A: Effects of deep hypothermia and guinea pig. Arch Phys Med Rehab 1972;53:328-332.
circulatory arrest on the auditory brain stem responses. Arch Otorhinolaryngol 198. Krieger KH, Spencer FC: Is paraplegia after repair of coarctation of the aorta
1979;225:199-205. due principally to distal hypotension during aortic cross-clamping? Surgery
169. Kai Y, Owen JH, Lenke LG, et al: Use of sciatic neurogenic motor evoked po- 1985;97:2-7.
tentials versus spinal potentials to predict early-onset neurologic deficits when 199. Kurthen M, Schramm J: Application of intraoperative spinal cord monitoring to
intervention is still possible during overdistraction. Spine 1993; 18:1134-1139. neurosurgery. In Jones SJ, Boyd S, Hetreed M, Smith MJ (eds): Handbook of
170. Kaieda R, MaekawaT, Takeshita H: Effects of diazepam on evoked electrospino- Spinal Cord Monitoring. London, KJuwer, 1994, pp 59-70.
gram and evoked electromyogram in man. Anesth Analg 1981;60:197-200. 200. Lam AM, Manninen PH, Ferguson GG, Nantua W: Monitoring electrophysio-
171. Kalkman CJ, Drummond JC, Ribberink AA: Low concentrations of isoflurane logic function during carotid endarterectomy: a comparison of somatosensory
abolish motor evoked responses to transcranial electrical stimulation during ni- evoked potentials and conventional electroencephalogram. Anesthesiology
trous oxide/opioid anesthesia in humans. Anesth Analg 1991 ;73:410—415. 1991;75:15-21.
172. Kalkman CJ, Drummond JC, Ribberink AA: Effects of propofol, etomidate, mi- 201. LaMont RL, Wasson SL, Green MA: Spinal cord monitoring during spinal
dazolam, and fentanyl on motor evoked responses to transcranial electrical or surgery using somatosensory spinal evoked potentials. J Pediatr Orthop
magnetic stimulation in humans. Anesthesiology 1992;76:502-509. 1983;3:31-36.
173. Kalkman CJ, Drummond JC, Patel PM: Effects of droperidol, pentobarbital, 202. Landi A, Copeland SA, Parry CBW, Jones SJ: The role of somatosensory
and ketamine on myogenic transcranial magnetic motor-evoked responses in evoked potentials and nerve conduction studies in the surgical management of
humans. Neurosurgery 1994;35:1066-1071. brachial plexus injuries. J Bone Joint Surg 1980;62B:492—496.
174. Kano T, Shimoji K: The effects of ketamine and neuroleptanalgesia on the 203. Lang EW, Chesnut RM, Beutler AS, et al: The utility of motor-evoked potential
evoked electrospinogram and electromyogram in man. Anesthesiology 1974; monitoring during intramedullary surgery. Anesth Analg 1996;83:1337-1341.
40:241-246. 204. Lang EW, Beutler AS, Chesnut RM, et al: Myogenic motor-evoked potential
175. Kaplan BJ, Gravenstein D, Friedman WA: Intraoperative electrophysiology in monitoring using partial neuromuscular blockade in surgery of the spine. Spine
treatment of peripheral nerve injuries. J Fla Med Assoc 1984;71:400-403. 1996:21:1676-1686.
176. Kaplan BJ, Friedman WA, Alexander JA, Hampson SR: Somatosensory evoked 205. Laschinger JC, Cunningham JN Jr, Nathan IM, et al: Experimental and clinical
potential monitoring of spinal cord ischemia during aortic operations. assessment of the adequacy of partial bypass in maintenance of spinal cord
Neurosurgery 1986;19:82-90. blood flow during operations on the thoracic aorta. Ann Thorac Surg 1983:36:
177. Kawaguchi M, Sakamoto T, Ohnishi H, et al: Intraoperative myogenic motor 417-426.
evoked potentials induced by direct electrical stimulation of the exposed motor 206. Laschinger JC, Cunningham JN Jr, Nathan IM, et al: Intraoperative identifica-
cortex under isoflurane and sevoflurane. Anesth Analg 1996;82:593-599. tion of vessels critical to spinal cord blood supply—use of somatosensory
178. Kearse LA, Brown EN, McPeck K: Somatosensory evoked potential sensitivity evoked potentials. Curr Surg 1984;41:107-109.
relative to electroencephalography for cerebral ischemia during carotid en- 207. Laschinger JC, Izumoto H, Kouchoukos NT: Evolving concepts in prevention
darterectomy. Stroke 1992:23:498-505. of spinal cord injury during operations on the descending thoracic and thora-
179. Kearse LA, Martin D, McPeck K, Lopez-Breshnahan M: Computer-derived coabdominal aorta. Ann Thorac Surg 1987;44:667-674.
density spectral array in detection of mild analog electroencephalographic is- 208. Laschinger JC, Owen J, Rosenbloom M, et al: Direct noninvasive monitoring of
chemic pattern changes during carotid endarterectomy. J Neurosurg spinal cord motor function during thoracic aortic occlusion: Use of motor
1993;78:884-890. evoked potentials. J Vase Surg 1988;7:161-171.
180. Keith RW, Stambough JL, Awender SH: Somatosensory cortical evoked poten- 209. Laschinger JC, Cunningham JN Jr, Isom OW, et al: Definition of the safe lower
tials: a review of 100 cases of intraoperative spinal surgery monitoring. J Spinal limits of aortic resection during surgical procedures on the thoracoabdominal
Dis 1990:3:220-226. aorta: use of somatosensory evoked potentials. J Am Coll Cardiol 1998;3:959-965.
210. Lederman RJ, Breuer AC, Hanson MR, et al: Peripheral nervous system com- 239. Manninen PH, Patterson S, Lam AM, et al: Evoked potential monitoring during
plications of coronary artery bypass graft surgery. Ann Neurol 1982; 12: posterior fossa aneurysm surgery: A comparison of two modalities. Can J
297-301. Anaesth 1994;41:92-97.
211. Lee EK, Seyal M: Generators of short latency human somatosensory-evoked 240. Markand ON, Dilley RS, Moorthy S, Warren C: Monitoring of somatosensory
potentials recorded over the spine and scalp. J Clin Neurophysiol 1998; 15: evoked responses during carotid endarterectomy. Arch Neurol 1984;41:
227-234. 375-378.
212. Lee VC: Spinal and cortical evoked potential studies in the ketamine-anes- 241. Markand ON: Continuous assessment of cerebral function with EEG and so-
thetized rabbit: Fentanyl exerts component-specific, naloxone-reversible matosensory evoked potential techniques during extracranial vascular recon-
changes dependent on stimulus intensity. Anesth Analg 1994;78:280-286. struction. In Loftus CM, Traynelis VC (eds): Intraoperative Monitoring
213. Lesnick JE, Michele JJ, Simeone FA, et al: Alteration of somatosensory evoked Techniques in Neurosurgery. New York, McGraw-Hill, 1994, pp 19-32!
potentials in response to global ischemia. J Neurosurg 1984;60:490-494. 242. Marsh RR, Frewen TC, Sutton LN: Resistance of the auditory brain stem re-
214. Lesser RP, Raudzens P, Liiders H, et al: Postoperative neurological deficits may sponse to high barbiturate levels. Otolaryngol Head Neck Surg 1984;92:685-688.
occur despite unchanged intraoperative somatosensory evoked potentials. Ann 243. Matsuda H, Shimazu A: Intraoperative spinal cord monitoring using electric re-
Neurol 1986;19:22-25. sponses to stimulation of caudal spinal cord or motor cortex. In Desmedt JE
215. Levine RA: Monitoring auditory evoked potentials during cerebellopontine (ed): Neuromonitoring in Surgery. Amsterdam, Elsevier, 1989, pp 175-190.
angle tumor surgery: Relative value to electrocochleography brainstem auditory 244. Matsui Y, Goh K, Shiiya N, et al: Clinical application of evoked spinal cord po-
evoked potentials and cerebellopontine angle recordings. In Schramm J, M0ller tentials elicited by direct stimulation of the cord during temporary occlusion of
AR (eds): Intraoperative Neurophysiologic Monitoring in Neurosurgery. New the thoracic aorta. J Thorac Cardiovasc Surg 1994;107:1519-1527.
York, Springer, 1991,*pp 193-204. 245. Matthies C, Samii M: Management of vestibular schwannomas (acoustic neuro-
216. Levy WJ, Shapiro HM, Maruchak G, Meathe E: Automated EEG processing for mas): The value of neurophysiology for intraoperative monitoring of auditory
intraoperative monitoring: A comparison of techniques. Anesthesiology function in 200 cases. Neurosurgery 1997;40:459-468.
1980;53:223-236. 246. Mavroudakis N, Vandesteene A, Brunko E: Spinal and brain-stem SEPs and H-
217. Levy WJ, Kraus KH, Gugino LD, et al: Transcranial magnetic evoked poten- reflex during enflurane anesthesia. Electroencephalogr Clin Neurophysiol
tial monitoring. In Loftus CM, Traynelis VC (eds): Intraoperative 1994;92:82-85.
Monitoring Techniques in Neurosurgery. New York, McGraw-Hill, 1994, pp 247. McDonnell DE, Faniginj HF, Sullivan HG: Somatosensory evoked potentials
251-255. (SEP) intraoperative monitoring during cranial vertebral compression and insta-
218. Lindsey RL: A simple solution for determining shunt flow during carotid en- bility. In Ducker TB, Brown RB (eds): Neurophysiology and Standards of
darterectomy. Anesthesiology 1984;61:215-216. Spinal Cord Monitoring. New York, Springer, 1986, pp 251-260.
219. Little JR, Lesser RP, Liiders H: Electrophysiological monitoring during basilar 248. McPherson RW, Niedermeyer EF, Otenasek RJ, Hanley DF: Correlation of tran-
aneurysm operation. Neurosurgery 1987;20:421—427. sient neurological deficit and somatosensory evoked potentials after intracranial
220. Livesay JJ, Cooley DA, Ventemiglia RA, et al: Surgical experience in descend- aneurysm surgery. J Neurosurg 1983:59:146-149.
ing thoracic aneurysmectomy with and without adjuncts to avoid ischemia. Ann 249. McPherson RW, Mahla M, Johnson R, Traystman RJ: Effects of enflurane.
Thorac Surg 1985:39:37^6. isoflurane and nitrous oxide on somatosensory evoked potentials during fen-
221. Loder RT, Thomson GJ, LaMont RL: Spinal cord monitoring in patients with tanyl anesthesia. Anesthesiology 1985;62:626-633.
nonidiopathic spinal deformities using somatosensory evoked potentials. Spine 250. Messick JM, Sharbrough F, Sundt T: Selective shunting on the basis of EEG
1991;16:1359-1364. and regional CBF monitoring during carotid endarterectomy. Int Anesthesiol
222. Loftus CM, Dyste GN, Reinarz SJ, Hingtgen WL: Distal cervical carotid dis- Clin 1984;22:137-152.
section after carotid endarterectomy: A complication of indwelling shunt? 251. Meyer PR Jr, Cotler HB, Gireesan GT: Operative neurological complications
Neurosurgery 1986;19:441-445. resulting from thoracic and lumbar spine internal fixation. Clin Orthop
223. Loftus C, Traynelis V (eds): Introperative Monitoring Techiques in Neuro- 1988;237:125-131.
surgery. New York, McGraw Hill, 1994. 252. Michenfelder JD: Intraoperative monitoring of sensory evoked potentials may
224. Loftus CM, Quest DO: Technical issues in carotid artery surgery 1995. be neither a proven nor an indicated technique. J Clin Monit 1987;3:45-47.
Neurosurgery 1995;36:629-647. 253. Mizoi K, Yoshimoto T: Permissible temporary occlusion time in aneurysm
225. Lopez JR, Chang SC, Steinberg GK: Predictive value of electrophysiological surgery as evaluated by evoked potential monitoring. Neurosurgery 1993;33:
monitoring in the treatment on cerebral vascular malformations. Electroenceph 434-440.
Clin Neurophysiol 1997; 103:120. 254. Mochida K, Shinomiya K, Komori H, Furuya K: A new method of multiseg-
226. Lopez JR, Chang SC, Steinberg GK: The use of electrophysiological monitor- ment motor pathway monitoring using muscle potentials after train spinal stim-
ing in the intraoperative management of intracranial aneurysms. J Neurol ulation. Spine 1995;20:2240-2246.
Neurosurg Psychiatry 1999;66:189-196. 255. Molaie M: False negative intraoperative somatosensory evoked potentials with
227. Lubicky JP, Spadaro JA, Yuan HA, et al: Variability of somatosensory cortical simultaneous bilateral stimulation. Clin Electroencephalogr 1986;17:6-9.
evoked potential monitoring during spinal surgery. Spine 1989;14:790-798. 256. M0ller AR: Evoked Potentials in Intraoperative Monitoring. Baltimore,
228. Lueders H, Gurd A, Hahn J, et al: A new technique for intraoperative monitor- Williams & Wilkins, 1988.
ing of spinal cord function: Multichannel recording of spinal cord and subcorti- 257. M0ller AR (ed): Intraoperative Neurophysiologic Monitoring. Luxembourg,
cal evoked potentials. Spine 1982;7:110-115. Harwood, 1995.
229. Maccabee PJ, Levine DB, Pinkhasov EI, et al: Evoked potentials recorded from 258. Momma F, Wand AD, Symon L: Effects of temporary occlusion on somatosen-
scalp and spinous processes during spinal column surgery. Electroenceph Clin sory evoked responses in aneurysm surgery. Surg Neurol 1987;27:343-352.
Neurophysiol 1983;56:569-582. 259. Moorthy SS, Markand ON, Dilley RS, et al: Somatosensory-evoked responses
230. MacEwen GD, Bunnell WP, Sriram K: Acute neurological complications in the during carotid endarterectomy. Anesth Analg 1982;61:879-883.
treatment of scoliosis: A report of the scoliosis research society. J Bone Joint 260. Morawetz RB, DeGiorlami U, Ojemann RG, et al: Cerebral blood flow deter-
SurgjAm] 1975;57:404-408. mined by hydrogen clearance during middle cerebral artery occlusion in
231. Machida M, Weinstein SL, Yamada T, et al: Monitoring of motor action poten- unanesthetized monkeys. Stroke 1978;9:143-149.
tials after stimulation of the spinal cord. J Bone Jt Surg 1988;70A:911-918. 261. More RC, Nuwer MR. Dawson EG: Cortical evoked potential monitoring
232. Macon JB, Poletti CE, Sweet WH, Zervas NT: Spinal conduction velocity mea- during spinal surgery: Sensitivity, specificity, reliability, and criteria for alarm. J
surement during laminectomy. Surg Forum 1980;31:453—455. Spinal Disord 1988;1:75-80.
233. Macon JB, Poletti CE: Conducted somatosensory evoked potentials during 262. MoriokaT, Fujii K, Tobimatsu Shozo, et al: Usefulness of epidurally evoked
spinal surgery. Part 1: Control conduction velocity measurements. J Neurosurg cortical potential monitoring during eervicomedullary glioma surgery. J Clin
1982;57:349-353. Monit 1991;7:30-34.
234. Maeda S, Miyamoto T. Murata H, Yamashita K: Prevention of spinal cord is- 263. Morota N, Deletis V, Constantini S, et al: The role of motor evoked potentials
chemia by monitoring spinal cord perfusion pressure and somatosensory during surgery for intramedullary spinal cord tumors. Neurosurgery 1997;41:
evoked potentials. J Cardiovasc Surg 1989:30:565-571. 1327-1336.
235. Maguire J, Wallace S, Madiga R, et al: Evaluation of intrapedicular screw posi- 264. Mostegl A, Bauer R: The application of somatosensory-evoked potentials in or-
tion using intraoperative evoked electromyography. Spine 1995;20:1068-1074. thopedic spine surgery. Arch Orthop Trauma Surg 1984; 103:179—184.
236. Mahla ME, Long DM, McKennett J, et al: Detection of brachial plexus dys- 265. Mostegl A, Bauer R, Eichenauer M: Intraoperative somatosensory potential
function by somatosensory evoked potential monitoring—a report of two cases. monitoring: a clinical analysis of 127 surgical procedures. Spine 1987; 13:
Anesthesiology 1984;60:248-252. 396-400.
237. Makarov MR, Delgado MR, Birch J, Samchukov ML: Monitoring peripheral 266. Murakami M, Mochizuki M, Okamoto Y, et al: Spinal cord monitoring during
nerve function during external fixation of upper extremities. J Pediatric Ortho- surgery for intramedullary spinal cord tumors. In Jones SJ, Boyd S, Hetreed M.
paedics 1997;17:663-667. Smith NJ (eds): Handbook of Spinal Cord Monitoring. London, Kluwer, 1994,
238. Manninen PH, Lam AM, Nicholas JF: The effects of isoflurane and isoflurane- pp 197-204.
nitrous oxide anesthesia on brainstem auditory evoked potentials in humans. 267. Mustain WD, Kendig RJ: Dissociation of neurogenic motor and somatosensory
Anesth Analg 1985; 64:43-47. evoked potentials. Spine 1991;16:851-853.
268. Myers RR, Stockard JJ, Saidman LJ: Monitoring of cerebral perfusion during 297. Owen JH, Bridwell KH, Grubb R, et al: The clinical application of neurogenic
anesthesia by time-compressed fourier analysis of the electroencephalogram. motor evoked potentials to monitor spinal cord function during surgery. Spine
Stroke 1977;8:331-337. 1991; 16(suppl):S385-S390.
269. Nagao S, Roccaforte P, Moody RA: The effects of isovolemic hemodilution and 298. Owen JH, Kostuik JP, Gornet M, et al: The use of mechanically elicited elec-
reinfusion of packed erythrocytes on somatosensory and visual evoked poten- tromyograms to protect nerve roots during surgery for spinal degeneration.
tials. J Surg Res 1978;25:530-537. Spine 1994;19:1704-1710.
270. Nagao S, Roccaforte P, Moody RA: Acute intracranial hypertension and audi- 299. Owen JH, Sponseller PD, Szymanski J, Hurdle M: Efficacy of multimodality
tory brainstem responses. Part 1. Changes in the auditory brainstem and so- spinal cord monitoring during surgery for neuromuscular scoliosis. Spine
matosensory evoked responses in intracranial hypertension in cats. J Neurosurg 1995;20:1480-1488.
1979;51:669-676. 300. Owen JH: Applications of neurophysiological measures during surgery of the
271. Nagle KJ, Emerson RG, Adams DC, et al: Intraoperative monitoring of motor spine. In Frymoyer JW (ed): The Adult Spine: Principles and Practice, 2nd ed.
evoked potentials: A review of 116 cases. Neurology 1996;47:999-1004. Philadelphia, Lippincott-Raven, 1997, pp 673-702.
272. Nainzadeh NK, Neuwirth MG, Bernstein R, Cohen LS: Direct recording of 301. Padberg AM, Wilson-Holden TJ, Lenke LG, Bridwell KH: Somatosensory- and
spinal evoked potentials to peripheral nerve stimulation by a specially modified motor-evoked potential monitoring without a wake-up test during idiopathic
electrode. In Ducker TB, Brown RH (ed): Neurophysiology and Standards of scoliosis surgery. Spine 1998;23:1392-1400.
Spinal Cord Monitoring. Springer-Verlag, 1988, pp 234-244. 302. Paiva T, Campos J, Baeta E, et al: EEG monitoring during endovascular em-
273. Nash CL Jr, Lorig RA, Schatzinger LA, Brown RH: Spinal cord monitoring bolization of cerebral arteriovenous malformations. Electroencephalogr Clin
during operative treatment of the spine. Clin Orthop 1977;126:100-105. Neurophysiol 1995;95:3-13.
274. Nash CL Jr, Schatzinger LH, Brown RH, Brodkey J: The unstable stable tho- 303. Palatinsky E, DiScenna A, McDonald H, et al: SSEP and BAEP monitoring of
racic compression fracture: Its problems and the use of spinal cord monitoring temporary clip application and induced hypotension during cerebrovascular
in the evaluation of treatment. Spine 1977;2:261-265. surgery. In Loftus CM, Traynelis VC (eds): Intraoperative Monitoring
275. Nash CL, Brown RH: Current concepts review spinal cord monitoring. J Bone Techniques in Neurosurgery. New York, McGraw-Hill, 1994, pp 61-71.
Joint Surg 1989;71:627-630. 304. Parenti G, Marconi F, Fiori L: Electrophysiological (EEG-SSEP) monitoring
276. Nelson KR: Use of peripheral nerve action potentials for intraoperative moni- during middle cerebral aneurysm surgery. J Neurosurg Sci 1996;40:195-205.
toring. In Gilmore R (ed): Evoked Potentials. Neurologic Clinics. Philadelphia, 305. Pathak KS, Brown RH, Nash CL Jr: Continuous opioid infusion for scoliosis
WB Saunders, 1988;6:917-933. fusion surgery. Anesth Analg 1983;62:841-845.
277. Nelson KR, Phillips LH: Neurophysiologic monitoring during surgery of pe- 306. Pathak KS, Brown RH, Cascorbi HF, Nash CL Jr: Effects of fentanyl and mor-
ripheral and cranial nerves, and in selective dorsal rhizotomy. Semin Neurol phine on intraoperative somatosensory cortical-evoked potentials. Anesth Analg
1990;10:141-149. 1984;63:833-837.
278. Nelson KR, Goodheart R, Salotto A, Tibbs P: Median nerve entrapment beneath 307. Paulsen RD, Steinberg GK, Norbash AM, et al: Embolization of rolandic cortex
the bicipital aponeurosis: Investigation with intraoperative short segment stimu- arteriovenous malformations. Neurosurgery 1999;44:479-484.
lation. Muscle Nerve 1994;17:1221-1223. 308. Pechstein U, Cedzich C, Nadstawek J, Schramm J: Transcranial high-frequency
279. Nercessian OA, Gonzalez EG, Stinchfield FE: The use of somatosensory repetitive electrical stimulation for recording myogenic motor evoked potentials
evoked potential during revision or reoperation for total hip arthroplasty. Clin with the patient under general anesthesia. Neurosurgery 1996;39:335-344.
Orthop 1989;243:138-142. 309. Pechstein U, Nadstawek J, Zentner J, Schramm J: Isoflurane plus nitrous oxide
280. North RB, Drenger B, Beattie C, et al: Monitoring of spinal cord stimulation versus propofol for recording of motor evoked potentials after high frequency
evoked potentials during thoracoabdominal aneurysm surgery. Neurosurgery repetitive electrical stimulation. Electroencephalogr Clin Neurophysiol 1998;
1991;28:325-330. 108:175-181.
281. Nuwer JM, Nuwer MR: Neurophysiologic surgical monitoring staffing patterns 310. Perez-Borja C, Meyer JS: Electroencephalographic monitoring during recon-
in the USA. Electroencephalogr Clin Neurophysiol 1997;103:616-620. structive surgery of the neck vessels. Electroencephalogr Clin Neurophysiol
282. Nuwer MR, Dawson EC: Intraoperative evoked potential monitoring of the 1965;18:162-169.
spinal cord: A restricted filter, scalp method during Harrington instrumentation 311. Phillips LH, Park TS: Electrophysiological monitoring during lipomyelo-
for scoliosis. Clin Orthop 1984;183:42-50. meningocele resection. Muscle Nerve 1990;13:127-132.
283. Nuwer MR, Dawson EC: Intraoperative evoked potential monitoring of the 312. Phillips LH II, Blanco JS, Sussman MD: Direct spinal stimulation for intraoper- •
spinal cord: Enhanced stability of cortical recordings. Electroencephalogr Clin ative monitoring during scoliosis surgery. Muscle Nerve 1995;18:319-325.
Neurophysiol 1984;59:318-327. 313. Phillips LH, Jane JA: Electrophysiologic monitoring during tethered spinal cord
284. Nuwer MR (ed): Evoked Potential Monitoring in the Operating Room. New release. In Loftus CM (ed): Clinical Neurosurgery; Proceedings of the Congress
York, Raven Press, 1986. of Neurological Surgeons, San Francisco, 1995. Baltimore, Williams &
285. Nuwer MR: Monitoring spinal cord surgery with cortical somatosensory Wilkins, 1996, pp 163-174.
evoked potentials. In Desmedt JE (ed): Neuromonitoring in Surgery. 314. Pitman MI, Nainzadeh N, Ergas E, Springer S: The use of somatosensory
Amsterdam, Elsevier, 1989, pp 158-164. evoked potentials for detection of neurapraxia during shoulder arthroscopy. J
286. Nuwer MR: Technical parameters, artifacts, and quality assessment of intraop- Arthrosc Rel Surg 1988;4:250-255.
erative evoked potentials. In Schramm J, M0ller AR (eds): Intraoperative 315. Porkkala T, Jantti V, Kaukinen S, Hakkinen V: Somatosensory evoked poten-
Neurophysiologic Monitoring in Neurosurgery. New York, Springer, 1991, pp tials during isoflurane anaesthesia. Acta Anaesthesiol Scand 1994; 38:206-210.
18-27. 316. Porkkala T, Kaukinen S, Hakkinen V, Jantti V: Effects of hypothermia and ster-
287. Nuwer MR, Daube J, Fischer C, et al: Neuromonitoring during surgery: Report na! retractors on median nerve somatosensory evoked potentials. Acta
of an IFCN committee. Electroencephalogr Clin Neurophysiol 1993:87: Anaesthesiol Scand 1997;41:843-848.
263-276. 317. Press JM, Rayner SL, Philip M, et al: Intraoperative monitoring of an unusual
288. Nuwer MR, Dawson EG, Carlson LG, et al: Somatosensory evoked potential brachial plexus tumor. Arch Phys Med Rehabil 1992;73:297-299.
spinal cord monitoring reduces neurologic deficits after scoliosis surgery: Re- 318. Prokop A, Meyer GP, Walter M, Erasmi H: Validity of SEP monitoring in carotid
sults of a large multicenter survey. Electroencephalogr Clin Neurophysiol surgery review and our own results. J Cardiovasc Surg 1996;37:337-342.
1995;96:6-11. 319. Rampil IJ, Holzer JA, Quest DO. el al: Prognostic value of computerized EEG
289. Nuwer MR: Spinal cord monitoring with somatosensory techniques. J Clin analysis during carotid endarterectomy. Anesth Analg 1983;62:186-192.
Neurophysiol. 1998;15:183-193. 320. Rampil IJ: Electroencephalogram. In Albin MA (ed): Textbook of Neuro-
290. Nuwer MR: Spinal cord monitoring. Muscle Nerve 1999;22:1620-1630. anesthesia with Neurosurgical and Neuroscience Perspectives. New York,
291. Oberle JW, Antoniadis G, Rath SA, Richter HP: Value of nerve action potentials McGraw-Hill, 1997, pp 193-220.
in the surgical management of traumatic nerve lesions. Neurosurgery 1997:41: 321. Raudzens PA: Intraoperative monitoring of evoked potentials. Ann NY Acad
1337-1344. Sci 1982;388:308-326.
292. Obi T, Mochizuki M, Isobe K, et al: Mechanically elicited nerve root discharge: 322. Redekop G, Ferguson G: Correlation of contralateral stenosis and intraoperative
Mechanical irritation and waveform. Acta Neurol Scand 1999;100:185-188. electroencephalogram changes with risk of stroke during carotid endarterec-
293. Ochiai N, Nagano A, Sugioka H, Hara T: Nerve grafting in brachial plexus in- tomy. Neurosurgery 1992;34:191-194.
juries. Results of free grafts in 90 patients. J Bone Joint Surg Br 1996:78: 323. Reilly EL, Kondo C, Brunberg JA, Doty DB: Visual evoked potentials during
754-758. hypothermia and prolonged circulatory arrest. Electroencephalogr Clin Neuro-
294. Ogilvie-Harris DJ, Wiley AM: Arthroscopic surgery of the shoulder. A general physiol 1978;45:100-106.
appraisal. J Bone Joint Surg 1986;68B:201-207. 324. Reuter DG, Tacker WA, Badylak SF, et al: Correlation of motor-evoked poten-
295. Ojemann RG, Levine RA, Montgomery WM, McGaffigan P: Use of intraopera- tial response to ischemic spinal cord damage. J Thorac Cardiovasc Surg 1992:
tive auditory evoked potentials to preserve hearing in unilateral acoustic neu- 104:262-272.
roma removal. J Neurosurg 1984;61:938-948. 325. Robertson SC, Traynelis VC, Yamada I T : Identification of the sensorimotor
296. Okamoto Y, Murakami M, NakagawaT, et al: Intraoperative spinal cord moni- cortex with SSEP phase reversal. In Loftus CM, Traynelis VC (eds):
toring during surgery for aortic aneurysm: Application of spinal cord evoked Intraoperative Monitoring Techniques in Neurosurgery. New York. McGraw-
potential. Electroencephalogr Clin Neurophysiol 1992;84:315-320. Hill, 1994, pp 107-111.
326. Robinson LR, Slimp JC, Anderson PA, Stolov WC: The efficacy of femoral 354. Shimoji K, Kano T: Evoked electrospinogram: Interpretation of origin and ef-
nerve intraoperative somatosensory evoked potentials during surgical treatment fects of anesthetics. In Phillips, MI (ed): Brain Unit Activity During Behavior.
of thoracolumbar fractures. Spine 1993;18:1793-1797. Springfield, IL, Charles C Thomas, 1973, pp 171-190.
327. Rodichok LD, Schwentker MC: Special problems in the operating room. In 355. Shimoji K: Ascending recordings: Update. In Ducker TB, Brown RH (eds):
Russell GB, Rodichok LD (eds): Primer of Intraoperative Monitoring. Boston, Monitoring the Spinal Cord. Berlin, Springer, 1987.
Butterworth-Heinemann, 1995, pp 241-244. 356. Shimoji K, Kurokawa T, Tamaki T, Willis W (eds): Spinal Cord Monitoring and
328. Romstock J, Watanabe E, Schramm J: Variability of epidural SEP from below Electrodiagnosis. New York, Springer, 1988.
and above spinal cord lesions—the effect of the lesion on spinal SEP. In Ducker 357. Slimp JC, Kliot M: Electrophysiological monitoring: Peripheral nerve surgery.
TL, Brown RH (eds): Neurophysiology and Standards of Spinal Cord In Andrews RJ (ed): Intraoperative Neuroprotection. Baltimore, Williams &
Monitoring. New York, Springer, 1988, pp 261-267. Wilkins, 1996, pp 375-392.
328a. Romstock J, Strauss C, Fahlbusch R: Continuous electromyography monitor- 358. Slimp JC: Intraoperative monitoring of nerve repairs. Hand Clinics 2000;16:25-36.
ing of motor cranial nerves during cerebellopontine angle surgery. J Neurosurg 359. Sloan TB, Koht A: Depression of cortical somatosensory evoked potentials by
2000;93:586-593. nitrous oxide. Br J Anaesthesiol 1985;57:849-852.
329. Rosner BS, Clark DL: Neurophysiologic effects of general anesthetics: II. 360. Sloan TB: American Society of Anesthesiologists: Clinical Applications of
Sequential regional actions in the brain. Anesthesiology 1973;39:59-67. Electrophysiologic Monitoring, Refresher Course Lectures, October 1985,
330. Roy EP III, Gutmann L, Riggs JE, et al: Intraoperative somatosensory evoked Lecture 2) 1.
potential monitoring in scoliosis. Clin Orthop 1988;229:94-98. 361. Sloan TB, Ronai AK, Toleikis JR, Koht A: Improvement of intraoperative so-
331. Russ W, Fraedrich G: Intraoperative detection of cerebral ischemia with somatosensory evoked potentials by etomidate. Anesth Analg 1988,67:582.
matosensory cortical evoked potentials during carotid endarterectomy-presenta- 362. Sloan TB, Kimovec MA, Serpico LC: Effects of thiopentone on median nerve
tion of a new method. Thorac Cardiovasc Surg 1984;32:124-126. somatosensory evoked potentials. Br J Anaesth 1989; 63:51-55.
332. Russ W, Fraedrich G, Hehrlein FW, Hempelmann G: Intraoperative somatosen- 363. Sloan TB, Fugina ML, Toleikis JR: Effects of midazolam on median nerve so-
sory evoked potentials as a prognostic factor of neurologic state after carotid matosensory evoked potentials. Br J Anaesth 1990;64:590-593.
endarterectomy. Thorac Cardiovasc Surgeon 1985;33:392-396. 364. Sloan TB: Mild hypothermia alters cortical magnetic evoked potentials. Anesth
333. Russ W, Thiel A, Schwandt HJ, Hempelmann G: Somatosensorisch evozierte Analg 1991;72:S260.
Potentiale unter thiopental und etomidat. Anaesthesist 1986;35:679-685. 365. Sloan TB, Levin D: Raised intracranial pressure alters magnetic cortical evoked
334. Russell GB, Graybeal JM: Direct measurement of nitrous oxide MAC and neu- potentials. J Neurosurg Anesth 1991;3:242.
rologic monitoring in rats during anesthesia under hyperbaric conditions. 366. Sloan TB, Erian R: Effect of atracurium-induced neuromuscular block on corti-
Anesth Analg 1992;75:995-999. cal motor evoked potentials. Anesth Analg 1993;76:979-984.
335. Russell GB: Motor evoked potentials. In Russell GB, Rodichok LD (eds): 367. Sloan TB, Angell D: Differential effect of isoflurane on motor evoked potentials
Primer of Intraoperative Monitoring. Boston, Butterworth-Heinemann, 1995, elicited by transcortical electric or magnetic stimulation. In Jones SS, Boyd S,
pp 159-169. Hetreed M, Smith NJ (eds): Handbook of Spinal Cord Monitoring. Hingham,
336. Russell GB, Rodichok LD (eds): Primer of Intraoperative Monitoring. Boston, MA, Kluwer Academic Publishers, 1993, pp 362-367.
Butterworth-Heinemann, 1995. 368. Sloan TB, Levin D: Etomidate amplifies and depresses transcranial motor
337. Ryan TP, Britt RH: Spinal and cortical somatosensory evoked potential monitor- evoked potentials in the monkey. J Neurosurg Anesth 1993;5:299.
ing during corrective spinal surgery with 108 patients. Spine 1986;11:352-361. 369. Sloan TB, Rogers J, Rogers H: MAC fractions of nitrous oxide and isoflurane
338. Sako K, Nakai H, Kawata Y, et al: Temporary arterial occlusion during anterior are not electrophysiological^ additive in the ketamine anesthetized baboon. J
communicating or anterior cerebral artery aneurysm operation under tibial nerve Neurosurg Anesth 1995;7:314.
somatosensory evoked potential monitoring. Surg Neurol 1998;49:316-323. 370. Sloan TB, Rogers JN: Inhalational anesthesia alters the optimal interstimulus
339. Samra SK, Vanderzant CW, Domer PA, Sackellares JC: Differential effects of interval for multipulse transcranial motor evoked potentials in the baboon. J
isoflurane on human median nerve somatosensory evoked potentials. Neurosurg Anesth 1996;8:346.
Anesthesiology 1987;66:29-35. 371. Sloan TB: Evoked potentials. In Albin, MA (ed): Textbook of Neuroanesthesia
340. Schepens MA, Boezeman EH, Hamerlijnck RP, et al: Somatosensory evoked with Neurosurgical and Neuroscience Perspectives. New York, McGraw-Hill,
potentials during exclusion and reperfusion of critical aortic segments in thora- 1997, pp 221-276.
coabdominal aortic aneurysm surgery. J Cardiac Surg 1994;9:692-702. 372. Sloan TB: Anesthetic effects on electrophysiologic recordings. J Clin
341. Schmid UD, Boll J, Liechti S, et al: Influence of some anesthetic agents on Neurophysiol 1998;15:217-226.
muscle responses to transcranial magnetic cortex stimulation: A pilot study in 373. Small NC: Complication in arthroscopy: the knee and other joints. Arthroscopy
humans. Neurosurgery 1992;30:85-92. 1986;2:253-258.
342. Schramm J, Watanabe E: Possibilities and limits of intraoperative functional as- 374. Spetzler RF, Selman WR, Nash CL Jr, Brown RH: Transoral microsurgical
sessment of the spinal cord. Klinische Wochenschrift 1988;66(Suppl XIV): odontoid resection and spinal cord monitoring. Spine 1979;4:506-510.
19-26. 375. Spielholz Nl, Benjamin MV, Engler GL, Ransohoff J: Somatosensory evoked
343. Schramm J, Koht A, Schmidt G, et al: Surgical and electrophysiological obser- potentials during decompression and stabilization of the spine. Spine
vations during clipping of 134 aneurysms with evoked potential monitoring. 1979;4:500-505.
Neurosurgery 1990;26:61 - 7 0 . 376. Spielholz, Nl: Intraoperative use of somatosensory evoked potentials. In
344. Schramm J, Moller AR (eds): Intraoperative Neurophysiologic Monitoring in Dumitru D (ed): Clinical Electrophysiology. Philadelphia, Hanley & Belfus,
Neurosurgery. New York, Springer, 1991. 1989, pp 823-840.
345. Schubert A, Licina MG, Lineberry PJ: The effect of ketamine on human so- 377. Spielholz Nl: Intraoperative monitoring using somato-sensory evoked poten-
matosensory evoked potentials and its modification by nitrous oxide. tials: A brief overview. Electromyogr Clin Neurophysiol 1994;34:29-34.
Anesthesiology 1990;72:33-39. 378. Spielholz, Nl: Intraoperative monitoring of the spinal cord and selected periph-
346. Schwartz DM, Drummond DS, Hahn M, et al: Prevention of positional brachial eral nerves. In Dumitru D (Ed): Electrodiagnostic Medicine. Philadelphia,
plexopathy during surgical correction of scoliosis. J Spinal Disord 2000; 13: Hanley & Belfus, 1995.
178-182. 379. Stephen JP, Sullivan MR, Hicks RG, et al: Cotrel-Dubousset instrumentation in
347. Schweiger H, Kamp HD, Dinkel M: Somatosensory-evoked potentials during children using simultaneous motor and somatosensory evoked potential moni-
carotid artery surgery: Experience in 400 operations. Surgery 1991; 109:602-609. toring. Spine 1996;21:2450-2457.
348. Schwentker MC, Forney DJ, Gieski R, Winters JI: Technical standards and 380. Stockard JJ, Sharbrough FW, Tinker JA: Effects of hypothermia on the human
techniques for basic electroencephalography. In Russell GB, Rodichok LD brainstem auditory response. Ann Neurol 1978;3:368-370.
(eds): Primer of Intraoperative Monitoring. Boston, Butterworth-Heinemann, 381. Stone RG, Weeks LE, Hajdu M, Stinchfield FE: Evaluation of sciatic nerve
1995, p 56. compromise during total hip arthroplasty. Clin Orthop 1985;201:26-31.
349. Seal D. Balaton J, Coupland SG, et al: Somatosensory evoked potential moni- 382. Strauss C, Fahlbusch R, Nimsky C, Cedzich C: Monitoring of visual evoked po-
toring during cardiac surgery: An examination of brachial plexus dysfunction. J tentials during para- and suprasellar procedures. In Loftus CM, Traynelis VC
Cardiothorac Vase Anesth 1997; 11:187-191. (eds): Intraoperative Monitoring Techniques in Neurosurgery. New York.
350. Seddon H: Surgical Disorders of the Peripheral Nerves, 2nd ed. London. McGraw-Hill, 1994, pp 135-140.
Churchill Livingstone, 1972. 383. Stuhmeier K-D, Grabitz K, Mainzer B, et al: Use of the electrospinogram for
351. Sekiya T, M0ller AR: Cochlear nerve injuries caused by cerebellopontine angle predicting harmful spinal cord ischemia during repair of thoracic or thoracoab-
manipulations. J Neurosurg 1987;67:244-249. dominal aortic aneurysms. Anesthesiology 1993;79:1170—1176.
352. Sharbrough F, Messick J, Sundt TM: Correlation of continuous electroen- 384. Sugioka H: Evoked potentials in the investigation of traumatic lesions of the pe-
cephalograms with cerebral blood flow measurements during carotid en- ripheral nerve and the brachial plexus. Clin Orthop 1984;184:85-92.
darterectomy. Stroke 1973;4:674-683. 385. Sunderland S: Nerves and Nerve Injuries, London, E&S Livingstone, 1968, pp
353. Shiiya N, Yasuda K, Matsui Y, et al: Spinal cord protection during thoracoab- 144-149.
dominal aortic aneurysm repair: Results of selective reconstruction of the criti- 386. Sundt TM, Sharbrough FW, Anderson RE, Michenfelder JD: Cerebral blood
cal segmental arteries guided by evoked spinal cord potential monitoring. J Vase flow measurements and electroencephalograms during carotid endarterectomy.
Surg 1995:21:970-975. J Neurosurg 1974;41:310-320.
387. Sundt TM Jr, Sharbrough FW, Piepgras DG, et al: Correlation of cerebral blood 414. Waring MD: Electrically evoked auditory brainstem response monitoring of au-
flow and electroencephalographic changes during carotid endarterectomy with ditory brainstem implant integrity during facial nerve tumor surgery. Laryngos-
results of surgery and hemodynamics of cerebral ischemia. Mayo Clin Proc cope 1992;102:1293-1295.
1981;56:533-543. 415. Wassmann H, Fischdick G, Jain KK: Cerebral protection during carotid en-
388. Sundt TM Jr: The ischemic tolerance of neural tissue and the need for monitor- darterectomy—EEG monitoring as a guide to the use of intraluminal shunts.
ing and selective shunting during carotid endarterectomy. Stroke 1983; 14: Acta Neurochir 1984;71:99-108!
93-98. 416. Watanabe E, Schramm J, Romstock J: Cortical and spinal intraoperative record-
389. Svensson LG, Patel V, Robinson MF, et al: Influence of preservation or perfu- ings in uneventful monitoring and in cases with neurological changes. In
sion of intraoperatively identified spinal cord blood supply on spinal motor Ducker TL, Brown RH (eds). Neurophysiology and Standards of Spinal Cord
evoked potentials and paraplegia after aortic surgery. J Vase Surg 1991; 13: Monitoring. New York, Springer, 1988, pp 226-233.
355-365. 417. Watt JWH, Fraser MH, Soni BM, et al: Total i.v. anaesthesia for transcranial
390. Symon L, Crockard HA, Dorsch NW, et al: Local cerebral blood flow and vas- magnetic evoked potential spinal cord monitoring. Br J Anaesth 1996;76:
cular reactivity in a chronic stable stroke in baboons. Stroke 1975;6:482-492. 870-871.
391. Symon L: The relationship between CBF, evoked potentials and the clinical fea- 418. Weight FF, Erulkar SD: Synaptic transmission and effects of temperature at the
tures in cerebral ischaemia. Proc 23rd Scand Neurol Congress, Acta Neurol squid giant synapse. Nature 1976;261:720-722.
Scand Suppl 1980;62:175-190. 419. Welch WC, Rose RD, Balzer JR, Jacobs GB: Evaluation with evoked and spon-
392. Symon L, Wang AD, Costa E Silva IE, Gentili F: Perioperative use of so- taneous electromyography during lumbar instrumentation: A prospective study.
matosensory evoked responses in aneurysm surgery. J Neurosurg 1984;60: J Neurosur 1997;87:397-402.
269-275. 420. Whittemore AD: Carotid endarterectomy and alternative approach. Arch Surg
393. Symon L, Momma F, MurotaT: Assessment of reversible cerebral ischaemia in 1980;115:940-942.
man: intraoperative monitoring of the somatosensory evoked response. Acta 421. Whittle IR, Johnston IH, Besser M: Spinal cord monitoring during surgery by
Neurochir Suppl 1988;42:3-7. direct recording of somatosensory evoked potentials: Technical note. J Neuro-
394. Symon L, Murota T: Intraoperative monitoring of somatosensory evoked poten- surg 1984;60:440-443.
tials during intracranial vascular surgery. In Desmedt JE (ed): Neuromonitoring 422. Whittle IR, Johnston IH, Besser M, et al: Intraoperative spinal cord monitoring
for Surgery. Amsterdam, Elsevier, 1989, pp 263-302. during surgery for scoliosis using somatosensory evoked potentials. Aust NZ J
395. Tamaki T, Tsuji H, Inoue S, Kobayashi H: The prevention of iatrogenic spinal Surg 1984;54:553-557.
cord injury utilizing the evoked spinal cord potential. Int Orthop 1981;4: 423. Whittle IR, Johnston IH, Besser M: Recording of spinal somatosensory evoked
313-317. potentials for intraoperative spinal cord monitoring. J Neurosurg 1986;64:
396. Tamaki T, Noguchi T, Takano H, et al: Spinal cord monitoring as a clinical uti- 601-612.
lization of the spinal evoked potential. Clin Orthop 1984;184:58-64. 424. Williams LV: Automatic detection of intraoperative neurologic injury. IEEE
397. Tamaki T. Spinal cord monitoring with spinal potentials evoked by direct stimu- Trans Biomed Engineering 1999;46(4):481-483.
lation of the spinal cord. In Desmedt JE (ed): Neuromonitoring for Surgery. 425. Wilber RG, Thompson GH, Shaffer JW, et al: Postoperative neurological
Amsterdam, Elsevier, 1989, pp 139-149. deficits in segmental spinal instrumentation: A study using spinal cord monitor-
398. Taniguchi M, Nadstawek J, Langenbach U, et al: Effects of four intravenous ing. J Bone Joint Surg [Am] 1984:66:1178-1187.
anesthetic agents on motor evoked potentials elicited by magnetic transcranial 426. Wilson GJ, Rebeyka IM, Coles JG, et al: Loss of the somatosensory evoked re-
stimulation. Neurosurgery 1993;33:407-415. sponse as an indicator of reversible cerebral ischemia during hypothermic, low-
399. Tator CH, Nedzelski JM: Preservation of hearing in patients undergoing exci- flow cardiopulmonary bypass. Ann Thorac Surg 1988;45:206-209.
sion of acoustic neuromas and other cerebellopontine angle tumors. J 427. Winters WD, Mori K, Spooner CE, Bauer RO: The neurophysiology of anesthe-
Neurosurg 1985;63:168-174. sia. Anesthesiology 1967;28:65-80.
400. Taylor BA, Fennelly ME, Taylor A, Farrell J: Temporal summation—the key to 428. Winters WD: Effects of drugs on the electrical activity of the brain: Anesthet-
motor evoked potential spinal cord monitoring in humans. J Neurol Neurosurg ics. Ann Rev Pharm Toxicol 1976;16:413-426.
Psychiatry 1993;56:104-106. 429. Wissel H, Nebelung W, Awiszus F: Intraoperative monitoring of common per-
401. Terzis JK, Dykes RW, Hakstain RW: Electrophysiological recordings in periph- oneal nerve function during surgery on the knee joint. Biomedizinische Technik
eral nerve surgery: A review. J Hand Surg 1976; 1:52-66. 1998;43:326-329.
402. Therapeutics and Technology Assessment Subcommittee of the American 430. Witzmann A, Reisecker F: Somatosensory and auditory evoked potentials mon-
Academy of Neurology: Assessment: Intraoperative neurophysiology. itoring in tumor removal and brainstem surgery. In Desmedt JE (ed):
Neurology 1990;40:1644-1646. Neuromonitoring for Surgery. Amsterdam, Elsevier, 1989, pp 219-241.
403. Thomas DGT, Jones SJ: Dorsal root entry zone lesions (Nashold's procedure) 431. Wolin LR, Massopust LC, Meder RN: Electoretinogram and cortical evoked
in brachial plexus avulsion. Neurosurgery 1984;15:966-968. potentials under hypothermia. Arch Ophthal 1964;72:521-524.
404. Tiberio G, Floriani M, Giulini SM, et al: Monitoring of somatosensory evoked 432. Yamada T, Muroga T, Kimura J: Tourniquet-induced ischemia and somatosen-
potentials during carotid endarterectomy: Relationship with different haemody- sory evoked potentials. Neurology 1981 ;31:1524—529.
namic parameters and clinical outcome. Eur J Vase Surg 1991 ;5:647-653. 433. Yarnamoto N, Takano H, Kitagawa H, et al: Monitoring for spinal cord is-
405. Tiel RL, Happel LT Jr, Kline DG: Nerve action potential recording method and chemia by use of the evoked spinal cord potentials during aortic aneurysm
equipment. Neurosurgery 1996;39:103-109. surgery. J Vase Surg 1994;20:826-833.
406. Toleikis JR, Carlvin AO, Shapiro DE, Schafer MF: The use of dermatomal 434. Yang LH, Lin SM, Lee WY: Intraoperative transcranial electrical motor evoked
evoked responses during surgical procedures that use intrapedicular fixation of potential monitoring during spinal surgery under intravenous ketamine or eto-
the lumbosacral spine. Spinel 993; 18:2401-2407. midate anaesthesia. Acta Neurochir (Wien) 1994; 127:191-198.
407. Trojaborg W, Boyson G: Relationship between EEG, regional cerebral blood 435. York DH, Pulliam MW, Rosenfeld JG, Watts C: Relationship between visual
flow and internal carotid artery pressure during carotid endarterectomy. evoked potentials and intracranial pressure. J Neurosurg 1981;55:909-916.
Electroencephalogr Clin Neurophysiol 1973;34:61-69. 436. York DH, Chabot RJ, Gaines RW: Response variability of somatosensory
408. Tsai R Y-C, Yang R-S, Nuwer MR, et al: Intraoperative dermatomal evoked po- evoked potentials during scoliosis surgery. Spine 1987;12:864-876.
tential monitoring fails to predict outcome from lumbar decompression surgery. 437. Young W, Mollin D. Intraoperative somatosensory evoked potentials monitor-
Spine 1997;22:1970-1975. ing of spinal surgery. In Desmedt JE (ed): Neuromonitoring for Surgery.
409. Tsuyama N, Tsuzuki N, Kurokawa T, Imai T: Clinical applications of spinal Amsterdam, Elsevier, 1989, pp 165-173.
cord action potential measurement. Int Orthop 1978;2:39-46. 438. Zampella E, Morawetz RB, McDowell HA, et al: The importance of cerebral is-
410. Ubags LH, Kalkman CJ, Been HD, Drummond JC: The use of a circumferential chemia during carotid endarterectomy. Neurosurgery 1991;29:727-731.
cathode improves amplitude of intraoperative electrical transcranial myogenic 439. Zentner J: Scalp recorded somatosensory evoked potentials in response to cauda
motor evoked responses. Anesth Analg 1996:82:1011-1014. equina stimulation in neurosurgical operations on the spinal cord. Br J
411. Van Beek A, Hubble B, Kinkead L, et al: Clinical use of nerve stimulation and Neurosurg 1989;3:39-44.
recording techniques. Plast Recontr Surg 1983:71:225-240. 440. Zentner J: Noninvasive motor evoked potential monitoring during neurosurgical
412. Vauzelle C, Stagnara P, Jouvinroux P: Functional monitoring of spinal cord ac- operations on the spinal cord. Neurosurgery 1989; 24:709-712.
tivity during spinal surgery. Clin Orthop 1973;93:173-178. 441. Zentner J: Motor evoked potential monitoring during neurosurgical operations
413. Veilleux M, Daube JR, Cucchiara RF: Monitoring of cortical evoked potentials on the spinal cord. Neurosurg Rev 1991; 14:29-36.
during surgical procedures on the cervical spine. Mayo Clin Proc 1987;62: 442. Zentner J, Albrecht T, Heuser D: Influence of halothane, enflurane, and isoflu-
256-264. " rane on motor evoked potentials. Neurosurgery 1992;31:298-305.
Chapter 13
Chemical Denervation
Joyce R. Grissom, M.D.
CHAPTER OUTLINE
Botulinum Neurotoxins: Physiology and The Role of Needle EMG and Electrical Stimulation in
Pharmacology BTX Therapy
Structure and Activation • Cell Intoxication Technique
• Presynaptic Neurophysiologic Effects • Postsynaptic
Therapeutic Uses of Botulinum Toxin in Specific
Effects • O t h e r Possible Effects • Autonomic Nervous
Disorders
System Effects
Cervical Dystonia • Spasmodic Dysphonia * Voice Tremor and
Stuttering • Cricopharyngeal Dysphagia • Blepharospasm
Commercial Preparations • Hemifacial Spasm • Oromandibular Dystonia • Limb Dys-
Pharmacology of Commercially Available Botulinum Toxin tonia • Nondystonic Tremor * Spasticity • Detrusor-Sphincter
• Resistance and Antibody Production • Contraindications Dyssynergia • Gastrointestinal Disorders • Hyperfunctional
and Side Effects • Toxicity Facial Lines • Hyperhidrosis • Illustrative Case
Chemical denervation using botulinum toxin (BTX) has revo- junction, but it also inhibits ACh release from pre- and postgan-
lutionized treatment for many disorders with excessive muscle glionic nerve endings of the autonomic nervous system. These
45
activity in common. These conditions include movement disor- sites of toxin activity predict the potential indications and side
ders such as dystonia and tremor; spasticity in multiple sclero- effect profiles of BTX as a therapeutic agent.
sis, cerebral palsy or stroke; and disorders of ocular motility, BTX-A is commercially produced and has been FDA-ap-
including strabismus and nystagmus. There is a preliminary proved for human use in the United States since 1984. BTX-
197
body of literature describing the use of BTX in pain syndromes, B is now commercially available.
including fibromyalgia and headache. BTX blocks the release
of the neurotransmitter acetycholine (ACh) from autonomic STRUCTURE AND ACTIVATION
nerve terminals as well as from the neuromuscular junction, and
is being used in novel ways to modulate autonomic functions Clostridial neurotoxins are synthesized as a single inactive
such as excessive or pathologic sweating or gastrointestinal dys- 150 kDa polypeptide chain released by bacterial lysis. Bacterial
motility. or tissue proteases cleave this polypeptide into active neurotox-
ins consisting of one heavy (H) 100 kDa chain and one light (L)
50 kDa chain (Fig. 13-1). BTX folds into three functionally dis-
PHYSIOLOGY A N D PHARMACOLOGY OF tinct domains. The carboxy-terminus of the H chain binds to
BOTULINUM NEUROTOXINS neuronal cells. The amino-terminus of the H chain is implicated
in membrane translocation. The L chain is a zinc-endopeptidase
The bacterium Clostridium botulinum produces seven differ- responsible for intracellular toxic activity. The H and L chains
ent serotypes of toxin (A, B, C, D, E, F, and G). These serotypes are held together by an inter-chain disulfide b o n d . -
54148199 218
are antigenically distinct, but have a common subunit struc- Reduction of this bond inside the neuronal cell activates the L
ture. Serotypes A, B, and E are linked to most cases of botu-
192
chain to express its zinc-endopeptidase activity. 148
lism in humans. Food-borne botulism is caused by ingestion of

undercooked food contaminated by neurotoxin-producing anaer- CELL INTOXICATION
obic spores of Clostridium botulinum. Spores generated in
anaerobic sections of the newborn intestine cause infant botu- There is a four-step process for cell intoxication by BTX: (1)
lism. BTX C has been cultured from wounds in patients with cell binding, (2) internalization, (3) membrane translocation,
wound botulism. These toxins bind to nerve cells and enter the
142
and (4) target modification in the cytosol. BTX receptors are
148
cytosol to interfere with nerve transmission by blocking the exo- located on the motor neuron plasmalemma at the neuromuscu-
cytotic release of ACh. BTX is most potent at the neuromuscular lar junction. Gangliosides, gangliolipids containing sialic acid
479
moieties found in abundance in nerve tissue, were first to be

considered as receptor substances in the neurotoxin binding
event on synaptosomal m e m b r a n e s . More recent studies 112
have suggested sialoglycoproteins as the more likely candi-

dates for toxin-specific receptors. Each serotype is believed 9
to have a unique receptor. Synaptogamin, a neuronal mem- 45
brane-associated protein has been proposed as the receptor for

BTX-B. 153
After binding to the nerve membrane there is a time interval

that precedes the onset of nerve transmission blockade. During
this period the toxin is screened from the action of anti-toxin,
and internalization of the toxin by receptor-mediated endocyto-
sis occurs. BTX inaccessibility to anti-toxin has a half-time
193
of approximately 5 minutes, an order of magnitude greater than

the rate at which transmission fails. This difference suggests
the presence of an intervening step between translocation
across the nerve membrane and the final lytic step resulting in
paralysis. After internalization, the L chain packaged in en-
192
dosomes must be moved across the endosome membrane into

the cytosol. The amino-terminus of the H chain is thought to
form transmembrane ion channels at an acidic pH allowing for
passage of the L chain and at least part of the H chain into the
cytoplasm. ' - ' - 16 75 105 179 188
The apparatus mediating exocytosis in the motor nerve termi-

nal consists of several protein components. The toxic intracellu-
lar activity of BTX is mediated by zinc-dependent specific
proteases targeting components of this apparatus. Different
FIGURE 13-1. Structure-function relationships in clostridial
BTX serotypes cleave the different components (Fig. 13-2).
neurotoxins.The 50-kDa carboxy-terminal domain of the H-chain is
BTX B, D, F, and G specifically recognize and cleave vesicle-
mainly responsible for specific binding t o receptors on the presynaptic
associated membrane protein (VAMP)/synapatobrevin.
membrane (top left panel). Binding is followed by internalization of the
BTX-A and E cause specific hydrolysis of SNAP-25 (synapto-
toxin-receptor complex inside vesicles.The amino-terminal domain of
somal-associated protein of 25 kDa). BTX-C cleaves syn-
the H-chain is thought t o be involved in t h e membrane translocation
taxin. BTX does not cleave the short polypeptides containing
148
of the L chain into t h e cytoplasm (top right).The L chain is freed in the

the cleavage site of the target proteins. Only long polypeptides
cytosol by t h e reduction of the interchain disulfide bond and can dis-
are cleaved, suggesting that the proteases recognize other seg-
play zinc-endopeptidase activity (lower panel). (From Montecucco C,
ments and/or require a specific amino acid conformation in the
Schiavo G: Mechanisms of action of tetanus and botulinum neurotox-
vicinity of the cleavage site that occurs only in larger polypep-
ins. Mol Microbiol 1994; 13:1-8, with permission.)
tides (Table 13-1). 189
PRESYNAPTIC NEUROPHYSIOLOGIC EFFECTS

BTX interferes with the quantal release of ACh. The quantal
content of the response to nerve stimulation, the endplate poten-
tial (EPP), is reduced, as are spontaneous miniature endplate
potentials (MEPPs) of quantal size. - Subminiature EPPs of 52 146
non-quantal size increase in frequency after BTX exposure. 66
Although BTX causes no direct structural effect on motor axons

or nerve terminals, it produces a change similar to that seen fol-
lowing nerve sectioning. This change includes bouton loss, de-
crease in and retraction of dendritic profiles, and increased
numbers of astrocytes. In addition to temporary blockade of
198
ACh release, BTX induces "overgrowth" of motor nerve end-

ings from the original terminal arborizations. Sprouting of 63
axon terminals is seen as early as 24 hours after treatment, and

continues even after functional reestablishment of the neuro-
SYNAPTIC CLEFT muscular junction (NMJ) has o c c u r r e d . Sprouting begins 546158
from the nodes of Ranvier of the preterminal motor axon, from

FIGURE 13-2. Schematic drawing of the binding of the terminal axon just proximal to the end plate and from the ar-
clostridial neurotoxins. Critical sequences of syntaxin and SNAP- borization over the endplate(Fig. 13-3). Growth of the new 106
25 involved in t h e binding of t h e positions of cleavage by different terminal axon sprouts proceeds preferentially along the longitu-
clostridial neurotoxins. (From Montecucco C, Schiavo G: Mechanisms dinal axis of muscle fibers. - New NMJs are formed and may 106 146
of action of tetanus and botulinum neurotoxins. Mol Microbiol be near the original NMJ or at some distance from it. New NMJ
1994; 13:1 - 8 , with permission.) formation enlarges the endplate region and results in muscle
Chapter 13 CHEMICAL DENERVATION — 481
TABLE 13-1. Botulinum Neurotoxins

Toxin Serotype Cellular Substrate Target Cleavage Site Cell Target Localization
BTX-A SNAP-25 Near C-terminus Neuron Presynaptic plasma membrane;
Glnl97-Argl98 possibly other regions
BTX-B VAMP/synaptobrevin Gln76-Phe77 Neuron Synaptic vesicle
BTX-C Syntaxin IA, IB Lys253-Ala254 Neuron Presynaptic plasma membrane;
Lys252-Ala253 ? + other regions
BTX-D SNAP 25 Lys59-Leu60 Neuron
VAMP/synaptobrevin Ala7-Asp68 Neuron Synaptic vesicle
Cellubrevin Unknown All cells Vesicles of endocytosing/
recycling system
BTX-E • SNAP-25 Argl08-llel8l Neuron Synaptic plasma membrane;
possibly other regions
BTX-F VAMP/synaptobrevin Gln58-Lys59 Neuron Synaptic vesicles
Cellubrevin Unknown All cells Vesicles of endocytosing/
recycling system
BTX-G VAMP/synaptobrevin Ala8l-Ala82 Neuron
VAMP, vesicle-associated membrane protein; SNAP-25, synaptosomal-associated protein of 25 kDa
From Brin MF: Botulinum toxin: Chemistry, pharmacology, toxicity, and immunology. Muscle Nerve 1997;20:S 146-S168, with permission.
fibers possessing multiple endplates that may be innervated by turns/amplitude ratio both increase after injection with BTX on
more than one axon. The nodal and terminal axon sprouts may
1
quantitiative electromyography (EMG). 77
be present for as long as 3 years following therapeutic use of

B T X - A . This axonal sprouting can be prevented by direct OTHER POSSIBLE EFFECTS
106
electrical stimulation of the muscle supplied by the BTX-ex-

posed nerve. 34
There is no question that BTX presynaptically blocks ACh
release at the neuromuscular junction; however, some clinical
POSTSYNAPTIC EFFECTS observations pertaining to therapeutic use of BTX for dystonia
are difficult to explain by BTX action at presynaptic motor ter-
Cholinergic transmission exerts trophic effects on striated minals alone. Clinical benefit following intramuscular injection
muscle and is necessary for normal muscle enzymatic activity is delayed by several days and increases over a period of weeks
and differentiation. Enzymes of energy metabolism are de- while a decrease in MEPPs can be demonstrated within a few
creased in BTX-treated muscle. Visually, the muscle size and h o u r s . Dystonic spasms are reduced in both frequency and
162
the color difference between slow and fast muscles become less strength following treatment with BTX. Clinical weakness cor-
apparent. - Type I (slow-twitch) fibers are characteristically
39 60
relates only poorly with improvement in dystonic activity and
rich in succinate dehydrogenase and poor in phosphorylase.
Type II (fast-twitch) fibers stain stongly for phosphorylase and
weakly for succinate dehydrogenase. - 57 129
Histochemically, after injection with BTX, each type of fiber

shows a greater decrease in the enzyme activity normally pre-
dominant in that fiber type. Atrophy involves all fiber types, but
is faster in onset and recovery in predominantly type I mus-
c ] e64.65.s94 Light and electron-microscopy show reversible
s
changes in all fiber types. These changes include myofilament

loss with reduction in cross-sectional area, dense body forma-
tion, lysosome accumulation, and formation of parallel tubule
clusters of sarcoplasmic reticulum origin. Levels of ACh at the
NMJ fall by 5 0 - 6 3 % following BTX administration. - A 61 202
comparision study of the histologic appearance of orbicularis

oculi muscle from myectomized blepharospasm patients treated
with BTX to orbicularis oculi muscle from untreated patients
found no consistent, long-lasting alterations in muscle fiber
morphology. 97
Electrophysiological^ single-fiber studies show increased FIGURE IBS. Axonal sprouting after BTX. Some sprouts (S)
jitter and blocking that is less marked with higher rates of motor from the preterminal axon (PTA) end in small dilations, presumed t o
unit discharge. Abnormalities in jitter can be demonstrated in
176
be the growth cone of the axon, in a muscle after 9 BTX injections (x
muscles distant to those injected even with the small doses used 250, SCI using an antibody t o PO). (From Borodic GE, Ferrante RJ,
for the treatment of hemifacial spasm. Fibrillations and posi-
85
Pearce LB, Alderson K: Pharmacology and histology of botulinum
tive sharp waves similar to those seen following surgical dener- toxin. In Jankovic J, Hallett M (eds):Therapy with Botulinum Toxin. N e w
vation are seen in muscles injected with B T X . Turns and the 168
York, Marcel Dekker, 1994, pp I 19-157, with permission.)
clinical benefit may persist for weeks or months after the reso- hemifacial spasm demonstrated mild abnormalities of cardio-
lution of localized weakness. These observations have
83
vascular reflexes as a distant effect.
85
prompted investigators to hypothesize that alteration in afferent

input mediated peripherally by BTX results in secondary cen-
tral reorganization. - - COMMERCIAL PREPARATIONS
37 83 84
BTX has direct effects on the intrafusal fibers of gamma mo-

torneurons in muscle s p i n d l e s . Physiologic changes have PHARMACOLOGY OF COMMERCIALLY AVAILABLE
74172
been described in the spinally mediated reciprocal inhibition of

BOTULINUM TOXIN
patients with arm dystonia treated with B T X . Patients with 167
upper limb dystonia demonstrate a decreased second phase of Dr. Hermann Somner first attempted to purify BTX-A in the
reciprocal inhibition as studied by conditioning the H-reflex in 1920s, but Dr. Carl Lammanna first chrystallized BTX-A bind-
forearm flexors with a radial nerve stimulus. Treatment with ing toxic units to nontoxic proteins in 1946. The 900-kD
197
BTX increased the second phase of reciprocal inhibition. There chrystallized form of BTX is used therapeutically today. This
was no change in the initial phase of reciprocal inhibition, form includes both the neurotoxin and nontoxic proteins, some
which is normal compared to controls. The authors suggest that of which agglutinate red blood cells (hemagglutinins). Only
this concurrent indirect effect of spinal inhibition may be medi- 20% of the protein in the chrystalline preparation is neuro-
ated through the effect of BTX on the intrafusal neuromuscular toxic. The toxin is produced through a process of anaerobic
53
junction. More recent work with repetitive transcortical mag- fermentation under highly controlled conditions. Botox, a com-
netic stimulation of the motor cortex has demonstrated that the mercially produced BTX-A product is FDA-approved for use in
intracortical deficiency in inhibition seen in dystonia is also the United States. It is derived from the Hall strain of
transiently normalized following treatment_with BTX. - 37 84
Clostridium botulinum in medium containing yeast and N - Z
The amplitude of the precentral P22/N30 median somatosen- amine. Toxin is precpitated and harvested by centrifugation.
128
sory evoked potential in frontal electrodes contralateral to the Crude toxin is purified by sequential precipitation and ion-ex-
direction of head movement is significantly higher than in change chromatography. The product is periodically assayed for
frontal electrodes ipsilateral to the direction of head movement toxin activity and purity and is ultimately mixed with serum al-
in patients with cervical dystonia. This side-to-side amplitude bumin and lyophilized. Lyophylization and filtration cause
86
difference is not found in normal s u b j e c t s . Like motor dis-

113114
some damage to the toxin structure, reducing the toxicity of the
inhibition, these somatosensory evoked potential abnormalities final commercial product. 177
in patients with cervical dystonia normalize jrfterjxeatment with Differences in production methods account for the differ-
BTX. The reduction in the contralateral P22/N30 may reflect ences in activity and weight of BTX produced by different com-
the change in the direction of head rotation, and the pattern and panies. Dysport is the BTX-A product produced commercially
force of muscle contraction after treatment, but a reduction in in the United Kingdom. The unit of measurement for BTX is
the abnormally high level of excitability of cortex by treatment the mouse unit (U). This is a unit of potency determined by
with BTX cannot be excluded. 114
mouse assay according to published specifications. One unit of
Some animal studies suggest that intramuscularly injected BTX-A is equal to the amount of toxin that will kill 5 0 % of a
BTX can reach the CNS. Iodine-labeled BTX-A injected into the group of 18-20 female Swiss-Webster mice. - A vial of Botox
29 98
gastrocnemius muscle of a cat produced histologic evidence of contains 100 U. A vial of Dysport contains 500 U. By definition
accumulated radioactivity in the ipsilateral spinal cord 3-4 days it would follow that one unit of Botox and one unit of Dysport
following injection, suggesting retrograde axonal transport of should have equivalent clinical effect; however, Botox contains
the toxin takes place. Despite the finding that BTX is bound
210
greater quantities of detoxified toxin than Dysport. As a result
177
by synaptosomes in the rat brain, there is at this time no evidence 1 nanogram (ng) of Botox equals 2.5 U whereas 1 ng of Dysport
that BTX has a physiologic effect on the human brain. - - 15 83 153
equals 40 U . Clinically, a bioequivalent ratio of one Botox
169
unit to three Dysport units has been recommended. 137155
AUTONOMIC NERVOUS SYSTEM EFFECTS In 1997 the FDA approved a new neurotoxin complex pre-
pared from a new bulk toxin source. This toxin has less toxicity
Autonomic symptoms are also characteristic of clinical botu- by weight per biologically active unit than the original Botox.
lism. Autonomic symptoms include constipation, nausea, vom- The decreased neurotoxin protein per effective dose appears to
iting, abdominal cramps, heartburn, blurred vision, dry mouth, produce a similar dose-related weakness with decreased neu-
micturational disturbance, and sexual dysfunction. In the au- 110
tralizing antibody formation tested in an animal model. This 4
tonomic nervous system BTX blocks ganglionic nerve endings, may have important implications for recommended dosing para-
post-ganglionic parasympathetic nerve endings and some post- digms targeted at minimizing risk of immunoresistance.
ganglionic sympathetic nerve e n d i n g s . A much higher
8 5 1 3 5 1 7 0
The lyophilized toxin must be reconstituted with preserva-
threshold level of electrical stimulation of parasympathetic tive-free sterile saline (Table 13-2). The reconstituted toxin is
nerves (vagus, chorda tympani, nervus erigens, and oculumotor) stable at room temperature for up to 4 hours. BTX-A is not
86
is required to produce a physiologic effect in animals intoxi- flammable or volatile, but spills may result in hazardous mists
cated with BTX compared to control animals. BTX blocks 59
or aerosols. Spills may be inactivated by 0.5% hypochlorite so-
contraction of isolated guinea pig vas deferens in response to lution. Toxin-containing solutions are also inactivated by heat-
hypogastric nerve stimulation and piloerection in the tail skin ing, boiling, or autoclaving to 121 °C for 20 minutes. BTX-A 178
obtained from cats. These responses were restored with direct

170
was approved by the FDA in the United States for treatment of
introduction of noradrenaline. There is experimental evidence strabismus, blepharospasm, and hemifacial spasm in 1 9 8 9 . 30127
in animals that BTX may additionally block nonadrenergic, at- For patients who become resistant to BTX-A an effective, anti-
ropine-resistant autonomic neuromuscular transmission in genically distinct alternative serotype could be a welcome alter-
some instances. Four of five patients injected with BTX for
135
native to phenol injections, rhizotomy, or brain surgery.
C h a p t e r 13 CHEMICAL DENERVATION — 483
T A B L E 13-2. B o t o x Dilution T a b l e have circulating antibodies than patients who continue to enjoy
ml Normal Saline/100 U Botox Vial Dilution a good clinical response. - Antibodies to BTX-A do not block
95 96
the effects of BTX-B and vice versa. 151
1 I0U/0.I ml
There are several assays available to detect the presence of
2 5 U/0.1 ml antibodies in serum. The in vivo mouse neutralization assay is
4 2.5 U/0.1 ml widely used. - In this regard, BTX-A is titrated with the serum
29 98
of the patient suspected of having antibodies and then injected

8 1.25 U/0.1 ml
into mice. If antibodies are present, they bind to the toxin and
10 1 U/0.1 ml protect the mouse from its lethal effects. If antibodies are
From Botox® package insert, with permission. absent, or present in low quantity, the mouse dies. A positive
mouse assay correlates well with lack of clinical response. A
BTX-B cleaves synaptobrevin, a vesicle-associated mem- negative response does not exclude the presence of antibody
brane protein. Unlike BTX-A, the initiation of ACh release
180
levels sufficient to cause human resistance, but insufficient to
by BTX-B is not blocked by aminopyridines, which increase protect the mouse. The mouse assay, therefore, underestimates
impulse-evoked C a by blocking presynaptic potassium chan-
2+
the presence of clinically significant immunoresistance. 29
nels. BTX-B is now commercially available in the United

7
Several immunoabsorbent assays have been developed. The cor-
States. This liquid preparation must be refrigerated but does relation between these assays and clinical resistance has not
not require reconstitution. The clinical effects are similar to
32
been established. The enzyme-linked immunosorbent assay
those produced by BTX-A, but are shorter-lived. The potency (ELISA) is positive in more than half of all patients treated with
of BTX-B is also determined by mouse assay. As with Botox BTX, including those who are still clinically responsive to treat-
and Dysport, the clinical effects of one unit are not equivalent. ment. It is possible that these assays detect antibodies to func-
190
In a double-blind placebo-controlled study BTX-B in doses of tionally unimportant parts of the toxin. Additionally, cross
2500-10,000 U in patients with cervical dystonia produced ef- reactivity between type A and B BTX occur in systems using
fects comparable to 150-200 U of Botox. Patients with and
127
polyclonal antibody reagents. 98
without resistance to BTX-A had similar responses. In a multi- It is also possible to clinically assay for BTX activity. In the
center, randomized, double-blind, placebo-controlled clinical frontalis type A test (F-TAT) 15 U of Botox are injected unilat-
study of 77 patients resistant to BTX-A, results of treatment erally in the frontalis muscle at 2 sites. If within 2 weeks the pa-
with 10,000 U of BTX-B were significantly better than for patient is unable to wrinkle the forehead on the injected side, he or
tients receiving placebo. Efficacy was determined by im-
32
she is "not resistant." These patients may desire to have the
provement on the Toronto Western Spasmodic Torticollis other frontalis muscle injected subsequently to maintain facial
Rating Scale (TWSTRS) and visual analog scales that assess symmetry. If the injected frontalis muscle moves normally, the
global patient benefit and pain. In this study, the overall esti- patient is deemed clinically "resistant." 29
mated duration of treatment effect was 12-16 weeks. Clinically There appears to be a correlation between the clinical and
relevant adverse events included dry mouth (17/39) and dys- electrodiagnostic results of BTX injection and the antibody
phagia (11/39). Dysphagia was self-limited and tolerable in af- status of patients suspected of antibody-mediated resistance.
fected study subjects. In a similarly designed study, 109 The amplitude of the peroneal CMAP recorded from the exten-
BTX-A-responsive patients were randomized to treatment with sor digitorum brevis (EDB) muscle was determined before and
placebo, 5,000 U BTX-B, or 10,000 U BTX-B. Efficacy signif- 3-5 weeks after BTX-A injection with 200 U of Dysport. In
icantly exceeding that of placebo was demonstrated with both serum-positive nonresponders the C M A P amplitude was un-
the 5,000 U and 10,000 U doses. Clinical benefit was greater changed 4 weeks post-injection. In serum-negative nonrespon-
with the 10,000 U dose. Duration of effect and adverse events ders a marked decrease in the C M A P amplitude occurred,
were similar in the BTX-A responsive study to those seen in indicating that the toxin remained capable of activity at the neu-
the BTX-A resistant study. BTX-C and BTX-F have also
27
romuscular junction in these patients. The lack of BTX efficacy
been studied as treatments for cervical d y s t o n i a . - The 70 91107
with regard to the dystonia being treated was likely related to an
clinical effects of BTX-F are not as long-lived as those pro- inadequate dose or selection of suboptimal injection sites in
duced by serotype and last approximately 1 month. Peroneal 29
these patients. Patients demonstrated to be nonresponsive by
119
CMAP amplitudes studied serially over 90 days in study subclinical assays such as the frontalis or EDB tests may benefit
jects injected in the extensor digitorum brevis muscle (EDB) from an alternate BTX serotype regardless of seronegativity or
with BTX-A and BTX-C showed a similar effect and ampli- seropositivity on immunologic assays.
tude profile. 70
Short intervals between injections, higher doses, and the anti-

genic load per dose appear to be related to an increased risk for
RESISTANCE AND ANTIBODY PRODUCTION development of antibodies. To minimize the risk of immunore-
92
sistance it is recommended that the smallest effective dose be

A small percentage of patients receiving therapeutic BTX used at the longest possible interval. Ideally injections should
become clinically resistant. Injections no longer have clinical be 3 or more months apart, and booster injections (defined as
efficacy and injected muscles do not become weak or develop reinjection within 3 weeks of an injection) should be avoided.
atrophy. Resistance is generally thought to be due to the devel- These precautions are of particular importance for doses in
opment of antibodies to the toxin. Antibody-mediated resistance excess of 100 U per session. - These recommendations have
29 92
occurs in 3-10% of patients. Some antibodies are directed

29
been based on clinical experience with the original Botox.
against the toxin, while others are directed against associated Current Botox preparations appear to be less immunogenic, p o -
nontoxic proteins. Only the antibodies that block the toxic tentially allowing higher doses without subtantial risk of devel-
action at the NMJ are important clinically. Patients who have
109
oping resistance. Figure 13-4 outlines an algorithmic approach
4
lost their clinical response to BTX therapy are more likely to to suspected immunoresistance. Plasma exchange has been used
reported in a small number of patients following remote BTX

Technical injections. -56
This weakness is generally self-limited and
175 1 8 7 2 0 9
Objectives Met? does not necessarily require discontinuation of BTX treatment.

Other reported idiosyncratic reactions include persistent local-
80-
ized rash, localized acute allergic reaction, and recurrent ptosis
following Botox injections in the cervical musculature for torti-
Reassess dose, collis. Neuromuscular transmission and autonomic function
128
muscle selection, can be altered at sites distant from local therapeutic BTX injec-
tecftftJqtM, and
adjunctive therapies
tions. Abnormal jitter can be demonstrated in muscles distant
85
to the site of BTX injection for blepharospasm. Generalized

176
weakness following BTX injection has been reported in one

Any reduction patient with ALS and in another patient with subclinical
In lone?
Lambert-Eaton syndrome. On the other hand, BTX has been
143
NO successfully used for treatment of cervical dystonia without ex-

acerbation of ptosis or weakness in muscles remote to the site of
injection in a patient with myasthenia gravis. BTX should be
73
to therapy used with caution in patients with diseases involving primary

muscle weakness, the NMJ, or motor neurons.
In 1990, the National Institutes of Health Consensus De-
Frontalis Iniectlon Antibody assay velopment Conference determined that the safety of BTX ther-
Review reinjection apy during pregnancy and breastfeeding and chronically during
strategy and childhood were unknown. There are insufficient data to con-
48
timing—modify Implement clude that BTX can be used during pregnancy without any risk
injection technique alternative
if needed and yiS->
treatment
of adverse outcome. Pregnancy, therefore, remains a relative
reassess role of strategies contraindication for BTX treatment. BTX injection should also
adjunctive therapies be avoided in patients with bleeding disorders, or taking antico-
agulant medication, because of the possibility of excessive
bleeding or hematoma formation.
Schedule for
follow-up
reassessment TOXICITY
and consideration
for reinjection
Toxic doses of BTX show variation between species.
Extrapolation from animal studies to humans must be accepted
FIGURE 13-4. Algorithm for evaluation of patients failing to with caution. The lethal dose for humans is unknown.
respond clinically t o injection with BTX-A. (From Brin MF: Maximum dosage recommendations for human clinical applica-
Botulinum toxin: Chemistry, pharmacology, toxicity, and immunology. tion are based on primate studies. The intravenous and intra-
Muscle Nerve l997;20(Suppl 6):SI46—SI68 with permission.) muscular L D in monkeys is between 38-42 U / k g .
5 0
1 0 2 1 8 6
Extrapolation from this primate data to the 70 kg human sug-

sucessfully to restore clinical response to BTX-A in an im- gests a human L D of approximately 3000 U. Most experts rec-
5 0
munoresistant patient severely disabled by cervical dystonia. 152

ommend a maximum dose of 300-400 U per session and a
Use of an alternative serotype is an additional option. maximum of 400 U over any 3-month period. Limited toxicity
29
data is available for intramuscularly injected BTX-B. In normal

CONTRAINDICATIONS AND SIDE EFFECTS monkeys, intramuscular doses up to 480 U/kg total body weight
of BTX-B produced no signs of toxicity. 151
Because of the small quantities of toxin used and its avid

binding to local nerve terminal receptor sites, side effects are
generally local. Like the therapeutic benefit, adverse side effects THE ROLE OF NEEDLE
are temporary. In more than a decade of clinical use, no cases of ELECTROMYOGRAPHY A N D ELECTRICAL
anaphylaxis or deaths have been attributed to BTX-A. STIMULATION IN BTX THERAPY
29
Prospective patients should be prepared for the possibility of

excessive local weakness, the most common side effect of BTX The role of needle EMG in BTX treatment of disorders of ex-
therapy. Depending on the injection site, weakness may result in cessive muscle activity has not been clearly established. EMG
functionally significant impairment such as ptosis or dysphagia. was frequently used as an adjunct to the clinical examination
With larger doses of BTX some patients may complain of non- when BTX was first used to treat cervical dystonia (CD). EMG
specific malaise, myalgia, or flu-like symptoms. These symp- was used both to confirm the pattern of muscle involvement,
toms may occur because some of the BTX can be taken up into and to confirm placement of BTX in targeted muscles. Over
the circulation and have effects on the autonomic nervous time, many experts became comfortable with observation and
system. Generalized pruritus without rash has been infrequently clinical examination alone to select and/or to inject muscles,
reported in patients treated with B T X . Patients receiving in-
68130
using EMG only when patients failed to benefit from clinically
jection into the orbicularis oculi for blepharospasm or hemifa- targeted injections, or when localization of the muscle was
cial spasm may experience a transient increase in intraocular anatomically difficult. In dystonia, selection of muscles for injec-
pressure due to toxin-induced effects on local cholinergic au- tion is clinically based on a number of factors. Patients may iden-
tonomic pathways. Plexopathy or radiculopathy have been
139
tify painful areas likely to be the locus of muscle hyperactivity.
Chapter 13 CHEMICAL DENERVATION — 4 8 5
The position of the resting dystonic limb should be noted.

Patients should be specifically asked not to use any "sensory
tricks" or compensatory measures. Walking, reading, or writing
may bring out or exaggerate the dystonia. In addition to the po-
sition of the dystonic limb, the presence of muscle hypertrophy
or tremor should be noted.
Some investigators continue to use EMG assistance to inject
BTX. More precise targeting of muscles using EMG may theo-
retically have particular value when the targeted muscles are in
close proximity to muscles in which weakness is not desired. In
cervical dystonia, the outcome of BTX treatment using EMG
guidance for injection has been compared to clinically directed
injection in a prospective, blinded study. Although the number
of patients demonstrating improvement was similar in the two
groups, the EMG-guided group had more patients having
marked improvement for the same dose of BTX. Adverse ef-
FIGURE 13-5. Hollow, Telfon-coated needles used for BTX
fects were similar in both groups. Investigators using EMG
47
Injection. A, Plastic hub with built in e l e c t r o d e cable B, Metal hub

guidance for CD in open studies have suggested that EMG guid-
with clip-on electrode cable.
ance results in lower doses of BTX than those cited in the litera-
ture due to improved accuracy of injection into targeted
muscles. 267 7
Speelman and Brans tested the accuracy of EMG rather than a crisp snap suggest suboptimal placement of the
needle placement into a dystonic muscle by an investigator with needle tip near the activated muscle or fascicle. EMG can simi-
9 years of experience with BTX injection for dystonia. 195
larly be used to avoid injection of nontargeted muscles.
Needle placement in the targeted muscle was confirmed only Confirming absence of crisp EMG activation with volitional
4 7 - 8 3 % of the time. Even in large superficial muscles such as contraction of a nontargeted muscle lying adjacent to a targeted
the sternocleidomastoid and the trapezius muscles placement muscle also helps to optimize toxin placement. Once placement
was incorrect 17% of the time. EMG guidance is recommended is confirmed, the syringe should be gently aspirated to ensure
for deep muscles not easily palpated from the surface or for any that the tip of the needle is not in a blood vessel; the desired dose
muscle whose localization is difficult. of toxin may then be injected through the electrode into the
Surface, needle, and wire electrodes have all been used to muscle. It is important to fill the syringe and needle with BTX
assess dystonic patients undergoing BTX treatment. Surface before needle placement to avoid inadvertent reduction in the
and wire electrodes do not permit EMG-directed injection of dose delivered due to the presence of dead space. Confirmation
toxin, and surface electrodes can only be used to study superfi- of needle placement by EMG can be difficult in patients where
cial muscles. Wire electrodes are useful for studying the simul- muscles are synergistically activated together, in children, cogni-
taneous muscle activation patterns of deep muscles in the neck tively impaired adults, or sedated patients who cannot voluntar-
or limbs where needle electrodes would be likely to be dis- ily activate muscles. With such patients, the treating physician
lodged by movement or cause exessive pain if left in the muscle may rely on passive motion of the muscle or may consider invol-
for an extended period of study. Teflon-coated, hollow monopo- untary activation of the muscle using electrical stimulation.
lar needle electrodes permit targeted injection into active dys- Electrical stimulation can activate an entire muscle through
tonic muscles. In addition to Teflon-coated needles, EMG- stimulation of a large nerve (motor nerve stimulation), or small
directed BTX injection requires a standard EMG instrument. fascicles through activation of small motor nerve branches
Teflon-coated needles are manufactured with an attached cable
to connect to the EMG instrument. Others require an alligator or
other customized clip designed to attach to the needle hub to
make this connection (Figs. 13-5 and 13-6). Standard EMG
filter settings are used. The reconstituted toxin is drawn up into
a tuberculin syringe attached to the hub of the hollow needle.
TECHNIQUE
Localization of target muscles using EMG is straightforward.
The needle tip is positioned so that a full recruitment pattern is
seen with activation of the targeted muscle. Motor unit action
potentials (MUAPs) should have normal amplitude and short rise
times making a distinctly crisp sound. It is important to empha-
size that the presence of crisp MUAPs indicates only that the
needle tip is in a contracting muscle fascicle, but does not con-
firm placement in the targeted muscle. Modulation of the EMG
by active contraction or passive movement of the target muscle
is required. Fibrillation potentials and positive sharp waves will
be seen in previously injected muscle after 10-20 days but are
usually not evident after 3 months. The presence of low-am-
168
FIGURE 13-6. EMG instrument setup with active, reference,
plitude, poorly defined units that fire with a muffled thump and ground electrodes.
tonic postures. Pharmacologic therapy may be helpful in some

patients, but is inadequately effective or limited by unaccept-
able side effects in most cervical dystonia patients.
Anticholinergic drugs such as Artane or Cogentin are occasion-
ally effective in dystonia, but often limited by side effects in-
cluding dry mouth, cognitive difficulties, blurred vision, nausea,
and urinary retention. Other medications with potential benefit
include Baclofen, Tegretol, benzodiazepines, Tetrabenazine,
and cyclobenzaprine. 31
FIGURE /3-7. Dystonic head and neck postures. A, Retrocollls. More than 20 pairs of muscles connect the skull to the shoulder
B, Laterocollis. CTorticollis. girdle, the skull to the vertebral column, and the cervical verte-
brae to one another. The head moves in three axes: flexion-exten-
sion; rotation; and lateral tilting. Involuntary turning with rotation
within the belly of the muscle (motor point stimulation). 154
of the chin toward one shoulder is referred to as torticollis. Other
Motor nerve stimulation is used for phenol neurolysis. For BTX dystonic postures may include neck flexion (anterocollis); neck
injection motor point stimulation may be preferred in order to extension (retrocollis), and head tilt with the ear deviated toward
place the needle electrode as close as possible to the endplate the ipsilateral shoulder (laterocollis) (Fig. 13-7). Dystonic cervi-
area. A surface electrode is plugged into the stimulator of an cal postures rarely result from overactivity of a single muscle.
EMG/NCV instrument and is positioned above the muscle/tendon Generally, the movements or postures are the result of complex
junction. The lead from the injection needle is also plugged into combinations of muscle activity. Successful treatment depends on
the stimulator. The EMG/injection needle is inserted into the the identification and injection of the muscles primarily involved
target muscle localized by palpation and passive motion. in producing each component of the observed head movement.
Stimulation is initiated at an intensity just sufficient to produce Six muscles are most frequently targeted for injection with BTX
a visible muscle contraction or twitch. This initial stimulus in- in the treatment of cervical dystonia. These muscles are the stern-
tensity is generally in the 1-3 mA r a n g e . The needle is then
154
ocleidomastoid (SCM), trapezius, splenius capitis, levator scapu-
positioned so as to produce the maximum twitch with the mini- lae, and deep posterior vertebral muscles (longissimus capitis and
mum possible stimulus. Final stimulus intensities are typically semispinalis capitis) (Table 13-3) Torticollis, therefore, may in-
47
between 0.025 and 0.5 mA. 154

volve abnormal activation of the ipsilateral splenius capitis
muscle and the contralateral SCM. Laterocollis may involve ab-
normal activation of the ipsilateral SCM, splenius capitis, scalene
THERAPEUTIC USES OF BOTULINUM complex, levator scapulae, and posterior vertebral muscles.
Retrocollis is produced by bilateral splenius capitis, upper trapez-
T O X I N IN SPECIFIC DISORDERS ius, and/or the deep cervical paraspinous muscles. Bilateral SCM,
scalene, and/or submental complex activation produce anterocol-.
CERVICAL DYSTONIA lis. Shoulder elevation is produced by abnormal activity of the ip-
Cervical dystonia is a focal dystonia involving the neck mus- silateral trapezius and levator scapulae muscles. In assessing a
cles. It may produce repetitive clonic head movements or sustained patient with cervical dystonia, the patient should be asked which
TABLE 13-3. Contributory Muscle Activity in Cervical Dystonia

Muscle Origin Insertion Action
Sternocleidomastoid Manubrium and medial Mastoid Rotation of chin contralaterally,
1/3 of clavicle anterior neck flexion
Trapezius Medial end of superior nuchal line; Scapula, acromion process, and lateral Scapular and shoulder elevation,
external occipital protuberance; third of the clavicle neck extension
ligamentum nuchae and spinous
processes C 7 - T 1 2
Splenius capitis Lower half of t h e ligamentum Mastoid process Ipsilateral rotation of t h e chin,
nuchae and spinous processes neck extension
C7-T5
Levator scapulae Transverse processes of C 1 - C 4 Superior angle of the scapula Shoulder and scapular elevation
Scalene complex Transverse processes of C I - C 7 Superior surface of first rib between Ipsilateral turn with anterocollis
the subclavian groove and the rib
tubercle
Longissimus capitis Transverse processes of the lower Mastoid process lateral t o splenius Ipsilateral tilt and neck extension
cervical and upper thoracic vertebrae insertion
Semispinalis capitis Lower cervical and upper thoracic Medial part of occiput Ipsilateral tilt and neck extension
vertebrae
From Brin MF: Cervical dystonia—syllabus for treatment of dystonia. Workshop demonstrating the use of botulinum toxin. American Academy of Neurology, New
Orleans, 1999, pp 67-88, with permission.
Digastric muscle
(anterior belly)
FIGURE 13-10. Diagram of muscles of the lateral neck.

Splenius capitus is located immediately posterior t o t h e upper (supe-
FIGURE 13-8. Diagram of the anterior neck muscles poten- rior) portion of t h e sternocleidomastoid.The levator scapulae is just
tially active in torticollis. N o t e the proximity of the carotid artery and inferior t o the splenius capitus muscle. (From Anderson TJ: Spasmodic
internal jugular vein t o these muscles. (From Anderson TJ: Spasmodic tor- torticollis. In Moore P (ed): Handbook of Botulinum Toxin Treatment.
ticollis. In Moore P (ed): Handbook of Botulinum Toxin Treatment. Oxford, Blackwell Science, 1995, pp 103-130, with permission.)
Oxford, Blackwell Science, 1995, pp 103-130, with permission.)
areas are painful, tight, or uncomfortable. The patient should be or hyperactive based on visible spasm, hypertrophy, or head po-
asked to relax and let the head and neck deviate without com- sition may be injected. In cases in which desired clinical results
pensation or resistance. The resultant head position, presence or cannot be achieved with muscle selection based on the clinical
absence of tremor, and muscle hypertrophy or apparent overac- examination alone, EMG can be used to better define the pattern
tivity should be noted. The patient should be asked to demon- of muscle involvement. In 72 patients with rotational torticollis
strate any maneuver he or she has identified to reduce or stop the on clinical exam, the EMG pattern of involvement looking at
abnormal posture (geste antagonist), or any activities, such as SCM, splenius capitis, and trapezius muscles was variable. 58
writing or reading, that may exacerbate or provoke cervical dys- Eight patients demonstrated electrical activity in one muscle, 39
tonia. Finally, the muscles of the neck should be palpated for had involvement of two muscles and 25 had involvement of three
tightness and tenderness. Muscles that are tight, tender, painful, muscles. It is important to reassess the pattern of muscle involve-
ment at each injection session as the pattern may change with
BTX injection. 81
The SCM arises from the mastoid process and the lateral half
of the superior nuchal line. The clavicular head inserts onto the
medial third of the clavicle and sternal head inserts onto the
manubrium sterni (Fig. 13-8). Unilateral SCM activation ro-
tates the chin to the contralateral side, or tilts the head down
toward the ipsilateral shoulder. Bilateral SCM activation may
propulse the head forward or pull the chin downwards. To
inject the SCM, the superior portion of muscle belly should be
grasped between two fingers. The needle is placed into the pos-
terolateral fibers as close to the mastoid process as possible
(Fig. 13-9).
The splenius capitis muscle arises from the mastoid process
and the occipital bone just below the lateral third of the superior
nuchal line. It extends downward, medially, and posteriorly to
3
insert into the lower half of the ligamentum nuchae and onto the
spines of the seventh cervical vertebrae and the upper thoracic
vertebrae. Its location is somewhat deep. It may be palpated
posterior to the upper posterior border of the sternocleidomas-
toid muscle (Fig. 13-10). Unilateral activation of the splenius
FIGURE 13-9. Site of sternocleidomastoid injection. Injection capitis muscles rotates the head ipsilaterally and tilts the head
is placed in t h e posterolateral fibers of t h e sternocleidomastoid as down toward the ipsilateral shoulder. Bilateral splenius capitis
close t o t h e mastoid process as possible. activation extends the neck and pulls the head posteriorly. The
FIGURE 13-13. Site of trapezius injection. (From Anderson TJ:

FIGURE 1 3 - 1 / . Site of splenius capitis injection.The splenius
Spasmodic torticollis. In Moore P (ed): Handbook of Botulinum Toxin
capitis muscle can be palpated just posterior t o the sternocleidomas-
Treatment. Oxford, Blackwell Science, I995, pp 103-130, with permission.)
toid at t h e apex of t h e angle between t h e sternocleidomastoid and
trapezius insertions.The injection is delivered at a depth of 1—2 cm
taking care t o avoid t h e venous plexus t h a t lies deep t o t h e splenius superior SCM may be isolated between two fingers and the
capitis muscle. (From Anderson TJ: Spasmodic torticollis. In Moore P splenius capitis is just posterior and deep to the SCM. The sple-
(ed): Handbook of Botulinum Toxin Treatment. Oxford, Blackwell nius muscle can be palpated here in the untreated patient and
Science, 1995, pp 103-130, with permission.) the injection is generally delivered 1-2.5 cm deep at this loca-
tion. Care should be taken to avoid the venous plexus that lies
deep to the splenius capitis muscle (Fig. 13-1l). 3
The levator scapulae muscle originates superiorly from the

transverse processes of the axis, atlas, and the third and fourth
cervical vertebrae. It inserts on the superior third of the medial
border of the scapulae. It is located behind the SCM muscle in-
ferior to the splenius capitis muscle and superior to the scalene i
complex. It elevates the ipsilateral shoulder and may contribute i
to lateral tilting of the head ipsilaterally. The levator scapulae I
muscle is localized for injection inferior to the splenius capitis !
muscle and superior to the scalenus complex. If hypertrophied

3
it may be palpable. This muscle may atrophy with serial injec-

tions, becoming more difficult to feel (Fig. 13-10).
The superior fibers of the trapezius arise from the medial
third of the superior nuchal line of the occiput, the external oc-
cipital protuberance, and the ligamentum nuchae and pass infe-
riorly to insert on the lateral third of the clavicle. Middle and j
inferior fibers arise from the thoracic vertbrae and attach to the j
acromion and to the scapula. The trapezius muscle elevates the j
ipsilateral shoulder. If the shoulder is fixed, contraction of the j
trapezius contributes to ipsilateral lateral tilting of the head. It I
may act with the contralateral splenius capitis and the ipsilateral
SCM to rotate the chin to the contralateral side. The trapezius
muscles may act in concert with other muscles to extend the
head and neck (Fig. 13-12). The trapezius muscle can be
FIGURE 13-12. Diagram of the muscles of the posterior grasped between the thumb and index finger at the base of the
neck. (From Anderson TJ: Spasmodic torticollis. In Moore P (ed): neck where an angle is formed between the neck and shoulder.
Handbook of Botulinum Toxin Treatment. Oxford, Blackwell Science, The injection is placed in the muscle between the two fingers
1995, pp 103-130, with permission.) (Fig. 13-13). 3
j
The semispinalis capitis muscle arises from the medial aspect

of the area between the inferior and superior nuchal lines of the
occipital bone and inserts on the lower cervical and upper tho-
racic vertbrae. It lies deep to the trapezius and splenius muscles
(Fig. 13-12). These muscles extend the neck and head and with
unilateral activation can modestly contribute to the lateral tilting
of the head down toward the ipsilateral shoulder and to con-
tralateral rotation of the chin. To locate the motor point, the pa-
tient is asked to hold the head erect and the midline nuchal
furrow is identified at a position approximately 2.5-4 cm below the
inion with one finger. The adjacent finger is rolled over the paraspi-
nus muscle complex. With the fingers framing the paraspinous
muscle complex the injecting needle is advanced through the
middle or thickest muscle portion past the outer muscle layers
into the semispinalis muscle, which lies approximately 2.5 cm
deep to the surface (Fig. 13-14). 3
Anterior vertebral muscles in the neck may pull the chin

down or forward. These muscles include the longus colli,
longus capitis, rectus capitis anterior, and rectus capitis lateralis.
Other than longus capitis, these muscles are deep and not
amenable to BTX injection. Unilateral contraction of the
platysma muscle can pull the chin down or down and laterally. 3
BTX is reconstituted as previously described. We use a 10

U/0.1 ml dilution. Injections can be made using a 27-gauge hy-
podermic needle in superficial neck muscles such as trapezius,
SCM, or splenius capitis. A Teflon-coated 24-gauge needle may
FIGURE 13-14. Site of semispinalis capitis injection. T h e
alternatively be used for EMG-guided injection, particularly in
needle is advanced through t h e outer layers of muscle approximately
deep muscles. BTX should be injected into the belly of the
2.5 cm deep to the surface of the skin. (From Anderson TJ: Spasmodic
muscle without EMG localization of the motor point. Table 165
torticollis. In Moore P (ed): Handbook of Botulinum Toxin Treatment.

13-4 provides recommended doses for muscles treated in cervi-
cal dystonia.
Response rates ranged from 53-90% in 17/19 studies pub-
lished between 1986 and 1992 involving more than 1000 pa- for about 2 w e e k s . 3164
Females are more likely to be affected
tients.165
Only 2 studies conducted by a single group of than males. 164
investigators had substantially lower response rates. In gen-

165
eral, studies performed using multiple injections per muscle had SPASMODIC DYSPHONIA
response rates comparable to those using single injection in
each targeted muscle. Borodic and colleagues demonstrated a
165
Spasmodic dysphonia is an action-induced laryngeal move-
higher rate of response to BTX injection for cervical dystonia ment disorder comparable to other focal dystonias such as
using multiple injection points per muscle compared to single writer's cramp, torticollis, and oromandibular dystonia.
injection points when patients were assessed for pain, posture Spasmodic dysphonias may be of either the adductor or the ab-
deformity, range of motion and activity. This is the only pub-
23
ductor type. Patients with adductor spasmodic dysphonia have
lished study comparing the techniques directly. Borodic postu- choked, strained-strangled phonation with voice arrests caused
lates that multiple point injections allow for more homogenous by abnormal intermittent or sustained vocal cord adduction.
diffusion of the toxin effect in the muscle targeted. Multiple
point versus single point injection strategy did not increase the
TABLE 13-4. Recommended Botox Doses
rate of response for muscle hypertrophy or dystonic tremor.
for Cervical Dystonia
Latency of benefit onset is approximately 7 days with a plateau
up to 3 weeks after injection. The effect lasts approximately 3 Muscle Starting Dose Range
months. 165
Sternocleidomastoid 50 U unilateral/ I5-75U

Dysphagia is caused by inadvertent spread of locally injected 25 U bilateral
toxin and is the major complication of BTX treatment for cervi-
Scalene complex 35 U I5-50U
cal dystonia. Approximately 20-25% of BTX injection sessions
are complicated by some degree of dysphagia and up to 57% of Splenius capitis 75 U 50-150 U
patients treated chronically with BTX experience dysphagia Levator scapulae 50 U 50-100 U
with one or more injections. Dysphagia resulting in upper
164
Semispinalis capitis 75 U 50-100 U

airway obstruction has been reported in one patient in a clinical
Longissimus capitis 75 U 50-150 U
study in North America. This patient was successfully treated
with the Heimlich maneuver. Studies have suggested that
22
Trapezius 75 U 50-100 U
doses into the SCM in excess of 100 U are particularly associ- Submental complex I0U 5-25 U
ated with this complication. Limitation of the SCM dose to less
From Brin MF: Cervical dystonia—syllabus for treatment of dystonia:Work-
than 100 U lowers the incidence of dysphagia to less than 2 % . 24
shop demonstrating the use of botulinum toxin. American Academy of

Dysphagia generally begins at about 7 days and typically persists Neurology, New Orleans, 1999, pp 67-88, with permission.
Abductor spasmodic dysphonia patients complain of weak,

whispery or breathy phonation. This is caused by abnormal in-
voluntary contraction of the posterior cricoarytenoid muscle
producing inappropriate abduction of the cords (Fig. 13-15).
Confirmation of the diagnosis should involve neurologic, oto-
laryngologic, and speech-language assessment. Cords should be
assessed for the presence of tremor, dystonia, and dyskinesia at
rest as well as with phonation.174
The intrinsic laryngeal muscles change the size of the glot-

tic opening by modulating the length position and tension of
the vocal cords. The intrinsic laryngeal muscles include the
A B lateral cricoarytenoid, interarytenoid, lateral thyroarytenoid,
posterior cricoarytenoid (PCA), cricothyroid, and thyroary-
FIGURE 13-15. Schematic to illustrate the action of the pos- tenoid (TA) muscles (Figs. 13-16 and 13-17, Table 13-5). A
terior cricoarytenoid (PCA) muscle. A, PCA muscles at rest. B, fine centrally mediated physiologic balance and sequencing
PCA muscles contracted and vocal c o r d s and glottis widely opened. exists between agonist and antagonist muscles. E M G activity
(From Brookes GB: Laryngeal dystonia. In Moore P (ed): Handbook of of the TA muscle increases with the onset of normal phona-
Botulinum Toxin Treatment. Oxford, Blackwell Science, 1995, pp tion. Approximately 100 ms later, contraction of the PCA
181 - 2 0 5 , with permission.) allows controlled adduction to take place. EMG activity in the
PCA continues after the TA has relaxed controlling the final
phase of vocal cord adduction. Modest EMG activation of the
TA muscle with forced inspiration is followed approximately
25 ms later by PCA contraction. A second small TA burst
occurs prior to PCA relaxation to control the concluding ab-
ductor movement (Fig. 13-18). 33
Recurrent laryngeal nerve sectioning as a treatment for spas-

modic dysphonia was first described in 1976. An initial success
rate of 97% at 6 months with this approroach intended to pro-
tuberculum
duce a slightly breathy, but less effortful voice. - The problem
6 55
has been recurrence of symptoms after a time due to hyperad-

duction of the intact vocal cord against the paralyzed cord.
Recurrence rates at 3 years range between 10 and 6 4 % . 6 5 5
Modification of this technique by crushing rather than section-

ing of the nerve produced only 13% improvement at 3 years. . 14
BTX injection through a hollow needle electrode was first d e - s
scribed to treat laryngeal dystonia in 1984. Initial doses 3.75 U

Lot. cricoarytenoid rn.
of Botox were injected into each TA with the hope that treat-
ment would cause enough weakness to resolve tonic spasms
FIGURE 13-16. Lateral schematic of larynx and attached
without completely paralyzing the muscles. Despite EMG tar-
17
muscles. (Modified from Hirano M, Ohala J: Use of hooked-wire elec-

geting of the TA muscle, some degree of diffusion into the adja-
t r o d e s for electromyography of t h e intrinsic laryngeal muscles. J
cent LCA muscle can generally be expected to occur. Two other
Speech Hearing Res 1969;12:362-373, with permission.)
groups subsequently published results of open-label, unilateral |
injection of full paralytic doses (20-30 U) of Botox into the \
TA. - The bilateral and unilateral approaches both resulted in
131 144 ;
Interorytenoid m. significant voice improvement lasting approximately 3 months.

Unilateral injection has been favored by some because there is
no risk of airway obstruction. The main benefit of bilateral in-
jection is the lower toxin dose required. Bilateral injections can
produce symptom improvement with doses that do not result in
visible reductions of vocal cord movement, in contrast to unilat-
eral injections that reduce symptoms only when reduced vocal
cord movement is observed. - Glottal incompetence mani-
131 217
fested clinically by breathiness and aspiration is greater in pa-

tients with bilateral injection than in those receiving unilateral
injection 1-2 weeks following treatment. This difference re-
solved by 1 month post-injection. Complications include tem-
217
porary vocal breathiness, mild dysphagia, and aspiration. 217
Patients develop symptoms of mild glottal incompetence 2 - 3

days after BTX injection and these usually resolve in 1-2
FIGURE 13-17. Schematic representation of intrinsic laryn- weeks. In 1991 a double-blind controlled study confirmed the
216
geal muscles. (Modified from Hirano M, Ohala J: Use of hooked-wire efficacy of BTX therapy suggested by initial open-label trials. 205
electrodes for electromyography of t h e intrinsic laryngeal muscles, j The percutanous approach for laryngeal BTX injection that fol-
Speech Hearing Res 1969;12:362-373, with permission.) lows has been adapted from a technique for recording laryngeal
Chapter 13 CHEMICAL DENERVATION — 4 9 !
TABLE 13-5. Intrinsic Laryngeal Muscles Involved in Spasmodic Dysphonia

Muscle Origin Insertion Action
Lateral cricoarytenoid Lateral surface of cricoid cartilage Lateral muscular process of Rotates vocal processes medially;
the arytenoids approximates vocal cords
Posterior cricoarytenoid Flat posterior surface of cricoid Muscular process of the arytenoids Rotates arytenoid around axis;
cartilage move vocal processes laterally;
abduct cords
Cricothyroid Oblique line on lateral surface of Anterior face of cricoid arch Approximate cricoid and thyroid
thyroid lamina anteriorly; increase distance
between thyroid prominence and
arytentoids; lengthen vocal cords
Thyroarytenoid Inner border of cricoid cartilage Vocal process posteriorly; thyroid Shorten cord; adjust vocal cord tension
prominence anteriorly
From Brookes GB: Laryngeal dystonia. In Moore P (ed): Handbook of Botulinum Toxin Treatment. Oxford, Blackwell Science, 1995, pp 181-205, with permission.
muscles with wire e l e c t o d e s . For injection, patients are

103132
is used for percutaneous injection. Reference and ground elec-
comfortably placed in a supine position with the neck hyperex- trodes are placed on the side of the neck or on the back. After
tended. A 1.5 inch, 27-gauge hollow Teflon-coated needle electrode cleansing with alcohol and skin infiltration with 2 % lidocaine
Phonation Normal
Phonation Laryngeal Dystonia
S8.0|iU 9 Needle ENC
2ee»u
FILTER : SBBHz
Medalec/T«ca " S a p p h i r e "

National Ho«pital
Hrdelec/Teca
National Hospital
Forced Inspiration Normal
Forced Inspiration Laryngeal Dystonia

3 Needle EMC
3 Needle EMC
I | 5*
280uU
Ss
FILTER : SB8Hz
Medelec/Teca " S a p p h i r e "

National H o s p i t a l
ttedelec/Teca "Sapphire'
National Hospital
FIGURE 13-18. Dynamic EMG in a normal subject. Phonation
and forced inspiration onset are marked by arrows. Upper traces: thy- FIGURE 13-19. Dynamic EMG in a patient with severe ad-
roarytenoid muscle activity (adducts, s h o r t e n s , and t e n s e s vocal ductor laryngeal dystonia. Upper traces: thyroarytenoid muscle ac-
cord). Lower t r a c e s : p o s t e r i o r cricoarytenoid muscle acitivity tivity (adducts, s h o r t e n s , and tenses vocal cord). Lower traces:
(abducts vocal c o r d ) . Time markers 500 ms; amplitude: 50 u.V/cm. posterior cricoarytenoid muscle acitivity (abducts vocal cord).Time
(From Brookes GB: Laryngeal dystonia. In Moore P (ed): Handbook of markers 500 ms; amplitude: 200 uV/cm. (From Brookes GB: Laryngeal
Botulinum Toxin T r e a t m e n t . Oxford, Blackwell Science, 1995, pp dystonia. In Moore P (ed): Handbook of Botulinum Toxin Treatment.
181 - 2 0 5 , with permission.) Oxford, Blackwell Science, 1995, pp 181-205, with permission.)
FIGURE 13-20. Percutaneous injection of the thyroary-
tenoid muscle. A, Point-touch technique for botulinum toxin injec-
tion for t h e t r e a t m e n t of spasmodic dysphonia. Ann O t o l Rhinol FIGURE 13-22. Location and direction of needle placement
Laryngol 1992; 101:883-887, with permission.) B.Teflon-coated needle for transcutaneous injection of botulinum toxin into the ip-
in the left thyroarytenoid muscle is directed through t h e cricothyroid silateral thyroarytenoid muscle through the thyroid carti-
m e m b r a n e slightly superiorly and laterally. (From Donovan D., lage. (From Green DC, Ward PH, Berke GS, Gerratt BR: Point-touch
Schwartz KJankovicJ: Unilateral injection of botulinum toxin in spas- technique for botulinum toxin injection for t h e t r e a t m e n t of spas-
modic dysphonia. In Jankovic J, Hallett M (eds): Therapy with modic dysponia.Ann Otol Rhinol Laryngol 1992; 101:883-887, with
Botulinum Toxin. N e w York, Marcel Dekker, 1994, p 467, with permis- permission.)
sion and from Green DC.Ward PH, Berke GS, Gerratt BR.
the needle is inserted through the cricothyroid membrane 2 - 8 but may also result in needle tip displacement. The TA, LCA,
33
mm to one side of midline (Fig. 13-20). Once through the mem- and cricothyroid muscles are continuous with one another and
brane, the needle must be angled sharply superiorly and tun- the wide field of the monopolar needle can make it difficult to
nelled submucosally laterally and superiorly toward the inferior determine in which muscle the needle tip is located. 132
fold of the vocal cord (Fig. 13-21). If the needle is in the glottic A technique for TA injection of patients with adductor laryn-
space or close to the edge of the fold, phonation will be seen and geal dysphonia without EMG guidance has been described. 90
heard on the EMG monitor as a 100-200 Hz complex waveform. The patient is seated and the neck is palpated to identify the out-
The needle should be directed more laterally in this event. If the line of the thyroid and cricoid cartilage. A flexible nasopharyn-,
needle is too far lateral, vibration of the thyroid cartilage will be goscope may be passed though the nose into the hypopharynx
picked up in the EMG signal. If the needle is placed too inferi-
132
after topical anesthesia of the nose and hypopharynx with 2 %
orly, the EMG will show increased activity with expiration and lidocaine spray. Toxin is injected transcutaneously through the
at the onset and termination, indicating the placement in the thyroid cartilage into the ipsilateral TA muscle. The needle is
LCA muscle. A crisp MUP interference pattern appears with
132
oriented at a 90° angle to the skin of the anterior neck and di-
prolonged vowel production and with effort closure of the vocal rected posteriorly in the sagittal plane. The anterior commissure
cords when the needle tip is well-placed in the posterior portion of the vocal cord is estimated to be midway between the thyroid
of the TA muscle. There is no predominant pattern with respira- notch and the bottom edge of the thyroid cartilage in the mid-
tion as the TA motor units fire during inspiration and expira- line. The needle is placed 5 mm lateral and 5 mm inferior to this
tion. Gentle phonation can confirm correct needle placement,
132
point (Fig. 13-22). Correct depth is determined by sensing the
depth at which the needle passes through the thyroid cartilage to
permit easy injection of toxin. If the needle is in the thyroid car-
tilage there will be high resistance to injection because of the
density of cartilage matrix. This resistance eases off as the
needle enters the TA muscle. This method is referred to as the
"touch-point" technique.
Indirect laryngoscopy can be used to inject BTX into the
vocal cord under direct visualization in patients who tolerate pe-
rioral procedures without excessive gagging. This approach ;
79
does not require EMG localization. The patient holds the tongue
out with a gauze pad while the larynx is visualized with a laryn- ?
geal mirror. The larynx is anesthetized with 1 ml of 4% cocaine
(lidocaine or tetracaine may also be used) dripped onto the mu-
cosal surface of the endolarynx. The vocal cord is penetrated on
the superior surface using a laryngeal injector device and BTX
is injected to multiple sites along the anterior-posterior axis.
Proponents of this technique used an initial unilateral dose of
FIGURE 13-21. Patient receiving transcutaneous injection of 2.5 U in 0.1 ml. If symptoms persist after 4 days, a patient re-
78
the thyroarytenoid muscle with EMG guidance. ceives 3.0 U in the contralateral vocal cord. Some patients required
a third injection. The total doses used during the initial treat-
ment were divided, with half being injected in each vocal cord
for subsequent treatments. Total doses ranged from 2.5 to 8 U
with an average effective dose of 4.5 U. Acoustic evaluation of
phonation involved having patients sustain the vowel "a" sound
at a comfortable intensity and pitch for as long as possible. 217
The digitized signal can be evaluated for a number of acoustic

parameters. Twenty-seven patients evaluated before and 1
month following injection showed no change in signal-to-noise
ratio, jitter (frequency perturbation), shimmer (amplitude per-
turbation), or standard deviation in fundamental frequency from
pretreatment measures. They did have significant differences
79
in the timing and distortion percentages of conversational

speech and reading of a standard passage. Abnormal words de-
creased by 4 1 % . Words with voice breaks decreased 26%.
Improvements were judged to be present on voice recordings in
80% of patients. These different techniques have not been di-
rectly compared against one another for efficacy.
The PCA muscle is targeted for injection in abductor spas-
modic dysphonia. Only one side is injected at a time as bilateral
PCA weakness can compromise inspiration. Reaching this
muscle located on the posterior aspect of the larynx is techni-
cally more challenging than TA injection in adductor dysphonia
and may be particularly difficult in muscular or obese patients.
The patient is reclined in a supine postion with the head rotated
j laterally away from the side to be injected. The larynx is manu-
ally rotated away from the side to be treated. After the skin and
FIGURE 13-23. Photograph of injection of the posterior
subcutaneous tissue is infiltrated with 2% lidocaine, a 2.5-3.0
cricoarytenoid muscle with EMG guidance. (From Blitzer A, Brin
inch, 18-gauge, Teflon-coated hollow needle electrode is intro-
MF.The evaluation and management of abductor laryngeal dystonia. In
duced at a point midway between the hyoid bone and the cricoid
Jankovic J, Hallett M (eds):Therapy with Botulinum Toxin. New York,
cartilage, just posterior to the anterior border of the SCM
Marcel Dekker, 1994, pp 451-459, with permission.)
muscle. The needle is directed obliquely in an inferior and
medial direction until it reaches the firm posterior wall of the
cricoid cartilage. The needle is then withdrawn slightly into the cases of exertional wheezing or stridor and 15 cases of dyspha-
belly of the PCA muscle (Fig. 13-23). The patient is asked to gia. These effects were mild and self-limited, lasting about a
activate the PCA muscle by sniffing or inspiring deeply. A cor- week. Tremor severity worsened in some cases after treatment.
responding augmentation of the crisp EMG interference pattern The investigators believe that the visible tremor present before
confirms the location of the needle tip in the PCA muscle. -19 33
treatment was made audible with PCA weakening and in-
Unilateral BTX injection of 56 patients with abductor dys- creased phonation.
phonia failed to produce significant improvement in 7 5 % of pa-
tients. If fiberoptic laryngoscopy revealed residual vocal cord VOICE TREMOR AND STUTTERING
19
abduction following an initial unilateral dose of 3.75 U of

Botox, patients received an additional 2.5-3.75 U on the side Work done at the National Institutes of Health with BTX in-
originally injected. If the injected PCA was observed to be par- jection into the TA muscle for patients with nondystonic voice
alyzed, small serial doses of 0.625-2.5 U were injected into the tremor and stuttering suggests that voice tremor patients experi-
contralateral PCA. No additional injections were given if pa- enced perceived benefit and reduction in speech effort of a dura-
tients developed stridor or significant narrowing of the space tion comparable to that experienced by patients injected for
between the vocal cords (glottic chink). For patients who have spasmodic dysphonia. BTX injection did not appear to be of
had injection of the PCA muscles bilaterally with narrowing of benefit in the treatment of stuttering. 196
the glottic opening and for patients with significant tremor, an

additional 2.5 U were injected into the cricothyroid muscle. In CRICOPHARYNGEAL DYSPHAGIA
these 56 patients, 13 had only 1 PCA injected, 13 patients had
both PCA and both cricothyroid muscles injected, and 9 had a BTX has also been used to treat swallowing disorders re-
unilateral type I thyroplasty in addition to BTX injection. Orr a lated to excess cricopharyngeal muscle activity in seven pa-
functional rating scale, the average pretreatment function was tients. Swallowing problems were caused by spasticity,
3 1 % of normal. The average best post-treatment function was hypertonus, or delayed relaxation of the upper esophageal
70% with an average improvement of 39%. Average pretreat- sphincter and were the result of a variety of underlying prob-
ment overall severity on a standardized 8-point rating scale was lems, including stroke, cancer resection, and acoustic neu-
5. Overall severity post-treatment decreased to an average of 3. roma. BTX was administered under general anesthesia. Five
Patients who had isolated vocal cord spasm did better than pa- of 7 patients experienced improvement or resolution of dys-
tients with combined dystonias. Patients with tremor or respi- p h a g i a . BTX was shown also to help dysphagia related to
181
ratory dyssynchrony prior to treatment also had a less cricopharyngeus muscle and pharyngoesophageal spasm in
satisfactory response to treatment. Adverse effects included 16 patients following laryngectomy. - 50 67
Before treatment. Right esotropia
FIGURE 13-25. Treatment of acquired VI nerve palsy with

BTX.Top, Patient I week after BTX injection, 5U right medial rectus.
N o t e full abduction of right eye, large right exotropia (center), and
palsy of medial rectus due t o BTX injection (right). Bottom, Same pa-
tient 2 months later. Patient is orthoptic in primary position. Excellent
abduction and adduction of the right globe are demonstrated. Patient
has 50° of binocular diplopia-free field. (From Rosenbaum AL:
Management of acute and chronic VI nerve palsy. In Jankovic J, Hallett
M (eds): Therapy with Botulinum Toxin. N e w York, Marcel Dekker,
I994,pp 387-393, with permission.)
BTX, a permanent antagonist contracture could be produced to

correct a misaligned visual axis (Fig. 13-24). The duration of
185
Post-injection. Right esotropia BTX-induced weakness (approximately 2 months) proved in-

sufficient to cause permanent antagonist contracture; therefore,
FIGURE 13-24. Rationale of strabismus treatment. Diagram the effects of BTX therapy in patients with fusion problems
of the effect of BTX injected into o n e horizontal rectus muscle. (Lee remain temporary, but it has proven to be an effective alternative
JP: Strabismus and o t h e r ocular motility disorders. In Moore P (ed): to traditional incisional strabismus surgery for many kinds of
Handbook of Botulinum Toxin Treatment. Oxford, Blackwell Science strabismus despite this temporal limitation. In acute VI nerve
125
1995, pp, 71 - 8 7 , with permission.) palsy, BTX injection of the ipsilateral medial rectus muscle can
restore an area of single binocular vision during the recovery ,
period (Fig. 13-25). Such intervention may allow a patient to '
173
STRABISMUS AND OTHER DISORDERS drive within a few days following treatment. It has been less
OF OCULAR MOTILITY clear whether secondary medial rectus contracture could be pre-
vented, thereby improving the rate of spontaneous recovery. In
Although patching, prisms, and orthoptic therapy may have a a prospective randomized and controlled study of patients with
role in some particular types of strabismus, surgery has been the acute VI nerve palsy, treament with BTX did not improve the
mainstay of strabismus treatment. Recession procedures to sur- rate of ultimate long-term recovery. BTX may be used before ,
124
gically weaken the action of a muscle are more effective than strabismus surgery to augment the surgical result, or following j
strengthening the action of a muscle. A muscle is disinserted
51
surgery if the result is suboptimal. BTX can be used diagnos- |
136
from its position on the globe and secured by sutures to a more tically to confirm the presence of residual lateral rectus function j
posterior position on the globe to effect recession. Thus, the prior to surgery for chronic VI nerve palsy. 173
j
combined effect of the surgery itself and the muscle shortening BTX additionally has a role in conditions where surgery is !
is to reduce the rotational and contractile force of the muscle. not indicated such as acute nerve palsy or acute Graves' dis-
Resection is an alternative procedure which surgically enhances e a s e . Use of BTX treatment and surgery in combination for
136
the effect of a muscle or tendon. Myotomy, myectomy, teno- chronic VI nerve palsy may reduce the number of muscles oper-
tomy, tenectomy, or transposition are some surgical alternatives ated on and the risk of anterior segment ischemia. Infants and 173
to achieve alignment. Use of adjustable sutures in some circum- children can be treated as effectively as adults. 136
j
stances can permit the adjustment of the surgical correction in BTX is effective at improving mild to moderate (up to 40 j
the immediate postoperative period by tightening or loosening prism diopters) nonparalytic horizontal strabismus. Larger
the sutures. - 51 125
angles of strabismus may benefit from BTX, but multiple injec-
Alan Scott became the father of the clinical use of BTX when tions are required and the results are usually less satisfactory
he described its use in the nonsurgical treatment of strabismus. than those achieved surgically. Long-standing paralysis or re-
136
He originally thought that the effect of BTX on ocular muscles strictive extraocular muscle conditions are contraindications to
might be similar to the long-term effect of a chronic sixth nerve BTX therapy. In acquired nystagmus BTX injections may be
136
palsy in which the antagonist medial rectus muscle develops used to preoperatively assess the potential effect of rectus
contracture during the period of acute paralysis. The medial muscle recession, or to treat inoperable nystagmus. BTX in- 126
rectus contracture could persist even after recovery of the lateral jections in patients with congenital nystagmus can produce os-
rectus muscle, permanently affecting ocular alignment. It was cillopsia, which is normally supressed by central compensatory
hypothesized that by paralyzing a muscle temporarily with mechanisms. 126
j
The patient is comfortably seated in a dental-type reclining

chair with adjustable headrest. A ground electrode is placed on
the forehead in the midline and a reference electrode is placed
over the brow of the eye to be injected. For adults and older
125
children anesthesia can be achieved with a local topical anes-

thetic. Conjunctival vessels can be constricted with 0 . 0 1 %
adrenaline (1 drop) just prior to injection. Smaller children
may require additional sedation. The monitor should be set on
free run. The filter can be adjusted to a bandpass of 150 Hz to
16 kHz with a gain of 200 jiV/div. The amplitude of the
MUAPs is usually 400-800 jiV for the medial or lateral rectus
muscles and 100-300 jiv for the inferior oblique or the infe-
rior rectus m u s c l e s . The injector separates the lids gently
125
with his fingers and the patient is asked to look away from the
muscle to be injected.
The tip of the needle electrode is introduced through the con-
junctiva to approximately the equator of the globe for injection
of the lateral rectus muscle. The patient is asked to look in the di- FIGURE 13-26. Extraocular muscles within the right orbit.
rection of action of the muscle being injected to produce EMG The inferior oblique muscle is not shown. (From C r u z OA, Flynn JT:
activity. The needle is advanced until the EMG signal is maximal Strabismus, other therapies. In Jankovic J, Hallett M (eds):Therapy with
and predominated by crisp MUAPs with a steep rise time. The Botulinum Toxin. N e w York, Marcel Dekker, 1994, pp 377-385, with
EMG signal should change as the patient moves the eye from permission.)
side-to-side becoming alternately activated and supressed. The
desired dose of BTX should be injected when the needle is in the
optimal position. The electrode should be left in place for 45 sec- particular benefit when there is a strong rotatory component of
onds to promote local diffusion of toxin rather than back-track- nystagmus. Patients who cannot have surgery require repeat
126
ing of toxin along the path of the needle electrode. 125

injection to maintain benefit (Fig. 13-26).
The inferior rectus muscle is injected through the lower lid. After injection, patients may be instructed to wear sunglasses
Anesthetic drops are not needed. The needle is advanced to the on the way home or to wear an eyepatch for 2 hours. Patients
likely position of the muscle optimizing the quality of the may take aspirin or acetaminophen for mild post-injection dis-
EMG needed as previously described. Care must be taken to comfort. Complications including subconjunctival hemorrhage
distinguish the inferior rectus muscle producing EMG activity and vertical eye deviation are generally minor and infrequent,
with downgaze from the inferior oblique muscle, which acti- occuring in less than 3 % of injections. However, others have
125
vates on upgaze. For superior oblique injections a transcuta-

128
noted higher complication rates: hemmorrrhage (1%), ptosis
neous approach through the upper lid can be used. Doses of 126
(15%), and spread to recti muscles (11%). Injection of the supe-
1.25-5 U of Botox per muscle are recommended in a volume rior rectus muscle and the oblique muscles carries a high risk of
of 0.05-0.1 ml. ptosis and/or unwanted eye deviation. Puncture of the globe
126
For patients with horizontal nystagmus, injected muscles are is possible.

selected to reduce nystagmus in the most prominent direction. Results have been reported for 1320 patients treated at the
Both rectus muscles may require injection. For vertical nystag- Toxin Clinic at Moorfields Eye Hospital in London over a
mus the inferior rectus or oblique muscle may be injected. nearly 10-year period. Patients were treated for concomitant,
125
Alternatively, retrobulbar injection of 2 0 - 2 5 U may be per- paralytic, restrictive, and muscular strabismus. Three patients
formed to produce partial ophthalmoplegia, which may have were treated for refractory convergence spasm. Seven and a half
C
IGURE 13-27. Injection points used to treat blepharospasm or hemifacial spasm. N o t e that lateral injections can affect the ipsilateral
7gomaticus major and minor muscles that are retractors of the nasolabial fold and the lateral angle of the mouth. (From Borodic GE, Ferrante RJ,
3
earce LB.AIderson K: Pharmacology and histology of botulinum toxin. In Jankovic J, Hallett M (eds):Therapy with Botulinum Toxin. New York,
iarcel Dekker, 1994, pp I 19-157, with permission.)
percent of patients achieved a functional cure defined as restored by different authors. Effective doses may range from 5 to 25 U
alignment persisting for at least a year after the last BTX injec- of Botox per eye in blepharospasm patients. The number of in-
tion. Thirty-five percent were discharged because they were sat- jections may vary from 4 - 6 injected sites per eye (Fig. 13-27).
isfied with their alignment or wished to stop treatment. Twenty This variability in technique does not appear to significantly
five percent went on to have surgery as an alternative to BTX affect results. The most common side effect of orbicularis
208
injections. Thirteen percent had ongoing injections. Five and a oculi injection is the development of ptosis. Ptosis is seen in ap-
half percent had surgery in combination with BTX. Less than proximately 12% of BTX treatments for blepharospasm. This 68
1% refused reinjection. Eighty-nine percent of patients had is usually mild and resolves in 1-2 weeks. If more significant
three or fewer injections. A group followed for long-term, re- weakness results in functional disability, lid props or crutches
current treatment received many more injections. can be fitted to eyeglass frames and will allow the patient to see
while waiting for toxin-induced weakness to resolve. Other
72
BLEPHAROSPASM side effects include local discomfort or bruising, diplopia

(2.1%), most often involving the inferior oblique muscle, that
Idiopathic blepharospasm is*a focal dystonia. It can be lies just posterior to the orbital septum, lower facial weakness
asymmetric or even unilateral at onset. It must be distin- (0.9%), and dry eyes (2.5%) or excessive tearing (epiphoria)
guished from hemifacial spasm when unilateral. Secondary (3.5%). 1 avoid injection of the medial aspect of the lower lid
68
blepharospasm may be seen with drugs, Wilson's disease, en- to minimize risk of lacrimal pump impairment and to avoid dif-
cephalitis, and Parkinson's d i s e a s e . Pharmacologic treat-
208
fusion into the inferior oblique muscle. Some authors advocate
ment has been attempted with a number of medications placing all patients on artificial tears while undergoing BTX
including anticholinergics, cholinergics, dopaminergics, anti- treatment. Dose reductions may be considered in patients de-
68
dopaminergics, gabaergics, and benzodiazepines. These drugs veloping problems with dry eyes due to decreased blinking. For
have not been particularly sucessful. Surgical interventions
208a
patients who do not achieve adequate control of spasm with
include facial nerve section and myectomy. BTX injec-
38 141
doses that avoid the dry eye problem, silicone punctal plugs pre-
tions about the eye have been used for blepharospasm since venting tear outflow may keep the eyes wetter and allow more
1983 with an overall response rate of approximately 9 3 % and effective doses of BTX to be used. Excessive orbicularis oculi
25
are the recognized treatment of choice. 68

weakness can result in incomplete eye closure during sleep.
The orbicularis oculi muscle is divided into 3 parts: orbital, Patients with this complication should keep the cornea moist at
preseptal, and pretarsal. Using the bony orbital rim as an night with Lacrilube and an eyepatch. BTX has been reported to
anatomic landmark, injections are made in the medial and lat- cause a transient increase in intraocular pressure (IOP) of 4
eral aspects of the upper lid. Keeping the injections close to the mmHg or more in 18 of 100 patients receiving injection into or- j
eyelash as medial and lateral as possible and directing the bicularis oculi for essential blepharospasm. The elevations
139 1
needle tip away from the center of the eye help to minimize the were demonstrated 13-15 days after treatment and resolved by
spread of toxin to the levator palpebrae superioris, and develop- 2 8 - 3 0 days after treatment in most subjects. The increase in
ment of unwelcome ptosis. Unintended penetration of the or-
25
IOP was with each injection in affected patients. In these cases,
bital septum, a thin membrane separating the eyelid proper from diffusing toxin is believed to affect autonomic cholinergic path-
the orbital compartment, can result in toxin deposition posteri- ways. Acetozolamide is given for 14 days to patients demon-
orly and levator weakness. Upper lid doses should be mini-
68
strating increased IOP after Botox injection.139
mized in patients whose treatment is frequently complicated by

ptosis. Limiting toxin injection to the brow may be sufficient. 68
HEMIFACIAL SPASM
There is some variability in the dose and injection sites reported
Hemifacial spasm (HFS) is characterized by brief involun-
tary, nonsupressible, synkinetic co-contraction of muscles in-
nervated by different branches of the facial nerve. Idiopathic
cases are believed to be caused by microvascular compression
of the 7th nerve at the root entry zone resulting in a focal area
of demyelination and ephaptic nerve impulse transmission. 25
HFS due to posterior fossa mass lesions (< 1% of cases) re-

quire neurosurgical intervention. There are reports of patients
72
benefiting from treatment with carbamazepine. Surgical de-

2
compression for idiopathic HFS is curative 90% of the time

with low morbidity and mortality. BTX therapy presents an
208
alternative to surgery that is simple, safe, and effective for the

majority of patients treated (Fig. 13-28). Primary HFS should
be distinguished from aberrant regeneration of the facial nerve
as a sequela of Bell's palsy that may appear clinically similar.
The injection technique and side effects of orbicularis oculi in-
jection for HFS is similar to that described for blepharospasm.
In hemifacial spasm, synkinetic spasms often involve lower
FIGURE 13-28. Moderate hemifacial spasm of the L face. facial muscles.
Before (left) and 2 weeks after t r e a t m e n t with BTX injection (right). There are a few pitfalls to avoid with treatment of lower facial
(From Biglan AW, Kim S: Management of hemifacial spasm with botu- muscles. Deep injections in risorius can result in weakness of the
linum A toxin. In Jankovic J, Hallett M (eds):Therapy with Botulinum underlying buccinator muscle with biting of the inside of the
Toxin. New York, Marcel Dekker, 1994, pp 353-359, with permission.) cheek. Injections into the orbicularis oris can produce an
208,1
FIGURE 13-29. Patient before and after contralateral injection of BTX to increase facial symmetry during active expression.
A, Asymmetric exposure of teeth and depressed excursions of the nasolabial fold and lateral angle of the mouth during smiling. B, Asymmetry is
corrected by contralateral injections. (From Borodic GE: Hemifacial spasm: Evaluation and management with empahsis on botulinum toxin therapy.
In Jankovic J, Hallett M (eds):Therapy with Botulinum Toxin. New York, Marcel Dekker, 1994, pp 3 3 1 - 3 5 1 , with permission.)
oddly shaped mouth or eversion of the lips. Injections of ele-

208a
tongue movements and jaw deviation. Jaw dystonias may be
| vators of the angle of the mouth such as zygomaticus major can classified as predominantly jaw opening (JO), jaw closing (JC),
I result in cosmetically undesirable smile asymmetry. While some or jaw deviating (JD). Physicians treating jaw dystonia with
degree of asymmetry may be present at rest, the difference is BTX must be very familiar with the local anatomy and should
generally much more apparent with active facial expression. This be prepared to manage complications of therapy. Intraoral injec-
phenomenon is referred to as facial dynamic asymmetry. The 25
tions should be undertaken only by a physician trained in the
patient should be reassured that the asymmetry is due to BTX in- anatomy and physiology of the mouth and pharynx with equip-
jection and will resolve as the effect of the medication wears off. ment to manage complications close at hand. 28
The contralateral side may be injected with BTX to produce a Pterygoid injections require EMG localization as* these mus-
more symmetric appearance when smiling if the degree of asym- cles are deep and cannot be palpated. EMG can be used in other
metry is distressing to the patient (Fig. 13-29). Patients who
25
muscles to improve accuracy and consistency and minimize
desire this intervention should be counseled that the characteris- dose. Recruitment of crisp MUAPs during activation of the tar-
tic appearance of their smile will change and that there may be geted muscle confirms correct placement. Toxin is diluted to 25
some impaired movement of the upper lip, likely to be most no- U/ml and is distributed over 3-5 sites per muscle. Doses recom-
table when speaking. Twitching in lower facial muscles will be mended for treatment of OMD are provided in Table 13-7.
adequately controlled by injections of the lower eyelids in some Serious side effects are not common and are largely related to
patients due to downward diffusion of the toxin or good control excessive local weakness. Treatment of jaw dystonia with BTX
of the eyelid "triggering" muscles. " If involvement of lower
208
may be complicated by dysphagia, particularly with targeting of
facial muscles is mild, it is probably best to avoid lower facial in- the digastric muscles. Excessive weakness of jaw closure may
28
jections. Injection of muscles of the lower face may be needed if occur with treatment of jaw-closure dystonia. One patient has
lower facial spasms cannot be acceptably controlled with eyelid
injections. Table 13-6 provides a guideline for starting doses in
TABLE 13-6. Recommened Botox Starting Doses for
several facial muscles commonly involved in HFS. Lower doses
Hemifacial Spasm
may be required in patients treated for HFS secondary to Bell's
palsy as residual weakness may be present. Muscle Starting Dose
Fifty-seven patients with HFS were treated with BTX injection Orbicularis oculi 12.5-25 U in 4-5 divided doses
in an open-label study. All reported some improvement in the fre- Levator labii superior 5.0 U
quency and intensity of their spasms. The mean disability score
11
Zygomaticus major 2.5 U

before treatment was 6.4 ± 1.2 on a rating scale for involuntary
Risorius 5.0 U
movements described by Marsden and Schachter. This scale as-
signs patients a rating between 0 (lowest disability) and 8 (highest Depressor labii inferior 5.0 U
disability). This improved to 1.5 ± 1.6 after treatment.
1373
Depressor anguli oris 5.0 U
Frontalis 10.0 U
OROMANDIBULAR DYSTONIA Platysma 2.5 U
From Tsui JKC: Blepharospasm and hemifacial spasm. Workshop demonstrating
Oromandibular dystonia (OMD) involves lower facial mus- the use of botulinum toxin. American Academy of Neurology, New Orleans, 1999,
cles of mastication and tongue musculature, producing involuntary pp 63-66, with permission.
TABLE 13-7. Recommended Doses of Botox for by the patient. Care must be taken to distinguish dystonic
Oromandibular Dystonia muscle activity from muscle activity consciously or uncon-
Muscle Median D o s e ± SD Range sciously compensating for the abnormal involuntary move-
ment. 168
Patients should be specifically asked to allow the
Masseter 24.5 ± 7.7 U 2.0-100.0 U
affected limb to deviate without any compensatory effort so that
Temporalis 18.5 ± i 1.9 U 2.0-75.0 U the dystonic movements can be observed. Having writer's
Medial pterygoid 16.3 ± 8 . 1 U 5.0-40.0 U cramp patients write with the unaffected hand can bring out dys-
2.5-60.0 U
tonic movements in the hand at rest. The muscles involved in
Lateral pytergoid I5.9±8.7U
this activity are particularly good to target for injection. Muscle
Anterior digastric 9.8 ± 4.6 U 3.75-30.0 U targeting can be successfully based on clinical observation and
Genio/hyoglossus 18.3 ± 1.5 U 10.0-27.0 U the patient's localization of pain or tightness in most cases. It
171
From Brin MF, Blitzer A, Herman S, Stewart C: Oromandibular dystonia: Treat- is important to reassess the pattern of muscle involvement
ment of 96 patients with botulinum toxin type A. In Jankovic J, Hallett M (eds): before each injection because BTX treatment may alter the pat-
Therapy with Botulinum Toxin. New York, Marcel Dekker, 1994, pp 429-425, tern of activation or unmask muscles not previously recognized
with permission. as involved. EMG should be considered to define the distribu-
115
tion of dystonic muscle activity only when injection of clini-

been reported to have developed j a w closure weakness severe cally selected muscles repeatedly fails to yield the desired
enough to require use of an elastic bandage wrapped around the clinical result despite increases in dose. This procedure is tech-
jaw for eating. Pterygoid injections may be complicated by rhi- nically difficult, time-consuming, and usually unnecessary.
nolalia or nasal regurgitation. Adverse effects occur in 11.8% There are many pitfalls. Dystonic muscle activity cannot be dis-
and 17.5% of patients in the JC and JO groups, respectively. tinguished from voluntary muscle activity on EMG. Discomfort
Patients with mixed JO and JD were categorized as JO. There caused by needle or wire electrodes may affect the pattern of
were 14 instances of dysphagia, but only one instance of dys- muscle involvement producing misleading results. 171
phagia was severe enough to require a change in diet. Other ad- The goal of BTX therapy should be clear in the mind of the
verse effects included hematoma, pain, weakness in chewing, patient and the treating physician before undertaking treatment.
nasal speech, breathy voice, and dysarthria. Thirty-seven to In a highly focal limb dystonia such as writer's cramp or a mu-
45% of 96 patients improved in function with BTX injection for sician's dystonia the goal might be to improve the quality of
jaw dystonia. The duration of benefit was approximately 3
28
writing or playing an instrument, to reduce pain, or to correct
months in patients treated for masticatory spasm. 11
abnormal movements. Improved range of motion, posture, or
pain reduction may be more appropriate goals than restoration
LIMB DYSTONIA of normal function in patients with segmental or generalized
dystonia. Tremor is often a prominent feature of dystonia and
Writer's cramp is the most common focal limb dystonia. muscles involved in dystonic tremor can be treated with the
Patients typically complain of hand or forearm muscle tigtness same outcome as the underlying dystonia. 168
while writing. Abnormal involuntary postures of the fingers, EMG confirmation of needle electrode placement during in-
wrist, or more proximal muscles may occur. The abnormality jections is of particular importance in the small, deep, overlap-
occurs only while writing and tends to worsen with longer writ- ping muscles of the arm and forearm. The anatomy of the FDS
ing efforts. The abnormality resolves with cessation of writing. muscle, a morphologically complex muscle has been de-
Writing becomes slow, effortful, and may be painful over time. scribed in detail to provide guidance on the optimal injection
The patient may have difficulty holding or controlling the pen. site for each fascicle when BTX is used to treat writer's or
Other focal occupational dystonias may afflict musicians, muscician's cramp involving flexion of specific fingers. 13
golfers, and even surgeons with devastating effect. Limb dystonia Passive stretch of a muscle or muscle fasicle with the needle
may be an isolated focal problem or may be part of a segmental correctly placed will produce visible movement. This tech-
or generalized dystonia. nique can be used as an alterative to EMG guidance in situa-
The treatment options in limb dystonia are similar to those tions when EMG is not available. Needle position can also be
for other forms of focal dystonia. Many patients with writer's confirmed by passing a small current through the needle to
cramp will attempt to switch writing to the opposite hand as an produce muscle contraction." 5
initial intervention. Twenty-five percent of those who do switch The most common side effect is functionally significant
hands subsequently develop dystonia in the nondominant weakness in either a targeted muscle or an adjacent muscle.
hand. BTX has been used to treat limb dystonia for over a
138
Development of finger extensor weakness is commonly re-
decade and has been demonstrated to be safe and effective. ported as an unintended complication of injection of the ECR. ' 17
Evaluation focuses on selection of muscles for injection once Weakness of ulnar finger flexors can be seen with injection of
the diagnosis of dystonia is established. The patient should be FCU, and injection of one fasicle of EDC, FDS, or FDP can
examined at rest and performing tasks that bring out the dys- result in weakness of adjacent fascicles. Table 13-8 provides
tonic movement such as writing or playing a musical instru- some recommended doses for muscles commonly treated for
ment. Writer's cramp patients should write long enough to writer's cramp.
reproduce abnormal symptoms. Examination of other activities Eleven of 12 writer's cramp patients treated in an open-label
such as sequentially tapping the fingers on a tabletop or drawing trial with EMG-localized BTX injections reported some degree
a spiral may be helpful. The affected limb should be observed of improvement. Five of 15 reported weakness adversely af-
171
for obvious distortions of posture such as finger or wrist flexion fecting function that was brief in duration (1-2 weeks). Clinical
or extension with writing. Some patients have pain or muscle benefit was sustained for 4-13 weeks, and one patient was free
tightness without any clear dystonic posturing. The limb can be of symptoms for 9 months. Twenty writer's cramp patients
palpated for muscle tightness in areas described as uncomfortable treated in a double-blind BTX investigation were assessed for
writing. Evaluation included speed and accuracy of pen con-

207
TABLE 13-8. Recommended Botox Doses for
trol, completion of Gibson's maze, speed and legibility of writ- Writer's Cramp
ing copying standard passage, and subjective assessment of Mean
writing. All had weakness in injected muscles after receiving
82
Starting Optimal
BTX. Speed and accuracy of pen control was significantly im- Muscle Dose Dose Range
proved in the BTX-injected group. There was no corresponding
Flexor digitorum sublimis I0U 20 U 2.5-^0 U
improvement in the placebo group. Improvement was also sig-
nificant regarding the speed of completion and reduction in Flexor digitorum profundus I0U 10 U 2.5-20 U
errors on Gibson's maze for BTX-injected patients. Writing Flexor pollicis longus I0U I0U 2.5-25 U
speed improved in 7 BTX-injected patients and writing was 20 U 10-40 U
Flexor carpi ulnaris 15 U
scored as better in 4 subjects. Patients showing improved writ-
ing skills were those with significant wrist-joint deviation. Extensor digitorum communis 15 U I0U 5-20 U
Writing worsened in one patient injected with BTX due to de- Flexor carpi radialis I5U 25 U 10-30 U
velopment of transiently excessive local weakness. Eight of 12 Extensor carpi ulnaris 15 U 20 U 10-50 U
patients reporting pain prior to injection had relief of pain.
Extensor pollicis longus 5U I0U 5-10 U
Patients receiving placebo experienced no change. There were
no other side effects described. In 1996 Wissel and his col- Extensor carpi radialis I0U 20 U 10-30 U
leagues used a standardized writer's cramp rating scale (WCRS) Extensor indicis proprius 2.5 U 5U 2.5-10 U
to demonstrate significant post-treatment improvement in 31 Flexor pollicis brevis 5U 5U 2.5-5 U
patients treated in an open-label with Dysport injections. Thirty-
Extensor pollicis brevis 2.5 U 2.5 U 2.5 U
one patients demonstrated significant post-treatment improve-
ment on the W C R S . The most commonly treated muscles in
212 Adductor pollicis 2.5 U 2.5 U 2.5 U
this study were FCU, FCR, FDS II and III, FPL, and ECU. Pronator teres I0U 20 U 10-30 U
Seventy-six percent of patients reported more than 20% subjec- Supinator I5U I0U 10-15 U
tive improvement. The mean response latency was 7.6 days
Interossei 5 U 5 U 5U
(range 1-15 days) with a mean duration of improvement of 59.6
days (range 6-195 days). Weakness was demonstrated in 27/31 Lumbricals I5U 30 U 15-30 U
patients but only functionally relevant in 3 instances. Computer- Pronator quadratus 5U 5U 5U
based writing speed analysis was significantly improved after From Karp Bl: Treatment of limb dystonia with botulinum toxin. Workshop
treatment. WCRS scores assessed through blinded videotape demonstrating the use of tobulinum toxin. American Academy of Neurology,
review by 4 independent raters showed good reliability between New Orleans, 1999, pp 119-132, with permission.
raters and significant improvement in patients before and after
treatment. frequently does not respond to medication. Botox therapy was
In an open-label study of 18 muscians with focal dystonia af- studied in an open-label trial in 78 patients with a variety of
Ifecting ability to play, 8 3 % of patients had improved ability to tremors. Tremor was felt to have been dystonic in 14, essential
108
play with at least one injection. Duration of the effect was 3-4
46
in 12, and mixed dystonic and essential in 52 patients. Two-
months. Unfortunately, for a majority of the treated patients, the thirds of the patients in this study experienced improvement in
degree of improvement was insufficient to permit the level of tremor severity of at least one point on a 0 - 4 scale. There was no
performance required or desired by the patient. It appears that change in tremor frequency. A single-blind, placebo-controlled
the degree of improvement obtained with BTX is generally less
than that required for a professional musician to return to his or
her premorbid level of play. 168 TABLE 13-9. Recommended Doses of Botox for
BTX injections have been employed in proximal arm mus- Treatment of Upper Limb Segmental Dystonia
cles in patients with segmental upper limb dystonia and in leg or Mean
foot dystonia. Recommendations for doses in upper and lower Starting Optimal
limb segmental dystonia are in Tables 13-9 and 13-10. Response Muscle Dose Dose Range
rates in these areas have been somewhat lower than in hand dys- Latissimus dorsi 25 U 75 U 20-180 U
tonia with greater than minimal benefit reported by patients in-
jected at the NIH of 50% and 60%, respectively. 1,6 Trapezius 35 U 85 U 20-120 U
Teres major 20 U 75 U 20-200 U
NONDYSTON1C TREMOR Triceps 20 U 30 U 10-80 U
Therapy for tremor must reduce the tremor and improve func- Pectoralis major 30 U 60 U 20-200 U
tion to be considered sucessful. Functional disability resulting Pronator teres 20 U 25 U 10-30 U
from treatment should not exceed the benefit in reduction of
Deltoid 20 U 40 U 15-60 U
tremor amplitude. The use of BTX to treat focal dystonia includ-
ing that complicated by tremor has been discussed. This section infraspinatus 30 U 50 U 30-60 U
primarily addresses the treatment of nondystonic limb tremor in- Teres minor 30 U 50 U 30-60 U
cluding essential and parkinsonian tremor. Essential tremor (ET)
Biceps I5U 50 U I5-80U
as well as rest and action tremors associated with Parkinson's
disease (PD) generally respond well to medical management. Levator scapulae 15 U 30 U 15-50 U
When medical management has proven inadequate, BTX has From Karp BhWorshop demonstrating the use of botulinum toxin. American
been studied as a therapeutic alternative. Essential head tremor Academy of Neurology, New Orleans, 1999, pp 119-132, with permission.
T A B L E 13-10. R e c o m m e n d e d B o t o x D o s e s for L o w e r (TCD), and 14 had head tremor (HT) without dystonia. Patients
Limb Dystonia in the HT group had received a variety of pharmacologic inter-
Mean ventions including propranolol, primidone, trihexyphenidyl, or
Starting Optimal tiapride without significant improvement. Clinical grading by
Muscle Dose Dose Range the Tsui scale and a 4-point pain scale was performed before
treatment and 2 - 3 weeks following injection along with quanti-
Tibialis posterior SOU 200 U 40-600 U
tative tremor recordings using a bidirectional accelerometer
Extensor hallucis longus 20 U 60 U 30-100 U placed on the forehead. Muscle selection was based on visible
206
Tibialis anterior 100 U 120 U 100-140 U head deviation and palpable oscillation of cervical muscles plus
analysis of simultaneous EMG recordings from 6 cervical mus-
Extensor digitorum longus I5U 55 U 50-60 U
cles. Forty patients reported subjective improvement while three
Extensor digitorum brevis 12.5 U 55 U 40-70 U TCD patients reported no improvement. Treatment significantly
Flexor digitorum longus 40 U 70 U 40-100 U improved both Tsui scores (10.2-5.2 in HT; 3.1-11 in TCD) and
Adductor digiti quinti 5U 50 U SOU pain scale scores (1.5-0.8 in TCD; 1.0-0.4 in HT). Tremor am-
plitude decreased in both groups without change in tremor fre-
Gastrocnemius/soleus 50 U 100 U 100 U
quency. Mild, transient side effects including local pain, neck
Semimembranosus 40 U 40 U 40 U weakness, and dysphagia occurred equally in both groups (39%
Flexor digitorum brevis 50 U 50 U 50 U TCD; 40% HT). No side effects required specific medical inter-
From Karp Bl: Treatment of limb dystonia with botulinum toxin. Workshop
vention. 213
demonstrating the use of botulinum toxin. American Academy of Neurology, BTX used in 6 PD patients and 1 ET patient produced mild to
New Orleans, 1999, pp I 19-132, with permission. marked improvement with a mean change of 2.6 on a 0-4 global
rating scale in an open-label pilot study. Three showed more
than 50% improvement in tremor severity on both clinical rating
study of BTX-A subsequently involved 21 patients with dis- scales and objective measurements. Twelve PD and 14 ET pa-
203
abling nondystonic tremors refractory to medical manage- tients were treated in a subsequent larger open-label trial. No
204
ment. Patients were divided into two groups: patients with

101
objective effect was seen in the PD patients as a group despite
Parkinson-like tremor and patients with ET. Half the patients ex- moderate to marked improvement reported subjectively by pa-
perienced > 30% benefit with placebo. Ten patients benefited by tients. Patients with distal tremors had more benefit than pa-
more than 30% beyond placebo effect after treatment with BTX. tients with proximal tremors. Seven patients treated with BTX
Statistically significant improvement was demonstrated in pos- in a double-blind study of parkinsonian rest tremor demon-,
tural tremor amplitude as measured by accelerometer, but there strated functional improvement of 2 4 % in time to spill water
was no statistically significant improvement in rest tremor am- and in maximum water spilled, and 38.6% improvement in
plitude. In this study, a standardized group of extensor and flexor maximum accelerometric displacement at one month. The
muscles was injected and a standardized initial dose of 50 U was normal saline control group experienced no effect. Benefits
divided equally between extensor and flexor muscles for the first were largely gone by 3 months. Side effects included a median
2 patients. This initial dose resulted in significant focal weakness 20% decrease in finger extensor strength that was not function-
in the first 2 patients studied, and the dose was subsequently ally significant.
8
halved to 25 U and distributed over a wider range of muscles. Case reports in the literature regarding the use of BTX to treat
This particular study involved a single treatment with placebo less common types of tremor include a description of successful
and with Botox for each patient. There was, therefore, no attempt treatment of hereditary trembling chin with BTX injection of
made to gradually titrate the dose upward or adjust the muscles the mentalis muscle. Vertical (yes-yes) cerebellar head tremor
87
selected over a series of injections to optimize the clinical result. resulting from a bilateral cerebellar infarct has been treated with
It is possible that some nonresponding patients may have im- Dysport. This tremor was a 2 - 3 Hz rest and postural tremor
proved at higher doses of toxin. which improved markedly after injection of 80 U into the right
Ten patients with ET were studied in a placebo-controlled SCM and 120 U each into both splenius capitis muscles. Tremor
trial of Botox therapy. All patients had side-to-side (no-no)
157
resolved completely 5 weeks later after injection of 80 U into
head tremors. Forty-unit doses were injected into each stern- the left SCM and 80 U each into the spenius capitis muscles. 76
ocleidomastoid (SCM) muscle, and 60 U doses were injected

into each splenius capitis. Placement was confirmed by EMG. SPASTICITY
Normal saline injections were used for placebo. Five of 10 pa-
tients had moderate to marked improvement after a single treat- Spasticity is a velocity-dependent increase in stretch reflex
ment with Botox. A single subject had mild improvement and 4 activity. Disruption of pyramidal pathways impairs or elimi-;
patients had no improvement. Only 1 subject had improvement nates descending inhibition of gamma motorneuron activity.
with placebo in this study. Eighty percent of subjects with mod- Spastic muscle overactivity may result in contracture, abnormal j
erate to marked benefit had associated objective neck weakness. posture, and pain. Both dystonia and spasticity share excessive 1
No subject had side effects requiring intervention. Patients with involuntary muscle contraction as the symptomatic conse-
nondystonic head tremor treated in an open-label trial with quence of disordered central nervous system physiology.
BTX-A demonstrated statistically significant improvement in Imbalance in muscle tone between agonist and antagonist mus-
the Webster tremor scale and in the Global Disability Scale. 206
cles produces postures characteristic for spasticity. BTX can be
Targeted muscles were determined by clinical examination and used to selectively weaken muscles to improve the balance in
by tremor analysis. EMG guidance was used for toxin place- muscle tone across one or more joints, but does not correct the
ment. Head tremor was quantitatively assessed in 43 patients underlying etiologic physiology and cannot directly improve
treated with Dysport. Twenty-nine had tremulous cervical dystonia function or performance. j
TABLE 13-1 I. Recommended Pediatric Botox Doses for Spasticity

Number of
Clinical Pattern Potential Muscles Involved Dose Units/Kg Injection Sites
Upper Limb
Adducted/internally rotated shoulder Pectoralis complex 2 2-3
Latissimus dorsi 2 2
Teres major 2 1-2
Subscapularis 1-2 1-2
Flexed elbow Brachioradialis I I
Biceps 2 2-3
Brachialis 2 2—3
Pronated forarm Pronator quadratus I I
Pronator teres I I
Flexed wrist Flexor carpi radialis 1-2 I
Flexor carpi ulnaris 1-2 I
Thumb-in-palm Flexor pollicis longus I I
Adductor pollicis longus I I
Opponens I I
Clenched fist Flexor digitorum profundus 1-2 1-2
Flexor digitorum superficialis 1-2 1-2
Intrinsic plus hand Lumbricales interossei 0.5-1 I
Lower Limb
Flexed hip Iliacus 1-2 I
Psoas
Rectus femoris % 3^4 2
Flexed knee Medial hamstrings 3-6 3^4
Gastrocnemius (as knee flexor) 3—6 2—4
Lateral hamstrings 2-3 1-2
Adducted thighs Adductor longus/brevis/magnus 3-6 1-2
Stiff (extended knee) Quadriceps mechanism 3-6 4
Equinovarus foot Gastrocnemius medial/lateral 3—6 1—2
Soleus 2-3 1-2
Tibialis posterior 1-2 I
Tibialis anterior 1-3 I
Flexor digitorum longus/brevis 1—2 I
Flexor hallucis longus 1-2 I
Striatal t o e Extensor hallucis longus 1-2 I
From Russman BS.Tilton A, Gormley ME: Cerebral palsy: A rational approach to a treatment protocol, and the role of botulinum toxin in treatment Muscle Nerve
1997;20:581 -593, with permission.
The traditional armamentarium for pharmacologic spasticity side effects. Four pharmacologic agents (diazepam, dantrolene,
treatment has included systemic, regional, and localized modal- baclofen, and tizanidine) are approved by the FDA for treatment
ities. Peripheral aggravators of muscle overactivity such as pres- of spasticity as a symptom of a CNS disorder in adults. 89
sure sores, ingrown toenails, urinary tract infections, or BTX can be used to selectively weaken muscles to improve the
excessively restrictive clothing should be sought out and elimi- balance in muscle tone across one or more joints, but does not cor-
nated before considering pharmacologic intervention. This ag- 88
rect the underlying physiology and cannot directly improve func-
gravation of muscle activity may result from stimulation of tion or performance. Targeted BTX injections may reduce
afferents such as flexor-reflex afferents. Physical therapy may
133
excessive muscle tone and associated pain, improve postural abnor-
be helpful in the early management of spasticity or in chronic malities reducing the need for splinting and bracing, prevent con-
patients as part of a combined approach in which physical meth- tractures, reduce nursing care, and improve hygiene maintenance.
ods are used to optimize the benefits of other therapies. 1903
BTX can be used together with other treatment modalities, includ-
Systemic pharmacologic treatments would be indicated in the ing systemic medication, intrathecal baclofen pump, or surgery.
setting of diffuse excessive muscle activity, as in anoxic, toxic, Effects and side effects are reversible so the injected muscles and
metabolic, inflammatory, degenerative, and some vascular or the dose can be adjusted over time to produce maximum benefit
traumatic conditions. Clinical conditions commonly resulting in and minimize iatrogenic functional deficits. BTX injections have
diffuse spasticity include multiple sclerosis, spinal cord injury, been studied in several neurologic disorders characterized by spas-
stroke, traumatic brain injury, and cerebral palsy. Numerous ticity including multiple sclerosis, traumatic brain injury, cerebral
medications have demonstrated efficacy in reduction of muscle palsy, spinal cord injury and stroke. - Recommended doses are
10 191
stretch reflexes, but none has been established as universally ef- listed in Tables 13-11 and 13-12 and Figures 13-30 and 13-31.
ficacious in reducing excessive muscle activity caused by le- Placebo or 400 U of BTX were injected into thigh adductor
sions of central motor pathways, and all have potentially serious muscles (100 U in adductor brevis; 100 U in adductor longus;
502 P A R T II BASIC A N D ADVANCED TECHNIQUES
T A B L E 13-12. R e c o m m e n d e d A d u l t B o t o x Dosing for Spasticity

Average Number of
Clinical Pattern Potential Muscles Involved Starting Dose Range Injection Sites
Upper Limb
Adducted/internally Pectoralis complex 100 U 75-100 U
rotated shoulder
Latissimus dorsi J00U 50-150 U 4
Teres major 50 U 25-75 U I
Subscapularis 50 U 25-75 U I
Flexed elbow Brachioradialis SOU 25-75 U 2
Biceps 100 U 50-200 U 4
Brachialis SOU 25-75 U 2
Pronated forearm Pronator quadratus 25 U 10-50 U I
Pronator teres 40 U 25-75 U 2
Flexed wrist Flexor carpi radialis 50 U 25-100 U 2
Flexor carpi ulnaris 40 U 10-50 U 2
Thumb-in-palm Flexor pollicis longus I5U 5-25 U I
Adductor pollicis longus I0U 5-25 U I
Opponens I0U 5-25 U I
Clenched fist Flexor digitorum profundus 50 U 25-75 U 4
Flexor digitorum superficialis 15 U 25-100 U 2
Intrinsic plus hand Lumbricales interossei I5U 10-50 U/hand 3
Lower Limb
Flexed hip Iliacus 100 U 50-150 U 2
Psoas 100 U 50-200 U 2
Rectus femoris 100 U 75-200 U 3
Flexed knee Medial hamstrings 100 U 50-150 U 3
Gastrocnemius (as knee flexor) 150 U 50-150 U 4
Lateral hamstrings 100 U 100-200 U 3
Adducted thighs Adductor longus/brevis/magnus 200 U/leg 75-300 U 6/leg
Stiff (extended knee) Quadriceps mechanism 100 U 50-200 U 4
Equinovarus foot Gastrocnemius medial/lateral 100 U 50-200 U 4
Soleus 75 U 50-100 U 2
Tibialis posterior 50 U 50-200 U 2
Tibialis anterior 75 U 50-150 U 3
Flexor digitorum longus/brevis 75 U 50-100 U 4
Flexor hallucis longus 50 U 25-75 U 2
Striatal toe Extensor hallucis longus SOU 20-100 U 2
Neck
Sternocleidomastoid* 40 U 15-75 U 2
Scalenus complex 30 U 15-50 U 3
Spenius capitis 60 U 50-150 U 3
Semispinalis capitis 60 U 50-150 U 3
Longissimus capitis 60 U 50-150 U 3
Trapezius 60 U 50-150 U 3
Levator scapulae 80 U 25-100 U 3
* SCM dose should be reduced by 50% if injections are bilateral.
From Brin MF: Cervical dystonia—syllabus for treatment of dystonia: Workshop demonstrating the use of botulinum toxin. American Academy of Neurology, New
Orleans, 1999, pp 67-88, with permission.
200 U in adductor magnus) in 10 nonambulatory multiple scle- In an open-label injection of 100-400 of BTX into thigh ad-
rosis patients with thigh adductor contractions severe enough to ductors, hamstrings, gastrocnemius-soleus, and posterior tibial
interfere with activities of daily living (ADLs) and hygiene. 191
muscle groups^ 10 of 11 myelopathic patients demonstrated sig-
Patients received injection with the alternative therapy at 3 nificant improvement on one or more scales during a 2 month
months in this double-blind, placebo-controlled, randomized assessment period. Patients were evaluated for tone, joint mo-
study. Patients receiving BTX had a reduction in spasticity score bility, gait, and global function. Seven of 12 spastic parapare-
20
from a mean of 7.9 to 4.7 at 6 weeks post-injection. The same pa- sis patients receiving open-label BTX injection of thigh
tients injected with placebo had an increase in spasticity score adductors reported decreased spasticity. Eight of 10 in this same
from a mean of 6.8 to 7.8 at 6 weeks. Hygiene scores also demon- study reported decreased speed in ambulation.
strated a significant benefit with BTX over placebo. No adverse An open-label study evaluated BTX injection in 15 children
effects were reported. Small open-label studies also suggested with cerebral palsy (5 hemiplegic, 5 diplegic, 5 quadriplegic). 5
that BTX was effective in reducing lower limb spasticity. 10 2 1

Patients had spasticity, dystonia, or both. Overactive muscles
were identified by physical examination. Up to 4 - 6 U/kg body pubic tubercle

weight were administered. A majority of patients showed im-
provement on a 4-point assessment scale at 6-8 weeks. Patients medial femoral
with spasticity alone appeared to benefit more than dystonic pa- epicondyle
tients or patients with mixed spasticity and dystonia. Twelve
cerebral palsy patients in a double-blind placebo-controlled
study received 1 U/kg for each leg divided between the medial
and lateral gastrocnemii. Five of 6 patients in the active treat-
120
ment group demonstrated improved gait compared to 2/6 pa-

tients in the placebo group. No adverse events were reported.
Spasticity in children presents a particular challenge because
spastic muscles may fail to grow normally. Hereditary spastic
mice can be used as an animal model for spasticity in children.
Gastrocnemius muscles in phenotypically spastic mice are sig-
nificantly shorter (P < 0.005) than in phenotypically normal
mice of the same breed. Treatment of spastic mice with BTX-A
FIGURE 13-30. Injection sites for thigh adductor muscles.
restored normal muscle length, while treatment of normal mice
(From Tsui JKC, O'Brien C: Clinical trials for spasticity. In Jankovic J,
with BTX had no effect on muscle length. Similarly, the stan-
Hallett M (eds); Therapy with Botulinum Toxin. N e w York, Marcel
dardized tendon length of the untreated spastic mice was signif-
Dekker, 1994, pp 523-533, with permission.)
icantly longer (P < 0.05) than that of the normal mice and the
spastic mice treated with BTX-A. The tibial length of the spas-
tic mice was 4 % shorter than that of the normal mice. BTX Electrical stimulation of the target muscle with the injection
treatment did not have a significant effect on tibial bone length needle can also be used to overcome this problem. Patients
1903
despite normalization of muscle length. In pediatric patients

49
with spasticity appear to require higher doses than dystonic pa-

BTX may not be able to prevent the need for corrective surgery. tients in order to achieve the desired reduction in muscle
There may be benefit in delaying surgery until growth is no tone. Fortunately, there is generally less concern about the
!90a
i longer a factor.
120
Seventeen patients with severe spasticity and a nonfunction-

i n g arm were studied in an open-label trial. Patients were in-
12
jected with 400-1000 U of Dysport or 100-200 U of Botox and

were assessed before and after injection for passive range of
motion at the shoulder, elbow, wrist, and fingers using the mod-
ified Ashworth spasticity score, and an eight-point scale grading
degree of difficulty experienced by the patient or caregiver for
each functional problem defined before treatment. Functional
problems included cleaning the palm, cutting fingernails,
putting the arm into a sleeve, balance with standing and walk-
ing, putting on gloves, and rolling over in bed. Muscles injected Areas of injection ^ _ Jp± v
included biceps brachii, flexor carpi ulnaris, flexor digitorum
Gastrocnemius
profundus, and flexor digitorum superficialis. Mean passive
range of motion improved 17° at the shoulder, 16° at the elbow,
and 30° at the wrist. The median Ashworth score improved from
5 to 3 for biceps brachii spasticity and from 5 to 4 for finger
flexor spasticity. Overall, 4/17 patients reported some functional
improvement post-injection. Benefits lasted 1-11 months, and Soleus
there were no adverse effects reported.
Five of 6 patients had improvements in ADL in an open-
label use of BTX-A in six patients with spasticity due to trau-
matic brain injury. 161
Specifically, 2 patients regained the
ability to write, one developed sufficient hand function to use
hand-held walking devices, and 3 patients became able to eat
and dress independently after treatment. All had improve-
ments in their modified Ashworth scale scores and range of
motion. No patients complained of weakness due to BTX in-
jections. The mean dose of Botox injected into upper limb
muscles was 96 U.
Technically, localization of the targeted muscle may be more
difficult in spasticity than in dystonia because a paretic patient
may be unable to voluntarily contract the involved muscle to
produce EMG activity. One method of localizing the needle tip FIGURE 13-31. Injection sites in the gastrocnemius/soleus
within the target muscle in such cases involves moving the distal group. (From Cosgrove AP, Graham, HK: Cerebral palsy. In Moore P
joint acted on by the target muscle passively through its full (ed): Handbook of Botulinum Toxin T r e a t m e n t . Oxford, Blackwell
range of motion and observing the resultant needle movement. 49
Sciences, 1995, pp 222-247, with permission.)
504 — PART I I BASIC A N D A D V A N C E D T E C H N I Q U E S
risk of excessive weakness as the muscles injected are often al- Ethyl alcohol in dilutions of 35-60% has been used to achieve
ready paretic and nonfunctional. temporary nerve block in patients with spasticity. Like phenol,
alcohol denatures the protein in nerve tissue. Treatment experi-
Chemical Neurolysis for T r e a t m e n t of Spasticity ence with alcohol for spasticity has predominantly involved in-
Chemical neurolysis with phenol and alcohol injections has tramuscular injection for the treatment of cerebral palsy in
also been used to treat spasticity. These agents can be adminis- children. Forty-five percent alcohol injected into the motor
88
tered locally to treat spasticity, avoiding systemic side effects. point of muscles was demonstrated to reduced spasticity without
Phenol (carboxylic acid) has been widely used as an injectable affecting voluntary strength in children with cerebral palsy in the
agent in the regional treatment of spasticity. Phenol can be in- 215
mid-1960s. Results lasted for 6-12 months with some patients
201
jected at the muscle's motor point or directly into nerve where it deriving benefit for as long as 2 - 3 years. Forty-five percent dilu-
denatures the protein in nerve fibers. Phenol, in low concentration of alcohol injected in large quantities (10-40 ml) into multi-
tion, has the properties of a local anesthetic. Protein coagulation ple sites in targeted muscles were also reported to improve
and necrosis are seen with concentrations greater than 5 % . 215
spasticity with preserved strength and sensation. Other investi-
156
Phenol causes protein coagulation and necrosis at concentra- gators have reported shorter duration of benefit of 1-6 weeks. 40
tions greater than 5 % . Alcohol can be used similarly to dehy-

117
Shortcomings of dilutions at 3 5 ^ 5 % include pain with injec-
drate nerve tissue resulting in sclerosis of the nerve fibers and tion, transient symptoms of systemic toxicity, and skin irritation
destruction of the myelin sheath. Disadvantages of phenol or al- or ulceration - Conscious sedation or general anesthesia may
40 121
cohol injections include pain and dysesthesias that may last be of benefit for the injection procedure. 121
weeks to months, variable duration of effect, and potential car- Local anesthetics with a short duration of action have been
diac arrythmias. In some cases, effects can be permanent. 1903
used prior to chemical neurolysis to diagnostically determine
Phenol or alcohol, injections may have particular benefit in large the mechanism of impairment (increased tone versus contrac-
proximal muscles that would require BTX doses exceeding rec- ture) and to predict expected benefit of a longer-lasting block. 88
ommended total doses to achieve benefit. Randomized, placebo-controlled studies are needed to confirm
Phenol has been used intrathecally, epidurally, intraneurally, the safety and efficacy of chemical neurolysis for spasticity.
perineurally, and intramuscularly for treatment of pain and spas-
ticity. Intrathecal and epidural injections introduced in the
215
Surgery for Spasticity
1950s are no longer in common use. Perineural, intraneural, and Surgical treatments of spasticity have been aimed at four sites
intramuscular injections are used in the treatment of spasticity of potential intervention: the brain, the spinal cord, peripheral
alone or in some cases in combination with BTX. Phenol injec- nerve, and muscle. Stereotactic surgery has been performed for
tion provides temporary reduction in spasticity producing joint symptomatic treatment of dyskinesia, rigidity, and spasticity.
contracture or interfering with function and rehabilitation. Targeting is controversial as the anatomic source for spasticity is
Temporary interruption of nerve function can prevent or correct unclear. Targets have included the globus pallidus, ventrolateral
deformities due to excessive muscle tone in patients with recently thalamic nuclei, and cerebellum. Complications of central surgi-
acquired spasticity in whom the ultimate neurologic outcome is cal approaches may include sensory loss, paresthesias, weakness,
uncertain and definitive orthopedic surgery is inappropriate. 80
and recurrent spasticity. Twenty-eight patients treated surgi-
1903
Peripheral nerve blocks provide more complete block than motor cally for cerebral palsy were followed for a mean of 21 years
point injections. - Phenol nerve injections may be performed
80 117
postoperatively. Results were good in patients with moderate to
percutaneously or using open surgical procedures with direct vi- severe dyskinetic cerebral palsy but poor in patients with quadri-
sualization of the nerve. Open procedures are preferred to percu- plegia or diplegia with spasticity. The diffuse nature of the under-
taneous injection for nerves containing both sensory and motor lying brain lesions seen in patients with spasticity suggest a
function so that the motor branches can be identified and isolated limited potential for success. Generally poor clinical results bear
to prevent post-block dysesthesia and sensory loss. Motor 149
this out and provide evidence that stereotactic surgery is not war-
branches are identified using a nerve stimulator. Three percent ranted in the current treatment of spasticity. Selective posterior
phenol in glycerine is injected into the exposed nerve until it be- rhizotomy can accelerate progression of hip subluxation causing
comes translucent and no motor response is elicited with electri- instability of the hip and the lumbar spine in cerebral palsy pa-
cal nerve stimulation. Neutralization with alcohol limits damage
80
tients with underlying hip dyplasia, a common comorbidity. 93
to surrounding tissue when phenol is applied directly to the

nerve. The affected limb may be treated with serial casting to
121
Patient Selection in Spasticity
correct any residual fixed contractures. Open treatment of the
80
Any limitation of range of motion to be treated with BTX

musculocutaneous and tibial nerves are described in the orthope- should be passively reduceable to some extent as BTX cannot
dic literature for the treatment of elbow flexion contracture and free a frozen joint or reduce contracture. Clear goals of treat-
equinovarus ankle posturing, respectively. Obturator blocks
80150
ment should be delineated and might include improved func-

decrease lower limb scissoring and facilitate hip abduction. tion, posture, range of motion, hygiene, and reduced pain.
Median nerve blocks relax wrist and finger flexors. Paravertebral Patients considered for BTX therapy should have been tried on
lumbar spinal nerve blocks reduce hip flexor spasticity. 117
medical therapies and demonstrate no benefit or unacceptable

Complications include sensory loss, dyesthesias (3-10% for sen- side effects. Muscles to be injected should not be excessively
sorimotor blocks), excessive weakness, and venous thrombo- wasted. Prior surgical intervention should not distort the
s j 100,134,215 A review of the use of phenol as a neurolytic agent
s anatomy of the area considered for injection. BTX treatment
published in 1978 concluded the following regarding peripheral can benefit carefully selected spastic adults and children.
nerve injections for spasticity: "Apparently anyone who under- Concurrent physical therapy can optimize the benefit of BTX
takes to do this should consider it only as an adjunct to other treatment. Studies comparing phenol, alcohol, and BTX to one
forms of treatment. The effect will be temporary around two another to determine if subsets of patients benefit from one
months to a year, with the shorter period more likely." 215
modality over another have yet to be performed.

FIGURE 13-32. Coordinated activity of the sphincter and de-

trusor to achieve normal voiding. (From Fowler CJ: Disorders of the
pelvic floor. In Moore P (ed): Handbook of Botulinum Toxin Treatment.
FIGURE 13-34. Technique used to inject toxin into the ex-
ternal urethral sphincter. A, 23-gauge, 35-cm Teflon-coated
Detrusor-Sphincter Dyssynergia m o n o p o l a r needle e l e c t r o d e . B, Cystoscope. C, External urethral
sphincter. D, Syringe containing toxin for injection. E, Eyepiece of cys-
Detrusor-sphincter dyssynergia can be a serious problem for toscope. F, Electrode t o EMG instrument. (Dykstra DD: Effects of bot-
patients with upper motor neuron injury or disease. The resul- ulinum toxin type A on detrusor-sphincter dyssyner in spinal c o r d
tant elevations in bladder pressure and impaired bladder empty- injury patients. In Jankovic J, Hallett M (eds):Therapy with Botulinum
ing can lead to renal damage, infection, and autonomic Toxin. New York, Marcel Dekker, 1994, pp 5 3 5 - 5 4 1 , with permission.)
dysreflexia (Figs. 13-32 and 13-33). Traditional management
includes medication, catheterization, electrical stimulation or
surgery to weaken sphincter function, and have in general been weakness lasting 2 - 3 weeks. Treatment does not improve con-
69
proven ineffective or unsatisfactory. BTX has recently been

69
tinence in incontinent patients. 184
used to relax the external urethral sphincter as an adjunct or al- A transurethral technique for BTX injection is described. 69
j tentative to these measures. An open-label BTX treatment study Botox (140 U) is diluted to 5 U/0.1 ml and is injected into the
was made of 11 men with spinal cord injury in 1988. Post-void rhabdosphinter through a cystoscope with a Teflon-coated
residuals (PVR) decreased in 8 men and autonomic dysreflexia monopolar needle electrode connected to an EMG instrument
decreased in 5. Urethral pressure profiles were decreased in the (Fig. 13-34). Subsequent weekly doses of toxin are 240 U.
7 men in whom it was measured. This study was followed by a
69
Patients required an average of 3 injections to produce a maxi-
double-blind treatment trial in 5 men with spinal cord injury. mal decrease in post-void residual volume. Urine was obtained
Again PVR, bladder pressure during voiding, and urethral pres- for culture prior to the procedure and patients received prophy-
sure profiles were reduced in the patients receiving active treat- lactic antibiotics for 4 days. Blood pressure was monitored
ment but not in controls. Men receiving saline injections as before and during cystoscopy out of concern for autonomic hy-
controls were subsequently injected with BTX with results sim- perreflexia caused by sympathetic stimulation. Patients w h o
ilar to the active treament group. Benefits lasted for approxi- became symptomatic during cystoscopy were treated with 10
mately 2 months. Three patients developed mild generalized mg of nifedepine by mouth prior to subsequent injection or 10
mg of nifedipine sublingually during the procedure.
An alternative technique is described in 6 patients who re-
ceived 100 U of Botox transperineally into the external ure-
thral sphincter. A 2-inch, 26-gauge Teflon-coated monopolar
needle is inserted in the midline 1.5-2.0 cm anterior to the
anus. The operator placed a gloved finger in the rectum to
monitor the position of the prostate gland while directing the
needle electrode toward the external urethral sphincter. In this
study Botox was diluted with 1 ml of gadopentetate and MRI
performed immediately after EMG-guided injection con-
firmed placement in the targeted muscle. Beneficial results
184
have been obtained in larger prospective studies by the same

group with amelioration of detrussor dyssynergia symptoms in
more than 8 5 % of treated patients. Fifty transurethral and
183
35 transperoneal BTX injections were performed in 24 pa-

tients. Both techniques had similar efficacy. These studies 184
have used a single (100 U Botox) injection vs. serial injections

1 week to 1 month apart. The major difference found between
the protocols was that the duration of effect was longer (9
months compared to 2 - 3 months) with the series of 3 monthly
FIGURE 13-33. Detrusor-sphincter dyssynergia. (From Fowler injections. BTX injection of the urinary sphincter may b e
183
CJ: Disorders of t h e pelvic floor. In Moore P (ed): Handbook of efficacious with single injections of relatively low doses of
Botulinum Toxin Treatment. Oxford, Blackwell Science, 1995, pp BTX (150 U Dysport) in this setting with an average duration
263-269, with permission.) of effect of 2-3 months. 163
i
GASTROINTESTINAL DISORDERS of 2 5 % without significant change in maximum voluntary pres-

sure indicating a lack of external anal sphincter involvement.
Botulinum toxin is believed to affect the cholinergic stimula- Resting pressures were approaching baseline by 2 months. The
tion of visceral smooth muscle as well as that of striated skeletal fissures healed in 7/10 patients by 2 months. One of these pa-
muscle. BTX blocks pre- and postganglionic parasympathetic tients relapsed at 3 months. Another patient had healed at one
nerve endings at which ACh is the transmitter. month but relapsed a month later. Long-lasting healing, there-
fore, occurred in 60% of patients. One patient had mild fecal in-
Achalasia continence lasting one day. There were no other complications.
Achalasia is a disorder of esophageal motility in which there is Unlike most conditions, BTX in this circumstance addresses the
absent esophageal body peristalsis and failure of the lower underlying etiology of the condition. The duration of the effect,
esophageal sphincter (LES) relaxation with swallowing. although limited, was sufficient to allow healing for the major-
Pharmacologic treatment, surgical myotomy, or balloon dilation ity of patients. There may be other Gl indications for use of
have been the traditional treatment options. Each has had draw- BTX. Further randomized and placebo-controlled studies are
backs in terms of efficacy or morbidity.' A pilot BTX treatment
60
needed in this area.
study of 5 patients with achalasia was reported in 1993. These
139
patients ranged in age from 44 to 85, suffered from severe dys- HYPERFUNCTIONAL FACIAL LINES
phagia, and met manometric, radiographic, and endoscopic diag-
nostic criteria for achalasia. Four of these patients had failed The observation has long been made that patients with Bell's
multiple balloon dilation attempts and 1 had failed cardiomy- palsy have few or no wrinkles. It was later noted that patients
otomy. Patients underwent upper endoscopy and 80 U of BTX di- receiving BTX injections of facial muscles for the treatment of
luted to 20 U/ml (divided into 4 sites) was injected into the LES facial movement disorders such as hemifacial spasm, ble-
via a 5-mm sclerotherapy needle under direct visualization. Four pharospasm, or Meige's syndrome lose some of their wrinkles
patients had complete relief and improvement in at least 2 objec- or "hyperfunctional lines." BTX has been studied for the cos-
tive parameters including LES pressure, cine-esophagography, or metic treatment of wrinkles created by hyperfuctional muscle
esophageal transit time. The remaining patient had a partial reactivity on the basis of these observations. The results of open-
sponse. Mild reflux esophagitis in the patient with prior history of label treatment of hyperfunctional lines in 26 patients were pub-
myectomy was the only side effect. Successful treatment with lished in 1994. Twenty of these patients had a dystonic facial
18
BTX of a patient with abdominal pain due to sphincter of Oddi movement disorder, 4 had hemifacial spasm, and 2 patients
dysfunction has also been reported. 160
were treated strictly for hyperfunctional facial lines. Doses were
individualized for each patient. Botox was reconstituted into di-
Anismus lutions ranging from 1.25 U/0.1 ml to 10 U/0.1 ml, and was
Botulinum toxin has also been used to treat intractable consti- divided into 0.1 ml aliquots. BTX was injected through a
pation due to anismus, a condition in which the puborectalis monopolar hollow Teflon-coated EMG needle connected to an
muscle and the anal spincter muscle contract paradoxically EMG instrument. All patients reported benefit, with effects last-
during defecation. There is no satisfactory treatment for this con- ing 3-6 months. Thirty-one subjects were injected with 10 U on
dition and patients are usually treated with an aggressive combi- each side for glabellar frown lines. Injections were placed both
nation of laxatives. Attempts at surgery dividing the puborectalis directly into the corrugator and under the wrinkle furrow. Direct
muscle have not enjoyed great success. BTX was first used for injections into the corrugator muscle lasted longer and were
this indication with a series of 7 patients reported in Lancet in preferred by a majority of patients. Treatment of the mental
41
1988. Sixty units of Dysport used on each side of the external

94
crease, crow's feet, the procerus line and dilator naris levator,
anal sphincter produced the best result. Four patients reported frontalis lines, nasolabial fold, upper lip muscles, and platysma
excellent results and returned for repeat injections. One patient have all been described in the literature. Complications have in-
had no improvement. Two patients had improved evacuation but cluded brow ptosis, lid ptosis, medial rectus paralysis, and
developed fecal incontinence. In a second series of four patients weakness of the upper lip. - Twenty-six patients were treated
18 41
treated with injections of the external anal sphincter, a hollow- for glabellar frown lines, crow's feet, horizontal forehead lines,
Teflon-coated 23-gauge EMG needle was used to confirm the di- and nasal crunch lines. The needle was inserted at one end of a
75
agnosis of paradoxical contraction and to confirm placement into hyperfunctional line and toxin was deposited as the needle was
the inappropriately contracting muscle. Six units of Botox
111
removed along the path of the treated muscle. EMG was used to
were injected into each half of the anal sphincter. All patients identify areas of greatest activity when the patient moved to ac-
were relieved of constipation 2 - 4 days after injection without centuate the treated line. These patients had no unwanted weak-
local or systemic side effects. Specifically no patients developed ness of facial muscles and no systemic complictions.
incontinence. Patients had normal bowel movements daily with- Ecchymosis and inadequate response subsequently requiring
out laxative use for 3 months. At 3 months two patients devel- repeat injection did occur. For experienced practitioners, the
oped recurrent symptoms. added benefit of EMG for this indication is unproven. Use of a
30-gauge needle without EMG is more comfortable for the pa-
Anal Fissures tient. A double-blind, placebo-controlled trial of treatment in-
Injection of 15 U of Botox in the internal anal sphincter has volving 12 patients demonstrated significant reduction in facial
been used to treat anal fissures as an alternative to sphinctero- wrinkles as judged by patients and by blinded surgeons review-
tomy. Anal fissures are believed to be maintained by contraction
983 ing slides of patients before and after BTX treatment. BTX
118
of the internal anal sphincter, and can cause severe pain. Toxin injections used together with collagen injections to treat wrin-
was divided into 3, 5 U aliquots placed bilaterally and posteri- kles are believed to increase the duration of cosmetic effect, and
orly into the internal anal spincheter localized by palpation. laser-peel procedures followed by BTX can delay the reappear-
Resting pressures on anal manometry were reduced a maximum ance of facial lines.
71
C h a p t e r 13 CHEMICAL DENERVATION — 5 0 7
HYPERHIDROSIS which remains unclear. Nonetheless, myofascial pain is

common, a major cause of suffering and disability and a treat-
BTX blockade of acetylcholine release at cholinergic auto- ment challenge. "Trigger points" are the clinical hallmark of this
nomic synapses has suggested the clicial utility of the toxin in condition. A trigger point is defined as a focal area of muscle
140
treatment of hyperhidrotic conditions such as gustatory sweat- tenderness in a palpably taut band of muscle. Brisk palpation
ing. It has been observed that patients receiving facial injections should produce a muscle twitch response in order to be consid-
for hemifacial spasm develop localized areas of anhidrosis. 35
ered abnormal or "active." Mechanical stimulation of the trigger
Gustatory sweating or Frey's syndrome is a result of parotidec- point should generate referred pain reproducing the patient's
tomy. Sprouting salivomotoric parasympathetic fibers from the chronic pain complaint. A double-blind, placebo-controlled,
otic ganglion to the parotid gland are believed to become misdi- crossover study of trigger point injections with BTX-A was per-
rected following parotidectomy." When these fibers reach the formed in six patients with chronic myofascial pain involving
severed distal ends of sympathetic nerve fibers innervating the cervical paraspinal and shoulder muscles. Trigger points were
44
sweat glands and subcutaneous blood vessels, a new, often a so- injected with 50 U of Botox diluted in 4 ml of normal saline and
cially distressing reflex stimulating facial sweat glands and divided among 2 or 3 sites or with normal saline alone as
blood vessels with mastication or other gustatory stimulation is placebo. Four patients reported a decrease in pain of 30% or
created. Postganglionic parasympathetic nerve fibers and sym- greater on at least 2 occasions with BTX-A injection but not with
pathetic nerve fibers are both mediated by acetylcholine. The placebo injections of saline. Patients with a positive response re-
first sucessful treatment of Frey's syndrome with BTX-A in- ported inprovement in all 5 variables, including pain visual
jected intracutaneously was reported in 1995. 62
analog scale, verbal descriptors of intensity, unpleasantness,
One 100 U ampule of Botox is diluted with 4 ml of normal spasm, and tenderness. Patients were not told when to expect an
saline to achieve a dilution of 2.5 U/0.1 ml. The distribution of effect and the first effects were reported within the first week,
the skin area to be treated is determined using the starch-iodine but not at 30 minutes. Statistically significant differences began
test. Intracutaneous injection of 0.1 ml of this solution is
145
between 2-4 weeks and the duration of benefit ranged from 4-11
made approximately every second centimeter in a grid covering weeks. BTX injections were well tolerated and no subject expe-
the affected area. A mean dose of 65 U with a range of 25-88 U rienced symptomatic weakness. This author points out that the
per session was used in a prospective study of 14 patients. 122
major advantage of BTX over other treatments is the potential
Mean doses of 31.3 U of Botox (range 2.5-100 U) were used in for a sustained effect lasting weeks or months. Longer response
another series of 19 patients. The unwanted sweating disap-
62
times could be possible with higher doses. Botox injections of
peared in 2 days. No side effects were reported. Symptoms 80-150 U in the psoas, pyriformis, and scalenus anterior mus-
reappeared in 12 patients. The mean duration of effect was 17.3 cles compared to similar volumes of lidocaine/methyl-
months. Three patients were reinjected. The recurrence of prenisolone in patients with myofascial pain syndrome. In this
symptoms suggests that successful reinnervation of the sweat study the reduction of pain was better in the BTX patients than
glands can take place. It is not clear why recovery of function in the anesthetic/steroid group. This trend did not become statis-
(or dysfunction) takes so much longer with sweating and va- tically significant until 60 days post-injection. 166
sodilation than with motor function. Some authors propose that Trigger points differ from the tender points in fibromyalgia in
this delay may be due to the presence of scar tissue forming a that tender points are not associated with a twitch on palpation.
natural barrier between the nerve ending and the sweat gland; Fibromyalgia is a more diffuse pain condition. Criteria for the
however, recurrent symptoms can occur well away from areas diagnosis of fibromyalgia outlined by the American College of
scarred due to parotidectomy. Patients may have positive results Rheumatology include pain in 3 or more regions above the
on starch-iodine testing despite subjective reports of absence of waist lasting for 3 or more months, and 11-18 or more tender
symptoms. It is possible that while some sweat glands may
123
points. BTX injection was of no benefit in 2 patients treated
214
regenerate, the numbers may be insufficient to reach a clinically as part of a planned randomized, double-blind study of 10 pa-
symptomatic threshold. At this time there is no fully adequate tients with fibromyalgia involving shoulder girdle muscles.
explanation for the long duration of BTX effect in this setting. Subjects were to receive either lidocaine or 25 U of BTX-A into
Three normal volunteers were injected with BTX in non- each of 4 trigger points in the trapezius muscle. Adverse events
facial regions to study its effect on sweating in a non-pathologic including confusion, increased pain across the shoulders, and
setting. Two volunteers had 20 U of Dysport injected in the
36
flu-like symptoms lasting 6 weeks devloped in 2 of the first 5
dorsum of the hand and a third had 30 U injected into the axilla. patients and the study was discontinued. The small size of the
Sweating ceased within 2 days and was still absent at 2 months study, its early discontinuation and the nature of some of the
in the hand and at 4 months in the axilla. These studies suggest side effects reported make it impossible to draw meaningful
that BTX could be a reasonable and long-lasting therapeutic al- conclusions from this effort.
ternative for severe axillary hyperhidrosis. BTX injection has been used to treat patients with chronic
pain attributed to cervical dynamic compression of the brachial
PAIN plexus or thoracic outlet syndrome. Seventy-seven limbs were
147
studied in 68 patients. Patients were diagnosed on the basis of

Most recently BTX injections have been used in the treatment history and physical examination including provocative maneu-
of pain syndromes that are likely related to increased muscle vers. Patients were also evaluated with somatosensory evoked
tension, or in pain syndromes in which the mechanism by which potentials, EMG, and nerve conduction studies. Seventeen pa-
BTX would be helpful is unclear. The most important of these tients (20 limbs) received BTX injection only. Twenty-nine pa-
disorders are myofascial pain syndrome, fibromyalgia, and tients (32 limbs) received BTX injection and surgical
chronic recurrent headache. intervention. Twenty-two patients (25 limbs) were treated with
Myofascial pain syndrome is a condition characterized by chronic, surgery only. Visual analog scales for pain and pain-related de-
focal muscle pain, stiffness, and fatigue, the pathophysiology of pression were assessed before and after treatment. The percentage
of patients reporting improvement in pain-related depression became more pronounced and persistent. Four years after symp-
was significantly greater for the BTX-only group compared to tom onset he underwent excision of a right talocalcaneonavicu-
the surgery-only group (80% and 6 4 % , respectively). Results lar osteophyte with interposition of the extensor digitorum
for the surgery-BTX group were intermediate (72%). brevis muscle. Six years after symptom onset, the patient was
Improvement in pain was similar for all groups (85% for the referred for a neurology consultation. The patient was noted to
BTX group, 88% for the surgery group, and 78% for the group have subtle posturing of the right hand with stressed gait, and
receiving both). Unfortunately no clinical or electrophysiologic decreased arm swing on the right in addition to focal dystonia
data are presented for the involved patients, treatment was not of the right foot. On note, at age 16 the patient had a history of
randomized, and the magnitude of the effect was not reported. head trauma with loss of consciousness lasting almost a month.
This study cannot be evaluated for the presence or absence of Evaluation included normal lower limb nerve conduction
confounding variables and selection bias. The clinical condition studies and needle EMG. MRI was remarkable for a small slit-
of the patients prior to treatement and the magnitude of the clin- like lesion in the left basal ganglia with hemosiderin products
ical benefit is not reported. This study contains insufficient in- suggesting distant history of hemorrhage. An angiogram was
formation to draw any conclusiomabout the efficacy of BTX for normal.
pain in purported thoracic outlet syndrome. The patient was felt to have a traumatic right hemidystonia
Some patients receiving BTX injection for the treatment of most marked in the lower limb. Pharmacologic treatment was
facial wrinkles coincidentally reported improvement in tension- initiated with Artane increased to a maximum dose of 32
type headaches. BTX was used to paralyze temporal muscles in
42
mg/day in 3 divided doses for the past several years. There was
6 patients with chronic tension headache (TH) to explore the role improvement in the focal dystonia, but the patient developed
of pericranial muscle tension in T H . Injection was unilateral,
219
drowsiness, fatigue, and lightheadness limiting upward titration.
leaving the contralateral side as a control. There was no signifi- The medical history is remarkable for Graves' disease for
cant reduction in pain intensity or pain threshold and investiga- which the patient takes synthyroid replacement medication.
tors concluded that in these muscles muscle tension played only Physical Examination. The patient is alert, oriented and co-
a minor role in headache pathogenesis. It was recommended that operative with the examination. Cranial nerve examination is
other neck and posterior head muscles be similarly studied. A remarkable for decreased mimetic movement of the right side of
28-year-old woman with a 5-year history of refractory cervico- the face. Motor examination is remarkable for 5/5 strength
genic headaches following a whiplash injury was reported to proximally and distally, slight right proximal upper limb fix and
have responded dramatically to a single BTX injection into the pronator drift. Muscle tone and mass are normal. There is nearly
symptomatic trapezius muscle in 1 9 9 7 . Injections were re-
104
constant dystonic eversion and dorsiflexion of the right foot and
quired every 3 months to maintain the benefit. In 1998 BTX was dorsiflexion of the toes that becomes more marked with walk-
revisited for refractory T H . In some patients TH seemed to
211
ing. Sensation to temperature and light touch is slightly de-
trigger secondary headaches falling in the migraine continuum. creased over the right face and arm. Deep tendon reflexes are
Four patients with predominantly TH who had failed extensive mildly increased in the right compared to the left arm without
efforts at conventional therapy were injected in symptomatic spread or clonus. Ankle reflexes are absent bilaterally. Plantar
areas. There was a reduction in associated myalgia and a reduc- responses are bilaterally flexor. Gait is remarkable for decreased
tion in the frequency and severity of migraine-type headches. arm swing on the right and dystonic posturing of the right foot.
Eight patients with TH unresponsive to amytriptyiine and physi- Balance and coordination are normal.
cal therapy were injected with 25 U of Dysport in frontal, tempo- Nerve Conduction Studies
ral, occipital, and SCM muscles (total dose 100 U ) . 182
Each DSL S Amp DML M Amp NCV
patient maintained a headache diary used to calculate the area (ms) (rlV) (ms) (mV) (m/s)
under the headache curve (AUC) 4 weeks before and following Nerve
treatment. The AUC was significantly reduced from 404 to 196 L peroneal 3.9 2.1 46.7
following treatment. No major side effects were reported. R peroneal Surgically absent
Headache is one of the most common complaints for which pa- R Sural 3.7 10
tients seek medical attention. BTX injection will never be first- R H reflex latency 30.4 ms
line therapy, as conventional medical therapy will often be L H reflex latency 30.3 ms
effective, well-tolerated, and inexpensive. If the benefit sug- Needle Electromygraphy. Right lower limb muscles were
gested by these small open pilot studies is confirmed in larger studied with a monopolar needle.
placebo-controlled, randomized studies there will doubltless be a Muscle Rest activity Recruitment
subset of refractory patients using large quantities of expensive Vastus medialis Normal Normal
medications or requiring sufficient emergency department, Tibialis anterior Normal Normal
clinic, and inpatient services to make efficacious use of BTX Extensor hallucis longus Normal Normal
cost-effective. The application to the treatment of headache has Extensor digitorum brevis Normal Normal
potential to greatly expand the clinical use of BTX. Medial gastrocnemius Normal Normal
L3-L5 paraspinus Normal Normal
Illustrative Case: Distal Lower Limb Dystonia Impression. Normal electrodiagnostic examination aside
Reason for Referral. Foot and ankle dystonia. from the anticipated absent right peroneal motor response sec-
History. A 36-year-old patient first noted the onset of right ondary to the previous surgery. Subtle post-traumatic sympto-
foot pain and involuntary eversion of the right foot and toe ex- matic hemiparesis with right lower limb dystonia based on
tension with walking beginning at age 22. Pain was also noted history and physical examination.
in the lateral aspect of the right lower limb. The patient was Recommendations. Botulinum toxin injections of the right
treated with a heel lift and physical therapy for a number of lower limb were recommmended to permit reduction or discon-
years without symptom resolution. Over time, the symptoms tinuation of pharmacologic therapy at age 29.
C h a p t e r 13 CHEMICAL DENERVATION — 5 0 9
Treatment Course. The patient was started with an injection 4. Aoki R, Merlino G, Spanyannis A, Wheeler L: Botox (botulinum toxin type A)
purified neurotoxin complex prepared from the new bulk toxin retains the same
of 50 U of Botox in the right peroneus tertius muscle with EMG
preclinical efficacy as the original but with reduced immunogenicity. Neurology
guidance. This resulted in decreased pain and reduced require- 1999;52(Suppl 2):A519-A520.
ment for Artane. After 3 injection cycles, the peroneus tertius 5. Arens LJ, Leary PM, Goldschmidt RB: Experience with botulinum toxin in the
(PT) dose was reduced to 40 U and 25 U doses were initiated in treatment of cerebral palsy. SAMJ 1997;87:1001-1003.
6. Aronson AE, DeSanto LW: Adductor spasmodic dysphonia three years after re-
peroneus longus (PL) and extensor digitorum longus (EDL). current laryngeal nerve section. Laryngoscope 1983;93:1-8.
This produced much better control of the dystonia initially with- 7. Ashton AC, Dolly JO: Microtuble dissociating drugs and A23187 reveal differ-
out clinically significant weakness and permitted the patient to ences in the inhibition of synaptosomal transmitter release by botulinum toxins
types A and B. J Neurochem 1991;56:827-835.
discontinue Artane altogether at peak benefit. Because of patient
8. Bain PG, Gregory R., Hyman N: Treatment of Parkinsonian tremor with botu-
discomfort with the injections, Botox injection to the PL muscle linum toxin: Results of a pilot study. Presented to the Dystonia Forum in
was temporarily delayed. However, injections without this London, 1994. Cited in Findley LJ: Tremor. In Moore P (ed): Handbook of
muscle proved less effective. The PL injection was reinstated. Botulinum Toxin Treatment. Oxford, Blackwell Science, 1995, pp 248-262.
9. Bakry N, Kamata Y, Simpson LL: Lectins from Triticum vulgaris and Umax
The combination of 25 U doses in EDL and PL after several in- flavus are universal antagonists of botulinum neurotoxin and tetanus toxin. J
jection cycles resulted in footdrop persisting between injections. Pharmacol ExpTher 1991;258:830-836.
EDL and PL doses were lowered. The patient has been managed 10. Benecke R: Botulinum toxin for spasm and spasticity in the lower limbs. In
consistently and effectively with injections every 3 months of 40 Jankovic J, Hallett M (eds): Therapy with Botulinum Toxin. New York, Marcel
Dekker, 1994, pp 557-565.
U in PT, 20 U in EDL, and 15 U in PL with modest weakness in 11. Berardelli A, Mercouri B, Priori A: Botulinum toxin for facial-oral-mandibular
toe extension that is not functionally significant. Injections are spasms and bruxism. In Jankovic J, Hallett M (eds): Therapy with Botulinum
repeated every 3 months. The patient usually notices some return Toxin. New York, Marcel Dekker, 1994, pp 361-367.
12. Bhakta BB, Cozens JA, Bamford JM, Chamberlain MA: Use of botulinum
of dystonia approximately a week before injection.
toxin in stroke patient with severe upper limb spasticity. J Neurol Neurosurg
Comment. This patient had a symptomatic mild hemiparesis Psychiatry 1996;61:30-35.
and hemidystonia most prominent in the right lower limb due to 13. Bickerton LE, Agur AMR, Ashby P: Flexor digitorum superficialis: Locations
of individual muscle bellies for botulinum toxin injections. Muscle Nerve
a basal ganglia lesion that was most likely caused by the signif- 1997;20:1043
icant head trauma suffered in adolescence. This patient was re- 14. Biller HF, Som ML, Lawson W: Laryngeal nerve crush for spastic dysphonia.
sponsive to systemic medication at a tolerable dose for many Ann Otol Rhinol Laryngol 1983;92:469^169.
years. As the dystonia became more prominent and persistent 15. Black JD, Dolly JO: Selective location of acceptors for botulinum neurotoxin A
in the central and peripheral nervous systems. Neuroscience 1987;23:767-779.
j over time, the patient could not tolerate high enough doses of 16. Blaustein RO, Germann WJ, Finkelstein A, DasGupta BR: The N-terminal half
anticholinergic medication to control his symptoms. Other sys- of the heavy chain of botulinum toxin A neurotoxin forms channels in planar
temic medications might have been tried, but most of these phospholipid bilayers. FEBS Lett 1987;226:115-120.
17. Blitzer A, Brin MF, Fahn S, et al: Botulinum toxin for the treatment of spastic
drugs have significant side effects and they are only effective in dysphonia as part of a trial of toxin injections for the treatment of other cranial
a small proportion of dystonia patients. Use of botulinum toxin dystonias. Laryngoscope 1986;96:1300-1301.
provided this patient with a high level of comfort and control 18. Blitzer A, Brin MF, Keen MS, Aviv JE: Botulinum toxin for the treatment of hy-
without need for systemic medication. Botulinum toxin at opti- perfunctional lines of the face. Arch Otolaryngol Head Neck Surg 1993; 119:
1018-1022.
mal titration requires high enough doses to control the toe ex- 19. Blitzer A, Brin MF: The evaluation and management of abductor laryngeal dys-
tension dystonia. This results in some weakness that is not tonia. In Jankovic J, Hallett M (eds): Therapy with Botulinum Toxin. New York,
functionally significant for the patient. Overall the patient is Marcel Dekker, 1994, pp 4 5 1 ^ 5 9 .
20. Bohlega S, Chaud P, Jacob PC: Mov Disord 1995; 10:399.
free of pain and medication side effects with improved function
21. Borg-Stein J, Pine JM, Miller JR, Brin MF: Botulinum toxin for the treatment
in walking, running, and exercise. of spasticity in multiple sclerosis. Am J Phys Med Rehabil 1993;72:364-368.
22. Borodic GE, Joseph M, Fay L, et al: Botulinum toxin A for the treatment of
spasmodic torticollis: Dysphagia and regional toxin spread. Head Neck 1990;
12:392-398.
CONCLUSION 23. Borodic GE, Pearce LB, Smith K, Joseph M: Botulinum A toxin for spasmodic
torticollis: Multiple vs. single injection points per muscle. Head Neck
The 1990 NIH consensus report on botulinum toxin found its 1992;14:33-37.
24. Borodic GE, Ferrante RJ, Pearce LB, Alderson K: Pharmacology and histology
use to be safe and effective in the treatment of just six condi- of botulinum toxin. In Jankovic J, Hallett M (eds): Therapy with Botulinum
tions: strabismus, blepharospasm, hemifacial spasm, adductor Toxin. New York, Marcel Dekker, 1994, pp 119-157.
spasmodic dysphonia, jaw-closing oromandibular dystonia, and 25. Borodic GE: Hemifacial spasm: Evaluation and management with empahsis on
botulinum toxin therapy. In Jankovic J, Hallett M (eds): Therapy with
cervical dystonia. In the intervening years experience with its
48
Botulinum Toxin. New York, Marcel Dekker, 1994. pp 331-351.

use for these conditions and other related conditions has led to 26. Brans JWM, de Boer I, Arimideh Meal: Botulinum toxin in cervical dystonia:
its novel use for many others. While much more work is needed Low dosage with electromyographic guidance. J Neurol 1995;242:529-534.
to prove the safety and efficacy for many of these conditions, 27. Brashear A, Lew MF, Dykstra DD, et al: Safety and efficacy of NeuroBloc (bot-
ulinum toxin type B) in type A-responsive cervical dystonia. Neurology
myriad physiologic processes mediated by ACh release will 1999;53:1439-1445.
assure expanding oppportunities for use of this novel medica- 28. Brin MF, Blitzer A, Herman S, Stewart C: Oromandibular dystonia: Treatment
tion to relieve pain and improve quality of life. of 96 patients with botulinum toxin type A. In Jankovic J, Hallett M (eds):
Therapy with Botulinum Toxin. New York, Marcel Dekker, 1994, pp 429-435.
29. Brin MF: Botulinum toxin: Chemistry, pharmacology, toxicity, and immunol-
ogy. Muscle Nerve 1997;20(Suppl 6):S146-S168.
30. Brin MF: Spasticity: Etiology, evaluation, management, and the role of botu-
REFERENCES linum toxin type A. (preface). Musle Nerve 1997; Suppl 6.
31. Brin MF: Cervical dystonia syllabus for treatment of dystonia: Workshop
1. Alderson K, Holds JB, Anderson RL: Botulinum-induced alteration of nerve- demonstrating the use of botulinum toxin, presented by the American Academy
muscle interactions in the human orbicularis oculi following treatment for ble- of Neurology, New Orleans, 1999.
pharospasm. Neurology 1991;41:1800-1805. 32. Brin MF, Lew MF, Adler CH, et al: Safety and efficacy of NeuroBloc (botu-
2. Alexander GE. Hamilton M: Carbamazepine for hemifacial spasm. Neurology linum toxin type B) in type A-resistant cervical dystonia. Neurology 1999;
2000;32:286-287. 53:1431-1438.
3. Anderson TJ: Spasmodic torticollis. In Moore P (ed): Handbook of Botulinum 33. Brookes GB: Laryngeal dystonia. In Moore P (ed): Handbook of Botulinum
Toxin Treatment. Oxford, Blackwell Science, 1999, pp 103-130. Toxin Treatment. Oxford, Blackwell Science, 1995, pp 181-205.
34. Brown MC, Goodwin GM, Ironton R: Prevention of motor nerve sprouting in 65. Duchen LW: Changes in the electron microscopic structure of slow and fast
botulinum toxin poisoned soleus muscles by direct stimulation of the muscle. J skeletal muscle fibers of the mouse after local injection of botulinum toxin. J
Physiol 1977;267:42P-43P. Neurol Sci 1971;14:61-74.
35. Bushara KO, Park DM: Botulinum toxin and sweating. J Neurol Neurosurg 66. Dunant Y, Loctin F, Marsal J, et al: Energy metabolism and quantal acetyl-
Psychiatry 1994;57:1437-1438. choline release: Effects of botulinum toxin, l-flouro-2,4-dinitrobenzene and di-
36. Bushara O, Jones JW, Park DM, Schutta HS: Botulinum toxin and sweating. amide in the Torpedo electric organ. J Neurochem 1988;50:431-439.
Mov Disord 1995;10:391-391. 67. Dunne J, Hayes M, Cameron D: Botulinum toxin A for cricopharyngeal dysto-
37. Byrnes ML, Thickbroom GW, Wilson SA, et al: The corticomotor representa- nia. Lancet 1993;342:559-559.
tion of upper limb muscles in writer's cramp and changes following botulinum 68. Dutton JJ: Acute and chronic effects of botulinum toxin in the management of
toxin. Brain 1998;121:977-988. blepharospasm. In Jankovic J, Hallett M (eds): Therapy with Botulinum Toxin.
38. Callahan A: Blepharospasm with resection of part of orbicularis nerve supply. New York, Marcel Dekker, 1994, pp 199-207.
Arch Ophthalmol 1963;70:508-511. 69. Dykstra DD: Effects of botulinum toxin type A on detrusor-sphincter dyssyner
39. Capra NF, Bernanke JM, Porter JD: Ultrastructural changes in the masseter in spinal cord injury patients. In Jankovic J, Hallett M (eds): Therapy with
muscle of Macaca fascicularis resulting from intramuscular injections of botu- Botulinum Toxin. New York, Marcel Dekker, 1994, pp 535-541.
linum toxin type A. Arch Oral Biol 1991;36:827-836. 70. Eleopra R, Tugnoli V, Rosseto O, et al: Botulinum neurotoxin serotype C: A
40. Carpenter EB, Seitz DG: Intramuscular alcohol as an aid in the management of novel effective botulinum therapy in human. Neurosci Lett 1997;224:91-94.
spastic cerebral palsy. Dev Med Child Neurol 1980;22:497-501. 71. Ellis AF, Tan AK: Cosmetic upper facial rejuvenation with botulinum. J
41. Carruthers A, Carruthers JDA: Botulinum toxin in the treatment of glabellar Otolaryngol 1997;26:92-96.
frown lines and other facial wrinkles. In Jankovic J, Hallett M (eds): Therapy 72 Elston JS: Idiopathic blepharospasm, hemifacial spasm, and therapeutic ptosis
with Botulinum Toxin. New York, Marcel Dekker, 1994, pp 577-595. induction. In Moore P (ed): Handbook of Botulinum Toxin Treatment. Oxford,
42. Carruthers A, Langtry JAA, Carruthers J, Robinson G: Improvement of tension- Blackwell Science, 1995, pp 90-100.
type headache when treating wrinkles with botulinum A toxin. Headache 73. Emmerson J: Botulinum toxin for spasmodic torticollis in a patient with myas-
1999;39:662-665. thenia gravis. Mov Disord 1994;9:937-937.
43. Chambers HG: The surgical treatment of spasticity. Muscle Nerve 1997;(Suppl 74. Filippi GM, Errico P, Santarelli R, et al: Botulinum A toxin effects on rat jaw
6):S121-S125. muscle spindles. Acta Otolaryngol 1993;! 13:400-404.
44. Cheshire WP, Abashian SW, Mann DJ: Botulinum toxin in the treatment of my- 75. Finkelstein A: Channels formed in phospholipid bilayer membranes by dipthe-
ofascial pain syndrome. Pain 1994;59:65-69. ria, tetanus, botulinum and anthrax toxin. J Physiol (Paris) 1990;84:188-190.
45. Coffield JA, Considine RV, Simpson LL: The site and mechanism of action of 76. Finisterer J, Muellbacher W, Mamoli B: Yes/yes head tremor without appendic-
botulinum neurotoxin. In Jankovic J, Hallett M (eds): Therapy with Botulinum ular tremor after bilateral cerebellar infarction. J Neurol Sci 1996;139:242-245.
Toxin. New York, Marcel Dekker, 1994, pp 3-13. 77. Finsterer J, Fuchs I, Mamoli B: Quantitative electromyography-guided treat-
46. Cole RA, Cohen LG, Hallett M: Treatment of muscian's cramp with botulinum ment of cervical dystonia. Clin Neuropharmachol 1997;20:42-48.
toxin. Med Probl Perf Art 1991;6:137-143. Cited in Karp BI, Hallett M: 78. Ford CN, Bless DM, Lowery JD: Indirect Iaryngoscopic approach for injection
Botulinum toxin treatment of focal hand dystonia. In Jankovic J, Hallett M of botulinum toxin in spasmodic dysphonia. Otolaryngol Head Neck Surg
(eds): Therapy with Botulinum Toxin. New York, Marcel Dekker, 1994, pp 1990;103:752-758.
299-306. 79. Ford CN: Indirect Iaryngoscopic approach for injection of botulinum toxin in
47. Cornelia C, Buchman AS, Tanner CM, et al: Botulinum toxin injection for spas- spasmodic dysphonia. In Jankovic J, Hallett M (eds): Therapy with Botulinum
modic torticollis: Increased magnitude of benefit with electromyographic assis- Toxin. New York, Marcel Dekker, 1994, pp 461-465.
tance. Neurology 1992;42:878-882. 80. Garland DE, Lucie RS, Waters RL: Curent uses of open phenol nerve block for
48. Consensus Statement: National Institutes of Health Consensus Development adult acquired spasticity. Clin Orthop 1982;165:217-222.
Conference No. 8. Bethesda, MD. National Institutes of Health, 1990. 81. Gelb DJ, Yoshimura DM, Olney RK, et al: Change in the pattern of muscle activity
49. Cosgrove AP, Graham, HK: Cerebral palsy. In Moore P (ed): Handbook of Bot- following boulinum toxin injections for torticollis. Ann Neurol 1991;29:370-376.
ulinum Toxin Treatment. Oxford, Blackwell Sciences, 1995, pp 222-247. 82. Gibson HB: The Gibson Spiral Maze Test: Retest data in relation to behavioral
50. Crary M, Glowasky A: Using botulinum toxin A to improve speech and swal- disturbance, personality and physical measures. Br J Psychol 1969;60:523-528.
lowing function following total laryngectomy. Arch Otolaryngol Head Neck 83. Giladi N: The mechanism of action of botulinum toxin type A in focal dystonia
Surg 1996;122:760-763. is probably through its dual effect on efferent (motor) and afferent pathways at
51. Cruz OA, Flynn JT: Strabismus, other therapies. In Jankovic J, Hallett M (eds): the injected site. J Neurol Sci 1997;152:132-135.
Therapy with Botulinum Toxin. New York, Marcel Dekker, 1994, pp 377-385. 84. Gilio F, Curra A, Lorenzano C, et al: Effects of botulinum toxin type A on intra-
52. Cull-Candy SG, Lundh H, Thesleff S: Effects of botulinum toxin on neuromus- cortical inhibition in patients with dystonia. Ann Neurol 2000;48:20-26.
cular transmission in the rat. J Physiol 1976:260:177-203. 85. Girlanda P, Vita G, Nicolosi C, et al: Botulinum toxin therapy: Distant effects
53. DasGupta BR, Boroff DA: Separation of toxin and hemagglutinin from chrys- on neuromuscular transmission and autonomic nervous system. J Neurology
talline toxin of Clostridium botulinum toxin type A by anion exchange chro- Neurosurg Psychiatry 1992;55:844-845.
matography and determination of their dimensions by gel filtration. J Biol 86. Gonnering RS: Pharmacology of botulinum toxin. Int Ophthalmol Clin
Chem 1968;243:1065-1072. 1993;33:203-226.
54. Das Gupta BR, Sugiyama H: A common subunit structure in Clostridium botu- 87. Gordon K, Cadera W, Hinton G: Sucessful treatment of hereditary trembling
linum type A, B and E toxins. Biochem Biophys Res Commun 1972;48: chin with botulinum toxin. J Child Neurol 1993;8:154-155.
108-112. 88. Gracies JM, Elovic E, McGuire J, Simpson DM: Traditional pharmachological
55. Dedo HH: Recurrent laryngeal nerve section for spasmodic dysphonia. Ann treatments for spasticity Part I: Local treatments. Muscle Nerve 1997;(Suppl
Otol Rhinol Laryngol 1976;85:451-459. 6):S61-S91
56. Defazio G, Lepore V, Melpignano C, et al: Cervical radiculopathy following 89. Gracies JM, Nance P, Elovic E, Simpson D: Traditional pharmacologic treat-
botulinum toxin therapy for cervical dystonia. Funct Neurol 1993;8:193— ments for spasticity. Part II: General and regional treatments. Muscle Nerve
196. 1997;(Suppl 6):S92-S103.
57. De Girolami U, Anthony DC, Frosch MP: Peripheral nerve and skeletal muscle. 90. Green DC, Ward PH, Berke GS, Gerratt BR: Point-touch technique for botu-
In Cotran RS, Kumar V, Collins T (eds): Pathologic Basis of Disease. linum toxin injection for the treatment of spasmodic dysponia. Ann Otol Rhinol
Philadelphia, W.B. Saunders, 1999, pp 1269-1292. Laryngol 1992;101:883-887.
58. Deuschl G, Heinan F, Kleedorfer B, et al: Clincal and polymyographic investi- 91. Greene PE, Fahn S: Use of botulinum toxin type F injections to treat torticollis
gation of spasmodic torticollis. J Neurol 1992;239:9-15. in patients with immunitiy to botulinum toxin type A. Mov Disord 1993:8:
59. Dickson E. Shevky R: Studies on the manner in which the toxin of Clostridium 479-483.
botulinum acts upon the body. I. The effect upon the autonomic nervous system. 92. Greene PE, Fahn S, Diamond B: Development of resistance to botulinum toxin
J Exp Med I923;37:711-731. type A in patients'with torticollis. Mov Disord 1994;9:213-217.
60. Drachman DB, Romanul FCA: Effect of neuromuscular blockade on enzymatic 93. Greene WB, Dietz FR, Goldberg MJ, et al: Rapid progression of hip subluxa-
activities of muscles. Arch Neurol 1970:23:85-89. tion in cerebral palsy after selective posterior rhizotomy. J Pediatr Orthop
61. Drachman DB: Neurotrophic regulation of muscle cholinesterase: Effect of bot- 1991;11:494-497.
ulinum toxin and denervation. J Physiol 1972:226:619-627. 93a. Gui D, Cassetta E, Anastasio G, et al: Botulinum toxin for chronic anal fissure.
62. Drobik C, Laskawi R: Frey's syndrome: Treatment with botulinum toxin. Acta Lancet 1994:344:67-88.
Otolaryngol 1995;115:459-461. 94. Hallan RI, Melling J, Womack NR, et al: Treatment of anismus in intractable
63. Duchen LW, Strich SJ: The effects of botulinum toxin of the pattern of innerva- constipation with botulinum A toxin. Lancet 1988;2:714-717.
tion of skeletal muscle in the mouse. Q J Exp Physiol Cogn Med Sci 1968;53: 95. Hambleton P. Cohen HE, Palmer BJ, Melling J: Antitoxins and botulinum toxin
84-89. treatment. Br Med J 1992:304:959-960.
64. Duchen LW: Changes in motor innervation and cholinesterase localiztion in- 96. Hambleton P, Moore AP: Botulinum neurotoxins: Origin, structure, molecular
duced by botulinum toxin in skeletal muscle of the mouse: Differences between actions and antibodies. In Moore P (ed): Handbook of Botulinum Toxin
slow and fast muscles. J Neurol Neurosurg Psychiatry 1970;33:40-54. Treatment. Oxford, Blackwell Science, 1995, pp 16-27.
97. Harris CP, Alderson KA, Nebeker J, et al: Histologic features of human orbicu- 128. LeWitt PA, Trosch RM: Idiosyncratic adverse reactions to intramuscular botu-
laris oculi treated with botulinum A toxin. Arch Ophthalmol 1991; 109:393-395. linum toxin type A injection. Mov Disord 1997;12:1064-1067.
98. Hatheway CG, Dang C: Immunogenicity of the neurotoxins of Clostridium bot- 129. LeWitt PA, Kupsky W, Ferrante MA, et al: Body fluid and tissue analysis. In
ulinum. In Jankovic J, Hallett M (eds): Therapy with Botulinum Toxin. New Goetz CG, Pappert EJ (eds): Textbook of Clinical Neurology. Philadelphia,
York, Marcel Dekker, 1994, pp 93-108. W.B. Saunders, 1999, pp 458-474.
99. Hellman-Laage JE: Treatment of gustatory flushing and sweating. Acta 130. Lorentz IT, Subramniam SS, Yiannikas C: Treatment of idiopathic spasmodic
Otolaryngol 1958;49:132-143. torticollis with botulinum toxin A: A double-blind study on 23 patients. Mov
100. Helweg-Larson J, Jacobsen E: Treatment of spasticity in cerebral palsy by Disord 1991;6:145-150.
means of phenol nerve blocks of peripheral nerves. Dan Med Bull 1969; 16: 131. Ludlow CL, Nauton RF, Sedory SE, et al: Effects of botulinum toxin on speech
20-25. in adductor spasmodic dysphonia. Neurology 1988;38:1220-1225.
101. Henderson JM, Ghika JA, Van Melle G, Haller E: Botulinum toxin A in non- 132. Ludlow CL, Rhew K, Nash EA: Botulinum toxin injection for adductor spas-
dystonic tremors. Eur Neurol 1996;36:29-35. modic dysphonia. In Jankovic J, Hallett M (eds): Therapy with Botulinum
102. Herrero BA, Ecklund AE, Street CS, et al: Experimental botulism in monkeys: Toxin. New York. Marcel Dekker, 1994, pp 437^149.
A clinical pathologic study. Exp Mol Pathol 1967;6:84-95. 133. Lund S, Lundberg A, Vylicky L: Inhibitory action from the flexor reflex affer-
103. Hirano M, Ohala J: Use of hooked-wire electrodes for electromyography of the ents on transmission to la afferents. Acta Physiol Scand 1965;64:345-355.
intrinsic laryngeal muscles. J Speech Hearing Res 1969;12:362-373. 134. Macek C: Venous thrombosis results from some phenol injections. JAMA
104. Hobson DE, Gladish DF: Botulinum toxin injection for cervicogenic headache. 1983;249:1807-1807.
Headache 1997;37:253-255. 135. MacKenzie I, Burnstock G< Dolly JO: The effects of purified botulinum neuro-
105. Hoch DH, Romero-Mira M, Ehrlich BE, et al: Channels formed by botulinum, toxin type A on cholinergic, adrenergic and non-adrenergic, atropine-resistant
tetanus, and diphtheria toxins in planar lipid bilayers: Relevance to transloca- autonomic neuromuscular transmission. Neuroscience 1982;7:997-1006.
tion of proteins across membranes. Proc Natl Acad Sci USA 1985;82: 136. Magoon EH: Clincal uses of botulinum toxin: Clinical trials for strabismus. In
1692-1696. Jankovic J, Hallett M (eds): Therapy with Botulinum Toxin. New York, Marcel
106. Holds JB, Alderson K, Fogg SG, Anderson RL: Motor nerve sprouting in a Dekker, 1994, pp 371-384.
human orbicularis oculi muscle after botulinum A injection. Invest Ophthalmol 137. Marion MH, Sheehy M, Sanglia S, Soulayrol S: Dose standardization of botu-
Vis Sci 1990;31:964-967. linum toxin. J Neurol Neurosurg Psychiatry 1995;59:102-103.
107. Houser MK, Sheean GL, Lees AJ: Further studies using higher doses of botu- 137a. Marsden CD, Schachter M: Assessment of extrapyramidal disorders. Br J Clin
linum toxin type F for torticollis resistant to botulinum toxin type A. J Pharmacol 1981;11:129-151.
Neurology Neurosurg Psychiatry 1998;64:577-580. 138. Marsden CD, Sheehy MP: Writer's cramp. Trends Neurosci 1990;13:148-153.
108. Jankovic JA, Schwartz K: Botulinum toxin treatment of tremors. Neurology 139. Maurri S, Brogelli S, Alfieri G, Barontini FI: Botulinum toxin injected into or-
1991;41:1185-1188. bicularis oculi may induce an increase of intraocular pressure. Mov Disord
109. Jankovic J: Botulinum toxin in movement disorders. Curr Opin Neurol 1995;10:388-388.
1994;7:358-356. 139a. Mayer NH: Clinicophysiologic concept of spasticity and motor dysfunction in
110. Jenzer G, Mumenthaler M, Luden H, Robert F: Autonomic dysfunction in botu- adults with an upper motor neuron lesion. Muscle Nerve 1997;20(suppl 6):S 1-S6.
lism B: A clinical report. Neurology 1975;25:150-153. 140. McClaflin RR: Myofascial pain syndrome: Primary care strategies for early in-
i 111. Joo JS, Agachan F, Wolff B, et al: Initial North American experience with botu- tervention. Postgrad Med 1994;96:56-70.
linum toxin type A for treatment of anismus. Dis Col Rectum 1996;39:1107-1111. 141. McCord CD, Shore J, Putnum JR: Treatment of essential blepharospasm. II: A
' 112. Kamata Y, Kimura Y, Hiroi T, et al: Purification and characterization of the gan- modification of exposure for the muscle stripping technique. Arch Ophthalmol
glioside-binding fragment of Clostridium botulinum type E neurotoxin. 1984;102:269-273.
Biochim BiophysActa 1993;1156:213-218. 142. Merson MH, Dowell VR Jr: Epidemiologic, clinical and laboratory aspects of
113. Kanovsky P, Streitova H, Dufek J, Rektor I: Lateralization of the P22/N30 com- wound botulism. N Engl J Med 1973;289:1105-1110.
ponent of somatosensory evoked potentials of the median nerve in patients with 143. Mezaki T, Kaji R, Kohara MKJ: Development of generalized weakness in a pa-
cervical dystonia. Mov Disord 1997;12:553-560. tient with amyotrophic lateral sclerosis after focal botulinum toxin injection.
114. Kanovsky P, Streitova H, Dufek J, et al: Change in lateralization of the P22/N30 Neurology 1996;46:845-846.
cortical component of median nerve somatosensory evoked potentials in pa- 144. Miller RH, Woodson GE, Jankovic J: Botulinum toxin injection of the vocal cord
tients with cervical dystonia after sucessful treatment with botulinum toxin A. for spasmodic dysphonia. Arch Otolaryngol Head Neck Surg 1987; 113:603-605.
Mov Disord 1998;13:108-117. 145. Minor V: Ein neues Verfahren zu der klinischen Untersuchung der Schweibabson-
115. Karp BK, Hallett M: Botulinum toxin treatment of focal hand dystonia. In derung. Dtch Z Nervenheilk 1927;101:302-306. Cited in Laskawi R, Drobik C,
Jankovic J, Hallett M (eds): Therapy with Botulinum Toxin. New York, Marcel Schonebeck C: Up to date report of botulinum toxin type A treatment in patients
Dekker, 1994, pp 299-305. with gustatory sweating (Frey's syndrome). Laryngoscope 1998;108:381-384.
116. Karp BI: Treatment of limb dystonia with botulinum toxin. Treatment of dysto- 146. Molgo J, Cornelia JX, Angaut-Petit D, et al: Presynaptic actions of botulinal neu-
nia: A workshop demonstrating the use of botulinum toxin, American Academy rotoxins at vertebrate neuromuscular junctions. J Physiol (Paris) 1990;84:152-166.
Neurology, New Orleans, 1999, pp 119-132. 147. Monsivais JJ, Monsivais DB: Botulinum toxin in painful syndromes. Hand
117. Katz RT: Management of spasticity. Am J Phys Med Rehabil 1988;67:108-112. Clinic 1996;12:787-788.
118. Keen M, Blitzer A, Aviv J, et al: Botulinum toxin A for hyperkinetic facial lines: 148. Montecucco C, Schiavo G: Mechanisms of action of tetanus and botulinum
Results of a double-blind, placebo controlled study. Plast Reconstr Surg neurotoxins. Mol Microbiol 1994;13:1-8.
1994;94:94-99. 149. Mooney V, Frykman G, McLamb J: Current status of intraneural phenol injec-
119. Kessler KR, Bednecke R: The EDB test—a clinical test for the detection of" an- tions. Clin Orthop 1969;63:122-129.
tibodies to botulinum toxin type A. Mov Disord 1997;12:95-99. 150. Moore TJ, Anderson RB: The use of open phenol blocks to the motor branches
120. Koman LA, Mooney JF, Smith BP: Botulinum toxin: Potential role in the man- of the tibial nerve in adult acquired spasticity. Foot Ankle 1991;11:219-221.
agement of cerebral palsy during childhood. In Jankovic J, Hallett M (eds): 151. Moyer E, Setler P: Botulinum toxin type B: Experimental and clinical experi-
Therapy with Botulinum Toxin. New York, Marcel Dekker, 1994, pp 511-521. ence. In Jankovic J, Hallett M (eds): Therapy with Botulinum Toxin. New York,
121. Koman LA, Mooney JF: Neuromuscular blockade in the management of cere- Marcel Dekker, 1994, pp 71-85.
bral palsy. J Child Neurol 1996; 11 (Suppl. 1):S23-S28. 152. Naumann M, Toyka KV, Mansouri Taleghani B, et al: Depletion of neutralising
122. Laccourreye O, Amuscatello L, Naude C, et al: Botulinum toxin type A for antibodies resensitizes a secondary non-responder to botulinum A neurotoxin. J
Frey's syndrome: A preliminary prospective study. Ann Otol Rhinol Laryngol Neurol Neurosurg Psychiatry 1998;65:924-927.
1998;107:52-55. 153. Nishiki T, Kamata Y, Nemoto Y, et al: Identification of protein receptor for
123. Laskawi R. Drobik C, Schonebeck C: Up-to-date report of botulinum toxin type Clostridium botulinum type B neurotoxin in rat brain synaptosomes. J Biol
A treatment in patients with gustatory sweating (Frey's syndrome). Laryngos- Chem 1994;269:10498-10503.
cope 1998;108:381-384. 154. O'Brien CF: Injection techniques using EMG and ES. Muscle Nerve
124. Lee J, Harris S, Cohen J, et al: Results of a prospective randomized trial of bot- 1997;(Suppl6):S176-S179.
ulinum toxin therapy in acute unilateral sixth nerve palsy. J Pediatr Ophthalmol 155. Odergren T, Hjaltason H, Kaakkola S, et al: A double-blind randomised, paral-
Strabismus 1994;31:286 lel group study to investigate the dose equivalence of Dysport and Botox in the
125. Lee JP: Strabismus and other ocular motility disorders. In Moore P (ed): treatment of cervical dystonia. J Neurol Neurosurg Psychiatry 1998;64:6-12.
Handbook of Botulinum Toxin Treatment. Oxford, Blackwell Science 1995, pp, 156. O'Hanlan JT, Galford HR, Bosley J: The use of 45% alcohol to control spastic-
71-87. ity. VA Med Monthly 1969;96:429^136.
126. Lennerstrand G, Nordb OA, Tian S. et al: Treatment of strabismus and nystag- 157. Pahwa R, Busenbark K, Swanson-Hyland EF, et al: Botulinum toxin treatment
mus with botulinum toxin type A: An evaluation of effects and complications. of essential head tremor. Neurology 1995;45:822-824.
Acta Ophthalmol Scand 1998;76:27-37. 158. Pamphlett R: Early terminal and nodal sprouting of motor axons after botu-
127. Lew MF, Adornato BT, Duane DD, et al: Botulinum toxin type B (BotB): A linum toxin. J Neurol Sci 1989;92:181-192.
double-blind, placebo-controlled, safety and efficacy study in cervical dystonia. 159. Pasricha PJ, Ravich TR, Kalloo AN: Botulinum toxin for achalasia. Lancet
Neurology 1997;49:701-707. 1993;8839:244-245.
160. Pasricha PJ, Kalloo AN: Botulinum toxin in the treatment of gastrointestinal 190. Siatkowski RM, Tyutyunikov A, Biglan AW, et al: Serum antibody production
disorders. In Jankovic J, Hallett M (eds): Therapy with Botulinum Toxin. New to botulinum A toxin. Ophthalmology 1993; 1861-1866.
York, Marcel Dekker, 1994, pp 543-555. 191. Simpson DM: Clinical trials of botulinum toxin in the treatment of spasticity.
161. Pavesi G, Bianti R, Medici D, et al: Botulium toxin type A in the treatment of Muscle Nerve 1997;(Suppl 6):S169-S175.
upper limb spasticity among patients with traumatic brain injury. J Neurol 192. Simpson LL: Kinetic studies on the interaction between botulinum toxin type A on
Neurosurg Psychiatry 1998;64:419^120. the cholinergic neuromuscular junction. J Pharmacol ExpTher 1980;212:16-21.
162. Pestronk A, Drachman DB, Griffen JW: Effect of botulinum toxin on trophic 193. Simpson LL: The study of clostridial and related toxins: The search for unique
regulation of acetylcholine receptors. Nature 1976;264:787-789. mechanisms and common denominators. J Physiol (Paris) 1990;84:142-151.
163. Petit H, Wiart L, Gaujard E, et al: Botulinum toxin treatment for detrussor- 194. Snellin LC: Postsnaptic effects of botulinum toxin at the neuromuscular junc-
sphincter dyssynergia in spinal cord disease. Spinal Cord 1998;36:91-94. tion. In Lewis GE (ed): Biomedical aspects of botulism. New York, Academe
164. Poewe W, Schelosky L, Kleedorfer B, et al: Treatment of spasmodic torticollis Press, 1981, pp 8 1 - 9 2 . Cited in Gonnering RS: Pharmacology of botulinum
with local injections of botulinum toxin: One year follow-up in 37 patients. J toxin. Int Ophthalmol Clin 1993;33:203-226.
Neurol 1992;239:21-25. 195. Speelman JD, Brans JWM: Cervical dystonia and botulinum treatment: Is elec-
165. Poewe W, Wissel J: Experience with botulinum toxin in cervical dystonia. In tromyographic guidance necessary? Mov Disord 1995;10:802-802.
Jankovic J, Hallett M (eds): Therapy with Botulinum Toxin. New York, Marcel 196. Stager SV, Ludlow CL: Responses of stutterers and vocal tremor patients to
Dekker, 1994, pp 267-287. treatment with botulinum toxin. In Jankovic J, Hallett M (eds): Therapy with
166. Porta M: A Comparative trial of botulinum toxin type A and methylpred- Botulinum Toxin. New York, Marcel Dekker, 1994, pp 481-490.
nisolone for the treatment of myofascial pain syndrome and pain from chronic 197. Stell R, Moore AP: History and current applications of botulinum toxin treat-
muscle spasm. Pain 2000;85:101-105. ment. In Moore P (ed): Handbook of Botulinum Toxin Treatment. Oxford,
167. Priori A, Berardelli A, Mereouri B, Manfredi M: Physiologic effects produced Blackwell Science, 1995, pp 3-15.
by botulinum toxin treatment of upper limb dystonia. Brain 1995;118:801-807. 198. Sumner BEH: Ultrstructural responses of the hypoglossal nucleus to the pres-
168. Pullman S: Limb dystonia: Treatment with botulinum toxin. In Jankovic J, ence in the tongue of botulinum toxin: A quantitative study. Exp Brain Res
Hallett M (eds): Therapy with botulinum toxin. New York, Marcel Dekker, 1977;30:313-321.
1994, pp 307-321. 199. Syuto B, Kubo S: Separation and characterization of heavy and light chains
169. Quinn N, Hallett M: Dose standardization of botulinum toxin. Lancet from Clostridium botulinum type C toxin and their reconstitution. J Biol Chem
1989;1:964-964. 1981;256:3712-3717.
170. Rand MJ, Whaler BC: Impairment of sympathetic transmission by botulinum 200. Takenaga S, Kawahigashi Y, Sonada Y, et al: [Treatment of spastic paraparesis
toxin. Nature 1965;208:588-591. with botulinum toxin with reference to beneficial effects, disease severity and
171. Rivest J, Lees AJ, Marsden CD: Writer's cramp: Treatment with botulinum long-term treatment]. Rinsho Shinkeigaku 1995;35:251-255. Cited in Simpson
toxin injections. Mov Disord 1991;6:55-59. DM: Clinical trials of botulinum toxin in the treatment of spasticity. Muscle
172. Rosales RL, Arimura K, Takenaga S, Osame M: Extrafusal and intrafusal muscle ef- Nerve 1997;(Suppl 6):S169-S173.
fects in experimental botulinum toxin-A injection. Muscle Nerve 1996;19:488-496. 201. Tardieu G, Tardieu C, Hariga J, Gagnard L: Treatment of spasticity by injection
173. Rosenbaum AL: Management of acute and chronic VI nerve palsy. In Jankovic of dilute alcohol at the motor point or by the epidural route. Clinical extension of
J, Hallett M (eds): Therapy with Botulinum Toxin. New York, Marcel Dekker, an experiment on the decerebrate cat. Dev Med Child Neurol 1968;10:555-568.
1994, pp 387-393. 202. Thesleff S, Molgo J, Tagerud S: Trophic relations at the neuromuscular junction as
174. Rosenfield DB: Clinical aspects of motor speech compromise. In Jankovic J, revealed by the use of botulinum neurotoxins. J Physiol (Paris) 1990,84:167-173.
Hallett M (eds): Therapy with Bothulinum Toxin. New York, Marcel Dekker, 203. Trosch R, Pullman SL: Botulinum toxin A injections for the treatment of hand
1994, pp 397-415. tremors. Ann Neurol 1992;32:250-250.
175. Sampaio C, Castro-Caldas A, Sales-Luis MI, et al: Brachial plexopathy after 204. Trosch RM, Pullman SL: Botulinum toxin A injections for the treatment of
botulinum toxin administration for cervical dystonia. J Neurol Neurosurg hand tremors. Mov Disord 1994;9:601-609.
Psychiatry 1993;56:220-220. 205. Truong DD, Rontal M, Rolnick M, et al: Double-blind controlled study of botu-
176. Sanders DB, Massey EW, Buckley EG: Botulinum toxin for blepharospasm: linum toxin in adductor spasmodic dysphonia. Laryngoscope 1991; 1991:
Single-fiber EMG studies. Neurology 1986;36:545-547. 630-634.
177. Schantz EJ, Scott AB: Use of Chrystalline type A botulinum toxin in medical 206. Tsui JKC, Eisen A, Stoessl AJ, Calne DB: Double-blind study of botulinum
research. In Lewis GE (ed): Biomedical Aspects of Botulism. New York, toxin in spasmodic torticollis. Lancet 1986;2:245-247.
Academic Press, 1981, 121-128. Cited in Gonnering RS: Pharmacology of bot- 207. Tsui JKC, Bhatt M, Calne S, Calne D: Botulinum toxin in the treatment of
ulinum toxin. Int Ophthalmol Clin 1993;33:203-226. writer's cramp: A double blind study. Neurology 1993;43:183-185.
178. Schantz EJ, Johnson EA: Preparation and characterization of botulinum toxin 208. Tsui JKC: Blepharospasm and hemifacial spasm. Treatment of dystonia: A
type A for human treatment. In Jankovic J, Hallett M (eds): Therapy with workshop demonstrating the use of botulinum toxin, American Academy
Botulinum Toxin. New York, Marcel Dekker, 1994, pp 41-61. Neurology, New Orleans, 1999, pp 63-66.
179. Schiavo G, Boquet P, DasGupta BR, Montecucco C: Membrane interactions of . Tsui JKC: Blepherospasm and hemifacial spasm—syllabus for: Treatment of
tetanus and botulinum neurotoxins: A photolabelling study with photoactivat- dystonia. Workshop demonstrating the use of botulinum toxin. American
able phopholipids. J Physiol (Paris) 1990;84:180-187. Academy of Neurology, Chicago, 2000, 60-64.
180. Schiavo G, Benefenati F, Poulain B, et al: Tetanus and botulinum-B neurotoxins 209. Vieregge P, Kompf D: Brachial plexopathy after botulinum toxin injection for
block transmitter release by proteolytic cleavage of synaptobrevin. Nature cervical dystonia. J Neurol Neurosurg Psychiatry 1993;56:1338-1339.
1992;359:832-835. 210. Weigand H, Erdmann G, Wellhoner HH: I-labeled botulinum A neurotoxin:
l25
181. Schneider 1, Thumfart WF, Potoschnig C, Eckel HE: Treatment of dysfunction Pharmacokinetics in cats after intramuscular injection. Naunyn Schmiedebergs
of the cricopharyngeal muscle with botulinum A toxin: Introduction of a new Arch Pharmacol 1976;292:161-165.
non-invasive method. Ann Otol Rhinol Laryngol 1994;103:31-35. 211. Wheeler AH: Botulinum toxin A adjunctive therapy for refractory headaches
182. Schulte-Mattler WJ, Wieser T, Zierz S: Treatment of tension-type headaches associated with pericranial muscle tension. Headache 1998;38:468-471.
with botulinum toxin: A pilot study. Eur J Med Res 1999;4:183-186. 212. Wissel J, Kabus C. Wenzel R, et al: Botulinum toxin in writer's cramp: Objective re-
183. Schurch B, Hauri D, Rodic B, et al: Botulinum-A toxin as a treatment of detru- sponse evaluation in 31 patients. J Neurol Neurosurg Psychiatry 1996;61:172-175.
sor-sphincter dyssynergia: A prospective study of 24 patients with spinal cord 213. Wissel J, Masuhr F, Schelosky L, et al: Quantitative assessment of botulinum
injury. J Urol 1996;155:1023-1029. toxin treatment in 43 patients with head tremor. Mov Disord 1999;12:722-726.
184. Schurch B, Hodler J, Rodic B: Botulinum A toxin as a treatment of detrusor- 214. Wolfe F, Smythe HA, Yunus MB, et al: The American College of Rheumatology
sphincter dyssnergia in patients with spinal cor injury: MRI controlled transper- 1990 criteria for the classification of fibromyalgia. Arthritis Rheum 1990,33:
ineal injections. J Neurol Neurosurg Psychiatry 1997;63:474-476. 160-172.
185. Scott AB: Botulinum toxin injections into extraocular muscles as an alternative 215. Wood KM: The use of phenol as a neurolytic agent. Pain 1978;5:205-229.
to strabismus surgery. Ophthalmology 1980;87:1044-1049. 216. Woodson GE, Zwhner P, Murray T, Swenson M: Functional assessment of pa-
186. Scott AB, Suzuki D: Systemic toxicity of botulinum toxin by intramuscular in- tients with spasmodic dysphonia. J Voice 1994;6:338-343.
jection in the monkey. Mov Disord 1988;3:333-335. Woodson GE, Murrry T, Zwirner P, Swenson MR: Acoustic, aerodynamic, and
187. Sheean GL, Murray NM, Marsden CD: Pain and remote weakness in limbs in- videoendoscopic and endoscopic assessment of unilateral thyroarytenoid
jected with botulinum toxin A for writer's cramp. Lancet 1995;346:154-156. muscle injection with botulinum toxin for spasmodic dysphonia. In Jankovic J,
188. Shone CC, Hambleton P, Melling J: A 50-kDa fragment from the NH -terminus
2
Hallett M (eds): Therapy with Botulinum Toxin. New York, Marcel Dekker,
of the heavy subunit of Clostridium botulinum neurotoxin forms channels in 2000, pp 421-427.
lipid vesicles. Eur J Biochem 1987:167:175-180. 218. Yang KH, Sugiyama H: Purification and properties of Clostridium botulinum
189. Shone CC, Quinn CP, Wait R, et al: Proteolytic cleavage of synthetic fragments type F toxin. Appl Microbiol 1975;29:598-603.
of vesicle associated membrane protein, isoform-2 by botulinum type B neuro- 219. Zwart JA, Bovim G, Sand T, Sjaastad O: Tension headache: Botulinum toxin
toxin. Eur J Biochem 1993;217:965-971. paralysis of temporal muscles. Headache 1994;34:458-462.

Lectrodiagnostic Medicine: Second Edition

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lectrodiagnostic Medicine: Second Edition

Uploaded by

Copyright:

Available Formats

LECTRODIAGNOSTIC

H A N LEY & BELFUS, I N C , / Philadelphia

Library of Congress Cataloging-in-Publication Data

Electrodiagnostic medicine / edited by Daniel Dumitru, Machiel J. Zwarts, Anthony A.

Electrodiagnostic Medicine, 2nd edition ISBN 1-56053-433-8

Last digit is the print number: 9 8 7 6 5 4 3 2 1

To my wife, Mary, and children, Joseph, Erin, Michael,

To my wife, Anita, and our children, Jelte, Iris, and Irene.

Someday When I A m an Old Woman The Mysterious Cats

Each moment flows from the fingers

Cyra S. Dumitru Daniel Amadeus Dumitru

2 Electrical Sources and Volume C o n d u c t i o n 27

Appendix T h e Leading/Trailing Dipole M o d e l a n d Near-Field/Far-Field 54

Appendix Basic Electricity Primer 98

4 Peripheral N e r v o u s S y s t e m ' s Reaction to Injury 115

5 Nerve Conduction Studies 159

6 Special N e r v e Conduction Techniques 225

7 Needle Electromyography 257

9 S o m a t o s e n s o r y Evoked Potentials 357

10 Magnetic Stimulation of the Central and Peripheral Nervous Systems . . 415

11 Quantitative Sensory Testing: Basic Principles and Clinical Applications 429

Part III P A T I E N T CARE-RELATED ISSUES

14 T h e Electrodiagnostic M e d i c i n e Consultation: Approach and Report .. 515

15 Electrodiagnostic M e d i c i n e Pitfalls 541

16 Disorders Affecting M o t o r Neurons 581

17 F o c a l Cranial Neuropathies 653

19 Brachial Plexopathies and Proximal M o n o n e u r o p a t h i e s 777

20 L u m b o s a c r a l Plexopathies and Proximal M o n o n e u r o p a t h i e s 837

21 A p p r o a c h to Peripheral Neuropathy 885

22 Hereditary Neuropathies 899

23 Acquired Neuropathies 937

24 Focal Peripheral Neuropathies 1043

25 N e u r o m u s c u l a r Junction Disorders 1127

26 Introduction to M y o p a t h i e s and M u s c l e Tissue's Reaction to Injury . . . 1229

27 Hereditary Myopathies 1265

29 Electrodiagnostic Medicine Evaluation of Children 1433

30 A A E M Glossary of Terms in Electrodiagnostic Medicine 1449

and Illustrations of Selected Waveforms

C D - R O M : Needle Electromyography Inside back c o v e r

Clinical Applications Manager, Oxford Instruments, Hawthorne, New York

Nervous Tissue Neuromuscular Transmission

2-10% of the cell membrane's molecular composition. 13

drophilic "head" group (choline and glycerol) and a hydrophobic "tail"

In addition to relatively large molecules, charged substances

mm or ions also have difficulty crossing the cell membrane, even

readily than water. The reason ions demonstrate great diffi-

Table I -1. Relationship of Effective Diameters of Different Transported

Urea molecule 0.36 0.0006

Protein Channels. Protein ion channels are water-filled pro-

0 0 there is no movement of CI" ions because the membrane is not

responding loss of positive charges. This creates an electrical

ther K ion loss (negative Cl~ charges attract positive K

charges). Eventually equilibrium is reached when the move-

taining t w o solutions of potassium chloride (K CI") with a greater

ganic anions. In the resting state, K flows into the extracellu-

concentration (100 millimolar) on the right half compared t o the left

does n o t measure a potential difference. B.The partition is replaced

with a divider that has many K pores only (semipermeable). A volt-

difference in electrical potential opposing K efflux. The mag- +

meter now records a potential difference with the higher concentra-

4B). The barrier is considered semipermeable to K . In this +

gradient. Because more K ions exist in the concentrated solu-

Intracellular Extracellular Equilibrium