(Contemporary Neurology Series, 66) Jasper R. Daube - Clinical Neurophysiology (Contemporary Neurology Series, 66) - Oxford University Press, USA (2002)

CLINICAL NEUROPHYSIOLOGY
SECOND EDITION
Series Editor:
Sid Gilman, M.D., F.R.C.P.
William J. Herdman Professor of Neurology
Chair, Department of Neurology
University of Michigan Medical Center
Contemporary Neurology Series

19 THE DIAGNOSIS OF STUPOR AND COMA, THIRD EDITION
Fred Plum, M.D. and Jerome B. Posner, M.D.
50 MULTIPLE SCLEROSIS
Donald W. Paty, M.D., F.R.C.P.C. and George C. Ebers, M.D., F.R.C.P.C., Editors
51 NEUROLOGY AND THE LAW: PRIVATE LITIGATION AND PUBLIC POLICY
H. Richard Beresford, M.D., J.D.
52 SUBARACHNOID HEMORRHAGE: CAUSES AND CURES
Bryce Weir, M.D.
53 SLEEP MEDICINE
Michael S. Aldrich, M.D.
54 BRAIN TUMORS
Harry S. Greenburg, M.D., William F. Chandler, M.D., and Howard M. Sandier, M.D.
55 THE NEUROLOGY OF EYE MOVEMENTS, THIRD EDITION
R. John Leigh, M.D. and David S. Zee, M.D.
56 MYASTHENIA GRAVIS AND MYASTHENIC DISORDERS
Andrew Engel, M.D., Editor
57 NEUROGENETICS
Stefan-M. Pulst, M.D., Editor
58 DISEASES OF THE SPINE AND SPINAL CORD
Thomas N. Byrne, M.D., Edward C. Benzel, M.D., and Stephen G. Waxman, M.D., Ph.D.
59 DIAGNOSIS AND MANAGEMENT OF PERIPHERAL NERVE DISORDERS
Jerry R. Mendell, M.D., David R. Cornblath, M.D., and John T. Kissel, M.D.
60 THE NEUROLOGY OF VISION
Jonathan D. Trobe, M.D.
61 HIV NEUROLOGY
Bruce James Brew, M.B.B.S., M.D., F.R.A.C.P.
62 ISCHEMIC CEREBROVASCULAR DISEASE
Harold P. Adams, Jr., M.D., Vladimir Hachinski, M.D., and John W. Norris, M.D., F.R.C.P.
63 CLINICAL NEUROPHYSIOLOGY OF THE VESTIBULAR SYSTEM, THIRD EDITION
Robert W. Baloh, M.D. and Vicente Honrubia, M.D.
64 NEUROLOGICAL COMPLICATIONS OF CRITICAL ILLNESS, SECOND EDITION
Eelco F.M. Wijdicks, M.D., Ph.D., F.A.C.P.
65 MIGRAINE: MANIFESTATIONS, PATHOGENESIS, AND MANAGEMENT, SECOND EDITION
Robert A. Davidoff, M.D.
66 CLINICAL NEUROPHYSIOLOGY, SECOND EDITION
Jasper R. Daube, M.D., Editor
CLINICAL
NEUROPHYSIOLOGY
Second Edition
Edited by
JASPER R. DAUBE, M.D.
Consultant, Department of Neurology
Mayo Clinic
Professor of Neurology
Mayo Medical School
Rochester, Minnesota
OXPORD
UNIVERSITY PRESS
2002
OXPORD
UNIVERSITY PRESS
Oxford New York

Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi
Sao Paulo Shanghai Singapore Taipei Tokyo Toronto
Copyright ©2002 by Mayo Foundation for Medical Education and Research

Published by Oxford University Press, Inc.
198 Madison Avenue, New York, New York, 10016
http://www.oup-usa.org
Oxford is a registered trademark of Oxford University Press
All rights reserved. This book is protected by copyright.

No part of it may be reproduced, stored in a retrieval system,
or transmitted in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise,
without written permission from Mayo Foundation.
Library of Congress Cataoging-in-Publication Data

Clinical neurophysiology/ [edited by] Jasper R. Daube.—2nd ed.
p. ; cm.—(Contemporary neurology series ; 66)
Includes bibliographical references and index.
ISBN 0-19-514080-X
1. Electroencephalography. 2. Electromyography. 3. Nervous
system—Diseases—Diagnosis. 4. Evoked potentials (Electrophysiology) 5.
Neurophysiology. I. Daube, Jasper R. II. Series.
[DNLM: 1. Nervous System Diseases—diagnosis. 2. Electroencephalography. 3.
Electromyography. 4. Evoked Potentials. 5. Nervous System Diseases—therapy. 6.
Neurophysiology. WL 140 C641 2002]
RC386.6.E43 C585 2002
616.8'047547—dc21 2002022459
Nothing in this publication implies that Mayo Foundation endorses any of the products mentioned in this book. Care
has been taken to confirm the accuracy of the information presented and to describe generally accepted practices.
However, the authors, editors, and publisher are not responsible for errors or omissions or for any consequences from
application of the information in this book and make no warranty, express or implied, with respect to the contents
of the publication.
The authors, editors, and publisher have exerted every effort to ensure that drug selection and dosage set forth in
this text are in accordance with current recommendations and practice at the time of publication. However, in view
of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy
and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and
dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new
or infrequently employed drug.
Some drugs and medical devices presented in this publication have Food and Drug Administration (FDA) clearance
for limited use in restricted research settings. It is die responsibility of health care providers to ascertain the FDA status
of each drug or device planned for use in their clinical practice.
23456789
Printed in the United States of America
on acid-free paper
FOREWORD
Clinical neurophysiology is a mature field. Many of its procedures are logical ex-
tensions of the neurologic examination and can add information that is helpful in
making a diagnosis. Because of its usefulness, clinical neurophysiology is practiced
by a large percentage of neurologists and physiatrists. Even the neurologists and
physiatrists who do not actively practice clinical neurophysiology are expected to
understand it. Therefore, it is important for practitioners to understand the fun-
damental facts and principles of the field and to be current with key advances.
Clinical neurophysiology is also a large field. Like neurology, it encompasses a
wide spectrum of issues and illnesses, ranging from the peripheral nervous system
to the central nervous system. As in neurology, it is difficult to be an expert in all
aspects of clinical neurophysiology, and most practitioners have a focused interest
in the field. However, as is true for neurology, clinical neurophysiology has an es-
sential unity. Problems are approached physiologically with methods that measure
the electric activity of the nervous system. This is another reason for practitioners
to be acquainted with the whole field even if they practice only a part of it.
Currently, there is considerable interest and activity in clinical neurophysiology.
Numerous societies in the United States and throughout the world are devoted to
this field, and their membership is growing. The umbrella organization for these
societies, the International Federation of Clinical Neurophysiology, has members
in 52 countries. Its journal, Clinical Neurophysiology, and the many national journals
have an enthusiastic readership. There are several examining bodies for compe-
tence in clinical neurophysiology. In the United States, the American Board of Psy-
chiatry and Neurology examines for competence in the broad field, the American
Board of Electrodiagnostic Medicine examines in the area commonly known as
electromyography, and the American Board of Clinical Neurophysiology examines
in the area of electroencephalography.
Where can a physician turn to learn the basics of clinical neurophysiology and
be sure the information is up to date? When Mayo Clinic neurologists speak about
clinical neurophysiology, they speak with special authority. The Mayo Clinic has
been a central force in the United States in many areas of the field. In the area of
electromyography, Dr. Edward Lambert, a pioneer in the field, has made many ba-
sic observations that still guide current practice, and, of course, he identified an
illness that now bears his name. He has trained many leaders of modern elec-
tromyography in the United States. In electroencephalography, Dr. Reginald Bick-
ford was a pioneer and was active in many areas, including evoked potentials and
even early attempts at magnetic stimulation of the brain. Many other leaders in
electroencephalography have been at the Mayo Clinic, and four of them, in addi-
tion to Dr. Bickford, have been presidents of the American EEG Society. No one
is better suited to orchestrate the writing of a textbook on clinical neurophysiol-
ogy than Dr. Jasper Daube. Dr. Daube is well recognized internationally as an ex-
pert in electromyography; he is very knowledgeable about all areas of the subject,
basic and applied. He is an outstanding leader with a gift for organization.
For all these reasons, it is nice to see this second edition of Clinical Neurophysi-
ology. Its many chapters cover the field in a broad way. The first several chapters
discuss the basic issues of neuronal generators, biologic electricity, and measure-
ment techniques central to all areas of clinical neurophysiology. Next, the indi-
v
vi Foreword
vidual areas of the field are discussed: areas including classic electromyography,
electroencephalography, and evoked potentials and extending to autonomic ner-
vous system testing, sleep, surgical monitoring, motor control, vestibular testing,
and magnetic stimulation. The text is organized for physicians who want to know
how to make an assessment of a particular symptom, of a particular system, or of
a particular disease. There is valuable information on the use of clinical neuro-
physiologic testing in a practical setting.
Clinical neurophysiology, like all fields of medicine, is evolving. Analysis and
management of data are becoming more heavily computerized. New methods of
quantification are now possible and are being used clinically. New techniques are
being developed. Perhaps most important, increasing emphasis is placed on how
to improve patient care with better integration of clinical neurophysiologic testing;
the third section of the book is devoted to these issues. This authoritative second
edition should serve both students and practitioners, keeping them up to date about
important new advances.
Mark Hallett, M.D.

Human Motor Control Section
National Institute of Neurological Disorders and Stroke
Bethesda, Maryland
Editor-in-Chief of Clinical Neurophysiology
PREFACE
Clinical Neurophysiology is the result of more than 50 years of experience at the Mayo
Clinic in training clinicians in the neurophysiologic methods for assessing diseases
of the central and peripheral nervous systems. The lectures and handouts that were
developed initially by Doctors Reginald Bickford and Edward Lambert in elec-
troencephalography and electromyography, respectively, were the seeds of what
has grown into the far-reaching field of endeavor of clinical neurophysiology at
Mayo Clinic. The clinical neurophysiology teaching program at Mayo Clinic has
continued to evolve into a formal, unified, 2-month course in clinical neurophysi-
ology that provides trainees with the knowledge and experience needed to apply
the principles of neurophysiology clinically.
The development of clinical neurophysiology at Mayo has paralleled develop-
ments in the field of medicine at large. The expansion during the past 25 years of
neurophysiology of diseases of the central and peripheral nervous system has been
recognized by the American Board of Psychiatry and Neurology of the American
Board of Medical Specialties with a Special Qualifications Examination in Clinical
Neurophysiology and by the Accreditation Council for Graduate Medical Educa-
tion Residency Review Committee for Neurology Accreditation of postresidency fel-
lowships in Clinical Neurophysiology.1
The Mayo course in clinical neurophysiology serves as an introduction to clini-
cal neurophysiology for residents, fellows, and other trainees. The course includes
lectures, small group seminars, practical workshops, and clinical experience in each
of the areas of clinical neurophysiology. The faculty for the course consists entirely
of Mayo Clinic staff members. These staff members are the authors of the chap-
ters of this textbook.
Over the years, the material for the clinical neurophysiology course was consol-
idated from individual lecture handouts into manuals. Persons outside Mayo who
had learned about these manuals by word of mouth increasingly requested them.
The success of these manuals prompted us to publish the first edition of Clinical
Neurophysiology in 1996. The continued evolution and expansion of the field of clin-
ical neurophysiology has resulted in this second edition.
The organization of our textbook is unique: it is built around the concept of test-
ing systems within the nervous system. The book consists of three major sections.
The first section is a review of the basics of clinical neurophysiology, knowledge
that is common to each of the areas of clinical neurophysiology. The second sec-
tion considers the assessment of diseases by anatomical system. Thus, methods for
assessing the motor system are grouped together, followed by those for assessing
the sensory system, higher cortical functions, and the autonomic nervous system.
The third section explains how clinical neurophysiologic techniques are used in
the clinical assessment of diseases of the nervous system.
This second edition includes new approaches, such as those described in the new
chapter on the clinical neurophysiology of pain. The basic methods of some of the
studies and the underlying physiologic and electronic principles in clinical neuro-
physiology required little change other than some simplification and clarification.
Other areas warranted major changes. The clinical problems in which each of the
clinical neurophysiologic approaches can add to the diagnosis and management
of neurologic disease have been detailed, especially the assessment of clinical symp-
vii
viii Preface
tom complexes with electroencephalography (EEG). The discussion of pediatric

EEG disorders, ambulatory EEG, new equipment and digital analyses, magneto-
EEG, electromyographic (EMG) techniques, motor unit number estimates, my-
oclonus on surface EMG, segmental sympathetic reflex, and postural normoten-
sion has been expanded. Chapters on EMG quantification and single fiber EMG
have been reorganized, and major revisions have been made in the discussion of
sensory potentials, somatosensory evoked potentials, acoustic reflex testing, car-
diovagal function, physiologic testing of sleep, and assessment of sleep disorders.
New approaches have been expanded in each of the four chapters on monitoring
neural function during surgery, particularly with motor evoked potentials.
Rochester, Minnesota J.R.D.
REFERENCE
1. Burns R, Daube J, Royden Jones H. Clinical neurophysiology training and certification in the United
States: 2000: American Board of Psychiatry and Neurology, Neurology Residency Review Committee.
Neurology 55:1773–1778, 2000.
ACKNOWLEDGMENTS
The authors of the second edition of Clinical Neurophysiology have made my work
as editor both educational and enjoyable. Each of the authors is active in clinical
neurophysiology practice, education, and research. They bring their experiences
to bear in the chapters they have written. Thus, the task of the editor was the re-
markably easy one of organizing and coordinating the material. The editor and au-
thors appreciate the skill and professionalism of the staff of the Sections of Scien-
tific Publications and Media Support Services at the Mayo Clinic; they have had an
integral part in the development of this textbook. Mayo neurology leadership has
continued to encourage and support the Division of Clinical Neurophysiology in
its combined efforts to provide trainees with the broad background of knowledge
they will need as they enter active practice. This support has provided strong en-
couragement for the book. All the staff in the Division of Neurophysiology have
contributed in a major way to the clinical neurophysiology course on which this
textbook is based. The laboratory directors have been particularly important: Dr.
Phillip Low, chair of the Division of Clinical Neurophysiology and director of the
Autonomic Reflex Laboratory and the Nerve Physiology Laboratory; Dr. C. Michel
Harper, director of the Electromyography Laboratory; Dr. Elson So, director of the
Electroencephalographic Laboratory; Drs. Lois E. Krahn and John W. Shepard, Jr.,
co-directors of the Sleep Disorders Center; Dr. Robert Brey, director of the
Vestibular/Balance Laboratory; Dr. Joseph Y. Matsumoto, director of the Move-
ment Disorder Laboratory; Dr. Robert Fealey, director of the Thermoregulatory
Sweat Laboratory; Dr. Peter J. Dyck, director of the Morphology Laboratory and
the Peripheral Neuropathy and Research Laboratory; and Dr. Andrew G. Engel,
director of the Muscle Laboratory. Special thanks must be given to Doctors Bar-
bara Westmoreland and Robert C. Hermann, who are the major organizers of the
clinical neurophysiology course; they provide much of the teaching for the trainees
participating in the course. The support of the Mayo Foundation has been critical
in the development of new directions and unique training programs in clinical neu-
rophysiology. We acknowledge not only this support but also the help given by
many others: the trainees who have participated in our clinical neurophysiology
program and the students in our courses in continuing medical education who
have given us feedback on our teaching material, the technicians who have been
a major part of our teaching program and who have provided a helpful critique of
our activities, Jean M. Smith and the other secretarial staff who have worked dili-
gently to keep the project on track, and other physicians at our institution who
have found our help in clinical neurophysiology useful in the care of their patients.
J.R.D.
ix
This page intentionally left blank
CONTENTS
CONTRIBUTORS xxiii
SECTION 1. ANALYSIS OF ELECTROPHYSIOLOGIC WAVEFORMS
1. ELECTRICITY AND ELECTRONICS FOR CLINICAL

NEUROPHYSIOLOGY 3
Terrence D. Lagerlund
BASIC PRINCIPLES AND DEFINITIONS IN ELECTRICITY 3
CIRCUIT ANALYSIS 5
RESISTIVE-CAPACITIVE AND RESISTIVE-INDUCTIVE CIRCUITS 7
CIRCUITS CONTAINING INDUCTORS AND CAPACITORS 8
FILTER CIRCUITS 12
TRANSISTORS AND AMPLIFIERS 13
2. ELECTRIC SAFETY IN THE LABORATORY AND HOSPITAL 17

ELECTRIC POWER DISTRIBUTION SYSTEMS 17
ELECTRIC SHOCK 17
LEAKAGE CURRENT 20
ELECTRIC SAFETY PRINCIPLES AND IMPLEMENTATION 24
3. VOLUME CONDUCTION 28
PRINCIPLES 28
ELECTRIC PROPERTIES OF VOLUME CONDUCTORS 30
CALCULATING POTENTIALS IN INFINITE HOMOGENEOUS
MEDIA 30
POTENTIALS IN NONHOMOGENEOUS MEDIA 34
APPLICATIONS OF VOLUME CONDUCTION PRINCIPLES 36
4. DIGITAL SIGNAL PROCESSING 41

DIGITAL COMPUTERS IN CLINICAL NEUROPHYSIOLOGY 41
xi
xii Contents
DIGITIZATION 44
COMMON USES OF DIGITAL PROCESSING 47
AVERAGING 48
DIGITAL FILTERING 49
TIME AND FREQUENCY DOMAIN ANALYSIS 50
5. ELECTROPHYSIOLOGIC GENERATORS IN CLINICAL

NEUROPHYSIOLOGY 53
PHYSIOLOGIC GENERATORS 53
STRUCTURAL GENERATORS 58
6. CLASSIFICATION OF WAVEFORM CHARACTERISTICS 63

Jasper R. Daube
CONTINUOUS WAVEFORMS 63
EVENT RECORDING 66
7. ALTERATION OF WAVEFORMS AND ARTIFACTS 69

Jasper R. Daube
PHYSIOLOGIC ALTERATION OF WAVEFORMS 69
ARTIFACTUAL WAVEFORMS 72
SECTION 2. ELECTROPHYSIOLOGIC ASSESSMENT OF

NEURAL FUNCTION
Part A. Cortical Function
8. ELECTROENCEPHALOGRAPHY: GENERAL PRINCIPLES AND

ADULT ELECTROENCEPHALOGRAMS 77
Donald W. Klass and Barbara F. Westmoreland
CLINICAL USEFULNESS OF
ELECTROENCEPHALOGRAPHY 77
DISPLAY OF ELECTROENCEPHALOGRAPHIC ACTIVITY 78
ACTIVATION PROCEDURES 80
ARTIFACTS 82
NORMAL ELECTROENCEPHALOGRAPHIC ACTIVITY OF
ADULTS 82
BENIGN VARIANTS 87
PATHOLOGIC ACTIVITY 90
Contents xiii
ELECTROENCEPHALOGRAPHIC MANIFESTATIONS OF FOCAL

INTRACRANIAL LESIONS 94
ELECTROENCEPHALOGRAPHIC MANIFESTATIONS OF
DIFFUSE ENCEPHALOPATHIES 100
EVALUATION FOR SUSPECTED BRAIN DEATH 104
9. ELECTROENCEPHALOGRAPHY: ELECTROENCEPHALOGRAMS
OF NEONATES, INFANTS, AND CHILDREN 108
Barbara F. Westmoreland
NEONATAL ELECTROENCEPHALOGRAPHIC PATTERNS 108
DEVELOPMENTAL CHANGES DURING INFANCY, CHILDHOOD,
AND ADOLESCENCE 113
BENIGN VARIANTS IN CHILDREN 115
ABNORMALITIES 116
10. AMBULATORY ELECTROENCEPHALOGRAPHY 124

Jeffrey R. Buchhalter
INDICATIONS 124
TECHNOLOGY 125
CLINICAL APPLICATIONS 126
11. PROLONGED VIDEO ELECTROENCEPHALOGRAPHY 129

Cheolsu Shin
EQUIPMENT 130
CLINICAL APPLICATION 131
12. ELECTROENCEPHALOGRAPHIC SPECIAL STUDIES 137

QUANTITATIVE METHODS OF ELECTROENCEPHALOGRAPHIC
ANALYSIS 137
MAGNETOENCEPHALOGRAPHY 144
13. ELECTROENCEPHALOGRAPHIC RECORDINGS FOR

EPILEPSY SURGERY 146
Gregory D. Cascino
CANDIDATES FOR EPILEPSY SURGERY 147
EXTRACRANIAL ELECTROENCEPHALOGRAPHY IN PARTIAL
EPILEPSY 148
CHRONIC INTRACRANIAL MONITORING 151
ELECTROCORTICOGRAPHY IN EPILEPSY SURGERY 157
xiv Contents
14. MOVEMENT-RELATED POTENTIALS AND EVENT-RELATED

POTENTIALS 161
Joseph Y. Matsumoto
MOVEMENT-RELATED CORTICAL POTENTIALS 161
EVENT-RELATED POTENTIALS 164
Part B. Sensory Pathways
15. NERVE ACTION POTENTIALS 169

EricJ. Sorenson
CLINICAL IMPORTANCE 169
PLANNING THE STUDY 172
RADICULOPATHY 172
PLEXOPATHY 173
COMMON MONONEUROPATHIES 173
TECHNICAL FACTORS 175
HUMAN FACTORS 176
METHODS 177
MEASUREMENTS 177
16. SOMATOSENSORY EVOKED POTENTIALS 181
J. Clarke Stevens
METHODS 182
SOMATOSENSORY EVOKED POTENTIAL INTERPRETATION 190
17. BRAIN STEM AUDITORY EVOKED POTENTIALS IN
CENTRAL DISORDERS 204
John N. Caviness
AUDITORY ANATOMY AND PHYSIOLOGY 205
AUDITORY EVOKED POTENTIALS IN NORMAL SUBJECTS 205
METHODS 207
USE OF BRAIN STEM AUDITORY EVOKED POTENTIALS IN
CLINICAL PROBLEM SOLVING 207
18. AUDIOGRAM, ACOUSTIC REFLEXES, AND EVOKED
OTOACOUSTIC EMISSIONS 214
Christopher D. Bauch and Wayne O. Olsen
AUDIOGRAM 215
ACOUSTIC REFLEX 216
EVOKED OTOACOUSTIC EMISSIONS 218
Contents XV
19. BRAIN STEM AUDITORY EVOKED POTENTIALS IN

PERIPHERAL ACOUSTIC DISORDERS 222
Christopher D. Bauch
STIMULI 222
INTERPRETATION 223
ELECTRODES 224
APPLICATIONS 224
20. VISUAL EVOKED POTENTIALS 227
John N. Caviness
VISUAL SYSTEM ANATOMY AND PHYSIOLOGY v227
VISUAL EVOKED POTENTIALS IN NORMAL SUBJECTS 228
METHODS 229
GENRAL PRINCIPLES OF INTERPRETATION OF CVISUAL
EVOKED POTENTIAL 230
USE OF VISUAL EVOKED POTENTIALS IN CLINICAL
PROBLEM SOLVING 230
LOCALIZATION OF VISUAL SYSTEM LESIONS 231
Part C. Motor Pathways
21. COMPOUND MUSCLE ACTION POTENTIALS 237

Jasper R. Daube
RECORDING COMPOUND MUSCLE ACTION POTENTIALS 238
STIMULATION 240
COMPOUND MUSCLE ACTION POTENTIAL MEASUREMENTS 242
F WAVES 245
PHYSIOLOGIC VARIABLES 249
COMPOUND MUSCLE ACTION POTENTIAL CHANGES IN
DISEASE 250
FINDINGS IN PERIPHERAL NERVE DISORDERS 251
22. ASSESSING THE NEUROMUSCULAR JUNCTION WITH
REPETITIVE STIMULATION STUDIES 268
Robert C. Hermann, Jr.
ANATOMY AND PHYSIOLOGY OF THE NEUROMUSCULAR
JUNCTION 268
TECHNIQUE 270
CRITERIA OF ABNORMALITY 276
xvi Contents
RAPID RATES OF STIMULATION 276

SELECTION OF NERVE-MUSCLE COMBINATIONS 276
CLINICAL CORRELATIONS 278
23. MOTOR EVOKED POTENTIALS 282
Jasper R Daube
MOTOR EVOKED POTENTIAL STIMULATION 282
MOTOR EVOKED POTENTIAL RECORDING 285
MOTOR EVOKED POTENTIAL VARIABLES 285
APPLICATIONS 286
Part D. Assessing the Motor Unit
24. ASSESSING THE MOTOR UNIT WITH NEEDLE

ELECTROMYOGRAPHY 293
Jasper R. Daube
KNOWLEDGE BASE OF NEEDLE ELECTROMYOGRAPHY 294
CONDUCTING THE NEEDLE EXAMINATION 294
NEEDLE EXAMINATION TECHNIQUES 295
ELECTROMYOGRAPHIC ANALYSIS 297
ORIGIN OF ELECTROMYOGRAPHIC POTENTIALS 299
NORMAL ELECTROMYOGRAPHIC ACTIVITY 299
RECRUITMENT 301
SIGNAL ANALYSIS 305
ABNORMAL ELECTRIC ACTIVITY 307
25. QUANTITATIVE ELECTROMYOGRAPHY 324
CHARACTERISTICS OF THE MOTOR UNIT POTENTIAL 325
CHARACTERISTICS OF THE RECORDING EQUIPMENT 326
PROPERTIES OF MOTOR UNIT POTENTIALS WITH STANDARD
ELECTRODES 328
MOTOR UNIT POTENTIAL COMPLEXITY (PHASES, TURNS,
CROSSINGS) 330
PROPERTIES OF MOTOR UNIT POTENTIALS MEASURABLE
ONLY WITH SPECIAL ELECTRODES 332
PROPERTIES OF INTERFERENCE PATTERN 333
MANUAL ANALYSIS OF SINGLE MOTOR UNIT POTENTIALS 334
Contents xvii
COMPUTER=-ASSISTED QUANTITATIVE ANALYSIS OF MOTOR

UNIT POTENTIALS 335
AUTOMATED METHODS OF ANALYSIS OF MOTOR UNIT
POTENTIALS 337
DECOMPOSITION OF THE INTERFERENCE PATTERN INTO
CONSTITUENT MOTOR UNIT POTENTIALS 337
TURNS AND AMPLITUDE ANALYSIS OF THE INTERFERENCE
PATTERN 339
POWER-SPECTRUM ANALYSIS 340
26. SINGLE FIBER ELECTROMYOGRAPHY 343
C. Michel Harper, Jr.
TECHNIQUE 345
PITFALLS OF SINGLE FIBER ELECTROMYOGRAPHY 349
CLINICAL APPLICATIONS OF SINGLE FIBER
27. ESTIMATING THE NUMBER OF MOTOR UNITS IN A MUSCLE 358
Jasper R. Daube
MOTOR UNIT NUMBER ESTIMATE BY STANDARD
MOTOR UNIT NUMBER ESTIMATE BY STANDARD NERVE
CONDUCTION STUDIES 360
QUANTITATIVE MOTOR UNIT NUMBER ESTIMATE 361
Part £. Reflexes and Central Motor Control
28. H REFLEXES 375

Kaihtyn A. Stolp-Smith
PHYSIOLOGIC BASIS 375
TECHNIQUE 377
PEDIATRIC H REFLEXES 378
29. CRANIAL REFLEXES 382
Raymond G. Auger and]. Clarke Stevens
BLINK REFLEX 382
JAW JERK (MASSETER REFLEX) 389
MASSETER INHIBITORY REFLEX 391
xviii Contents
30. LONG LATENCY REFLEXES AND THE SILENT PERIOD 394

Joseph Y. Matsumoto
LONG LATENCY REFLEXES 394
THE SILENT PERIOD 396
31. SURFACE ELECTROMYOGRAPHIC STUDIES OF MOVEMENT
DISORDERS 399
James H. Bower and Joseph Y. Matsumoto
TECHNIQUES 400
NORMAL PATTERNS 401
TREMOR 402
MYOCLONUS 405
DYSTONIA 409
TICS, CHOREA, AND ATHETOSIS 411
SURFACE ELECTROMYOGRAPHIC STUDY OF VOLUNTARY
MOVEMENT DISORDERS 411
32. VERTIGO AND BALANCE 413
Robert H. Brey
OFFICE PROCEDURES AND HISTORY TAKING 414
PHYSICAL EXAMINATION OF THE VESTIBULAR SYSTEM 415
LABORATORY EXAMINATION: ELECTRONYSTAGMOGRAPHY
TEST BATTERY 420
OTOLITH ORGAN TESTING 427
SUPERIOR SEMICIRCULAR CANAL DEHISCENCE 427
COMPUTERIZED ROTARY CHAIR TEST (HARMONIC
ACCELERATION TESTING) 428
COMPUTERIZED DYNAMIC POSTUROGRAPHY 430
VESTIBULAR REHABILITATION 433
Part F. Autonomic Function
33. CLINICAL PHYSIOLOGY OF THE AUTONOMIC NERVOUS

SYSTEM 437
Eduardo E. Benarroch
SYMPTOMS AND DISEASES 437
GENERAL ORGANIZATION OF THE AUTONOMIC SYSTEM 438
SYMPATHETIC FUNCTION 439
SYMPATHETIC INNERVATION OF THE SKIN 440
Contents xix
MUSCLE SYMPATHETIC ACTIVITY 440

AUTONOMIC CONTROL OF HEART RATE 441
CARDIOVASCULAR REFLEXES 441
MAINTENANCE OF POSTURAL NORMOTENSION 442
34. QUANTITATIVE SUDOMOTOR AXON REFLEX TEST AND

RELATED TESTS 445
Phillip A. Low
LABORATORY EVALUATION OF AUTONOMIC FUNCTION 445
QUANTITATIVE SUDOMOTOR AXON REFLEX TEST 445
35. ADRENERGIC FUNCTION 451

Phillip A. Low
SKIN VASOMOTOR REFLEXES 451
36. THERMOREGULATORY SWEAT TEST 458

Robert D. Fealey
ROLE OF THERMOREGULATORY SWEAT TESTING: CLINICAL
SYNDROMES AND PROBLEMS EVALUATED 458
METHOD 459
NORMAL THERMOREGULATORY SWEAT DISTRIBUTIONS 462
REPORTING RESULTS 463
DIFFICULTIES AND PITFALLS IN INTERPRETATION 464
37. CARDIOVAGAL AND OTHER REFLEXES 467

Eduardo E. Benarroch and Phillip A. Low
HEART RATE RESPONSE TO DEEP BREATHING 467
HEART RATE RESPONSE TO STANDING 469
OTHER TESTS OF CARDIOVAGAL FUNCTION 469
POWER SPECTRUM ANALYSIS 470
THE VALSALVA MANEUVER 470
OTHER TESTS OF AUTONOMIC FUNCTION 474
38. ELECTROPHYSIOLOGY OF PAIN 477

Rose M. Dotson
QUANTITATIVE SENSORY TEST 477
AUTONOMIC TESTS 481
MICRONEUROGRAPHY 483
LASER EVOKED POTENTIALS 484
XX Contents
Part G. Sleep and Consciousness
39. PHYSIOLOGIC ASSESSMENT OF SLEEP 493

Peter J. Hauri, Cameron D. Harris, and Michael H. Silber
DEFINITIONS 493
EVALUATION OF SLEEP 497
ASSESSING RESPIRATION DURING SLEEP 504
ASSESSING PERIODIC LIMB MOVEMENTS 510
ASSESSING OTHER PHYSIOLOGIC VARIABLES 511
40. ASSESSING SLEEP DISORDERS 513

Peter J. Hauri, Cameron D. Harris, and Michael H. Silber
INDICATIONS FOR POLYSOMNOGRAPHY 513
PARTIAL EVALUATIONS 514
PERFORMANCE OF A SLEEP STUDY 515
DISORDERS OF EXCESSIVE DAYTIME SOMNOLENCE 516
PARASOMNIAS 517
Part H. Intraoperative Monitoring
41. CEREBRAL FUNCTION MONITORING 523

Elson L. So and Frank W. Sharbrough
TECHNICAL FACTORS IN INTRAOPERATIVE
ELECTROENCEPHALOGRAPHIC MONITORING 523
SYMMETRICAL ELECTROENCEPHALOGRAPHIC PATTERNS
DURING ANESTHESIA 527
PREOPERATIVE FOCAL ABNORMALITIES 527
FOCAL ELECTROENCEPHALOGRAPHIC CHANGES DURING
CAROTID ENDARTERECTOMY 528
ELECTROENCEPHALOGRAPHY AND OTHER MONITORING
TECHNIQUES IN RELATION TO SURGICAL STRATEGY
DURING CAROTID ENDARTERECTOMY 529
ELECTROENCEPHALOGRAPHIC MONITORING DURING
CARDIAC SURGERY 530
42. BRAIN STEM AND CRANIAL NERVE MONITORING 532

METHODS 532
APPLICATIONS 535
Contents xxi
43. SPINAL CORD MONITORING 539

Jasper R. Daube
APPLICATIONS 539
MONITORING METHODS 540
ELECTROMYOGRAPHIC AND NERVE CONDUCTION STUDIES 546
TYPES OF SURGERY 547
PRIMARY VASCULAR DISEASE 552
44. PERIPHERAL NERVOUS SYSTEM MONITORING 555
METHODS 555
APPLICATIONS 556
SECTION 3. APPLICATIONS OF CLINICAL NEUROPHYSIOLOGY:

ASSESSING SYMPTOM COMPLEXES AND DISEASE ENTITIES
45. APPLICATION OF CLINICAL NEUROPHYSIOLOGY: ASSESSING

SYMPTOM COMPLEXES 565
Jasper R. Daube and Elson L. So
CLINICAL NEUROPHYSIOLOGY IN THE ASSESSMENT OF
DISEASE 566
WHEN CLINICAL NEUROPHYSIOLOGY CAN HELP 567
SYMPTOM COMPLEXES AND NEURAL SYSTEMS 568
LOCALIZATION OF DISEASE 570
IDENTIFYING DISEASE TYPES 571
PROGNOSIS 572
ASSESSING CLINICAL DISORDERS: ASSESSMENT WITH
ELECTROENCEPHALOGRAPHY 575
ASSESSING CLINICAL DISORDERS: ASSESSMENT WITH
ELECTROMYOGRAPHY AND NERVE CONDUCTION STUDIES 583
GLOSSARY OF ELECTROPHYSIOLOGIC TERMS 599
INDEX 629
CONTRIBUTORS
RAYMOND G. AUGER, M.D. JEFFREY R. BUCHHALTER, M.D.
Consultant Consultant
Department of Neurology Section of Child and Adolescent Neurology
Mayo Clinic Mayo Clinic
Associate Professor of Neurology Associate Professor of Neurology
Mayo Medical School Mayo Medical School
Rochester, Minnesota Rochester, Minnesota
CHRISTOPHER D. BAUCH, PH.D. GREGORY D. CASCINO, M.D.

Division of Audiology Department of Neurology
Associate Professor of Audiology Professor of Neurology
JOHN N. CAVINESS, M.D.

eDUARDO E. BENARROCH, M.D.
Consultant
Consultant
Department of Neurology
Department of Neurology
Mayo Clinic
Mayo Clinic
Scottsdale, Arizona
Professor of Neurology
Associate Professor of Neurology
Mayo Medical School
Mayo Medical School
JAMES H. BOWER, M.D. JASPER R. DAUBE, M.D.

Department of Neurology Department of Neurology
Assistant Professor of Neurology Professor of Neurology
ROBERT H. BREY, PH.D. ROSE M. DOTSON, M.D.

Division of Audiology Department of Neurology
Professor of Audiology Assistant Professor of Neurology
xxiii
xxiv Contributors
ROBERT D. FEALEY, M.D. TERRENCE D. LAGERLUND, M.D., PH.D.

Assistant Professor of Neurology Associate Professor of Neurology
C. MICHEL HARPER, JR., M.D. PHILLIP A. Low, M.D.

Associate Professor of Neurology Professor of Neurology
CAMERON D. HARRIS, B.S., R.P.S.G.T. JOSEPH Y. MATSUMOTO, M.D.

Associate Consultant
Sleep Disorders Center Department of Neurology
Assistant Professor of Medicine Associate Professor of Neurology
PETER J. HAURI, PH.D.

Consultant WAYNE O. OLSEN, PH.D.
Sleep Disorders Center Emeritus Member
Consultant Division of Audiology
Section of Behavioral Medicine Mayo Clinic
Mayo Clinic Emeritus Professor of Audiology
Professor of Psychology Mayo Medical School
Mayo Medical School Rochester, Minnesota
FRANK W. SHARBROUGH, M.D.

ROBERT C. HERMANN, JR., M.D. Consultant
Consultant Department of Neurology
Department of Neurology Mayo Clinic
Mayo Clinic Professor of Neurology
Assistant Professor of Neurology Mayo Medical School
Mayo Medical School Rochester, Minnesota
CHEOLSU SHIN, M.D.

DONALD W. KLASS, M.D. Consultant
Emeritus Member Department of Neurology
Section of Electroencephalography Mayo Clinic
Mayo Clinic Associate Professor of Neurology and Assistant
Emeritus Professor of Neurology Professor of Pharmacology
Contributors XXV
MICHAEL H. SILBER, M.B., CH.B. J. CLARKE STEVENS, M.D.

Associate Professor of Neurology Professor of Neurology
KATHRYN A. STOLP-SMITH, M.D.

ELSON L. So, M.D. Consultant
Director Department of Physical Medicine and
Section of Electroencephalography Rehabilitation
Professor of Neurology Associate Professor of Physical Medicine and
Mayo Medical School Rehabilitation
Rochester, Minnesota Mayo Medical School
ERIC J. SORENSON, M.D. BARBARA F. WESTMORELAND, M.D.

Assistant Professor of Neurology Professor of Neurology
SECTION 1
Analysis of
Electrophysiologic
Waveforms
Clinical neurophysiology is an area of medical practice focused primarily on
measuring function in the central and peripheral nervous systems, includ-
ing the autonomic nervous system, and in muscles. The specialty identifies
and characterizes diseases of these areas, understands their pathophysiol-
ogy, and, to a limited extent, treats them. Clinical neurophysiology relies en-
tirely on the measurement of ongoing function—either spontaneous or in
response to a defined stimulus—in a patient. Generally, the techniques used
are noninvasive and do not require the removal of tissue for testing. Each
of the clinical neurophysiology methods measures function by recording al-
terations in physiology as manifested by changes in electrical waveforms,
electromagnetic fields, force, or secretory activities. Each of these variables
is measured as a waveform that changes over time. Electrical measurements
in which the voltage or current flow associated with neural activity is plot-
ted on a temporal basis are the most common measurements used in clini-
cal neurophysiology. Knowledge of the generation, recording, measurement,
and analysis of such waveforms is critical in learning how to perform and
apply the methods of clinical neurophysiology in the study of disease.
The first section of this introductory textbook reviews the generation,
recording, and analysis of the waveforms studied in the practice of clinical
neurophysiology. The principles of electricity and electronics needed to
make the recordings are reviewed in Chapter 1. To make the appropriate
measurements, clinical neurophysiologists rely on equipment with technical
specifications; this requires that they understand the basic principles re-
viewed in Chapter 1. All electrical stimulation and recording methods re-
quire applying electrical connections that pass small amounts of current
through human tissue. Although the risks of harm from this current flow
are small, they must be understood. The principles of electrical safety nec-
essary to minimize, reduce, or eliminate any risk are discussed in Chapter 2.
These unchanging principles are of critical importance to the practice of
clinical neurophysiology.
The rules reviewed in Chapter 1 describe the familiar forms of electricity
found in the home and business. Electrical recordings made from human
tissue are distinctly different because the electric currents are carried by
charged ions that are present throughout the tissue, rather than by electrons
in wires. Electric currents flowing throughout the human body vary with the
resistance and capacitance of the tissues. This widespread flow of electricity
is referred to as volume conduction. Volume conduction produces the unique
aspects of the generation and recording of physiologic waveforms recorded
2 Analysis of Electrophysiologic Waveforms
from human tissue. The immutable principles of volume conduction de-

scribed in Chapter 3 are applicable to the many forms of electrical record-
ing used in clinical neurophysiology, whether the waveforms are recorded
from the head (electroencephalography), nerves (nerve conduction stud-
ies), muscles (electromyography), or skin (autonomic function testing).
Measurement of current flow, or potential differences, between areas of
the body was first made using analog electronic devices that have been re-
placed entirely by digital recordings throughout clinical neurophysiology.
The basic principles of digital techniques for selecting, displaying, and stor-
ing the waveforms are described in Chapter 4.
Virtually all tissues (and the cells composing them) in the human body
have electric potentials associated with their activities. These potentials are
much larger for nerve and muscle tissue and can easily be recorded for an-
alyzing the function of these tissues and their alteration with disease. The
generators of electricity in the human body and the variety of normal and
abnormal waveforms associated with their activities are described in Chap-
ter 5. The common, basic features of all electrical waveforms, whatever their
sources, are summarized in Chapter 6. The range of alterations that occur
in these waveforms in disease and the electric artifacts that occur in associ-
ation with the physiologic waveforms are reviewed in Chapter 7.
Chapte1r 1
ELECTRICITY AND ELECTRONICS
FOR CLINICAL NEUROPHYSIOLOGY
BASIC PRINCIPLES AND DEFINITIONS IN CIRCUITS CONTAINING INDUCTORS AND

ELECTRICITY CAPACITORS
Electric Charges and Force Inductive-Capacitive Circuits
Electric Potential Inductive-Resistive-Capacitive Circuits
Electric Current and Resistors Root-Mean-Square Potentials or Currents
Capacitors Calculation of Reactance
Coils (Inductors) Calculation of Impedance and the
CIRCUIT ANALYSIS Phenomenon of Resonance
KirchhofPs First Law FILTER CIRCUITS
Rules for Seats of Electromotive Force, High-Pass Filters
Resistors, Capacitors, and Inductors Low-Pass Filters
Kirchhoff's Second Law TRANSISTORS AND AMPLIFIERS
RESISTIVE-CAPACITIVE AND RESISTIVE- Semiconductors and Doping
INDUCTIVE CIRCUITS Diodes and Rectification
Resistive-Capacitive Circuits and Time Transistors and Amplification
Constant Differential Amplifiers
Resistive-Inductive Circuits and Time SUMMARY
Constant
BASIC PRINCIPLES AND given by F — qE. Thus, the electric field can
DEFINITIONS IN ELECTRICITY be thought of as the electric force per unit
charge (E= F/q).
Electric Charges and Force Ordinary matter consists of atoms con-
taining a nucleus composed of positively
Electric charges are of two types, designated charged protons and uncharged neutrons.
positive and negative. Charges exert electric Negatively charged electrons occupy the
forces on each other: like charges repel, op- space around the nucleus, to which they nor-
posite charges attract. An electric charge mally are bound by the attractive electric
can be thought of as generating an electric force between them and the nucleus. In
field in the surrounding space. The field unionized atoms, the net charge of the elec-
around a positive charge points radially trons is equal and opposite to the charge
outward in all directions from that charge, of the nucleus, so that the atom as a whole
whereas the field around a negative charge is electrically neutral. The charge carried
points radially inward. Numerically, the by 6.25 X 1018 protons is one coulomb (the
electric force F on charge q in a region of SI [Systeme International] unit of electric
space in which there is an electric field E is charge).
3
Electric Potential the movement of positive charges and,

hence, is in the direction of the applied elec-
Energy (work) is required to move a charge tric field. A metal contains free electrons.
in an electric field because of the electric When an electric field is applied, the elec-
force acting on that charge. The energy re- trons (being negatively charged) move in
quired, U, is proportional to the charge q; the direction opposite to the electric field,
thus, it makes sense to talk about the energy but the current by convention is still taken
per unit charge. This quantity is called the to be in the direction of the field (that is,
electric potential (V = U/q) and is measured in opposite to the direction of charge move-
volts; one volt is one joule of energy per ment) . The current flowing in a conductor,
coulomb of charge. The energy required to for example, a wire, divided by the cross-
move a charge in a uniform electric field is sectional area A of that conductor is called
also proportional to the distance moved. It the current density ( J ) , that is, J= i/A.
can be shown that the difference in electric Movement of charges in an ordinary con-
potential between two points a distance / ductor is not completely free; there is fric-
apart in a region of space containing an elec- tion, which is called resistance. Many con-
tric field E is given by V= El. Electric po- ductors are linear, that is, the electric field
tential in a circuit is somewhat analogous to that causes current flow is proportional to
pressure in fluid dynamics. the current density in the conductor. The re-
To have continuous movement of charges, sistivity (p) of a substance determines how
as in an electric circuit, energy must be sup- much current it will conduct for a given ap-
plied continuously by a device such as an plied electric field and is numerically equal
electrochemical cell, or a battery of such cells to the ratio of the electric field to the cur-
(which convert chemical energy to electric rent density (p = E/J). The resistivity is a
energy), or an electric generator (which constant for any given substance. In contrast,
converts mechanical energy to electric en- the resistance R of an individual conductor
ergy). Such a device is called a seat of electro- (also called a resistor in this context), which
motive force (EMF). The electromotive force is equal to the ratio of the potential differ-
of a battery or generator is equal to the en- ence (V) across the resistor to the current
ergy supplied per unit of charge and is mea- flow i in the resistor (R = V/i, called Ohm's
sured in volts. The rate at which energy U law), depends on the geometry of the resis-
is supplied is called power (P = dU/dt); it is tor as well as on the material of which it is
measured in watts. One watt is one joule per made. Resistance is measured in volts per
second. ampere, or ohms (ft). The resistance R of a
long cylindrical conductor of length I, cross-
sectional area A, and resistivity p is given by
Electric Current and Resistors R = pi/A.
Sometimes it is more convenient to discuss
The movement of electric charges is electric the conductance of a substance, <r = I/p.
current. The current i is numerically equal This is the ratio of the current density to the
to the rate of flow of charge q(i= dq/dt) and electric field (a- = J/E). Similarly, the con-
is measured in amperes. One ampere is one ductance of a resistor is the reciprocal of the
coulomb per second. Current in a circuit is resistance (G = 1/K). Ohm's law in terms of
somewhat analogous to flow in fluid dy- conductance may be written G = i/V.
namics. A conductor is a substance that has
free charges that can be induced to move
when an electric field is applied. For exam- Capacitors
ple, a salt solution contains sodium and chlo-
ride ions. When such a solution is immersed A capacitor is a device for storing electric
in an electric field, the sodium ions move in charge; it generally consists of two charged
the direction of the field, while the chloride conductors separated by a dielectric (insu-
ions move in the opposite direction. The di- lator) . A capacitor has the property that the
rection of flow of current is determined by charge q stored is proportional to the po-
Electricity and Electronics for Clinical Neurophysiology 5
tential difference V across the capacitor: in imagination, go around the circuit in ei-
q = CVwhere C, the charge per unit poten- ther direction, adding up algebraically the
tial, is called the capacitance. Capacitance is changes in potential that we encounter, we
measured in coulombs per volt, or farads must arrive at the same potential when we
(F). return to the starting point. In other words,
the algebraic sum of the changes in poten-
tial encountered in a complete traversal of
Coils (Inductors) a circuit loop must be zero. This is Kirch-
hoff s first law.1
A coil (also known as an inductor or electro-
magnet) is a device that generates a magnetic
field when a current flows in it. Physically, it Rules for Seats of Electromotive
consists of a coil of wire that may be wrapped Force, Resistors, Capacitors,
around a magnetic core, for example, a fer- and Inductors
romagnetic substance such as iron, cobalt,
or nickel. A coil has a property, called in- To apply Kirchhoffs first law to a circuit, the
ductance, that is analogous to the mechani- following rules must be used to determine
cal property of inertia; it resists any change the algebraic signs of the potentials across
in current flow (either increase or decrease). circuit components:
More precisely, a coil is capable of generat- 1. If a seat of EMF e is traversed in the di-
ing an EMF in response to any change in the rection of the EMF (that is, from the
current flowing in it; the direction of the negative to the positive terminal), the
EMF is always such as to oppose the current change in potential is +s; in the op-
change, and numerically the potential dif- posite direction, it is — e.
ference V across a coil is proportional to the 2. If a resistor is traversed in the direction
rate of change of the current i in the coil: of the current, the change in potential
V= — L (di/dt), where Lis the inductance of is — iR, in the opposite direction, it is
the coil and the minus sign indicates that the + iR
potential is in the direction opposite to the 3. If a capacitor is traversed in the direc-
current change. Inductance is measured in tion of its positively charged plate to its
volt-seconds per ampere, or henrys (H). negatively charged plate, the change in
potential is —q/C; in the opposite di-
rection, it is + q/C.
CIRCUIT ANALYSIS 4. If an inductor is traversed in the di-
rection of the current flow, the change
A circuit is a closed loop or series of loops in potential is —L(di/dt); in the oppo-
composed of circuit elements connected by site direction, it is +L(di/dt).
conducting wires. Circuit elements include Figure 1-1 shows a simple circuit con-
a source of EMF (power supply), resistors, taining a source of EMF and a resistor, to
capacitors, inductors, and transistors. which Kirchhoffs first law may be applied to
determine the current flow, as follows:
Kirchhoff's First Law

For any loop of a circuit, the energy per unit
charge (potential) imparted to the loop
must equal the energy per unit charge dis-
sipated (principle of conservation of en-
ergy) . Stated another way, if electric poten- Thus, the current is given by the EMF di-
tial is to have any meaning, a given point can vided by the resistance. A similar analysis can
have only one value of potential at any given be made when a circuit contains multiple
time. If we start at any point in a circuit and, sources of EMF, or multiple resistors, con-
These equations are solved for the four cur-

rents as follows:
Figure 1-1. A simple electric circuit containing an elec-

tromotive force, e, and a resistor, R. a, Starting location
for application of Kirchhoff s first law to the circuit; B,
battery (seat of electromotive force); i, current. (From
Halliday D, Resnick R. Physics, Part II, Rev. 2nd ed. John
Wiley & Sons, New York, 1962, p 790. By permission of
the publisher.)
nected in series. The effective net EMF is the

algebraic sum of the individual EMFs. The
effective net resistance is the sum of the in-
dividual resistances.
Equivalent resistance:
Kirchhoff s Second Law
A node is a junction point of two or more
conductors in a circuit. The node cannot act
as a repository of electric charge; therefore,
the sum of all the current flowing into a The net current i in the circuit can be cal-
culated from Ohm's law, using an effective
node must equal the sum of all the current
flowing out of the node (otherwise, charge net resistance given as the reciprocal of the
would be constantly accumulating at the sum of the reciprocals of the three resis-
node). This is Kirchhoff s second law. By us- tances in parallel.
ing this law for each node in the circuit, to-
gether with the first law for each loop in the
circuit, any arbitrary circuit problem, no
matter how complex, can be expressed as
mathematical equations. Whether the equa-
tions can be solved and how difficult it is to
obtain a solution depend on the nature of
the circuit and the elements it contains and
on the mathematical skills available for the
task.
Figure 1-2 shows a circuit containing a
single source of EMF connected to three re-
sistors in parallel. Kirchhoff s second law is
applied to the junction of the three resistors, Figure 1-2. A circuit containing an electromotive force
and Kirchhoff s first law is applied to each (EMF), e, and three resistors (Ri, R%, R$) in parallel,
a, Junction point (node); b, junction point (node); i,
branch of the circuit, to give four indepen- current. (From Halliday D, Resnick R. Physics, Part II,
dent equations in the four current variables Rev. 2nd ed. John Wiley & Sons, New York, 1962, p 800.
i, ii, 12, and i$, as follows: By permission of the publisher.)
RESISTIVE-CAPACITIVE AND for the current to fall to l/e (37%) of its ini-
RESISTIVE-INDUCTIVE tial value; it is also the time needed for the
charge to rise to 1 — l/e (63%) of its final
CIRCUITS value; e is the base of the natural logarithm
(e= 2.718282). The time constant is equal
Resistive-Capacitive Circuits toRC.
and Time Constant When the switch shown in Figure 1-3 is
placed in position b, the direction of current
Figure 1-3 shows a resistive-capacitive (RC) flow is reversed and the capacitor is dis-
circuit containing a single source of EMF charged. Application of Kirchhoffs first law
connected to a resistor and capacitor in se- to this circuit yields a differential equation.
ries. When the switch is placed in position a, Its solution is an exponential, as follows:
the current flows in such a way as to charge
the capacitor. Because of the presence of the
resistor, the capacitor is not charged all at
once but gradually over time. When Kirch-
hoff s first law is applied to this circuit and
use is made of the fact that the current is
charging the capacitor at the rate i — dq/dt,
a differential equation results. Its solution is
an exponential curve, as follows:
In this situation, both the charge on the
capacitor and the current (slope of the curve
representing q against time) fall exponen-
tially with time with the same time constant,
RC. The current is negative in this case (that
is, it flows counterclockwise in the circuit as
the capacitor is discharged).
Resistive-Inductive Circuits
Figure 1-4A shows the exponential rise in
the charge q on the capacitor, and Figure and Time Constant
1-4J5 shows that the current i (which is the
slope of the curve representing q against A circuit containing a resistor R and an in-
time) falls exponentially to zero as the ca- ductor L (an RL circuit) as well as a source
pacitor becomes fully charged. The time con- of EMF can be studied by similar methods.
stant of the RC circuit is the time required Because of the inductance, the current flow
in this circuit does not rise immediately to
its eventual value when a switch is closed but
rather rises exponentially with a time con-
stant (time to reach 63% of final value) given
by L/R This is caused by the "inertial" effect
of the inductor. The form of the exponen-
tial rise in current is similar to the shape of
the exponential rise in charge shown for the
previous RC circuit in Figure 1-4A. Simi-
larly, when a switch bypasses the source of
Figure 1-3, A circuit containing an electromotive force EMF in this circuit, the current will not drop
(EMF), e, a resistor, R, and a capacitor, C. With the to zero immediately but will fall off expo-
switch, S, in position a, the capacitor is charged. In po- nentially in time with time constant L/R be-
sition b, it is discharged, x = Starling location for ap- cause of the effect of the inductor. The equa-
plication of Kirchhoffs first law to the circuit. (From tions describing the RL circuit are similar to
Halliday D, Resnick R. Physics, Part II, Rev. 2nd ed. John
Wiley & Sons, New York, 1962, p 802. By permission of those describing the RC circuit, with i re-
the publisher.) placing q, L replacing R, and R replacing 1/C.
Analysis of Electrophysiologic Waveforms
Figure 1-4. A, The variation of charge q, with time, t, during the charging process, B, The variatiion of current, i,
with time, t. For R= 2000 C = 1.0 uf, and = 10 V. (From Halliday D, Resnick R. Physics, Part II, Rev. 2nd ed.
John Wiley & Sons, New York, 1962, p 804. By permission of the pu
CIRCUITS CONTAINING (a), the capacitor is fully charged and there

INDUCTORS AND CAPACITORS is no current flow. All the energy present in
the circuit is stored as electric energy (U E )
in the capacitor. By stage (b), a current flow
Inductivc-Capacitive Circuits has partially discharged the capacitor but at
the same time has created a magnetic field
Figure 1-5 shows an ideal circuit containing in the inductor. The total energy is split be-
a capacitor and an inductor, an inductive- tween electric energy in the capacitor and
capacitive, or LC, circuit. The circuit is ideal magnetic energy in the inductor (UB). At
because it contains no resistance. In stage stage (c), the current flow has reached its
Figure 1-5. A simple LC circuit showing eight stages (a-h) in one cycle of oscillation. The bar graphs under each
stage show the relative amounts of magnetic (Us,) and electric (Uz) energy stored in the circuit at any time. (From
Halliday D, Resnick R. Physics, Part II, Rev. 2nd ed. John Wiley & Sons, New York, 1962, p 944. By permission of
the publisher.)
maximum and the capacitor is fully dis-

charged. The inductance effect, however,
causes the current to continue to flow, now
charging the capacitor in the opposite di-
rection, as shown in stages (d) and (e). In
stages (/) through (h), the scenario is re-
peated, with the capacitor discharging and
the current flowing in the opposite direc-
tion. Finally, stage (a) is reached again, and where the frequency / of oscillations is
the entire cycle repeats. Thus, this LC cir- given by
cuit is an oscillator, and the charge on the
capacitor is a cosine function of time; simi-
larly, the current in the circuit is a sine func-
tion of time. This is, in fact, the solution to
the equations that result from applying The time constant of this circuit is VZc.
Kirchhoff s first law to the circuit, as follows: The symbol <f> represents a phase angle that
If, on the other hand, the AC generator

supplies energy continuously at a fixed fre-
quency/ as expressed by the equation
then the current flow in the circuit also

varies sinusoidally with the frequency / of
the driving EMF,
Figure 1-6. An AC circuit containing a generator, e(t), although there generally is a phase shift (f>
inductor, L, resistor, R, and capacitor, C. (From Halli- (effectively, a time delay) between the cur-
day D, Resnick R. Physics, Part II, Rev. 2nd ed. John rent and the driving EMF. For such a circuit,
Wiley & Sons, New York, 1962, p 952. By permission of it can be shown that the amplitude of the
the publisher.)
current flow i$ is directly proportional to the
amplitude of the driving EMF e0> that is, IQ =
SQ/Z. This is analogous to the situation in a
determines how much of the initial energy direct current (DC) circuit containing a bat-
at time t = 0 is electric and how much is tery and a resistor, in which current is pro-
magnetic; for the special case shown in Fig- portional to the EMF of the battery (i= e/R,
ure 1-5 (initially, all the energy is electric), Ohm's law). The quantity in the AC circuit
4> = 0. that is analogous to resistance in a DC cir-
cuit is impedance (Z); like resistance, it is mea-
sured in ohms. Impedance is determined by
Inductive-Resistive-Capacitive all three circuit elements (L, C, and R); also,
unlike DC resistance, impedance is a func-
Circuits tion of the frequency/of the AC generator.
A more realistic circuit is shown in Figure
1-6; it contains a resistor, capacitor, and in- Root-Mean-Square Potentials
ductor. It also contains a source of alternat-
ing current EMF (an AC generator). This is
or Currents
a device that generates an EMF s(t) that Often, when measuring potentials and cur-
varies sinusoidally over time (which is exactly
rents in AC circuits that vary sinusoidally
what the electric power company generators with time, it is more convenient to deal with
do) with a frequency of/(typically 60 Hz for an average potential or current than the am-
line power in the United States). This cir- plitude (the peak positive or negative value)
cuit is governed by the following general over a cycle. A simple average is not useful,
equation:
however, because it is always zero; that is, the
values are positive for half the cycle and neg-
ative for the other half. The most useful "av-
erage" quantity that can be used is the root-
mean-square (rms) potential or current. This
is defined as the square root of the average
of the squares. Because the square of a quan-
tity, whether negative or positive, is always
If the AC generator is omitted, the circuit positive, the rms value over a cycle is non-
behaves as a damped oscillator; if the ca- zero. It can be shown that the rms value of
pacitor is initially charged up and then a sinusoidally varying quantity is equal to the
switched into the circuit, the current in the amplitude divided by V2 (approximately
circuit will vary sinusoidally over time, but its 0.707 times the amplitude). When it is said
amplitude decays exponentially to zero be- that the line voltage for electric service in
cause of power dissipation in the resistor. the United States is 120 V, an rms value is
implied; the amplitude of the voltage varia-

tion is actually ±170 V. Similarly, the rating
of a fuse or circuit breaker is an rms current
rather than current amplitude.
Calculation of Reactance
In general, impedance (Z) is made up of
three parts: the resistance (R), the reactance
of the capacitor (Xc), and the reactance of
the inductor (XL). Reactance, which is mea- Figure 1-7. A right triangle symbolizing the relation-
ship among resistance (R), inductive and capacitive re-
sured in ohms, is the opposition that a ca- actance (XL - Xc), and impedance (Z).
pacitor or inductor offers to the flow of AC
current; it is a function of frequency. The re-
actance of a capacitor may be calculated as The right triangle shown in Figure 1-7 (with
follows: impedance 2 being the hypotenuse) sym-
bolizes in geometric form the relationship
among impedance, resistance, and reactance.
Note that the capacitive reactance Xc is sub-
tracted from the inductive reactance XL in cal-
Similarly, the reactance of an inductor is calculating impedance. This, together with the
culated as follows: frequency dependence of the reactances de-
scribed above, leads to an important phe-
nomenon in AC circuits: there is always one
frequency at which the impedance of an
The reactance of a capacitor is least at LRC circuit is a minimum. This frequency
high frequencies, becomes progressively may be calculated by setting Xc = XL, be-
greater at lower frequencies, and is infinite cause when this is true, impedance Z equals
at zero frequency (DC), because an ideal ca- resistance R (the smallest possible value):
pacitor uses a perfect insulator between the
plates that is not capable of carrying any di-
rect current. (The only reason a capacitor
appears to conduct AC current is that AC
current is constantly reversing direction;
the capacitor in this case is merely being
charged, discharged, and charged again in
the opposite polarity [Fig. 1-5].) The reac- Thus, the frequency at which impedance
tance of an inductor is zero at zero frequency is a minimum is exactly equal to the fre-
(DC) and increases progressively with in- quency of oscillations of the LC or LRC cir-
creasing frequency. This happens because cuit without an AC generator (the frequency
the effect of an inductor is to oppose of the circuit shown in Figure 1-5). This can
changes in current, and the more rapidly the be restated as follows: when an LRC circuit
current changes, the greater the induced is driven by an AC source of EMF, the largest
EMF opposing that change will be. current flow occurs when the frequency of
the driving EMF is exactly equal to the nat-
ural, or resonant, frequency of the circuit.
Calculation of Impedance and the The current flow at driving frequencies
Phenomenon of Resonance above or below the resonant frequency is
less; that is, the impedance of the circuit is
After the reactances and resistances have greater. This phenomenon is known as reso-
been calculated, the impedance is calculated nance; it is exploited in all tuner circuits to
as follows: select a signal of one particular frequency
(that is, one broadcast station) and to reject
signals of all other frequencies. Similar cir-
cults can be used as narrow band-pass filters pacitor charges, and it becomes negative and
to eliminate all but a narrow range of fre- decreases exponentially when the input po-
quencies from a signal or as notch filters to tential is "turned off' and the capacitor dis-
eliminate a narrow range of frequencies charges. This accounts for the shape of the
from a signal. output in response to a square-wave calibra-
tion pulse.
Alternatively, one can imagine applying a
FILTER CIRCUITS sinusoidal AC potential to the input of this
filter circuit. As demonstrated by the fol-
High-Pass Filters lowing equations, the current i is then equal
to the input potential divided by the im-
Figure 1—8A shows a simple high-pass (low- pedance Z of the circuit, which can be cal-
frequency) filter circuit, which consists of a culated from the resistance and capacitive
capacitor C in series with and a resistor R in reactance. With some algebraic manipula-
parallel with the output circuit. The input tion, the ratio of the output to the input po-
potential Mn is applied between the input tentials can be calculated as a function of
terminal and ground, and the output po- frequency:
tential Vout is developed across the resistor
R, This circuit may be analyzed in two ways.
First, it may be treated as an RC circuit sim-
ilar to that shown in Figure 1-3, with the
input potential V{n taking the place of the
battery EMF. The output potential Vût is
proportional to the current i in the circuit.
This decreases exponentially to zero with
time constant (TC) equal to RC when the
input potential is "turned on" and the ca-
figure 1-8. A, A high-pass (low-

frequency) filter circuit. B, A low-pass
(high-frequency) filter circuit. C, Capac-
itor; R, resistor; TC, time constant.
(From Fisch BJ. Spehlmann's EEC
Primer, Rev. 2nd ed. Elsevier Science
Publishers, Amsterdam, 1991, p 56. By
permission of the publisher.)
From this formula, it is apparent that the

output is strongly attenuated at low fre-
quencies (when / is near zero) but is essen-
tially equal to the input at high frequencies
(when / is large). The cut-off frequency / of
the high-pass filter is usually specified as the
frequency at which the attenuation factor
Vout/Vin is 1/V2, or 0.707; this occurs when
/= 1/27T/RC. Equivalendy, die time constant
of die filter is given by 1/27T/ where /is die
filter cut-off frequency.
From this formula, it may be seen that die
output is attenuated at high frequencies but
becomes nearly equal to the input at low fre-
Low-Pass Filters quencies. The cut-off frequency /of die low-
pass filter is usually specified as die fre-
Figure 1-85 shows a simple low-pass (high- quency at which the attenuation factor
frequency) filter circuit, which consists of a Vout/Vin is 1/V2, or 0.707; this occurs when
resistor R in series witii and a capacitor C in /= 1/277-flC, as for the high-pass filter.
parallel with die output circuit. This is also Note that die only essential difference be-
an RC circuit, but the output potential Vout tween a high-pass filter and a low-pass filter
in this case is developed across die capacitor is die source of the output potential (to be
and is proportional to the charge on die ca- fed to the amplifier); die high-pass filter de-
pacitor. Comparison witii Figure 1-4 shows velops its output potential across die resistor
that die output potential increases expo- R, whereas the low-pass filter develops its out-
nentially to a maximum with time constant put potential across die capacitor C.
equal to RC when die input potential is
"turned on" and die capacitor charges, and
it decreases exponentially when die input
potential is "turned off' and die capacitor
discharges. This accounts for the shape of TRANSISTORS AND
die output in response to a square-wave cal- AMPLIFIERS
ibration pulse. The time constant of a high-
frequency filter is discussed less often than Semiconductors and Doping
that of a low-frequency filter in practice, be-
cause it is much smaller, for example, only Transistors are constructed of materials
2 ms for a 70-Hz filter, and cannot be mea- called semiconductors, which have resistivities
sured on electroencephalographic tracings intermediate between those of good con-
made at standard paper speeds by visual inductors (such as metals) and insulators
spection of die calibration pulse. (most nonmetals). Silicon and germanium
This filter circuit can also be analyzed in are the most frequentiy used substances.
the context of a sinusoidal AC input po- They are very poor conductors when in pure
tential. The output potential Vout is equal form, but when doped with trace quantities
to the current flow times the capacitive re- of elements capable of acting as electron
actance Xc (the equivalent of Ohm's law for donors or acceptors, diey become semicon-
a capacitor, with reactance substituting for ductors. The resistivity of die semiconductor
resistance in an AC circuit), and the cur- can be altered by controlling die doping
rent i is equal to the input potential V\n di- process.
vided by the total impedance Z, which is the Doping die tetravalent base material, sili-
same as before. Using these facts and per- con or germanium, widi a pentavalent ele-
forming some algebraic manipulation, the ment such as arsenic provides extra "free"
ratio of the output to the input potential electrons diat can conduct an electric cur-
can be calculated as a function of fre- rent. Because diese electrons carry negative
quency, as follows: charge, die semiconductor that results is re-
ferred to as an n-type semiconductor. Alterna-

tively, the base material can be doped with a
trivalent element such as gallium. An ab-
sence of sufficient electrons to fill all of the
orbitals is the result; the unfilled, or electron-
deficient, areas are called holes and behave
as positive charges that are free to move and,
thus, conduct a current. The resulting semi-
conductor is referred to as a p-type semicon- Figure 1-9. A, A forward-biased pn semiconductor junc-
ductor. (What actually happens is that elec- tion permits current flow. B, A reverse-biased junction
prevents current flow. V, Potential across the semicon-
trons from a neighboring atom move to fill ductor junction. (From Misulis.2 By permission of the
in the hole, resulting in a hole moving to a American Clinical Neurophysiology Society.)
new position.) Thus, n-type semiconductors
have electrons available for conducting cur-
rent, whereas p-type semiconductors have device, rectification results—the current flow
holes (a potential space for electrons) avail- is pulsatile but constrained to a single di-
able for conducting current. rection. This is the first step that occurs in
power supply circuits that convert AC line
voltage to DC voltages suitable for use in
Diodes and Rectification most electronic circuits.
A useful electronic device can be made when
two or more dissimilar semiconductors are Transistors and Amplification
adjacent. When an n-type semiconductor
slab is fused along one face with a p-type A transistor is a device that controls the trans-
semiconductor, electrons diffuse from the n fer of electric charge across a resistor. Junc-
region to the p region, filling some of the tion bipolar transistors (the most common
empty holes of the p region, up to the point type) can be made in two forms called npn
at which the relative attraction of the holes and pnp. Both are composed of a three-layer
for electrons is exactly counterbalanced by sandwich of semiconductors of different
the effect of the electric field set up between types. The names npn and pnp refer to the
the regions by the migration of electrons. types of semiconductors and their order in
This leaves the p region with a net negative the sandwich; npn transistors, for example,
charge and the n region with a net positive have a positively doped material in the mid-
charge. If such a device is connected in a cir- dle layer and negatively doped materials in
cuit to a source of EMF with the positive po- the outer layers. In an npn transistor, the pri-
tential applied at the p region, a process mary movement of electrons is from one of
called forward biasing, the electric field across the outer layers (the emitter), which is con-
the junction is reduced and further migra- nected to the negative terminal of an exter-
tion of electrons from n to p occurs, which nal battery or power supply, to the other
constitutes a current flowing (by conven- outer layer (the collector), which is connected
tion) from the p to the n terminal (Fig. 1-9). to the positive terminal, through the middle
If, however, the positive potential is applied layer (the base). Under resting conditions,
at the n region, a process called reverse bi- little current flow occurs, because although
asing, the electric field across the junction the emitter-base np junction is forward bi-
is actually increased and current flow is ased and can conduct current, the base-
blocked, because in the region of the junc- collector pn junction is reverse biased and
tion, the electrons that would carry current blocks current (Fig. 1—10). However, if a
in the n region have been depleted and the small positive potential is applied to the base
"holes" that would carry current in the p re- through an external connection, the junc-
gion have also been filled. tional electric field at the base-collector pn
This type of device is called a diode; it al- junction is reduced, because some electrons
lows current flow only in one direction. If a entering the p-type base are allowed to leave
sinusoidal AC potential is applied to such a through the external connection, prevent-
plifier. This type of amplifier is constructed

to amplify only the difference in the potential
between its two inputs. This is one way of re-
ducing contamination of the physiologic sig-
nal by electrical noise, for example, 60 Hz
noise from line voltage devices, because this
noise tends to be the same at all electrode
positions and cancels out when a difference
in potential is formed. Physiologic signals,
on the other hand, are usually different at
different electrode positions.
Figure 1-10. An npn junction bipolar transistor show- A differential amplifier is actually com-
ing the potential applied between the emitter, E, and posed of two amplifier circuits (two transis-
base, B, that controls the flow of current between the tors) , one for the Gl input and one for the
emitter and collector, C. The figure on the right shows
the circuit, using the conventional symbol for the tran- G2 input (the nomenclature Gl and G2 is
sistor. (From Misulis.2 By permission of the American still in common use, even though it origi-
Clinical Neurophysiology Society.) nated in the early days of EEG and EMG
when vacuum tube amplifiers were used; G
was used to indicate a grid in the vacuum
ing the filling of holes in the base. This al- tubes). The G2 transistor is connected to a
lows a continuous movement of electrons negative power supply voltage, and the Gl
from the emitter through the base into the transistor is connected to a positive power
collector. The middle, or base, layer is very supply voltage. The outputs of both transis-
thin, so that even a small positive potential tors are connected together (Fig. 1-11). In
applied to the base is sufficient to facilitate this fashion, increased current flow in the
a large conductance between collector and emitter-to-collector circuit of the Gl transis-
emitter. Thus, a small controlling voltage tor produces a positive change in the output
across base and emitter governs the flow of potential, whereas increased current flow in
a much larger current through the collector-
to-emitter circuit. This, in effect, is the basis
of an electric amplifier.
Because the gain, or amplification, of a
single transistor is limited, multiple stages
of amplification are used in an electroen-
cephalographic (EEG) or electromyographic
(EMG) machine. For example, six stages of
amplification, each with a gain of 10, give an
overall amplification factor of 1,000,000. A
complete EEG or EMG amplifier contains
preamplification stages, which typically boost
the signal from the patient by a factor of
10 to 1000, followed by low-frequency, high-
frequency, and 60 Hz notch filter circuits,
followed by driver amplification stages that
provide the remaining amplification and
produce an output signal capable of driving
the oscilloscope display or the pen motors
of the paper display unit. Figure 1-11. A differential amplifier constructed from
two single-ended amplifiers, with input potentials Vi
and Va with respect to the reference (Ref), the second
Differential Amplifiers of which produces an inverted output; the net output
is given by the product of the amplifier gain (10 in this
case) and the difference in potential between the two
The type of amplifier used predominantly in inputs. (From Misulis.2 By permission of the American
clinical neurophysiology is the differential am- Clinical Neurophysiology Society.)
the emitter-to-collector circuit of the G2 location of the ground electrode does not
transistor produces a negative change in the matter. Because each input of the differen-
output potential. tial amplifier receives a potential that is rel-
Although an ideal differential amplifier ative to the same ground and these poten-
would be sensitive only to the difference in tials are subtracted in the output, the
potential between the two inputs, in practice potential of the ground electrode cancels
a large enough signal applied to both inputs out. However, if there is a very poor (that is,
simultaneously, called common mode, pro- one with high impedance) electrode con-
duces a small output signal. This occurs be- nection or, in the extreme case, if an elec-
cause the input impedance and the gain and trode is left unconnected, the differential
frequency response of the two transistors in amplifier input effectively becomes the
the differential amplifier are not quite iden- ground electrode potential. In addition to
tical. The common mode rejection ratio introducing more 60 Hz and other noise
(CMRR) of a differential amplifier can be into the recorded signal, artifacts and mis-
calculated as the applied common input po- localization of cerebral electric activity can
tential divided by the output potential.2 For result by the unexpected introduction of a
modern amplifiers, this ratio is approxi- signal coming from an EEG ground elec-
mately 10,000. However, if the electrode im- trode into one or more channels.
pedances are high or differ significantly be-
tween the Gl and the G2 inputs, the effective
signal perceived by the two transistors in the SUMMARY
differential amplifier can differ significantly
and the CMRR can be drastically reduced, This chapter reviews the basic principles of
thus allowing significant amounts of noise to electric and electronic circuits that are im-
contaminate the signals being recorded. portant to clinical neurophysiology. Knowl-
Even though the differential amplifier edge of these basic principles and how to
output depends on the difference in the po- solve simple circuit problems is necessary for
tentials at its two inputs, each input poten- a complete understanding of the proper op-
tial as perceived by the amplifier is relative eration of equipment used in clinical neu-
to a common reference, or ground, potential. rophysiology and of the terminology and
This ground potential is in practice equal to specifications given in equipment manuals.
the potential at a single ground electrode
that must be attached to the patient. For
example, the ground electrode for EEG
recording is typically placed on the mastoid REFERENCES
process; occasionally other locations, such as
1. Halliday D, Resnick R. Physics, Part II, Rev. 2nd ed.
the frontal area, are used. As long as the elec- John Wiley & Sons, New York, 1962.
trode-to-patient connections are adequate 2. Misulis KE. Basic electronics for clinical neurophys-
(that is, have low enough impedance), the iology. J Clin Neurophysiol 6:41-74, 1989.
Chapter 2
ELECTRIC SAFETY IN THE
LABORATORY AND HOSPITAL
ELECTRIC POWER DISTRIBUTION Methods to Reduce Leakage Current

SYSTEMS Reaching Patients
ELECTRIC SHOCK ELECTRIC SAFETY PRINCIPLES AND
Requirements for Electric Current to Flow IMPLEMENTATION
Through the Body Equipment Grounding
Physiologic Effect of Electric Current Tests for Equipment Grounding and
Factors Reducing Risk of Electric Shock Leakage Current
Factors Increasing Risk of Electric Shock in Rules for Electric Safety
Hospitals Electric Safety Procedures for Technicians
LEAKAGE CURRENT SUMMARY
Origin
Methods by Which Leakage Current
Reaches Patients
ELECTRIC POWER line whose potential is 120 V from ground.

DISTRIBUTION SYSTEMS This hot line incorporates a circuit breaker
diat limits current flow to a level (for exam-
The electric systems of buildings are de- ple, 20 A) that will not cause excessive heat-
signed to distribute electric energy from one ing in wiring in the building. For reasons ex-
central point of entrance to all die electric plained later, each receptacle also includes
appliances and receptacles. Power compa- a ground contact connected to earth ground
nies provide electric energy at high voltage through a conductor separate from the
(typically 4800 V for a hospital or medical "neutral" conductor (see Fig. 2-1, wherein
clinic) to minimize transmission losses. A this separate conductor is the metal conduit
step-down transformer converts the high- that houses the wiring).
voltage energy to safer, usable voltages (usu-
ally 120 V and 240 V). Figure 2-1 shows die
wiring of a typical 120 V circuit. The sec- ELECTRIC SHOCK
ondary coil of the transformer has a center
tap that acts as the return path ("neutral") Electric shock is the consequence of the flow
for the circuit; it is connected to earth of current through the body. The effect of
ground through a grounding stake at the electric shock depends both on the magni-
transformer site. Each of the two outer ends tude of the current flow and the path taken
of the 240 V secondary coil can be used to by the current in the body, which is deter-
drive one 120 V circuit; this provides a hot mined by the points of entry and exit.
17
Figure 2-1. Scheme of a building's electric power distribution system showing a step-down transformer, circuit
breaker, grounding stake, and equipment ground carried by metal conduit. (From Cromwell L, Weibell FJ,
Pfeiffer EA. Biomedical Instrumentation and Measurements, 2nd ed. Prentice-Hall, Englewood Cliffs, New Jersey,
1980, p 437. By permission of the publisher.)
Requirements for Electric Current terminals, is in direct continuity with the hot
to Flow Through the Body line through a very low resistance path
caused by some fault such as a mechanical
When an externally generated current flows break in insulation or fluid spilled into the
through the body, it has a point of entrance circuit or inadvertent direct connection of
and a point of exit. The current may be electrode lead wires to energized, detach-
thought of as originating from some electric able power-line-cord plugs,1 or, more com-
apparatus, or source, flowing through a con- monly, (2) because of a low-level leakage of
ducting material from the apparatus to the current through a moderate resistance path,
body, flowing through the body, and finally which may be inherent in the design of the
flowing through a second conducting mate- apparatus. A further requirement for signif-
rial from the body to ground. Thus, to have icant electric shock is that the entire path-
an electric shock, there must be at least two way to, through, and out of the body must
connections to the body: one to the current have a sufficiently low resistance.
source and the other to ground. An appa- An additional requirement for a lethal
ratus can act as a source of current either electric shock is that the current must take
(1) because a point of connection between a path through the body that includes the
it and the body, such as an exposed metal heart (for example, when current enters
part of the chassis or other metal contacts or through one arm and exits through the
Electric Safety in the Laboratory and Hospital 19
Figure 2-2. Effects of 60 Hz AC elec-

tric current flow of various magni-
tudes produced by a 1-second exter-
nal contact with the body. (From
Cromwell L, Weibell FJ, Pfeiffer EA.
Biomedical Instrumentation and Mea-
surements, 2nd ed. Prentice-Hall,
Englewood Cliffs, New Jersey, 1980,
p 434. By permission of the publisher.)
other), because the mechanism of lethal mal perception (500 /xA) to severe burns
shock is almost invariably the induction of and physical injury (> 10 A). Because cur-
ventricular fibrillation. rent flow in a limb leads to involuntary mus-
cle contraction through direct depolariza-
tion of muscle fibers, a victim of electric
Physiologic Effect of shock may not be able to let go of the source
Electric Current of the current when it exceeds about 25 mA.
The threshold for induction of ventricular
For currents that enter and leave the body fibrillation is approximately 100 mA. The
through the skin, the usual situation outside externally applied current spreads out as it
of hospitals, Figure 2-2 shows the approxi- passes through the body, so that the frac-
mate amounts of current associated with var- tion passing through the heart is small, less
ious physiologic effects, ranging from mini- than 0.1%.
In hospitals, one of the two required concurrent sources), conducting fluids
tacts between an external source or ground may get into instruments through
and the body may be an intracardiac cathe- spillage or leakage, and instruments
ter. If current enters or leaves through this may be used by many persons or used
device, essentially the entire current flows in many locations (or both), thus in-
through the myocardium. In this case, the creasing the chance of fault caused by
threshold for inducing ventricular fibrillation misuse or wear. In the operating room,
is far less than for externally applied current. instruments such as electrosurgical
In humans, this threshold is estimated to be units may present special risks to the
approximately 50 ûA, but experiments in patient if proper precautions for elec-
dogs have shown that as little as 20 //A is suf- tric safety are not followed.4'5
ficient.2 Furthermore, the results of a recent 2. Through attached electric instruments,
study have suggested that the threshold for patients are often grounded or they
induction of cardiovascular collapse (which may easily contact grounded objects,
is less than the threshold for inducing ven- for example, metal parts of beds,
tricular fibrillation) is the more relevant lamps, and instrument cases.
quantity, and this threshold is only 20 //A.3 3. Contacts with the source of leakage cur-
The threshold may be significantly lower in rent and with ground are often low re-
persons with preexisting heart disease. sistance because connections to moni-
toring devices purposely minimize skin
resistance (for example, electrodes ap-
Factors Reducing Risk of plied with conducting paste) or bypass
it altogether (for example, indwelling
Electric Shock catheters). Furthermore, patients with
conductive intracardiac catheters, such
The risk of electric shock or electrocution
as pacemaker leads and saline-filled
from appliances is reduced by several fac-
catheters, have a direct low-resistance
tors, including the following:
pathway to the heart. Because only tiny
1. Leakage currents that are available
currents flowing in such a path may
from most electric appliances are rela-
induce lethal ventricular fibrillation,
tively small.
such patients are called electrically sus-
2. People using appliances are often not
ceptible.
connected to ground.
4. Weakened or comatose patients cannot
3. Contacts with the source of leakage cur-
withdraw from a source of current.
rent and with ground usually have high
5. Patients' hearts may be more susceptible
resistance, for example, dry, intact skin.
(through disease) to electric current-
4. Healthy, alert people can withdraw
induced ventricular fibrillation.
from a source of current in most cases.
5. The hearts of healthy people require
significant electric currents to induce
ventricular fibrillation.
LEAKAGE CURRENT
Origin
Factors Increasing Risk of Electric Leakage current in an electric apparatus
Shock in Hospitals may originate in several ways, including the
following:
The risk of electric shock or electrocution 1. There is always a finite internal circuit
from appliances in hospitalized patients is resistance between the power line (hot
significantly greater because of the following wire) and the instrument chassis,
factors: known as instrument ground; this may be
1. Leakage currents that may be available decreased by faults in the wiring or by
from appliances are relatively large be- breakdown of insulation. A resistance
cause patients may be attached to many as large as 5 Mfl still allows 24 ^(A to
instruments (thus providing multiple flow between the "hot" conductor and
ground, which may be enough to in- through the round pin of electric
duce ventricular fibrillation in an elec- plugs. Any leakage currents that would
trically susceptible patient. otherwise flow to a subject in contact
2. The capacitance between the "hot" with the chassis are instead shunted
conductor and the chassis resulting through this low-resistance pathway to
from internal circuitry or external ca- ground. Because the leakage currents
bling may provide a relatively low-im- in properly functioning equipment are
pedance pathway for alternating cur- small, the ground wire in the building
rent. A capacitance as small as 440 power distribution system usually car-
picofarad (pF) still allows 20 fJiA to ries very little current, unlike the neu-
flow between the "hot" conductor and tral wire, which carries the full operat-
ground. ing current. Hence, the potential drop
3. The inductive coupling between between the equipment chassis and the
power-line circuits and other circuit earth ground connection located at the
loops, such as ground loops when there electric distribution panel of the build-
are multiple ground connections to ing is minimal, and the equipment
the patient, can induce ground-path ground potential remains very close to
current flow as well. In addition to the earth ground potential (Fig. 2-3).
the leakage currents available from 2. Hospital rooms that have exposed
equipment-to-patient ground connec- metal parts, for example, window
tions, leakage currents may be intro- frames, bathroom plumbing, shelving,
duced by similar mechanisms into other and door frames, may also connect
leads or connections to the patient. these to earth ground through the
same grounding system used for elec-
tric outlets. All such grounded points
Methods by Which Leakage in one room should be connected to a
single ground wire, an equipotential
Current Reaches Patients grounding system. Also, all biomedical
equipment connected to a patient
Leakage currents may reach patients when
should draw power from the same
contact is made either directly or through
group of outlets to avoid large ground
another person to exposed metal parts or to
loops (Fig. 2-4).
the chassis of electric equipment. Leakage
3. When necessary, isolation transformers
currents may also reach patients through a
may be used to eliminate the neutral-
direct connection of the chassis (equipment
to-ground connection entirely, thereby
ground) to the patient; such a patient ground
connection is necessary for noise-free record-
ing of physiologic signals. Finally, leakage
currents may reach patients through resis-
tive or capacitive (or possibly inductive) cou-
pling to leads other than the patient ground.
Methods to Reduce Leakage

Current Reaching Patients
Many methods are used in modern hospital
electric distribution systems and in biomed-
ical instruments to decrease the risk of elec-
tric shock by reducing the available leakage
currents, including the following: Figure 2-3. The power cord ground wire conducts leak-
age current from an electric apparatus chassis to earth
1. The chassis and all exposed metal parts ground. (From Seaba P. Electrical safety. Am J EEC
of electric appliances are grounded Technology 20:1-13, 1980. By permission of the Amer-
through a separate ground wire, ican Society of Electroneurodiagnostic Technologies.)
reducing the risk of shock when a pa-

tient connected to a biomedical in-
strument comes in contact with an
earth ground (Fig. 2-5). However, iso-
lation transformers do not eliminate
entirely the risk of shock, because they
may have significant leakage currents;
furthermore, they are expensive, and
they also generate heat, noise, and elec-
tric interference. Therefore, their use
is usually limited to special settings,
such as operating rooms.
Figure 2-4. A faulty floor polisher produces excessive 4. Appliances can be constructed with non-
leakage current, which flows through the building metallic cases to minimize the chance
ground wiring. A portion of this current also flows of patients contacting the equipment
through the large ground loop involving the patient
and the two electric devices connected to the patient chassis.
(that is, an electrocardiogram [EGG] monitor attached 5. Appliances should have short line
to one leg and an electroencephalogram [EEG] ma- cords, and the use of extension cords
chine) . This ground loop was created by the use of two should be avoided to minimize capaci-
outlets physically distant from one other. (From Seaba
P. Electrical safety. Am J EEG Technology 20:1-13, tive and resistive leakage currents be-
1980. By permission of the American Society of Elec- tween the hot and the ground wires.
troneurodiagnostic Technologies.) Note that each foot length of cord un-
avoidably introduces about 1 juA of
Figure 2-5. An isolation transformer used to reduce equipment leakage current. The equipment is connected to
the secondary coil of the transformer, which is electrically isolated from the power line hot and neutral conduc-
tors. (From Cromwell L, Weibell FJ, Pfeiffer EA, Usselman LB. Biomedical Instrumentation and Measurements.
Prentice-Hall, Englewood Cliffs, New Jersey, 1973, p 387. By permission of the publisher.)
leakage current into the ground con- diac catheters) should be avoided by
nection. using nonconductive materials.
6. If at all possible, direct connections of 7. Current that may flow through all con-
patients to ground should be avoided. nections between a patient and the
In particular, inadvertent electric paths equipment, both signal and ground,
between ground and patients that by- should be limited to no more than 20
pass the normally high skin resistance /xA through the use of current limiters
(especially paths provided by intracar- (Fig. 2-6) or inductive or optical cou-
Flgure 2-6. Current-limiting devices. Top, Circuit used in older biomedical equipment includes a 5 mA fuse in the
patient ground connection; current flow through this one connection is limited to 5 mA, which is insufficient pro-
tection for electrically susceptible patients. Middle, Circuits in modern biomedical equipment use special current
limiters. Bottom, Electric characteristics of current limiters. In the operating range, current is proportional to volt-
age (Ohm's law), but currents outside this cannot exceed 10 /jA. (From Cromwell L, Weibell FJ, Pfeiffer EA. Bio-
medical Instrumentation and Measurements, 2nd ed. Prentice-Hall, Englewood Cliffs, New Jersey, 1980, p 442. By
permission of the publisher.)
Figure 2-7. The input circuit of an inductively isolated biomedical instrument. Isolated patient leads are achieved
by using a carrier wave amplifier with transformer coupling. (From Cromwell L, Weibell FJ, Pfeiffer EA. Biomed-
ical Instrumentation and Measurements, 2nd ed. Prentice-Hall, Englewood Cliffs, New Jersey, 1980, p 443. By per-
mission of the publisher.)
pling devices (Fig. 2-7). Alternatively, system. Failure of this ground connection
when practical, battery-powered equip- may occur in several ways. There can be a
ment that has no direct connection to failure of attachment of the grounding wire
line voltage or to ground can be used. in the line cord to the equipment chassis, a
break in continuity of the grounding wire
within the cord, or a failure of connection
of the grounding wire to the grounding pin.
ELECTRIC SAFETY PRINCIPLES Also, the grounding pin may make poor con-
AND IMPLEMENTATION tact with the wall receptacle because of a re-
duction in contact tension caused by me-
Equipment Grounding chanical wear. The grounding pin can also
be deliberately bypassed using a so-called
Proper grounding of electric equipment cheater (3-prong to 2-prong) adapter. De-
(that is, providing a low-resistance pathway fects also can occur in building wiring, such
from the equipment chassis to an earth as an improper or omitted connection of the
ground) is usually accomplished through wall receptacle's grounding terminal to a
the grounding wire in the line cord that con- ground wire or an interruption of the
nects to the round pin in the plug and ground connection somewhere in the build-
thence to the building electric grounding ing's wiring. This is particularly likely if the
Eletiric Safety in the Laboratory and Hospital 25
metal conduit, which is subject to corrosion Tests that should be performed on each
and loss of mechanical contact, provides the biomedical instrument at the time of pur-
ground connection. Particularly in newly chase and periodically thereafter, for ex-
constructed or remodeled rooms and build- ample, every 6 to 12 months, include the
ings, it is advisable to visually inspect and to following:
electrically test the ground connection in all 1. Visually inspect the line cord and plug
wall receptacles. Because the ground con- for signs of damage, wear, or breakage.
nection is only for electric safety purposes, 2. Measure the resistance between the
the lack of it in no way affects operation ground pin of the plug and the instru-
of the electric equipment and, therefore, ment chassis. This should be less than
will remain undetected if not specifically o.i a
checked. 3. Measure the chassis-to-earth ground
leakage current using the circuit shown
in Figure 2-8. This measurement
should be made with the equipment's
Tests for Equipment Grounding grounding pin disconnected (to en-
and Leakage Current sure safety even if the building ground-
ing system is faulty) and under four
Each hospital, laboratory, or clinic should separate conditions. This should in-
establish an electric safely program that clude both normal and reverse polarity
includes selecting equipment that meets of the hot and neutral wires (to ensure
appropriate safety standards, testing new safety even if the wall receptacle is er-
equipment after purchase to verify that stan- roneously wired with opposite polar-
dards are met, inspecting and retesting ity) and with the equipment power
equipment periodically thereafter to ensure switch "on" and "off." In all four con-
that damage through use and misuse does ditions, this current should not exceed
not compromise safety, educating all those 100 /AA.
who use the equipment (especially techni- 4. Measure the leakage current from each
cians) in electric safety principles, and en- terminal that connects to a patient, in-
suring that certain basic minimal safety tests cluding the patient ground to earth
are performed each time a biomedical ap- ground, under the same four condi-
paratus is plugged in, turned on, and con- tions. This is the maximal leakage cur-
nected to a patient. rent that the equipment can supply to
Tests that should be performed on build- a patient who is grounded through a
ing wiring at the time of installation include second connection. For use with elec-
the following: trically susceptible patients, this cur-
1. Visually inspect the wiring of all wall re- rent should not exceed 20 jjiA.
ceptacles to ensure that it is correct. 5. Measure the leakage current from the
2. Measure the resistance between each power-line hot wire to each terminal
wall receptacle ground (and other that connects to a patient, including
grounded objects in the room) and a the patient ground, under the same
ground known to be adequate, such as four conditions. This is the maximal
a cold water pipe or an independent current that can be absorbed by the
grounding bus. This resistance should equipment from a patient who acci-
be less than 0.1 H. dentally comes in contact with a 120 V
3. Measure the contact tension provided power line. For use with electrically sus-
by the wall receptacle, that is, the force ceptible patients, this current should
required to withdraw the ground pin of not exceed 20 //A.
a test plug from the receptacle. This
should be at least 10 ounces.
These tests, except the first, should also Rules for Electric Safety
be performed periodically, for example,
every 6-12 months, and the receptacles In addition to a program of periodic testing
whose contact tension has degraded below and inspection, electric safety requires that
10 ounces should be replaced. all persons using electric equipment in the
Figure 2-8. Circuit used to test leakage current in a biomedical instrument (an electroencephalogram [EEG] ma-
chine). The reversing switch allows the test to be made with both standard and reverse polarity of line voltage. The
three leakage currents measured are (1) from the power line hot wire to each patient terminal, (2) from the equip-
ment chassis to earth ground, and (3) from each patient terminal to earth ground. (From Grass ER. Electrical Safety
Specifically Related to EEG [Bulletin #X757C78]. Grass Instrument Company, Quincy, Massachusetts, 1978, p 9. By
permission of Grass Instrument Company.)
laboratory or hospital are familiar with the equate earth ground, such as through
following rules: use of an approved three-prong
1. Do not ever directly ground patients grounded plug.
or allow patients to come in contact 3. Use only safe, properly designed, and
with grounded objects while con- pretested electric equipment. All bio-
nected to a biomedical instrument. medical devices directly connected to
Patient-ground connections should patients must have isolation or cur-
be made to only one instrument at a rent-limiting circuits if they are to be
time; thus, if a patient is already con- used with electrically susceptible pa-
nected to a device such as an electro- tients. All line-powered equipment
cardiogram (EGG) monitor, a sepa- should have three-prong grounded
rate EEG or EMG ground should be plugs. In general, patients should not
omitted unless it is absolutely neces- be allowed to bring their own electric
sary for obtaining a recording free of appliances from home for use in a
artifact and noise. hospital room.
2. Ensure that every electric device or 4. Ensure that all electric equipment in
appliance, for example, lamps, elec- use has had a safety inspection re-
tric beds, electric shavers, and radios, cently (within 6-12 months), as indi-
that a patient might accidentally come cated by a dated electrical safety in-
in contact with is connected to an ad- spection tag or sticker.
5. Connect all patient-connected equip- sticker? If there is already a patient

ment to outlets in the same area or ground connection to one of these
cluster to avoid large ground loops. other instruments, try testing without
6. Never use an extension cord on pa- any other patient ground connection.
tient-connected equipment because 4. Choose an outlet in the same area or
this adds leakage current through its cluster used by other patient-con-
internal capacitance and resistance nected devices. Before plugging in the
and, thus, provides another chance equipment, check the contact tension
for ground connection failure. of the chosen receptacle with a simple
7. Cover all electric connections to in- device that should be carried with all
tracardiac catheters with insulation, portable equipment.
such as a surgical rubber glove, to 5. Turn on the instrument and calibrate
eliminate electric continuity between it before connecting it to the patient.
external devices or ground and the Major electric problems may show up
catheter whenever possible. during calibration; furthermore, elec-
8. Have a defibrillator available at all tric surges occur as the instrument is
building locations where patients turned on and leakage currents may be
have cardiac catheters in place. higher while it is warming up.
9. Do not ignore the occurrence of any 6. Disconnect the patient from the in-
electric shocks, however minor; inves- strument before turning it off.
tigate their causes. Thoroughly test
any equipment that may have been in-
volved before putting it back in ser-
vice. Also, do not ignore any abnor- SUMMARY
mal 60 Hz interference or artifact in
an electrophysiologic recording; this This chapter reviews the principles of elec-
finding may indicate that some device tric safety that are relevant to clinical neu-
is leaking current into the patient. rophysiologic studies. Knowledge of these
10. Follow certain safety procedures, in- principles is necessary both for those in-
cluding routine safety checks, each volved in evaluating and purchasing test in-
time an electric device is to be con- struments and for those involved in main-
nected to a patient. taining and using them. All those who order,
perform, interpret, or supervise electro-
physiologic testing share the legal responsi-
Electric Safety Procedures bility for patient safety, including electric
for Technicians safety.
The following procedures should be fol-

lowed by technicians while performing an REFERENCES
electrophysiologic test requiring line-pow-
ered equipment on a patient, especially 1. Anonymous. Risk of electric shock from patient mon-
portable studies performed in a patient's itoring cables and electrode lead wires. Health De-
room: vices 22:301-303, 1993.
2. Starmer CF, Mclntosh HD, Whalen RE. Electrical
1. Check the physical condition of the hazards and cardiovascular function. N Engl J Med
equipment. Is there any evidence of 284:181-186, 1971.
liquid spills, cord wear, or damage? Is 3. Swerdlow CD, Olson WH, O'Connor ME, Gallik DM,
the plug bent or broken? Is the equip- Malkin RA, Laks M. Cardiovascular collapse caused
ment labeled with a current electric by electrocardiographically silent 60-Hz intracradiac
leakage current. Implications for electrical safety.
safety inspection sticker? Circulation 99:2559-2564, 1999.
2. Inspect the patient area for any two- 4. Litt L, Ehrenwerth J. Electrical safety in the operat-
wire ungrounded appliances. Have ing room: important old wine, disguised new bottles.
them unplugged and removed. Anesth Analg 78:417-419, 1994.
5. McNulty SE, Cooper M, Staudt S. Transmitted ra-
3. Inquire about any other instruments at- diofrequency current through a flow directed pul-
tached to the patient. Are they labeled monary artery catheter. Anesth Analg 78:587-589,
with a current electric safety inspection 1994.
Chapter 3
VOLUME CONDUCTION
PRINCIPLES Homogeneous Sphere Model

ELECTRIC PROPERTIES OF VOLUME Multiplanar and Multiple Sphere Models
CONDUCTORS APPLICATIONS OF VOLUME
CALCULATING POTENTIALS IN INFINITE CONDUCTION PRINCIPLES
HOMOGENEOUS MEDIA Stationary Potentials Produced by
Monopole, Dipole, and Quadrupole Sources Propagating Generators
Spatial Distributions of Potentials from Eelectroencephalographic Applications
Various Sources Dipole Source Localization
POTENTIALS IN NONHOMOGENEOUS SUMMARY
MEDIA
Planar Interfaces and Hemi-Infinite
Homogeneous Media
PRINCIPLES reach the skin surface, where, according to

Ohm's law, the current causes a potential
Electrophysiologic studies involve recording drop across the space between two elec-
potential differences from electrodes in con- trodes. This drop in potential can be de-
tact with the body; usually, these electrodes tected and amplified by a differential ampli-
are placed on the skin surface. The bioelec- fier. The properties of the volume conductor
tric potentials are generated by sources in- determine the potentials recorded by a given
side the body, for example, brain, peripheral array of electrodes in response to a given set
nerve, and muscle that may be some distance of generators.
away from the recording electrodes. These The potential generated by a population
sources may be either active or passive. Ac- of neurons is equal to the sum of the po-
tive sources are ionic channels that open or tentials generated by the individual neurons.
close in response to changes in transmem- In the case of some neurons, such as cere-
brane potential, neurotransmitter binding, bral cortical neurons that have extensive
intracellular calcium, other second messen- dendritic trees in which multiple synapses
gers, and so forth, allowing small currents to may be active simultaneously and in which
flow into or out of the cell body, axon, or multiple regions of active membrane are ca-
dendrite. Passive sources are areas of neu- pable of generating action potentials, the
ronal membrane that permit current flow potentials generated by each neuron are also
into or out of the cell by passive leakage or a sum of the potentials generated by multi-
capacitive effects. These current sources or ple active and passive areas of membrane.
sinks (active or passive) lead to widespread Only if the responsible neuronal generators
extracellular currents flowing in the con- are arranged regularly and activated more
ducting medium surrounding the neurons, or less synchronously is sufficient summation
called the volume conductor. Some of the ex- obtained to allow recording of potentials at
tracellular currents in the volume conductor a considerable distance from the genera-
28
Volume Conduction 29
tors. For nerve conduction studies, this is dipoles oriented opposite to one another
achieved by the synchronous volley of action (Fig. 3-1).
potentials in closely grouped parallel axons At a macroscopic level, the potential field
of a compound nerve (the compound nerve ac- generated by synchronous activation of
tion potential). many cortical pyramidal cells behaves like
In the cerebral cortex, which is the gen- that of a dipole layer. This has been called
erator of spontaneous EEC activity, the pyra- an open field configuration, in contrast to the
midal neurons are arranged in a regular fields generated by neurons with dendritic
manner, with the main axes of the dendritic arborizations that are distributed radially
trees parallel to one another and perpen- around the cell body and called closed fields.
dicular to the cortical surface. Furthermore, Closed-field potentials are equivalent to the
thousands of these cortical pyramidal neu- field produced by a set of radially oriented
rons are activated more or less simultane- dipoles at the surface of a sphere; such a
ously by synapses made by a single axon or field is negligible at a distance because both
small groups of axons, producing significant the radial and tangential components of
extracellular current flow. Under these cir- current flow cancel each other in this con-
cumstances, the longitudinal components of figuration.
current flow from different neurons add to- Because the dendrites of cortical pyrami-
gether, and the transverse components of dal cells are perpendicular to the cortical, or
flow cancel out, producing a laminar current
along the main axes of the neurons. De-
pending on whether the activated synapse is
excitatory or inhibitory, the direction of cur-
rent flow across the cell membrane is either
inward or outward. The synaptic transmem-
brane current flow is accompanied by an op-
posite outward or inward current flow at
another location along the dendritic tree,
called passive source or sink, which produces
a dipolar configuration, as described in the
following section. An excitatory postsynaptic
potential (EPSP) occurs when positive ions
flow intracellularly, called inward current flow,
and an inhibitory postsynaptic potential
(IPSP) occurs when negative ions flow in-
tracellularly, called outward current flow. Figure 3-1. Patterns of current flow near a neuron
Thus, the local extracellular potential pro- caused by synaptic activation. E, Current flow caused by
duced by an EPSP is negative and that pro- activation of an excitatory apical dendritic synapse de-
polarizes the cell membrane, producing a current sink.
duced by an IPSP is positive. The extracellular potential, shown on the left, has a neg-
The orientation of the dipole created by ative polarity at the synapse. A distributed passive source
synaptic activity in the cerebral cortex de- at the cell body produces an extracellular potential with
pends on both the type of synaptic activity, a positive polarity. /, Current flow caused by activation
of an inhibitory synapse near the cell body hyperpolar-
whether an EPSP or IPSP, and the location izes the cell membrane, producing a current source.
of the synapses, whether superficial or deep. The extracellular potential, shown on the right, has a
An EPSP located superficially in the cere- positive polarity at the synapse. A distributed passive
bral cortex, that is, along the distal branches sink at the more superficial level of a distal dendrite
of the pyramidal cells, produces a dipole produces an extracellular potential with a negative po-
larity. Thus, a deep inhibitory postsynaptic potential pro-
with a superficial negative and a deep posi- duces an extracellular potential field similar to that of
tive pole. A deep EPSP, for example, caused a superficial excitatory postsynaptic potential. (From
by a synapse near the cell body or on the Lopes da Silva F, Van Rotterdam A. Biophysical aspects
basal dendrites, produces a dipole with a of EEG and magnetoencephalogram generation. In
Niedermeyer E, Lopes da Silva F [eds]. Electroen-
superficial positive and a deep negative cephalography: Basic Principles, Clinical Applications,
pole.1 The IPSPs and EPSPs located at sim- and Related Fields, 3rd ed. Williams & Wilkins, Balti-
ilar depths in the cerebral cortex produce more, 1993, pp 78-91. By permission of the publisher.)
pial, surface, many dipole-like sources of electrodes are embedded. In this situation,
EEG and evoked potential waveforms are ra- the recorded potential can be easily calcu-
dial in direction, that is, they are perpen- lated from the configuration of source
dicular to the surface of the scalp. These currents.
generators typically reside at the apex of cor-
tical gyri. However, cortical generators lo- Monopole, Dipole, and
cated in the walls of sulci—where the corti- Quadrupole Sources
cal surface is perpendicular to the scalp
surface—may create potential fields that cor- A single source, or sink, of current is re-
respond to a tangentially oriented dipole. A ferred to as a monopole. The potential rela-
classic example of this is the potential field tive to a distant reference at distance r from
of the centrotemporal spike discharges of- a monopole source in an infinite homoge-
ten seen in benign rolandic epilepsy of neous medium of conductivity cr (resistivity
childhood. p = I/O") is given by
ELECTRIC PROPERTIES OF
VOLUME CONDUCTORS
where / is the magnitude of the monopolar
In general, a volume conductor such as the current source. Thus, the potential of a
body can be characterized by its conductiv- monopole falls off inversely with distance
ity (or resistivity) and its dielectric constant from the source, and the equipotential lines
(capacitive properties), which may vary from for a monopole form circles around the cur-
tissue to tissue. This may be modeled by di- rent source or sink (Fig. 3-2A).
viding the entire volume conductor into Two adjacent current monopoles of op-
many small regions, each of which is as- posite polarity constitute a current dipole.
sumed to be homogeneous, that is, the con- This is a more realistic generator than an iso-
ductivity and dielectric constant are the lated monopole because the current ema-
same throughout. For most purposes, at the nating from the source can flow through the
frequencies of interest in neurophysiologic medium to the sink where it is absorbed. In
recordings, the capacitive properties of the respect to the potentials they produce at dis-
volume conductor may be ignored and only tant recording sites, many neuronal current
a conductivity value for each region, to- generators may be well described in terms
gether with the geometry of the region, is of a current dipole. For example, as noted
needed to characterize the volume conduc- above, the main contributors to spontaneous
tor. In a noncapacitive (purely resistive) vol- EEG activity are the excitatory and inhibitory
ume conductor, the recorded potentials are postsynaptic potentials in the dendritic trees
always in phase, or synchronous, with the of cortical pyramidal neurons. The arrange-
current sources, and the conductive prop- ment of synapses on the dendritic trees pro-
erties of the medium are independent of duces a current source and sink separated
frequency. However, in the more general by a significant distance, and this constitutes
case of a resistive-capacitive medium, volume an electric dipole. The characteristic orga-
conduction is frequency dependent, and the nization of the cortical pyramidal neurons,
source currents and recorded potentials are that is, oriented parallel to each other and
out of phase. perpendicular to the cortical surface, allows
the potentials from many such dipole
sources to summate effectively.
CALCULATING POTENTIALS IN The potential. relative to a distant refer-
INFINITE HOMOGENEOUS ence measured at distance r from a dipole
MEDIA source in an infinite homogeneous medium
of conductivity a is given by
The simplest form of a volume conductor is
a homogeneous medium without bound-
aries in which generators and recording
Figure 3-2. Equipotendal lines in a volume conductor for various current source distributions: A, A point source
or monopole; B, a dipole; C, a quadrupole (two oppositely directed dipoles); and D, an action potential source
propagating along an axon. In D, the arrows represent some of the lines of current flow. (From Stein.2 By permis-
sion of Plenum Press.)
where / is the magnitude of the dipole cur-

rent source, d is the pole separation, and 0
is the angle between the dipole axis and the
line from the dipole to the measurement
point (Fig. 3-3A); this formula is valid only
for r» d. Thus, the potential of a dipole
falls off inversely with the square of the dis-
tance from the source. The lines of current
flow around a dipole form curved paths (Fig.
3-SB). The equipotential surfaces are per-
pendicular to the lines of current flow and
have a figure 8 configuration around the di-
pole.2 The zero potential surface is a plane
halfway between the two poles of the dipole,
because on this plane 0 = 90° and cos 6 =
0° (Fig. 3-35).
Two adjacent current dipoles of opposite
orientation placed end-to-end constitute a
current quadrupole. The potential of a Figure 3-3. A current dipole. A, Coordinate system
showing definition of r and 6. B, Lines of current flow
quadrupole falls off inversely with the cube (solid) and equipotential lines (dashed) in a volume con-
of the distance from the source, and the ductor (From Nunez.3 By permission of Oxford Uni-
equipotential surfaces around the quadru- versity Press.)
calculate and plot the potential recorded

from successive points along a given line or
axis in the volume conductor surrounding
the source. For a monopole source, the po-
tential recorded relative to a distant refer-
ence at points along a line at distance I from
the source in an infinite medium of con-
ductivity <r is
where I is the source current magnitude and

x is the distance along the line between the
source and the recording point. The poten-
tial distribution is a single peak whose sharp-
ness increases with decreasing distance I
from the source (Fig. 3-4A).
For a dipole source, the potential recorded
relative to a distant reference at points along
Figure 3-4. Potentials recorded along a line located at

various distances from a current source (1,2, and 5 cm),
as a function of position along the line: A, a current
monopole, B, a dipole, and C, an action potential
source. (From Stein.2 By permission of Plenum Press.)
pole have a cloverleaf configuration (Fig.

3—2 C). A quadrupole is a fair approximation
of the potential generated by an action po-
tential propagating along an axon: the ax-
onal membrane has a negative polarity out-
side and a positive polarity inside at the peak
of the action potential. However, on either
side of this peak, the membrane is positive
outside and negative inside (Fig. 3-2D).
Spatial Distributions of Potentials

from Various Sources
Figure 3-5. A, Coordinate system for potentials
recorded along a line perpendicular to the axis of a cur-
To understand the spatial distributions of rent dipole source. B, Coordinate system for potentials
potentials generated by various types of recorded along a line parallel to the axis of a current di-
sources, the above formulas can be used to pole source.
a line perpendicular to the dipole axis in an spaced electrodes (bipolar recording) at

infinite medium of conductivity <r is points along a line perpendicular to the di-
pole axis is
where /is the dipole magnitude, dis the pole

separation, / is the distance from the dipole where A# is the electrode separation and
to the line, and x is the distance along the dV/dx (the derivative of Vwith respect to x)
line between the dipole axis and the record- represents the slope of the Vversus x curve.
ing point (Fig. 3-5A). This situation applies The potential distribution obtained with this
to scalp potentials produced by a radially recording method is biphasic (Fig. 3-6.8).
oriented cortical dipole generator. For ex- For a dipole source, the potential
ample, a radially oriented cortical dipole is recorded relative to a distant reference at
seen frequently in EEC recordings because points along a line parallel to the dipole axis
of the radial orientation of the cortical pyra- in an infinite medium of conductivity a is
midal neurons. The potential distribution is
a single peak whose sharpness increases with
decreasing distance / from the source (Fig.
3-6A).
For the same dipole source, the potential where / is the dipole magnitude, d is the
difference recorded between two closely pole separation, Us the distance from the
Figure 3-6. Potentials recorded along a line located at various distances from a current dipole (solid curve, 1 cm;
dashed curve, 2 cm; dotted curve, 3 cm) as a function of position along the line. A, Referential recording, with the
line perpendicular to dipole axis; B, bipolar recording for line perpendicular to dipole axis; C, referential record-
ing, with the line parallel to dipole axis; D, bipolar recording for line parallel to dipole axis.
dipole to the line, and y is the distance

along the line between the dipole center
and the recording point (Fig. 3-5B). This
situation, for example, applies to scalp po-
tentials produced by a tangentially oriented
cortical dipole generator. A dipole of this
orientation is often present when the cor-
tical generator region is deep in a sulcus or
fissure in which the cortical surface is per-
pendicular to the scalp surface. This is of- A B
ten the case, for example, for the spikes Figure 3-7. Lines of current flow caused by a monopo-
lar source below a planar interface when, A, the con-
seen in EEG recordings of subjects with be- ductivity of the upper region is larger than that of the
nign rolandic epilepsy of childhood. The lower region (erg > o^) or, J5, when the conductivity of
potential distribution for a tangential di- the upper region is smaller than that of the lower re-
pole generator is biphasic, and its sharpness gion (cr2 < <TI) . (From Nunez.3 By permission of Ox-
increases with decreasing distance I from ford University Press.)
the source (Fig. 3-4B and 3-6 C).
For the same dipole source, the potential a planar interface between them, the con-
difference recorded between two closely figurations of the lines of current flow
spaced electrodes (bipolar recording) at change at the interface. This occurs because
points along a line parallel to the dipole the current density (current per unit area)
axis is flowing in the direction parallel to the in-
terface is less in the region of higher con-
ductivity. Consequently, if the source is lo-
cated in the region of lower conductivity, the
lines of current flow bend outward as they
where ky is the electrode separation and enter the region of higher conductivity (Fig.
dV/dy (the derivative of V with respect to y) 3-7A). If the source is located in the region
represents the slope of the V vs. y curve. The of higher conductivity, the lines of current
potential distribution obtained with this flow bend inward as they enter the region of
recording method is triphasic (Fig. 3-6D). lower conductivity (Fig. 3-7B).
For a quadrupole source, the potential dis- One of the major inhomogeneities in the
tribution recorded relative to a distant ref- body as a volume conductor is the interface
erence at points along a line parallel to the between the body and the external environ-
quadrupole axis in an infinite medium of ment. For sources located a short distance
conductivity a is triphasic, and its sharpness under the skin surface, for example, action
increases with decreasing distance from the potentials propagating in fairly superficial
source (Fig. 3—4C). This situation approxi- nerves or EEG activity generated by cortical
mates the potential expected from an action sources approximately 2-3 cm deep, this in-
potential propagating along an axon paral- terface can be approximated as a plane, with
lel to the line of the recording electrodes. a region of high conductivity on one side
containing the embedded source and a re-
gion of essentially zero conductivity on the
other side. In this situation, no current flow
POTENTIALS IN penetrates from the high conductivity re-
NONHOMOGENEOUS gion to the zero conductivity region; thus,
MEDIA the lines of current flow are completely re-
flected at the interface. Consequently, po-
Planar Interfaces and Hemi-Infinite tentials measured at the interface, that is,
Homogeneous Media surface-recorded potentials, caused by un-
derlying generators are twice as large as they
When a monopolar source is located in a would be for the same generators and
volume conductor divided into two hemi- recording site in an infinite homogeneous
infinite regions of differing conductivity with medium (Fig. 3-8A).
Figure 3-8. The approximate effects on measured surface potentials caused by a dipole source of various inhomo-
geneities in the volume conductor. <J>H is the potential that would have been recorded at the same point if the same
source dipole were immersed in an infinite homogeneous medium; $ is the potential actually recorded in the in-
homogeneous medium. A-C are based on a planar interface and apply to superficial, or cortical, dipole sources: A,
body-air interface; B, brain-cerebrospinal fluid (CSF) and CSF-skull interfaces; C, brain-CSF-skull-scalp-air in-
terfaces. D and E are based on spherical interfaces and apply to deep, or brain stem, dipole sources: D, body-air
interface; E, brain-skull, skull-scalp, and scalp-air interfaces. (From Nunez.3 By permission of Oxford University
Press.)
Homogeneous Sphere Model where 7 is the magnitude of the dipole and

d is die pole separation. At points near the
For sources located in the head, such as cor- center of the sphere (r= 0), the potential is
tical and subcortical generators of EEC and the same as that expected in an infinite ho-
evoked potentials, a spherical volume con- mogeneous medium, but at the surface of
ductor model is a reasonable approximation the sphere, that is, for scalp surface record-
to the actual geometry. The simplest model ings, it is three times as great (Fig. 3-8D).
assumes a uniform conductivity a- within a This is because the lines of current flow are
sphere of radius a and zero conductivity out- confined to the spherical volume and the
side the sphere. For a dipole source located current density (and hence electric poten-
at the center of the sphere (such as a brain tial) is correspondingly greater.
stem generator of one of the short latency
auditory-evoked potential peaks), the po-
tential at a point inside the sphere located Multiplanar and Multiple
at distance r from the dipole and at an an- Sphere Models
gle 6 from the dipole axis is
A single air-body interface surface is only one
of the inhomogeneities that affects volume
conduction. For EEG and scalp recorded-
evoked potentials, the other inhomogeneities APPLICATIONS OF VOLUME

of importance are the differing conductivi- CONDUCTION PRINCIPLES
ties of brain, cerebrospinal fluid (CSF),
skull, and scalp. To investigate the effects of Stationary Potentials Produced by
these regions, both multiplanar and multi-
ple sphere models have been used. Propagating Generators
For dipole sources located in the cerebral
cortex and for subdural recording elec- When recording from various electrodes
trodes, a model using two planar interfaces placed at different locations along the path
(brain-CSF and CSF-skull) can be used. This of a complex propagating generator, such as
model predicts that the measured potentials an action potential source, the time at which
would be approximately equal to those that the propagating potential is seen at each lo-
would be recorded in an infinite homoge- cation is different because of the finite ve-
neous medium (Fig. 3-85). For cortical di- locity of propagation of the generator. (Note
pole sources with scalp surface recording that volume conduction of electric potentials
electrodes, a model using five regions from the generator to the recording elec-
(brain, CSF, skull, scalp, and air) and four trode is essentially instantaneous, because
planar interfaces predicts that the measured electric disturbances propagate at the speed
potentials would be approximately equal to of light in a conducting medium.) However,
one-fourth of those that would be recorded when a propagating generator passes through
in an infinite homogeneous medium (Fig. an interface between volume conducting re-
3—8C). This may be compared with the fac- gions of different sizes or conductivities, a
tor-of-2 augmentation of potentials pre- potential can be induced simultaneously at
dicted by the single planar interface model all recording electrodes during the time at
(Fig. 3-8A); the predicted relative attenua- which the generator is crossing the bound-
tion (by a factor of 8) is caused mainly by aries between regions with differing proper-
the effects of the poorly conducting skull, ties.4 Such a potential, which does not ap-
whose conductivity is only about 1/80 that pear at different times in different recording
of brain or scalp. The effect of the skull may locations, has been referred to as a station-
be diminished markedly in subjects who have ary potential (Fig. 3—9). This effect may be ob-
a skull defect, for example, because of preserved in somatosensory evoked potential
vious surgery. The EEG activity in the vicin- recordings as a consequence of the change
ity of such a defect may be several times in geometry of the volume conductor as a
greater in amplitude than the EEG activity in propagating nerve impulse travels from a
surrounding regions where the skull is intact. limb to the trunk. The same effect may also
For deep dipole sources, for example, influence the morphology of a brain stem
brain stem auditory evoked potential gener- auditory evoked potential recording because
ators, a multiple sphere model with four re- of changes in volume conductor properties
gions (brain, skull, scalp, and air), three along the central auditory pathways caused
spherical interfaces, and a dipole in the cen- by the complicated anatomy of the posterior
ter is appropriate. For scalp surface record- fossa. Such a stationary potential can be seen
ing electrodes, this model predicts that the only in derivations in which the first and sec-
measured potentials would be approxi- ond electrodes between which the potential
mately equal to twice those that would be is measured are on opposite sides of the
recorded in an infinite homogeneous boundary between the regions with differing
medium (Fig. 3-8E). This maybe compared sizes or conductivities; generally, this occurs
with the factor-of-3 augmentation of poten- only when recording with respect to a rela-
tials predicted by the homogeneous sphere tively distant reference electrode.5
model; the predicted relative attenuation
(by a factor of 2/3) caused by the poorly
conducting skull is not nearly as great for Electroencephalographic
deep sources as it is for superficial, or corti- Applications
cal, generators.3 The effect of skull defects
on potentials from deep sources is corre- In recording spontaneous scalp EEG activity
spondingly less than for superficial sources. or the late components of the somatosensory
Figure 3-9. Action potential propagation from a larger to a smaller volume conductor region, modeled here as two
joined cylinders of unequal diameter but equal conductivities. The expected potentials as a function of time at var-
ious recording positions at the surface of the cylinders are shown, recorded with respect to a reference electrode,
ref., on the larger cylinder, for an action potential propagating upward at 40 m/second. Recording locations on the
surface of the smaller cylinder, above the junction, show a "stationary" potential (arrows) at the time correspond-
ing to the propagating potential traversing the junction region. (From Stegeman et al.5 By permission of Elsevier
Science Ireland.)
and visual evoked potential that are re- not possible to find an inactive reference.
corded from scalp electrodes, it is desirable Even if a physically distant reference posi-
to measure the potentials at each scalp elec- tion were chosen, as on a limb, volume con-
trode position with respect to a distant, to- duction between the head and the distant
tally inactive reference electrode. In fact, it is position would make the reference active;
Figure 3-10. The futility of using a distant reference for scalp electroencephalographic (EEG) recording. The right
arm reference (A) is electrically equivalent to a neck reference (B), except for a slight additional resistance in se-
ries. AMP, amplifier. (From Nunez.3 By permission of Oxford University Press.)
that is, the reference electrode would be ness of the conducting medium between in-
electrically equivalent to a reference at the tracranial generators and overlying elec-
neck (still relatively close to intracranial gen- trodes may lead to a significant focal atten-
erators), with a slight additional resistance uation of electric activity, as in the case of a
that is negligible in its effect because of the subdural hematoma or a collection of fluid.
very large input resistance of the EEG am-
plifier (Fig. 3-10). In addition, such a dis-
tant reference would have unacceptable
characteristics in that very large artifacts, for
example, those produced by the electrocar-
diogram, movement, and muscle, would
probably be seen in the recording.
Properties of the volume-conducting
medium between intracranial generators
and scalp electrodes can have a major effect
on the recorded potentials. When the poorly
conducting skull is breached by openings,
for example, naturally occurring openings
such as the orbits or external auditory me-
atus or iatrogenic openings such as cran-
iotomy defects, long current paths through
the opening may cause appreciable electric
potentials to be recorded in areas that are,
in fact, far from the generators (Fig. 3-11).
Amplitude asymmetries, that is, differences
between homologous regions on the oppo- Figure 3-11. The effect of skull openings on scalp-
site side of the head, are the most commonly recorded potentials; the large skull resistivity—80 times
that of scalp or brain—leads to long current paths
observed effects of skull defects, with higher through skull openings that may cause appreciable po-
amplitudes occurring on the side of the tentials to be recorded far from the generators. (From
opening. A regional increase in the thick- Nunez.3 By permission of Oxford University Press.)
For extracranial generators such as the eyes, fortunately, for any given potential distribu-
which have a constant electric dipole motion recorded from the scalp surface, the
ment that induces changing electric poten- number of possible configurations of gener-
tials when they move, the effect of skull ators that could equally well produce that
openings is reversed; that is, the amplitude distribution is infinite; in general, the in-
of potentials caused by these generators is verse problem does not have a unique solu-
usually attenuated in the region of a large tion. If the problem can be constrained by
skull defect. independent anatomical or physiologic data,
however, then a solution to the inverse prob-
lem may be possible.6 For example, if there
Dipole Source Localization is reason to believe that a particular EEG
waveform or evoked potential peak is gen-
In many clinical neurophysiologic studies, erated by a localized intracranial source that
particularly EEG and evoked potential stud- may be well represented as a single electric
ies, conducted for clinical or research pur- dipole, then a model of the volume con-
poses, localization of the generators of a par- ductor properties of the head, such as the
ticular waveform or activity is of paramount three-sphere model discussed above under
importance. Volume conductor theory, as Multiplanar and Multiple Sphere Models
discussed in this chapter, provides a way to and in Figure 3-8, can be coupled with an
calculate the surface potential distribution appropriate mathematical algorithm to find
that would result from a known configura- the location, orientation, and strength of the
tion of intracranial generators, the forward single dipole whose predicted scalp-poten-
problem. However, it is also desirable to have tial distribution best fits the observed scalp
a way to determine the type, location, potential.7'8
strength, and orientation of all of the gen- The inherent uncertainties in the geome-
erators of a given scalp surface-recorded try and electric properties of the volume
waveform or activity, the inverse problem. Un- conductor limit the accuracy with which di-
Figure 3-12. A magnetoencephalographic recording device consists of a magnetic field gradiometer (two coils with
opposite polarities connected in series) and a superconducting quantum interference device (SQUID) amplifier.
(From Nunez.3 By permission of Oxford University Press.)
pole localization based on scalp-recorded recorded in clinical neurophysiologic stud-

potentials can be performed. A new tech- ies. Knowledge of these principles is neces-
nique for improving source localization de- sary for proper interpretation of neurophys-
rives from the magnetic field that any elec- iologic recordings in order to extract
trical current generates. Thus, intracranial information concerning the function and lo-
current sources generate magnetic fields cation of the neural structures that generate
that, with appropriate sensing devices, may the recorded activity or waveforms.
be detected at various locations outside the
head. These magnetic fields, unlike scalp-
recorded potentials, are unaffected by the REFERENCES
intervening medium, and calculations of
source locations from magnetic field maps 1. Niedermeyer E, Lopes da Silva F (eds). Electroen-
thus may be performed without a need for cephalography: Basic Principles, Clinical Applica-
complex and possibly inaccurate volume tions, and Related Fields, 3rd ed. Williams &
Wilkins, Baltimore, 1993.
conduction models. With a sensitive mag- 2. Stein RB (ed). Nerve and Muscle: Membranes,
netic detector, for example, a magnetome- Cells, and Systems. Plenum Press, New York, 1980,
ter or magnetic gradiometer (Fig. 3-12), pp 65-86.
and a special type of high-gain, low-noise am- 3. Nunez PL (ed). Electric Fields of the Brain: the
Neurophysics of EEG. Oxford University Press, New
plifier, specifically, a superconducting quan- York, 1981.
tum interference device, or SQUID, it is pos- 4. Dumitru D, DeLisa JA. AAEM Minimonography
sible to record the magnetic equivalent of #10: volume conduction. Muscle Nerve 14:605-624,
the EEG or evoked potentials known as the 1991.
magnetoencephalogram (MEG) or the evoked 5. Stegeman DF, Van Oosterom A, Colon EJ. Far-field
evoked potential components induced by a propa-
magnetic fields (EMF). Although this tech- gating generator: computational evidence. Elec-
nique is still confined largely to research troencephalogr Clin Neurophysiol 67:176-187,
rather than clinical applications, it has al- 1987.
lowed intracranial generators to be localized 6. van Oosterom A. History and evolution of methods
for solving the inverse problem. J Clin Neurophys-
accurately and noninvasively.9"12 Clinical ap- iol 8:371-380, 1991.
plications may include localizing the seizure 7. Scherg M. Functional imaging and localization of
focus by determining the generators of in- electromagnetic brain activity. Brain Topogr
terictal spikes and focal slow waves in pa- 5:103-111, 1992.
tients with intractable partial epilepsy who 8. Wong PK. Source modelling of the rolandic focus.
Brain Topogr 4:105-112, 1991.
are being considered for epilepsy surgery 9. Cohen D, Cuffin BN. EEG versus MEG localization
and using somatosensory magnetic evoked accuracy: theory and experiment. Brain Topogr
fields to accurately locate the somatosensory 4:95-103, 1991.
cortex as part of the planning process for a 10. Stok CJ, Spekreijse HJ, Peters MJ, Boom HB, Lopes
da Silva FH. A comparative EEG/MEG equivalent
surgical procedure, for example, resection dipole study of the pattern onset visual response.
of a brain tumor. Electroencephalogr Clin Neurophysiol Suppl
41:34-50, 1990.
11. Papanicolaou AC, Simos PG, Breier JI, et al. Mag-
netoencephalographic mapping of the language-
SUMMARY specific cortex. J Neurosurg 90:85-93, 1999.
12. Nakasato N, Yoshimoto T. Somatosensory, auditory,
and visual evoked magnetic fields in patients with
This chapter reviews the principles of vol- brain diseases. J Clin Neurophysiol 17:201-211,
ume conduction as applied to the potentials 2000.
Chapter 4
DIGITAL SIGNAL PROCESSING
DIGITAL COMPUTERS IN CLINICAL Repetitive Transient Waveforms

NEUROPHYSIOLOGY Movement-Associated Potentials
Utility DIGITAL FILTERING
Digital Clinical Neurophysiology Types of Digital Filters
Digital Electroencephalography Characteristics of Digital Filters
Construction of Digital Systems TIME AND FREQUENCY DOMAIN
DIGITIZATION ANALYSIS
Principles Interval Analysis
Analog-to-Digital Conversion Autocorrelation Analysis
Quantization Fourier (Spectral) Analysis
Sampling Statistical Analysis
Aliasing Pattern Recognition
COMMON USES OF DIGITAL PROCESSING SUMMARY
AVERAGING
Evoked Potentials and Nerve Conduction
Studies
DIGITAL COMPUTERS IN and evoked potential studies, and it also lends

CLINICAL NEUROPHYSIOLOGY itself to serial comparisons between studies
performed on the same subject at different
Utility times or between two groups of subjects in sci-
entific investigations. Digital computers may
Digital computers can perform types of sig- also partially automate the interpretation of
nal processing not readily available with ana- clinical neurophysiology studies. This chapter
log devices such as ordinary electric circuits. discusses the uses of digital signal processing
Because of their large storage capacities and and storage that are common to many types
rapid, random-access retrieval, they can of physiologic studies.
make the process of obtaining, storing, re-
trieving, and viewing clinical neurophysiol-
ogy data easier. Also, because of their so- Digital Clinical Neurophysiology
phisticated computational abilities, they
may aid in extracting from waveforms in- In recent years, digital instruments have
formation that is not readily obtainable with largely replaced analog instruments. The ad-
visual analysis alone. Furthermore, they are vantages of digital over the analog record-
well suited for quantification of key features ings that had been used for the early work
of waveforms. This may be useful in accu- in each of the fields of clinical neurophysi-
rate clinical diagnosis of electroencephalo- ology derive from the unique capabilities of
graphic (EEG), electromyographic (EMG), digital recording technology. These capabil-
41
ities include1 (1) convenient storage and re- fields of clinical neurophysiology are moving
trieval of records, (2) montage reformatting, steadily toward digital technology. The
(3) filter, sensitivity, and time base changes, greatest impact of digital recordings has
(4) reliability of interpretation, (5) rapid lo- been in EEG, as discussed in the following
cation of events and features of interest, (6) section.
annotating recordings, and (7) quantitative
analysis of background activity and tran-
sients. Digital Electroencephalography
The disadvantages of digital instruments
include the following: Although the accuracy of visual reproduc-
1. Cost—Digital instruments may be more tion of the EEG waveforms on digital EEG
expensive, paticularly in the long term, instruments was limited in the past by the
because with the rapid evolution of resolution of the screen display, this limita-
computer technology, digital instru- tion has been largely overcome by the ready
ments become obsolete more rapidly availability of personal computer graphics
than their analog counterparts did. cards and monitors with resolutions of
2. Maintenance—Repair of digital instru- 1280 X 1024 pixels or higher. Furthermore,
ments requires more knowledge than digital EEG may be combined with digital
is required for analog machines, and video recording for the evaluation of pa-
trouble-shooting is more complex. tients with seizures or spells. With the com-
Maintenance personnel must be knowl- bination of EEG and video, the EEG can be
edgeable about computers and com- correlated with clinical behavior during
puter software as well as hardware. transient spells or seizures. Moreover, by
Digital instruments may be less fault- recording both the EEG and video in a dig-
tolerant, and equipment failures may ital format, events of interest can be located
be more catastrophic with digital sys- quickly during prolonged recordings and
tems, with possible loss of an entire the video can be displayed nearly instanta-
study because of system failure. neously, compared to analog video record-
3. Incompatible data formats—In marked ings that require time-consuming tape
contrast to the relatively standard data searches for the segment of interest. Also,
formats used in the personal computer digital recording of video significantly fa-
industry that facilitate sharing of data, cilitates the editing and copying of video
digital instruments use data formats segments.
that are proprietary to each manufac- Applications of the unique capabilities of
turer, and in general, studies recorded digital recording technology are illustrated
on the instruments of one manufac- in the following discussion of digital EEG.1
turer cannot be read on those of an- 1. Linear display—In contrast to pen-
other. This limits the ability to share based analog recordings on paper, in
studies between laboratories. To sur- which the movement of the pen along
mount this difficulty, some companies the arc of a circle (rather than per-
now offer reader programs for perpendicular to the direction of paper
sonal computers that are capable of movement) causes a nonlinear distor-
reading the data formats used by many tion of waveform morphology when
different manufacturers, but these high-amplitude pan excursions occur,
programs are an additional expense. a digital display accurately represents
4. Obsolescence of data formats—As dig- the waveform morphology indepen-
ital systems evolve and new models dently of signal amplitude.
are released, recording formats may 2. Convenient storage and retrieval of
change over time even with the same records—Multiple digital recordings
manufacturer; thus, eventually, it may (typically hundreds of EEG studies)
be impossible to use current systems may be kept on-line for quick retrieval,
to review studies acquired on older and larger numbers of older record-
instruments. ings (thousands of studies) may be
However, the advantages of digital record- archived on digital media (such as CD-
ing outweigh the disadvantages, and all ROM or DVD-ROM) that require very
Digital Signal Processing 43
little storage space and from which they time the EEG is reviewed, rather than
may be readily retrieved when needed. at the time of recording.
This significantly reduces storage space 5. Reliability of interpretation—A recent
requirements compared with analog study comparing the accuracy of inter-
recordings on paper and eliminates the pretation of digital vs. analog EEG
need for microfilming paper record- recordings demonstrated a clear ad-
ings. With standard computer net- vantage of digital EEG review,2 which
works, recordings (including digital most likely is related to the ability to
video, when applicable) may be viewed view the same EEG segment using sev-
on appropriately configured personal eral different montages, filters, and
computers located at sites remote from sensitivities. In this study, two experi-
the instruments used for recording enced board-certified electroencepha-
without a need to physically transport lographers each read 89 pediatric EEGs
the record. Wide-area networks allow recorded digitally. The studies were
records to be accessed at essentially un- read either in conventional analog pa-
limited distance from the recording lo- per format, using a digital display but
cation, and currently available high- without use of digital tools such as
speed network connections to homes montage reformatting, digital filtering,
allow reading of emergency and after- time base or sensitivity adjustment at re-
hours recordings in the reviewer's view time, or using all the features of a
home. digital system. The inter-reader agree-
3. Montage reformatting—On digital in- ment (kappa) was calculated for each
struments, the EEG montage is se- reading condition. Kappa values of
lected at the time the EEG is reviewed, 0-0.39 represent poor agreement,
rather than at the time of recording. 0.40-0.59 fair agreement, 0.60-0.74
Digital instruments record all data us- good agreement, and 0.75-1.00 excel-
ing a referential montage with a single lent agreement. As shown in Table 4-1,
common reference electrode (such as the inter-reader agreement in classifi-
Cz or an average ear reference). All cation of records as normal vs. abnor-
other montages then can be recon- mal and focal vs. nonfocal was best
structed by simple arithmetic opera- when interpretation was done using
tions on the recorded referential data. digital tools.
In addition to the routine bipolar and 6. Rapid location of events and features
referential montages, special montages of interest—Typical digital EEG instru-
such as a common average reference ments allow rapid paging or scrolling
or a laplacian (source) montage may through the record in the forward or
be used. This is discussed more fully in reverse direction as well as skipping di-
Chapter 12. rectly to specific times or specific events
4. Filter, sensitivity, and time base (marked by the technologist or an-
changes—In a similar fashion, the other person reviewing the EEG).
high- and low-frequency filters and 7. Annotating recordings—During the
notch filter, the vertical display scale recording, technologists may enter tex-
(sensitivity), and the horizontal display tual comments about the recording
scale (time base) are selected at the conditions or the patient's behavior;
Table 4-1. Inter-reader Agreement (Kappa) in Classification of

Electroencephalographic Records
Reading Condition Normal Versus Abnormal Focal Versus Nonfocal
Paper (analog) 0.69 0.46
Digital without reformatting 0.61 0.5
Digital with reformatting 0.81 0.65
these replace the comments that would to represent electric or nonelectric physio-
be written on a paper record. Also, the logic quantities (such as potential in micro-
reviewer may mark entire "pages" of volts, current in milliamperes, pressure in
the EEC record or mark individual millimeters of mercury, or oxygen saturation
waveforms or features and label these in percentage) in one of two ways. An ana-
with text descriptions. log signal takes on any potential (voltage)
8. Quantitative analysis of background ac- within a specific range (for example, — 3-
tivity and transients—This may include 3 V); the potential generally is directly pro-
interval analysis, autocorrelation analy- portional to the physiologic quantity repre-
sis, spectral analysis, statistical analysis, sented by the signal; therefore, that poten-
and pattern recognition (such as auto- tial is an analog of the physiologic quantity.
matic spike or seizure detection) as well Analog signals are generally continuous in the
as cross-correlation and cross-spectral sense that the potential varies continuously
analyses, interpolation, topographic as a function of time. In contrast, a single
displays, multivariate statistical meth- digital signal may take on only one of two
ods, cortical projection techniques, possible potentials (for example, 0 or 3 V);
and source localization. These tech- such a signal may represent one of two pos-
niques are discussed below in this chap- sible states (on or off; yes or no) or one of
ter and in Chapter 12. two possible digits (0 or 1) and is said to rep-
resent one bit (binary digiZ) of information.
Multiple digital signals may be used to rep-
resent a physiologic quantity as a binary
Construction of Digital Systems number (a series of Os and Is forming a
quantity in a base 2 number system; that is,
A digital (computerized) system for acquisi-
tion, storage, and display of physiologic the rightmost digit has a value of 2° = 1, the
second digit from the right has a value of
waveforms has the following key compo-
21 = 2, the third digit has a value of 22 = 4,
nents: (1) electrodes, (2) amplifiers and fil- etc.). Digital signals are discrete and discon-
ters, (3) analog-to-digital converters, (4}
tinuous (that is, they have only two possible
solid state digital memory, (5) digital proces- states), and the nearly instantaneous transi-
sor (central processing unit), (6) magnetic tion from one state to another is made only
or optical disk (or tape) storage, and (7) at specific times. This is the only format in
screen or printer for waveform display.
which digital computers can store and pro-
The electrodes, amplifiers, and filters in a
cess information, and it is most suited to
digital system are essentially identical to performing complex and accurate arith-
those in an all-analog system. The amplified metic operations (for example, adding, sub-
signal for each channel is sent to an analog- tracting, multiplying, dividing) or logical op-
to-digital converter (ADC), which converts it erations (logical conjunction, disjunction,
by the process of digitization to digital form negation). Analog representations are more
and stores it in solid state memory. A digital suited for human interpretation; for exam-
processor is capable of moving digital data
around in memory and processing or ma- ple, a waveform display generally uses verti-
nipulating it; it may also send data to a mag- cal displacement as an analog to the physi-
ologic quantity, such as the potential being
netic or optical disk or tape storage media
for permanent storage, or it may generate displayed, and horizontal displacement as an
analog to elapsed time.
displays of waveforms and related textual an-
notations on a screen or printer.
Analog-to-Digital Conversion
DIGITIZATION Digitization, or analog-to-digital conversion,
is the process by which analog signals are
Principles converted to digital signals. It is the trans-
formation of continuous potential changes in
Electric signals derived from an electrode or an analog signal representing a physiologic
some other type of transducer may be used quantity to a sequence of discrete digital
numbers (binary integers). Digitization is the potential input to the ADC. A simple ex-
performed by a complex circuit known as an ample is shown in Figure 4—1, which shows
analog-to-digital converter (ADC). There a 4-bit ADC, whose input is an analog signal
are two aspects to digitization, quantization in the range 0-16 V, and whose output is a
and sampling. 4-digit binary number that can take on the
values 0, 1, 2, . . . , 15. In this example, any
input potential between 12 and 13V will re-
Quantization sult in the same output (12); thus, the reso-
lution of the ADC (also known as the quan-
Quantization describes the assignment of a tum size) is 1 V. The input range of the ADC
digital number to the instantaneous value of is 0-16 volts.
Figure 4-1. Scheme of a four-bit analog-to-digital converter (ADC). Inputs consist of the continuous signal to be
digitized (range 0-16 V) and a start digitization pulse from a clock that is used to initiate digitization at appropri-
ate times. Outputs consist of four digital signals (+3 or 0 volt representing "1" and "0") that together can encode
a four-bit integer (range 0-15).
In general, the following three terms closely as possible the expected range of
characterize quantization: amplified potentials. If the range of a signal
1. Quantum size (ADC resolution)—This exceeds the ADC range, the ADC will either
determines the minimum potential overflow or underflow and the signal will be
change that can be detected by the distorted (clipped), but if the range of a sig-
ADC and corresponds to a change of 1 nal is too small compared with the ADC
in the least significant bit (2°). A typi- range, much of the resolution of the ADC
cal value might be 1 mV (for the am- will be wasted and the effective resolution
plified signal reaching the ADC). may be insufficient, again distorting the sig-
2. Number of bits in ADC (n)—This de- nal significantly (Fig. 4-2).
termines the range of digitized (out-
put) values. For an ADC that can ac-
cept positive or negative inputs, 1 bit is Sampling
required for sign (+ or —), and the
fractional resolution is then 1 part in In digitization, the conversion of the con-
2n-1. A typical value might be 9-16 bits tinuous analog signal to digital form is usu-
(corresponding to ± 1 part in 256 to 1 ally performed at discrete equidistant time
part in 32,768). intervals. The following two terms charac-
3. Input range—This determines the terize sampling:
maximum and minimum input poten- 1. Sampling interval—This determines
tials. Input potentials above or below the temporal resolution of the digi-
the maximum/minimum are called tizer. A typical value may range from
overflow or underflow, respectively. A typ- 0.01 ms (for brain stem auditory
ical value might be ± 2 volts. evoked potentials) to 5 ms or more (for
The three quantization parameters are re- EEC).
lated by the formula: 2. Sampling frequency—This is the recip-
rocal of the sampling interval and is
Input Range = ± ADC Resolution X (2 n-l _
1) measured in hertz (Hz) (s-1).
In addition to determining the temporal
where n — number of bits. Note that the in- resolution of the digitizer, the sampling fre-
put range of the ADC should match as quency determines the maximum frequency
Figure 4-2. Effect of quantization parameters, that is, analog-to-digital converter (ADC) resolution and input range,
on the fidelity with which an analog signal can be represented digitally. In A, the signal exceeds the input range,
so that its digital representation (D) is clipped. In B, the signal uses more than 50% of the input range and is rel-
atively well represented (E). In C, the signal uses less than 15% of the input range and, because of the limited res-
olution of the analog-to-digital converter, it is poorly represented (F). (From Spehlmann R. Evoked Potential Primer:
Visual, Auditory, and Somatosensory Evoked Potentials in Clinical Diagnosis. Butterworth Publishers, Boston, 1985,
pp 35-52. By permission of the publisher.)
Figure 4-3. Effect of sampling interval and aliasing on the fidelity with which an analog signal can be represented
digitally. In A, the sampling frequency is 14 times that of the signal frequency and the signal is well represented
(D). In B, the sampling frequency is only 6 times the signal frequency, and the representation is less accurate but
still acceptable (E). In C, the sampling frequency is only 1.5 times the signal frequency, and thus less than the
Nyquist frequency; the consequent aliasing causes the digital representation (F) to be entirely misleading in that
it appears to have a frequency that is approximately half the true frequency. (From Spehlmann R. Evoked Poten-
tial Primer: Visual, Auditory, and Somatosensory Evoked Potentials in Clinical Diagnosis. Butterworth Publishers,
Boston, 1985, p 44. By permission of the publisher.)
in the signal to be digitized that can be ad- 25 Hz, not 75 Hz. Aliasing must always be
equately represented. The sampling theorem avoided or else the digitized data will be a
(Nyquist theorem} states that if a signal con- gross misrepresention of the true signal. In
tains component frequencies ranging from practice, aliasing is avoided by filtering the in-
0 to /N, then the minimum sampling fre- put signal before digitization to remove all
quency that can be used for the digitized frequencies above the Nyquist frequency
data to adequately represent the frequency (Fig. 4-3).
content of the original signal is 2/N, where For example, if the sampling interval in
/N is the Nyquist frequency. The Nyquist fre- use is 5 ms, die Nyquist frequency is 100 Hz.
quency can be calculated from the sampling A 70 Hz low-pass filter widi 6 dB per octave
interval as /N = I/(2 X sampling interval). slope would attenuate frequencies of 100 Hz
For example, if/N = 50 Hz, then the sam- to 0.57 of their original amplitude, which
pling frequency must be at least 100 Hz may not be enough. A 50 Hz low-pass filter
(sampling interval of 0.01 second or less). with 12 dB per octave slope would attenuate
This sampling frequency is the minimum nec- frequencies of 100 Hz to 0.2 times their orig-
essary to avoid gross distortion of the input inal amplitude, which may be enough to
signal; a larger sampling frequency (by a fac- prevent significant contamination of the
tor of 3-5) may be necessary in many appli- digitized signal by aliased frequency com-
cations to achieve adequate resolution of ponents, provided that the amplitude of the
fine details in the waveforms being digitized. faster components in the original signal is
relatively small.
Aliasing
COMMON USES OF DIGITAL
Sampling at a frequency lower than 2/N pro- PROCESSING
duces aliasing. Aliasing is distortion of a sig-
nal caused by folding of frequency compo- One common use of digital signal process-
nents in the signal higher than /N onto lower ing in clinical neurophysiology is signal
frequencies. For example, a sine wave of averaging, particularly in evoked-potential
75 Hz, if sampled at 100 Hz, will appear in and sensory nerve conduction studies. Aver-
the digitized data as a sine wave of frequency aging may also be applied to repetitive tran-
sient waveforms and event-related potentials nerve conduction studies. Epoch lengths of
(such as movement-associated potentials). A 10-20 ms are typical for brain stem audi-
second major use of digital signal processing tory evoked potentials and electrococh-
is for digital filtering. Less common but still leograms.
important uses are in time-frequency analy- The basic operation of an averager is
sis, including interval and Fourier (spectral) shown in Figure 4-4. After each stimulus,
analysis, autocorrelation analysis, statistical the input signal is digitized at several discrete
analysis, and automated pattern recognition. sampling times within a fixed length epoch
Other uses tend to be more specialized to that begins at the time of the stimulus. Dig-
particular types of clinical neurophysiologic itized values of potential at each discrete
studies; some of these are discussed else- sample time, each characterized by its la-
where in this book. tency (time after the stimulus), are averaged
for many stimuli; the resulting averaged sig-
nal may be displayed on a screen or printed
AVERAGING on paper. The stimulus-dependent portions
of the signal (the evoked potential or nerve
Evoked Potentials and Nerve action potential) are similar in amplitude
Conduction Studies and latency in each epoch averaged and ap-
pear in the averaged result, whereas the
Digital averaging devices for nerve conduc- stimulus-independent (random) portions of
tion studies and evoked potentials are used the signal (noise and background neuronal
routinely in clinical neurophysiology. Their activity among others) differ substantially
function is similar regardless of the type of from epoch to epoch and are suppressed by
signal averaged, although for different types averaging. The suppression factor, which of-
of studies, the epoch length for averaging ten is called the signal-to-noise ratio, for truly
differs significantly. Epoch lengths of random signals is Vn, where n is the num-
200-500 ms are typical for visual and long- ber of epochs averaged.3 For example,
latency auditory evoked potentials. Epoch achieving a signal-to-noise ratio of 20 re-
lengths of 30-100 ms are typical for middle- quires averaging 400 epochs. The required
latency auditory evoked potentials and for signal-to-noise ratio and, hence, the number
Figure 4-4. Operation of an averager. Analog signals recorded after each stimulus are digitized by an analog-to-
digital converter during a fixed-length time window, or epoch, after the stimulus. The resulting digital representa-
tions are totaled and divided by the number of epochs averaged. The digital result can be displayed by an analog
device such as an oscilloscope after conversion from digital to analog form. (From Spehlmann R. Evoked Potential
Primer: Visual, Auditory, and Somatosensory Evoked Potentials in Clinical Diagnosis. Butterworth Publishers, Boston,
1985, p 37. By permission of the publisher.)
of epochs, depends on the type of signal be- phisticated processing than ordinary forward
ing averaged and the amount of background averaging, because the signal being averaged
activity, or noise. For example, typical brain must be digitized continuously and stored in
stem auditory evoked potentials are about memory so that when a "trigger" occurs, dig-
0.5 fiV in amplitude, whereas background itized data for the previous 0.5-1 second may
EEG activity may be 50 ;u,V or more, requir- be included in the average.
ing a signal-to-noise ratio of 100 (10,000
epochs averaged) . In contrast, sensory nerve
action potentials are typically 10 fJiV or more DIGITAL FILTERING
in amplitude, with noise that is comparable,
requiring a signal-to-noise ratio of only 2-3 Types of Digital Filters
(4-9 epochs averaged) .
A digital filter is a computer program or al-
gorithm that can remove unwanted fre-
Repetitive Transient Waveforms quency components from a signal.4 Just as
for analog filters, they may be classified as
Repetitive transient waveforms that are not low-pass, high-pass, band-pass, or notch fil-
stimulus-related may also be averaged, such ters. Most digital filters function by forming
as epileptic spikes in an EEG or iterative a linear combination (weighted average) of
EMG discharges. The epoch for averaging in signal amplitudes at the current time and
this case is a time "window" around the wave- various past times. The two types of com-
form, for example, from a specified time be- monly used digital filters are the finite im-
fore to a specified time after the peak of the pulse response (FIR) filter and the infinite im-
waveform. The waveforms to be averaged pulse response (IIR) filter. The FIR filter
and the reference times defining the "win- output is a linear combination only of the
dows" around them may be determined input signal at the current time and past
manually by positioning a cursor over the times. This type of filter has a property such
peak of each successive waveform to be av- that its output necessarily becomes zero
eraged or automatically using sophisticated within a finite amount of time after the in-
transient detection programs capable of put signal goes to zero. The IIR filter output
identifying all waveforms of interest and lois a linear combination of both the input sig-
cating their peak and onset. After the epochs nal at the current time and past times (feed-
have been defined, averaging proceeds in forward data, flow) and the output signal at
the same way as for evoked potentials. past times (feedback data flow). This type of
filter has the property that its output may
persist indefinitely in the absence of any fur-
Movement-Associated Potentials ther input, because the output signal itself
is fed back into the filter. Infinite impulse
Movement-associated potentials—one class of response filters can be unstable and also
event-related potentials—are a cerebral ac- have the undesirable property of noise
tivity associated with, and generally preced- buildup, because noise terms created by
ing, a movement (voluntary or involuntary). arithmetic round-off errors are fed back into
They are obtained by simultaneously record- the filter and amplified. For these reasons,
ing several EEG channels and one or more FIR filters are easier to design. However, IIR
EMG channels, the latter to determine the filters often require less computation than
time of occurrence and other characteristics FIR for comparable sharpness in their fre-
of the movement. An EMG channel may act quency responses and, hence, are often used
as a "trigger" for the averager, but the epoch for filtering signals in "real time."
for averaging usually begins before the onset
of the muscle activity as recorded by the
EMG channel and may extend up to or be- Characteristics of Digital Filters
yond the time of the muscle activity. Hence,
this type of averaging is often called back- Digital filters have several characteristics that
averaging. It requires somewhat more so- distinguish them from analog filters. First,
Figure 4-5. Example of digital signal filtering.

An electroencephalogram (EEG) contami-
nated, A, by scalp electromyographic (EMG)
artifact was filtered, B, using a low-pass digital
filter. C, The frequency spectra of the unfil-
tered signals show the large high-frequency
(20 Hz) muscle activity components in the last
two channels before filtering. (From Gotman
et al.5 By permission of Elsevier Science Ire-
land.)
they can be constructed and modified easily or vice versa. The number of zero crossings
because they are software programs rather or other level crossings per unit time is re-
than hardware devices. Second, they can eas- lated to the dominant frequency of the sig-
ily be designed to have relatively sharp fre- nal (Figs. 4-6d and 4-6e). For example, a si-
quency cut-offs if desired, for example, nusoidal signal that crosses zero 120 times
much sharper than the typical 6 dB per oc- every second has a frequency of l/% (120) =
tave roll-off of an analog filter. Third, they 60 Hz.
need not introduce any time delay (phase
shift) in the signal, as invariably happens
with ordinary analog filters; thus, time rela- Autocorrelation Analysis
tionships between different channels can be
preserved even if different filters are used Autocorrelation analysis may be used to rec-
for each.5 ognize the dominant rhythmic activity in a
An example of a segment of EEG con- signal and to determine its frequency. It is
taminated by muscle artifact as it appears be- based on computing the degree of interde-
fore and after application of a digital filter pendence (correlation) between successive
is shown in Figure 4-5. values of a signal. A signal that is truly ran-
dom, such as white noise, will have no cor-
relation between successive values. In con-
TIME AND FREQUENCY trast, a signal with rhythmic components has
DOMAIN ANALYSIS an autocorrelation significantly different
from zero. The autocorrelation function (ACF)
Interval Analysis is defined as the correlation between a sig-
nal and that same signal delayed by time t,
Interval analysis is a method of determining expressed as a function of t. The ACF at
the frequency or repetition rate of wave- t = 0 is always 1, that is, 100% correlation.
forms, which is similar to what is done by vi- For a periodic signal (one with rhythmic
sual inspection. It is based on measuring the components), it is an oscillating function of
distribution of intervals between either zero t with a frequency like that of the dominant
or other level crossings or between maxima rhythmic component in the original signal
and minima of a signal.6 A zero crossing (Fig. 4-6c).
occurs when the potential in a channel
changes from positive to negative or vice-
versa. A level crossing (used less often than Fourier (Spectral) Analysis
a zero crossing) occurs when the potential
in a channel changes from greater than to Fourier analysis is the representation of a pe-
less than a given value (for example, 50 riodic function as a Fourier series, a sum of
Figure 4-6. Examples of several types of signal analysis: a, an electroencephalographic (EEC) signal; b, its power
spectrum; c, its autocorrelation function; d, the distribution density function of intervals between any two succes-
sive zero crossings; and e, the distribution density function of the intervals between successive zero crossings at which
the signal changes in the same direction, that is, from positive to negative or vice versa. (From Lopes da Silva.6 By
permission of John Wiley & Sons.)
trigonometric functions, that is, sines and times 0, T, 2T, . . . , (N- 1) T, where T is the
cosines. A Fourier series may be used to sampling interval. The output is the ampli-
approximate any periodic function. The tudes of AT/2 - 1 sine waves at frequencies/
greater the number of terms in the series, 2/ . . . , (AT/2 - I)/and of A//2 + 1 cosine
the greater the accuracy of the approxima- waves at frequencies 0, f, 2/ . . . , (AT/2)/
tion will be. The Fourier transform is a math- Note that there is no "data reduction"; the
ematical method to analyze a periodic func- number of input values (N) equals the num-
tion into a sum of a large number of cosine ber of output values.
and sine waves with frequencies of f, 2f, 3f, By convention, both the cosine and sine
4f, etc., where / is the lowest, or fundamen- waves of a given frequency are often lumped
tal, frequency in the function analyzed, and together into one quantity describing the
2/ 3f, and so forth, are harmonics. The input amount of that frequency present in the sig-
of the Fourier transform is the periodic func- nal. This quantity is called the spectral in-
tion, or signal, to be analyzed. The output is tensity or power, and a plot of this as a func-
the amplitude of each sine and cosine wave tion of frequency is the power spectrum (Fig.
in the series. 4-6b). The intensity, or power, is the square
In working with discrete data, that is, a sig- of the amplitude. The phase (phase angle) for
nal sampled at equal time intervals 0, T, 2T, any given frequency describes how much of
3T, . . . , (AT- 1)T, a discrete Fourier trans- that frequency is in the form of a cosine wave
form is used. In this case, the fundamental and how much is a sine wave. The following
frequency can be shown to be/= 1 /NT. The formulas relate these quantities; here, C is
input of the discrete Fourier transform is the cosine wave amplitude and S is sine wave
N digitized values representing the signal at amplitude:
Intensity trols. (3) Determine by statistical analysis

Amplitude A whether it is possible to reliably separate the
two groups of waveforms on the basis of the
Phase calculated features. For example, one could
Arctan
calculate the rising and falling slope of can-
didate sharp waves, compute these slopes for
Note: (f> = 0 for pure cosine wave true epileptic sharp waves and for other tran-
(f> = 90° for pure sine wave. sients such as muscle artifacts or nonepilep-
tic sharp transients in background activity,
and determine whether a certain range of
Statistical Analysis slopes characterizes the true sharp waves.
These techniques have been used with some
Statistical analysis of a digitized signal may
be a useful data reduction technique. In ef- success to detect spikes and sharp waves,
spike-and-wave bursts, sleep spindles and
fect, it treats digitized values at successive
K-complexes, and seizure discharges in EEG
time points as independent values of a ran-
recordings7 and to detect motor unit po-
dom variable. In this technique, one may
plot the amplitude distribution—the num- tentials, fibrillation potentials, and other it-
ber of digitized samples of the signal having erative discharges in EMG recordings.
a given amplitude value vs. the amplitude it-
self—and visually inspect the shape of the
distribution. Alternatively, one may calculate
SUMMARY
the moments of the probability distribution of sig-
nal amplitudes, including the following:6 This chapter reviews the principles of digiti-
zation, the design of digitally based instru-
First moment, mean voltage m\ (center of
ments for clinical neurophysiology, and sev-
distribution)
eral common uses of digital processing,
Second central moment, variance m?, and
including averaging, digital filtering, and
standard deviation cr = Vw 2 (width of
distribution) some types of time-domain and frequency-
domain analysis. An understanding of these
Third central moment m$ and skewness
principles is necessary to select and use dig-
)8i = w 3 /(w2) 3/2 (asymmetry of distribu-
itally based instruments appropriately and to
tion)
understand their unique features.
Fourth central moment m^ and kurtosis
excess j8g = m^/(m^ (peakedness or flat-
ness of distribution).
REFERENCES
Pattern Recognition 1. Fisch BJ. Fisch and Spehlmann's EEG Primer: Basic
Principles of Digital and Analog EEG, 3rd ed. Elsev-
Pattern recognition algorithms are designed ier, Amsterdam, 1999.
2. Levy SR, Berg AT, Testa FM, Novotny EJ Jr, Chiappa
to detect a specific waveform in a signal that KH. Comparison of digital and conventional EEG in-
has characteristic features, such as a motor terpretation. J Clin Neurophysiol 15:476-480, 1998.
unit potential in an EMG or a sharp wave in 3. Misulis KE. Spehlmann's Evoked Potential Primer:
an EEC. The characteristic features may be Visual, Auditory, and Somatosensory Evoked Poten-
tials in Clinical Diagnosis, 2nd ed. Butterworth-
defined in the time domain (for example, Heinemann, Boston, 1994, pp 25-37.
durations, slopes, and curvature of wave- 4. Maccabee PJ, Hassan NF. AAEM minimonograph
forms), in the frequency domain (after fil- #39: Digital filtering: basic concepts and application
tering signal), or in both. One common ap- to evoked potentials. Muscle Nerve 15:865-875,1992.
proach in developing a pattern recognition 5. Gotman J, Ives JR, Gloor P. Frequency content of
EEG and EMG at seizure onset: possibility of removal
algorithm is as follows: (1) Define a set of of EMG artefact by digital filtering. Electroen-
candidate features and how to calculate cephalogr Clin Neurophysiol 52:626-639, 1981.
them. (2) Calculate the chosen features for 6. Lopes da Silva FH. Computerized EEG analysis: a tu-
a visually selected collection of waveforms of torial overview. In Halliday AM, Butler SR, Paul R
(eds): A Textbook of Clinical Neurophysiology. John
the type to be detected, the learning set, and Wiley & Sons, Chichester, England, 1987, pp 61-102.
for a collection of similar waveforms deter- 7. Gotman J. Autonomic detection of seizures and
mined not to be of the required type, con- spikes. J Clin Neurophysiol 16:130-140, 1999.
Chapter 5
ELECTROPHYSIOLOGIC
GENERATORS IN CLINICAL
NEUROPHYSIOLOGY
PHYSIOLOGIC GENERATORS Muscles

Resting Membrane Potential Sweat Glands
Electrotonic Membrane Potentials Spinal Cord
Action Potentials Brain Stem
Postsynaptic Potentials Special Sensory Receptors
Other Membrane Potentials Optic and Auditory Pathways
STRUCTURAL GENERATORS Cerebral Cortex
Peripheral Nerves SUMMARY
PHYSIOLOGIC GENERATORS tion gradients across the semipermeable cell

membrane.
The bioelectric potentials recorded during Ionic concentrations inside and outside a
electrophysiologic studies are generated by cell are characteristically different. For ex-
sources inside the body that may be some ample, potassium is more concentrated in-
distance from the recording electrodes. tracellularly and sodium extracellularly. The
Sources may be classified according to mode value of the resting membrane potential of
of generation at the cellular level as elec- the cell depends on the concentration gra-
trotonic membrane potentials, action po- dient for each ion species as well as the rel-
tentials, postsynaptic potentials, or other ative permeability of the membrane to each
membrane potentials. These sources all de- ionic species. The Nernst potential for each
rive from ionic current flow into or out of ionic species is the resting membrane po-
cells, which is made possible because all ex- tential that would result if the membrane
citable cells, such as neurons and muscle were permeable to that species alone. Nernst
fibers, have a resting membrane potential. potentials for four ionic species are shown
in Table 5-1.
Driving forces of two types are responsible
Resting Membrane Potential for fluxes of ions across the cell membrane.
The chemical driving force causes ions to move
The resting membrane potential of neurons and from a region of high concentration to one
muscle fibers is defined as the electric po- of lower concentration. The electrical driving
tential inside the cell minus that outside the force causes positively charged ions to move
cell. It typically has a value of approximately from a region of higher potential to one of
—66 mV and is caused by ionic concentra- lower potential, whereas negatively charged
53
Table 5-1. Ncrnst Potential for Four Ionic Species

Intracellular Extracellular
Concentration Concentration
Ion (mmol/L) (mmol/L) Nernst Potential (mV)
K+ 68 4 -76
Na+ 15 142 +60
cr 9 105 -66
Ca2+ 0.0005 0.125 +74
ions move in the opposite direction. The net membrane potential from its equilibrium
driving force is the sum of the chemical and value.
electrical driving forces. The net ionic flux
across the membrane is the product of the
net driving force and the membrane per- Electrotonic Membrane Potentials
meability to the ion. In the resting state,
there is a net flux of sodium into the cell Ekctrotonic membrane potentials result from the
and a net flux of potassium out of the cell, passive electrical properties of the cell.
but there is no net flux of chloride because These are the membrane properties men-
the net driving force is zero. The steady-state tioned above (total membrane conductance
concentration gradient of sodium and potas- per unit area gtot and membrane capaci-
sium ions across the cell membrane is main- tance per unit area Cm) as well as the intra-
tained by an active, energy-consuming pro- cellular axial resistance per unit length ra.
cess that pumps potassium into the cell and The membrane time constant determines
sodium out of the cell. the behavior of a localized portion of the
An electrical equivalent circuit for the membrane when an additional current is in-
membrane can be constructed in which the jected into the cell. For example, this cur-
chemical driving force for each ionic species rent could be physiologic in origin (caused
is represented as an electrical battery (a seat by activation of ion channels) or could be
of electromotive force given by the Nernst introduced experimentally by a microelec-
potential for that ion) and the membrane trode impaling the membrane. Because of
permeability for each ionic species is repre- its resistive and capacitive properties, the
sented as a resistor. The flux of each ionic membrane potential changes exponentially
species represents a flow of electric current from its resting value to a new equilibrium
across the membrane. value when the current is turned on and re-
Important membrane electrical proper- turns exponentially to its resting value when
ties include the electromotive force for each the current is turned off (Fig. 5-1). The time
ion (that is, the Nernst potential) deter- required for the potential to reach 63% of
mined by the ionic concentration, the con- the new value is the time constant, given by
ductance per unit area (g"Na, gcl, £K) for each T = Cm/gtot = RC. Here, R = l/gtotA is the
ion (the product of the number of ion chan- total cell input resistance, and C = CmA is
nels per unit area and the conductance of the total cell capacitance (A is the total cell
each channel), and the capacitance per unit membrane area).
area (Cm) of the membrane (typically, 20 Cable properties of axons and dendrites
mF/m2). or muscle fibers determine the behavior of
The equilibrium, or resting, potential of a long cylindrical cell process when an ad-
the membrane can be calculated from the ditional current is injected into one location.
equivalent circuit for the membrane. Any Because of the resistive and capacitive prop-
change in membrane permeability to one or erties of the process, the membrane poten-
more ionic species or any externally injected tial falls off exponentially with increased
current across the membrane alters the distance from the site of current injection
Electrophysiologic Generators in Clinical Neurophysiology 55
Figure 5-1. The actual shape (c) of the response of the membrane to a rectangular current pulse is intermediate
between that of a pure resistive element (a) and a pure capacitive element (b). The product of the membrane's re-
sistance (R) and the capacitance (C) is called the membrane time constant (T). The total membrane current (7m) is
shown by the solid line in the lower half of the figure; dotted lines /; and 7C show the time course of ionic and ca-
pacitive current, respectively. (From Koester J. Passive membrane properties of the neuron. In Kandel ER, Schwartz
JH, Jessell TM [eds]. Principles of Neural Science, 3rd ed. Appleton & Lange, Norwalk, Connecticut, 1991, p 97.
By permission of The McGraw-Hill Companies.)
(Fig. 5-2). The distance in which the po- The propagation velocity of an action po-
tential change from the resting value tential along myelinated axons is faster than
reaches 37% of its maximum is the mem- that along unmyelinated axons because of
brane space constant A, given by the effect of the myelin sheath on the cable
properties of the axon. Specifically, the
myelin sheath decreases the membrane ca-
pacitance and conductance, decreases the
time constant, and increases the space con-
where a is the radius of the axon or dendrite stant of the segment of axon between the
cylinder. nodes of Ranvier. For unmyelinated axons,
the action potential propagation velocity
v <* a1/2, whereas for myelinated axons, v « a,
Action Potentials where a is the axon radius.
The voltage clamp technique is used ex-
Action potentials are regenerative changes in perimentally to measure the membrane
membrane potential that propagate along a conductance changes underlying an action
cell process, for example, an axon, dendrite, potential. In this technique, an external ap-
or muscle fiber. They have a depolarizing paratus connected to a microelectrode im-
phase, in which the membrane potential re- paling the cell maintains the membrane
verses sign from negative to positive, fol- potential clamped at a fixed value and mea-
lowed by a repolarizing phase, in which the sures the net current flow into or out of the
potential returns to its resting value. cell, from which conductance can be calcu-
Figure 5-2. A, Current injected into

a neuronal process with a microelec-
trode follows the path of least resis-
tance to the return electrode in the
extracellular fluid. B, The change in
Vm produced by focal current injec-
tion decays exponentially with dis-
tance along the length of the process.
(From Koester J. Passive membrane
properties of the neuron. In Kandel
ER, Schwartz JH, Jessell, TM [eds].
Principles of Neural Science, 3rd ed.
Appleton & Lange, Norwalk, Con-
necticut, 1991, p 99. By permission of
The McGraw-Hill Companies.)
lated. To separate the contributions of vari- through an intrinsic inactivation time con-
ous ionic channels to the net current flow, stant also contributes to the process and is
pharmacologic agents are used to block probably the dominant mechanism in mam-
channels selectively; for example, tetrodo- malian neurons.
toxin blocks sodium channels and tetraeth- Hodgkin and Huxley1 formulated a math-
ylammonium blocks potassium channels. ematical model of the action potential in
With these techniques, the basis of the giant squid axons based on their experi-
ionic fluxes that occur during action po- mental measurements of the time and volt-
tentials has been determined. Rapidly open- age dependence of sodium and potassium
ing voltage-sensitive sodium channels are re- conductances. This model successfully re-
sponsible for the depolarization phase of produced the observed time course of the
the action potential. Because of the factor action potential (Fig. 5-3). Similar models
of 400 increase in membrane sodium per- have been developed for many types of ver-
meability at the peak of the action potential tebrate neurons.
compared with its resting value, the mem-
brane potential comes close to the Nernst
potential for sodium. More slowly opening Postsynaptic Potentials
voltage-sensitive potassium channels are
partially responsible for the repolarization Postsynaptic potentials (PSPs) are one type of
of the membrane, and may produce a tran- nonregenerative and nonpropagating cell
sient hyperpolarization, after the action po- potential. They are membrane potentials
tential. The closing of the sodium channels caused by neurotransmitter-induced open-
Figure 5-3. The action potential can be reconstructed from the changes in gfra and g^ that result from the open-
ing and closing of Na+ and K+ voltage-gated channels. (From Koester J. Voltage-gated ion channels and the gen-
eration of the action potential. In Kandel ER, Schwartz JH, Jessell TM [eds]. Principles of Neural Science, 3rd ed.
Appleton & Lange, Norwalk, Connecticut, 1991, p 110. By permission of The McGraw-Hill Companies, as adapted
from Hodgkin AL. The Conduction of the Nervous Impulse. Liverpool University Press, Liverpool, 1964, p 63. By
permission of the author.)
ing or closing of ion channels in a postsyn- Other Membrane Potentials

aptic neuron or muscle fiber and underlie
normal synaptic transmission. Other types of membrane potentials include
Excitatory PSPs (EPSPs) are depolarizing those caused by other ionic channel types,
potentials in the postsynaptic membrane especially in neurons in the central nervous
caused by the effects of an excitatory neu- system. These include calcium channels and
rotransmitter. Commonly, EPSPs are pro- other types of sodium and potassium chan-
duced by the opening of channels to sodium nels. These additional channel types are re-
and calcium and, less commonly, by the clos- sponsible for the rich variety of behaviors
ing of potassium channels. seen in central nervous system neurons and
Inhibitory PSPs (IPSPs) are hyperpolarizing probably contribute to such phenomena as
potentials in the postsynaptic membrane rhythmic oscillations and epileptic bursts in
caused by the effects of an inhibitory neu- cortical neurons and adaptation and re-
rotransmitter. Commonly, IPSPs are pro- cruitment patterns in spinal motor neurons,
duced by the opening of potassium or chlo- although interneuronal connectivity is also
ride channels and, less commonly, by the important in these phenomena.
closing of sodium or calcium channels. Note Slow calcium spikes are depolarizing poten-
that postsynaptic inhibition may occur in the tials mediated by slow calcium channels.
absence of a detectable IPSP, because open- These channels are activated by membrane
ing of potassium or chloride channels may depolarization, and they are inactivated by
merely decrease the cell input resistance— repolarization, by an intrinsic slow inactiva-
thereby making the cell harder to depolar- tion time constant, and by increasing intra-
ize—without significantly hyperpolarizing cellular calcium concentration. Slow cal-
the cell. cium spikes occur in cortical pyramidal
tions caused by activation of slowly inacti-

vating, or persistent, sodium channels. Be-
cause late afterdepolarizations may lead to
oscillations, they may be responsible for
many types of rhythmic cortical activity.
Afterhyperpolarizations are prolonged hy-
perpolarizing potentials that follow a so-
dium- or calcium-mediated depolarizing po-
tential. They commonly are mediated by
slow or calcium-dependent potassium chan-
nels. These channels are activated by in-
creasing intracellular calcium concentration
or by membrane depolarization, and they
are inactivated by decreasing intracellular
calcium concentrations, membrane repolar-
ization, and a relatively long intrinsic inacti-
vation time constant. Afterhyperpolariza-
tions may last for hundreds of milliseconds
in some neuron types, for example, spinal
motor neurons, and thus limit the maximal
firing rate of the cell. Even more prolonged
afterhyperpolarizations may occur because
of activation of so-called early potassium
Figure 5-4. Relationship between surface-recorded channels, which are activated by membrane
electroencephalographic (EEG) discharges and intra-
cellular and extracellular activity in a cortical epileptic hyperpolarization and act to delay the return
focus in an experimental animal. (From Martin JH. The of the membrane potential to equilibrium.
collective electrical behavior of cortical neurons: the Dendritic spikes are small depolarizing po-
electroencephalogram and the mechanisms of epilepsy. tentials recorded from dendrites. They are
In Kandel ER, Schwartz JH, Jessell, TM [eds]. Princi-
ples of Neural Science, 3rd ed. Appleton & Lange, Nor- mediated primarily by a type of slow calcium
walk, Connecticut, 1991, p 787. By permission of The channel in certain regions of dendritic
McGraw-Hill Companies, as adapted from Ayala GF, membrane, the active or booster zones, and
Dichter M, Gumnit RJ, Matsumoto H, Spencer WA. they provide amplification and regeneration
Genesis of epileptic interictal spikes. New knowledge of of postsynaptic potentials on dendritic
cortical feedback systems suggests a neurophysiological
explanation of brief paroxysms. Brain Res 52:1-17, branches far from the soma (that is, more
1973. By permission of Elsevier Science Publishers.) than 1 or 2 length constants distant) which
would otherwise not be able to have a sig-
nificant effect on the soma potential if
neurons and other neuron types. They are only electrotonic (passive) conduction were
involved in intrinsic bursting behavior, in involved.2
which multiple high-frequency fast sodium
spikes, or action potentials, occur on the slow
calcium spike, and in the epileptic paroxys- STRUCTURAL GENERATORS
mal depolarization shift (Fig. 5-4).
Afterdepolarizations are prolonged or late The variety of clinical neurophysiologic stud-
depolarizing potentials that follow a sodium- ies corresponds to a variety of structural gen-
mediated action potential. Early afterdepo- erators in the body, including muscles, sweat
larizations commonly are mediated by the glands, peripheral nerves, and various com-
slow calcium channels described above. Sev- ponents of the central nervous system. Each
eral types of late afterdepolarizations have structural generator may have associated
also been described, including rebound cal- with it several different types of physiologic
cium spikes (which occur after hyperpolar- potential. In some cases, the activity result-
izing events and are caused by activation of ing from different physiologic potentials can
low-threshold inactivating calcium chan- be easily distinguished, but in many cases
nels) and persistent long-lasting depolariza- (particularly sensory pathways in the central
nervous system assessed by evoked potential is the intracellular trigger for the contractile
studies), complete knowledge of the physio- process.
logic generator underlying any recorded Muscle end plate potentials are muscle fiber
waveform is lacking. EPSPs that originate at the end plate, or neu-
romuscular junction. Both sodium and
potassium ions flow through the channel
Peripheral Nerves opened by acetylcholine-receptor binding,
so that the reversal potential, that is, the
Peripheral nerves consist of axons and sup- membrane potential in a voltage clamp ex-
porting structures, including myelin-pro- periment at which the ionic current flow re-
ducing Schwann cells, connective tissue, and verses from net inward to net outward, for
bloodvessels. They contain three types of ax- this channel is near zero volts, intermediate
ons. Motor axons originate from neurons in between the Nernst potentials of sodium and
the spinal cord (spinal motor neurons) or potassium. Normal end plate potentials are
brain stem (cranial nerve motor neurons) caused by the simultaneous release of hun-
and synapse on muscle fibers. Sensory axons dreds of quanta, or packets, of acetylcholine,
originate from neurons in spinal dorsal root which occurs when an action potential
ganglia or cranial nerve ganglia; these axons reaches the nerve terminal. However, minia-
terminate in skin, muscle, or other organs ture end plate potentials are much smaller and
in specialized sensory receptors, for exam- are caused by the spontaneous, random re-
ple, pacinian corpuscles or muscle spindles, lease of a single packet of acetylcholine from
or as "bare" nerve terminals. Autonomic ax- the nerve terminal. End plate noise re-
ons originate either in neurons of the spinal corded from needle electrodes in the vicin-
cord or brain stem (preganglionic neurons) ity of the muscle end plate is caused by
or in neurons in autonomic ganglia (post- miniature end plate potentials. End plate
ganglionic neurons located in the sympa- spikes are action potentials of muscle fibers
thetic trunk ganglia for many sympathetic caused by mechanical activation, for exam-
neurons or in visceral ganglia near the end- ple, by the electromyographic needle, of
organ innervated for all parasympathetic nerve terminals in the end plate region and
and some sympathetic neurons). The sum of are mediated by normal neuromuscular
the propagating action potentials of all stim- synaptic transmission.
ulated sensory axons in a motor or mixed Muscle action potentials are similar to
nerve can be recorded as a sensory nerve ac- nerve action potentials but have a generally
tion potential during sensory nerve conduc- slower propagation velocity. They propagate
tion studies. These primarily involve the outward—often in two opposite directions si-
large diameter, or IA and IB, sensory axons multaneously—from the vicinity of the neu-
in the nerve, because only they are stimu- romuscular junction to the ends of the mus-
lated by conventional electrical stimuli. Gen- cle fiber. Muscle action potentials may occur
erally, motor and autonomic axons are spontaneously in individual muscle fibers
tested only indirectly by stimulating the (fibrillation potentials), simultaneously in
nerve and observing the postsynaptic effects all muscle fibers that are part of the same
in muscle and sweat glands. motor unit (for example, voluntary motor
units or involuntary fasciculation poten-
tials) , or nearly simultaneously in all muscle
Muscles fibers supplied by one motor or mixed nerve
(leading to the surface-recorded compound
Muscles consist of muscle fibers and con- muscle action potential in motor nerve con-
nective tissue; both motor and sensory axons duction studies).
traverse muscles. Contraction of muscle
fibers is initiated by neuromuscular synaptic
transmission (with acetylcholine as the neu- Sweat Glands
rotransmitter), which leads to a propagated
muscle action potential. This in turn causes Sweat glands are end organs innervated by
calcium to enter the muscle fibers. Calcium sympathetic postganglionic neurons that
ons of central and peripheral neurons ar-

ranged in various ascending and descending
tracts. Many synapses occur on spinal cord
neurons. Thus, potentials originating from
generators in the spinal cord include EPSPs,
IPSPs, and action potentials. Although the
health and quantity of spinal motor neurons
in the ventral horn can be assessed to some
degree by standard electromyographic and
nerve conduction studies, specific potentials
produced by spinal motor neurons them-
selves are not routinely measured. During
somatosensory evoked potential studies, re-
figure 5-5. Physiology of two postganglionic axon recordings from cervical or cervical-cranial
flexes: the sudomotor axon reflex and the veno-arteri- derivations may demonstrate an Nil wave-
olar reflex. The former is the basis of the quantitative form that is thought to be generated pri-
sudomotor axon reflex test, but the latter is not used marily at the root entry zone in the dorsal
in any clinical test. (From Low PA. Quantitation of au-
tonomic function. In Dyck PJ, Thomas PK, Grifiin JW,
horn at the level of spinal segments C5 and
Low PA, Poduslo JF [eds]. Peripheral Neuropathy, 3rd C6. Because this waveform manifests a la-
ed. WB Saunders, Philadelphia, 1993, p 731. By per- tency shift with recording at progressively
mission of the publisher.) higher levels, it may be because of a propa-
gated action potential in the sensory axons
ascending in the dorsal columns.
have acetylcholine as their neurotransmitter.
Many of these axons have multiple branches
in the skin, and a single axon may innervate Brain Stem
many sweat glands. Release of acetylcholine
by sympathetic nerve terminals in response The brain stem somatosensory pathways in-
to various stimuli, for example, an electric clude the dorsal column nuclei in the lower
stimulus applied to a contralateral limb, medulla and the medial lemniscus. Record-
leads to a prolonged depolarization of sweat ings during somatosensory evoked potential
glands that can be recorded diffusely from studies demonstrate an N13 waveform
the skin surface in certain areas, a periph- thought to be generated by the upper cer-
eral autonomic surface potential. Function vical cord dorsal columns (presynaptic ac-
of peripheral sympathetic axons and sweat tion potential) or by the dorsal column nu-
glands is also assessed with the quantitative clei (postsynaptic potential) or by both. A
sudomotor axon reflex test. In this test, a PI4 waveform may also be seen; it is thought
controlled release of acetylcholine into one to be generated by the medial lemniscus (ac-
skin region leads to reflex depolarization of tion potential).
autonomic axons that propagate antidromi-
cally to a "Y" branch point and orthodromi-
cally from there to activate sweat glands Special Sensory Receptors
synaptically in a nearby region of skin, thus
producing sweat. Sweat gland activity (sweat
production) is quantified by measuring the Sensory receptors in the visual and auditory
systems generate characteristic potentials
water output of the glands3 (Fig. 5-5). that can be recorded with appropriate
evoked potential studies.
The retina contains photoreceptors, the
Spinal Cord rods and cones, and other types of neurons,
including bipolar, horizontal, amacrine, and
The spinal cord contains cell bodies of mo- ganglion cells. Retinal visual evoked poten-
tor, autonomic, and other neurons and ax- tials, the electroretinogram, can be recorded
from electrodes placed near or on the eye reversal visual evoked potential. Cortical
and are thought to be caused by summed neurons include both pyramidal cells (exci-
postsynaptic potential activity in retinal tatory neurons that provide the major out-
neurons. put of the cerebral cortex) and stellate cells
Hair cells are sensory receptors in the (excitatory or inhibitory interneurons). The
cochlea. They release a neurotransmitter long apical dendrites of pyramidal cells are
that activates the peripheral axons of the perpendicular to the cortical surface. Each
bipolar cells of the spiral ganglion, which in pyramidal cell has an extensive dendritic
turn conduct action potentials to the central tree on which 1000-100,000 synapses may
axons that make up the auditory nerve and occur.
synapse in the cochlear nucleus in the lower The neocortex consists of six cellular lay-
pons. Some components of the electro- ers; layer I is most superficial and layer VI is
cochleogram, that is, the cochlear micro- the deepest layer. Layer I contains mainly
phonic and the summating potential, are glial cells and axonal and dendritic pro-
thought to be produced by sensory receptor cesses. Layer IV is most developed in sensory
potentials in the hair cells. Wave I of the areas of the cortex and receives much of the
brain stem auditory evoked potential (BAEP) specific thalamocortical projections. Layer V
is caused by the propagated action potentials is most developed in motor areas of the cor-
in the auditory nerve. tex in which many of the pyramidal cells are
exceptionally large and project particularly
to distant sites, including the brain stem and
spinal cord.
Optic and Auditory Pathways Brodmann divided the cortex into 52 cor-
tical areas on the basis of cell size, neuron
The optic and auditory pathways are gener-
density, myelinated axon density, and num-
ators of later components of the visual and
ber of layers. The primary somatosensory
auditory evoked potentials. Propagated ac-
cortex (areas 3, 1, and 2) is the likely gen-
tion potentials in the optic tracts or optic ra-
erator of the scalp components of the so-
diations may contribute to some variable
matosensory evoked potential. Primary vi-
early components of diffuse light-flash visual
sual cortex (area 17) and visual association
evoked potentials, but these have little clin-
cortex (areas 18 and 19) are the likely gen-
ical usefulness. Wave II of the BAEP is
erators of the PI00 component of the visual
thought to be generated either by propa-
evoked potential. Auditory cortex (areas 41
gated action potentials in the auditory nerve
as it enters the brain stem or by postsynap- and 42) may be the generator of the late
components of the long-latency auditory
tic potentials in the cochlear nucleus. Wave
evoked potential.
III may be generated by postsynaptic poten-
It is known that postsynaptic potentials—
tials in the superior olivary nucleus and
not action potentials—in cortical neurons
waves IV and V by the lateral lemniscus
are responsible for all scalp-recorded elec-
(propagated action potentials) or the infe-
trical activity. This is because postsynaptic
rior colliculus (postsynaptic potentials).4
potentials are of long duration (10s-100s of
Later waves (VI and VII) may arise from the
milliseconds), involve large areas of mem-
medial geniculate body and the auditory
brane surface, occur nearly simultaneously
radiations.
in thousands of cortical pyramidal cells, and
occur especially in pyramidal cell dendrites
that are uniformly perpendicular to the cor-
Cerebral Cortex tical surface. These properties all allow post-
synaptic potentials to summate effectively to
The cerebral cortex is the generator of es- produce a detectable scalp potential. In con-
sentially all electroencephalographic (EEG) trast, action potentials are brief (1 ms), in-
activity recorded without averaging as well as volve small surface areas of membrane (ax-
the late components of evoked potentials, in- ons), occur at random and widely spaced
cluding the PI00 component of the pattern- intervals in various neurons, and propagate
along axons that are oriented in many di- pathologic changes demonstrated in the un-
rections, all of which make effective sum- derlying generators.
mation impossible.
REFERENCES
SUMMARY
1. Hodgkin AL, Huxley AF. A quantitative description
This chapter reviews the generators of elec- of membrane current and its application to con-
trophysiologic potentials in terms of basic duction and excitation in nerve. J Physiol (Lond)
cellular electrophysiology and the anatomi- 117:500-544, 1952.
cal structures that generate electrophysio- 2. Kandel ER, SchwartzJHJessell TM (eds). Principles
of Neural Science, 4th ed. McGraw-Hill, New York,
logic potentials of clinical interest. Knowl- 2000.
edge of the generators of the potentials 3. Dyck PJ, Thomas PK, Griffin JW, Low PA, Poduslo
recorded in clinical neurophysiologic stud- JF (eds). Peripheral Neuropathy, 3rd ed. WB Saun-
ies is helpful in understanding the charac- ders, Philadelphia, 1993.
4. Melcher JR, Guinan JJ Jr, Knudson IM, Kiang NY.
teristics and distribution of the recorded po- Generators of the brain stem auditory evoked po-
tentials and is the first step in correlating the tential in cat. II. Correlating lesion sites with wave-
alterations seen in disease states with the form changes. Hear Res 93:28-51, 1996.
Chapter 6
CLASSIFICATION OF WAVEFORM
CHARACTERISTICS
Jasper R. Daube
CONTINUOUS WAVEFORMS
EVENT RECORDING
SUMMARY
The waveforms that make up the electric is hertz (Hz). A 10 Hz signal is a waveform
signals generated by the central, periph- that has continuous variation 10 times per
eral, and autonomic nervous systems and second. The traces in Figure 6-1 have wave-
muscle described in Chapter 5 are classified forms varying at five basic rates (3 Hz, 3.5
according to the variables that characterize Hz, 6 Hz, 10 Hz, and 20 Hz). The continu-
them. These variables identify the signals ous waveforms generated by physiologic
and demonstrate the abnormalities that oc- generators can be described by their com-
cur in disease. Each waveform is a change ponent frequencies. The change of a con-
over time in the electric potential differ- tinuous waveform from one frequency to an-
ence between two recording points. If no other may occur because a single generator
change occurs in the potential difference, changes its rate of activity or because one
a flat line with no signal is recorded, even generator working at 3 Hz becomes inactive
if there is a voltage difference between the while another one working at 6 Hz becomes
points. Changes in the difference in po- active.
tential are broadly classified as continuous If the recording electrodes are located
and intermittent. near two structures that simultaneously gen-
erate signals of different frequencies, a more
complex signal is recorded. Signals illustrat-
ing the combination of two frequencies are
CONTINUOUS WAVEFORMS shown in Figure 6-1. Note that the combi-
nation of signals of both frequencies is still
A continuously varying signal is described by recognized if the frequencies are widely dif-
the rate of change of the signal (cycles per ferent. Combinations of waveforms with sim-
second), the size or amplitude of the signal ilar frequencies result in less recognizable
(peak-to-peak), the character of the wave- waveforms and even temporary obliteration
form, and the consistency of the signal over of the signal (3 Hz and 3.5 Hz). Such com-
time. Continuous waveforms recorded from plex physiologic waveforms can still be de-
living tissue are usually sinusoidal, as shown scribed in terms of the component fre-
in Figure 6-1. The rate of change of a sig- quencies of the signal. Waveforms become
nal is known as its frequency and is measured even more complex when differently shaped
in cycles per second. The term more com- waves of the same frequency summate, as
monly used to describe this rate of alteration shown in Figure 6-2.
63
Figure 6-1. Simple, regular, sinusoidal waves of different frequencies combine to form complex waveforms. A, A 3
Hz and a 20 Hz waveform summate into a waveform in which both are still recognizable. B, A 3 Hz and a 10 Hz
waveform summate to form a more complicated waveform in which the components are less recognizable. C, Sum-
mation of a 3 Hz and a 6 Hz waveform results in an apparently regular, more complex waveform. D, Summation
of a 3 Hz and a 3.5 Hz waveform results in fluctuation of the waveform as the components go in and out of phase.
All continuously varying signals in physio-

logic recordings are described in terms of
their frequency components, as illustrated in
Figure 6-3. Many of them, such as elec-
troencephalographic waveforms, often have
a predominant frequency that at any given
time characterizes the signal at a pair of elec-
trodes. The ability to dissect a waveform into
its component frequencies does not mean
that distinct structures generate each of the
frequencies. For example, a neuron with
synaptic input from multiple sources dis-
plays postsynaptic potentials that summate at
the cell body and produce a complex, vary-
ing intracellular potential. Many frequencies
would be identified by frequency analysis,
but no single generator would be active at
any of the component frequencies. Fre-
quency analysis with automated electronic
systems can define the frequency compo-
nents of any signal as a histogram, as shown
in Figure 6-4. Frequency analysis of either
the first or the second trace in A, B, C, and
D in Figure 6-1 would show only one fre-
quency component, whereas frequency anal-
ysis of the third trace in each panel would
show two frequency components. A fre-
quency analysis of this type can be made on
signals of different time duration.
Figure 6-2. Summation of waveforms with different
shapes but the same frequency (3 Hz) results in more Therefore, a continuously varying signal is
complex waveforms. A sine wave is combined with, A, described by its frequency components,
a square wave, B, a triangular wave, and, C, a ramp wave. their amplitudes, and the intervals of time
Figure 6-3. The frequency components of electric activity recorded in clinical neurophysiology are shown on a log-
arithmic scale. T and t are the upper and lower time constants of each frequency cutoff. EGG, electrocardiogram;
EEG, electroencephalogram; EMG, electromyogram. (From Geddes LA, Baker LE. Principles of Applied Biomed-
ical Instrumentation. John Wiley & Sons, New York, 1968, p 317. By permission of the publisher.)
Figure 6-4. Frequency analysis of, la, normal and, Ib, abnormal electroencephalographic recordings. The frequency
components of la are shown in Ib as a predominantly 10 Hz signal. Tracing 2a is a combination of the 10 Hz ac-
tivity, with the abnormal 1-4 Hz activity shown in 2b. (From Fisch BJ. Spehlmann's EEG Primer, 2nd ed. Elsevier
Science Publishers, Amsterdam, 1991, p 132. By permission of the publisher.)
65
Rgure 6-5. Single spikes occur as events at different rates. A, A spike changes from 5/second to 10/second. B, A
spike occurring at 1/second combines with another, independent spike occurring at 2/second. C, A single spike
occurs in bursts of 8 spikes at 10/second every 2 seconds.
over which they occur. This approach to de- in time, and (3) the pattern of recurrence.
scribing signals is used most commonly with In Figure 6-5 C, groups of eight spikes recur
electroencephalographic potentials.1 A de- at 2-second intervals. This is described as a
scription of neurophysiologic signals always burst pattern of 10/second spikes recurring
includes the location at which the signal is regularly at 2-second intervals.
recorded. For example, the frontal region of A recording from a pair of electrodes of-
the head may have a 30 mV, 20 Hz signal oc- ten includes the activity of multiple genera-
curring for 10 seconds of a 20-second record- tors, each of which may be generating events
ing, whereas the occipital region simultane- in different patterns. A single generator may
ously has a 10 Hz signal throughout the 20 generate regularly recurring events that oc-
second recording epoch. cur in a changing but definable pattern or
events that recur in an unpredictable, or
random, pattern. In Figure 6—55, a second
EVENT RECORDING independent generator becomes active after
6 seconds. Such superimposition gives a
A second approach to the description of the complex pattern of events that can be sepa-
waveforms seen in clinical neurophysiology rated into the recurrence of distinct events
is used primarily in recordings made up of coming from different generators. This dis-
a sequence of well-defined events, as shown tinction is made most accurately by analyz-
in Figure 6-5. Recurrent, brief spikes in a ing the pattern of events.2 Also, recognizing
recording that includes no other activity are the unique appearance of the individual
shown in Figure 6-5A. The description and events can help to distinguish them, but this
classification of events use terminology sim- becomes unreliable when the individual
ilar to that used for continuous recording, waveforms are similar.
but with some differences in meaning. In In summary, characterizing the occur-
Figure 6-5A, the spikes occurring in the first rence of events requires (1) identifying the
5 seconds of the recording are described as patterns of individual events as bursting or
occurring at a frequency of 3/second; those nonbursting, (2) describing those that occur
in the next 4 seconds have a frequency of in bursts according to their rate of firing dur-
6/second. The entire recording of 37 events ing a burst and the recurrence rate of the
in 10 seconds could be described as spikes burst, and (3) describing those that are non-
occurring 3.7 times per second. However, bursting by the rate and pattern of firing.
this is an average of their occurrences in- Each event is characterized further on the
stead of the actual recurrence of 3 and 6 basis of its own variables.3 The event, often
times per second. Therefore, events are bet- called a discharge or spike, has an amplitude
ter described as (1) the number per unit that is measured either from the baseline to
time, (2) the regularity of their occurrence the peak or from peak-to-peak (Fig. 6—6).
Classification of Waveform Characteristics 67
figure 6-6. Each single event, such as the spikes in Figure 6-5, can be described quantitatively by the variables il-
lustrated here. The same event could also be analyzed into its component frequencies, which could range from
fewer than 1 Hz to more than 10,000 Hz.
The discharge has a duration from onset to chrony of their firing. Greater synchrony of
termination. Discharges have configurations firing produces a simpler waveform of larger
that may be monophasic, biphasic, triphasic, size.
or more complex with multiple phases.4 If Individual events can be characterized
there is more than one phase, each phase also according to their frequency spectrum.
can be described according to its amplitude, The analysis of an event—regardless of its
duration, and configuration. Each compo- size, shape, or configuration—can break it
nent of a discharge has a rise time, or a rate down into a summation of the activity of dif-
of rise, from the positive peak to the negative ferent frequencies. Event recording is used
peak. The rise time is a direct function of in recording from single axons and single
the distance of the recording electrodes neurons and in clinical electromyography.
from the generator and can be used to de- In these settings, the measurements are gen-
termine how close a generator is to the erally those of the pattern of firing and the
recording electrodes. Short duration or characteristics of the individual discharges.
rapid rise times occur when the recording The frequency component of such poten-
electrodes are close to the generator. tials is not a useful measure.
A typical discharge is triphasic when it is The division of electric activity into con-
recorded from a nearby generator. If the tinuous waveforms and events is somewhat
waveform is moving, the initial positive por- arbitrary, and they may be found together,
tion of the discharge is recorded from the such as spike discharges in a sinusoidal
potential when the potential is distant from electroencephalographic waveform. Similar
the recording electrode. A negative compo- measurements can be made with either
nent is obtained when the potential is adja- spikes or continuous waveforms. The fre-
cent to the active electrode, and the late pos- quency of a continuous waveform is the in-
itive component when the potential leaves verse of the duration of a cycle. For exam-
the electrode. The complexity of a discharge ple, a 10 Hz signal has a 100-ms interval
is a function of the number of generators between recurrences. The broad sine wave
that contribute to the discharge and the syn- of the 10 Hz signal when repeated through
one cycle could be considered an event and rations, and configurations as well as by their
characterized as such. In contrast, the des- patterns of occurrence and distribution.
ignation of events and their patterns of re-
currence and complexity of appearance oc-
casionally requires characterizing diem in REFERENCES
terms of frequency, amplitude, and rate of
recurrence. 1. Samar VJ, Bopardikar A, Rao R, Swartz K. Wavelet
analysis of neuroelectric waveforms: a conceptual tu-
torial. Brain Lang 66:7-60, 1999.
SUMMARY 2. Kobayashi K, James CJ, Nakahori T, Akiyama T, Got-
man J. Isolation of epileptiform discharges from un-
averaged EEC by independent component analysis.
The waveforms recorded in clinical neuro- Clin Neurophysiol 110:1755-1763, 1999.
physiology are divided into continuous wave- 3. Dumitru D. Physiologic basis of potentials recorded
forms and discrete waveforms, or events. in electromyography. Muscle Nerve 23:1667-1685,
Continuous waveforms are described by 2000.
4. Dumitru D, KingJC, Stegeman DF. Normal needle
dieir frequency components, amplitudes, electromyographic insertional activity morphology: a
and distributions. Discrete waveforms are de- clinical and simulation study. Muscle Nerve 21:910-
scribed by their individual amplitudes, du- 920, 1998.
Chapter 7
ALTERATION OF WAVEFORMS
AND ARTIFACTS
Jasper R. Daube
PHYSIOLOGIC ALTERATION OF ARTIFACTUAL WAVEFORMS

WAVEFORMS Physiologic Artifacts
Single Potential Nonphysiologic Artifacts
Continuous Waves SUMMARY
Signal Display
Abnormalities of the waveforms generated the single event. In contrast, changes in con-
by the central nervous system, nerves, or tinuously varying signals, such as electroen-
muscles can be assessed only in terms of a cephalographic (EEG) waves, are described
change in the waveform of a specific gener- by the characteristics of a series of waves. The
ator. No waveform itself can be defined as distinction between a single event and con-
abnormal without reference to the genera- tinuous waves is not always clear, but usually
tor. Normal waveforms arising from one gen- they can be separated. The alterations of sin-
erator would be abnormal if they arose from gle potentials and waves are considered sep-
a different generator. For example, the spike arately below.
activity normally generated by muscle has
features that are similar to those of an epilep-
tic spike generated by the cerebral cortex. Single Potential
Therefore, alterations in waveform must be
considered in relation to the categories de- Electric activity generated by nerve or mus-
scribed in Chapter 6. In contrast, electric ar- cle tissue often appears as a single discrete
tifacts often have distinct waveforms that do event, a single potential, with no activity or
not arise from any physiologic generator. Ar- only unrelated activity around it. Single
tifacts are best defined as electric activity of potentials may be normal or abnormal.
no clinical significance originating from Changes in individual potentials are de-
nonphysiologic sources. Artifacts are con- scribed by measuring the variables of the po-
sidered separately in the last section of this tential to determine whether they are out-
chapter. side the normal range (Table 7-1). To
describe single potentials, four sets of vari-
ables are measured. The first set describes
PHYSIOLOGIC ALTERATION the size and includes amplitude (peak-to-
OF WAVEFORMS peak or baseline-to-peak), area, and dura-
tion of the potential. The second set de-
Changes in single potentials, such as a sin- scribes the waveform configuration and
gle well-defined electromyographic (EMG) includes the rate of change of the compo-
spike, are described by the characteristics of nents of the potential, the number and tim-
69
Table 7-1. Measurable Variables of Single Potentials

Size Amplitude, area, duration
Configuration Rate of change, direction, number, and timing
of reversal of direction
Recurrence Rate, pattern, timing
Distribution Field, area, location
Relationship to other Time-locked, latency, order of interpotential
waveforms or events interval
Stability Pattern and type of change with recurrence
over time
ing of changes in the direction of the cur- measuring these variables must be defined
rent flow, and the components of the po- because the results can vary with the method
tential. The components include the phases of measurement.
and turns of the potential. The third set de-
scribes the pattern and frequency of occur-
rence of the potential. A spike might occur Continuous Waves
at a regular, low rate (for example, a volun-
tary motor unit potential) or as high fre- Much of the electric activity that is gener-
quency, short bursts (for example, a ated by neural tissue occurs as continuously
myokymic discharge). The fourth set de- varying potentials that may persist over long
scribes the distribution or field of occur- periods. These potentials usually have a si-
rence. For example, an epileptic spike might nusoidal configuration. Recurrent single
occur in the frontal lobe or in the temporal events recorded at a considerable distance
lobe. A waveform may have different vari- from the generator may also appear as con-
ables in different parts of its field. A poten- tinuously varying waves. Continuous waves
tial may be described by its relationship to are characterized by variables similar to—
other events, such as the latency of a re- but different from—the variables that char-
sponse. Disease can alter the variables of ex- acterize single events (Table 7-2).
isting waveforms, eliminate a normal wave- Variables that are used to measure the size
form, or initiate a new waveform. of continuous waves include amplitude (peak-
If a single potential recurs over time, an- to-peak or base-to-peak), root mean square
other set of variables is measured, including (square root of mean amplitude over time),
stability, rate, pattern, and the type of and power (square of the amplitude). In most
change that occurs with time. The alteration situations, the major variable to measure is
of waveforms with disease is defined by the frequency, or the number of cycles of
which variable is outside the normal range. the wave per second. Frequency can be mea-
In disease, each variable should be consid- sured simply as baseline or zero-crossings
ered for measurement. The methods for per second. More complex automated analy-
Table 7-2. Measurable Variables of Continuous Waves

Size Peak-to-peak amplitude, root mean
square, power
Frequency Cycles per second, zero crossing
Appearance Usually sinusoidal, frequency bands
Distribution Field or area, symmetry
Relationship to other waves Phase relation, synchrony
Alteration of Waveforms and Artifacts 71
ses of frequency spectra are the fast Fourier Measurement of the timing, frequency, and
transforms and autoregressive modeling. spatial distribution of the waves can provide
Continuous waves may be simple, with a sin- valuable information about the presence
gle frequency, or they may be complex, with and the stage of disease.
more than one frequency contributing to
the waveform. The addition of multiple fre-
quencies changes the appearance of the Signal Display
wave from a simple sinusoidal pattern to a
more complex, varying one. Continuous The single potentials and continuous waves
waves can be analyzed with regard to their generated by neural and muscle tissues can
frequency components and the power of each be recorded as analog or digital signals.
component. Polarity or direction is seldom Modern equipment uses a digital format that
described because the waves are continuous. allows the signals to be readily stored for sub-
Frequency analysis can provide a precise sequent review and analysis. This capability
measurement of the waveform, but it re- makes it possible to analyze signals without
quires defining the amount of each of the displaying the raw data, showing only the
component frequencies. This is sometimes processed data. Although this can improve
done with frequency bands (Table 7-3). the recording efficiency, it has the risk of
The distribution of continuous waves is an- recording and analyzing unwanted signals,
other important variable to measure. It is such as the artifacts described in the follow-
usually described as broad areas, and coming section. Thus, it is preferable to display
parisons are made between homologous ar- the raw, unprocessed signal for review be-
eas of the body for symmetry. The relation- fore proceeding to analyze the information.
ship of continuous waves to other waves in The human eye and ear are better than au-
the same or other areas is another impor- tomated systems for recognizing artifact. For
tant variable that is measured to identify al- example, the raw signal recorded during
terations produced by disease. Waveforms evoked potential testing should be displayed
may occur in synchrony for defined periods along with the averaged potential during
or they may not be in synchrony but still have data collection.
a definable time relationship, the phase rela- Unprocessed signals are best displayed as
tion. Waves may be in phase or out of phase. a horizontal trace in which the horizontal
Table 7-3. Voltages, Display Times, and Frequency of Common Signals in

Clinical Neurophysiology
Voltage (mV) Time (ms) Frequency (Hz)
Electromyography 50-1000 20-1000 32-16,000
Nerve conduction studies 1-20,000 10-500 1-8000
Electroencephalography 1-2000 5000-200,000 0.1-1000
Brain stem auditory 0.1-2 5-20
evoked potentials
Somatosensory evoked 0.1-20 50-200 20-3000
potentials
Visual evoked potentials 1.0-200 100-200 20-3000
Skin potentials 100-5000 1000-10,000 0.1-100
Electrocardiography 1000-5000 10,000-50,000 0.5-100
Respiratory movements 50-2000 5000-200,000 32-10,000
Electronystagmography 1000-5000 5000-100,000 1-2000
Electroretinography 1000-5000 500-2000 1-500
Vascular reflexes 1000-5000 1000-5000 0.1-100
axis (sweep) is time and the vertical axis is ARTIFACTUAL WAVEFORMS

voltage change. The sweep speed and am-
plification vary widely with the many differ- Artifacts are unwanted signals generated by
ent forms of signals (Table 7-3). Multiple sources other than those of interest. They
signals from different areas are often are not of clinical value. Artifacts can be clas-
recorded simultaneously as vertically sepa- sified as signals from living tissue, physiologic
rated lines. artifact, or as signals from other sources, non-
There are many formats for displaying physiologic artifact.
processed data. The most common one is a
line format, as used for averaged signals. Re-
sults may also be shown as histograms, bar Physiologic Artifacts
graphs, numerical tables, topographic maps,
or frequency plots (for example, com- Physiologic artifacts are unwanted noise that
pressed spectral arrays). Statistical analysis of in other settings are the signals of interest.
the data is used with many of these displays. These include (1) the electrocardiogram—a
The assumptions of any statistical analysis relatively high-amplitude, widely distributed
performed must be understood and appro- potential generated by heart muscle that can
priate for the problem to be solved. interfere with any clinical neurophysiology
Table 7-4. Common Forms of Artifact and Interference

Source Appearance
Movement of Charged Structures

Eye movement Slow positive, lateralized
Eye blink V-shaped positive
Tongue movement Slow positive
Eye flutter Rapid, rhythmic, alternating
Normal Activation
Muscle potentials Rapid, recurrent spikes
Perspiration Very slow oscillation
Electrocardiogram Sharp and slow, regular
Dental fillings touching Short spikes
Transcutaneous stimulator 70-150 Hz spikes
Cardiac pacemaker 1 Hz spike
Paging and radio signals Intermittent, recognizable sound
Recording System
Electrode movement Irregular, rapid spike
Wire movement Irregular, slower waves
Poor electrode contact Mixture of rapid spikes and 60 Hz
Rubbing materials, static Sharp spikes
Display terminal 300 Hz, regular
White thermal noise Random, high-frequency
Electromagnetic, External
Equipment 60 cycle Regular, 60 Hz
Switch artifact Rapid spikes
Diathermy Complex, 120 Hz
Cautery Dense, high-frequency spikes
Alteration of Waveforms and Artifacts 73
Figure 7-1. Rhythmic physiologic artifacts in electroencephalographic (EEC) recordings. EGG, electrocardiogram.
recording; (2) EMG signals that accompany trodes, the electric amplification and display
muscle contractions during EEG and evoked system, electric stimulation, and the external
potential recordings; (3) potentials that oc- electric devices or wiring.4 The most com-
cur with the movement of electrically mon source of such artifacts is movement of
charged structures (for example, tongue the wires that connect the electrodes to the
movement, eye movement, or blink); and (4) equipment or movement of the electrodes
autonomic nervous system potentials, such on the skin. Movement of the electrodes on
as those arising from changes in skin im- the skin causes change in the electrical
pedance with perspiration.1-3 Common arti- charge and capacitance that exist at the in-
facts are listed in Table 7-4. Each of these terface of the electrode and the skin. Alter-
signals and other waveforms that may be ation in static field or electromagnetic in-
recorded to study a particular structure in duction with wire movement can produce
one setting may be an artifact that interferes large artifacts. Artifacts also can arise from
with the recording of a different signal in the opening and closing of switches on
another setting (Fig. 7-1). Although physio- equipment; from poor connections of the
logic artifacts are phenomena that cannot be recording electrodes, with high resistance of
dissociated from normal function, they must the electrodes; and from the use of dissimi-
be circumvented as much as possible. For ex- lar metals. Spurious signals generated within
ample, decreasing the level of muscle activ- the recording apparatus are usually a 60 or
ity can help circumvent EMG artifact. An- 300 Hz signal.
other method is to filter out unwanted Several external power sources generate
frequencies. In some cases, physiologic arti- specific artifacts; examples include the 60-
facts need only to be recognized and men- cycle signal caused by electromagnetic radi-
tally discounted or subtracted electronically, ation from power lines; the modified 60-
for example, eye movement artifact on EEG cycle signal caused by fluorescent lights; the
records. high-frequency, complex discharges from
cautery and diathermic equipment; and the
irregular waveforms from radio sources,
Nonphysiologic Artifacts and magnetic resonance imaging power5
(Fig. 7-2).
Nonphysiologic artifacts are from technical Artifacts sometimes are referred to as in-
sources, for example, the recording elec- terference because they interfere with record-
determines the likelihood of eliminating the

artifact by averaging the signal.
SUMMARY
Any of the variables used to describe the con-
tinuous and discrete waveforms recorded in
clinical neurophysiology can be altered.
Changes in amplitude, frequency, and dis-
tribution of waveforms occur in continuous
waveforms. Frequency change may include
the addition of new, abnormal frequencies,
the loss of normal frequencies, and either
an increase or decrease in amplitude. Dis-
crete events themselves may be abnormal.
The configuration, distribution, size, and
pattern of normally occurring discrete
Figure 7-2. Nonphysiologic artifacts recorded during
events may be changed by disease.
surgical monitoring of muscle activity.
REFERENCES
1. Croft RJ, Barry RJ. Removal of ocular artifact from
the EEC: a review. Neurophysiol Clin 30:5-19, 2000.
ing the activity of interest. By recognizing the 2. Picton TW, van Roon P, Armilio ML, Berg P, Ille N,
nature and source of an artifact, clinical neu- Scherg M. The correction of ocular artifacts: a topo-
rophysiologists can often reduce it or elimi- graphic perspective. Clin Neurophysiol 111:53-65,
2000.
nate it by changing the electrodes or by 3. Jung TP, Makeig S, Humphries C, et al. Removing
changing the location of the equipment or electroencephalographic artifacts by blind source
its relationship to the power source. At separation. Psychophysiology 37:163-178, 2000.
times, averaging can reduce activity if it is 4. Wichmann T. A digital averaging method for re-
moval of stimulus artifacts in neurophysiologic ex-
not time-locked to the stimulus. Differential periments. J Neurosci Methods 98:57-62, 2000.
amplification that is used in all modern 5. Allen PJ, Josephs O, Turner R. A method for re-
recording equipment markedly reduces ex- moving imaging artifact from continuous EEG
ternal artifacts. Appropriate grounding can recorded during functional MRI. Neuroimage 12:
also help.6 230-239, 2000.
6. Tenke CE, Kayser J. A convenient method for de-
Continuously occurring artifacts are some- tecting electrolyte bridges in multichannel elec-
times referred to as noise and compared with troencephalogram and event-related potential
the signal as a signal-to-noise ratio. This ratio recordings. Clin Neurophysiol 112:545-550, 2001.
SECTION 2
Electrophysiologic
Assessment of
Neural Function
Part A
Cortical
Function
In clinical neurophysiology, neural function is assessed by measuring the

electric waveforms generated by neural tissue and the changes in these wave-
forms produced by disease. The characteristics of the waveforms and their
alteration with disease are a function of the neural generators producing
the waveform. A particular modality of recording in clinical neurophysiol-
ogy reflects only the alteration in the area of the nervous system generating
the activity. For example, electroencephalography (EEG) records the wave-
forms arising from cerebral cortex; the waveforms are recorded using elec-
trodes applied to the scalp. These waveforms may change indirectly with dis-
ease elsewhere (for example, the slowing that occurs in cerebral ischemia
because of reduced cardiac output). The EEG recordings described in Chap-
ters 8 and 9 reflect the disease processes that directly involve the cerebral
cortex. Some diseases involving the cortex, such as Alzheimer's disease, do
not cause a change in electric activity because the disease does not affect
the neural generators. The EEG records both the ongoing spontaneous elec-
tric activities of the cerebral cortex and the cortical response to external
stimuli. These patterns of responses can provide important clues to the un-
derlying disease process.
Variations of standard EEG recordings have been developed that provide
unique information for specific situations. Longer recordings are needed to
document infrequent episodes and to define their nature, character, and
spread. Some abnormalities may not be detected with a standard 30- to 60-
minute EEG recording. For abnormal electric activity that occurs only in an
outpatient setting or under specific circumstances, ambulatory recordings
from a few scalp electrodes are made continuously during activities at home
76
76 Cortical Function
or work (Chapter 10) to obtain a full picture of their natures. To help de-
fine the nature and origin of frequent seizures, a patient undergoes pro-
longed EEG recording with multiple electrodes left in place for several days
(Chapter 11). Electroencephalograms can also be recorded in other spe-
cialized situations, such as the intensive care unit or operating room, or with
computerized quantitation described in Chapter 12, which also describes ad-
ditional information that can now be obtained from magneto-EEG. Patients
being considered for epilepsy surgery require highly specialized recordings,
including new correlations with magnetic resonance imaging (Chapter 13).
Cortical function can also be assessed with nonspontaneous potentials, such
as those that occur before a planned movement or in response to external
stimulation (Chapter 14).
Chapter 8
ELECTROENCEPHALOGRAPHY:
GENERAL PRINCIPLES AND ADULT
ELECTROENCEPHALOGRAMS
Donald W. Klass
CLINICAL USEFULNESS OF PATHOLOGIC ACTIVITY

ELECTROENCEPHALOGRAPHY Nonspecific Types
DISPLAY OF Distinctive Epileptiform Patterns
ELECTROENCEPHALOGRAPHIC ELECTROENCEPHALOGRAPHIC
ACTIVITY MANIFESTATIONS OF FOCAL
ACTIVATION PROCEDURES INTRACRANIAL LESIONS
ARTIFACTS Principles
NORMAL ELECTROENCEPHALOGRAPHIC Importance of Correct Identification
ACTIVITY OF ADULTS Electroencephalographic Manifestations
Awake State ELECTROENCEPHALOGRAPHIC
Drowsiness MANIFESTATIONS OF DIFFUSE
Sleep State ENCEPHALOPATHIES
BENIGN VARIANTS EVALUATION FOR SUSPECTED BRAIN
Variants During Wakefulness DEATH
Variants During Drowsiness and Sleep SUMMARY
CLINICAL USEFULNESS OF ness and suspected encephalopathies, EEG

ELECTROENCEPHALOGRAPHY can help determine whether there is an or-
ganic disturbance of function, indicate the
Ekctroencephalography (EEG) is a dynamic degree of disturbance of function, make or
"noninvasive, benign, and relatively inexpen- confirm the diagnosis of specific disorders,
sive technique that assesses brain function."1 indicate whether there is a focal process, and
The most common referrals for EEG are for help determine the prognosis.
the evaluation of patients with suspected sei- With regard to prognosis, favorable find-
zure disorders. However, EEG is also useful ings in EEG include variability, reactivity,
for assessing many other conditions associ- varying wake and sleep patterns, and a pro-
ated with an alteration of cerebral activity. gressive increase in background frequencies.
In epilepsy, EEG can confirm the diagnosis Patterns that indicate a poor prognosis for
and, depending on the type of pattern, help return of useful neurologic function include
indicate the type of seizure disorder the pa- an invariant monorhythmic pattern with lit-
tient has. In disorders of altered conscious- tle reactivity, the burst-suppression pattern,
77
generalized periodic discharges, and gener- DISPLAY OF

alized suppression of activity. ELECTROENCEPHALOGRAPHIC
The Glossary of the International Federa-
tion of Clinical Neurophysiology defines ACTIVITY
EEC as "(1) The science relating to the elec- Preference for the use of particular mon-
trical activity of the brain. (2) The technique tages to display EEG activity varies consider-
of recording electroencephalograms."2 These ably among different electroencephalog-
definitions encompass EEG both for use as raphers. The American EEG Society has rec-
a research tool and as a clinical diagnostic ommended that all EEG laboratories in-
procedure. The purpose of this chapter, clude a minimum number of standard mon-
however, is to survey the clinical applications tages for basic recording to facilitate com-
of the product (electroencephalograms) munication among different laboratories.6
without consideration of the actual record- Included in the minimum recommenda-
ing technique.3""5 Two additional points tions are choices for longitudinal bipolar
should be mentioned. First, in practice it is (Fig. 8-1 A), transverse bipolar (Fig. 8-1J3),
often important to monitor other physio- and referential displays (Fig. 8-1 C). All EE
logic variables simultaneously with the EEG. laboratories are strongly encouraged to com-
Second, detection and interpretation of the ply with these recommendations. Because
EEG data derived from visual analysis involv any one type of montage may be inadequate
matters of judgment and experience, which to solve a particular problem, most elec-
render clinical EEG an art as much as a troencephalographers consider it advisable
science. to use a combination of display systems for
Figure 8-1. Electroencephalogram from a 4V2-year-old girl, showing the appearance of, A, focal right frontocentral
(F4,C4) spikes in a longitudinal (anteroposterior) bipolar montage, B, focal right frontocentral spikes in a trans-
verse bipolar montage, C, focal right frontocentral spikes in a referential montage, and, D, focal right frontocen-
tral spikes in a montage combining bipolar and referential recording, (continued)
Figure 8-1. (Continued)
79
each patient and to emphasize whatever ar- ACTIVATION PROCEDURES

ray seems to be most useful in each case (Fig.
8-1D). One should be able to select the most Activation procedures are used commonly
advantageous arrays for particular types of in standard EEG practice to determine
activity to localize that activity adequately or whether the response to provocation is nor-
to display it more prominentiy if it seems mal or abnormal.4'7"9 Commonly, these pro-
equivocal. Therefore, proper facility and cedures are used to elicit abnormal activity,
flexibility in recording require, first, a clear particularly paroxysmal activity, when the
understanding of the rationale for the dif- basic resting record has been uninformative.
ferent display systems and, second, consid- The simple act of eye opening is used uni-
erable practice in visualization and record- versally in EEG practice to test for reactivity
ing of the actual EEGs. of normal or abnormal activity, but the act
The advent of digital EEG recording has of eye closure can sometimes provoke parox-
proved to be a boon for the electroen- ysmal discharges of the spike-wave variety.
cephalographer to conveniently select any of Eye closure can also provoke a nonspecific
several different montages for the most ap- abnormality in the form of bisynchronous
propriate display of a particular segment of rhythmic posterior slow waves (posterior
the EEG (see Chapter 4). phi).
EEG: General Principles and Adult Electroencephalograms 81
Figure 8-2. Electroencephalogram from a 6-year-old boy awake and at rest (segment on left) and normal response
to hyperventilation (HV) (segment on right).
A second commonly used activation pro- Two other principal types of phenomena
cedure is hyperventilation. In patients who may be activated by photic stimulation. The
can cooperate adequately, hyperventilation photoparoxysmal response (previously termed
is performed for 3-5 minutes and recording photoconvulsive) usually takes the form of a
is continued for at least 2-3 minutes after- generalized and bilaterally synchronous
ward. In many normal persons, but particu- burst of spike-and-slow-wave complexes that
larly in children, hyperventilation induces may or may not outlast the photic stimulus.
bursts of generalized slow waves (Fig. 8-2). These paroxysmal abnormalities have a high
This so-called hyperventilation buildup usu- correlation with clinical seizures, particularly
ally subsides promptly after cessation of the in patients who have a clinical history of light
overbreathing. With hypoglycemia, the re- sensitivity in their natural environment.9'10
sponse may be exaggerated in magnitude The discharges may be associated with clin-
and persistence. The most common abnor- ical manifestations of myoclonus or absence
mality elicited by hyperventilation is the attacks. The technologist must terminate the
diffuse and bilaterally synchronous 3-Hz photic stimulation when a photoparoxysmal
spike-and-slow-wave discharge, which may response occurs; otherwise, prolonged stim-
sometimes be accompanied by the clinical ulation may precipitate a generalized
manifestations of typical absence seizures. tonic-clonic seizure. Lesser degrees of pho-
Less frequently, hyperventilation may acti- toparoxysmal responses may be confined to
vate focal spikes or increase the prominence the posterior head regions; they are less
of focal slow waves. likely to be associated with clinical manifes-
Another activation procedure used in tations. Rarely, the photoparoxysmal re-
most EEG laboratories is stroboscopic inter- sponse may be initiated by a focal discharge
mittent photic stimulation. Stimulation in the occipital or frontal region.11
should include frequencies from 1 to 30 Hz Another type of response that needs to be
and must include frequencies between 10 differentiated from the photoparoxysmal re-
and 20 Hz, conducted first with the patient's sponse may resemble cerebral activity but is
eyes open and then with them closed. Most actually caused by myogenic potentials that
normal subjects show some evidence of flash- follow the flash frequency, particularly in the
related responses over the posterior head re- frontal regions. This latter type of response
gions (photoentrainment or photic driving), is known as the photomyogenic or photomyologic
but absence of grossly detectable responses response (previously termed photomyoclonid) .9
or transient asymmetries of the driving re- Typically, it shows recruitment during the
sponse are not considered abnormal. period of intermittent photic stimulation
and never outlasts the flashes. Although this ARTIFACTS

type of response is less common than the
photoparoxysmal variety, it does not signify In EEG, artifact refers to any electric signal
an abnormality unless the response is grossly that is not generated directly by the brain.
exaggerated, widespread, and is associated Artifacts are frequent contaminants of EEG
with generalized myoclonus. recordings because of the high sensitivity of
Rarely, photic stimulation can activate fo- the instrumentation required to amplify the
cal spikes in the occipital regions. Exagger- EEG, and artifacts can mimic almost every
ated responses to low flash frequencies re- kind of EEG pattern.13
corded from the occipital regions also may The technologist and electroencephalog-
be indicative of an encephalopathy, such rapher need to be constantly alert to the pos-
as ceroid lipofuscinosis in children and sibility of artifact.5 Excessive artifact can ren-
Creutzfeldt-Jakob disease in adults. In some der the EEG uninterpretable; even worse,
patients with sensitivity to light, activation of confusion of artifact with brain wave activity
paroxysmal discharges may also occur when can lead to serious misinterpretation.
scanning geometric patterns, watching tele-
vision, or playing video games.12
Recording during sleep is an important
means of activating focal or generalized NORMAL
paroxysmal abnormalities. Significant gen- ELECTROENCEPHALOGRAPHIC
eralized spike-wave discharges and focal tem- ACTIVITY OF ADULTS
poral sharp waves are usually activated best
during deeper levels of nonrapid eye move- Awake State
ment (NREM), slow wave, sleep. Daytime
sleep in the laboratory can usually be The activity seen in the EEGs of awake adults
achieved after partial overnight sleep depri- consists of frequencies in the alpha and beta
vation and by foregoing stimulating bever- ranges, with the alpha rhythm constituting
ages or medications before the recording. the predominant background activity.
Total overnight sleep deprivation, however,
may be an activator in itself. If the patient ALPHA RHYTHM
cannot fall asleep spontaneously, sedation
may be required. The most commonly used Alpha activity refers to any activity in the
sedative in.EEG laboratories is chloral hy- range between 8 and 13 Hz, whereas the al-
drate, because it does not induce beta (fast) pha rhythm is a specific rhythm consisting of
activity that may complicate interpretation alpha activity occurring over the posterior
of the tracing, whereas barbiturates and head regions when the person is awake and
most benzodiazepines typically induce beta relaxed and has the eyes closed (Fig. 8-3);
activity and are generally avoided for that it is attenuated by attention. The alpha fre-
reason. quency is very stable in a person, rarely vary-
Other activation procedures may need to ing by more than 0.5 Hz14>15 over many years
be used in special circumstances, as de- of adult life, and the frequency is identical
scribed in the comprehensive review of sen- on the two sides of the head.
sory activation.9 Every effort should be made The usual alpha amplitude in an adult is
to attempt to reproduce the particular stim- 15-50 joV. The maximal amplitude occurs
ulus that may have triggered the patient's over the occipital region, with variable
symptoms. In addition to the light sensitiv- spread to the parietal, temporal, and, at
ity mentioned above, activation in the visual times, central leads. Often, the alpha activ-
domain may result from reading, from stim- ity has a higher voltage and wider distribu-
uli in the auditory system (simple sounds or tion over the right hemisphere.
complex music), from somatosensory stim- The alpha rhythm should attenuate bilat-
ulation in a localized region of the body, erally and promptly with eye opening, alert-
or from mental activity such as arithmetic ing stimuli, or mental concentration. Some
calculation, to mention a few examples. alpha rhythm may return when the eyes re-
Activation may involve highly individual main open for more than a few seconds. Fail-
stimulus—response characteristics.9 ure of the alpha rhythm to attenuate on one
EEC: General Principles and Adult Electroencephalograms 83
Figure 8-3. Normal electroencephalogram from a 47-year-old man, showing symmetric alpha rhythm predominantly
in the occipital regions (Oi and O2).
side with either eye opening or mental alert- component of the background activity. Ado-
ing indicates an abnormality on the side that lescents retain somewhat more theta activity
fails to attenuate.16 than do middle-aged adults. In older pa-
tients, theta components can occur as single
BETA ACTIVITY transients or as part of a mixed alpha-theta
burst over the temporal regions.1*
Beta activity has a frequency greater than 13
Hz.2 The average voltage is between 10 and DELTA ACTIVITY
20 /iV.17 The three main types of beta activ-
ity, based on distribution, are the following: Delta activity (< 4 Hz) is the predominant ac-
(7) the precentral type occurs predomi- tivity of infants but is not a normal compo-
nantly over the frontal and central regions, nent of the EEG in young or middle-aged
increases with drowsiness, and may attenu-
ate with bodily movement; (2) posterior
dominant beta activity can be seen in chil-
dren up to 1-2 years old; it also is enhanced
by drowsiness; and (3) generalized beta ac-
tivity maximal over the frontocentral region
is induced or enhanced by certain drugs,
such as benzodiazepines and barbiturates,
and may attain an amplitude greater than
25 /u,V14'17 (Fig. 8-4). Focal accentuation of
beta activity can result from a lesion or de-
fect in the skull.18
THETA ACTIVITY
Theta activity (4-7 Hz) is only a minimal com-
ponent of the EEC of a normal awake adult, Figure 8-4. Electroencephalogram containing diffuse
but it is more prevalent during drowsiness. beta activity in a 14-year-old girl receiving diazepam
In young children, theta activity is a common therapy.
Figure 8-5. Normal electroencephalogram showing bilateral mu rhythm in the central regions; the rhythm persists
when the eyes are opened and attenuates with movement of the thumbs. This is in contrast to the alpha rhythm
(Oi and Og), which is attenuated by eye opening.
adults. Delta waves intermixed with the al- 8-6). The amplitude is usually between 20
pha rhythm over the occipital regions are a and 50 juV, and the duration is from ap-
common finding in normal teenagers and proximately 100-250 ms. Lambda waves are
are called the posterior slow waves of youth. bilateral and synchronous but may be asym-
metrical. Surface-negative apiculate waves
arising from the occipital regions while the
MU RHYTHM
subject scans a geometric pattern can occur
Mu rhythm consists of arch-shaped waveforms in normal children and young adults and are
with a frequency of 7-11 Hz. It typically oc- probably related to lambda waves.14'21
curs independently on the two sides of the
head over the central regions.17 Mu activity
ELECTROENCEPHALOGRAPHY IN
is functionally related to the sensorimotor
OLDER ADULTS
cortex and is attenuated by active or passive
movement of the extremities or by the Previously, it was thought that a shift to
thought of movement (Fig. 8-5). It can oc- slower background frequencies occurred in
cur in an asymmetrical fashion or predomi- the older population. However, several re-
nate over one hemisphere. Mu rhythm can cent studies have shown that the alpha fre-
be quite striking if there is an overlying skull
defect (breach activity); this needs to be dis-
tinguished from pathologic spike activity.18'19
LAMBDA WAVES
The lambda wave has a configuration resem-
bling the Greek letter A and occurs over the
occipital regions when the subject actively
scans a picture.14 Lambda waves appear to
represent an evoked cerebral response to vi-
sual stimuli produced by movements of im-
ages across the retina with saccadic eye
movements.20 The waveforms are monopha- Figure 8-6. Electroencephalogram showing normal
sic or diphasic, with the most prominent lambda waves maximal in Oj and Og when the patient's
component usually surface-positive (Fig. eyes are open and looking around the room.
quency remains greater than 8 Hz in normal

elderly subjects.15-22 There is a tendency for
the alpha rhythm to be of slightly lower volt-
age in older subjects and perhaps to show
less reactivity.15'23
Benign temporal slow transients consist of
sporadic delta waves that occur singly or in
pairs over the temporal region.24 Temporal
transients can be seen in normal subjects,
usually after the age of 60 years. They have
a left-sided preponderance and appear to be
related to a normal aging process. Drowsi-
ness facilitates the appearance of these tem-
poral transients.
Figure 8-7. Electroencephalogram from a 20-year-old

Drowsiness woman during sleep, showing normal 14-Hz sleep
spindles.
In adults, drowsiness is typically associated
with slowing of the background frequency,
followed by disappearance of alpha activity
and enhancement of theta activity. At times, mately 12 Hz and maximal amplitude is lo-
bursts of generalized moderate-to-high- cated more anteriorly. More continuous
amplitude rhythmic 5-7 Hz theta activity can spindle activity may be seen in some patients
be present.25 In older subjects, there may be who are receiving drug therapy, particularly
an enhancement of theta and delta waves benzodiazepines.
over the temporal regions. Admixed sharply Changes during drowsiness and sleep oc-
contoured waveforms, called wicket spikes or cur in normal elderly subjects.24 During
wicket waves, may also be present over the drowsiness, some subjects exhibit rhythmic
temporal regions.26 This activity can occur trains of high-amplitude delta activity that
in an asymmetrical fashion and may be max- need to be carefully distinguished from the
imal over the left temporal region. pathologic intermittent rhythmic delta ac-
The beta activity over the frontocentral re- tivity that they may closely resemble. During
gions often increases in prominence during stage 2 of NREM sleep, the V waves and K
drowsiness. This usually has a frequency of complexes are lower in amplitude and less
16-20 Hz, but occasional bursts of faster fre- sharp in appearance than in young adults,
quencies may occur. and sleep spindles are less prominent. Delta
Mu activity may also be seen during activity is lower in voltage than in young
drowsiness and may persist after the alpha adults during stages 3 and 4 of NREM sleep.
rhythm disappears.
VERTEX SHARP TRANSIENTS
Sleep State Vertex sharp transients (Vwaves) are sharp-con-
toured transients that occur maximally over
Sleep activity consists of slow waves, spindles, the central vertex region during sleep (Fig.
V waves, K complexes, and positive occipital 8—8). In children and young adults, V waves
sharp transients of sleep (POSTS). may have a sharp or spiky appearance and
In adults, the sleep spindles of stage 2 attain high voltages. They are typically sym-
NREM (slow wave) sleep usually have a fre- metrical in the central leads but may show
quency of 14 Hz and occur in a symmetrical transient asymmetries at the time of sleep
and synchronous fashion over the two hemi- onset. F waves, or frontally dominant V waves,
spheres in the central regions (Fig. 8-7). In are often broader than the centrally domi-
a slightly deeper level of stage 2 sleep, the nant V waves and may extend asymmetrically
spindle frequency decreases to approxi- into the lateral frontal regions.
Figure 8-8. Electroencephalogram from a 34-year-old woman during sleep, showing normal V waves maximal in C3
and C4 and PS and P$.
The K complex is a diphasic or polyphasic the eyes are open, and it also shows inter-
wave that is maximal at the vertex and is usu- mittent groups of irregular eye movement
ally longer than 500 ms. It is frequently as- artifacts. In addition, rhythmic groups of
sociated with spindle activity.2 The K com- saw-toothed waves may occur intermittently
plex represents a nonspecific response to over the frontal and central leads and may
afferent stimulation and is generally linked precede the rapid eye movements.
to the arousal mechanism.
POSITIVE OCCIPITAL SHARP

TRANSIENTS OF SLEEP
Positive occipital sharp transients of sleep are
sharp-contoured surface-positive transients
that occur singly or in clusters over the oc-
cipital regions (Fig. 8-9). They are usually
bilaterally synchronous but may be some-
what asymmetrical. They are predominantly
seen during light-to-moderate levels of sleep
and should not be mistaken for abnormal
sharp waves.
RAPID EYE MOVEMENT SLEEP

During rapid eye movement (REM) sleep, Rgure 8-9. Electroencephalogram during sleep show-
the EEG shows a low-voltage pattern that has ing prominent normal positive occipital sharp tran-
some similarities to an awake pattern when sients of sleep (POSTS) maximal in Oi and Og.
Figure 8-10. Onset of subclinical rhythmic electrographic discharge of adults in a 57-year-old woman.
BENIGN VARIANTS discharge but is never accompanied by any

clinical symptoms and has no significance
Variants During Wakefulncss for the diagnosis of epileptic seizures (Fig.
8-10). The characteristics of SREDA are
ALPHA VARIANTS listed in Table 8-1.
The alpha-variant patterns consist of activity
over the posterior head regions. This activ- BREACH RHYTHM
ity has a harmonic relationship to the alpha Various normal rhythms are enhanced in
rhythm and shows a reactivity and a distri- amplitude when recorded over a skull de-
bution similar to those of the alpha rhythm. fect. The term breach rhythm has been used
The slow alpha variant appears as dicrotic to refer to a focal increase in the amplitude
or notched waveforms that result from a sub- of sharp-contoured EEG activity over or near
harmonic component of the alpha rhythm, the area of a skull defect.18'19 Particularly
usually in the range of 4—5 Hz. The fast al- when the mu and beta rhythms are involved,
pha variant contains a frequency twice that
of the resting alpha activity.
PHOTIC RESPONSES Table 8-1. Characteristics of Subclinical

Rhythmic Electrographic Discharge
Complex waveforms may be induced by of Adults
photic stimulation when harmonics or sub-
harmonic components are admixed with the Feature Characteristic
fundamental frequency of the driving re-
sponse. Occasionally, the resultant mixture Onset Segmented or abrupt
of frequencies can produce waveforms that Repetition rate Theta
simulate epileptiform spikes or spike-wave Duration Average 1 minute
complexes. Distribution Maximal parietal-
posterior temporal
SUBCLINICAL RHYTHMIC Laterali ty Symmetrical or
ELECTROGRAPHIC DISCHARGE asymmetrical
OF ADULTS State Awake at rest or during
hyperventilation
Subclinical rhythmic electrographic discharge of Background Often visible during the
adults (SREDA) is an uncommon phenome- activity discharge
non that occurs mainly in the elderly.27 The Delta aftermath None
pattern consists of a mixture of dieta and Clinical None
delta frequencies, but most often predomi- accompaniment
nating in the theta frequency range.28'29 It Patient age Mainly older adults
closely resembles an epileptogenic seizure
pendently over the two temporal regions,

with a shifting emphasis from side to side
(Fig. 8-12). This pattern differs from a true
seizure discharge in that it does not evolve
into other frequencies or waveforms. It is
present predominantly in young or middle-
aged adults.
14&6-HZ POSITIVE BURSTS

The 148c6-Hz positive bursts (previously
known as 14 and 6 per second positive spikes}
Figure 8-11. Electroencephalogram from a 21-year-old are displayed best on long interelectrode dis-
man, showing breach rhythm (C4) over the area of a tance referential montages and are most
skull defect. prominent over the posterior temporal re-
gion during light sleep. As the name implies,
the activity can resemble epileptiform spikes the bursts occur at a rate of 14 Hz (Fig.
(Fig. 8-11). 8-13A) or between 6 and 7 Hz (Fig. 8-135)
and are from 0.5 to 1 second in duration.
Variants During Drowsiness

and Sleep
Several patterns that have distinctive char-
acteristics but little or no clinical signifi-
cance occur principally during drowsiness or
light sleep.30 Some of these have an ap-
pearance suggestive of epileptiform abnor-
mality, but they have no importance for the
diagnosis of seizures or cerebral lesions.31
RHYTHMIC TEMPORAL THETA

BURSTS OF DROWSINESS
The rhythmic temporal theta bursts of drowsiness
(previously known as the psychomotor-variant
pattern) frequently assume a flat-topped or
notched appearance because of the har-
monics of the fundamental theta frequency.
The bursts may occur bilaterally or inde-
Figure 8-13. Electroencephalogram from a 12-year-old

Rgure 8-12. Electroencephalogram from a 43-year-old boy during sleep. A, A 14-Hz positive burst (maximal in
woman, showing rhythmic temporal theta activity dur- TS and Te) and, B, 6-Hz positive bursts (maximal in T
ing drowsiness. andT6).
They usually occur independently over the Table 8-2. Characteristics of Benign
two hemispheres and vary from side to side Sporadic Sleep Spikes
in occurrence.30 They are most frequendy
seen in subjects between 12 and 20 years old. Feature Characteristic
Amplitude Low
BENIGN SPORADIC SLEEP SPIKES
Duration Short
Benign sporadic sleep spikes (BSSS),30 also Morphology Sharp, diphasic
known as small sharp spikes (SSS) or benign (steep descent)
epileptiform transients of sleep (BETS), occur Associated slow wave None or minimal
mainly in adults during drowsiness and light Background activity No disruption
levels of sleep (Fig. 8-14). They are usually Distribution Widespread
low-voltage, short-duration diphasic spikes Lateral! ty
with a steep descending limb. They occur Single Maximal unilateral
typically as single spikes, rarely as doublets,
Multiple Bilateral
and never in repetitive trains. The BSSS may
have a single low-voltage aftercoming slow- Occurrence Mainly adults
wave component, but they do not distort die Event Sporadic
background and they are not associated with State Drowsiness or
slow-wave activity, as temporal sharp waves light sleep
are. They are best seen with long interelec- Patient age Adult
trode distance derivations, including the Clinical None
temporal and ear leads. Provided a long accompaniment
enough recording is obtained, they almost
always have a bilateral representation, oc-
curring either independently or synchro-
nously over the two hemispheres. Their repetition rate, or the phantom spike-and-
characteristics are summarized in Table 8-2. wave, because of its usual low-amplitude
They need to be carefully distinguished from spike component. It occurs mainly during
more important types of spikes because the drowsiness and disappears during deeper
BSSS have no significance for the diagnosis levels of sleep (Fig. 8-15). It has no associ-
of epileptic seizures.32 ated clinical manifestations and has no use-
ful correlation with clinical seizures or other
6-HZ SPIKE-AND-WAVE symptoms.30 Its typical characteristics are
listed in Table 8-3.
The 6-Hz spike-and-wave pattern has also been
called the fast spike-and-wave, because of its
WICKET SPIKES
Wicket spikes, a pattern described by Reiher
and Lebel,26 consist of single spike-like wave-
forms and appear as a monophasic fragment
of a mu-like rhydim (Fig. 8-16). Wicket
spikes, or wicket waves, have a frequency of
6-11 Hz and an amplitude ranging from 60
to 200 /xV and are seen mainly in adults.
They occur during drowsiness and light
sleep and become apparent when the alpha
and other awake patterns drop out. Wicket
spikes are present over the temporal regions,
occurring bilaterally or independently over
die two temporal regions, and they may oc-
cur more frequently on one side, usually the
left. When wicket spikes occur as a single
Figure 8-14. A typical benign sporadic sleep spike in a waveform, they may be mistaken for a tem-
66-year-old patient. poral spike discharge; however, wicket spikes
Figure 8-16. Wicket spikes in the left temporal region

of a 44-year-old man.
Figure 8-15. A 6-Hz spike-wave burst during drowsiness by the slow component. They may resem-
in a 14-year-old girl with headaches but no seizures. ble—but should not be mistaken for—a
spike-and-wave discharge.
are not accompanied by aftercoming slow
waves or a distortion or slowing of the back- PATHOLOGIC ACTIVITY
ground that occurs with a true epileptogenic
temporal spike. Nonspecific Types
MITTEN PATTERNS Abnormal amounts of diffuse slow-wave ac-
tivity are an indication of the severity of dis-
Mitten patterns, waveforms originally de- turbed cerebral function but are considered
scribed by Gibbs and Gibbs,33 are seen dur- nonspecific because they give little clue as to
ing sleep and consist of fast-wave and slow- the cause.34 For example, similar changes
wave components that resemble a mitten, may be caused by intrinsic cerebral degen-
with the thumb of the mitten formed by the erative disease, cerebral dysfunction from
last wave of a spindle and the hand portion hypoxia, inflammatory disease involving the
central nervous system, the effects of exter-
nal toxins, or various systemic electrolyte or
Table 8-3. Characteristics of 6-Hz metabolic disorders. In general, with the
Spike-and-Waves more severe cerebral disturbance, the aver-
age frequency of abnormal activity is slower,
Feature Characteristic spontaneous variability is lessened, abnor-
mal patterns are less reactive to external
Frequency 6± 1 Hz stimuli, and the likelihood that normal phys-
Repetition rate Regular iologic activity is disrupted or lost is greater.
Burst duration Brief, < 1-2 seconds Voltage increases to a point with increasing
Spike duration Brief severity, but with the most severe impair-
Amplitude Low ment, voltage is decreased and, with cerebral
Distribution Diffuse, maximal anterior death (see later under Evaluation for Sus-
or posterior pected Brain Death), is eventually lost en-
Laterality Bisynchronous, tirely. Even though nonspecific slow-wave ab-
symmetrical normalities may provide no clue about the
Clinical state Drowsiness cause, they can be helpful in indicating or-
ganic cerebral dysfunction in subjects in
Clinical None
accompaniment whom a distinction may need to be drawn
Patient age Young adult between psychiatric and organic causes of
mental symptoms.
Distinctive Epileptiform Patterns the awake recording. This increases to

80%-90% when the patient goes to sleep.
Epileptiform discharges are paroxysmal Therefore, the EEG evaluation for someone
waveforms with distinctive morphology that with suspected complex partial seizures
stand out from the ongoing background ac- should include a sleep recording. Detailed
tivity.35'36 They are recorded predominantly accounts of the EEG manifestations of com-
from patients with epileptic seizures. Interic- plex partial seizures have been provided by
tal activity refers to activity recorded between Klass.37
seizures. Ictal discharges are those occurring
during a clinical seizure. The main types of
epileptiform discharges are spikes, sharp Frontal Spike Discharge
waves, and spike-and-slow-wave complexes. Frontal spike discharges may occur at any age.
Spike discharges are potentials with steep as- This is another highly epileptogenic spike,
cending and descending limbs, a pointed and 70%-80% of patients with a frontal
peak, and a duration less than 70 ms.2 Sharp spike discharge have seizures.33'38 Often,
waves are broader waveforms that have a du- some underlying pathologic condition can
ration of 70-200 ms.2 A spike-and-slow-wave be demonstrated.
complex consists of a spike followed by a slow
wave and is often referred to as a spike-wave
pattern. Epileptiform discharges may be fo- Central-temporal (Rolandic or Sylvian)
cal or generalized in distribution. Spikes
The central-temporal (C-T) spike is one of the
INTERICTAL DISCHARGES—FOCAL more common types of focal spike dis-
Anterior Temporal Spikes charges in children39'40 (see Fig. 9—11). With
standard electrode placement, the discharge
The anterior temporal spike, or sharp wave, is is seen primarily in the central and midtem-
the most frequent type of focal discharge in poral leads, but it is often maximal in leads
adolescents and adults (Fig. 8-17). Tempo- halfway between these two positions (that is,
ral spikes and sharp-wave discharges have a C5 and C6). The site of origin appears to be
very high correlation with the presence of the lower rolandic area just above the sylvian
clinical partial seizures (90%-95%).33 Sleep fissure.40 The spike discharges have a char-
markedly potentiates the presence of these acteristic appearance: they often are high-
discharges, especially moderately deep levels amplitude, diphasic, blunt spikes, with an
of NREM sleep.37 Approximately 30%-50% aftercoming slow wave.41 They may be fre-
of patients with temporal lobe epilepsy may quent and occur in brief clusters or trains,
show spikes or sharp-wave discharges during unilaterally, bilaterally, or shift from side to
side. During sleep, C-T spikes are often ac-
tivated or enhanced.38'40 The EEG back-
ground activity is usually normal. Fre-
quently, the surface distribution of the spike
corresponds to a tangential dipole source
with peak negativity in the tempororolandic
region and positivity in the frontal leads.
The C-T spike discharge is seen primarily
between the ages of 4 and 12 years. About
60%-80% of children with the C-T spike
discharge have seizures,40 and the condition
is termed benign rolandic epilepsy of childhood
(BREC). Seizures typical of this disorder,
which have been termed sylvian seizures, con-
sist of twitching of one side of the face or
Figure 8-17. Focal epileptiform sharp waves arising hand (or both); motor-speech arrest; exces-
from the left temporal region in a 35-year-old man. sive salivation or drooling because of diffi-
culty swallowing; tingling of the side of the the discharge lasts longer than 3 to 4 sec-
mouth, tongue, or cheek; progression to a onds, there is usually some type of clinical
generalized seizure; and more frequent oc- accompaniment such as staring, sursumver-
currence during the night.40 The seizures gence, clonic movements of the face, motor
are easily controlled with anticonvulsant arrest, and unresponsiveness—all indicative
medications. The EEG pattern and seizures of absence seizures. The discharges and clin-
usually disappear spontaneously in the sec- ical seizures are enhanced by hyperventila-
ond decade of life, and no gross focal lesion tion and hypoglycemia. The interictal EEG
can be demonstrated. record is usually normal. During sleep, the
discharges occur in a more fragmented and
less sustained fashion, often consisting of sin-
Occipital Spikes
gle spike-wave complexes or multispike-and-
Occipital spikes can be seen in young children, wave discharges of varying complexity that
usually those younger than 3-5 years, and of- increase in number during deepening stages
ten resolve as the child gets older.38 The of NREM sleep.35 The pattern is seen most
spike discharges may be unilateral or bilat- often in children between the ages of 3 and
eral. Occipital spikes in this age group are 15 years.
not highly epileptogenic; only 30%-50%
of children with occipital spikes have Slow Spike-and-Wave
seizures.38'42 Instead, the presence of these
spikes tends to correlate more with early- The slow spike-and-wave pattern has also been
onset visual deprivation, and approximately referred to as sharp-and-slow-wave complexes,
40% of the children with occipital spikes may because the spike component has a duration
have some associated visual problem. that conforms more to the definition of a
In older children, occipital spikes can be sharp wave.2'44 The complexes occur rhyth-
seen with the entity of benign epilepsies of mically with a frequency of 1.5-2.5 Hz (see
childhood with occipital paroxysms in which Fig. 9-10). The trains of slow spike-and-wave
occipital spikes are present when the eyes discharges often are not associated with any
are closed and are attenuated with eye open- apparent clinical manifestation. However, if
ing42-43 (see Fig 9-12). appropriate testing is performed, there may
Focal occipital spikes with associated sei- be some type of subtle impairment of psy-
zures can also occur in association with an chomotor performance. The interictal back-
underlying pathologic condition such as vas- ground between the spike-and-wave bursts is
cular lesions, tumors, or cerebral dysgenesis. often abnormal. Seizures in patients with the
slow spike-and-wave pattern consist of vari-
INTERICTAL DISCHARGES—DIFFUSE ous types and combinations of generalized
seizures, such as tonic seizures, atypical ab-
The main types of generalized epileptiform sences, tonic-clonic seizures, akinetic sei-
discharges are the 3-Hz spike-and-wave, zures, or myoclonic seizures.44'45 The slow
slow spike-and-wave, and atypical spike-and- spike-and-wave pattern is most often seen in
wave discharges and paroxysmal rhythmic children between 2 and 6 years of age, but
fast activity. it may persist through adolescence.35'44'46
' The slow spike-and-wave pattern usually
3-Hz Spike-and-Wave occurs in patients with some type of under-
lying organic pathologic condition and who
The 3-Hz spike-and-wave pattern consists of also have clinical signs of cerebral damage.
stereotyped, generalized, bilaterally syn- Many patients with this pattern of discharge
chronous, symmetrical spike-and-slow-wave have a severe convulsive disorder, with sev
complexes that have an average repetition eral types of seizures, poor response to anti-
rate of 3 Hz (see Fig. 9-9). Usually, there i convulsants, and signs of mental and motor
maximal voltage in the superior frontal dysfunction. This constellation of clinical
(Fs>4) regions. These complexes are dis- signs and EEG pattern has been referred to
played best in ear reference montages. If as the Lennox-Gastaut syndrome.45'46
Atypical Spike-and-Wave myoclonic seizures that are associated with a

generalized high-amplitude spike- or sharp-
The term atypical spike-and-wave refers to a wave discharge in the EEG. The symptom
generalized spike-and-wave discharge that complex of hypsarrhythmia and infantile
lacks the regular repetition rate and stereo- spasms with arrest of psychomotor develop-
typed appearance of the 3-Hz spike-and-wave ment, West's syndrome, is not a specific disease
or slow spike-and-wave pattern. Atypical entity but reflects a response of the imma-
spike-and-wave complexes occur with vary- ture brain to a severe cerebral insult or dys-
ing frequencies between 2 and 5 Hz; there function, usually occurring before 1 year of
may be admixed multiple spike components. age. The cause is unknown in approximately
When multiple spikes are a prominent fea- one-third of the patients, but in the others,
ture, the discharges are known as multispike- this symptom complex may be the result of
and-slow-wave complexes. The paroxysms are diverse prenatal, perinatal, or postnatal dif-
usually brief, ranging from 1 to 3 seconds in ficulties, such as encephalitis, congenital de-
duration. Sleep often activates these dis- fects, or various biochemical or metabolic
charges.8 The atypical spike-and-wave dis- derangements.39 One of the most frequent
charge may be seen in children or adults and identifiable causes is tuberous sclerosis.
occurs in patients with different types of gen-
eralized seizures, especially myoclonic and
tonic-clomc seizures. ICTAL DISCHARGES
Ictal EEG discharges consist of repetitive ac-
Paroxysmal Rhythmic Fast Activity tivity that has an abrupt onset and termina
Paroxysmal rhythmic fast activity, or paroxysmal tion. Morphology of the activity can take any
tachyrhythmia, consists of rhythmic fast activ- shape or form—spikes, sharp waves, spike-
ity or repetitive spike discharges with a fre- and-wave discharges, or rhythmic activity in
quency of from 12 to 20 Hz, predominately the beta, alpha, theta, and delta frequency
visible in the parasagittal regions.47 The ranges,35'50 or electrodecremental episodes,
paroxysms are usually synchronous but may as mentioned above. The ictal discharge pat-
be asymmetrical. They often culminate in tern may consist of stereotyped waveforms
one or more slow waves. This pattern often throughout the discharge, as with the gen-
occurs with tonic or tonic-clonic seizures and eralized 3-Hz spike-and-wave pattern during
is seen most frequently on recordings made absence seizures, or the seizure pattern may
during sleep. evolve with changing frequencies, ampli-
tudes, and waveforms and spreading distri-
Hypsarrhythmia bution, as with most seizures of focal origin.
In contrast to the 3-Hz spike-and-wave pat-
The term hypsarrhythmia refers to a high-volt- tern of absence seizures, a generalized tonic-
age continuous or nearly continuous pattern clonic seizure evolves in several phases.51
consisting of a chaotic admixture of multi- The first phase consists of rhythmic fast ac-
focal spikes or sharp waves and arrhythmic tivity (repetitive spikes) that is associated
slow waves (see Fig. 9-8). The hypsarrhyth- with the tonic phase of the seizure. The next
mic pattern is seen mainly in patients be- phase consists of spike discharges and slow
tween the ages of 4 months and 4 years and waves associated with the clonic phase of the
is often associated with the clinical seizures seizure. These spike and slow-wave parox-
known as infantile spasms.48 The EEG ac- ysms gradually become separated by in-
companiment of infantile spasms usually creasingly longer pauses and stop suddenly.
consists of an initial high-amplitude diffuse Postictally, there is a temporary period of
spike, slow wave, or both, followed by an diffuse flattening followed by diffuse slow
abrupt generalized decrease of voltage with waves, which gradually subside over a vari-
low-amplitude fast activity, called the elec- able period.
trodecremental pattern (a descriptive term for Ictal discharges of focal origin may begin
the EEG seizure pattern coined by Bickford in a localized fashion (Fig. 8-18), but fre-
and Klass49). The patient may also have brief quently the onset is marked by an abrupt dif-
Figure 8-18. Focal seizure discharge arising from the left central region (Cs) and accompanied by the clinical man-
ifestations of tonic version of the head to the right.
fuse or asymmetrical decrease in voltage. be generalized (Fig. 8-19), convulsive or

Surface attenuation at onset is particularly nonconvulsive, or partial (Fig. 8-20).
common with discharges beginning in the
temporal lobe.37 Postictal slowing is typically
focal unless the seizure has secondarily
evolved into a generalized seizure. Ictal sei- ELECTROENCEPHALOGRAPHIC
zure discharges may last from a few seconds MANIFESTATIONS OF FOCAL
to 2 or 3 minutes. When a typical discharge INTRACRANIAL LESIONS
occurs without clinical seizure manifesta-
tions, it is known as a subdinical electrographic The practice of EEG has changed consid-
seizure discharge. Status epilepticus consists of erably since the advent of neuroimaging
continuous or frequently repeated seizure techniques, and EEG is no longer the pri-
discharges and only brief interruptions with- mary noninvasive screening device for focal
out clinical recovery. Status epilepticus may intracranial lesions it once was. Neverthe-
Figure 8-19. Continuous generalized spike-and-wave discharges in a 49-year-old woman in nonconvulsive status
epilepticus.
Figure 8-20. Electroencephalogram showing continuous focal spike discharges over the right temporal region in
an obtunded 50-year-old man who is in complex partial status.
less, electroencephalographers and refer- not refer to the physical properties of the le-
ring physicians need to be aware of the elec- sion but to the concentration of its effects
trophysiologic changes that can occur with on the adjacent brain. Because lesions such
focal lesions because EEG offers a different as neoplasms and cysts are electrically silent,
measure of cerebral function than do com- the EEG abnormalities they produce are
puted tomographic and most magnetic res- caused by the disturbance of nearby neu-
onance imaging studies. The two types of ronal structures. Therefore, local pressure
procedures are complementary rather than from a concentrated mass generally pro-
mutually exclusive. The EEG is one measure duces focal EEG abnormalities more readily
of what the brain is capable of doing rather than even a highly malignant neoplasm,
than a depiction of what damage has been which infiltrates among neurons more dif-
done. The EEG may provide early evidence fusely. A third spatial factor to consider is
for a previously unsuspected focal intracra- the location of the lesion. In general, the
nial lesion. Furthermore, when clinically in- more superficial a lesion is in relation to the
dicated, frequent sequential EEG exami- convexity of the cerebral hemisphere, the
nations are facilitated by virtue of being more likely it is to produce focal EEG ab-
available at the bedside and of having neg- normalities. Independent of pathologic type
ligible risk and modest cost. and size, lesions near the base of the skull
or in the posterior fossa may provoke no
change or nonfocal abnormality in the scalp
Principles EEG.
Of the temporal factors, one needs to con-
Several different but interrelated factors sider when the tracing is made with respect
influence the EEG expression of focal in- to the stage of evolution of the lesion. A sin-
tracranial lesions. This concept may be sum- gle EEG recorded very early in the develop-
marized by the term spatial-temporal biody- ment of a neoplasm may be entirely normal
namics. This term indicates that the EEG and of no help diagnostically, but sequential
expression of a lesion is a dynamic physio- recordings could be extremely helpful by
logic process involving the dimension of contributing positive evidence for diagnosis
time as well as of space. if a focal abnormality emerges later. By way
Of the spatial factors, the first to consider of contrast, sequential recordings made af-
is the extent of the lesion. A lesion, irre- ter infarction of a similar cerebral region
spective of type, may not produce any EEG would be expected to show maximal abnor-
abnormality until it attains sufficient size to mality in the acute stage and diminishing
be detectable in the usual scalp recording. abnormality during the course of resolution.
A second spatial factor is density. This does The type and magnitude of the EEG abnor-
malities also depend on the age of the pa- abnormality at all, depending on the spatial-
tient at the time the disease occurs. Some temporal biodynamics of the individual case.
types of EEG abnormality are expressed only Furthermore, the EEG manifestations of dif-
during a particular stage of maturation. The ferent pathologic types of focal lesions may
location and magnitude of some abnormal- be indistinguishable from one another.
ities also can differ between children and
adults. Intermittent rhythmic delta activity LOCAL OR REGIONAL EFFECTS
produced by a deep-seated lesion, for ex- ON NORMAL ACTIVITY
ample, is apt to be dominant over the oc-
cipital regions in children and over the Focal lesions located near the convexity of
frontal regions in adults. Cerebral insults of- the cerebral hemispheres frequently pro-
ten produce more prominent EEG abnor- duce localized attenuation of normal back-
malities in children than might be expected ground activity. Depending on the location
to result from a process of similar severity and size of the lesion, this may involve alpha
in adults. Finally, the balance between de- activity, beta activity, sleep activity, or all
structive and reparative forces greatly affects background rhythms on one side of the
the EEG findings. For example, a rapidly ex- head. Less frequently, focal lesions may re-
panding neoplasm typically produces a pre- sult in an increase in amplitude of the back-
dominance of slow polymorphic and highly ground activity, a circumstance that may oc-
persistent delta activity, in contrast to the cur with indolent lesions or those involving
slowly growing neoplasm whose only EEG subcortical nuclei. Before attributing an in-
manifestation may be a spike focus indistin- crease in amplitude to the lesion itself, one
guishable from the effects of scar formation. needs to exclude the effects of unequal in-
terelectrode distances and underlying skull
defects. One should remember also that
Importance of Correct some activity, such as the mu rhythm, may
Identification normally be asymmetrical.
Asymmetry of frequency is less common
An accurate description is essential for cor- than asymmetry of amplitude. A consistent
rect identification of any type of EEG activ- asymmetry of frequency of alpha activity of
ity, whether normal, abnormal, or artifac- more than 1 Hz is abnormal and usually in-
tual. Mistakes in identification can usually be dicates a lateralized disturbance on the side
avoided if the basic physical and physiologic of the slower frequency.
elements are considered as completely as Lesions may cause defective reactivity of
possible in formulating the description: (1) normal rhythms. A focal lesion may produce
frequency, (2) voltage, (3) phase relation- ipsilateral defective attenuation of alpha ac-
ships, (4) quantity, (5) morphology, (6) tivity with eye opening or with mental activ-
topography, (7) reactivity, (8) manner of oc- ity when the eyes are closed.16 A focal lesion
currence, and (9) polarity. These factors are may cause depression of voltage or slowing
important not only for identifying abnor- of frequency of sleep spindles on the side of
malities but also for determining their de- the lesion.5^
gree of severity and significance. 2 The significance of an asymmetry depends
When attempting to determine the iden- on its degree with respect to the limits of
tity of any EEG phenomenon, another im- normal variation, its consistency during the
portant consideration is its context. The recording, the concurrent asymmetry of
matter of context includes other ongoing other activity in the same region, and the
EEG activity and also the clinical state of the presence of associated abnormal activity.
patient at the time the recording is made.
LOCAL EFFECTS—ABNORMAL
Electroencephalographic ACTIVITY
Manifestations Polymorphic focal delta activity is the hall-
mark of focal lesions affecting the cerebral
Focal intracranial lesions may produce al- hemispheres (Fig. 8-21). Although this type
most any type of EEG abnormality or no EEG of activity does not arise directly from the le-
97
Figure 8-21. Electroencephalogram from a 20-year-old man with a left frontal astrocytoma, showing persistent poly-
morphic delta activity predominantly in the left frontal region (Fpi).
sion itself, it generally indicates proximity to metastatic brain tumors may cause bilateral
the site of the maximal or active destructive independent delta foci, that circumstance is
process. Factors in judging the degree of uncommon.
severity include the following: morphology Because of the important consequences of
(irregular), frequency (slowest), persistence focal delta activity, artifacts that may mimic
(high), amplitude (increased to a maximum this type of abnormality need to be carefully
and then decreased), reactivity (deficient), excluded. Localized slow-wave artifacts can
and associated destruction of underlying be generated by faulty electrodes, eye or
background activity. To some extent, these head movement, vascular pulsation, and
factors also represent acuteness. Focal poly- electrodermal activity. Focal slow-wave ab-
morphic delta activity is detected most normalities can also be masked by higher
clearly during alert wakefulness. Often, it is amplitude normal background rhythms, by
delineated less clearly during sleep, because diffuse abnormal rhythms, and by artifacts
of the presence of normal, more widespread (Fig. 8-22). These masking effects can some-
delta activity. Persistence during sleep, how- times be diminished by maintaining the
ever, has been related to superficial location patient's alertness and by making ample
and epileptogenicity of the lesion. In a pa- recording with the patient's eyes open. Op-
tient with seizures and focal polymorphic timal techniques of recording focal slow
delta activity during wakefulness, sleep may waves include use of a long time constant, a
be a useful adjunct to determine whether fo- high-frequency filter if faster activity is in-
cal distinctive epileptiform discharges arise trusive, a slow speed, and montages that
from the same area or independently from display homologous derivations in adjacent
a different area. Focal polymorphic delta ac- channels. Various perceptual techniques have
tivity does not help to distinguish the patho- been used by individual electroencepha-
logic type of the lesion responsible for gen- lographers to aid in the detection of low-
erating the abnormal activity. Although amplitude abnormal slow waves. According
Figure 8-22. Focal polymorphic delta activity in the right frontal region (Fp2) that is clearly evident when the pa-
tient's eyes are open but masked by eye movement when the eyes are closed.
to the prevalent view, focal polymorphic ing instrument and appropriate adjustments
delta activity is related most closely to lesions of sensitivity. Artifacts that can simulate
that involve the superficial white matter of spikes or sharp waves include those result-
the cerebral hemisphere and that interfere ing from faulty electrodes, electrostatic dis-
with cortical connections. charges, and chance superimposition of
Focal spikes or sharp waves may accurately electrocardiographic artifact. Not all spo-
represent the focal origin of seizures, but radic spikes have serious pathologic impli-
they usually do not arise from the lesion cations, however, as already discussed.
itself. These distinctive epileptiform dis-
charges may be slightly removed from the
REGIONAL EFFECTS—ABNORMAL
site of maximal destruction and may repre-
ACTIVITY
sent the pathophysiologic balance between
destructive and reparative processes. Al- Some types of focal abnormal slow wave are
though these types of paroxysmal discharges likely to be somewhat more distant from the
usually signify a relatively chronic lesion, maximal site of the lesion, including focal
they may also be the first manifestation in delta activity that is more regular, more re-
the course of slowly progressive lesions. Well- active, and less continuous than the persis-
defined spikes superimposed on normal tent polymorphic delta activity already men-
background activity are more usually associ- tioned. Generally, this is also true of focal
ated with chronic lesions, whereas ill-formed theta activity. Temporal intermittent rhythmic
sharp waves superimposed on a disrupted delta activity (TIRDA) is a focal abnormality
background more often represent acute le- that is highly significant for the diagnosis of
sions. These paroxysmal abnormalities may partial seizures and is thought to represent
appear only during drowsiness or sleep. Less the surface manifestation of a spike focus
frequently, they are activated by hyperventi- situated deeply within the temporal lobe54
lation and only rarely by photic stimulation. (Fig. 8-23). Different types of slow-wave
Optimal detection requires the use of max- abnormalities may coexist, and such con-
imal high-frequency response of the record- tiguous rhythms need to be sorted out to de-
REMOTE EFFECTS—ABNORMAL
ACTIVITY
Intermittent rhythmic delta activity is sometimes
referred to as a projected rhythm, because
when it results from a focal intracranial le-
sion it represents a distant effect of the le-
sion rather than a local superficial source.
When intermittent rhythmic delta activity
occurs in isolation, the lesion is generally sit-
uated deep within the cranium. However,
this type of abnormality does not necessarily
Figure 8-23. Temporal intermittent rhythmic delta ac-
tivity (TIRDA) over right temporal region in a 75-year-
signify an intracranial lesion, because it is in-
old man with complex partial seizures. distinguishable from abnormalities resulting
from systemic metabolic or electrolyte dis-
orders. This type of abnormality may also be
termine their relative importance for local- a transient postictal effect. Characteristically,
ization. Artifacts such as those generated by it is bilateral and diffusely distributed, bi-
abnormal electrocardiographic paroxysms synchronous, monorhythmic and monomor-
or by rhythmic eye movements may simulate phic, and reactive to eye opening and alert-
intermittent rhythmic delta activity or con- ing. It often is increased by hyperventilation,
tinuous focal theta activity. Some acute le- and it usually disappears during sleep. Some-
sions are associated with periodic lateralized times the waveforms may be saw-toothed
epileptiform discharges (PLEDs), an acronym rather than sinusoidal. Frontally predomi-
coined by Chatrian and colleagues,55 which nant intermittent rhythmic delta activity is
are helpful for demonstrating a unilateral known as FIRDA. Artifacts that need to be
hemispheric lesion (Fig. 8-24). In cases of distinguished from this type of abnormality
ischemic cortical infarction, PLEDs and include the electro-oculographic and glos-
delta activity are typically present before sokinetic potentials. When diffuse intermit-
computed tomography shows evidence of an tent rhythmic delta activity is caused by a
anatomical lesion. focal hemispheric lesion, it is often asym-
Figure 8-24. Electroencephalogram from a 71-year-old man with recent head trauma, showing left-sided periodic
lateralized epileptiform discharges (PLEDs).
Figure 8-25. Electroencephalograms from a 67-year-old man, showing focal persistent polymorphic delta activity in
the right temporal region (F8 and T4) (segment on left) and the addition 2l/z months later (segment on right) of
intermittent rhythmic delta activity caused by a rapidly progressive brain tumor.
metrical (more frequently of higher ampli- to obtain a postoperative baseline EEG. Gen-
tude on the side of the lesion). erally, this is best accomplished after about
a week, when the scalp wound has healed
SEQUENTIAL ALTERATIONS
and the acute effects of the surgical proce-
dure have subsided. Sequential recordings
In a patient with seizures, development of may be helpful for detecting early or late
focal delta activity despite seizure control complications and may also be useful for as-
may indicate the presence of a neoplasm. A sessing the effects of radiation therapy or
progressive focal lesion should be suspected chemotherapy for intracranial lesions.
if focal polymorphic delta activity appears
after previous EEGs have been normal or if
previous EEGs have contained a nonfocal,
nonspecific abnormality. A progressive le- ELECTROENCEPHALOGRAPHIC
sion in sequential recordings may show in- MANIFESTATIONS OF DIFFUSE
creased magnitude of focal polymorphic ENCEPHALOPATHIES
delta activity (factors of severity mentioned
above), enlarged topographic distribution of The EEG is helpful in the evaluation of dif-
the abnormality, ipsilateral attenuation of fuse disorders of cerebral function and
background activity, or the addition of dif- serves as a measurement of the severity of
fuse intermittent rhythmic delta activity (Fig. the disturbance.34'56'57 Diffuse encephalop-
8-25). In contrast to the usual situation with athies can be caused by various conditions,
neoplasms, after cortical infarction, the EEG including metabolic, toxic, inflammatory,
typically shows maximal focal slow wave ab- post-traumatic, hypoxic, and degenerative
normality acutely, and sequential recordings disorders.
show decreasing focal abnormality. When The type of diffuse disorder and whether
PLEDs occur with acute lesions (vascular or it involves white or gray matter influence the
neoplastic), these discharges usually subside EEG pattern.58 Processes that predomi-
within a few days despite continued presence nantly affect superficial white matter usually
of the anatomical lesion. cause polymorphic delta slowing in the EEG,
For sequential follow-up of patients after whereas processes that involve cortical and
intracranial surgical procedures, it is useful subcortical gray matter are more likely to
It is characterized by dementia, motor dys-

function, myoclonus, and, when the disease
is fully developed, a characteristic periodic
EEG pattern consisting of generalized, bi-
synchronous, and periodic sharp waves re-
curring at intervals of 0.5-1 second, with a
duration of 200-400 ms, the prototype of
the periodic short-interval diffuse discharges
(PSIDDs)59 (Fig. 8-27). Myoclonic jerks are
often associated with the periodic sharp
waves; however, there is not always a constant
relationship between the two. Although oc-
Figure 8-26. Severe diffuse slow-wave (delta) abnor- casionally other degenerative or toxic disor-
mality in a 9-year-old boy with encephalitis. ders24'60 may be associated with a quasiperi-
odic sharp-wave pattern, the presence of
periodic sharp waves, progressive dementia,
cause intermittent bilaterally synchronous and myoclonus is strongly suggestive of
paroxysmal slow-wave activity. Epileptiform Creutzfeldt-Jakob disease.
abnormalities are seen more commonly in Subacute sclerosing panencephalitis (SSPE) is
gray matter disease than in white matter dis- a degenerative disorder that occurs in chil-
ease. Other factors that influence the degree dren and adolescents and is believed to be
and type of EEC abnormalities include the caused by the measles virus. This degenera-
age and clinical state of the patient, the stage tive disorder is characterized by abnormal
of the disease process, and other complicat- movements, intellectual deterioration, and
ing factors such as infectious processes, a diagnostic, periodic EEG pattern. This
metabolic derangements, or drug effects. consists of repetitive stereotyped high-volt-
The most common type of EEG finding in age sharp-and-slow-wave complexes recur-
diffuse disorders, or encephalopathies, con- ring every 4-15 seconds, the prototype of
sists of slowing of varying degrees34'56 (Fig. the periodic long-interval diffuse discharges
8—26). This may involve background activity, (PLIDDs)59 (Fig. 8-28). This pattern usually
the theta frequency range, or generalized is present during the intermediate stages of
polymorphic delta range. Intermittent bursts the disease. In a single recording from a sin-
of bilaterally synchronous rhythmic slow gle patient, the morphology of the com-
waves can occur in a generalized fashion or plexes is stereotyped; however, the shape of
have a maximal expression over the anterior the complexes can vary in different patients
or posterior head regions. Usually the de- and change from time to time in the same
gree of slowing parallels the degree of dis- patient at different stages of the disease. The
turbance of function or alteration in level of complexes are usually generalized and bi-
consciousness (or both). These findings can synchronous, but at times they may be asym-
be caused by various diffuse disorders and, metrical or more lateralized. Stereotyped
therefore, are considered nonspecific motor jerks or spasms are often associated
changes in that they are not diagnostic of with the periodic complexes.
any single condition, as mentioned above. Periodic lateralized epileptiform discharges
At times, however, the EEG may show a are often seen with herpes simplex encephali-
more specific pattern, such as periodic pat- tis. They consist of periodic sharp waves that
terns or the various distinctive coma pat- occur in a focal or lateralized manner over
terns. The periodic patterns include those one hemisphere, particularly involving the
associated with Creutzfeldt-Jakob disease, temporal regions. They may also be seen
subacute sclerosing panencephalitis, and with other viral encephalitides such as La
herpes simplex encephalitis. Crosse encephalitis.
Creutzfeldt-Jakob disease is a diffuse, suba- The coma patterns include the triphasic
cute, and progressive disorder of the cen- wave, alpha- and beta-frequency, spindle,
tral nervous system that occurs predomi- and burst-suppression.
nantly in middle-aged patients and is In hepatic coma, the EEG often shows a
thought to be a transmissible prion disease. triphasic wave pattern consisting of medium-
Figure 8-27. Diffuse periodic sharp waves in a 71-year-old man with Creutzfeldtjakob disease.
Figure 8-28. Diffuse periodic complexes in a 12-year-old girl with subacute sclerosing panencephalitis.
102
103
Rgure 8-29. Triphasic waves in a 77-year-old woman in hepatic coma.
to high-voltage broad triphasic waves, a pat- A spindle coma pattern resembles a sleep
tern described by Bickford and Butt,61 that EEG and consists predominantly of spindle
occur rhythmically or in serial trains at a rate activity with some V waves, but it shows no
of 1-2 Hz in a bilaterally synchronous and reactivity.64 This type of pattern can result
symmetrical fashion over the two hemi- from various causes, including head trauma,
spheres and have a fronto-occipital or oc- hypoxic insults, or brain stem lesions. De-
cipitofrontal time lag.61 The triphasic waves pending on the type of underlying cause and
usually have a frontal predominance and severity of damage to the central nervous sys
consist of a short-duration, low-voltage sur- tem, the pattern indicates that the potential
face-negative component followed by a for improvement exists. In many types of
prominent positive sharp-contoured wave coma, spontaneous variability of EEG activ-
and then a longer duration surface-negative ity, including the sleep-like pattern, indi-
slow wave61'62 (Fig. 8-29). Although tripha- cates a better prognosis than a prolonged in-
sic waves are often associated with liver dys- variant pattern.
function, atypical triphasic waves can be seen The burst-suppression pattern consists of pe-
in other conditions, including metabolic de- riodic or episodic bursts of activity, usually
rangements, electrolyte disturbances, toxic irregular mixtures of sharp waves or spikes,
states, and degenerative processes, or after a alternating with intervals of attenuation
hypoxic episode. The triphasic wave pattern (Fig. 8-30). This pattern is often seen after
needs to be carefully distinguished from a severe insult to the brain, such as a hy-
epileptiform activity. poxic or anoxic insult, in which case the pat-
The alpha-frequency coma pattern consists of tern usually indicates a poor prognosis.
diffusely distributed invariant alpha activity However, the burst-suppression pattern can
that shows little or no reactivity or variabil- also be seen with potentially reversible con-
ity. This type of pattern has been seen after ditions, such as anesthesia, drug intoxica-
cardiac arrest or hypoxic insult to the brain tion, and hypothermia.
and with significant brain stem lesions.63 In summary, in patients with diffuse dis-
When the alpha-frequency coma pattern is orders, the EEG is useful in documenting a
seen in the context of a hypoxic insult, it usu- disturbance of cerebral function, in deter-
ally indicates a poor prognosis. mining the degree of the disturbance, in
The beta-frequency coma pattern consists of monitoring changes and trends in the
generalized beta activity superimposed on un- course of the disease process, and in help-
derlying delta slowing. This pattern is usually ing to establish the diagnosis in certain con-
associated with drug toxicity or anesthesia. ditions in which a characteristic EEG pattern
Figure 8-30. Diffuse burst-suppression pattern after cardiac arrest.
is present. Also, the EEC sometimes helps to tivity of cerebral origin (Fig. 8-31). Electro-
detect the presence of an additional, more cerebral inactivity (ECI) is defined as "no EEG
focal cerebral process. activity over 2 ;uV."65 There are important
minimal technical criteria for recording in
patients with suspected cerebral death.
EVALUATION FOR SUSPECTED These criteria include the following:65
BRAIN DEATH 1. A minimum of 8 scalp electrodes
should be used.
The EEG can provide confirmatory evidence 2. Interelectrode impedances should be
of brain death, which is manifested by an ab- less than 10,000 Q, but more than
sence of spontaneous or induced electric ac- 100 ft.
Figure 8-31. Electrocerebral inactivity.

3. The integrity of the entire recording REFERENCES

system must be verified. 1. American Clinical Neurophysiology Society indica-
4. Interelectrode distances should be at tions for obtaining an electroencephalogram. J Clin
least 10 cm. Neurophysiol 15:76-77, 1998.
5. The sensitivity should be at least 2 2. Chatrian GE, Bergamini L, Dondey M, Klass DW,
/xV/mm for at least 30 minutes of Lennox-Buchthal M, Petersen I. A glossary of terms
most commonly used by clinical electroencephalo-
recording. graphers. Electroencephalogr Clin Neurophysiol
6. Appropriate filter settings should be 37:538-548, 1974.
used. 3. McGee FE Jr, White RJ. EEG instrumentation. In
7. Additional monitoring techniques Henry CE (ed). Current Clinical Neurophysiology:
should be used when necessary. Update on EEG and Evoked Potentials. Else-
vier/North-Holland Biomedical Press, Amsterdam,
8. There should be no EEG reactivity to 1980, pp 1-49.
afferent stimulation. 4. Reilly EL. EEG recording and operation of the ap-
9. The recording should be made by a paratus. In Niedermeyer E, Lopes da Silva F (eds).
qualified technologist. Electroencephalography: Basic Principles, Clinical
Applications, and Related Fields, 4th ed. Williams
10. A repeat EEG should be performed if & Wilkins, Baltimore, 1999, pp 122-142.
there is doubt about the presence of 5. Tyner FS, Knott JR, Mayer WB (eds). Fundamen-
electrocerebral silence. tals of EEG Technology. Vol. 1: Basic Concepts and
Because temporary and reversible ECI Methods. Raven Press, New York, 1983.
can be caused by drug overdose and hy- 6. American Electroencephalographic Society. Guide-
lines in EEG and evoked potentials, 1986. J Clin
pothermia, these conditions should be ex- Neurophysiol 3 (Suppl 1):1-152, 1986.
cluded before reaching a conclusion of 7. Bickford RG. Activation procedures and special
brain death.66 In young infants, because of electrodes. In Klass DW, Daly DD (eds). Current
uncertainties about the significance of ECI, Practice of Clinical Electroencephalography. Raven
Press, New York, 1979, pp 269-305.
one should exercise caution in the inter- 8. Fisch BJ (ed). Fisch and Spehlmann's EEG Primer:
pretation of this finding.66 The guidelines Basic Principles of Digital and Analog EEG, 3rd ed.
of the Task Force for the Determination of Amsterdam, Elsevier, 1999.
Brain Death in Children67 recommend that 9. Klass DW, Fischer-Williams M. Sensory stimulation,
for infants between 7 days and 2 months old, sleep and sleep deprivation. In Remond A (ed).
Handbook of Electroencephalography and Clinical
two EEGs demonstrating ECI be performed Neurophysiology. Vol. 3: Techniques and Methods
at least 48 hours apart and that for infants of Data Acquisition of EEG and EMG. Part D: Ac-
between 2 months and 1 year old, two tivation and Provocation Methods in Clinical Neu-
records showing ECI be done at least 24 rophysiology. Elsevier Scientific Publishing Com-
pany, Amsterdam, 1976, pp 3D-5-3D-73.
hours apart. 10. Kasteleijn-Nolst Trenite DG, Guerrini R, Binnie
CD, Genton P. Visual sensitivity and epilepsy: a pro-
posed terminology and classification for clinical
and EEG phenomenology. Epilepsia 42:692-701,
SUMMARY 2001.
11. Sunku AJ, Gomez MR, Klass DW. Epileptic seizures,
EEG abnormalities, and neuronal heterotopia in
This chapter provides an overview of the di- the Dubowitz syndrome. Am J Electroneurodiagn
verse patterns of activity seen in clinical elec- Technol 38:156-163, 1998.
troencephalography. For accurate appraisal 12. Hormes JT, Mellinger JF, Klass DW. Testing for
electroencephalographic activation with video
of the EEG, the many individual differences games in patients with light sensitivity. Am J EEG
in normal activity need to be considered Technol 35:37-45, 1995.
first, with appropriate attention to the limits 13. Klass DW. The continuing challenge of artifacts in
of normal variability at different ages and at the EEG. Am J EEG Technol 35:239-269, 1995.
different states of wakefulness and sleep. De- 14. Kellaway P. An orderly approach to visual analysis:
characteristics of the normal EEG of adults and chil-
tection and correct identification of abnor- dren. In Daly DD, Pedley TA (eds). Current Prac-
mal patterns of activity require considerable tice of Clinical Electroencephalography, 2nd ed.
practice and experience. Finally, the useful- Raven Press, New York, 1990, pp 139-199.
ness of the EEG depends on an enlightened 15. Markand ON. Alpha rhythms. J Clin Neurophysiol
7:163-189, 1990.
interpretation of the findings with regard to 16. Westmoreland BF, Klass DW. Defective alpha reac-
the specific clinical problem that the EEG is tivity with mental concentration. J Clin NeUrophys
being used to solve. iol 15:424-428, 1998.
17. KozelkaJW, Pedley TA. Beta and mu rhythms. J Clin tion of epileptic seizures and nonepileptic episodic
Neurophysiol 7:191-207, 1990. disorders. In Klass DW, Daly DD (eds). Current
18. Radhakrishnan K, Silbert PL, Klass DW. Breach ac- Practice of Clinical Electroencephalography. Raven
tivity related to an osteolytic skull metastasis. Am J Press, New York, 1979, pp 221-268.
EEC Technol 34:1-5, 1994. 36. Pedley TA. EEG patterns that mimic epileptiform
19. Radhakrishnan K, Chandy D, Menon G, Sarma S. discharges but have no association with seizures. In
Clinical and electroencephalographic correlates of Henry CE (ed). Current Clinical Neurophysiology:
breach activity. Am J Electroneurodiagn Technol Update on EEG and Evoked Potentials. Elsevier/
39:138-147, 1999. North-Holland Biomedical Press, Amsterdam,
20. Chatrian GE. The lambda waves. In Remond A 1980, pp 307-336.
(ed). Handbook of Electroencephalography and 37. Klass DW. Electroencephalographic manifestations
Clinical Neurophysiology. Vol. 6: The Normal EEG of complex partial seizures. Adv Neurol 11:113-140,
Throughout Life. Part A: The EEG of the Waking 1975.
Adult. Elsevier Scientific Publishing, Amsterdam, 38. Kellaway P. The incidence, significance and natural
1976, pp 6A-123-6A-149. history of spike foci in children. In Henry CE (ed).
21. Sunku AJ, DonatJF, Johnson JA, Klass DW. Occip- Course in Clinical Electroencephalography: Cur-
ital responses to visual pattern stimulation (ab- rent Clinical Neurophysiology; Update on EEG and
stract). Electroencephalogr Clin Neurophysiol 103: Evoked Potentials. Elsevier/North-Holland Bio-
28P, 1997. medical Press, Amsterdam, 1980, pp 151-175.
22. Katz RI, Harner RN. Electroencephalography in ag- 39. Gomez MR, Klass DW. Epilepsies of infancy and
ing. In Albert ML (ed). Clinical Neurology of Ag- childhood. Ann Neurol 13:113-124, 1983.
ing. Oxford University Press, New York, 1984, pp 40. Lombroso CT. Sylvian seizures and midtemporal
114-138. spike foci in children. Arch Neurol 17:52-59, 1967.
23. Bennett DR. Electroencephalographic and evoked 41. Kellaway P. The electroencephalographic features
potential changes with aging. Semin Neurol 1:47- of benign centrotemporal (rolandic) epilepsy of
51, 1981. childhood. Epilepsia 41:1053-1056, 2000.
24. Klass DW, Brenner RP. Electroencephalography of 42. Niedermeyer E, Lopes da Silva F (eds). Electroen-
the elderly. J Clin Neurophysiol 12:116-131, 1995. cephalography: Basic Principles, Clinical Applica-
25. Santamaria J, Chiappa KH (eds). The EEG of tions, and Related Fields, 4th ed. Williams &
Drowsiness. DEMOS Publications, New York, 1987. Wilkins, Baltimore, 1999.
26. Reiher J, Lebel M. Wicket spikes: clinical correlates 43. Kivity S, Ephraim T, Weitz R, Tamir A. Childhood
of a previously undescribed EEG pattern. Can J epilepsy with occipital paroxysms: clinical variants
Neurol Sci 4:39-47, 1977. in 134 patients. Epilepsia 41:1522-1533, 2000,
27. Westmoreland BE, Klass DW. A distinctive rhythmic 44. Blume WT, David RB, Gomez MR. Generalized
EEG discharge of adults. Electroencephalogr Clin sharp and slow wave complexes. Associated clinical
Neurophysiol 51:186-191, 1981. features and long-term follow-up. Brain 96:289-
28. Miller CR, Westmoreland BE, Klass DW. Subclini- 306, 1973.
cal rhythmic EEG discharge of adults (SREDA): fur- 45. Gastaut H, Roger J, Soulayrol R, et al. Childhood
ther observations. Am J EEG Technol 25:217-224, epileptic encephalopathy with diffuse slow spike-
1985. waves (otherwise known as "petit mal variant") or
29. Westmoreland BF, Klass DW. Unusual variants of Lennox syndrome. Epilepsia 7:139-179, 1966.
subclinical rhythmic electrographic discharge of 46. Markand ON. Slow spike-wave activity in EEG and
adults (SREDA). Electroencephalogr Clin Neuro- associated clinical features: often called 'Lennox'
physiol 102:1-4, 1997. or "Lennox-Gastaut" syndrome. Neurology 27:746—
30. Klass DW, Westmoreland BF. Nonepileptogenic 757, 1977.
epileptiform electroencephalographic activity. Ann 47. Westmoreland BF, Klass DW. Unusual EEG pat-
Neurol 18:627-635, 1985. terns. J Clin Neurophysiol 7:209-228, 1990.
31. Maulsby RL. EEG patterns of uncertain diagnostic 48. Jeavons PM, Bower BD. Infantile spasms. In Vinken
significance. In Klass DW, Daly DD (eds). Current PJ, Bruyn GW (eds). Handbook of Clinical Neu-
Practice of Clinical Electroencephalography. Raven rology. Vol. 15: The Epilepsies. North-Holland Pub-
Press, New York, 1979, pp 411-419. lishing Company, Amsterdam, 1974, pp 219-234.
32. Reiher J, Lebel M, Klass DW. Small sharp spikes 49. Bickford RG, Klass D. Scalp and depth electro-
(SSS): reassessment of electroencephalographic graphic studies of electro-decremental seizures (ab-
characteristics and clinical significance (abstract). stract). Electroencephalogr Clin Neurophysiol 12:
Electroencephalogr Clin Neurophysiol 43:775, 263, 1960.
1977. 50. Daly DD. Epilepsy and syncope. In Daly DD, Ped-
33. Gibbs FA, Gibbs EL (eds). Medical Electroen- ley TA (eds). Current Practice of Clinical Elec-
cephalography. Addison-Wesley, Reading, Massa- troencephalography, 2nd ed. Raven Press, New
chusetts, 1967. York, 1990, pp 269-334.
34. Sharbrough FW. Nonspecific abnormal EEG pat- 51. Drury I, Henry TR. Ictal patterns in generalized
terns. In Niedermeyer E, Lopes da Silva F (eds). epilepsy. J Clin Neurophysiol 10:268-280, 1993.
Electroencephalography: Basic Principles, Clinical 52. Klass DW. Identifying the abnormal EEG. In Halli-
Applications, and Related Fields, 4th ed. Williams day AM, Buder SR, Paul R (eds). A Textbook of
& Wilkins, Baltimore, 1999, pp 215-234. Clinical Neurophysiology. John Wiley & Sons,
35. Daly DD. Use of the EEG for diagnosis and evalua- Chichester, 1987, pp 189-199.
107
53. Reeves AL, Klass DW. Frequency asymmetry of sleep 61. Bickford RG, Butt HR. Hepatic coma: the elec-
spindles associated with focal pathology. Electro- troencephalographic pattern. J Clin Invest 34:790-
encephalogr Clin Neurophysiol 106:84-86, 1998. 799, 1955.
54. Normand MM, Wszolek ZK, Klass DW. Temporal 62. Fisch BJ, Klass DW. The diagnostic specificity of
intermittent rhythmic delta activity in electroen- triphasic wave patterns. Electroencephalogr Clin
cephalograms. J Clin Neurophysiol 12:280-284, Neurophysiol 70:1-8, 1988.
1995. 63. Westmoreland BF, Klass DW, Sharbrough FW,
55. Chatrian GE, Shaw C-M, Leffman H. The signifi- Reagan TJ. Alpha-coma. Electroencephalographic,
cance of periodic lateralized epileptiform dis- clinical, pathologic, and etiologic correlations.
charges in EEG: an electrographic, clinical and Arch Neurol 32:713-718, 1975.
pathological study. Electroencephalogr Clin Neu- 64. Chatrian GE, White LE Jr, Daly D. Electroen-
rophysiol 17:177-193, 1964. cephalographic patterns resembling those of sleep
56. Vas GA, Cracco JB. Diffuse encephalopathies. In in certain comatose states after injuries to the head.
Daly DD, Pedley TA (eds). Current Practice of Clin- Electroencephalogr Clin Neurophysiol 15:272-280,
ical Electroencephalography, 2nd ed. Raven Press, 1963.
New York, 1990, pp 371-399. 65. American Electroencephalographic Society Guide-
57. Young GB. The EEG in coma. J Clin Neurophysiol lines in Electroencephalography, Evoked Poten-
17:473-485, 2000. tials, and Polysomnography. J Clin Neurophysiol
58. Gloor P, Kalabay O, Giard N. The electroen- 11:1-147, 1994.
cephalogram in diffuse encephalopathies: elec- 66. Chatrian G-E. Coma, other states of altered re-
troencephalographic correlates of grey and white sponsiveness, and brain death. In Daly DD, Pedley
matter lesions. Brain 91:779-802, 1968. TA (eds). Current Practice of Clinical Electroen-
59. Brenner RP, Schaul N. Periodic EEG patterns: clas- cephalography, 2nd ed. Raven Press, New York,
sification, clinical correlation, and pathophysiol- 1990, pp 425-487.
ogy. J Clin Neurophysiol 7:249-267, 1990. 67. Task Force for the Determination of Brain Death
60. Hormes JT, Benarroch EE, Rodriguez M, Klass DW. in Children. Guidelines for the determination of
Periodic sharp waves in baclofen-induced en- brain death in children. Neurology 37:1077-1078,
cephalopathy. Arch Neurol 45:814-815, 1988. 1987.
Chapter 9
ELECTROENCEPHALOGRAPHY:
ELECTROENCEPHALOGRAMS
OF NEONATES, INFANTS
AND CHILDREN
NEONATAL 0 Waves, or Cone-Shaped Waves

ELECTROENCEPHALOGRAPHIC Posterior Slow Waves of Youth
PATTERNS 14&6-Hz Positive Spike Bursts
Normal ABNORMALITIES
Abnormal Epileptiform Abnormalities
DEVELOPMENTAL CHANGES DURING Slow-Wave Abnormalities
INFANCY, CHILDHOOD, AND Conditions Giving Rise to Abnormal
ADOLESCENCE Electroencephalographic Patterns in
Hyperventilation Children
Photic Responses Transient Abnormalities
Drowsiness SUMMARY
Sleep
BENIGN VARIANTS IN CHILDREN
Posterior Slow Waves Associated With Eye
Blinks, or Slow Lambdas of Youth
NEONATAL epileptiform abnormalities and, often, indi-

ELECTROENCEPHALOGRAPHIC cate the type of seizure disorder. The EEG
PATTERNS is helpful in the evaluation of nonepileptic
spells and transient CNS symptoms by indi-
Electroencephalography (EEG) is an im- cating whether there is an associated EEG
portant part of the evaluation of many dis- change and, if so, the degree, location, and
orders in infants and children, including type of change. Ambulatory and prolonged
seizures, spells, transient central nervous sys- video EEG monitoring can be of additional
tem (CNS) symptoms, behavioral disorders, help in the evaluation of seizures, spells, and
altered states of consciousness, and lesions transient symptoms.
or conditions resulting in a disturbance of Another common referral is for evalua-
cerebral function. The most common refer- tion of children who have altered con-
ral for EEG is for evaluation of seizures, be- sciousness or are comatose. The EEG can
cause the EEG can show the presence of help determine the degree and extent of the
108
EEGs of Neonates, Infants, and Children 109
disturbance and, if specific patterns are pres- 1. Occipital-dominant slow waves, which
ent (as described in Chapter 8), help indi- are broad, high-amplitude slow waves
cate the diagnosis or prognosis (or both).1 over the occipital head regions, shift-
The EEG is helpful in evaluating condi- ing from side to side in prominence.
tions or lesions causing a disturbance of 2. The spindle delta brush pattern, which
cerebral function, in determining whether consists of moderate-to-high-amplitude
the process is focal or generalized, and in delta waves with superimposed 8- to 20-
identifying the extent of the disturbance. Hz activity; it is seen over the rolandic,
The EEG reflects the degree and extent of temporal, and occipital head regions
the disturbance and, if certain diagnostic (Fig. 9-1).
EEG patterns are present, helps to make the 3. Multifocal sharp transients or random
diagnosis. focal sharp waveforms occurring in
Children with behavioral disturbances, at- multiple locations but most frequently
tention deficits, or learning disorders are over the frontal head regions (frontal
also referred for EEGs to rule out an un- sharp transients) (Fig. 9-2).
derlying organic process. 4. Anterior slow waves, which consist of
rhythmic delta slow waves over the an-
terior head regions; they are usually as-
Normal sociated with frontal sharp transients.
5. Bursts or trains of sharp-contoured
Recordings in newborn infants usually re- theta waves over the temporal and cen-
quire monitoring of physiologic variables tral vertex regions. In the younger pre-
such as the electrocardiogram (EGG), res- mature infant, these patterns are seen
pirations, eye movement, and chin myogram in both the awake and sleep states. As
in addition to EEG recording to help deter- the child matures, these patterns are
mine the state of the infant.2-4 seen primarily during sleep and then
In infants of less than 32 weeks' concep- disappear after about 44 weeks' con-
tional age, which is the age since the first day ceptional age.
of the mother's last menses, the EEG con- Infants 38-42 weeks old show four basic
sists of an intermittent or discontinuous pat- patterns:3"5
tern, with bursts of activity alternating with 1. A low-voltage irregular pattern that is
long quiescent periods. There is little dis- present during wakefulness and active
tinction between the EEG of the awake and sleep.
sleep states.3"5 2. A high-voltage slow-wave pattern that is
Between 32 and 37 weeks' conceptional seen during quiet sleep.
age, there is an EEG distinction between the 3. A trace alternant pattern that is also
awake state and the two types of sleep states, seen during quiet sleep.
active and quiet sleep.3"5 Active sleep, which 4. A mixed pattern of theta and delta
is similar to rapid eye movement (REM) waves seen during drowsiness and ac-
sleep in adults, manifests eye movements, tive sleep and as a transitional pattern
body twitches, grimaces, reduction in mus- between the various states.
cle tone, and irregular respirations. During
this state, the EEG shows a more continuous
pattern, similar to that of the awake state. Abnormal
Quiet sleep has reduced eye and body move-
ments, increased muscle tone, regular res- The EEG in premature newborn infants
pirations, and a regular EGG. During quiet should be interpreted with care. The age of
sleep, the EEG shows a discontinuous pat- the infant is important because the EEG ac-
tern, with bursts of mixed sharp and slow- tivity changes with maturation, and what
wave activity alternating with periods of flat- might be normal at one age may be abnor-
tening of the background. This is referred mal for a more mature infant. The back-
to as the trace alternant pattern. ground rhythms are the most significant
Other types of activity that are present at EEG finding.6 If these are appropriate for
this age include:3"5 the infant's age, the infant usually has a fairly
Figure 9-1. Electroencephalogram (EEG) from a normal premature infant at 32 weeks' conceptional age (C.A.)
during sleep, showing the delta brush pattern.
good prognosis; if abnormal, then the de- ally disappear within a few days. Mild and
gree of abnormality usually reflects the de- transient focal abnormalities in the EEG
gree of disturbance and the ultimate out- usually are not associated with any obvious
come of the infant.6 Abnormalities may be focal pathologic condition. However, persis-
divided into mild and significant types. tent focal EEG abnormalities are often as-
sociated with structural lesions such as
MILD ABNORMALITIES intracranial hemorrhage or congenital
defects.3'5
Mild abnormalities, such as excessive multi-
focal sharp transients or immature and dys-
SIGNIFICANT ABNORMALITIES
mature patterns, can be seen in stressed pre-
mature or term infants. These findings are Significant abnormalities in the EEG are
nonspecific and rarely suggest a specific di- usually associated with an important distur-
agnosis.3'5 They often are transient and usu- bance of brain function and often indicate
Figure 9-2. Electroencephalo-

gram (EEG) from a normal pre-
mature infant at 34 weeks' con-
ceptional age (C.A.) during sleep,
showing a frontal sharp transient
(Fp2). Less prominent sharp tran-
sients also occur in T% and ^4.
a poor prognosis or poor neurologic out- of neonates and consists of focal or multifo-
come. The more abnormal the pattern, the cal interictal and ictal discharges.5'7'8 The in-
more severe the underlying encephalopathy terictal discharges usually take the form of
or disturbance of brain function.3-8 The fol- spikes, sharp waves, and broad slow waves.
lowing patterns are significantly abnormal: The ictal discharges consist of rhythmic ac-
tivity that may take the form of spikes, sharp
waves, slow waves, or rhythmic activity in the
Isoelectric EEC alpha, beta, theta, or delta range and may
An isoekctric EEC is a flat record that meets evolve and persist for relatively long periods.
the criteria for electrocerebral inactivity. In- Ictal electrographic discharges often occur
fants with a single flat EEG may survive the in association with clinical seizures but may
neonatal period but usually suffer severe be present without any clinical accompani-
long-term neurologic sequelae.5 ment (Fig. 9-3). If associated with seizures,
the seizures usually take the form of clonic
or tonic movements, but there may be di-
Burst-Suppression Pattern verse and subtle manifestations that may not
A burst-suppression pattern consists of diffuse be easily recognizable as epileptic.7'8
bursts of abnormal activity superimposed on
an isoelectric or very low-amplitude back-
Positive Rolandic Sharp Waves
ground. This is an invariant pattern that
does not change with state of sleep-wakeful- Positive rolandic sharp waves occur unilaterally
ness or in response to stimuli. It, too, is as- or bilaterally and are most common in the
sociated with severe encephalopathy and rolandic and midline areas.5 They were de-
poor long-term prognosis.5 scribed initially in infants with intraventric-
ular hemorrhage; however, they also occur
in patients who have periventricular leuko-
Persistent Low Voltage
malacia and deep white matter lesions.5
Persistent low voltage can occur in a general-
ized fashion in association with a diffuse dis-
turbance of function or, in a more focal fash-
Asymmetry
ion, in association with focal lesions such as An excessive and persistent asymmetry (> 50%)
porencephaly, subdural collection of fluids, of the activity during both the wake and
or congenital abnormalities. sleep states, occurring focally or lateralized
to one hemisphere, is a significant abnormal
Epileptiform Activity finding in an infant's EEG.5 This can occur
with congenital lesions, porencephalic cysts,
Epileptiform activity is one of the most fre- vascular insults, or subdural collections of
quent types of abnormalities seen in EEGs fluid.
Figure 9-3. Focal subclinical

EEG-seizure discharge aris-
ing from the left frontal re-
gion (Fs) in a 7-day-old girl
with neonatal seizures.
Periodic Discharges glycemia and hypocalcemia, which previ-

ously were the most frequent causes of
Periodic lateralized epileptiform discharges neonatal seizures. Metabolic disorders can
(PLEDs) can occur with an acute or sub- be associated with focal or multifocal epilep-
acute process, most often caused by an isch- tiform abnormalities in the EEG; the EEG
emic, hypoxic, or vascular insult, or neona- usually improves after the metabolic disor-
tal herpes simplex encephalitis.5'9'10 der has been corrected.3'5'10
Persistent Slowing Drugs and Drug Withdrawal

Persistent focal or generalized slowing in the Drug withdrawal is becoming a more fre-
delta and theta range with decreased reac- quent cause of seizures in newborn infants.
tivity can occur with a focal lesion or a dif- The EEG in these infants often shows evi-
fuse process.5 dence of cortical irritability, as manifested by
focal or multifocal epileptiform activity and
CONDITIONS GIVING RISE seizure discharges. Drugs such as anticon-
TO ABNORMAL vulsants can also cause changes in an infant's
ELECTROENCEPHALOGRAPHIC EEG, including a burst-suppression pattern
PATTERNS IN INFANTS or generalized suppression of activity.5
Hypoxic-Ischemic Insult
Infectious Diseases
This gives rise to severe EEG abnormalities
and is the most common cause of neonatal Infectious diseases involving the CNS are
seizures.3"6 frequently associated with abnormalities in
the EEG. The most characteristic finding is
seen in neonatal herpes simplex encephali-
Intraventricular Hemorrhage
tis and consists of PLEDs.9 Other findings as-
Intraventricular hemorrhage, which can sociated with infectious processes include
occur in premature infants who have a significant asymmetries, interhemispheric
hypoxic-ischemic insult, is associated with asynchronies, multifocal sharp waves, and
rolandic positive sharp waves in the EEG.3 seizure discharges.
Metabolic Disorders Inborn Errors of Metabolism

The most common metabolic disorders that Inborn errors of metabolism, biochemical
produce abnormal EEG patterns are hypo- disorders, and aminoacidurias can be asso-
Figure 9-4. Focal spikes in an 8-

month-old boy with focal right-
sided motor seizures caused by
tuberous sclerosis complex.
Figure 9-5. Repetitive spike discharges over the left hemisphere in a 2-year-old boy with left hemisphere pachygyria.
elated with abnormal EEG patterns. Phe- Congenital Abnormalities

nylketonuria previously was a common
cause, but other types of aminoacidurias and These may be associated with various types
inborn errors of metabolism can present and combinations of abnormal patterns re-
with neonatal seizures and epileptiform flecting the abnormality.3'5
discharges.5
DEVELOPMENTAL CHANGES
Dysgenetic Disorders or DURING INFANCY,
Neurocutaneous Disorders CHILDHOOD, AND
Tuberous sclerosis complex presenting in in- ADOLESCENCE
fancy may be associated with a hypsarrhyth-
mic pattern or focal or multifocal EEG ab- In the first 3 months of life, a transition oc-
normalities that may or may not be related curs from the neonatal to the infant EEG
to the location of the tubers11 (Fig. 9-4). pattern.3'4 The background consists of ir-
Sturge-Weber syndrome is associated with regular low-amplitude delta activity. Rhyth-
an asymmetry of background activity and mic activity in the range of 5-6 Hz is pres-
epileptiform activity on the side of the facial ent in the central regions and is probably a
nevus. precursor of the mu rhythm. At 3 months of
age, rhythmic occipital activity in the range
Cortical Malformations of 3-4 Hz is present. This activity can be at-
tenuated with eye opening and represents a
Cortical malformations, particularly cortical precursor of the alpha rhythm. Between 4
dysplasia, may be highly epileptogenic and and 6 months of age, the central rhythm be-
can result in focal or multifocal epileptiform comes better developed and shows a fre-
abnormalities with frequent interictal and ic- quency of 5-8 Hz; a better defined occipital
tal discharges. At times almost continuous rhythm in the range of 5-6 Hz is present
trains of spike discharges or rhythmic epilep- when the eyes are closed. Between 6 months
tiform discharges may be present12"15 (Fig. and 2 years of age, the central rhythm is well
9-5). developed at a frequency of 5-8 Hz. After
6 months, the occipital rhythms become Drowsiness

more prominent, and there is a gradual shift
to higher amplitude and faster frequency ac- Drowsiness in young children is character-
tivity, ranging from 6 to 8 Hz. In patients be ized by high-amplitude sinusoidal 4-5 Hz
tween 2 and 5 years of age, the central and theta activity that is maximal over the frontal,
occipital rhythms are further differentiated. central, and parietal regions. These slow
By 3 years of age, the occipital alpha rhythms waves initially occur in prolonged rhythmic
range from 6 to 8 Hz, and the amplitude of trains and, in children between 1 and 9 years
this activity gradually increases.16'17 Between old, in bursts16'17 (Fig. 9-7). Sometimes the
6 and 16 years of age, there is a progressive slow waves may have a notched or sharp ap-
increase in the alpha frequencies, and the pearance because of superimposed faster
typical adult frequency range of 9-10 Hz is frequencies.17 During adolescence, charac-
usually reached by 10-12 years of age. Some teristic trains of monorhythmic sinusoidal
interspersed theta activity may still be pres- theta activity occur over the frontal regions
ent, predominantly over the anterior head and may precede the disappearance of the
regions. During the first decade of life, there alpha rhythm.
is considerable variability among children of
the same age with regard to the amount of
alpha, theta, and delta activity present.16'17 Sleep
During the first few months of life, the trace
Hyperventilation alternant pattern of newborns (see above un-
der Normal) is replaced by generalized slow-
In younger children, the slowing produced wave activity during quiet sleep, and the per-
by hyperventilation is often maximal over centage of time spent in nonrapid eye
the posterior head regions (Fig. 9-6), movement (NREM) sleep is increased. Spin-
whereas in older children, the buildup re- dles become apparent by 1-3 months of age
sponse is usually maximal over the anterior and are well developed and bisynchronous
head regions.16 by 1-2 years of age. The spindles in the pa-
tient's first year of life may have a charac-
teristic arciform or comb-like appearance
Photic Responses and occur in prolonged and asynchronous
trains. V waves are apparent by 3-5 months
Young children show responses in the oc- of age. In children from 2 to 5 years old,
cipital regions predominantly at slower flash V waves have a high amplitude and a sharp
frequencies. In older children, driving can or spiky appearance and occur in groups. O
be seen at the faster flash frequencies. waves are large, broad, bioccipital delta
Figure 9-6. A, Normal EEC from a 6-year-old boy during resting wakefulness and, B, during hyperventilation.
Figure 9-7. Normal burst of rhythmic slow waves during drowsiness in a SVg-year-old boy.
waves that are present over the occipital re- surface-negative. The waveform may have a
gions during drowsiness and sleep. Occa- sharp-contoured appearance, but it should
sionally they may have a somewhat sharply not be misinterpreted as abnormal epilepti-
contoured monophasic or diphasic waveform activity. These waveforms can be seen
form.17 Occipital slow waves are most promi- in children who are between 6 months and
nent at 1-2 years of age but may persist to 10 years old and are most prominent in
at least age 5 years. those 2-3 years old.
BENIGN VARIANTS O Waves, or Cone-Shaped Waves

IN CHILDREN
In young children, high-voltage slow-wave
Several benign variants can be seen in the transients, 0 waves, varying from a cone-
EEGs of children and adolescents. These in- shaped appearance to diphasic slow-wave
clude the slow lambdas of youth, O waves, transients may be present over the occipital
posterior slow waves of youth, and 14&6-Hz head regions and interspersed with occipital-
positive bursts. dominant slow delta waves of sleep.17 This
activity should not be mistaken for abnormal
sharp wave or slow wave activity.
Posterior Slow Waves Associated
With Eye Blinks, or Slow Lambdas
of Youth Posterior Slow Waves of Youth
A phenomenon similar to that of the lambda Single delta frequency waves, referred to as
waves is the posterior slow waves associated posterior slow waves of youth or slow fused tran-
with eye blinks in some children (slow lamb- sients, are common over the posterior head
das of childhood, or shut-eye waves) . They are regions in children and adolescents.16'17
single, broad, and monophasic or diphasic The posterior slow waves of youth occur spo-
waveforms that occur bilaterally over die oc- radically rather than in consecutive trains,
cipital head regions after eye blinks or eye do not protrude much above the average
movements.18 The amplitude is often 100- amplitude of the alpha rhythm, and attenu-
200 iiV, and the duration is approximately ate together with the alpha rhythm when the
200-400 ms. The predominant polarity is eyes are opened.
116
Figure 9-8. Hypsarrhythmia in a 9-month-old boy with infantile spasms.
14&6-Hz Positive Spike Bursts Epileptiform Abnormalities

The / 48c 6-Hz positive spike bursts consist of brief Almost any type of epileptiform abnormality
bursts or trains of positive spikes that occur can be seen in children. Some types of
during drowsiness at a rate of 14 and 6 Hz, epileptiform abnormalities and associated
mainly over the posterior temporal regions. seizure disorders are unique or seen more
They can be seen in children but are seen commonly in children.20 The types of epilep-
most often in adolescents19 (see Fig. 8-13). tiform abnormalities most common in chil-
dren include the following:
1. Hypsarrhythmia—a pattern seen in chil-
ABNORMALITIES dren from 4 months to 4 years old. It
consists of high-amplitude multifocal
Electroencephalograms in children can spikes, sharp waves, and slow waves.
show a wide variety of abnormalities. The This type of epileptiform activity is of-
most common are epileptiform abnormali- ten seen in association with infantile
ties and slowing. spasms10'19'20 (Fig. 9-8).
Figure 9-9. Absence seizure accompa-

nied by typical paroxysm of 3-Hz spike
and slow-wave complexes during hyper-
ventilation in a 10-year-old girl.
Figure 9-10. Slow spike-and-

wave pattern (sharp and slow-
wave complexes) in a 5-year-old
boy with seizures and mental re-
tardation (Lennox-Gastaut syn-
drome) .
2. 3-Hz Generalized Spike-and-Wave—a pat- tal retardation (Lennox-Gastaut syn-

tern usually seen in children between drome)10'19'20 (Fig. 9-10).
3 and 15 years old. It is associated with 4. Central-Temporal Spikes—a pattern seen
absence seizures10-19'20 (Fig. 9-9). in children 4-12 years old. The spikes
3. Generalized Slow Spike-and-Wave—a pat- have a blunt spike-and-wave appear-
tern (generalized sharp and slow- ance. This pattern occurs in children
wave discharges) seen in young chil- with benign rolandic epilepsy of child-
dren with frequent seizures and men- hood (BREC)20-22 (Fig. 9-11). The
Rgure 9-11. Central-temporal spikes (maximal in C4 and T4) in a 7-year-old boy with a history of a single noctur-
nal seizure.
Figure 9-12. Occipital spike discharges attenuated with eye opening in a 10-year-old boy with benign occipital
seizures of childhood.
above epileptiform patterns are de- Slow-Wave Abnormalities

scribed more fully in Chapter 8.
5. Occipital Spikes and Seizures—a pattern Slow-wave abnormalities may be focal or dif-
seen in children with occipital lesions fuse and are often more prominent and take
because of birth injury, vascular lesions, longer to resolve in children than in adults.
congenital malformations, cortical dys- Slow-wave abnormalities in children also
genesis, Sturge-Weber syndrome, tuber- tend to have maximal expression over the
ous sclerosis, tumors, or trauma.10'20'23 posterior head region.10'19
Occipital spikes are also seen in the
benign epilepsies of children with oc-
cipital paroxysms23'24 (Fig. 9-12). This Conditions Giving Rise to Abnormal
is a seizure disorder associated with el-
ementary visual phenomena; it may Electroencephalographic Patterns
progress to secondarily generalized in Children
tonic-clonic seizures. The child may
also have nocturnal seizures with head Electroencephalographic abnormalities can
and eye deviation, nausea, and vomit- be seen in many disorders, including de-
ing. The seizures sometimes are fol- generative disorders, inflammatory diseases,
lowed by migraine headaches. The in- tumors, vascular insults, head trauma, and
terictal EEG shows spike-and-wave various types of encephalopathies. Transient
discharges over the occipital head re- abnormalities also occur with migraine head-
gions. These discharges occur in a uni- aches and postictal states.
lateral, bilaterally independent, or bi-
laterally synchronous manner, are
DEGENERATIVE DISORDERS
attenuated with eye opening, and re-
occur with eye closure. Ictal discharges Degenerative disorders, various aminoacid-
consist of low-voltage fast activity over urias, and inborn errors of metabolism may
the occipital head regions, but it can be associated with slowing and multifocal
spread more widely. The child often epileptiform abnormalities in the EEG, par-
outgrows the seizures and the spike dis- ticularly if the child has seizures.
charges.21'23'25 Focal occipital spike The type of degenerative process and
discharges can also be seen in young whether it involves the white or gray matter
children with amblyopia, without asso- influences the EEG pattern. Processes that
ciated seizures.10'19'20 affect predominantly the white matter usu-
ally cause polymorphic delta slowing in the INFLAMMATORY DISORDERS

EEC, whereas processes involving the corti-
cal or subcortical gray matter are more likely Meningitis
to cause bilaterally synchronous paroxysma This is associated with differing degrees of
activity.26 Epileptiform abnormalities are slowing in the EEG depending on the type
more common in gray matter disease but of meningitis and the degree of involvement
also can occur in white matter disease. Other of the CNS. Purulent meningitis is often as
factors influencing the degree and type of sociated with moderate to severe generalized
EEG abnormalities include the age and sta- slow-wave abnormalities, and epileptiform
tus of the patient, the stage of the disease discharges may be present in patients who
process, and other complicating factors, in- have seizures.27
cluding infectious, metabolic, and drug
effects.
Encephalitis
Gray Matter Disease The EEG abnormalities in encephalitis often
are more severe than those in meningitis,
Gray matter disease such as the progressive with the EEG showing high-voltage arrhydi-
myoclonic epilepsies is associated with gen- mic or rhythmic delta slowing (see Fig.
eralized epileptiform abnormalities and 8-26). Focal or multifocal epileptiform ab-
slowing in the EEG of patients following the normalities may occur in patients who have
onset of seizures.10'26 seizures.27
Herpes simplex encephalitis can occur in in-
White Matter Disease fants and children. As in adults, PLEDs are
a prominent feature in the EEG.9 After res-
White matter disease, including the various olution of the infection, the EEG may show
leukodystrophies, is associated widi the loss localized areas of attenuation overlying cys
of background activity and moderate- to tic areas of the brain or multifocal epilepti-
high-amplitude delta slowing, which is of- form abnormalities or both.
ten maximal over the posterior head re- Rasmussen encephalitis, a syndrome of
gions10'19'26 (Fig. 9-13). chronic smoldering encephalitis, is charac-
Figure9-13 Delta slowing overthe posteriorheadregionns i a 3 year old boy with metachromic leukiodystrophy
terized by progressive neurologic and in- HEMICONVULSIONS, HEMIPLEGIA,

tellectual deterioration and recurrent AND EPILEPSY
seizures.28 The patients have varying types of
seizures, which may progress to epilepsia In the hemiconvukions, hemiplegia, and epilepsy
partialis continua. The EEG shows various syndrome, the infant or child has a series of
types of epileptiform and slow-wave abnor- seizures or hemiconvulsive status epilepticus
malities that occur in different locations dur- during an acute febrile illness.10 Following
ing the different stages of the disease pro- the acute episode of seizures the child has
cess. Often, Rasmussen encephalitis initially hemiparesis, and later, chronic epilepsy de-
involves one hemisphere and then may velops. In the acute stage of the syndrome,
spread more widely. the EEG shows frequent or continuous spike
Subacute sclerosing panencephalitis, a slow discharges or spike-and-wave discharges over
virus disorder believed to be caused by the the involved side. After the acute stage, there
measles virus that occurs in children and is persistent attenuation of the EEG activity
adolescents, is associated with repetitive over the affected side. Focal, unilateral, or
stereotyped high-voltage sharp and slow- multifocal epileptiform discharges subse-
wave complexes that recur every 4 or 5 sec- quently occur over the affected side.
onds and are associated with stereotyped
motor jerks or spasms27 (see Fig. 8-28). TUMORS
The EEGs of children with tumors show a
Brain Abscess greater predominance of slow-wave abnor-
Focal polymorphic delta slowing is often malities over the posterior head regions than
present over the site of the abscess if the le- those of adults with tumors. This may partly
sion is located close to the surface of the reflect two things: 1) children have a greater
brain (Fig. 9-14). Although epileptiform ab- incidence of posterior fossa tumors than
normalities are not usually seen in acute adults and 2) the predominance of slow-
stages, they may develop in the later stages wave abnormalities over the posterior head
of resolution of the abscess.27 region in children is an age-related phe-
nomena. In supratentorial tumors, the EEG
SUBDURAL EFFUSION shows focal or lateralized slow-wave abnor-
AND EMPYEMA malities, asymmetry, or epileptiform activity
over the involved area.10
Subdural effusions, hygroma, and empye-
mas act like subdural hematomas and cause VASCULAR LESIONS
an attenuation or decreased amplitude of ac-
tivity and slowing over the involved hemi- Vascular insults, including infarcts and hem-
sphere.10 orrhage, are less common in children than
Figure 9-14. Focal delta slowing over the

right frontal region in an 8-year-old boy
with a right frontal abscess.
Figure 9-15. Moderate diffuse slow-wave

abnormalities after head trauma in a 10-
year-old girl.
adults, but when present, they are associated be out of proportion to the degree of head
with focal slowing and loss of background ac- injury.10
tivity. Periodic lateralized epileptiform dis-
charges may be seen in the acute stage and HYDROCEPHALUS
focal epileptiform abnormalities, in the
chronic stage.10 The EEG abnormalities in patients with hy-
drocephalus may consist of focal or gener-
HEAD TRAUMA alized slow-wave abnormalities, epileptiform
abnormalities, asynchronous sleep activity,
The EEG may show focal, lateralized, or gen- or asymmetry of the background activity.
eralized slow-wave abnormalities or an asym- There may be an increase in the slow-wave
metry of activity (or both slow-wave abnor- and epileptiform abnormalities with ob-
malities and asymmetry) depending on the structive hydrocephalus because of mal-
extent of the head injury (Fig. 9-15). The function of the shunt (Fig. 9-16). The inci
slow-wave abnormalities often are maximal dence of EEG abnormalities is higher in
over the posterior head regions. A moderate children who have had a ventricular shunt,
degree of slowing is not uncommon after rel- and focal abnormalities are often present in
atively minor head injury, and slowing may the area of the shunt.
Figure 9-16. Intermittent rhythmic

slow-wave abnormalities in a 12-year-old
child with obstructive hydrocephalus
with a blocked shunt.
122
Figure 9-17. Prominent delta slowing over the right hemisphere during a hemiplegic migrainous episode in a
10-year-old girl.
Transient Abnormalities lying organic disorder, indicates the degree

and extent of cerebral dysfunction, confirms
MIGRAINE HEADACHE or helps make the diagnosis of a specific dis-
order such as seizures, indicates whether
In between migrainous attacks, the EEC may there is a focal or generalized process, and,
show nonspecific abnormalities. During and in certain instances, aids in determining the
after a migraine episode, asymmetry and fo- prognosis.
cal, lateralized, or generalized slowing may be
present. Slow-wave abnormalities are usually
more prominent and widespread in children REFERENCES
than in adults10 (Fig. 9-17). The abnormal-
ities may persist longer in children than in 1. Shewmon DA. Coma prognosis in children. Part II:
adults and may take up to a week to resolve. Clinical application. J Clin Neurophysiol 17:467-
472, 2000.
2. American Electroencephalographic Society. Guide-
POSTICTAL SLOWING line two: minimum technical standards for pediatric
electroencephalography. J Clin Neurophysiol 11:
Postictal slowing and attenuation of the 6-9, 1994.
background may be more prominent and 3. Hahn JS, Tharp BR. Neonatal and pediatric elec-
persist longer in children than in adults. troencephalography. In AminofF MJ (ed). Electro-
diagnosis in Clinical Neurology, 4th ed. Churchill
Livingstone, New York, 1999, pp 81-127.
4. Stockard-Pope JE, Werner SS, Bickford RG. Adas of
SUMMARY Neonatal Electroencephalography, 2nd ed. Raven
Press, New York, 1992.
5. Scher MS. Electroencephalography of the newborn:
This chapter discusses the normal and abnormal and abnormal features. In Niedermeyer E,
normal EEG patterns in neonates, infants, Lopes da Silva F (eds). Electroencephalography:
and children. Because the EEG is a neuro- Basic Principles, Clinical Applications, and Related
physiologic test that reflects a disturbance of Fields, 4th ed. Williams & Wilkins, Baltimore, 1999,
pp 896-946.
cerebral function, it is an important tool in 6. Holmes GL, Lombroso CT. Prognostic value of
evaluating infants and children. The EEG background patterns in the neonatal EEC. J Clin
helps determine whether there is an under- Neurophysiol 10:323-352, 1993.
7. Mizrahi EM. Pediatric electroencephalographic dren. In Daly DD, Pedley TA (eds). Current Prac-
video monitoring. J Clin Neurophysiol 16:100-110, tice of Clinical Electroencephalography, 2nd ed.
1999. Raven Press, New York, 1990, pp 139-199.
8. Mizrahi EM, Kellaway P. Characterization and clas- 18. Westmoreland BF, Klass DW. Unusual EEG pat-
sification of neonatal seizures. Neurology 37:1837- terns. J Clin Neurophysiol 7:209-228, 1990.
1844, 1987. 19. Daly DD, Pedley TA (eds). Current Practice of Clin-
9. Mizrahi EM, Tharp BR. A characteristic EEC pat- ical Electroencephalography, 2nd ed. Raven Press,
tern in neonatal herpes simplex encephalitis. Neu- New York, 1990.
rology 32:1215-1220, 1982. 20. Engel J, Pedley TA, Aicardi J (eds). Epilepsy: A
10. Niedermeyer E, Lopes da Silva F (eds). Electroen- Comprehensive Textbook. Lippincott-Raven, Phil-
cephalography: Basic Principles, Clinical Applica- adelphia, 1998.
tions, and Related Fields, 4th ed. Williams & 21. Loiseau P. Benign focal epilepsies of childhood. In
Wilkins, Baltimore, 1999. Wyllie E (ed). The Treatment of Epilepsy: Princi-
11. Westmoreland BF. The electroencephalogram in ples and Practice. Lea & Febiger, Philadelphia,
tuberous sclerosis. In Gomez MR (ed). Tuberous 1993, pp 503-512.
Sclerosis Complex, 3rd ed. Oxford University Press, 22. Kellaway P. The electroencephalographic features
New York, 1999, pp 63-74. of benign centrotemporal (rolandic) epilepsy of
12. Gambardella A, Palmini A, Andermann F, et al. Use- childhood. Epilepsia 41:1053-1056, 2000.
fulness of focal rhythmic discharges on scalp EEG 23. Panayiotopoulos CP (ed). Benign Childhood Par-
of patients with focal cortical dysplasia and in- tial Seizures and Related Epileptic Syndromes. John
tractable epilepsy. Electroencephalogr Clin Neuro- Libbey, London, 1999.
physiol 98:243-249, 1996. 24. Kivity S, Ephraim T, Weitz R, Tamir A. Childhood
13. Palmini A, Andermann F, Olivier A, et al. Focal neu- epilepsy with occipital paroxysms: clinical variants
ronal migration disorders and intractable partial in 134 patients. Epilepsia 41:1522-1533, 2000.
epilepsy: a study of 30 patients. Ann Neurol 25. Ferric CD, Beaumanoir A, Guerrini R, et al. Early-
30:741-749, 1991. onset benign occipital seizure susceptibility syn-
14. Raymond AA, Fish DR. EEG features of focal mal- drome. Epilepsia 38:285-293, 1997.
formations of cortical development. J Clin Neuro- 26. Schaul N. The fundamental neural mechanisms of
physiol 13:495-506, 1996. electroencephalography. Electroencephalogr Clin
15. Raymond AA, Fish DR, Boyd SG, Smith SJ, Pitt MC, Neurophysiol 106:101-107, 1998.
Kendall B. Cortical dysgenesis: serial EEG findings 27. Westmoreland BF. The EEG in cerebral inflamma-
in children and adults. Electroencephalogr Clin tory processes. In Niedermeyer E, Lopes da Silva F
Neurophysiol 94:389-397, 1995. (eds). Electroencephalography: Basic Principles,
16. Blume WT, Kaibara M (eds). Atlas of Pediatric Elec- Clinical Applications, and Related Fields, 4th ed.
troencephalography, 2nd ed. Lippincott-Raven, Williams & Wilkins, Baltimore, 1999, pp 302-316.
Philadelphia, 1999. 28. Andermann F (ed). Chronic Encephalitis and
17. Kellaway P. An orderly approach to visual analysis: Epilepsy: Rasmussen's Syndrome. Butterworth-
characteristics of the normal EEG of adults and chil- Heinemann, Boston, 1991.
Chapte 10
AMBULATORY
ELECTROENCEPHALOGRAPHY
Jeffrey R. Buchhalter
INDICATIONS
TECHNOLOGY
CLINICAL APPLICATIONS
SUMMARY
INDICATIONS during school. A routine awake and

sleep EEG is normal.
Ambulatory electroencephalography (AEEG) is a Clinical Scenario 3: A 27-year-old woman
clinical neurophysiology tool that is most describes a numb feeling in her left arm
useful when a diagnosis cannot be made on and states that she "blanks out" 3 or 4
the basis of the clinical history, neurologic times a week. A routine awake and sleep
examination, and routine, outpatient awake EEG is normal.
and sleep EEG studies. Recently, AEEG has Clinical Scenario 1 poses the problem of an
been used to define the frequency of epilep- infant who may have nocturnal seizures, in-
tiform discharges in a nonepileptic popula- termittent aspiration, or respiratory obstruc-
tion.1 The advantages of AEEG include the tive disease. Ambulatory electroencephalog-
ability to perform prolonged recording in raphy could clarify the event by demonstrating
the patient's natural environment and to epileptiform abnormalities associated with the
record multiple cycles of sleep that have not event or by recording the lack of chest wall
been induced with medication. Also, the cost movement and nasal airflow, with the use of
saving compared with the cost of inpatient additional polysomnogram channels, or by
evaluation is substantial. The likelihood of showing no abnormalities at all.
recording useful information depends on Clinical Scenario 2 describes the frequent
the clinical scenario that prompted the rou- problem of the "staring child." The differ-
tine EEG and the frequency of the event. ential diagnoses that could be refined by
The following are situations in which AEEG recording the event with AEEG in the class-
may be indicated: room setting include normal daydreaming,
Clinical Scenario 1: Parents report that complex partial seizures, and absence sei-
their 9-month-old child intermittently zures. If the event is a seizure, precise quan-
stiffens and stops breathing for 10 sec- titation is possible, as is the detection of sub-
onds within 2 or 3 hours after the on- clinical (electrographic) seizures.
set of sleep each night. Routine EEG Clinical Scenario 3 could be caused by a
with sleep recording for 15 minutes is complex partial seizure, a pseudoseizure,
normal. or cardiovascular or migraine phenomena.
Clinical Scenario 2: Teachers report that Recording the event with AEEG could dem-
a 6-year-old boy stares 1 or 2 times a day onstrate epileptiform abnormalities consis-
124
Ambulatory Electroencephalography 125
tent with a seizure or suggest vascular insuf- duction of computerized, off-line analysis of
ficiency as indicated by focal or generalized the analog data. These algorithms were dem-
slowing. onstrated to provide excellent detection of
epileptiform rhythmic activity.7
Analog recording on cassettes was re-
placed by the application of digital, com-
TECHNOLOGY puter-based technology that has rendered
the recording quality of AEEG comparable
Although single-channel ambulatory cardiac to that of in-laboratory EEG. Since the ini-
monitoring was reported in 1949, it was not tial report of this type of system,8 an im-
until the 1970s that the miniaturization of pressive array of commercially available
cassette recording allowed AEEG to become products with various modifications has
a reality. Devices capable of recording 4 become available. All systems use digital
channels of physiologic data were modified recording that allows computer-based refor-
to record EEGs. The addition of pream- matting of the montage at the time of re-
plifiers reduced the noise level from 30 to view. Basic recording features vary slightly
50 /AV peak-to-peak to 5 juV peak-to-peak, between systems with regard to the number
so that EEG background activity and higher of recording channels (16—32), sampling
amplitude epileptiform transients could be rate (1-500 Hz), analog-to-digital conversion
distinguished from background noise.2 The (12-22 bit), frequency bandwidth (DG-70
problem of prolonged playback times Hz), and common mode rejection of greater
caused by the limited response time of me- than 100 dB. Continuous recording can be
chanical pens was solved by adaptation of an performed from 24 to 60 hours depending
ink jet printer that allowed playback of a on the battery life and data compression
24-hour recording in 24 minutes. Thus, by techniques used. Seventy-two hours of re-
1980, it was possible to record 4 channels for cording is possible if events or samples are
24 hours on a lightweight, analog cassette stored in an intermittent rather than a con-
recorder. This system was used to quantify tinuous mode. Most systems incorporate
3-Hz spike-wave discharges in children with epileptiform transient and seizure detection
childhood absence epilepsy, to lateralize algorithms. Systems vary significantly with re-
EEG abnormalities, and to distinguish pseu- gard to the availability of pulse oximetry, spe-
doseizures from epileptic seizures. cific channels for polysomnography, and si-
However, substantial technical problems multaneous video recording. Clinical events
included limited spatial resolution and dif- are recorded in an event calendar and indi-
ficulty with artifact detection because of the cated by pushing an event marker button.
limited number of channels, inability to re- One system provides an audio channel for
format montages, and inferior recording the patient or attendant to indicate the time
characteristics when compared with those of and nature of the event. The data are down-
standard hardwired systems. To estimate the loaded from the AEEG storage medium to a
effect of limited channels, a series of rigor- standard desktop personal computer (PC)
ous studies was performed in epilepsy pa- on which proprietary software allows review
tients undergoing standard, inpatient EEG with montage reformatting, ability to alter
monitoring. The distribution of epilepti- the filter and sensitivity settings, and analy-
form discharges was determined,3 the mon- sis of detected events.
tage of the 4-channel AEEG was accordingly Despite the significant technical advances,
optimized,4 and the detection of epilepti- several limitations of AEEG remain. Elec-
form abnormalities by the two techniques trode stability cannot be assured in the home
was compared.5 It was found that even a lim- setting as it can in the EEG laboratory or
ited number of recording channels could de- epilepsy monitoring unit. A trained EEG
tect approximately 75% of the abnormali- technologist or nurse can test the patient's
ties. The subsequent introduction of an response to the external environment in a
8-channel system markedly improved the manner not feasible with AEEG. Also, the re-
ability to identify artifacts and improved lo- duction or discontinuation of antiepileptic
calization.6 This was enhanced by the intro- medication to facilitate seizure recording in
the inpatient setting would not be safe in the To compare the utility of recording tech-
home. niques in typical absence epilepsy, 25 chil-
dren with this epilepsy syndrome had a rou-
tine awake EEG, followed by an 8-hour
CLINICAL APPLICATIONS AEEG in the awake state.13 These studies
were repeated 1 month after the initiation
There are two fundamental types of clinical of treatment. Although the initial prolonged
applications of AEEG. The first type is de- study did not add to diagnostic accuracy, at
termining whether an event is of epileptic 1 month, only 4 children had spike-wave ab-
etiology (including persons with known normalities on the routine EEG and 10 chil-
seizures). Common examples include loss or dren had epileptiform discharges on the
impairment of consciousness, behavioral dis- AEEG despite having valproic acid levels in
turbances, motor phenomena, and sensory the therapeutic range. It appears that in this
experiences.9 The second type is in persons syndrome AEEG is very useful when the clin-
with known seizure disorders in which the ician and parents believe that seizures have
AEEG may clarify partial vs. generalized on- been eliminated.
set, quantification of electrographic sei- Ambulatory electroencephalography has
zures, and localization of ictal onset for pos- been shown to provide diagnostic and prog-
sible epilepsy surgery. The utility of AEEG is nostic information in children with infantile
best demonstrated when comparison can be spasms. This severe symptomatic epilepsy
made with the same population previously syndrome of childhood is characterized by
(or simultaneously) studied with either rou- frequent clinical and electrographic sei-
tine awake/sleep EEG or inpatient, video- zures. In a cohort of 74 infants with infan-
EEG recording. The following review high- tile spasms, AEEG detected partial seizures
lights the applications noted above. in 51% in addition to the generalized
An early study with the 4-channel cassette spasms, and the partial seizures were associ-
recorder of 100 children and adults with ated with an unfavorable outcome.14 This
temporal lobe epilepsy indicated that AEEG study contributed to the recognition that
was at least 3 times more effective in record- a generalized seizure disorder can be associ-
ing a seizure than routine EEG.10 Laterality, ated with focal seizures that indicate a rela-
but not precise localization of the seizure fo- tively poor outcome. More recently, the util-
cus, could be determined in many of the pa- ity of the computerized, 16-channel, digital
tients. Applications of AEEG in the pediatric AEEG with a seizure detection algorithm was
population have included distinguishing, demonstrated in a population of children to
epileptic from nonepileptic spells, predict- differentiate epileptic from nonepileptic
ing outcome after neonatal seizures, quanti- events.15 The seizure was detected by the
fying absence seizures, and characterizing computer only and not by clinical observa-
infantile spasms. An 8-channel system was tion in 5 of 26 recordings.
used to study 95 infants and children who One issue that remains unclear is how
had clinically likely (n = 40) or unlikely long ambulatory monitoring should.be per-
(n = 55) seizures.11 In the known seizure formed to clarify the epileptic etiology of an
group, recorded seizures aided in the quan- event. This question was approached in a
tification of seizure frequency. Ambulatory large study of 2221 patients (most of them
recordings captured events in 24 of the sus- adults).16 Ambulatory electroencephalogra-
pected pseudoseizure group and demon- phy recorded typical clinical events in ap-
strated no ictal EEG changes. A study of proximately one-third of the study popula-
infants who had neonatal seizures demon- tion. Of these patients, one-third had EEG
strated that a combination of findings on the findings that correlated with the ictus. The
routine EEG performed at the time of the recording duration ranged from 1 to 8 days,
seizure followed by subsequent AEEG pre- with 90% of patients having an event within
dicted the risk of seizure recurrence at 4 2 days. This is the only study that provides
months of age in 92% of infants.12 This study information about the likely yield of AEEG
showed how routine EEG and AEEG might monitoring with time, although interpreta-
complement each other. tion is limited because of the lack of clinical
Ambulatory Electroencephalography 127
information about the patients, thereby lim- electrophysiologic evaluation.21 The authors
iting the ability to predict who may benefit were careful to note that AEEG cannot be
from more prolonged study. used in circumstances that would compro-
Relatively few recent investigations have mise the safety of the patient, for example,
directly compared routine EEG and AEEG. when the dose of an antiepileptic drug is de-
In a study of 344 patients, 16-channel, creased to facilitate the recording of seizures
computer-assisted AEEG proved "clinically or when intracranial electrodes are required.
useful" (defined as detecting an epileptiform
abnormality or recording a clinical event
that did not have an epileptiform EEG cor- SUMMARY
relate) in 74% of the study population.17 A
similar proportion (67%) of AEEG studies Ambulatory electroencephalography is an
was useful in 191 patients who had normal or effective tool for determining the possible
nonspecific routine EEG recordings. A sub- epileptic etiology of events involving alter-
sequent investigation revealed that manage- ations in conscious, behavioral, sensory, and
ment was affected directly by the results of the motor activity when sufficient information
16-channel AEEG recording in approxi- has not been obtained from routine
mately 80% of the patients referred. awake/sleep EEG. Ambulatory electroen-
A potential application of AEEG is the de- cephalography is also useful in the classifi-
tection of abnormalities that occur during cation, quantification, and localization of
sleep. A recent study compared the detec- seizures. Technical barriers to high-quality
tion of abnormalities in routine sleep-de- recording have been eliminated by com-
prived EEGs vs. AEEG in patients with nor- puter-assisted digital systems with seizure
mal or nonspecific awake recordings.19 detection algorithms. Out-of-hospital AEEG
Approximately the same percentage of in- should not be used when the dose of med-
terictal epileptiform transients was found ication is decreased or intracranial electrode
by both techniques during sleep. However, implantation is required. Ambulatory elec-
whereas the sleep-deprived EEG recorded troencephalography is likely to provide sub-
no seizures, the AEEG recorded seizures in stantial cost savings compared with the cost
15% of patients. Furthermore, for approxi- of misdiagnosis of nonepileptic events, in-
mately half of the seizures, patients did not appropriate medication use, physician uti-
indicate awareness of their seizures by acti- lization, and inpatient EEG evaluation, but
vating a push button. This study highlights these have not been evaluated formally.
the value of prolonged recording and sei-
zure detection algorithms.
The role of AEEG in the selection of can- REFERENCES
didates for epilepsy surgery is evolving. As
noted above, the early 4-channel system did 1. Schachter SC, Ito M, Wannamaker BB, et al. Inci-
dence of spikes and paroxysmal rhythmic events in
not provide the necessary localizing infor- overnight ambulatory computer-assisted EEGs of
mation.5 Similar results were found in a normal subjects: a multicenter study. J Clin Neuro-
study in which an 8-channel AEEG record- physiol 15:251-255, 1998.
ing with sphenoidal electrodes was com- 2. Ives JR. 4-Channel 24 hour cassette recorder for
long-term EEG monitoring of ambulatory patients.
pared directly with a standard 16-channel Electroencephalogr Clin Neurophysiol 39:88-92,
recording in an inpatient population un- 1975.
dergoing study for epilepsy surgery.20 The 3. Leroy RF, Ebersole JS. An evaluation of ambulatory,
AEEG system was comparable to the stan- cassette EEG monitoring: I. Montage design. Neu-
dard system for hemispheric lateralization, rology 33:1-7, 1983,
4. Ebersole JS, Leroy RF. An evaluation of ambulatory,
but it was not as reliable for precise local- cassette EEG monitoring: II. Detection of interictal
ization of seizure onset. However, it has been abnormalities. Neurology 33:8-18, 1983.
suggested recently that the use of AEEG sys- 5. Ebersole JS, Leroy RF. Evaluation of ambulatory cas-
tems with at least 16 channels in a carefully sette EEG monitoring: III. Diagnostic accuracy com-
pared to intensive inpatient EEG monitoring. Neu-
selected population of patients may provide rology 33:853-860, 1983.
the necessary information to proceed di- 6. Ebersole JS, Bridgers SL. Direct comparison of 3-
rectly to epilepsy surgery without inpatient and 8-channel ambulatory cassette EEG with in-
tensive inpatient monitoring. Neurology 35:846- 15. Foley CM, Miles DK, Legido A, Grover W. Diag-
854, 1985. nostic value of computerized outpatient long-term
7. Koffler DJ, GotmanJ. Automatic detection of spike- EEG monitoring in children and adolescents (ab-
and-wave bursts in ambulatory EEG recordings. stract). Epilepsia 34 (Suppl 6): 139, 1993.
Electroencephalogr Clin Neurophysiol 61:165-180, 16. Tuunainen A, Nousiainen U, Mervaala E, Riekki-
1985. nen P. Efficacy of a 1- to 3-day ambulatory elec-
8. Ives JR, Mainwaring NR, Schomer DL. An 18-chan- troencephalogram in recording epileptic seizures.
nel solid-state ambulatory EEG event recorder for Arch Neurol 47:799-800, 1990.
use in the home and hospital environment (ab- 17. Morris GL III, Galezowska J, Leroy R, North R. The
stract). Epilepsia 33 (Suppl 3):63, 1992. results of computer-assisted ambulatory 16-channel
9. Gilliam F, Kuzniecky R, Faught E. Ambulatory EEG EEG. Electroencephalogr Clin Neurophysiol 91:
monitoring. J Clin Neurophysiol 16:111-115, 1999. 229-231, 1994.
10. Ives JR, Woods JF. A study of 100 patients with fo- 18. Morris GL. The clinical utility of computer-assisted
cal epilepsy using a 4-channel ambulatory cassette ambulatory 16 channel EEG. J Med Eng Technol
recorder, ISAM 1979: Proceedings of the Third In- 21:47-52, 1997.
ternational Symposium on Ambulatory Monitoring, 19. Liporace J, Tatum WT, Morris GL III, French J.
1979. Academic Press, London. Clinical utility of sleep-deprived versus computer-
11. AminoffMJ, Goodin DS, Berg BO, Compton MN. assisted ambulatory 16<hannel EEG in epilepsy pa-
Ambulatory EEG recordings in epileptic and non- tients: a multi-center study. Epilepsy Res 32:357-
epileptic children. Neurology 38:558-562, 1988. 362, 1998.
12. Kerr SL, Shucard DW, Kohrman MH, Cohen ME. 20. Tuunainen A, Nousiainen U. Ictal recordings of
Sequential use of standard and ambulatory EEG in ambulatory cassette EEG with sphenoidal elec-
neonatal seizures. Pediatr Neurol 6:159-162, 1990. trodes in temporal lobe epilepsy: comparison with
13. de Feo MR, Mecarelli O, Ricci G, Rina MF. The util- intensive videomonitoring. Acta Neurol Scand
ity of ambulatory EEG monitoring in typical ab- 88:21-25, 1993.
sence seizures. Brain Dev 13:223-227, 1991. 21. Schomer DL, IvesJR, Schachter SC. The role of am-
14. Plouin P, Dulac O, Jalin C, Chiron C. Twenty-four- bulatory EEG in the evaluation of patients for
hour ambulatory EEG monitoring in infantile epilepsy surgery. J Clin Neurophysiol 16:116-129,
spasms. Epilepsia 34:686-691, 1993. 1999.
Chapter 11
PROLONGED VIDEO
Cheolsu Shin
EQUIPMENT Classification of Seizure Type

CLINICAL APPLICATION Prolonged Video Electroencephalography
Epileptic Versus Nonepileptic Events and Surgical Evaluation
Psychogenic Seizures or Nonepileptic SUMMARY
Behavioral Events
Prolonged video electroencephalography (PVEEG) ing out psychosocial issues that may con-
has become an essential tool in the evalua- tribute to the patient's condition.
tion of patients with epilepsy and other Prolonged video electroencephalography
paroxysmal or episodic events.1"3 Often, the can be used to differentiate epileptic from
clinical history alone is not adequate for nonepileptic events, and it can help classify
making an accurate diagnosis, pardy because seizure type. Prolonged video electroen-
of inadequate observer history—die patients cephalography is also important in localiz-
frequendy are unaware and have to report a ing die epileptogenic region in patients be-
second-hand history that usually is sketchy ing considered for epilepsy surgery. With
at best. Even widi activation procedures modifications of this technique, neonates,
such as sleep, sleep deprivation, hyperventi- infants and children, patients in intensive
lation, and photic stimulation, epileptiform care units, and patients with other episodic
abnormalities may not appear on routine events such as sleep disorders can be evalu-
electroencephalographic (EEG) recordings, ated. This chapter describes various tech-
which may last an hour at most. Accurate di- niques of PVEEG and discusses its clinical
agnosis often requires capturing the clinical applications.
spell in conjunction with simultaneous elec- In 1949, Hunter and Jasper,4 then Schwab
trophysiologic and behavioral monitoring. et al.,5 and Stewart et al.6 described systems
Prolonged video electroencephalography al- for cinematographic and EEG recordings.
lows behavioral correlation of the patient's The early split-screen systems used mirrors
clinical spells widi die EEG. The most ex- to record simultaneously die patient's be-
perienced examiners may miss subtle or havioral manifestations and the EEG. They
even obvious clinical manifestations of sei- were bulky and could record only 1 hour of
zures when they view the event only once. cinema before the film reel had to be
Capturing the seizure on video allows re- changed, making die systems quite imprac-
peated review of die event and comparison tical. The development of the videotape
widi subsequent events. Prolonged video player and more compact camera units led
electroencephalography provides physicians to greater acceptance of the procedure.
and allied health staff an extended period Even today, a simple prolonged EEG can be
of observation that is often helpful in search- obtained using a video camcorder and a
129
standard EEC. An obvious disadvantage of With rapidly advancing computer and net-
this technique is that paper must be printed working technology, both in hardware and
for the entire monitoring session. Currently, software, most systems now use digital tech-
video EEG units with digitally encoded EEG nology at least for EEG signal processing.
stored on a computer system or the video- Electroencephalographic data can be stored
tape itself are commercially available from on optical disks or other computer storage
many sources. Most systems use telemetered devices. The cost of these digital PVEEG sys-
EEG with either cable or radio telemetry and tems can be high, but because of their su-
remote control video cameras, allowing rel- perior recording, data processing, and anal-
atively free movement of patients in the ysis capabilities, they are gaining wider
monitoring unit. acceptance as an industry standard. Digital
PVEEG allows on-line processing of EEG ac-
tivity, with automatic detection of seizures
EQUIPMENT and interictal epileptiform activity.9'10 Many
software programs are available for this on-
The variability among the PVEEG systems line detection, but all of them require veri-
used in laboratories throughout the world is fication off-line because movement artifacts
considerable because of the different man- or rhythmic or sharp sleep transients can
ufacturers.7 Each manufacturer uses custom- trigger the detection paradigm.11'12 Mon-
designed hardware and software to encode tage reformatting, filtering, frequency anal-
and decode the EEG signal during acquisi- ysis, timing analysis, and correlation studies
tion and analysis. This makes it difficult to can be performed off-line with these systems.
exchange the raw EEG data between differ- The ability to analyze a single seizure care-
ent systems. Miniaturization of electronics fully in many ways and to view it with several
and computer equipment allows PVEEG on different montages has decreased the num-
an outpatient basis because the patient and ber of actual seizures that need to be
family can easily transport the system. This recorded. Because of the risks involved in
can be used for patients with frequent events recording seizures and in tapering seizure
that may not require detailed testing by med- medications and the expense of the proce-
ical personnel.8 However, in most patients, dure, it is important to obtain as much in-
PVEEG is performed in an inpatient epilepsy formation from as few seizures as possible.
unit where trained medical personnel pro- Video signals are also being digitized for
vide continuous surveillance along with network transmission as well as for digital
video and EEG monitoring. storage. This enables remote monitoring of
Currently, several systems are available patients, for example, from an intensive care
commercially for PVEEG. Most of them use unit, and the monitoring includes both
a VHS video system with either multiplexed video and EEG. However, because video sig-
EEG recorded on one of the audio channels nals are such large data sets, a larger and
or a system that records the EEG digitally on faster capacity network system is necessary
the videotape or a computer with time syn- for on-line access and the quality of the video
chronization. Infrared cameras are used at may be inferior to that of an analog video
night to achieve reasonable quality video system. These problems are likely to be over-
recording even in a darkened room. The pa- come with rapidly advancing technology.
tient may be connected to the recording Portable PVEEG systems can also be used
equipment with a long EEG cable or the in intensive care units, neonatal units, and
EEG data may be sent through a radio trans- psychiatric facilities. These portable systems
mitter. With cable telemetry, patients have can be helpful in capturing events that are
limited mobility. However, the technical too infrequent to be seen on routine EEG
quality of cable or hardwired systems is usu- but do not need on-line monitoring. The
ally superior to that of radiotelemetered sys- systems can include a video camera and
tems. With radiotelemetry, patients have recorder synchronized to the EEG. The data
greater mobility, which may or may not be are analyzed off-line later, as in PVEEG mon-
an advantage given the circumstances of a itoring of outpatients.
patient's seizures, seizure frequency, and Most of these systems are designed so that
severity. continuous video EEG monitoring can pro-
Prolonged Video Electroencephalography 131
ceed without the continuous attendance of Various movement disorders can be con-
technical personnel. However, most inpa- fused with seizures. Simultaneous surface
tient PVEEG monitoring units have techni- electromyographic (EMG) recording of ag-
cal personnel dedicated to monitoring sev- onist and antagonist muscles with continu-
eral patients continuously around the clock. ous EEG recording can be helpful in diag-
The personnel perform essential functions nosing movement disorders and ruling out
such as testing and examining patients dur- an epileptic cause.
ing and after the seizure or spell and oper- Sleep disorders and seizures may be con-
ating remote control cameras to focus or fused with one another. Prolonged video
zoom in on the patient, and they ensure high electroencephalography allows behavior to
quality EEG recordings by repairing faulty be correlated with the EEG and other vari-
electrodes or connections. Also, the techni- ables such as respiration, eye movements,
cal personnel can recognize subtle or sub- surface EMG, EGG, and oxygen saturation.
clinical electrographic seizures and test the Prolonged video electroencephalography
patient during the seizure and alert the treat- may also help differentiate transient isch-
ment team to possible status epilepticus emic attacks, hypoglycemic attacks, and mi-
when overt clinical activity may not be obvi- graine attacks from seizures. Blood glucose
ous. Competent technical personnel are es- monitoring can be performed at the time of
sential for conducting specialized tests such the attacks to rule out hypoglycemia.
as ictal single-photon emission computed to- Prolonged video electroencephalography
mography (SPECT), in which it is vital to rec- can be especially helpful in differentiating
ognize the seizure rapidly, either by behav- types of childhood spells. Daydreaming,
ior or by EEG. breathholding, migraines, night terrors, and
other parasomnias can be confused with
seizures. Occasionally, nocturnal epileptic
CLINICAL APPLICATION seizures are misdiagnosed clinically as phys-
iologic parasomnias. Often, outpatient mon-
Epileptic Versus Nonepileptic itoring under the supervision of parents is
Events adequate for classifying these types of
spells.14
Prolonged video electroencephalography of- Repetitive motor activities are common in
ten can help answer the fundamental ques- patients with altered mental status in inten-
tion of whether a spell is an epileptic event.13 sive care units. Many of these movements are
Many different paroxysmal events can mimic manifestations of underlying cerebral or
seizures clinically. Distinguishing epileptic spinal lesions or reflections of toxic and
and nonepileptic events is critical for de- metabolic derangement. However, some of
termining effective therapy, because anti- these movements may represent epileptic
seizure medications are rarely beneficial for events, possibly subtle status epilepticus,
conditions other than seizures and have po- which would explain the altered mental sta-
tentially significant risks. tus. Because untreated generalized status
Recurrent episodes of loss of conscious- epilepticus can cause permanent neuronal
ness can result from various nonepileptic damage, this possibility needs to be kept in
causes, including syncopal attacks from aor- mind and pursued for appropriate diagno-
tic stenosis, cardiac arrhythmias, vasovagal sis and treatment with antiseizure medica-
depression, orthostatic hypotension, and hy- tions. Prolonged video electroencephalog-
perventilation. Vertigo and nonspecific dizzy raphy with a portable system or on-line
spells may also be difficult to differentiate remote monitoring can be helpful in diag-
from seizures clinically. Prolonged video nosing these movements.
electroencephalography with simultaneous
electrocardiography (EGG) and blood pres-
sure monitoring may help to delineate the Psychogenic Seizures or
nonepileptic nature of these events. Rarely, Nonepileptic Behavioral Events
cardiac arrhythmias can be triggered by an
epileptic discharge; this can be documented Psychogenic seizures are a difficult diagnos-
by PVEEG. tic problem.15'16 Accurate diagnosis of psy-
chogenic seizures can lead to appropriate seizures. Muscle and movement artifacts are
psychiatric treatment and discontinuation usually prominent, but there is no ictal EEG
of unnecessary and potentially dangerous change. After the movement stops, the nor-
medical therapy. Prolonged video elec- mal background EEG usually returns imme-
troencephalography may be advantageous diately. Postictal slowing and suppression of
economically in the diagnosis and treatment the EEG, which typically are present after a
of psychogenic seizures, because patients true generalized tonic-clonic seizure, do not
with such a condition consume a significant occur after a psychogenic seizure. A caveat
amount of medical resources, including the is that the patient needs to be tested during
costs of repeated visits to the emergency the ictal events to assess the responsiveness
department.17 In most cases, psychogenic and memory processing. If there is no alter-
seizures are not under conscious control. ation of consciousness, the possibility of a
Factitious disorders, such as Munchausen's simple partial seizure cannot be ruled out.
syndrome or actual willful malingering, are Surface EEG changes may not be seen with
rare. Gates and colleagues18 compared the the simple partial seizures, and this may be
clinical manifestations of psychogenic gen- true of the majority of cases.21
eralized tonic-clonk events with those of The timing of the spells is also helpful in-
true generalized tonic-clonic seizures and formation because psychogenic seizures are
found that out-of-phase clonic movements of more likely to occur during the day. If they
the extremities, pelvic thrusting, lack of eye occur at night, it is during wakefulness. Al-
manifestations, side-to-side head move- though some patients may claim that the
ments, and early vocalizations were more spells occur out of sleep, they invariably wake
common in the psychogenic group. Psy- up first (as shown by the change in EEG pat-
chogenic seizures also tend to last longer tern from sleep to wakefulness) and then
than 2 minutes; patients with psychogenic have the spell.22.Epileptic seizures during
seizures often have long attacks, with fre- sleep tend to occur directly out of sleep,
quent rest periods during attacks. Also, these without intervening wakefulness. Laboratory
patients often respond to some degree to studies on changes in the serum levels of
verbal or noxious stimulation during the prolactin or neuron-specific enolase can cor-
generalized movements. However, caution is roborate the PVEEG studies.23'24 Epileptic
needed in making a clinical judgment on the generalized tonic-clonic seizures or com-
basis of the history alone, because many of plex partial seizures—but not psychogenic
these ictal features can occur in epileptic seizures—are associated with a significant in-
seizures, especially ones with an extratem- crease in the levels of these substances.
poral focus.19 Induction or provocation of psychogenic
Although a spectrum of psychiatric diag- seizures is a matter of controversy.25 Certain
noses are represented in patients with psy- triggering situations can be used if the his-
chogenic seizures and as many as one-third torical information about the reliable trig-
of them have a history of sexual or physical gering factors is clear. Some clinicians have
abuse or assault, some may not have any eas- used saline injections or a tuning fork with
ily identifiable psychiatric or psychologic a strong suggestion that a seizure will occur.
problem. In addition, these psychiatric and However, to interpret the results, it has to be
psychologic factors are also associated with verified that the induced spell is the same
patients with epilepsy. To complicate mat- type as the noninduced spell.
ters further, some epileptic patients also
have psychogenic seizures, and in some se-
ries, 10%-40% of patients with psychogenic
seizures also had epileptic seizures.20 There- Classification of Seizure Type
fore, the correlation of each type of ictal
event with the EEG is essential to exclude Prolonged video electroencephalography
epilepsy as a cause. may lead to a reclassification of the seizure
Prolonged video electroencephalography type in many patients with uncontrolled
allows careful analysis of behavioral and EEG seizures, especially if the diagnosis is in ques-
manifestations in patients with psychogenic tion, thus improving medical management.
Distinguishing primary generalized from mimic absence seizures and, at times, have
secondary generalized seizures is often diffi- been referred to as pseudoabsences. Frontal
cult only on the basis of the clinical history lobe seizures often begin with low-amplitude
and routine EEG. Some patients have rapid fast activity but can be associated with some
secondary spread and may have an interic- frontal sharp waves or spikes. Prolonged
tal EEG that shows generalized spike and video electroencephalography, especially with
wave discharges (secondary bilateral syn- computer-assisted recordings, can often
chrony) . However, ictal recording may dem- help identify these frontal lobe seizures.
onstrate that seizure onset is focal. Other pa- Inpatient PVEEG can be helpful in dis-
tients with true absence or true generalized tinguishing various spells and multiple sei-
tonic-clonic seizures may mistakenly be zure types in patients with Lennox-Gastaut
thought to have focal epilepsy, because of syndrome.31 Patients with this syndrome can
the asymmetrical manifestation of general- have different types of epileptic seizures as
ized discharges. By clinical history alone, ab- well as stereotyped mannerisms, tics, and
sence seizures may be indistinguishable other movements that are not epileptic. Be-
from complex partial seizures. In virtually all cause these phenomena tend to occur al-
patients with untreated absence seizures, hy- most daily in this patient population, only a
perventilation will activate 3-Hz spike-and- relatively short session of inpatient PVEEG
wave discharges. The diagnosis may be more monitoring is needed. After the various
difficult to make in patients with infrequent spells and seizures have been classified, out-
spells or those taking medication and may patient management is simpler because
benefit from PVEEG. Medications can be ta- antiepileptic drug adjustments are not nec-
pered when the patient is in the hospital and essary for recurrent nonepileptic spells.
being carefully observed. Through education and reassurance of par-
Monitoring can also help to differentiate ents and caretakers, emergency department
temporal from extratemporal seizures.26"28 visits or unnecessary rectal administration of
Many patients with simple partial seizures diazepam can be avoided.
have no EEG accompaniment. However,
most patients with complex partial seizures
have an ictal EEG change. Temporal lobe
seizures often begin with an attenuation of Prolonged Video
scalp activity, followed by a rhythmic dis- Electroencephalography and
charge, usually in the theta range that in- Surgical Evaluation
creases in amplitude and becomes more
widespread. Postictally, a focal slowing often Surgical treatment frequently eliminates or
occurs over the temporal region where the decreases the frequency and severity of
seizure began. Tachycardia is observed in seizures in many partial epilepsies.32'33
most patients with complex partial seizures Seizures arising from the temporal lobe are
of temporal lobe origin, even before any sig- especially amenable to surgical treatment.34
nificant motor activity is apparent. Lateral- A comprehensive evaluation is performed
ized ictal posturing of the upper extremity on patients who are being considered for
contralateral to the ictal onset is observed epilepsy surgery.35 The most important part
with many seizures of temporal onset.29 Also, of the presurgical evaluation probably is the
there is forced turning of the head away recording of the patient's typical seizures on
from the ictal focus just before secondary continuous EEG with video monitoring.36
generalization.30 This serves two purposes at the outset: first,
Frontal lobe seizures tend to be shorter to establish that the refractory habitual
and to cause less postictal confusion than seizures are indeed epileptic and not a
temporal lobe seizures. Frontal lobe seizures nonepileptic behavioral spell and, second, to
are associated with frequent falls, because of establish the localization of the epileptic fo-
the rapid bilateral spread. Focal tonic, or cus electrophysiologically. Because seizures
fencing, postures may be seen and may sug- occur sporadically and unpredictably, treat-
gest a focus in the supplementary motor ment with antiepileptic medications is usu-
area. Certain frontal lobe seizures may ally withdrawn rather rapidly to expedite the
recording of seizures.37 Some concern exists visual phenomena can be observed or re-
about whether seizures recorded during ported by the patient. For the language area,
acute withdrawal of medication faithfully stimulation causes the arrest of language
represent the patient's habitual seizure pat- function, such as a pause in reading a sen-
tern. An additional risk is the possibility of tence aloud or naming objects. Electroen-
a secondarily generalized tonic-clonic sei- cephalographic recording must be per-
zure in patients who had only complex par- formed to verify that the stimulation did not
tial seizures and the possibility of a general- cause an afterdischarge (electrographic sei-
ized tonic-clonic or complex partial status zure), because that would suggest a propa-
epilepticus. The PVEEG monitoring unit gated effect away from die stimulated elec-
should be well equipped and the personnel trodes. This can be done in the PVEEG
should be expertly trained in the manage- monitoring unit outside the operating room
ment of diese neurologic emergencies. Ready and obviate intraoperative cortical mapping
intravenous access must be maintained by with the patient awake.
way of a heparin-lock system, and intra- In evaluating medically refractory partial
venous formulations of benzodiazepines epilepsy, PVEEG from the scalp electrodes
(lorazepam or diazepam) or fosphenytoin alone is not always adequate for localizing
should be immediately available. Patients seizure onset. Subtraction ictal-interictal
and families need to be well informed about SPECT coregistered on magnetic resonance
these issues before or at the time of admis- images (SISCOM) can be useful for either
sion to the monitoring unit. avoiding further invasive intracranial moni-
Prolonged video electroencephalography toring or guiding the placement of in-
allows careful analysis of clinical and EEG tracranial electrodes.38'39 The ictal SPECT
changes and can be used with scalp, sphe- scan needs to be performed with exquisite
noidal, foramen ovale, or implanted depth temporal urgency. The sooner the tracer is
or subdural strip or grid electrodes. With injected after the onset of the seizure (usu-
computer-assisted recordings, montage really within tens of seconds), the more likely
formatting, and off-time analysis of seizures, the scan will reflect the foci of ictal increase
there is less need for invasive EEG record- in blood flow. Therefore, it is critical to mon-
ings. For surgical monitoring, digital video itor the patient closely for signs of seizure
systems with 64-128 channels are commonly onset, either by behavior or by EEG. This is
used. For focal cortical resection, typical possible only in the setting of inpatient
seizures must be shown to arise from a sin- PVEEG with appropriately trained monitor-
gle region of the brain. An adequate num- ing personnel and ready access to and avail-
ber of seizures must be recorded so that the ability of the radioactive tracer from the nu-
seizure can be localized confidendy. If the clear medicine department.
patient has multiple seizure types, then all
of them need to be captured and analyzed
to determine whether the seizures have mul- SUMMARY
tifocal onset or a single focus with different
propagation paths. Prolonged video electroencephalography
When intracranial electrodes are im- has become an essential diagnostic tool in
planted and the identified epileptogenic the management of epileptic seizures, psy-
zone is near a functional cortical region (for chogenic seizures, and other paroxysmal
example, motor, sensory, language, or visual events. Various monitoring systems are avail-
cortex), the determination of cortical func- able, with the dominant trend for digital sys-
tion is feasible. A train of electric pulses (usu- tems. Prolonged video electroencephalogra-
ally biphasic rectangular pulses at 60 Hz phy monitoring allows psychogenic seizures
from a stimulation isolation unit) can be ap- and other nonepileptic paroxysmal events to
plied through pairs of electrodes through be differentiated from epileptic seizures. It
the special interface device coupled to the also enables accurate classification of most
PVEEG input stage. During the delivery of seizure types and is a necessary part of the
the stimulation, positive motor, sensory, or evaluation of patients being considered for
surgical treatment of epilepsy. Also, this type 14. Connolly MB, Wong PK, Karim Y, Smith S, Farrell
of monitoring allows extended observation K. Outpatient video-EEG monitoring in children.
Epilepsia 35:477-481, 1994.
of patients and can lead to a better under- 15. Krumholz A. Nonepileptic seizures: diagnosis and
standing of their medical problems and any management. Neurology 53:S76-S83, 1999.
superimposed psychosocial issues. All these 16. Scott DF. Recognition and diagnostic aspects of
factors contribute to improved quality of life nonepileptic seizures. In Riley TL, Roy A (eds).
Pseudoseizures. Williams & Wilkins, Baltimore,
for patients and their families and more ef- 1982, pp 21-33.
ficient delivery of appropriate medical care 17. Martin RC, Gilliam FG, Kilgore M, Faught E,
in these often very complex situations. Kuzniecky R. Improved health care resource uti-
lization following video-EEG-confirmed diagnosis
of nonepileptic psychogenic seizures. Seizure
7:385-390, 1998.
18. Gates JR, Ramani V, Whalen S, Loewenson R. Ictal
REFERENCES characteristics of pseudoseizures. Arch Neurol
42:1183-1187, 1985.
1. EngelJJr, BurchfielJ, Ebersole J, et al. Long-term 19. Geyer JD, Payne TA, Drury I. The value of pelvic
monitoring for epilepsy. Report of an IFCN com- thrusting in the diagnosis of seizures and pseudo-
mittee. Electroencephalogr Clin Neurophysiol 87: seizures. Neurology 54:227-229, 2000.
437-458, 1993. 20. Ramani SV, Quesney LF, Olson D, Gumnit RJ. Di-
2. Kaplan PW, Lesser RP. Long-term monitoring. In agnosis of hysterical seizures in epileptic patients.
Daly DD, Pedley TA (eds). Current Practice of Clin- Am J Psychiatry 137:705-709, 1980.
ical Electroencephalography, 2nd ed. Raven Press, 21. Bare MA, Burnstine TH, Fisher RS, Lesser RP. Elec-
New York, 1990, pp 513-534. troencephalographic changes during simple partial
3. Mizrahi EM. Pediatric electroencephalographic seizures. Epilepsia 35:715-720, 1994.
video monitoring. J Clin Neurophysiol 16:100-110, 22. Bazil CW, Walczak TS. Effects of sleep and sleep
1999. stage on epileptic and nonepileptic seizures. Epilep-
4. Hunter J, Jasper HH. A method of analysis of sei- sia 38:56-62, 1997.
zure pattern and electroencephalogram: a cine- 23. Rabinowicz AL, Correale J, Boutros RB, Couldwell
matographic technique. Electroencephalogr Clin WT, Henderson CW, DeGiorgio CM. Neuron-
Neurophysiol 1:113-114, 1949. specific enolase is increased after single seizures
5. Schwab RS, Schwab MW, Withee D, Chock YC. Syn- during inpatient video/EEC monitoring. Epilepsia
chronized moving pictures of patient and EEG. 37:122-125, 1996.
Electroencephalogr Clin Neurophysiol 6:684-686, 24. Tumani H, Otto M, Gefeller O, et al. Kinetics of
1954. serum neuron-specific enolase and prolactin in pa-
6. Stewart LF, Jasper HH, Hodge C. Another simple tients after single epileptic seizures. Epilepsia 40:
method for simultaneous cinematographic record- 713-718, 1999.
ing of the patient and his electroencephalogram 25. Dericioglu N, Saygi S, Ciger A. The value of provo-
during seizures. Electroencephalogr Clin Neuro- cation methods in patients suspected of having
physiol 8:688-691, 1956. nonepileptic seizures. Seizure 8:152-156, 1999.
7. Lagerlund TD, Cascino GD, Cicora KM, Shar- 26. Marks WJJr, Laxer KD. Semiology of temporal lobe
brough FW. Long-term electroencephalographic seizures: value in lateralizing the seizure focus.
monitoring for diagnosis and management of Epilepsia 39:721-726, 1998.
seizures. Mayo Clin Proc 71:1000-1006, 1996. 27. Williamson PD, Thadani VM, French JA, et al. Me-
8. Gilliam F, Kuzniecky R, Faught E. Ambulatory EEG dial temporal lobe epilepsy: videotape analysis of
monitoring. J Clin Neurophysiol 16:111-115, 1999. objective clinical seizure characteristics. Epilepsia
9. Gotman J. Automatic recognition of epileptic 39:1182-1188, 1998.
seizures in the EEG. Electroencephalogr Clin Neu- 28. Geyer JD, Payne TA, Faught E, Drury I. Postictal
rophysiol 54:530-540, 1982. nose-rubbing in the diagnosis, lateralization, and lo-
10. Gotman J. Seizure recognition and analysis. Elec- calization of seizures. Neurology 52:743-745, 1999.
troencephalogr Clin Neurophysiol Suppl 37:133- 29. Kotagal P, Bleasel A, Geller E, Kankirawatana P,
145, 1985. Moorjani BI, Rybicki L. Lateralizing value of asym-
11. Salinsky MC. A practical analysis of computer based metric tonic limb posturing observed in secondar-
seizure detection during continuous video-EEC ily generalized tonic-clonic seizures. Epilepsia
monitoring. Electroencephalogr Clin Neurophysiol 41:457-462, 2000.
103:445-449, 1997. 30. Theodore WH, Porter RJ, Albert P, et al. The sec-
12. SpattJ, Pelzl G, Mamoli B. Reliability of automatic ondarily generalized tonic-clonic seizure: a video-
and visual analysis of interictal spikes in lateralising tape analysis. Neurology 44:1403-1407, 1994.
an epileptic focus during video-EEG monitoring. 31. Bare MA, Glauser TA, Strawsburg RH. Need for
Electroencephalogr Clin Neurophysiol 103:421- electroencephalogram video confirmation of atyp-
425, 1997. ical absence seizures in children with Lennox-Gas-
13. Mohan KK, Markand ON, Salanova V. Diagnostic taut syndrome. J Child Neurol 13:498-500, 1998.
utility of video EEG monitoring in paroxysmal 32. EngelJ (ed). Surgical Treatment of the Epilepsies,
events. Acta Neurol Scand 94:320-325, 1996. 2nd ed. Raven Press, New York, 1993.
33. Penry JK, Porter RJ. Intensive monitoring of pa- proves patient lateralization in temporal lobe
tients with intractable seizures. In Penry JK (ed). epilepsy. Epilepsia 41:1567-1573, 2000.
Epilepsy: The 8th International Symposium. Raven 37. Quiroga RC, Pirra L, Podesta C, Leiguarda RC, Ra-
Press, New York, 1977, pp 95-101. binowicz AL. Time distribution of epileptic seizures
34. Kilpatrick C, Cook M, Kaye A, Murphy M, Matkovic during video-EEG monitoring. Implications for
Z. Non-invasive investigations successfully select pa- health insurance systems in developing countries.
tients for temporal lobe surgery. J Neurol Neuro- Seizure 6:475-477, 1997.
surg Psychiatry 63:327-333, 1997. 38. O'Brien TJ, So EL, Mullan BP, et al. Subtraction peri-
35. Sutula TP, Sackellares JC, Miller JQ, Dreifuss FE. ictal SPECT is predictive of extratemporal epilepsy
Intensive monitoring in refractory epilepsy. Neu- surgery outcome. Neurology 55:1668-1677, 2000.
rology 31:243-247, 1981. 39. So EL, O'Brien TJ, Brinkmann BH, Mullan BP. The
36. Series W, Caramanos Z, Lindinger G, Pataraia E, EEG evaluation of single photon emission com-
Baumgartner C. Combining ictal surface-elec- puted tomography abnormalities in epilepsy. J Clin
troencephalography and seizure semiology im- Neurophysiol 17:10-28, 2000.
Chapter 12
ELECTROENCEPHALOGRAPHIC
SPECIAL STUDIES
QUANTITATIVE METHODS OF Interpolation Techniques

ELECTROENCEPHALOGRAPHIC Topographic Displays (Mapping)
ANALYSIS Multivariate Statistical Methods of
Fourier (Spectral) Analysis Topographic Analysis
Spike, Sharp-Wave, and Seizure Detection Cortical Projection Techniques
Montage Reformatting Source Dipole Localization
Cross-Correlation Analysis MAGNETOENCEPHALOGRAPHY
Cross-Spectral Analysis SUMMARY
QUANTITATIVE METHODS OF in normal and disease states, including de-

ELECTROENCEPHALOGRAPHIC mentia, cerebral infarction, cerebral neo-
plasms, and various toxic and metabolic dis-
ANALYSIS orders. Most abnormal conditions increase
the amount of slow-wave activity in an EEG
Digital computers can aid in extracting in- and may decrease the amount of faster fre-
formation from electroencephalographic quency activity. Quantitative measures that
(EEG) waveforms that is not readily obtain- may be used include (1) the percentage
able with visual analysis alone. These com- power in the delta, theta, or delta + theta
puters may also be used for quantification of bands; (2) the percentage power in the al-
key features of waveforms, which may be use- pha, beta, or alpha + beta bands; (5) various
ful in accurate EEG interpretation and also other power ratios (for example, alpha to
in making serial comparisons between EEGs delta); (4) the mean frequency in the entire
performed on the same subject at different spectrum or in a portion of the spectrum,
times or between two subject groups in sci- such as the alpha band; and (5) the spectral
entific investigations. Digital computers may edge frequency, which is often taken to be
also partially automate the interpretation of the frequency below which 95% of the power
EEGs, particularly in prolonged monitoring in the entire spectrum occurs. Spectral anal-
for epilepsy. This chapter discusses some of ysis has also been used in intraoperative
the uses of computers in EEG. monitoring to assess changes in depth of
anesthesia or cerebral ischemia. Although
spectral analysis per se can be applied only
Fourier (Spectral) Analysis to one channel at a time, useful comparisons
can be made between the spectra of record-
The technique of spectral analysis is de- ings in different regions of the head. As ex-
scribed in Chapter 4. Electroencephalo- pected, focal brain lesions tend to produce
graphic spectral analysis has been used to as- focal changes in power spectra, whereas dif-
sess the dominant background frequencies fuse processes produce generalized changes.
137
Quantitative measures of asymmetry can be toring systems usually do not rely exclusively
calculated, such as the left-to-right ratio of on computer detection of seizures but also
power in the entire spectrum or in individ- make use of trained observers or patient and
ual frequency bands, such as alpha, delta, family members to recognize and log seizure
and so forth. Some studies of power spectra occurrence. The specificity of currently used
during transient cerebral events, such as algorithms is also fairly good; however, in
epileptic seizures, have also been conducted. many prolonged monitoring situations, a
The power spectra of a partial complex sei- large variety of artifacts arising from patient
zure of temporal lobe origin, for example, activities and various waveforms of cerebral
tend to be complex, with several frequency origin, for example, V waves in sleep, cause
components (often a fundamental with one false detections that may outnumber true
or more harmonics, that is, integral multi- epileptic discharges by an order of magni-
ples of the fundamental frequency) whose tude for some patients. Thus, a physician or
amplitudes in various regions of the head technician must review all discharges de-
may differ. The spectra also evolve over time tected by the computer, but the ability to
during the course of the seizure. markedly decrease the amount of data to be
reviewed still makes automated detection al-
gorithms of great practical value.
Spike, Sharp-Wave, and
Seizure Detection
Montage Reformatting
The detection of interictal and ictal epilep-
tic discharges in an EEG by computer algo- Montage reformatting allows EEG montages
rithms has received much attention because to be selected when the data are reviewed,
of the increasing use of prolonged EEG independently of the montage used to ac-
monitoring for epilepsy. In prolonged EEG quire and to store the data. The same EEG
monitoring, many EEG channels (21 or segment can be viewed using a variety of dif-
more in most cases) are recorded over many ferent montages. To generate a derivation
hours, days, or weeks, generating huge such as Xi — X%, which does not exist in the
amounts of data that must be reviewed to as- EEG as recorded, the computer looks for two
sess the frequency, nature, and localization existing channels, one of which records X\
of epileptic discharges. Computer algo- against a reference electrode and another
rithms may be used to locate candidate in- which records X% against the same reference,
terictal or ictal discharges for further study and then subtracts the two. That is, X\ —
by an electroencephalographer and, thus, X2 = (Xi - R) - (X2 - R), where R is the
the need to scan many pages of EEG tracings reference. This allows new referential and
manually may be avoided. The techniques of bipolar montages to be formed.1 For exam-
pattern recognition are discussed in Chapter ple, suppose that recorded EEG data include
4. Spike and sharp-wave detection generally the following channels:
relies on "sharpness" criteria applied to in-
dividual waveforms on a channel-by-channel Channel Derivation
basis, although more advanced algorithms 1 Fpl-Cz
may use context information and multi-
2 Fp2-Cz
channel correlation to improve specificity.
Seizure detection algorithms are more com- 3 Fs-Cz
plex, because of the great variability in types 4 F4-CZ
of ictal discharges, and generally are based 5 CS-GZ
on detecting a sudden change in rhythmic- 6 C4~CZ
ity and amplitude of background activity si- V Ps-Cz
multaneously in several channels. 8 P4-CZ
Current state-of-the-art commercially 9 Oi-C2
available software has a reasonably good 10 Og-Cz
overall sensitivity, although for the seizures 11 A!-CZ
of certain patients the sensitivity may be in- 12 A2-CZ
adequate. Consequently, prolonged moni-
Electroencephalographic Special Studies 139
Then, a new referential montage with ipsi- be calculated as follows: O\ — Ai/g = (O\ —
lateral ear reference can be created by sub- Cz) - 0.5 (A! - Cz) - 0.5 (A2 - C z ). Other
tracting pairs of channels as follows: types of montages, such as a common aver-
age reference montage (in which the refer-
Channels Derivation ence for each electrode is the average po-
tential of all recorded electrodes) or a
1-11 Fpj-Ai laplacian (source) montage2'3 (in which the
2-12 Fp2-A2 reference for each electrode is the average
3-11 FS-A! of the four nearest neighbors to that elec-
4-12 F4-A2 trode) , may also be generated easily by the
5-11 CS-AI computer. The same ictal discharge viewed
6-12 C4-A2 with various montages (one as recorded,
7-11 PS-A! three after reformatting) is shown in Figure
12-1.
8-12 P4-A2
9-11 Oi-Ai
10-12 O2-A2
Cross-Correlation Analysis
New reference electrodes can also be cre- Cross-correlation analysis quantifies the rela-
ated by averaging data from two or more ex- tionship between EEG signals recorded from
isting electrodes. For example, suppose an different derivations. The cross-correlation func-
EEG is recorded with a Cz reference. A new tion is the correlation between EEG signal
derivation such as O\ — A\/%, where Ai/g rep- number 1 and EEG signal number 2 delayed
resents the average of the ear electrodes, can by time t, expressed as a function of t. If EEG
Figure 12-1. An electroencephalogram recorded during a seizure. A, As recorded from a C3/C4 average reference
(AVon figure) that is active. B, Reformatted to average ear reference (Al/2 in figure). C, Reformatted to lon-
gitudinal bipolar montage. D, Reformatted to longitudinal laplacian montage (Nav, 4-neighbor average; N3,
3-neighbor average). (From Lagerlund.1 By permission of Lippincott Williams & Wilkins.)
form of the cross-correlation function. From

the cross-power spectrum, one can calculate
two additional functions:
1. Coherence function—a normalized func-
tion that takes on values between 0 and
1. It is defined as the absolute square
of the cross-power spectrum of signals
numbers 1 and 2, divided by the prod-
uct of the power spectrum of signal
number 1 and that of signal number 2.
It is a convenient measure of the de-
gree of correlation between similar fre-
quency components in the two signals
(0%-100%), expressed as a function of
the frequency.
2. Phase spectrum—the relative phase an-
gle (0°-360°) between corresponding
frequency components of signal num-
ber 1 and those of signal number 2; it
is expressed as a function of the fre-
quency.
The phase angle, in turn, is related to tim-
ing differences between the two signals. Be-
cause the phase angle between two sine or
Figure 12-2. A, Autocorrelation functions and, B, cross- cosine waves delayed by a fixed time Tequals
correlation functions for a 1-second epoch of alpha ac- 360° multiplied by the product of the time
tivity recorded using a C^/C^ average reference (AVon
figure). The autocorrelations demonstrate the rhyth- delay T and the frequency / the time delay
micity of the signal but cannot be used to assess inter- between two signals can be calculated from
channel time differences. The cross-correlations dem- the phase spectrum (f> by the formula4
onstrate a systematically increasing time delay between
channels as one goes from right, Te, to left, T5, across
the head.
signal number 2 is identical to number 1, Thus, if a plot of phase angle vs. frequency
the cross-correlation function is the same as is linear over some range of frequencies, the
the autocorrelation function described in time difference between the signals can be
Chapter 4. If signal number 2 is similar to derived by taking the slope of the plot di-
number 1 but delayed by time T, then the vided by 360° (Fig. 12-3).
cross-correlation function is a maximum at
t= T, and periodically thereafter if both sig-
nals are periodic. Thus, examination of Interpolation Techniques
cross-correlation functions may be used as a
means to assess small interchannel time dif- Interpolation techniques are used to esti-
ferences (Fig. 12-2); one possible use of this mate the electric potential at intermediate
is in determining whether a bilateral epilep- scalp positions from its known value at each
tic discharge in fact represents secondary electrode position. This procedure is neces-
bilateral synchrony or is generalized from sary when constructing maps of scalp po-
onset. tential, and it may form a preliminary step
in the application of various techniques of
spatial, or topographic, analysis of EEG, such
Cross-Spectral Analysis as calculating accurate estimates of the lapla-
cian of the scalp potential, multivariate sta-
The cross-fower spectrum between EEG signal tistical analyses, cortical projection tech-
number 1 and number 2 is the Fourier trans- niques, and source dipole localization. Some
Figure 12-3. A, Onset of a seizure discharge recorded with a bipolar montage from intracerebral electrodes in a
patient with epilepsy; A 3-5 are in amygdala, B 3-5 are in hippocampus. B, Power spectra of each channel (left) and
phase angle and coherence spectra (right); the upper coherence graph is the coherence itself, the lower graph is
the lower boundary of the 99% confidence interval of the coherence. For the range of frequencies indicated by
the arrows, the phase is linearly related to frequency with slope 4.7°/Hz, corresponding to a delay of 13 ms between
channels. Positive slope indicates that the first channel is leading. (From Gotman.4 By permission of Elsevier Sci-
entific Publishers.)
interpolation methods that are commonly Topographic Displays (Mapping)

used include the following:
1. Nearest neighbor inverse distance Electroencephalographic topographic, or
weighted.5'6 spatial, maps are a way of displaying EEG
2. All-electrode inverse distance weighted.7 data that differ significantly from the con-
3. Rectangular two-dimensional splines, ventional multichannel amplitude-vs.-time
which conceptually minimize the plots (montages). However, topographic
"bending energy" of an infinite, elastic mapping is not in itself a method of EEG
plate constrained to pass through analysis, because it only displays the "raw"
known points.8'9 data in a different way. In its simplest form,
4. Rectangular three-dimensional splines, a topographic map of scalp potential is a
which interpolate potentials in diree "snapshot" of the EEG at one instant in time;
dimensions, after which results may be it shows the distribution of potentials on the
projected as needed onto spherical or head surface at that time and may facilitate
ellipsoidal surfaces.10'11 localization of EEG abnormalities (while
5. Spherical surface splines, which are ap- making it more difficult to appreciate their
plied to a spherical surface instead of variation in time). Because potentials are ac-
a rectangular one.5'8'12 tually measured at only a few points on the
6. Spherical harmonic expansion, the head (that is, at the electrode locations), an
equivalent of a Fourier series in spher- interpolation technique must be used to es-
ical instead of rectangular coordi- timate potentials at all other scalp points;6
nates.13'14 thus, the information conveyed by a topo-
7. Single or multidipole source models, graphic display is only as accurate as the in-
which localize source dipoles and then terpolation technique used (Fig. 12-4). Var-
predict the scalp potential these sources ious methods of topographic display include
would generate. three-dimensional plots (potential on z axis
Note that methods 1-3 are based on a rec- vs. x and y coordinates), contour plots (con-
tangular planar model of the scalp surface, necting all points on the head with die same
and methods 5 and 6 assume a spherical value of potential), gray scale intensity plots
head and should be more accurate. Only (degree of darkness at each point on a head
method 1 (the simplest and least accurate) map corresponds to the potential at that
has been used widely in commercial EEG point), and color plots (color at each point
systems. on a head map corresponds to the potential
Hgune I2~4, Example of topographic map construction for visual evoked potential signals recorded referentially
from 20 scalp electrodes. Each evoked potential is divided into 128 intervals of 4 ms. A, Individual evoked poten-
tials for the indicated electrode locations. B, Mean voltage values at each electrode location for the 4-ms interval
beginning 192 ms after the stimulus (corresponding to the vertical line in A), C, The grid of interpolation points
(64 X 64) used; each of the 4096 points is assigned a voltage value by linear interpolation from the three nearest
known points. D, The topographic map of this evoked potential, using an equal interval intensity scale to represent
voltage at each location. (From Duffy et al.6 By permission of the American Neurological Association )
at that point). As a general rule, topographic Multivariate Statistical Methods

maps should be used only as an adjunct to
ordinary time series EEC displays and not as of Topographic Analysis
a replacement for them, because 75% or The purpose of multivariate statistical meth-
more of the information derived from an ods of EEG topographic analysis is to achieve
EEC depends on the temporal rather than data reduction from many simultaneously
the spatial characteristics of waveforms. recorded EEG signals. This is possible be-
In addition to maps of unprocessed EEG, cause of the redundancy of multichannel
topographic mapping has been used to dis- EEG data—many activities or waveforms ap-
play the spatial distribution of the results of pear simultaneously in many different chan-
various techniques of multichannel EEG nels. Mathematical techniques such as factor
analysis. For example, maps of power spec- analysis, principal component analysis, or
tra—power within a specific frequency band eigenvector analysis are used to reduce the
at various scalp locations—can be generated. observed EEG signals in multiple channels
Similarly, maps of the iaplacian of the scalp to a minimum number of independent, or
potential, of principal components of the orthogonal, component signals. Individual
scalp potential (from multivariate statistical components may be displayed spatially as
analysis), or of estimated cortical potentials topographic maps or temporally as derived
(from a cortical projection technique) can EEG channels. Although these methods may
also be produced. provide data reduction, their major draw-
back is that the resultant independent sig- dian nerve somatosensory evoked potentials
nals are not always recognizable as tradi- and interictal epileptic discharges on the cor-
tional pure EEG activities such as alpha or tical surface, with confirmation by recordings
mu, and comparison between analyses made from subdural electrode grids.
on the same EEG at different times or on
different subjects is difficult. Also, these
methods generally do not provide informa- Source Dipole Localization
tion on the nature and location of physio-
logic generators of EEG. However, principal The ultimate goal of localization of abnor-
component analysis may be used to create a mal activity, such as epileptic discharges,
type of spatial filter that can aid in the re- from EEG is to find the intracranial sources
moval of certain types of artifact, such as oc- generating a given distribution of scalp po-
ular movement and electrocardiographic ar- tentials. This is sometimes called the inverse
tifact, from an EEG recording.15 problem (the forward problem refers to finding
the distribution of scalp potentials resulting
from a known distribution of intracranial
Cortical Projection Techniques sources). Although the forward problem has
a unique solution, the inverse problem does
Cortical projection techniques such as spatial not; that is, there are an infinite number of
deconvolution16 and deblurring17 are de- different sets of intracranial generators that
signed to reverse the smearing effect of the could produce any given distribution of
skull on scalp EEG by using a model of vol- scalp activity. To constrain the problem, cer-
ume conduction in the head (such as a three- tain physiologically based assumptions must
sphere model of brain, skull, and scalp or an be made about the number and approxi-
anatomically based boundary element or fi- mate location of generators. Most ap-
nite element model constructed from mag- proaches to solving the inverse problem
netic resonance scans of the patient's head). have concentrated on finding the location,
The electric potential at selected points on orientation, and strength of a single dipole gen-
the brain surface is calculated from the elec- erator whose potential field best matches ac-
tric potential at the scalp surface, thus non- tual data. This is done with a least-squares
invasively providing a distribution of electri- minimization algorithm, which varies the di-
cal activity at the cortical surface. Cortical pole coordinates and direction to minimize
potentials may be displayed as a time series the sum of squares of the differences be-
(montage format) or as a topographic display tween the predicted and actual potentials at
(map format). This technique has proved ca- each electrode location on the head.18-19
pable of resolving two adjacent dipole sources The assumption of a single dipole generator
in the cerebral cortex that could not be reis most useful for small generators, such as
solved by inspection of the scalp EEG signals the generators of certain evoked potential
and has been used successfully to localize me- peaks or of some epileptic spikes (Fig. 12-5).
figure 12-5. Dipole modelling of a spike discharge recorded with a sternoclavicular reference from a patient with
epilepsy. A, Map of measured spike discharge distribution. B, Map of distribution of potential based on fitted di-
pole. C, Fitted dipole located in right frontocentral region (long arrow), with short arrow indicating orientation of
dipole negativity. (From Thickbroom et al.19 By permission of Elsevier Scientific Publishers.)
More recently, distributed source models The major advantage of MEG is in source
such as low resolution electromagnetic to- localization. The accuracy of localization of
mography (LORETA) have been proposed intracranial sources by MEG is not limited
and have been used to estimate the location by the smearing effects of the volume-con-
of generators of some EEG waveforms.20 ducting medium, especially the poorly con-
ducting skull, on electric potentials, as oc-
curs in EEG, because all the tissues between
MAGNETOENCEPHALOGRAPHY the sources and the magnetic field detectors
are transparent to magnetic fields. Thus,
Magnetoencephalography (MEG) is the re- when a dipole localization algorithm is used
cording of the small magnetic fields pro- with MEG data, a simple homogeneous
duced by the electric activity of neurons in sphere model of the volume conductor is
the brain. These magnetic fields are gener- usually sufficient to obtain accurate local-
ated by current flowing in neurons, with a ization of the source dipole.21 In addition to
small contribution from extracellular cur- recording the spontaneous MEG, one can
rent flow in the volume conducting medium record evoked magnetic fields in response
around the brain (generally less than the to visual, auditory, and somatosensory stim-
contribution of intracellular currents). uli, and these may also be submitted to di-
These magnetic fields are extremely small, pole localization algorithms to determine
typically in the femptotesla or picotesla the location of visual, auditory, and so-
range (10~15 to 10~12 T). They must be de- matosensory cortical areas. This may be used
tected by a magnetic gradiometer connected as part of the surgical planning process in
to a special type of extremely sensitive am- patients with tumors or vascular malforma-
plifier called a superconducting quantum in- tions, in whom the sensory cortical areas may
terference device (SQUID), which must be be significantly displaced from their usual
cooled by liquid helium. To eliminate noise or expected location. Magnetoencephalog-
signals caused by the much larger magnetic raphy has been used for localization of an
fields associated with electrical equipment, epileptic spike focus before performing
power lines, and the earth's magnetic field, resective surgery for intractable partial
a special magnetically shielded room is re- epilepsy.22
quired. For all of these reasons, MEG is a
very expensive tool. Another disadvantage of
MEG, compared with EEG, is that it cannot
be used readily for the long-term recordings SUMMARY
needed to capture and to localize an epilep-
tic seizure, because the subject's head must This chapter reviews several quantitative
be kept immobilized near the magnetic gra- analysis techniques that may be applied to
diometer array during the entire recording. digitized EEG data. The technique of mag-
Until recently, the number of channels avail- netoencephalography is also discussed. Cur-
able in commercial MEG instruments was rently, many of these techniques are used
relatively small, although some systems now primarily as research tools, but as they be-
available have more than 100 channels; the come more widely available, they probably
spatial resolution of these devices is quite will have an increasing effect on EEG inter-
good. Because magnetic fields created by a pretation and diagnosis.
current source are always oriented along a
tangent to a circle around the line of cur-
rent flow, MEG is insensitive to radially ori- REFERENCES
ented currents in cerebral cortex and is sen-
sitive only to tangential currents, in contrast 1. Lagerlund TD. Montage reformatting and digital
to EEG, which is sensitive to both (although filtering. In Luders H (ed). Epilepsy Surgery. Raven
more sensitive to radial than to tangential Press, New York, 1991, pp 318-322.
2. Hjorth B. An on-line transformation of EEG scalp
currents). Thus, in practice, MEG record- potentials into orthogonal source derivations. Elec-
ings are often combined with simultaneous troencephalogr Clin Neurophysiol 39:526-530,
conventional EEG recordings. 1975.
3. Wallin G, Stalberg E. Source derivation in clinical computation of laplacian of scalp potential. Elec-
routine EEG. Electroencephalogr Clin Neurophys- troencephalogr Clin Neurophysiol 95:178-188,
iol 50:282-292, 1980. 1995.
4. Gotman J. Measurement of small time differences 14. Pascual-Marqui RD, Gonzalez-Andino SL, Valdes-
between EEG channels: method and application to Sosa PA, Biscay-Lino R. Current source density es-
epileptic seizure propagation. Electroencephalogr timation and interpolation based on the spherical
Clin Neurophysiol 56:501-514, 1983. harmonic Fourier expansion. Int J Neurosci 43:
5. Babiloni F, Babiloni C, Fattorini L, Carducci F, 237-249, 1988.
Onorati P, Urbano A. Performances of surface 15. Lagerlund TD, Sharbrough FW, Busacker NE. Spa-
Laplacian estimators: a study of simulated and real tial filtering of multichannel electroencephalo-
scalp potential distributions. Brain Topogr 8:35—45, graphic recordings through principal component
1995. analysis by singular value decomposition. J Clin
6. Duffy FH, Burchfiel JL, Lombroso CT. Brain elec- Neurophysiol 14:73-82, 1997.
trical activity mapping (BEAM): a method for ex- 16. Nunez PL. Methods to estimate spatial properties
tending the clinical utility of EEG and evoked po- of dynamic cortical source activity. In Pfurtscheller
tential data. Ann Neurol 5:309-321, 1979. G, Lopes da Silva FH (eds). Functional Brain Imag-
7. Lemos MS, Fisch BJ. The weighted average refer- ing. Hans Huber Publishers, Toronto, 1988, pp 3-9.
ence montage. Electroencephalogr Clin Neuro- 17. Gevins A, Le J, Brickett P, Reutter B, Desmond J.
physiol 79:361-370, 1991. Seeing through the skull: advanced EEGs use MRIs
8. Nunez PL. Estimation of large scale neocortical to accurately measure cortical activity from the
source activity with EEG surface Laplacians. Brain scalp. Brain Topogr 4:125-131, 1991.
Topogr 2:141-154, 1989. 18. Salu Y, Cohen LG, Rose D, Sato S, Kufta C, Hallett
9. Perrin F, Pernier J, Bertrand O, Giard MH, Echal- M. An improved method for localizing electric
lierJF. Mapping of scalp potentials by surface spline brain dipoles. IEEE Trans Biomed Eng 37:699-705,
interpolation. Electroencephalogr Clin Neurophys- 1990.
iol 66:75-81, 1987. 19. Thickbroom GW, Davies HD, Carroll WM,
10. Law SK, Nunez PL, Wijesinghe RS. High-resolution Mastaglia FL. Averaging, spatio-temporal mapping
EEG using spline generated surface Laplacians on and dipole modelling of focal epileptic spikes.
spherical and ellipsoidal surfaces. IEEE Trans Bio- Electroencephalogr Clin Neurophysiol 64:274-277,
med Eng 40:145-153, 1993. 1986.
11. Srinivasan R, Nunez PL, Tucker DM, Silberstein RB, 20. Pascual-Marqui RD, Michel CM, Lehmann D. Low
Cadusch PJ. Spatial sampling and filtering of EEG resolution electromagnetic tomography: a new
with spline laplacians to estimate cortical potentials. method for localizing electrical activity in the brain.
Brain Topogr 8:355-366, 1996. Int J Psychophysiol 18:49-65, 1994.
12. Perrin F, Pernier J, Bertrand O, Echallier JF. Spher- 21. Hari R. Comment: MEG in the study of epilepsy.
ical splines for scalp potential and current density Acta Neurol Scand Suppl 152:89-90, 1994.
mapping. Electroencephalogr Clin Neurophysiol 22. Minassian BA, Otsubo H, Weiss S, Elliott I, Rutka
72:184-187, 1989. JT, Snead OC III. Magnetoencephalographic local-
13. Lagerlund TD, Sharbrough FW, Busacker NE, ization in pediatric epilepsy surgery: comparison
Cicora KM. Interelectrode coherences from near- with invasive intracranial electroencephalography.
est-neighbor and spherical harmonic expansion Ann Neurol 46:627-633, 1999.
Chapter 13
ELECTROENCEPHALOGRAPHIC
RECORDINGS FOR
EPILEPSY SURGERY
Gregory D. Cascino
CANDIDATES FOR EPILEPSY SURGERY Prognostic Importance of Chronic

EXTRACRANIAL Intracranial Electroencephalographic
ELECTROENCEPHALOGRAPHY IN Monitoring
PARTIAL EPILEPSY ELECTROCORTICOGRAPHY IN EPILEPSY
CHRONIC INTRACRANIAL MONITORING SURGERY
Depth Electrode Studies SUMMARY
Subdural Electrode Studies
Partial epilepsy, also called focal or localization- disorders.3'4 An estimated 75,000 patients in
related epilepsy, is the most common seizure the United States are potential candidates
disorder encountered.1 Partial seizure activ- for epilepsy surgery.1 Successful surgical
ity occurs in more than 90% of the incident therapy may render the person seizure-free
cases of epilepsy in adults. Approximately and produce substantial improvement in
80% of all partial seizures emanate from the cognitive, social, and behavioral func-
anterior temporal lobe, and most extratem- tions.3"6 The most common surgical proce-
poral seizures are of frontal lobe origin.1 An dure performed for intractable partial
estimated 1 million or more people in the epilepsy is focal cortical resection, or cor-
United States have partial epilepsy, and ticectomy, of epileptic brain tissue in the tem-
nearly 45% of them have seizures that are poral lobe.1'3'4'7 Long-term outcome stud-
refractory to antiepileptic drug (AED) med- ies have shown that approximately 60% of
ication.1 A seizure disorder is considered in- patients are rendered seizure-free after an-
tractable when the seizures are refractory to terior temporal lobe surgery.4 Nearly 90% of
AED therapy and the patient's quality of life highly favorable candidates may have an ex-
is impaired.1'2 The disabling seizure types as- cellent operative outcome.3'7'15"17
sociated with partial epilepsy include com- Preoperative electroencephalography (EEG)
plex partial seizures and secondarily gener- and electrocorticography (ECoG) help to lo-
alized tonic-clonic seizures. calize the epileptogenic zone, that is, the site
Surgical therapy is highly effective and of seizure onset and initial seizure propaga-
well tolerated in selected patients with in- tion.1'7'8'11'15'16'18"28 This chapter describes
tractable partial epilepsy and is an alterna- the clinical applications of long-term EEG
tive to AED medication in patients with phys- monitoring and ECoG in selecting surgical
ically and socially disabling partial seizure candidates and in tailoring focal cortical
146
Electroencephalographic Recordings for Epilepsy Surgery 147
resection in patients with intractable partial anterior temporal lobectomy,3'4 which typi-
epilepsy. cally involves lateral temporal neocortical
excision and amygdalohippocampectomy.3'4
The underlying lesion is perhaps the most
CANDIDATES FOR EPILEPSY prognostically important preoperative factor
SURGERY in estimating the probability of surgical
success.3'4'27 Magnetic resonance imaging
Several factors that are considered in select- (MRI) may serve as an in vivo surrogate for
ing candidates for epilepsy surgery include pathology in patients being considered for
the adverse effect of the seizures on the pa- surgical treatment.3'7'15'16'27'29'30 Finally, the
tient's quality of life, seizure type and site of purpose of epilepsy surgery is not only to ren-
seizure onset, the likely pathologic substrate der the patient seizure-free but also to allow
of the epileptogenic zone, and the capabil- the patient to become a participating and
ity of the patient to sustain substantial psy- productive member of society.6 The poten-
chosocial rehabilitation after successful sur- tial for vocational rehabilitation may be a fac-
gical treatment.1-5'7-9'10'14'17'29-32 Patients tor in determining the appropriate medical
with medically refractory partial seizure dis- or surgical management for patients with in-
orders who are disabled by their seizures or tractable partial epilepsy.6
AED toxicity, or both, may be potential can- A comprehensive presurgical evaluation is
didates for epilepsy surgery.2'5'9'17'33 performed to select appropriate surgical
The effect of intractable partial epilepsy candidates.1'3'5'7-13'15-30'^'35"37 Preoperative
on a patient's lifestyle is an individual de- testing includes neuropsychologic studies,
termination that may vary considerably speech and language evaluation, visual
among surgical candidates. Common issues perimetry, and MRI of the head.1'7'9'15-17
that are often considered in motivating the Magnetic resonance imaging is a reliable in-
patient to proceed with a presurgical evalu- dicator of the site of the epileptic brain tis-
ation include the desire to drive a motor ve- sue in patients with a structural intra-axial
hicle, to become gainfully employed, and to abnormality.3'7'15"17'27'29 The epileptogenic
live independently.6 The type of partial sei- pathologic substrates commonly identified
zure and the localization of the epilepto- in patients being considered for epilepsy sur-
genic zone also need to be considered. Pa- gery include mesial temporal sclerosis, pri-
tients with auras or simplex partial seizures mary brain tumor, vascular malformation,
associated with a visceral or experiential and malformations of cortical develop-
phenomenon alone, for example, abdomi- ment.3'7'15"17'27'29 Magnetic resonance imag-
nal discomfort, may not be appropriate can- ing is also of predictive value for operative
didates for epilepsy surgery because of the outcome.3'7'15"17'27 An MRI-identified struc-
limited adverse effect of this type of seizure tural lesion that is concordant with the
on a patient's quality of life. The risk of sur- epileptogenic zone is a strong predictor of
gical treatment in these patients must be an excellent operative outcome. Medial tem-
weighed carefully against any potential poral lobe epilepsy and lesional epilepsy are sur-
benefit.34 gically remediable epileptic syndromes, because
Most surgical procedures are performed the seizures associated with these disorders
for complex partial seizures. Secondarily are almost invariably medically refractory
generalized tonic-clonic seizures may be as- and remit with surgical treatment.1'10'17 Se-
sociated with a significant risk of physical lected patients may require additional diag-
trauma, but they are more responsive to nostic studies to determine surgical candi-
AED medication than are complex partial dacy. These procedures include positron
seizures.1 The localization of the epilepto- emission tomography (PET) and single-
genic zone is an important determinant of photon emission computed tomography
the efficacy of the operative procedure as (SPECT).30'37
well as the likelihood of surgical morbidity.9 The major contraindication to epilepsy
The most effective and "safest" surgical pro- surgery is tile presence of important medical
cedure for intractable partial epilepsy is problems or psychiatric disease that does not
allow safe completion of the presurgical eval- dromes, for example, lesional epilepsy, early
uation or performance of the operation.1 A in the course of medical treatment.
coexistent psychiatric illness such as depres-
sion or anxiety is not necessarily a reason for
excluding a surgical candidate. Factors that
are not of major importance in determining EXTRACRANIAL
surgical candidacy include age at operation ELECTROENCEPHALOGRAPHY
and duration of epilepsy.3 Although the ma- IN PARTIAL EPILEPSY
jority of patients who have surgical treatment
are young adults or children, epilepsy sur- Electroencephalography is the most fre-
gery has been shown to be safe and effective quently performed neurodiagnostic study in
in selected older patients.3 The duration of the evaluation of patients with intractable
a seizure disorder is less important in se- partial epilepsy."-13'19'21'23-25-27'28'36'38-40
lecting surgical candidates than the specific These studies usually are performed be-
medical therapy administered. Most of the tween seizure episodes—interictally—be-
persons considered have had a medically cause of the episodic nature of the disorder
refractory seizure disorder for at least 6 (Fig. 13-1).19'21'38 Interictal EEG recordings
months to 1 year before undergoing presur- in patients with epilepsy should be per-
gical evaluation. Epilepsy surgery can be formed according to the methods estab-
considered in highly favorable candidates lished by the American EEG Society.21'38
with surgically remediable epileptic syn- Standard activation procedures, for exam-
Figure 13-1. Scalp-recorded interictal EEG during sleep. Bitemporal independent spike discharges maximal at the
T7 and T8 electrode positions are evident.
pie, hyperventilation and photic stimulation, charges) during brief recording periods, de-
should be included.21'38 Sleep deprivation spite serial EEG studies.23'24'38 Epileptiform
and the recording of nonrapid eye move- activity generated in cortical areas remote
ment sleep may be used to increase the from the scalp electrodes, for example, the
sensitivity of the EEG to demonstrate inter- amygdala and hippocampus, may not be as-
ictal epileptiform alterations.21'38'40 The sociated with interictal extracranial EEG al-
EEG technologist should obtain information terations.23 Attenuation of spike activity by
about the type(s) of seizures, the timing of the dura mater, bone, and scalp limits the
the last seizure, and the identification of sensitivity of extracranial EEG recordings.38
seizure-precipitating events, for example, Approximately 20% to 70% of cortical spikes
reading epilepsy. Current AED medication are recorded on a scalp EEG. Patients with
and recent drug levels should also be noted. intractable partial epilepsy may have repeti-
Extracranial EEG has several limitations tively normal interictal EEG studies.23'24'38
that must be recognized for correct inter- Most surgical epilepsy centers perform ic-
pretation of these studies in potential surgi- tal EEG recordings, that is, during seizure
cal candidates.11'13'22'24'25'27'32'39 Interictal activity, before considering focal corticec-
scalp-recorded EEG studies in patients with tomy, because of the limitations of interictal
partial epilepsy may fail to demonstrate spe- EEG studies1'8-10'19 (Fig. 13-2). Ictal ex-
cific epileptiform activity (for example, tracranial EEG, however, may also provide
spike, sharp-wave, or spike-and-wave dis- inaccurate information about the localiza-
Figure 13-2. A, Awake scalp-recorded EEG showing normal background before seizure. B, Initial ictal EEG pattern
during an aura involves a subtle discharge in the right temporal lobe region with preservation of alpha activity on
the left. C, Approximately 1 minute after the recording in B, there is a right temporal lobe seizure discharge asso-
ciated with a complex partial seizure. Note prominent slowing on the left, (continued)
tion of the epileptogenic zone.23"25'27 Scalp- The sensitivity and specificity of extracra-
recorded EEC studies may fail to detect spe- nial EEG studies may increase with the use
cific alterations localized to the site of sei- of supplementary electrodes. Sphenoidal
zure onset, for example, the amygdala, only and inferior lateral temporal scalp (Tl, T2,
to reveal distant, more widespread cortical F9, F10) electrodes and closely placed scalp
excitability, such as the frontotemporal cor- electrodes may be useful to delineate the
tex. The sensitivity and specificity of ictal ex- topography of interictal and ictal activ-
tracranial EEG in partial epilepsy depend on ity.23"26 Sphenoidal electrodes may record
the seizure type(s) and the localization of epileptiform activity emanating from the
the epileptogenic zone.23"25 Most patients mesiobasal limbic region and assist in local-
have a focal or generalized scalp-recorded izing the epileptogenic zone before anterior
EEG alteration during complex partial temporal lobectomy is performed.25 Results
seizures.26 However, simple partial seizures are conflicting about the sensitivity of sphe-
may not be associated with an EEG change.39 noidal electrodes compared with that of
Ictal extracranial EEG recordings are also scalp electrodes in patients with temporal
more sensitive and specific in patients with lobe epilepsy.26 Nasopharyngeal electrodes
seizures of temporal lobe origin.23"25 The are artifact-prone, are poorly tolerated by
scalp-recorded EEG demonstrates a local- patients (and, thus, may interfere with the
ized epileptiform abnormality in only a small sleep recording), and have not been dem-
proportion of patients with frontal lobe onstrated to be more specific or more sen-
epilepsy.23"25 sitive than inferior lateral scalp electrodes.36
Computer-assisted EEG monitoring may be ings with implanted electrodes can be used
used for automatic seizure recognition and with EEG telemetry and video monitoring
off-line seizure analysis. over several days to several weeks to evaluate
spontaneous EEG activity.8'11-13'19'20'22'24'27'28
Before being considered for intracranial
recordings, patients should be shown to have
CHRONIC INTRACRANIAL a medically refractory seizure disorder and
MONITORING to be appropriate surgical candidates. Iden-
tifying an epileptogenic lesion with neu-
Chronic intracranial EEG monitoring (CIM) is roimaging may obviate CIM in some pa-
a generic term for invasive EEG recordings tients. Intracranial electrodes must be
that are used preoperatively to evaluate pa- inserted with the patient under general anes-
tients with intractable partial epilepsy.27 thesia. The risk of hemorrhage or infection
These studies are performed after scalp- associated with CIM is approximately 3%.27
recorded EEG studies alone have been de- Several different methods are available for
termined to be inadequate for localizing the CIM studies.9'27 The techniques used most
epileptogenic ZOne.8>11'12'22'24'27'28 Chronic frequently are depth electrode and multi-
intracranial EEG monitoring studies are electrode grid recordings. The results of the
conducted only in selected patients before presurgical evaluation determine the ap-
focal cortical resection. Prolonged record- propriate intracranial technique and the
Table 13-1. Indications for Chronic Intracranial Electroencephalographic

Monitoring*
Indication Subdural Strips Subdural Grid Depth Electrodes
Lateralization+
Localization*
Mapping^
*0, not used; +, limited role; ++, useful technique; + ++, primary indication.
+Lateralizing the epileptogenic zone, for example, right vs. left temporal lobe epilepsy.
^Localizing (not lateralizing) the epileptogenic zone, for example, temporal lobe vs. frontal lobe epilepsy.
Îdentify the functional cortical anatomy before resection, for example, multilobar neocortical epilepsy.
anatomical location of the electrodes. The studies confirmed that the sensitivity and
common indications for performing CIM specificity of these recordings were in-
are summarized in Table 13—1. creased compared with those of extracranial
EEC monitoring13 (Fig. 13-3). Depth elec-
trodes are used most commonly to deter-
Depth Electrode Studies mine the lateralization of seizure onset in
patients with scalp-recorded bitemporal
The first form of CIM used in the evaluation seizures.11'12 Depth electrodes can be im-
of patients with intractable partial epilepsy planted stereotaxically with MRI and PET
was depth electrode recordings.7'8'13 Early guidance.27'37 Common artifacts identified
Figure 13-3. Depth electrode recording shows right anterior temporal lobe spike discharges (R1-R8). Scalp-
recorded EEG is unremarkable. AV, C3-C4 reference.
with extracranial recordings, such as muscle form discharges do not necessarily preclude
activity during seizures, do not occur with a successful surgical outcome after anterior
depth electrode studies. In patients with temporal lobectomy.11"13'27'28
temporal lobe epilepsy, depth electrode Ictal EEG is the most important factor
studies are performed with electrodes im- used to determine the site of seizure onset
planted in the mesiobasal limbic region.11"13 with depth electrodes27 (Fig. 13-4). The pat-
Montages are used that contain surface der- tern of seizure onset recorded with depth
ivations, for example, scalp or multielec- electrodes may have an important predictive
trode grid electrodes and recordings from value in determining the response to focal
the limbic region. corticectomy. A regional pattern of seizure
Interpretation of depth electrode studies onset indicates a less favorable seizure out-
involves examination of interictal and ictal come.8 Focal seizure onset with an electro-
epileptiform discharges and assessment of graphic alteration at a single electrode con-
background EEG activity. Depth-recorded tact implies more precise localization of
interictal epileptiform activity may not be re- the site of seizure onset.8 Initial ictal EEG
liable for determining the region of seizure changes may include attenuation of back-
onset. In 10%-20% of patients, interictal ground EEG activity, with the development
spiking with depth electrodes may be more of a low-amplitude, high-frequency rhythmi-
prominent contralateral to the epilepto- cal discharge. Spread to other electrode
genic zone. Bitemporal interictal epilepti- contacts may then occur, with the develop-
Figure 13-4. A right depth electrode (RTD) and subdural electrode (RG) recording showing a right anterior tem-
poral lobe seizure (RTD 1 and 3) characterized by a high-frequency discharge. A subdural grid was placed over the
right temporoparietal region. HF, 35 Hz; LF, 1 Hz.
ment of clinical symptoms. In most patients, agreement between scalp-recorded EEG and
several (5-10) spontaneous seizures are depth localization has varied from 50% to
recorded with depth electrodes before a de- 100%.8'22 Studies that combined the diag-
cision is made about surgical candidacy.19 nostic accuracy of extracranial EEG with
The propagation of seizure activity to the op- tests of focal functional deficit (PET or neu-
posite cerebral hemisphere may also have ropsychologic studies or both) have shown
important prognostic value in determining that the results of depth electrode and scalp
the response of focal corticectomy.20 A recordings are discordant in only 13% of pa-
shorter interhemispheric conduction time tients. Patients with favorable surgical out-
may be associated with a less favorable sur- comes may be more likely to have an agree-
gical outcome after anterior temporal lobec- ment between scalp EEG and depth
tomy.20 Background activity is not as valu- electrode recordings.11"13'26 The predictive
able as ictal EEG in determining the value of depth electrode recordings has also
localization of the epileptiform zone. Com- been compared with extracranial EEG
mon nonepileptiform alterations in the re- recordings.22 A statistically significant differ-
gion of the epileptogenic zone include at- ence has not been found in surgical out-
tenuation of EEG activity and focal slowing.8 come between patients monitored with the
The diagnostic yield of depth electrode two techniques.9'22 In selected patients, ex-
studies has been assessed in patients with intracranial EEG, performed as part of a com-
tractable partial epilepsy.8'22 The reported prehensive presurgical evaluation, may be
Rgure 13-5. A, Subdural grid (RG) and bitemporal depth electrode (RD and LD) recording showing a neocorti-
cal seizure. HF, 35 Hz; LF, 1 Hz. B, The amplitude of the seizure discharge gradually increases. HF, 35 Hz; LF, 1
Hz. C, Note the progressive ictal pattern, with subsequent propagation to the right temporal lobe. HF, 35 Hz; LF,
1 Hz. (continued)
adequate without using CIM to localize the onset is already known through presurgical
epileptogenic zone.27 Depth electrode stud- evaluation.9 Subdural grid recordings may
ies, however, may demonstrate that a patient show the extent of neocortical involvement
is an appropriate surgical candidate when and provide information about the precise
the scalp recording provided nonlocalizing localization of seizure onset. The grids con-
information or showed bitemporal epilepti- tain multiple, evenly spaced electrode con-
form abnormalities.11"13 tacts in a plastic insulating material.8'9 The
electrodes can be placed either under the
temporal lobe to record from the inferior
Subdural Electrode Studies temporal region or in the interhemispheric
fissure to record electrographic activity of
Multielectrode grid recordings are a later mesiofrontal origin. Multielectrode grids
development in CIM.8 Multielectrode grids may not be as sensitive as depth electrodes
are used to evaluate patients with neocorti- for detecting epileptiform abnormalities em-
cal epilepsy of temporal or extratemporal anating from the mesiobasal temporal lobe.8
origin (Fig. 13-5). Typically, subdural elec- Extraoperative cortical stimulation studies
trodes are placed over the cortical surface may be performed with multielectrode grids
and used to localize the epileptogenic zone to determine the localization of functional
and to delineate functional cortical anat- cortex.35 These studies may be useful in pa-
omy. These studies are most useful in patients who cannot cooperate with intraoper-
tients in whom the lateralization of seizure ative cortical stimulation. Intraoperative

studies are limited in duration and may be intractable partial epilepsy.15 The presence
difficult to perform in some patients, such of MRI-identified atrophy of the hippocam-
as children. Extraoperative cortical stimula- pal formation and a prolonged interhemi-
tion during intracranial monitoring may be spheric propagation time correlated with an
used to delineate the limits of primary cor- excellent operative outcome.15 Schiller and
tical areas that may directly affect the extent colleagues27 examined the surgical outcome
of cortical resection.35 of 108 consecutive patients at Mayo Clinic
who had CIM and epilepsy surgery, and the
main predictive indicator of an excellent op-
Prognostic Importance of erative outcome was the presence of an MRI-
identified lesion or hippocampal formation
Chronic Intracranial atrophy that was concordant with the epilep-
Electroencephalographic togenic zone. Approximately 80% of pa-
Monitoring tients with abnormal MRI findings were sei-
zure-free after surgical treatment. However,
The prognostic importance of depth elec- a small proportion of patients (22%) with a
trode and subdural electrode recordings in normal MRI study were seizure-free.27 The
patients having surgical treatment for in- electrographic ictal pattern and spatial ex-
tractable partial epilepsy has been evalu- tent of the initial ictal discharge were not
ated.15'27 The relationship of depth elec- predictive of long-term operative outcome.
trode studies to quantitative MRI and A multifocal independent initiation of the
operative outcome was assessed in 30 pa- seizure discharge was associated with a poor
tients undergoing temporal lobe surgery for operative outcome.
ELECTROCORTICOGRAPHY IN type (s) of extraoperative EEG monitoring.9

EPILEPSY SURGERY If the epileptogenic zone has been localized
with CIM, ECoG may be less important in
Intraoperative intracranial EEG recordings determining the region of cortical resec-
were used initially at several epilepsy centers tion.9 Patients in whom depth electrode
to guide surgical resection of epileptic brain studies show a mesiobasal limbic origin of
tissue.7'9 The authoritative work of Wilder seizures may subsequently have en bloc re-
Penfield and Herbert Jasper in Montreal and section of the temporal lobe without ECoG.
A. Earl Walker in Baltimore demonstrated Traditional focal corticectomies have used a
the importance of ECoG during focal corti- "tailored resection," with ECoG localizing
cal resective surgery for patients with in- the site of seizure onset.41
tractable partial epilepsy.1*'41 Penfield and Several techniques for ECoG are available
Jasper relied on predominantly interictal ex- that include the use of subdural strip elec-
tracranial EEG and ECoG to determine the trodes or a rigid electrode holder with
region of cortical resection, because long- graphite tip or cotton wick electrodes.32 Sub-
term EEG monitoring was still in its in- dural strip electrodes are particularly useful
fancy.41 The methods for ECoG established in recordings obtained from the inferior
by the investigators in Montreal and Balti- temporal lobe and the mesiofrontal re-
more continue to be used in the surgical gion.32 The preexcision ECoG at Mayo
treatment of epilepsy.14'41 Clinic uses three subdural strips (each with
The need for ECoG depends on the re- eight electrode contacts) placed on the su-
sults of the presurgical evaluation, the lo- perior and inferior temporal lobe gyri and
calization of the epileptogenic zone, and the in the suprasylvian region (Figs. 13-6 and
Figure 13-6. Electrocorticography performed at the time of left anterior temporal lobectomy. The three upper
channels represent recording from the mesiotemporal region with depth electrodes. Prominent spiking is noted
in the mesiotemporal region and the lateral temporal cortex.
Figure 13-7. Postexcision electrocorticography performed with a subdural strip placed posterior to the margin of
the resection. No definite residual spiking is noted.
13-7). Three depth electrodes are also ECoG and the presurgical evaluation. The
placed by hand in the region of the amyg- relationship between the localization of the
dala and hippocampus. A baseline extracra- epileptogenic zone and the eloquent cortex
nial EEC recording is performed before the is considered at the time of the operation.
surgical procedure with the patient awake, A postexcision recording is made, with a sub-
and a contralateral ear reference is used for dural strip electrode placed posterior to the
the intracranial recordings. Samples of EEG margin of the surgical resection (Fig. 13-6).
activity are obtained before and after corti- Electrocorticography has several impor-
cal resection. Electrocorticography may be tant limitations that must be recognized for
performed with the patient under light lo- appropriate interpretation of these studies.9
cal or general anesthesia. The traditional Information about the localization and lat-
procedure of the Montreal group was to ob- eralization of the epileptogenic zone must
tain the ECoG recordings in awake pa- be determined before the ECoG study.
tients.41 Electrical stimulation was then used Other potential disadvantages of ECoG in-
intraoperatively to assist in localizing the clude the sampling of predominant interic-
epileptogenic zone and functional anatomy. tal EEG activity, the restricted spatial distri-
At Mayo Clinic, ECoG is performed with bution of the EEG recording, and usually the
the patient under general anesthesia. The brief duration of the monitoring. Electro-
anesthetic agents used include nitrous ox- corticography has been used mainly to assess
ide, fentanyl, or a low-volume percentage of neocortical epileptiform activity, and it may
isoflurane. A 16- or 32-channel EEG record- be restricted in sampling from the or-
ing is made. Methohexital given intra- bitofrontal, mesiofrontal, and mesiotempo-
venously may be used to enhance or to acti- ral regions. Electrocorticography predomi-
vate epileptiform activity in the preexcision nantly records epileptiform activity from the
recordings. The strategy for intraoperative lateral surface of the temporal lobe, and this
EEG monitoring is discussed with the surgi- may not correlate with the region of seizure
cal team, the anesthesiologist, and the EEG onset, for example, the mesiobasal temporal
technician. The electroencephalographer is lobe.9 The neurosurgical team must delay
present in the operating room for immedi- the cortical resection until ECoG is per-
ate interpretation qf the EEG data. The ex- formed and the studies are interpreted. Gen-
tent of the focal corticectomy is based on the eral anesthesia used during the surgical pro-
cedure may suppress epileptiform activity. 2. Camfield PR, Camfield CS. Antiepileptic drug ther-
High concentrations of certain anesthetic apy: When is epilepsy truly intractable? Epilepsia 37
(Suppl 1):S60-S65, 1996.
drugs, such as isoflurane, may make it diffi- 3. Radhakrishnan K, So EL, Silbert PL, et al. Predic-
cult to document the neocortical extent of tors of outcome of anterior temporal lobectomy for
the epileptogenic zone. Enflurane may acti- intractable epilepsy: a multivariate study. Neurology
vate or increase interictal epileptiform ac- 51:465-471, 1998.
4. Walczak TS, Radtke RA, McNamaraJO, et al. An-
tivity; however, this may be nonspecific and terior temporal lobectomy for complex partial
may not correlate with the site of seizure seizures: evaluation, results, and long-term follow-
onset. The false positivity of methohexital- up in 100 cases. Neurology 40:413-418, 1990.
activated ECoG has also been observed. Cer- 5. National Institutes of Health Consensus Confer-
tain ECoG-recorded postexcision spike dis- ence. Surgery for epilepsy. JAMA 264:729-733,
1990.
charges are not prognostically important.9 6. Reeves AL, So EL, Evans RW, et al. Factors associ-
Evidence on the prognostic importance of ated with work outcome after anterior temporal
ECoG in determining long-term seizure con- lobectomy for intractable epilepsy. Epilepsia
trol has been conflicting.18 A recent study of 38:689-695, 1997.
7. Cascino GD, Trenerry MR, Sharbrough FW, So EL,
165 patients showed that the topography of Marsh WR, Strelow DC. Depth electrode studies in
the preexcision ECoG in patients with tem- temporal lobe epilepsy: relation to quantitative
poral lobe epilepsy was not predictive of op- magnetic resonance imaging and operative out-
erative outcome and the location, morphol- come. Epilepsia 36:230-235, 1995.
ogy, and distribution of spiking in the 8. Engel JJr, Crandall PH. Intensive neurodiagnostic
monitoring with intracranial electrodes. Adv Neu-
postexcision recording may not affect sei- rol 46:85-106, 1987.
zure outcome.15 The presence of residual 9. Engel J Jr, Ojemann GA. The next step. In Engel J
spiking may be associated statistically with Jr (ed). Surgical Treatment of the Epilepsies, 2nd
unfavorable seizure outcome. However, the ed. Raven Press, New York, 1993, pp 319-329.
10. Engel JJr, Shewmon DA. Overview: Who should be
presence of residual spiking may not pre- considered a surgical candidate? In EngleJJr (ed).
clude a successful surgical outcome after fo- Surgical Treatment of the Epilepsies, 2nd ed. Raven
cal corticectomy.18 Press, New York, 1993, pp 23-34.
11. So N, Gloor P, Quesney LF, Jones-Gotman M,
Olivier A, Andermann F. Depth electrode investi-
gations in patients with bitemporal epileptiform ab-
SUMMARY normalities. Ann Neurol 25:423-431, 1989.
12. So N, Olivier A, Andermann F, Gloor P, Quesney
LF. Results of surgical treatment in patients with
Extraoperative and intraoperative EEG stud- bitemporal epileptiform abnormalities. Ann Neu-
ies are only part of a comprehensive presur- rol 25:432-439, 1989.
gical evaluation of patients with intractable 13. Spencer SS, Spencer DD, Williamson PD, Mattson
RH. The localizing value of depth electroencephalog-
partial epilepsy. The level of experience of raphy in 32 patients with refractory epilepsy. Ann
an individual surgical epilepsy program is Neurol 12:248-253, 1982.
important in selecting the electrodiagnostic 14. Walker AE. Temporal lobectomy. J Neurosurg
methods used. Ultimately, the decision re- 26:642-649, 1967.
garding surgical candidacy depends on 15. Cascino GD, Trenerry MR, Jack CRJr, et al. Elec-
trocorticography and temporal lobe epilepsy: rela-
demonstrating a convergence of the results tionship to quantitative MRI and operative out-
of diagnostic studies in the presurgical eval- come. Epilepsia 36:692-696, 1995.
uation. The preoperative EEG evaluation 16. Cascino GD, Trenerry MR, So EL, et al. Routine
must be correlated with the rest of the EEG and temporal lobe epilepsy: relation to long-
term EEG monitoring, quantitative MRI, and op-
presurgical investigation, especially the neu- erative outcome. Epilepsia 37:651-656, 1996.
roimaging studies, in selecting surgical can- 17. Cascino GD. Neuroimaging in partial epilepsy:
didates and in predicting the likely outcome structural magnetic resonance imaging. J Epilepsy
of focal cortical resection. 11:121-129, 1998.
18. Fiol ME, Gates JR, Torres F, Maxwell RE. The prog-
nostic value of residual spikes in the postexcision
electrocorticogram after temporal lobectomy. Neu-
rology 41:512-516, 1991.
REFERENCES 19. Gates JR. Epilepsy presurgical evaluation in the era
of intensive neurodiagnostic monitoring. Adv Neu-
1. Dreifuss FE. Goals of surgery for epilepsy. In Engel rol 46:227-247, 1987.
J Jr (ed). Surgical Treatment of the Epilepsies. 20. LiebJP, Babb TL, Engel J Jr, Jann Brown W, Pre-
Raven Press, New York, 1987, pp 31-49. torius J, Crandall PH. Interhemispheric propaga-
tion time of hippocampal seizures: cell density and 31. Taylor DC. Epileptic experience, schizophrenia,
surgical outcome correlates (abstract). Electroen- and the temporal lobe. McLean Hosp J Special Vol-
cephalogr Clin Neurophysiol 58:38P, 1984. ume:22-39, 1977.
21. Niedermeyer E, Lopes da Silva F (eds). Electroen- 32. Lesser RP, Gordon B. Methodologic considerations
cephalography: Basic Principles, Clinical Applica- in cortical electrical stimulation in adults. In Luders
tions, and Related Fields, 2nd ed. Baltimore, Urban HO, Noachtar S (eds). Epileptic Seizures: Patho-
& Schwarzenberg, 1987. physiology and Clinical Semiology. Churchill Liv-
22. Olivier A, Gloor P, Quesney LF, Andermann F. The ingstone, New York, 2000, pp 153-165.
indications for and the role of depth electrode 33. Ojemann LM, Dodrill CB. Natural history of drug
recording in epilepsy. Appl Neurophysiol 46:33-36, resistant seizures: clinical aspects. Epilepsy Res
1983. Suppl 5:13-17, 1992.
23. Quesney LF, Constain M, Fish DR, Rasmussen T. 34. Pilcher WH, Roberts DW, Flanigin HF, et al. Com-
Frontal lobe epilepsy—A field of recent emphasis. plications of epilepsy surgery. In Engel J Jr (ed).
Am J EEC Technol 30:177-193, 1990. Surgical Treatment of the Epilepsies, 2nd ed. Raven
24. Quesney LF, Gloor P. Localization of epileptic foci. Press, New York, 1993, pp 565-581.
Electroencephalogr Clin Neurophysiol Suppl 37: 35. Lesser RP, Luders H, Morris HH, et al. Electrical
165-200, 1985. stimulation of Wernicke's area interferes with com-
25. Quesney LF, Risinger MW, Shewmon DA. Ex- prehension. Neurology 36:658-663, 1986.
tracranial EEG evaluation. In EngelJJr (ed). Sur- 36. Sperling MR, Engel J Jr. Electroencephalographic
gical Treatment of the Epilepsies, 2nd ed. Raven recording from the temporal lobes: a comparison
Press, New York, 1993, pp 173-195. of ear, anterior temporal, and nasopharyngeal elec-
26. Sharbrough FW. Electrical fields and recording trodes. Ann Neurol 17:510-513, 1985.
techniques. In Daly DD, Pedley TA (eds). Current 37. Theodore WH, Sato S, Kufta CV, Gaillard WD, Kel-
Practice of Clinical Electroencephalography, 2nd ley K. FDG-positron emission tomography and in-
ed. Raven Press, New York, 1990, pp 29-49. vasive EEG: seizure focus detection and surgical
27. Schiller Y, Cascino GD, Sharbrough FW. Chronic outcome. Epilepsia 38:81-86, 1997.
intracranial EEG monitoring for localizing the 38. Daly DD. Epilepsy and syncope. In Daly DD, Ped-
epileptogenic zone: an electroclinical correlation. ley TA (eds). Current Practice of Clinical Elec-
Epilepsia 39:1302-1308, 1998. troencephalography, 2nd ed. Raven Press, New
28. Schiller Y, Cascino GD, Busacker NE, Sharbrough York, 1990, pp 269-334.
FW. Characterization and comparison of local on- 39. Devinsky O, Sato S, Kufta CV, et al. Electroen-
set and remote propagated electrographic seizures cephalographic studies of simple partial seizures
recorded with intracranial electrodes. Epilepsia with subdural electrode recordings. Neurology
39:380-388, 1998. 39:527-533, 1989.
29. Jack CRJr, Theodore WH, Cook M, McCarthy G. 40. Ellingson RJ, Wilken K, Bennett DR. Efficacy of
MRI-based hippocampal volumetrics: data acquisi- sleep deprivation as an activation procedure in
tion, normal ranges, and optimal protocol. Magn epilepsy patients. J Clin Neurophysiol 1:83-101,
Reson Imaging 13:1057-1064, 1995. 1984.
30. O'Brien TJ, So EL, Mullan BP, et al. Subtraction peri- 41. Penfield W, Jasper H (eds). Epilepsy and the Func-
ictal SPECT is predictive of extratemporal epilepsy tional Anatomy of the Human Brain. Little Brown,
surgery outcome. Neurology 55:1668-1677, 2000. Boston, 1954.
Chapter 14
MOVEMENT-RELATED POTENTIALS
AND EVENT-RELATED POTENTIALS
Joseph Y. Matsumoto
MOVEMENT-RELATED CORTICAL Jerk-Locked Averaging

POTENTIALS Contingent Negative Variation
Technique EVENT-RELATED POTENTIALS
Normal Waveforms The P300
Abnormalities in Disease SUMMARY
MOVEMENT-RELATED in animals indicate that this activity reflects

CORTICAL POTENTIALS the feed-forward processing of motor com-
mands that are projected from the cerebel-
Special electroencephalographic (EEG) stud- lum through the thalamus to the motor and
ies have been designed to explore the corti- premotor areas of the cerebral cortex.2
cal processes underlying movement, atten-
tion, and cognition. As a first step, these
studies must define an "event" that is either Technique
the cause or the effect of the higher cortical
process under investigation. This event One of the most challenging tasks in clini-
serves as the temporal reference point for cal neurophysiology is to obtain movement-
computerized EEG averaging. Microcom- related cortical potentials. The minimal
puters and software programs are available scalp electrode montage should include the
that average the cortical activity that pre- Fz, Cz, C3, and C4 positions referenced to
cedes the event ("back averaging") and fol- linked ears. It is critical that the electro-ocu-
lows it. Although these potentials have been lographic activity also be recorded. By ap-
most important in understanding the corti- plying additional electrodes in grids over
cal activity underlying normal and abnormal sensorimotor cortex, the potentials can be
movements, their clinical application is lim- mapped topographically. A low-frequency
ited. Jerk-locked averaging can help identify filter with a cut-off in the range of approxi-
a cortical origin for some abnormal move- mately 1-0.05 Hz must be used to record the
ments, and the Bereitschaftspotential can slow premovement negative wave.
help identify some functional movement A brisk voluntary movement, usually of a
disorders. digit, acts as the timing event, but the char-
Kornhuber and Deecke1 defined the cor- acter of the movement must be defined.3
tical activity that surrounds a self-paced, vol- The subject is instructed to stare straight
untary movement and reported that the ahead and to make self-paced, repetitive
premotor cortical activity begins 1.0-1.5 sec- movements 3-10 seconds apart. The initial
onds before the onset of movement. This rise of rectified electromyographic (EMG)
activity is a gradually rising negativity that activity from the movement triggers data col-
they termed the Bereitschaftspotential. Studies lection. The computer buffer is configured
161
to collect the data for 2 seconds before and Bereitschaftspotential, a slowly rising nega-
for 1 second after the trigger. Data from 100 tivity with a maximal amplitude at the ver-
to 200 such movements are collected and tex, beginning 1.0-1.5 seconds before the
stored for later analysis. onset of movement.1 Approximately 500 ms
Averaging must be performed off-line to before movement onset, the slope of the
ensure proper artifact rejection. Each trace negativity turns more sharply upward. This
is reevaluated to detect any contamination period, termed NS', localizes more focally
by eye movement artifact. The tracing also to the contralateral central region.5 Closely
must display a clear EMG take-off point. If surrounding the EMG activity, the motor po-
acceptable, the tracing is aligned at the EMG tential appears as a peaked wave in a scalp
onset and computed into the ongoing aver- distribution over the contralateral motor
aging process.4 cortex. The final wave, called the reafferante
potentially Kornhuber and Deecke,1 appears
90 ms after movement. In fact, this is a
Normal Waveforms series of positive and negative waves pre-
dominantly over the frontal and parietal ar-
Movement-related potentials consist of three eas reflecting sensory feedback from the
major waves (Fig. 14-1). The earliest is the movement.
Figure 14-1. Movement-related potential recorded from multiple scalp locations, with right anterior tibial muscle
(TA) relaxation in a normal subject with no soleus movement (TS). ACC is a simultaneous accelerometer record-
ing. The initial slow negative Bereitschaftspotential (BP) begins 1.7 seconds before movement and is followed by
the negative slope (AW) 650 ms before movement. Later positivity represents the reafferente potential. Electro-
oculogram (EOG) shows that no eye movement occurred. (From Terada K, Ikeda A, Yazawa S, Nagamine T,
Shibasaki H. Movement-related cortical potentials associated with voluntary relaxation of foot muscles. Clin Neu-
rophysiol 110:397-403, 1999. By permission of Elsevier Science.)
Movement-Related Potentials and Event-Related Potentials 163
The precise identity of neural generators to define the muscle that leads the jerk. This
of each of the movement-related potentials muscle may then be used as the EMG trig-
is a subject of controversy.6 In particular, ger event. For this technique to be success-
there is much speculation about the activity ful, the jerks must occur frequently enough
of the supplementary motor area (SMA),7 to collect 50-100 tracings.
because of its central role in motor planning Jerk-locked averaging is most helpful in
and control. The vertex predominance of uncovering the cortical event preceding a
the Bereitschaftspotential led to the hy- myoclonic jerk when the standard EEG is un-
pothesis that it reflected planning activity in revealing. In cortical myoclonus, a focal or
the SMA.8'9 Evidence supporting this hy- generalized transient potential is recorded
pothesis comes from high-resolution EEG that precedes the myoclonic jerk by 20-60 ms
mapping studies and subdural grid record- (Fig. 14-2). In reticular reflex myoclonus,
ings that implicate mesial wall cortical areas, jerk-locked averaging actually obscures the
including the SMA, as strong generators of projected cortical transient potential be-
the Bereitschaftspotential.10'11 cause of the irregular interval between the
jerk and the EEG event.21 In nonepileptic
myoclonus, jerk-locked averaging fails to re-
Abnormalities in Disease veal time-locked cortical activity.
Early studies reported that the movement-

related potentials in Parkinson's disease
were normal, but this likely reflected trig-
gering difficulties in patients who make slow,
small movements. With the manual averag-
ing technique, it was found that an early
stage of negativity (650 ms before move-
ment) is depressed in parkinsonian patients,
perhaps reflecting poor activation in the
SMA1^ and impaired motor learning.13 Sim-
ilarly, the Bereitschaftspotential associated
with gait showed decreased activation in
Parkinson's disease.14 Levodopa augments
the activity in this abnormal portion of the
Bereitschaftspotential.15 In tardive dyskine-
sia, the Bereitschaftspotential is increased in
amplitude.16 Patients with focal hand dysto-
nia have abnormalities in the distribution
and amplitude of the NS' and Bere-
itschaftspotential that indicate impaired cor-
tical activation during voluntary movement
and relaxation.17'18 In cerebellar disease or
after ventral thalamotomy, the Bere-
itschaftspotential is reduced in amplitude or
absent, indicating loss of the feed-forward
message in the cerebellothalamocortical
loop.19 Figure 14-2. Averaged scalp potential triggered and
back averaged from extensor indicis proprius record-
ing of the EMG in a patient with lipidosis and my-
oclonus. A, EEG traces show no evidence of a cortical
Jerk-Locked Averaging discharge corresponding to the myoclonic jerks recorded
in the EMG. B, 100 averaged responses recorded from
Movement-related potentials can be used to the same scalp electrodes when triggered by the mus-
cle activity show a well-defined cortical sharp wave 16-18
evaluate involuntary movements through a ms before the EMG that is maximal at the C4 and P4
technique termed jerk-locked averaging.20 The electrodes. (From Shibasaki and Kuroiwa.20 By permis-
examiner first performs a surface EMG study sion of Elsevier Science.)
Jerk-locked averaging has a limited role in the disciplines of physiology, psychology,

evaluating other involuntary movements. In psychiatry, and neurology—is beyond the
Tourette's syndrome, either no cortical ac- scope of this book.
tivity or a rudimentary premovement poten-
tial precedes spontaneous tics.22'23 However,
when the patient voluntarily mimics a tic, the The P300
normal Bereitschaftspotential appears. Sim-
ilarly, in Huntington's disease, propriospinal The P300 is the most commonly recorded
myoclonus, and periodic leg movements in event-related potential.34 Generally, an odd-
restless legs syndrome, involuntary move- ball technique of auditory stimulation is used,
ments are not preceded by the Bereitschafts- in which a standard stimulus, also called fre-
potential.24"26 Psychogenic myoclonus, how- quent stimulus, is replaced at infrequent in-
ever, may be distinguished by the presence tervals by a stimulus of different tone,
of a Bereitschaftspotential before an appar- termed the oddball stimulus or rare stimulus.
ently involuntary muscle jerk.27 The subject is instructed to attend to or to
count the oddball stimuli. Only trials trig-
gered by this rare event are averaged.
Contingent Negative Variation At times, the P300 is visible on a single raw
tracing. Averaging clearly defines a wave with
In the contingent negative variation testing a peak latency of approximately 300 ms and
paradigm, a stimulus such as a click warns an amplitude of approximately 10 /tV. The
the subject to prepare to move. After 1-2 sec- amplitude of the wave is increased by many
onds, a second stimulus, such as a flash of factors, including the subject's attentive ness
light, signals the patient to begin moving. and the unpredictability of the oddball stim-
Contingent negative variation is a slow neg- ulus. The P300 has a bilateral, mid-parietal
ative potential that appears in the interval distribution. However, a single generator for
between the warning stimulus and the sec- the potential cannot be defined; the wave
ond stimulus. The distribution of this wave likely reflects activity in several areas of the
is predominately bilateral and frontal, but it brain.35 The role of the wave in cognition is
may shift with variations in the testing pro- also debated. It may be the electrophysio-
cedure. The neural generators of contingent logic correlate of selected attention.
negative variation appear to be different The P300 is abnormal in many diseases in
from those of the Bereitschaftspotential.28 which cortical processing is impaired. The
amplitude is decreased and the latency is pro-
longed in all types of dementia.36 Prolonga-
EVENT-RELATED POTENTIALS tion of the P300 latency may also be a pre-
clinical finding in those at risk for developing
Whereas standard somatosensory or visual Alzheimer's disease.37 Abnormalities have
evoked potentials map the cortical response also been reported in Parkinson's disease,38
to a simple sensory stimulus, event-related mild metabolic encephalopathies, drug in-
potentials record the cortical activity evoked toxications, multiple sclerosis, autism, and
by a stimulus charged with cognitive signifi- schizophrenia.34
cance. As such, event-related potentials are
more sensitive to the "endogenous" reaction
to a stimulus than to the physical nature of SUMMARY
the stimulus. Sutton et al.29 were among the
first to note a large late cortical positivity in Special EEC averaging techniques may be
reaction to stimuli to which the subject at- used to study the cortical processes underly-
tached importance. Since that time, numer- ing movement and cognition. Movement-
ous techniques have been devised to record related potentials and contingent negative
the cortical activity surrounding processes variation are observed before a voluntary
such as selective attention,30 memory,31 ol- movement occurs. Jerk-locked averaging may
faction,32 and facial recognition.33 A com- detect cortical activity associated with invol-
prehensive review of this topic—spanning untary movements. The P300 and other
Movement-Related Potentials and Event-Related Potentials 165
event-related potentials provide electrophysi- and initiation of gait in Parkinson's disease. Neu-
ologic correlates of perception and cognition. rology 43:1784-1788, 1993.
15. Dick JP, Cantello R, Buruma O, et al. The Bere-
itschaftspotential, L-DOPA and Parkinson's disease.
Electroencephalogr Clin Neurophysiol 66:263-274,
1987.
REFERENCES 16. Adler LE, Pecevich M, Nagamoto H. Bere-
itschaftspotential in tardive dyskinesia. Mov Disord
1. Kornhuber HH, Deecke L. Hirnpotentialanderun- 4:105-112, 1989.
gen bei Willkurbewegungen und passiven Bewe- 17. Deuschl G, Toro C, Matsumoto J, Hallett M. Move-
gungen des Menschen: Bereitschaftspotential und ment-related cortical potentials in writer's cramp.
reafferente Potentiale. Pflugers Arch 284:1-17, Ann Neurol 38:862-868, 1995.
1965. 18. Yazawa S, Ikeda A, Kaji R, et al. Abnormal cortical
2. Sasaki K, Gemba H, Hashimoto S, Mizuno N. In- processing of voluntary muscle relaxation in pa-
fluences of cerebellar hemispherectomy on slow po- tients with focal hand dystonia studied by move-
tentials in the motor cortex preceding self-paced ment-related potentials. Brain 122:1357-1366,
hand movements in the monkey. Neurosci Lett 1999.
15:23-28, 1979. 19. Shibasaki H, Shima F, Kuroiwa Y. Clinical studies
3. Slobounov S, Rearick M, Chiang H. EEC correlates of the movement-related cortical potential (MP)
of finger movements as a function of range of mo- and the relationship between the dentatorubrothal-
tion and pre-loading conditions. Clin Neurophysiol amic pathway and readiness potential (RP). J Neu-
111:1997-2007, 2000. rol 219:15-25, 1978.
4. Barrett G, Shibasaki H, Neshige R. A computer-as- 20. Shibasaki H, Kuroiwa Y. Electroencephalographic
sisted method for averaging movement-related cor- correlates of myoclonus. Electroencephalogr Clin
tical potentials with respect to EMG onset. Elec- Neurophysiol 39:455-463, 1975.
troencephalogr Clin Neurophysiol 60:276-281, 21. Hallett M, Chadwick D, Adam J, Marsden CD. Retic-
1985. ular reflex myoclonus: a physiological type of hu-
5. Shibasaki H, Barrett G, Halliday E, Halliday AM. man post-hypoxic myoclonus. J Neurol Neurosurg
Components of the movement-related cortical po- Psychiatry 40:253-264, 1977.
tential and their scalp topography. Electroen- 22. Karp BI, Porter S, Toro C, Hallett M. Simple mo-
cephalogr Clin Neurophysiol 49:213-226, 1980. tor tics may be preceded by a premotor potential.
6. Slovounov SM, Rearick MP, Simon RF, Johnston JA. J Neurol Neurosurg Psychiatry 61:103-106, 1996.
Movement-related potentials are task or end-effec- 23. Obeso JA, Rothwell JC, Marsden CD. Simple tics in
tor dependent: evidence from a multifinger exper- Gilles de la Tourette's syndrome are not prefaced
iment. Exp Brain Res 135:106-116, 2000. by a normal premovement EEG potential. J Neurol
7. Stancak A Jr, Feige B, Lucking CH, Kristeva-Feige Neurosurg Psychiatry 44:735-738, 1981.
R. Oscillatory cortical activity and movement-re- 24. Berardelli A, Noth J, Thompson PD, et al. Patho-
lated potentials in proximal and distal movements. physiology of chorea and bradykinesia in Hunting-
Clin Neurophysiol 111:636-650, 2000. ton's disease. Mov Disord 14:398-403, 1999.
8. Deecke L, Kornhuber HH. An electrical sign of par- 25. Brown P, Thompson PD, Rothwell JC, Day BL,
ticipation of the mesial 'supplementary' motor cor- Marsden CD. Axial myoclonus of propriospinal ori-
tex in human voluntary finger movement. Brain Res gin. Brain 114:197-214, 1991.
159:473-476, 1978. 26. Trenkwalder C, Bucher SF, Oertel WH, Proeckl D,
9. Toro C, Matsumoto J, Deuschl G, Roth BJ, Hallett Plendl H, Paulus W. Bereitschaftspotential in idio-
M. Source analysis of scalp-recorded movement-re- pathic and symptomatic restless legs syndrome.
lated electrical potentials. Electroencephalogr Clin Electroencephalogr Clin Neurophysiol 89:95-103,
Neurophysiol 86:167-175, 1993. 1993.
10. Cui RQ, Deecke L. High resolution DC-EEG anal- 27. Terada K, Ikeda A, Van Ness PC, et al. Presence of
ysis of the Bereitschaftspotential and post move- Bereitschaftspotential preceding psychogenic my-
ment onset potentials accompanying uni- or bilat- oclonus: clinical application of jerk-locked back av-
eral voluntary finger movements. Brain Topogr eraging. J Neurol Neurosurg Psychiatry 58:745-747,
11:233-249, 1999. 1995.
11. Ikeda A, Yazawa S, Kunieda T, et al. Cognitive mo- 28. Ikeda A, Shibasaki H, Nagamine T, et al. Dissocia-
tor control in human pre-supplementary motor area tion between contingent negative variation and
studied by subdural recoding of discrimination/ Bereitschaftspotential in a patient with cerebellar
selection-related potentials. Brain 122:915-931, efferent lesion. Electroencephalogr Clin Neuro-
1999. physiol 90:359-364, 1994.
12. Dick JP, Rothwell JC, Day BL, et al. The Bere- 29. Sutton S, Braren M, Zubin J, John ER. Evoked-
itschaftspotential is abnormal in Parkinson's dis- potential correlates of stimulus uncertainty. Science
ease. Brain 112:233-244, 1989. 150:1187-1188, 1965.
13. Fattapposta F, Pierelli F, Traversa G, et al. Prepro- 30. Heinze HJ, Mangun GR, Burchert W, et al. Com-
gramming and control activity of bimanual self- bined spatial and temporal imaging of brain activ-
paced motor task in Parkinson's disease. Clin Neu- ity during visual selective attention in humans. Na-
rophysiol 111:873-883, 2000. ture 372:543-546, 1994.
14. Vidailhet M, Stocchi F, Rothwell JC, et al. The Bere- 31. Begleiter H, Porjesz B, Wang W. A neurophysio-
itschaftspotential preceding simple foot movement logic correlate of visual short-term memory in hu-
mans. Electroencephalogr Clin Neurophysiol 87: 35. Neshige R, Luders H. Recording of event-related
46-53, 1993. potentials (P300) from human cortex. J Clin Neu-
32. Evans WJ, Cui L, Starr A. Olfactory event-related rophysiol 9:294-298, 1992.
potentials in normal human subjects: effects of age 36. Goodin DS, Squires KG, Starr A. Long latency event-
and gender. Electroencephalogr Clin Neurophys- related components of the auditory evoked poten-
iol 95:293-301, 1995. tial in dementia. Brain 101:635-648, 1978.
33. Hertz S, Porjesz B, Begleiter H, Chorlian D. Event- 37. Green J, Levey AI. Event-related potential changes
related potentials to faces: the effects of priming in groups at increased risk for Alzheimer disease.
and recognition. Electroencephalogr Clin Neuro- Arch Neurol 56:1398-1403, 1999.
physiol 92:342-351, 1994. 38. Rumbach L, Tranchant C, VielJF, WarterJM. Event-
34. Picton TW. The P300 wave of the human event-re- related potentials in Parkinson's disease: a 12-
lated potential. J Clin Neurophysiol 9:456-479, month follow-up study. J Neurol Sci 116:148-151,
1992. 1993.
SECTION 2
Electrophysiologic
Assessment of
Neural Function
Part B
Sensory
Pathways
The sensory axons that conduct information from the periphery to the cen-
tral nervous system generate electricity. Because the signals from single ax-
ons are difficult to record, most electrophysiologic recordings from humans
undergoing testing for possible neurologic or neuromuscular disease are
summated responses made from specific generators in response to con-
trolled external stimulation.
Somatic sensory and somatic motor axons can be tested by stimulating
along the length of a nerve while recording the sensory or motor responses
from peripheral nerve or muscle. Sensory axons can be isolated for testing
by selective stimulation of sensory structures or by selective recording from
generators that are purely sensory. The potentials recorded from sensory
structures in response to specific stimulation are called sensory evoked poten-
tials. Sensory evoked potentials are classified as nerve conduction studies that
test peripheral nerves (Chapter 15), somatosensory evoked potentials that test
central somatic sensory pathways (Chapter 16), brain stem auditory evoked po-
tentials that test peripheral and central auditory pathways (Chapters 17, 18,
and 19), and visual evoked potentials that test peripheral and central visual
pathways (Chapter 20).
Combinations of these sensory potential recording methods allow precise
localization of damage to different levels of the nervous system and, in some
cases, help to define the type of underlying lesion. Advances in the method
of stimulation and recording have extended the applications of some of
these techniques.
Mechanical components of peripheral auditory function can be tested sep-
arately with audiography, acoustic reflexes, and evoked otoacoustic emis-
168 Sensory Pathways
sions. Movement-related potentials and event-related potentials (see Chap-

ter 14 in Part A, Cortical Function) are also sometimes referred to as evoked
potentials. Evoked potentials obtained with stimulation of motor axons are
considered in Part C, Motor Pathways. Sensory evoked potentials each have
unique and specific waveforms that are altered in characteristic ways by dis-
ease. They may be recorded directly or averaged if the signal is small.
Chapter 15
NERVE ACTION POTENTIALS
Eric J. Sorenson
CLINICAL IMPORTANCE METHODS

PLANNING THE STUDY Nerve Stimulation
RADICULOPATHY Recording the Potential
PLEXOPATHY Averaging
COMMON MONONEUROPATHIES MEASUREMENTS
TECHNICAL FACTORS SUMMARY
HUMAN FACTORS
Electrically evoked nerve action potentials NAP studies, including the intrinsic and ex-
(NAPs) is the only way clinically to study di- trinsic electrical factors and their effects on
rectly the function of peripheral nerves. these studies. Common human errors made
These potentials are not affected by sec- in conducting the studies and how these er-
ondary factors such as transmission at the rors affect the studies are reviewed. Finally,
neuromuscular junction or the electrical ex- the chapter explains how NAPs are mea-
citability of the muscle, as are compound sured and the function of a nerve is quanti-
muscle action potentials. Except perhaps for fied.
sympathetic sweat gland skin potentials,1 the
electrical excitability of sensory nerve target
receptors cannot be measured reliably. Al- CLINICAL IMPORTANCE
though somatosensory evoked potentials
provide information about the proximal pe- The main value of assessing peripheral NAPs
ripheral sensory pathways (see Chapter 16), is in studying sensory nerves. Nerve action
NAPs are the only practical way to assess sen- potentials are still among the most reliable
sory peripheral nerves reliably. Because of means of studying peripheral sensory
these factors, the measurement of electri- nerves.2 Nerve action potentials are sensitive
cally evoked peripheral NAPs provides in- to pathologic conditions in a nerve. Often,
valuable information about the physiology alterations in the amplitudes, terminal la-
and pathology of the peripheral nervous sys- tencies, and conduction velocities of evoked
tem, particularly about sensory nerves. responses are the earliest abnormalities de-
This chapter discusses the clinical rele- tected in a peripheral neuropathic process.3
vance of NAPs and describes strategies for The studies provide invaluable data for the
planning appropriate sensory testing to max- localization and classification of a peripheral
imize diagnostic yield. Also, the recording of neuropathy, and the study results may indi-
NAPs is frequently complicated by unique cate a specific pathologic condition.
technical difficulties, which are discussed Most of the peripheral nerves that are
with a strategy for troubleshooting such studied are mixed nerves that have motor
problems. The chapter also includes a brief and sensory axons. Whenever possible, it is
review of the electrical principles relevant to best to isolate the sensory and motor com-
169
technique. The main advantage of the or-

thodromic technique is that it eliminates vol-
ume conduction from muscle action poten-
tials because no motor fibers are activated.
Another, and less optimal technique, is to
stimulate a mixed nerve and to record at a
fixed distance over the nerve where it con-
tains both motor and sensory fibers. With
this technique, it is assumed diat because the
nerve has a small number of motor axons
and the amplitude of the NAP is directly pro-
portional to the total number of axons acti-
vated, the contribution of the motor fibers
Figure 15-1. Triphasic appearance of a nerve action po-
tential. Note initial positivity (A) followed by the dom- to the NAP is negligible. Because of the vari-
inant negative peak (B). ation from subject to subject in the motor
and sensory components of mixed nerves,
normal data obtained with this technique
ponents, because studying a pure popula- are more variable than comparable data ob-
tion of fibers provides information that is tained with the antidromic and orthodromic
more meaningful for interpretation. For techniques.
practical purposes, only the sensory compo- Some of the peripheral nerves that are
nent can be isolated. Because the waveforms available for testing are pure sensory nerves.
are generated from the depolarizing nerve, For these nerves, it is necessary only to stim-
the electrodes will record a traveling wave. ulate the nerve and to record at a fixed dis-
Thus, it is common for sensory waveforms to tance along it, using either the antidromic
have an initial positivity and a triphasic (in- or orthodromic technique. Examples of pure
stead of biphasic) appearance (Fig. 15-1). sensory nerves are the sural, superficial per-
Two main strategies have been developed oneal, saphenous, and medial and lateral an-
for studying sensory axons. The first strategy tebrachial sensory nerves. Although superfi-
is to stimulate the mixed motor and sensory cially the technique for studying these nerves
nerve and to record the NAP at a site that is appears straightforward, technical factors
distal to where the nerve splits into motor may make it difficult, as discussed below.
and sensory components. This is called the A major advantage of studying sensory
antidromic technique, because the direction of NAPs is their sensitivity for detecting an un-
the action potential is opposite (anti-) that derlying abnormality. Frequendy, sensory
of the physiologic action potential. The ad- potential abnormalities are the earliest find-
vantage of the antidromic technique is that ings in a pathologic state. To interpret the
it ensures adequate supramaximal stimula- findings accurately, the temporal profile and
tion of the nerve and, thus, larger ampli- evolution of the changes in NAPs must be
tudes.4'5 However, this technique also acti- understood. This is particularly important in
vates the motor fibers and generates a cases of acute nerve injury. Following an
muscle action potential that, through vol- acute axonal lesion, the NAP initially is nor-
ume conduction, may interfere with the sen- mal. Over the ensuing days, as wallerian de-
sory NAP that is being recorded. generation begins and the degenerating ax-
The second strategy for isolating sensory ons lose their electrical excitability, the
fibers in a mixed nerve is to stimulate the amplitude of the NAP begins to decrease,
nerve distal to the point where it splits into reaching its nadir approximately 1 to 2 weeks
sensory and motor components and to after the injury.
record proximally over the mixed nerve. Alternatively, if the lesion affects only the
This is called the orthodromic technique, be- myelin sheath and causes a focal area of de-
cause the direction of the action potential is myelination, the sensory NAPs distal to the
the same as that of the physiologic action po- lesion may be normal. Stimulating and
tential. However, the number of fibers acti- recording across the area of demyelination
vated and the amplitude of the responses may show a delay in the conduction velocity
are more variable than with the antidromic across the site; the delay will be seen as ei-
Nerve Action Potentials 171
ther a prolonged latency or slowed conduc- be slightly decreased because the largest,
tion velocity. Unlike compound muscle ac- fastest conducting fibers in the area of de-
tion potentials, focal conduction block, as myelination are affected.
defined by a discrete loss of amplitude across In diffuse demyelination, in which the
a focal segment usually cannot be estab- nerve is affected all along its course, distal
lished with NAPs. This is caused by phase can- latencies are prolonged, conduction veloci-
cellation, which results from sensory fibers ties are slowed, and amplitudes are reduced.
having different conduction velocities. Nor- However, the electrophysiologic features
mally, phase cancellation causes a substan- that are used to differentiate an acquired de-
tial reduction in amplitudes within sensory myelinating process from an inherited pro-
nerve action potentials with increasing dis- cess, that is, conduction block and disper-
tance between the stimulating and record- sion, cannot be assessed accurately with
ing electrodes. In chronic neuropathic le- sensory NAPs.
sions, sensory NAPs are more sensitive in Testing NAPs over long segments is more
detecting abnormality than compound mus- sensitive for detecting diffuse demyelina-
cle action potentials because the former are tion; because the effects of subtle changes
unaffected by reinnervation that occurs with are additive, the changes become more ap-
the compound muscle action potential. parent when a long segment of nerve is
Understanding how pathologic condi- tested. Conversely, in the case of focal de-
tions affect NAPs helps to classify the neu- myelination, testing over the shortest seg-
ropathic process. This classification has im- ment of nerve possible provides the greatest
portant implications for the differential sensitivity, because subtle areas of focal con-
diagnosis of neuropathy.2 The usual finding duction slowing are not "averaged" with the
in axonal neuropathy is a decrease in am- normally conducting segments. The differ-
plitude. Because the neuropathic process ence between focal and diffuse disorders be-
preferentially can affect the largest fibers comes important when selecting the sensory
and, hence, the fastest conducting ones, the NAPs to sample in response to a specific clin-
conduction velocity may be slightly below ical situation. For example, if focal median
the limit of normal. Generally, axonal loss mononeuropathy at the wrist is suspected,
alone should not decrease conduction ve- sampling the sensory NAP across a short dis-
locity less than approximately 70% of the tance, as in the palmar (orthodromic) stim-
lower limit of normal.6 Similarly, the termi- ulation technique, provides the greatest
nal distal latency of the response may be sensitivity. However, if a diffuse disorder is
slightly prolonged. suspected, a median antidromic technique
Jf the nerve has a focal area of demyeli- with proximal and distal stimulation is pre-
nation, the findings depend on the sites of ferred because the amplitude is more re-
stimulation. If the stimulation sites are prox- producible and the conduction velocity is
imal and distal to the area of demyelination, sampled over a long segment of nerve.
the conduction velocity usually is decreased Because sensory NAPs are sensitive to pe-
substantially. However, the distal latency is ripheral neuropathic lesions, they can pro-
normal. If the area of demyelination causes vide important information for localization.
pronounced conduction block or disper- The distribution of abnormalities can sug-
sion, no proximal response may be obtained. gest a focal lesion, a multifocal process, or a
However, true conduction block, defined as diffuse disease. Also, because of the unique
loss of amplitude over a discrete segment, anatomy of sensory neurons, sensory NAPs
cannot be determined reliably by sensory are extremely helpful in differentiating an
nerve conduction studies alone. intraspinal process from a more peripheral
If the focal demyelinating lesion occurs one. The cell bodies of the sensory neurons
only in the terminal segment of the nerve, form dorsal root ganglia, which lie within the
the distal latency is prolonged. The ampli- intervertebral foramina, where the spinal
tudes are frequently reduced if the lesion is roots exit from the spinal canal. Thus, a pro-
associated with conduction block or phase cess that is localized within the spinal canal
cancellation. A conduction velocity obtained is described as preganglionic. In a pregan-
by stimulating or recording at two sites prox- glionic lesion, the distal sensory axon re-
imal to the lesion in a terminal segment may mains intact and connected with the cell
body. The sensory nerve action potentials re- tients referred to the electrophysiology lab-
main normal, even if the sensory loss is se- oratory. Sensory NAPs are most useful in
vere. Conversely, any lesion that affects the confirming that the lesion is preganglionic
nerve by interrupting its axons distal to the (that is, intraspinal). The most common
intervertebral canal causes a loss of ampli- lumbar radiculopathies are at the L5 and SI
tude in sensory NAPs. This provides invalu- levels, followed by the L4 and L3 levels. Be-
able information for differentiating a pre- cause a peroneal neuropathy may mimic an
ganglionic lesion such as a radiculopathy L5 radiculopathy, the superficial peroneal
from a postganglionic lesion such as a plex- sensory nerve should be studied. As men-
opathy or mononeuropathy. However, re- tioned above, the dorsal root ganglia are
member that a postganglionic lesion that within the intervertebral foramina, and an
does not affect the axons (that is, pure con- intraspinal lesion does not disrupt the con-
duction block) will not affect the amplitude tinuity between the cell body and its axon;
of sensory NAPs if the stimulation and therefore, the sensory NAPs are not affected.
recording sites are distal to the lesion. However, the location of the dorsal root gan-
glia can vary. The dorsal root ganglia of
lower lumbar and upper sacral roots may ac-
PLANNING THE STUDY tually be located within the spinal canal in
40% of patients.7 Thus, in these patients,
radiculopathy caused by lateral herniation of
The sensory NAPs to be sampled must be se- an intervertebral disk may cause axonal dam-
lected on the basis of the clinical findings age peripheral to the dorsal root ganglion,
and the differential diagnosis of the pre- reducing the amplitude of sensory NAPs.
senting complaints. The general rule is that This has been demonstrated in the superfi-
the sensory distribution that is affected clin- cial peroneal sensory nerve in L5 radicu-
ically should be tested. Although this seems lopathies.7 In an SI radiculopathy, the sural
intuitive, often electromyographers are sensory nerve should be selected for testing
tempted to restrict testing to the more com- because it is located within the dermatomal
mon nerves because these nerves are more distribution of the SI nerve root. Because a
familiar and less complicated. This tempta- femoral neuropathy may mimic an L3 or L4
tion has to be avoided because, as in many radiculopathy, the saphenous sensory nerve
cases, testing the less familiar nerves can pro- should be studied to exclude a postgan-
vide invaluable localizing information that glionic lesion.
often cannot be obtained otherwise. This re- The superficial peroneal sensory, sural
quires knowledge of the techniques unique sensory, and saphenous sensory nerves are
to these nerves and the technical problems pure sensory nerves, and anatomical land-
that tend to occur with these studies. The marks have been established for several tech-
importance of maintaining skill in this area niques for stimulating and recording from
cannot be overemphasized. Retaining these these nerves. However, the anatomical loca-
skills requires that less familiar sensory tion of these nerves varies, and amplitudes
nerves be tested regularly, not only once or vary significantly from person to person.
twice a year. Examples of these less familiar Also, for each of these nerves, the normal
sensory nerves are the superficial peroneal amplitude values diminish with age, and with
sensory nerve, the lateral and medial ante- age, the amplitudes become increasingly dif-
brachial nerves, and the dorsal ulnar cuta- ficult to obtain. Because of this, it is impor-
neous nerve. Testing of each of these nerves tant to compare the responses with those of
in the appropriate clinical setting adds sub- the opposite side in any case in which re-
stantially to the quality of the study. sponses cannot be obtainable or the ampli-
tude is equivocal for a person of that age.
Normally, the side-to-side asymmetry may be
RADICULOPATHY as much as 50%.8 These less common sen-
sory NAP studies should be conducted for
Cervical and lumbar radiculopathies are the common referral diagnoses in order to
among the most common diagnoses of pa- maintain the skill needed to perform such
tests with confidence and to obtain valid and have not been developed to sample the cu-
reliable results. taneous branches in the upper lumbar der-
matomes. Techniques for dermatomal so-
matosensory evoked potentials have been
PLEXOPATHY developed9 and are described in Chapter 16.
Localization of an upper lumbar plexopa-
The selection of nerves to be tested in a per- thy often relies on the findings of needle
son with suspected plexopathy should be electromyography.
based on the most likely localization deter-
mined on routine neurologic examination.
In cases of brachial plexopathy, the specific COMMON MONONEUROPATHIES
site of involvement often cannot be localized
on the basis of clinical findings alone. Tai- Median and ulnar neuropathies are among
loring the study to the areas of suspected in- the most common diagnoses of patients with
volvement increases substantially the yield of an upper extremity problem who are re-
the nerve conduction studies. Although bra- ferred to the electrophysiology laboratory.
chial plexus lesions can be patchy in distri- Prolongation of the distal latency of the me-
bution, a clinical examination often suggests dian nerve across the carpal tunnel at the
one of three patterns: upper trunk/lateral wrist is the most common abnormality in a
cord, middle trunk/posterior cord, or lower median neuropathy at the wrist. Several
trunk/medial cord. In the upper trunk/ techniques have been described that assess
lateral cord distribution, the lateral ante- slowing of conduction in the median nerve
brachial cutaneous sensory nerve needs to at the wrist. Two techniques are used almost
be studied in addition to the median nerve. exclusively in clinical practice: an anti-
The lateral antebrachial cutaneous sensory dromic technique and an orthodromic tech-
nerve represents the termination of the mus- nique. In the antidromic technique, the
culocutaneous nerve and, in all cases, is a recording site is over one of the digits sup-
branch from the upper trunk and lateral plied by the median nerve, commonly the
cord. If a middle trunk/posterior cord le- second digit (index finger), and the stimu-
sion is suspected, a superficial radial sensory lation sites are proximal to the wrist and at
response in addition to the median sensory the elbow. As mentioned above under Clin-
response will enable a more complete as- ical Importance, the advantage of this tech-
sessment of the cutaneous distribution from nique is that the amplitudes are less variable.
this of the brachial plexus. If a lower However, because the antidromic technique
trunk/medial cord lesion is suspected, a me- involves a longer distance, it is less sensitive
dial antebrachial cutaneous nerve study in to subtle slowing of conduction across the
addition to an ulnar sensory nerve study is elbow. In the orthodromic technique, the
necessary to adequately assess the cutaneous stimulation site is the palm and the record-
distribution of the lesion. As with some sen- ing sites are proximal to the wrist and at the
sory nerves in the lower extremity, these un- elbow. This technique is more sensitive for
common nerve studies become increasingly focal slowing because the distal distance is
difficult to perform the older the patient shorter. Also, the recordings can be ob-
is, and side-to-side comparisons should be tained with one set of electrical stimulations,
made for any responses that cannot be ob- decreasing the number of shocks compared
tained or have an equivocal amplitude. with that of the antidromic technique. How-
In lumbosacral plexopathy, the anatomi- ever, with the orthodromic technique, the
cal patterns are not as discrete as they are in number of axons activated at supramaximal
the upper limb. Clinically, lumbosacral plex- stimulation is more variable, as is the ampli-
opathies often can be divided into two der- tude of the sensory NAPs.
matomal distribution patterns: upper lum- For clinically advanced median mononeu-
bar and lower lumbosacral. In most cases, ropathy at the wrist, the antidromic study is
the lower lumbosacral distribution can be preferable to the palmar orthodromic study,
sampled with the sural and superficial pe- because the reduced sensitivity of the an-
roneal sensory nerves. Reliable techniques tidromic response is not an issue and the
studying median sensory NAPs is not often

used to assess ulnar neuropathy, because it
does not appreciably increase the sensitivity
of the ulnar studies. The explanation mainly
is that the ulnar antidromic distance is not
markedly longer than the palmar ortho-
dromic distance.
Conduction block in the ulnar nerve
across the elbow cannot be proven reliably
with sensory studies, but if conduction
block is present, the proximal sensory am-
plitudes may be unobtainable. In the case
of pure conduction block, the distal sensory
amplitudes and latencies are normal. Stim-
ulation above and below the elbow may be
helpful in demonstrating focal slowing of
conduction across the elbow by isolating
Figure 15-2. Example of median (upper tracing) and ul- the slowing across a much shorter segment.
nar (lower tracing) palmar sensory studies in mild me- If the axons are disrupted, the only finding
dian mononeuropathy at the wrist. Note that the dif- on an ulnar antidromic sensory study may
ference in the peak latencies between the two studies be a reduction in amplitude, a finding that
is nearly 1 ms.
is not helpful in localization. If this is the
case, additional sensory studies can be used
to further localize the lesion. For example,
amplitude values are more meaningful. For the dorsal ulnar cutaneous nerve branches
clinically mild median mononeuropathy at from the main ulnar nerve in the forearm;
the wrist, the palmar orthodromic study is thus, if it is abnormal, the lesion can be lo-
preferable to the antidromic study because calized to a site above the wrist. However,
of the greater sensitivity of the orthodromic the reliability of dorsal ulnar cutaneous re-
response; furthermore, in this case, the am- sponses is less than that of other nerves
plitude of the sensory NAP is not as relevant more commonly tested.11 Occasionally, a
as the distal latency (Fig. 15-2). If the clini- lower trunk or medial cord brachial plex-
cal scenerio is indeterminate, the median opathy may appear similar clinically to an
motor nerve should be tested first. If the mo- ulnar neuropathy. If both the ulnar an-
tor responses show slowing across the wrist, tidromic study and the dorsal ulnar cuta-
it is likely that the sensory responses will, too; neous sensory nerves are abnormal, the me-
therefore, the antidromic technique should dial antebrachial cutaneous nerve should
be used. If there is no slowing of conduction be tested. Abnormal findings in this nerve
across the wrist, the more sensitive palmar suggest a more proximal lesion of the me-
orthodromic technique should be used. Var- dial cord or lower trunk of the brachial
ious grading scales have been developed to plexus.
quantify severity of the median mononeu- Peroneal, sciatic, and femoral neurop-
ropathy at the wrist, using nerve conduction athies are common diagnoses of patients
studies.10 with a lower extremity problem who are re-
Ulnar neuropathies occur most com- ferred to the electrophysiology laboratory.
monly at the elbow; however, definitive lo- The sensory nerves to be tested for each of
calization of the lesion often can be difficult, these mononeuropathies are the superficial
especially the longer the condition has been peroneal, sural, and saphenous nerves, re-
chronic. The most common technique for spectively. The antidromic technique is
studying the ulnar sensory nerve is the an- used. If the lesion causes a pure conduction
tidromic method, with the recording site block above the level of stimulation, the sen-
over the 5th digit and the stimulation sites sory NAP may be normal. This is not un-
proximal to the wrist and above the elbow. common in a peroneal neuropathy at the
The palmar orthodromic technique used in head of the fibula. If the amplitude in any
of these sensory studies is abnormal or bor- between the recording and stimulation sites
derline, the opposite side should be tested. affects the amplitude. Third, the tempera-
ture of the nerve at the time of the study
and, fourth, the distance between the elec-
TECHNICAL FACTORS trodes and the nerve affect the amplitude.
Each of these factors is discussed below. To
The sensory NAPs measured in the electro- obtain reproducible and comparable results,
physiology laboratory represent a summa- the stimulation distance, nerve temperature,
tion of individual action potentials of all the and recording distance must be controlled
large myelinated sensory axons in the stim- and standardized for each sensory nerve
ulated nerve. Because the responses are tested.
recorded from the nerve and not from the The first factor, the number of axons de-
muscle, the amplitudes are much lower than polarized—The primary reason NAP ampli-
those of compound muscle action poten- tudes are measured is they directly reflect
tials. This causes several problems that are the number of functioning axons in the
not usually encountered in motor conduc- nerve. For this reason, sensory NAPs are sen-
tion studies. The main difficulty is the back- sitive to any denervating process that affects
ground electrical noise. Because the sensory sensory axons. Abnormalities detected with
amplitudes are low, the background noise sensory nerve conduction studies are often
appears proportionally larger; this is re- the most sensitive markers for peripheral
ferred to as a lower signal-to-noise ratio. Thus, neuropathies.
it is imperative that background electrical ac- The second factor, the distance between
tivity be minimized. This includes proper im- the recording and stimulation sites—Be-
pedence of the electrodes to avoid any im- cause the conduction velocity of sensory ax-
pedence mismatch that would distort the ons varies markedly, the longer the con-
common mode rejection between the elec- ducting segment the more the responses
trodes. Also, background voluntary muscle tend to disperse. Thus, the closer a depo-
activity that interferes with the baseline must larizing stimulus is applied to the recording
be minimized. The shock must be delivered site, the less the dispersion and the larger
so that shock artifact is minimized. This re- the amplitude. The longer the distance, the
quires proper ground placement, and it greater the dispersion and the lower the am-
often requires rotating the stimulator to plitude. Although the distance between the
minimize the effects of shock artifact. It is distal stimulating site and recording site can
important to place the stimulator as near the be maintained at a fixed interval, the dis-
nerve as possible to allow supramaximal de- tance between the proximal sites cannot be.
polarization with the minimal amount of Therefore, it is preferable to record and to
current. This will benefit the study by mini- compare the distal amplitudes of the re-
mizing shock artifact and by reducing the sponses instead of the proximal ones.
risk of overstimulation and creating muscle The third factor, nerve temperature—If
artifact. Occasionally, a needle cathode the limb is cool, the action potential is pro-
needs to be placed near the nerve to prolonged, lengthening the duration of depo-
vide the appropriate amount of stimulation. larization, which in turn increases the am-
Sources of external electrical activity need plitude of the response. This effect is not
to be eliminated, including any electrical insignificant clinically, and limb tempera-
equipment within the vicinity of the study, ture cannot be ignored in performing nerve
particularly fluorescent lighting. Incandes- conduction studies. A cool limb also has the
cent lighting does not produce the same effect of slowing conduction, which is ap-
electrical interference and, thus, is preferred parent in the prolonged distal latency and
to fluorescent lighting for rooms in which slowed conduction velocity. A mathematical
nerve conduction studies are performed. formula is used in some electrophysiology
The amplitude of sensory NAPs depends laboratories to correct for differences in
on several factors. First, the amplitude is di- limb temperature, but it is more reliable to
rectly proportional to the number of axons warm a cool limb before performing the
that are depolarized. Second, the distance studies.
The fourth factor, the distance between

the recording electrodes and the nerve—
The amplitude decreases proportionally to
the square of the distance between the
recording electrode and the nerve. For most
standard sensory studies, this is not a critical
factor. However, in some obese patients or
those with exceptionally large hands or dig-
its, the increased distance between electrode
and nerve may affect the amplitude of the
response. In less commonly performed sen-
sory conduction studies, for example, lateral
and medial antebrachial cutaneous nerves
and the saphenous nerve, careful attention
to anatomical landmarks is required to en-
Figure 15-3. Sensory nerve action potential with Gl and
sure that the recording is made from the G2 separated by, top, 3.5 cm, and, bottom, 1 cm. Note
same location each time, because even the the decrease in amplitude. Bar markers indicate mea-
slightest movement can affect the distance surements for latency and amplitude.
between the electrode and the nerve and,
thus, affect the amplitude.
ing electrodes will alter the response ampli-
tudes noticeably. Because the amplitude of
HUMAN FACTORS the response is inversely proportional to the
square of the distance between Gl and the
Human error often affects and alters sensory nerve, an error of even a few millimeters will
NAPs. These are perhaps the easiest factors alter the amplitude markedly (Fig. 15-4).
to correct and to control. Measuring dis-
tance is a source of human error. The dis-
tances that are chosen affect the amplitude,
latency, and conduction velocity, and care-
ful attention has to be paid to measuring dis-
tances because an error of even a few mil-
limeters can change the conduction velocity
calculation. Recall that relatively short dis-
tances are used to calculate conduction ve-
locity. Therefore, standard positioning of
the limb and standard anatomical landmarks
need to be used in every study.
The interelectrode distance is often over-
looked. The Gl and G2 electrodes need to
be far enough apart to prevent phase can-
cellation and a reduction in the amplitude
of the sensory NAP12 (Fig. 15-3). However,
placing the electrodes too far apart increases
the problem with electrical interference by
reducing the effectiveness of the common-
mode rejection. The interelectrode distance
must be kept constant.
Placement of the recording and stimulat-
ing electrodes needs to be standardized. In
most nerve conduction studies, specific land- Figure 15-4. Sensory nerve action potential with, top to
bottom, Gl and G2 moved away from the nerve at 5 mm
marks are used for placing the stimulating increments. Note the decrease in amplitude and pro-
and the recording electrodes, and any devi- longation of the rise time. Bar markers indicate mea-
ation from these locations with the record- surements for latency and amplitude.
Also, stimulating the nerve from a distance near-needle stimulation, the monopolar
requires excessive current, which can cause needle serves as the cathode and the surface
excessive shock artifact, nonspecific excita- electrode, as the anode.
tion of other nerves, or direct muscle stim-
ulation, leading to problems with volume
conduction. Recording the Potential
Recording sensory NAPs requires appropri-
METHODS ate placement of the Gl and G2 electrodes.
When recording over the digits, ring elec-
Nerve Stimulation trodes are used that record the electrical po-
tentials circumferentially from the digit.
The sensory nerve should be stimulated with Standard small plate electrodes should be
a current that is sufficient to activate all the used in the limb. An interelectrode distance
sensory axons in the nerve but not to cause of 3.5-4.0 cm maximizes the sensory re-
overstimulation. This balance requires that sponse and minimizes the amount of electri-
the stimulator be as close to the nerve as cal interference. If the electrical background
possible. To initiate the study, the stimula- noise is excessive, assess the impedance of
tor is placed over the approximate location the electrodes. Any impedance mismatch in
of the nerve, with the cathode distally. The the electrodes causes problems with the
nerve should be depolarized with the cath- common-mode rejection, and the signals
ode, not the anode. Depolarization with the generated by background noise will be am-
anode distally can cause an electrical con- plified. If no satisfactory response can be ob-
duction block, called anodal block, of some tained, confirm that the location of the elec-
of the axons and result in a submaximal re- trodes is correct.
sponse. Also, if the anode is placed distally,
the distance measurements will not be ac-
curate. As the current is increased gradually, Averaging
a sensory NAP becomes apparent. At this
stage, slide the stimulator laterally to iden- Sensory potentials have much lower ampli-
tify where the response is maximal. At the tudes than motor responses; thus, the signal-
site of the maximal response, gradually in- to-noise ratio is much lower. Because of this,
crease the current until the amplitude the sensory responses are often affected by
reaches its maximum. the background, even with superior testing
If stimulation creates a large shock arti- technique. The low signal-to-noise ratio can
fact, confirm that the ground is in the ap- be improved by averaging, which can reduce
propriate location. Check the impedance of or eliminate random background noise. The
the recording and stimulating electrodes, improvement in the signal-to-noise ratio is
and if necessary, apply conduction paste to directly proportional to the square root of
improve the impedance values. After this has the number of responses averaged. This
been done, rotate the anode off the nerve means that improvement is greatest with the
while stimulating it to decrease the shock first few responses that are averaged and lit-
artifact. tle more is gained after averaging the first
If an acceptable response cannot be ob- four or five responses.
tained in large limbs, consider near-nerve
stimulation with a monopolar needle elec-
trode. Near-nerve stimulation has several ad- MEASUREMENTS
vantages. The largest amount of impedance
arises from transcutaneous stimulation. Clinically, the most relevant measurements
Thus, placing the needle within the subcu- are amplitude, distal latency, and conduc-
taneous tissue eliminates the transcutaneous tion velocity. Amplitude is measured as a
resistance. Also, placing the needle much peak-to-peak amplitude, from the nadir of
closer to the nerve allows supramaximal the initial positivity to the peak negativity
stimulation at a lower level of current. With (Fig. 15-5). Because amplitude decreases
comparisons. This is not optimal, because it

introduces a source of human error.
Conduction velocity is the last variable
that is assessed clinically. Conduction veloc-
ity is obtained by dividing the distance be-
tween two stimulus sites by the conduction
time. The onset latencies are measured and
subtracted to obtain conduction time (Fig.
15-7). The resulting time represents con-
duction time. In this instance, the onset la-
tencies are chosen because they represent
the fastest conducting fibers, and the calcu-
lated conduction velocity reflects the speed
of conduction in these fibers (Fig. 15-6).
The normal values used at the Mayo Clinic
are listed in Table 15—1.
SUMMARY
Figure 15-5. Amplitude measurement of a potential, Nerve conduction studies are an invaluable
with, top, and without, bottom, an initial positive peak.
addition to clinical electrophysiology test-
ing. Sensory NAPs are a sensitive and spe-
cific measure of function in the peripheral
substantially with increasing length of the sensory pathways. These studies confirm
segment, the amplitude that is recorded and whether large myelinated axons are affected
compared is the distal amplitude. by an underlying abnormality. When an area
The distal latency that is recorded and that is affected clinically is tested, nerve con-
compared is the peak latency, not the onset la- duction studies can help to distinguish be-
tency of the response. Peak latency is the time tween a preganglionic (that is, root level or
from the stimulus to the peak negative de- higher) and a postganglionic (that is, pe-
flection. Onset latencies are more variable ripheral) process.
and difficult to measure (Fig. 15-6). The dis- Sensory NAPs are small and technically
tances for the distal response should be kept difficult to record; therefore close attention
fixed so that the latencies can be compared has to be given to proper technique, in-
directly with normal values. If the distance cluding minimizing electrical interference
is variable, then the latency must be divided and using proper stimulating and recording
by the distance (generating a figure analo- methods. When stimulating a nerve, stimu-
gous to a conduction velocity) to allow for late as close to the nerve as possible to allow
a supramaximal response but to minimize
the electrical artifacts. When recording, ad-
here to fixed anatomical landmarks, main-
tain proper interelectrode distance, and be
attentive in measuring stimulating and record-
ing distances. If satisfactory responses can-
not be obtained with transcutaneous stimu-
lation, consider near-needle stimulation in
the appropriate setting. Because the am-
plitudes are low, averaging responses en-
hances the signal and eliminates the random
ambient electrical activity. The measure-
Figure 15-6. Onset latency at 2.5 ms and peak latency ments that are most relevant clinically and
at 3.2 ms. should be noted on all studies are the distal
Figure 15-7. Measurement of conduction velocity in a proximal nerve segment. The difference in latencies between
the proximal and distal stimulation sites is divided by the distance between the two sites. Latencies are measured
to the initial negative deflection (onset latency).
Table 15-1. Nerves Used for Routine Sensory Nerve Conduction Studies,
With Representative Normal Values for Amplitude, Distal Latency, and
Conduction Velocity Over Proximal Segments
Proximal
Conduction
Amplitude Velocity Distal Latency
Nerve Stimulate Record (/*V) (m/s) (ms)
Median Palm Wrist > 50 < 2.3
Elbow > 55
Median Wrist Digit2 > 14 < 3.6
Elbow >56
Ulnar Palm Wrist > 15 < 2.3
Elbow > 56
Ulnar Wrist Digit5 > 10 < 3.1
Elbow > 56
Superficial radial Forearm Dorsum of > 20 < 2.8
hand
Elbow >50
Sural Calf Lateral ankle >6 <4.5
Proximal calf >40
Medial plantar Plantar foot Medial ankle <7 <4.0
179
amplitude, distal peak latency, and conduc- 5. Wilbourn AJ. Sensory nerve conduction studies.
tion velocity of the nerve. It must be em- J Clin Neurophysiol 11:584-601, 1994.
6. Albers JW. Clinical neurophysiology of generalized
phasized that careful technique is needed to polyneuropathy. J Clin Neurophysiol 10:149-166,
produce reliable and valid results. The in- 1993.
terpretations drawn from the results of a 7. Levin KH. L5 radiculopathy with reduced superfi-
study can be only as good as the information cial peroneal sensory responses: intraspinal and ex-
on which they are based. traspinal causes. Muscle Nerve 21:3-7, 1998.
8. Bromberg MB, Jaros L. Symmetry of normal motor
and sensory nerve conduction measurements. Mus-
cle Nerve 21:498-503, 1998.
REFERENCES 9. Dumitru D, Dreyfuss P. Dermatomal/segmental so-
matosensory evoked potential evaluation of L5/S1
1. Handwerker HO. Sixty years of C-fiber recordings unilateral/unilevel radiculopathies. Muscle Nerve
from animal and human skin nerves: historical 19:442-449, 1996.
notes. Prog Brain Res 113:39-51, 1996. 10. Bland JD. A neurophysiological grading scale for
2. Donofrio PD, Albers JW. AAEM minimonograph carpal tunnel syndrome. Muscle Nerve 23:1280-
#34: Polyneuropathy: classification by nerve con- 1283, 2000.
duction studies and electromyography. Muscle 11. Dutra de Oliveira AL, Barreira AA, Marques WJr.
Nerve 13:889-903, 1990. Limitations on the clinical utility of the ulnar dor-
3. Buchthal F, Rosenfalck A. Sensory potentials in sal cutaneous sensory nerve action potential. Clin
polyneuropathy. Brain 94:241-262, 1971. Neurophysiol 111:1208-1210, 2000.
4. Melendrez JL, MacMillan LJ, Vajsar J. Amplitudes 12. Wee AS, Ashley RA. Effect of interelectrode record-
of sural and radial sensory nerve action potentials ing distance on morphology of the antidromic sen-
in orthodromic and antidromic studies in children. sory nerve action potentials at the finger. Elec-
Electromyogr Clin Neurophysiol 38:47-50, 1998. tromyogr Clin Neurophysiol 30:93-96, 1990.
Chapter 16
SOMATOSENSORY EVOKED
POTENTIALS
J. Clarke Stevens
METHODS Dermatomal Somatosensory Evoked

Averaging Potentials
Factors That Affect the Amplitude of the Trigeminal Nerve Somatosensory Evoked
Evoked Response Potentials
Peak Nomenclature Standard Methods and Recording Montages
Montage and Near-Field and Far-Field SOMATOSENSORY EVOKED POTENTIAL
Potentials INTERPRETATION
Neuroanatomy and Origin of Peripheral and Variables Affecting Latencies and Amplitudes
Central Waveforms Localization
Median and Ulnar Mixed Nerve Amplitude
Somatosensory Evoked Potentials CLINICAL APPLICATIONS
Posterior Tibial Mixed Nerve Somatosensory Disorders of the Peripheral Nervous System
Evoked Potentials Disorders of the Central Nervous System
Cutaneous Nerve Stimulation SUMMARY
Somatosensory Evoked Potentials
Somatosensory evoked potentials (SEPs) are pre- vide objective evidence of central nervous
synaptic and postsynaptic responses recorded system (CNS) dysfunction when sensory
over the limbs, spine, and scalp following the symptoms are vague and the findings on
stimulation of peripheral nerve trunks and neurologic examination are normal or of
cutaneous nerves. Somatosensory evoked uncertain significance. Patients who are con-
potentials are analogous to standard pe- vinced their symptoms are caused by multi-
ripheral sensory nerve conduction studies. ple sclerosis are reassured by normal SEP
The main value of SEPs is to provide a mea- recordings.
surement of sensory conduction in proximal Spinal cord dysfunction due to causes
peripheral nerves, the spinal cord, and the other than multiple sclerosis is the second
brain. Although evoked potentials can be most common indication for recording
elicited by physiologic stimuli such as a fin- SEPs. The localization of sensory symptoms
ger tap or tendon stretch, electric stimula- to a proximal peripheral nerve, the spinal
tion produces consistently higher amplitude cord, or a cerebral site is helpful in diag-
evoked potentials and is the only stimulus nosis and can suggest where an imaging
useful for clinical application. study may show abnormality. Somatosensory
Somatosensory evoked potentials are used evoked potentials are normal if sensory
most frequently in the evaluation of patients complaints are caused by a conversion reac-
with suspected multiple sclerosis to obtain tion or malingering. Magnetic resonance
evidence for a second lesion. They also pro- imaging (MRI) of supratentorial structural
181
lesions is so sensitive and reliable that SEPs problem at these locations. However, record-
are rarely indicated for investigation of this ing over the lumbar or cervical spine is dif-
area of the CNS. ficult because of the motor unit activity of
the paraspinal muscles and the distance
from the generators. Audio monitoring of
METHODS all channels is essential; however, the spine
derivations are most important, because they
Averaging are usually the noisiest channels. Muscle ar-
tifact in the scalp leads is rarely a problem.
The main technical limitation to recording Sedation of the patient is helpful, especially
SEPs is that their amplitude is low compared patients who are tense or spastic; diazepam
with the noise of motor activity, movement is routinely given for sedation unless it is
artifacts, the electrocardiogram, the electro- contraindicated. Carefully monitored se-
magnetic activity in the environment, and dated infants and children sometimes sleep
the electroencephalogram (EEG). Gener- during the test.
ally, 500-2000 stimuli are necessary to dis-
play well-defined, reproducible waveforms of
1-10 /xV. Averaging summates activity that is ELECTRIC ARTIFACT
time-locked to the stimulus trigger, while The two main sources of electric artifact in
gradually subtracting random background recordings of SEPs are stimulus artifact and
noise. If the noise is excessive, increasing the 60 Hz alternating current transmitted to the
number of stimuli averaged does not help to amplifier by the machine used to record the
extract a better signal because the signal-to- evoked potential or through electromag-
noise ratio is too small. For example, artifact netic radiation. Stimulus artifact can be de-
in the form of large quasirandom triphasic creased by using a stimulus-isolation device
motor unit potentials may produce continu- and a fast-recovery amplifier, by maintaining
ous variation in the averaged waveform that proper orientation and contact of the stim-
does not improve with continued averaging. ulating electrodes, and by avoiding higher
If the noise becomes time-locked with the than necessary stimulus intensity. Maintain-
signal, it will be enhanced and may be mis- ing recording electrode impedance less than
taken for a physiologic signal. Sixty cycle 5000 H by cleaning the skin and proper
electrical artifact can be reduced by using a grounding eliminates most 60 Hz noise from
stimulation rate that is not a factor of 60 (for the SEP. If different types of electrodes, for
example, 2.1 Hz). Averagers deal with inter- example, surface and needle electrodes, are
mittent artifacts by rejecting sweeps that con- used at recording and reference sites, an im-
tain waveforms exceeding the fixed maximal pedance mismatch is created, thus amplify-
amplitude. This helps to decrease or to elim- ing 60 Hz interference.
inate most artifacts. In addition, inspection
of the signal being averaged allows the tech-
nician to interrupt averaging if excessive ar- FILTER SETTINGS
tifact occurs. Averaging is resumed after the
technical problem has been corrected. Correct filter settings decrease noise without
reducing the waveforms of interest. A low-
frequency filter setting of 30 Hz and a high-
Factors That Affect the Amplitude frequency setting of 3 kHz are usually satis-
of the Evoked Response factory. Restricting low frequencies with a
filter setting of 150 reduces 60 cycle artifact
MUSCLE ARTIFACT and may be useful in some situations. It also
allows better visualization of certain peaks
Muscle artifact can be controlled by having (for example, Nil). However, the 150 kHz
the patient relax in a reclining chair or bed. setting has the disadvantage of reducing the
Because evoked responses recorded at the amplitude of most peaks (for example, N13)
elbow, Erb's point, or the knee have a high and slightly shortening peak latencies. Use
amplitude, muscle activity is not usually a of a 60 Hz "notch" filter is not recommended
Somatosensory Evoked Potentials 183
because SEPs in this range contain impor-

tant physiologic information.
STIMULATION VARIABLES
The stimulus applied should be of sufficient
intensity to produce a small visible twitch of
the muscle. Stimulus intensities higher than
this are painful (making it difficult for the
patient to relax) and do not increase the am-
plitude of the SEP. Intensities twice motor
threshold are sometimes necessary to
achieve maximal central responses. If the
small foot muscles are atrophied, a twitch
may not be visible except at high stimulus
intensities. In this case, use a stimulus that Figure 16-1. Tibial nerve stimulation. Upper, Unilateral
and lower, bilateral stimulation at the same amplifica-
is 2—2.5 times sensory threshold. If sensory tion in the same patient. Note that bilateral stimulation
loss is marked, higher stimulus intensities enhances the amplitudes of N22, N30, and P38.
can be used without producing patient dis-
comfort and may be needed to exceed
threshold. In the upper limb, stimulus rates subcortical potentials and allow central con-
of 2-5 Hz are well tolerated by most patients, duction time or at least the absolute scalp la-
but rates of 1-2 Hz are used in the lower tencies to be estimated. The response to bi-
limb. A slower rate of 0.5 Hz may be required lateral stimulation may not be enhanced if
to avoid a flexor withdrawal reflex in a spas- the peripheral and central nerve conduction
tic limb. Rates greater than 10 Hz may cause velocities are not similar on the two sides. Bi-
an increase in the latency and a decrease in lateral median nerve stimulation is rarely
the amplitude of some components. necessary.
NERVE STIMULATED
Peak Nomenclature
In the upper limb, the highest amplitude
SEPs are obtained with stimulation of the ul- It is a function of differential amplifiers that,
nar and median nerves. In the lower limb, if both grids are active, an upward deflection
stimulation of the tibial nerve produces may be produced by a negativity in grid 1 or
more reliable SEPs than stimulation of the a positivity in grid 2 and a downward de-
peroneal nerve. The lowest amplitudes are flection may be produced by a positivity in
obtained with stimulation of cutaneous or grid 1 or a negativity in grid 2. Therefore,
dermatomal nerves. After cutaneous nerve avoid associating the polarity of a particular
stimulation, spinal potentials usually are ab- peak with the actual polarity of the genera-
sent in normal subjects; thus, disease in the tor or assuming that grid 1 is the most active
CNS cannot be distinguished from that in electrode. It is customary to label upward de-
the peripheral nervous system. flections as N (negative) and downward de-
flections as P (positive) when recording with
UNILATERAL AND BILATERAL the grid-1 negative-up convention. The num-
STIMULATION ber following the N or P refers to the aver-
age latency at which the particular potential
Tibial spinal, brain stem, and cerebral is recorded in normal subjects. Thus, with a
evoked potentials increase in amplitude with bipolar montage, the negative potential that
bilateral stimulation (Fig. 16-1). Therefore, is recorded over the brachial plexus ap-
bilateral stimulation is helpful when the tib- proximately 9 ms after the median nerve is
ial scalp SEPs are absent or poorly formed stimulated at the wrist is termed N9 (Table
or spine potentials are absent. In this case, 16-1). However, peak nomenclature has not
bilateral stimulation may produce adequate been standardized, and slightly different
Table 16-1. Standard Methods for Recording Somatosensory Evoked Potentials

Median or Ulnar Nerve Tibial Nerve
Stimulation Bipolar at wrist, cathode Bipolar at ankle, cathode
proximal proximal
Standard recording, Bipolar at elbow: N5 Bipolar at knee: N8
montage: potential EPi-EPc:N9 Ll-ICc: N22
C5S-Fz: Nil, N13, N14 C5S-Fz: N30
C3' (C4')-Fz: N20/P25 Cz-Fz: N33/P38
Ground: lead plate arm Ground: lead plate leg
Optional recording Anterior neck-C5S: PI3 C3'-C4': N33/P38
C3' (C4')-noncephalic:
P9, P13, P14 far-field
N20/P25 parietal
F3' (F4')-noncephalic:
N22/P30 frontal
Machine settings Stimulus rate, 1-5 Hz Stimulus rate, 0.5-2 Hz
Stimulus intensity, slight Stimulus intensity, slight
muscle twitch muscle twitch
Amplifier sensitivity Amplifier sensitivity
10 fjiV spine and scalp 10 fjuV spine and scalp
20 pV peripheral 20 /i-V peripheral
Analysis time, 50 ms Analysis time, 80 ms
Filter, 30-3000 Hz Filter, 30-3000 Hz
Number averaged, 500 Number averaged, 500
Measurement Height and F-distance Height and F-distance
Limb temperature Limb temperature
Waveform amplitude Waveform amplitude
N5, N9, N13: onset-peak N8, N22, N30: onset-peak
N20: N20-P25 N33: N33-P38
P25: P25-N35 P38: P38-N46
EPi-EPc, Erb's point ipsilateral to contralateral; C5S, spine of C5 vertebra; Ll-ICc, spine of LI vertebra to con-
tralateral iliac crest.
numbering systems are used to identify the recording point. If the recording electrode
same evoked potential. Evoked potentials is close to the generator, a second electrode
are measured only when two superimposed that is a long distance away will act as an in-
averages reveal consistent responses. If it is different electrode and a high-amplitude
clear that reproducible waveforms are pres- potential will be obtained. If the recording
ent when looking at the two averages, view- electrode is far from the source, another
ing a single combined average may make electrode at a similar distance will be almost
it easier to identify and to measure the as active as the recording electrode itself, in
waveforms. which case the potentials of the two elec-
trodes will cancel in a differential amplifier
so that little or no potential will be recorded.
Montage and Near-Field and Another important principle in the
Far-Field Potentials recording of SEPs is that all electrodes—
regardless of where they are placed on the
It is important to remember that in a vol- body—are relatively active. This is analogous
ume conductor the amplitude of the poten- to the electric activity of the heart, which also
tial is related inversely to the square of the can be recorded anywhere on the body. Ide-
distance between the generator and the ally, one electrode should be as close to the
generator as possible and the other elec- referential montage, such as the scalp to
trode, as far away as possible to obtain the Erb's point, preferentially detects FFPs. The
maximal potential difference between the advantage of using far-field recordings is
electrodes. However, increasing electrode that information from many different levels
distance also increases noise, especially mus- of the nervous system can be obtained from
cle artifact, and may introduce additional a single recording montage. The disadvan-
generators. Therefore, it is necessary to com- tage is the excess noise introduced by long
promise between the cancellation effect of interelectrode distances, which makes it dif-
closely spaced electrodes and the noise in- ficult to obtain accurate latency measure-
troduced by long distance between record- ments. Averaging several thousand responses
ing electrodes. If the generator is proximal in a cooperative and young, thin patient may
to the shoulder or hip, moving the reference produce readable responses.
electrode distally along a limb does not im- With antidromic stimulation of the me-
prove the signal. dian and radial nerves, a bipolar and refer-
As in peripheral nerve conduction studies, ential montage shows stationary waves that
a bipolar electrode montage that detects an are generated in the hand and fingers at
approaching or departing depolarization points where a change in the shape or vol-
records a positive waveform, a positivity, and ume of the limb (volume conductor) oc-
the electrode overlying the depolarization curs.2 Previously, the FFP was regarded as
records a negative waveform, a negativity. In a monophasic positivity reflecting the ap-
montages in which both electrodes overlie proaching wave front of depolarization.
the same afferent pathway, the two elec- However, stationary activity from a moving
trodes frequently record potentials of oppo- source often contains a major negative com-
site polarity and summate to produce a ponent that sometimes exceeds the preced-
recording of higher amplitude. The situa- ing positivity.3 Because the size and shape of
tion is different when the reference elec- the volume conductor change in three di-
trode is remote from the generator. For mensions, the charge density may change
example, median subcortical potentials gen- and, thus, a traveling dipole moving through
erally are positive in polarity when a scalp- the volume conductor will create voltage dif-
noncephalic reference is used, whereas the ferences. The polarity of the FFP may be re-
same components have a negative polarity in lated to the size of the volume conductor
the C5S-scalp montage. Therefore, the po- that the traveling wave leaves and enters.
larity of evoked potentials varies with the After stimulation of the median nerve, the
electrode montage used. Also, potentials FFP (P9) is frequently recorded in a scalp-
with the same latency recorded in different noncephalic montage, with a latency similar
montages do not necessarily have the same to that of the traveling wave recorded from
generators. Although the waveforms recorded the median nerve in the brachial plexus. The
by neck and scalp electrodes sometimes ap- P9 potential consists of two separate poten-
pear to be a single peak, they usually are the tials: one arises at the point where the arm
result of the combined activity of several dif- meets the trunk and the other, as it leaves
ferent generators, not all of which can be the trunk and reaches the root of the neck.
recorded as separate waveforms. The presence of stationary waves at several
Nerve action potentials that travel along points along the median nerve indicates that
nerves or fiber tracts are called traveling some potentials arise from changes in the
waves. Potentials that remain localized in ar- volume conductor rather than from con-
eas of nuclei or synapses are called station- duction through synapses or along myeli-
ary waves. Near-field potentials (NFPs) repre- nated fiber tracts.
sent a propagating nerve action potential
that is recorded as it passes under the record-
ing electrodes. The recording electrodes Neuroanatomy and Origin of
3 cm apart that are used in routine nerve Peripheral and Central Waveforms
conduction studies primarily record NFPs.
The term far-field potential (FFP) refers to sta- The neuroanatomy of the sensory pathways
tionary potentials generated by nerve action has been known for years, but the exact ori-
potentials distant to the recording site.1 A gin of many of the components of the SEP
is still not clear and it is evident that some

peaks have overlapping generator sources.4
The potentials recorded with low current
stimulation of a mixed nerve represent ac-
tivity in the proprioceptive system. This
activity is conducted peripherally by large-
diameter, myelinated, fast-conducting cuta-
neous and muscle afferents and is con-
ducted centrally by the dorsal column-
medial lemniscus and the spinocerebellar
pathways. There are numerous collaterals to
the gray matter at all levels. The axons of
third-order neurons go from the ventral pos- Figure 16-2. Normal 8-channel median somatosensory
terolateral nucleus of the thalamus to the evoked potentials in a 13-year-old child. AC, anterior
primary somatosensory cortex. Stimulation cervical electrode placed just above the thyroid carti-
lage.
of cutaneous nerves or dermatomes activates
large cutaneous afferents, but the SEPs have
a lower amplitude than those produced by trode in the same location contralaterally
stimulation of mixed nerves, because fewer represents orthodromic activity in sensory
fibers are excited. Also, the peak latencies fibers and antidromic activity in motor fibers
obtained with the stimulation of sensory passing through the brachial plexus. The P9
nerves are slightly longer than those ob- far-field potential recorded in a scalp-non-
tained when a mixed nerve is stimulated, be- cephalic montage is probably caused by an
cause fast-conducting group la muscle af- abrupt change in current flow as the action
ferents are not present in a sensory nerve. potential passes from the arm to the trunk.
Lesions affecting sensory modalities trans- Stimulation of the median nerve activates cu-
mitted by small-diameter sensory fibers or by taneous sensory fibers that enter the spinal
central pathways in the ventral half of the cord through the upper and middle trunks
spinal cord usually do not produce SEP ab- and posterior roots of C6 and C7. An-
normalities. Pain-related SEPs have been tidromic median motor and spindle afferent
recorded by stimulation of small-diameter potentials pass through the medial cord and
A5 pain fibers with a carbon-dioxide laser lower trunk of the plexus to enter the spinal
and high-intensity electric stimulation, but cord through the anterior and posterior
the acceptance of these techniques has been roots of C8 and Tl. With ulnar nerve stim-
limited/'6 ulation, activity is confined to the C8 and Tl
segments. Occasionally, the ulnar N9 po-
tential is difficult to record in normal sub-
Median and Ulnar Mixed Nerve jects older than 60 years. Most of the po-
Somatosensory Evoked Potentials
Following stimulation of the median or ul-
nar nerve at the wrist, activity can be
recorded at the elbow, Erb's point, cervical
spine, and scalp (Figs. 16-2 and 16-3). The
N5 potential recorded with a bipolar elec-
trode at the elbow represents the propagat-
ing nerve action potential in the median
or ulnar nerve. This potential indicates
whether the stimulation is adequate and pro-
vides an estimate of peripheral conduction
velocity.
The N9 potential recorded with an elec-
trode at Erb's point (2 cm superior to the Figure 16-3. Normal 8-channel ulnar somatosensory
midpoint of the clavicle) referred to an elec- evoked potentials.
tential is generated in sensory fibers, because tend to be abnormal in cases of brain stem
the N9 potential is prominent in patients dysfunction.14'15 The N13-P14 interpeak la-
with avulsion of the roots of the brachial tency assesses cervical cord-brain stem con-
plexus. Conversely, if a peripheral sensory duction time.
deficit is substantial, the N9 peak may rep- N18 is a broad, subcortically generated
resent antidromic activity in motor, not sen- FFP best recorded in an ipsilateral scalp-to-
sory, fibers. noncephalic montage. Evidence points to
An electrode over the spine of C5 or C7 this potential being postsynaptic activity aris-
referred to Fz is the most common montage ing from several generator sources in the
for recording activity arising from the cervi- brain stem.14'16 Studies of patients with brain
cal spine and brain stem. This montage stem lesions suggest that N18 reflects exci-
records three negative potentials: Nil, N13, tatory postsynaptic potentials evoked by dor-
and N14. The Nil potential is likely a pre- sal column axons in the cuneate nucleus or
synaptic traveling wave that arises from ac- accessory inferior olive (or both) or, possi-
tivity near the root entry zone of C6 and C7 bly, presynaptic afferent depolarization in
and action potentials ascending in the dor- the cuneate nucleus.17"19
sal columns. Nil is also referred to as the Several small potentials occasionally are
dorsal column volley (DCV). visible on the rising phase of the N20 po-
The evidence is convincing that N13 is a tential and may originate in thalamic relay
standing dipole that is negative when re- nuclei and the thalamocortical radiation.
corded over the posterior neck, and positive Also it has been suggested that they are gen-
when recorded prevertebrally.7"10 N13/P13 erated in the cerebral cortex in a closely sit-
is a dorsal horn postsynaptic potential that uated polysynaptic network.20
is elicited by collaterals of the primary af- Scalp potentials with a latency of 20 ms
ferent fibers in the lower cervical cord. A sec- and longer reflect the postsynaptic poten-
ond potential with the same latency occurs tials generated by neurons in the hand area
at the level of the cervicornedullary junction; of the primary somatosensory cortex in re-
it possibly arises from the cuneate nu- sponse to the afferent thalamocortical vol-
cleus.11"13 The C5S-Fz montage records a ley. N20 is superimposed on the widespread
large N13 potential that is likely an average bilateral N18 potential.21 There is disagree-
of the standing dorsal cord potential and the ment about the exact identity of cortical
P13/P14 FFP recorded by the scalp elec- generators. Whether the N20/P25 peaks are
trode. The N13 potential can be recorded in mediated by separate thalamocortical pro-
all normal subjects, whereas the Nil peak is jections or by sequential activity in one path-
recorded in approximately 75% of normal way is uncertain. Somatosensory evoked po-
subjects and N14 in approximately 15%- tentials may occur with selective involvement
20%. Separation of the spinal N13/P13 dor- of early or late cortical peaks. This phe-
sal horn potential from PI 4 is facilitated by nomenon in combination with the effect of
recording from the anterior neck at the su- •tourniquet-induced ischemia of peripheral
perior border of the thyroid cartilage with a nerve on SEPs has led to the speculation that
contralateral elbow reference or with a C5S- the N20 peak is related to vibration and po-
anterior neck montage. Loss of N13/P13, sition sense (large myelinated fibers). In
but not P14 and N20, may occur when le- contrast, the N35 peak is attributed to pain
sions interrupt collateral axons to dorsal and temperature sense (small myelinated
horn neurons without affecting fibers as- fibers).22
cending in the dorsal columns. The N20/P25 complex, as recorded with
P14 (scalp-to-noncephalic montage) is a the bipolar C3' or C4'-Fz montage, may be
subcortically generated FFP with a wide- an average of independent posterior frontal
spread scalp distribution. In the C5S-Fz (P22/N30) and parietal (N20/P30) genera-
montage, N14 is seen sometimes as a small tors, each of which can be eliminated selec-
negative potential on the falling phase of tively. This may go unnoticed in the bipolar
N13. N14/P15 potentials probably arise in montage because potentials are recorded
the caudal medial lemniscus because they from either the parietal or the frontal elec-
are preserved in cases of thalamic lesion and trodes. When a cortical lesion is suspected,
it has been suggested that recordings be quentially from area 3b to area 1. The scalp
made from a parietal site and a prerolandic potentials recorded depend on the relative
site, using an earlobe or noncephalic refer- strengths of the tangential and radial gen-
ence for both.23 In practice, we do not find erators. Scalp frontal electrodes record pri-
this necessary, because imaging studies are marily P20 and N30, the central electrodes
effective in localizing cortical lesions. near C3 or C4 record primarily P25 and N35,
Conclusions different from those above and the parietal electrodes record primarily
have been drawn from another analysis of N20 and P30.
the cortical generators after median nerve Little is known about the generators in-
stimulation.24 Depolarization of a sheet of volved in subsequent SEP peaks. They may
cortical pyramidal cells produces an extra- represent the sequential activation of corti-
cellular potential field of one polarity at and cal areas through corticocortical or thalamo-
distal to the apical dendrites and a potential cortical connections.
field of opposite polarity at and proximal to
the cell bodies.25
Neurons at the cortical surface, with a ra- Posterior Tibial Mixed Nerve
dial orientation, produce a potential field of Somatosensory Evoked Potentials
one polarity at the cortical surface and a po-
tential field of the opposite polarity in the Following stimulation of the tibial nerve at
lower layers of the cortex and white matter. the ankle, the peripheral nerve action po-
Neurons in the wall of the central sulcus, tential recorded at the popliteal fossa is la-
with a tangential orientation, produce a po- beled N8 (Fig. 16-4). An electrode over ver-
tential field of one polarity on one side of tebra LI referred to the iliac crest records a
the sulcus and the opposite polarity on the negative, sometimes bifid, potential desig-
other side. The tangential + somatosensory ra- nated N22. N22 represents postsynaptic po-
dial theory postulates that Brodmann areas 3b tentials generated in the dorsal horn of the
and 1 contain generators. The cortex in the spinal cord, analogous to the stationary
anterior wall of the postcentral gyrus gener- N13/P13 potential recorded over the neck
ates the primary evoked response P20/N30, with stimulation of the median nerve. An ini-
and, a few milliseconds later, the cortex at tial, small, rarely recorded negative peak,
the crown of the postcentral gyrus (mainly N18, is a traveling wave that represents con-
Brodmann area 1) generates the primary duction through the cauda equina and
evoked response P25/N35. Intracranial root.27 It is best displayed with an LlS-to-L4S
recordings of P25 are maximal from a small montage; however, because its amplitude is
region in the medial portion of the hand small, the potential cannot be recorded in
area of the somatosensory cortex. P25 may many patients.
reflect patchy activation of discrete regions
of area I.26
The complexity of the potential fields
recorded from the surface of the so-
matosensory cortex or the scalp reflects the
different orientation of the generators in ar-
eas 3b and 1 and their asynchronous activa-
tion. The assumption is that independent
frontal and parietal generators do not exist
and electrodes in these locations are record-
ing opposite ends of the radial and tangen-
tial dipole generators located in areas 3b
and 1. The generators in area 3b are acti-
vated first because the thalamocortical fibers
that project to area 3b may have faster con-
duction velocities than those projecting to
area 1. Also, area 3b projects to area 1, sug- Figure 16-4. Normal 8-channel left tibial somatosen-
gesting that information is processed se- sory evoked potentials in a 27-year-old woman.
The N30 potential recorded over the cer- peripheral and central pudendal nerve
vical spine (G5S-Fz montage) represents ac- pathways that may be associated with urinary
tivity in the fasciculus gracilis, spinocerebel- and bowel incontinence and impotence. In
lar pathways, and possibly the gracile men, pudendal nerve SEPs are obtained with
nucleus.28'2^ It is difficult to record in many stimulation of the penis by bipolar ring elec-
subjects because of muscle artifact and pro- trodes.38 Scalp evoked potentials are easily
gressive dispersion of the ascending volley. obtained, but it is difficult to record re-
Stimulation of the posterior tibial nerve bi- sponses from the lumbar region, making it
laterally is frequently helpful in eliciting impossible to localize lesions in the periph-
the tibial lumbar and cervical responses eral and central pudendal nerve pathways.
(Fig. 16-4).
N34 (Fz-C5S montage) is a subcortically
generated FFP analogous to N18 following Dermatomal Somatosensory
stimulation of the median nerve. It may re- Evoked Potentials
flect postsynaptic activity from many gener-
ator sources in the brain stem and, perhaps, Dermatomal stimulation is used occasionally
thalamus. to assess function of the lumbosacral and
The activity of the foot area in primary so- cervical nerve roots. Stimulation sites are the
matosensory cortex is ascribed to P38 (also thumb (C6), adjacent sides of the index and
known as P37).30 This potential is usually middle fingers (C7), little finger (C8), the
maximal somewhere between the midline dorsal surface of the foot between the first
and the centroparietal scalp locations, con- and second toes (L5), and the lateral side of
tralateral to the simulated leg. The primary the foot (SI). Stimulation sites and normal
component may consist of at least two values are available for the cervical, thoracic,
dipoles.31 Because the orientation of the di- and lumbosacral levels.34
pole inside the longitudinal fissure is vari-
able, P38 sometimes is maximal over the ip-
silateral scalp. This is known as paradoxical Trigeminal Nerve Somatosensory
localization. To be certain that P38 is absent,
record from the ipsilateral scalp as well as Evoked Potentials
from the usual midline location.
Somatosensory evoked potentials have been
reported in response to electric or mechan-
Cutaneous Nerve Stimulation ical stimulation of the trigeminal nerve.
Somatosensory Evoked Potentials These recordings are difficult to obtain be-
cause of shock artifact and contamination of
Stimulation of cutaneous nerves such as the response by muscle artifact and the blink
the sural, superficial peroneal, and lateral reflex. 5
femoral cutaneous nerves in the lower ex-
tremity and the digital, superficial radial,
and other nerves in the upper extremity Standard Methods and
readily elicits a scalp SEP in normal subjects. Recording Montages
However, the amplitude of the potentials is
much smaller than those obtained with The montages used in Mayo neurophysiol-
mixed nerve stimulation, and responses are ogy laboratories to study SEPs are listed in
not obtained over the spine. Cutaneous Table 16-1. Helpful guidelines for conduct-
nerve stimulation is used (1) to assess the ining SEP studies are available from the Amer-
tegrity of specific cutaneous nerves that are ican Association of Electodiagnostic Medi-
not readily studied with conventional nerve cine.36 For routine recordings, we prefer an
conduction study techniques, (2) to evaluate Fz reference.37 Surface electrodes are used
isolated root function, and (3) to assess for stimulation and recording. Scalp elec-
patchy numbness for medical-legal reasons.32 trodes are fixed with collodion, and spinal
Stimulation of the pudendal nerve is help- and Erb's point electrodes are taped in
ful in the evaluation of disturbances in the place. Stimulating and peripheral recording
electrodes are fixed in place with an elastic upon for interpretation even though the ab-
strap. Bipolar electrodes have a fixed inter- normalities are nonlocalizing.
electrode distance of 35 mm. Electrolyte gel A low temperature of the limb decreases
is applied to all electrodes, and impedance peripheral nerve conduction velocity and
is maintained less than 5 kfl. Despite differ- prolongs the latency of spinal and cortical
ences between electrophysiology laborato- evoked potentials. Therefore, it is necessary
ries in the number of channels and record- to monitor limb temperature to avoid errors.
ing montages used, the major recognized If the temperature of the arm is less than
potentials are consistent. When SEP laten- 32°C and that of the leg, less than 30°C, the
cies are prolonged, a motor conduction limbs should be warmed. However, central
study in the arm or leg (or both) is per- conduction velocity is affected only if hy-
formed to check for slowing of peripheral pothermia is profound. To assist with inter-
nerve conduction. pretation, median and tibial nerve conduc-
tion studies are performed if the SEPs are
abnormal. For example, a peripheral neu-
SOMATOSENSORY EVOKED ropathy can markedly affect the absolute la-
POTENTIAL INTERPRETATION tencies and morphology of evoked poten-
tials. Sedation given to reduce muscle
Examine the record to determine whether artifact may allow the patient to sleep dur-
all the normally appearing components are ing the test, but it can mildly prolong the
present. The absence of a main ulnar or me- scalp latencies.
dian SEP component almost always indicates Somatosensory evoked potential latencies
an abnormality. The absence of a waveform and amplitudes are affected by age. Values
that is easily recorded on the contralateral in children do not reach those of adults un-
side also indicates an abnormality. Occa- til age 8 years. In older age groups, there is
sionally, the lumbar and cervical responses a small decrease of peripheral sensory nerve
following tibial nerve stimulation are absent conduction and amplitude, which is most
in normal subjects and frequently absent in marked distally. According to one study,
older and obese subjects, particularly if they median nerve central conduction time
have difficulty relaxing. The lack of super- (N13-N20) was constant between the ages
imposable tracings at the lumbar and cer- of 10 and 49 years, increased by 0.3 ms be-
vical levels often represents a technical tween the 5th and 6th decades, and then re-
limitation rather than an abnormality. Sub- mained stable in normal subjects up to 79
cortical or peripheral potentials may be low years old.38 Mild prolongation of N9-N13
in amplitude or absent, but because of cen- and N11-N13 transit times has been found
tral amplification and several parallel central in comparing subjects 15-39 years old with
pathways, a relatively normal scalp response those 40—60 years old.39
may still be obtained. Avoid making state-
ments about pathologic conditions, because
disease-specific changes are not observed Localization
with SEP studies.
For purposes of clinical interpretation, SEP
waveforms are assumed to represent the se-
Variables Affecting Latencies quential activation of ascending levels of the
and Amplitudes somatosensory pathway. Interpeak latency
prolongations indicate a defect between the
The latencies of central SEPs are a function generators of the two peaks involved. In-
of body height and limb length. Therefore, terpeak latency determinations are most de-
the use of absolute latencies has major lim- sirable because the effects of height, limb
itations. The use of interpeak latencies that length, and temperature are eliminated.
are not related to body size eliminates the With the stimulation of the median or ul-
effect of height. However, when interpeak nar nerve, the absence or delay of N13, with
latencies cannot be measured because all a normal N9, suggests a lesion central to the
peripheral or subcortical evoked potentials brachial plexus and caudal to the foramen
are absent, absolute latencies must be relied magnum. The loss of N13 is also consistent
with a lesion of the low-to-mid cervical cord. Disorders of the Peripheral

Because collaterals from the main pathway Nervous System
generate the N13 dorsal horn potential, it is
not uncommon for a lesion of the dorsal PERIPHERAL NEUROPATHY
horn to eliminate N13, while dorsal column
function and the N14 and N20 potentials are Peripheral neuropathies generally cause
preserved. If N13 is normal but N20 is de- prolongation of the absolute latencies of
layed or absent, a lesion rostral to the mid- brachial plexus, neck, and scalp evoked
cervical cord is indicated and is either a cor- potentials, but central conduction is nor-
tical lesion or a subcortical lesion of the mal. When the neuropathy is marked, only
ascending somatosensory pathways. low-amplitude and poorly formed scalp po-
The absence of a lumbar potential fol- tentials are obtained, and other compo-
lowing tibial nerve stimulation suggests a le- nents are absent (Fig. 16—5). In some cases,
sion at or distal to this level. The presence no response is obtained at any level, par-
of the lumbar N22 potential, with delay of ticularly with stimulation of the posterior
the cervical N30 potential, suggests a lesion tibial nerve. Routine nerve conduction
between these two areas. In the absence of studies are sufficient for the evaluation of
a cervical potential, the presence of a lum- most neuropathies. Somatosensory evoked
bar potential with a delayed or absent scalp potentials are useful for evaluating the
component suggests a nonlocalized lesion proximal segments of nerve if slowing of
rostral to the lumbar spinal cord. conduction through the plexus or roots
Side-to-side interpeak latency differences is suspected. Occasionally, patients with
are also sensitive indicators of abnormality. chronic inflammatory demyelinating neu-
Dispersion of SEPs suggests desynchroniza- ropathy may have slowing primarily at the
tion of the nerve action potential analogous root level, with relatively normal peripheral
to that found in demyelinating disease of pe- nerve conduction.
ripheral nerves; however, this is difficult to In the neuropathy associated with scle-
quantify and should be interpreted cau- rosing myeloma, prolonged interpeak la-
tiously. Morphological peculiarities of wave- tencies of N9-N13 have been found and
forms, unaccompanied by latency prolon- correlated with demyelination in the dorsal
gation, should not be interpreted as an root. Similar changes have been reported
abnormality but rather as an atypical feature in Guillain-Barre syndrome, but only occa-
of uncertain clinical significance. sionally in the absence of changes in F-wave
latencies (Fig. 16-6).40 In patients with
severe sensory neuropathies, a relatively
Amplitude normal scalp SEP can often be obtained
because of central amplification of the pe-
A decrease in the amplitude of SEP wave- ripheral afferent volley, even when the pe-
forms is helpful; however, the range of nor- ripheral sensory nerve action potential is
mal values is broad, making this measure- absent. If the scalp latencies are within nor-
ment less useful than latency measurements. mal limits, it can be inferred that periph-
Also, a general attenuation of cortical SEP eral sensory nerve conduction velocities
amplitude may be encountered with a lesion probably are not markedly slowed. Another
at any level of the somatosensory pathway way to obtain an indirect measure of pe-
from the periphery to the cerebral cortex. It ripheral conduction is to record the shift in
has been suggested that a 50% or greater side- scalp SEP latencies while stimulating the
toîde difference indicates a substantial cen- median nerve at the wrist and elbow. Simi-
tral conduction block or axonal loss or both. larly, SEPs can be used to follow peripheral
nerve regeneration after trauma or surgical
anastomosis at a time when sensory nerve
CLINICAL APPLICATIONS action potentials are too small to be
recorded. Central conduction times in
The usefulness of SEPs in disorders of the hereditary motor and sensory neuropathy
central and peripheral nervous system has type I are usually normal.41 Slowing of
been reviewed recently.32 spinal sensory conduction occurs in some
Figure 16-5. Median somatosensory evoked potentials (SEP) in a patient with a severe axonal peripheral neuropathy
from prolonged ingestion of 2 g of vitamin Be daily. N5 has a very low amplitude and N9 is absent. The N13 am-
plitude is very small, but N20 is normal because of central amplification of the signal. Central conduction (N13-
N20) is normal. The antidromic median sensory nerve action potential (SNAP) recorded from the index finger was
absent.
patients with diabetes mellitus and is more TRAUMATIC BRACHIAL

prominent after stimulation of the tibial PLEXOPATHY
than the median nerve. The abnormality is
likely the result of a distal axonopathy that Following traumatic plexopathy, recordable
affects the terminal segments of the dorsal scalp SEPs and the attenuation or absence
column axons.42 of sensory nerve action potentials indicate
In our laboratory, the most common in- continuity between peripheral and central
dication for dermatomal SEPs is meralgia structures. Conversely, the presence of a nor-
paresthetica (Fig. 16-7). mal Erb's point potential and the absence of
cervical and scalp responses suggest avulsion
of the roots of the plexus (Fig. 16-8). How-
ever, localization often is not possible for se-
vere lesions that involve both preganglionic
and postganglionic elements of the plexus.
The absence of responses without paraspinal
denervation may suggest the need for surgi-
cal exploration of the plexus. Complete SEP
study of an extensive brachial plexus injury
may require stimulation of several nerves,
including the ulnar, median, and radial
Figure 16-6. Guillain-Barre syndrome in a 69-year-old nerves.
woman. Note markedly prolonged scalp response after
sural nerve stimulation. Subcortical evoked potentials With more restricted injuries, it is possible
are absent. Sural nerve action potential at the ankle was to be more selective in planning the study.
normal (amplitude 9.8 /iV, latency 4.1 ms at 14 cm). The nerve chosen should have roots near the
tive recording of SEPs over the scalp during

direct stimulation of roots or other elements
of the brachial plexus at the time of brachial
plexus surgery can provide more accurate in-
formation about the integrity of the root,
trunk, or spinal cord.
THORACIC OUTLET SYNDROME

Ulnar SEPs have limited usefulness in the di-
Figure 16-7. Left meralgia paresthetica. Stimulation of agnosis of neurogenic thoracic outlet syn-
the skin of the right and left anterolateral thigh. The drome.43"45 Several patterns have been de-
scalp response after stimulation on the left is absent. scribed, including a low-amplitude N9, with
a prolonged N9—N13 interpeak latency, and
a low-amplitude N13, with or without atten-
site of injury, as determined clinically and uation of the N9 potential. The ulnar nerve
electromyographically. If only one or two should be stimulated because stimulation of
roots or trunks are involved, stimulation of the median nerve usually gives normal re-
the median nerve can give normal results, sults. However, median nerve stimulation is
because median nerve afferents enter the useful to calculate the ulnar-to-median N9
spinal cord through several roots (C6, 7, 8, amplitude ratio, which is decreased in many
and Tl). Stimulation of the musculocuta- patients with neurogenic thoracic outlet syn-
neous nerve has been advocated for the drome.46 Many studies are flawed because
study of upper trunk lesions, but this nerve the definition of thoracic outlet syndrome is
is difficult to stimulate selectively without ac- vague and the criteria for identifying SEPs
tivating the radial nerve. If avulsion of C7 is as abnormal are not adequate. Somatosen-
suspected, stimulate the radial nerve, and for sory evoked potentials are usually normal in
C8 and Tl lesions, stimulate the ulnar nerve. patients with vascular or symptomatic tho-
A more precise but time-consuming alter- racic outlet syndrome.
native to stimulation of mixed nerve trunks
is stimulation of "segmental" cutaneous
nerves. Stimulation of the lateral cutaneous RADICULOPATHY
nerve of the forearm and digital nerve af- Radiculopathies are evaluated most easily
ferents permits evaluation of nerve fibers and reliably by routine electrophysiologic
confined to nerve roots C5-C8. Intraopera- methods; however, needle electromyo-
graphic abnormalities only indicate dys-
function of the motor root. The SEPs evoked
by stimulation of a mixed nerve are usually
normal in the case of a single root lesion,
because major nerve trunks are formed from
several roots. Segmental sensory stimulation
or stimulation of individual digits or pe-
ripheral nerve branches innervated by a sin-
gle nerve root would appear to be an at-
tractive method for the evaluation of
patients with disk disease. However, because
of the time and effort involved and the low
Figure 16-8. The median somatosensory evoked po- amplitude of the responses, SEP studies are
tential in a 34-year-old man injured in a motorcycle used rarely for the diagnosis of cervical
accident shows a normal N5, a poorly formed N9, and radiculopathies.
absence of N13 and N20 components, consistent with Some investigators have reported a good
root avulsion. The median antidromic sensory ampli-
tude was 15.6 /nV. The thenar compound muscle action correlation between lumbosacral derma-
potential was absent. The ulnar somatosensory evoked tomal SEP abnormalities and findings on
potential showed the same pattern. electromyography, myelography, and at sur-
tials. The overlap is so marked that patho-

logic conditions cannot be predicted reliably
by changes in SEPs. The following discussion
is limited to disorders in which SEPs appear
to aid in diagnosis or management.
CERVICAL SPONDYLOTIC
MYELOPATHY
In cases of cervical spondylosis with myelopa-
thy, scalp evoked potentials after tibial nerve
stimulation are abnormal in approximately
75% of cases, ulnar SEPs are abnormal in
60%, and median SEPs in 25%.53~55 The ab-
Figure 16-9. Left median somatosensory evoked po- normalities consist of loss of amplitude,
tentials in a 63-year-old man with a slowly progressive degradation of waveforms, or slight in-
quadriparesis caused by cervical spondylotic myelopa-
thy. The Nl 3-N20 interpeak latency is mildly prolonged terpeak delays (Figs. 16-9 and 16-10). The
at 7.0 ms (normal, < 6.7 ms). See MRI in Figure 16-10. low-amplitude N9 potentials that are found
occasionally may be caused by disease of the
dorsal root ganglia. Cases of cervical spondy-
gery, but others have found the method less
useful.47"49 A careful study of patients with
a unilateral L5 or SI radiculopathy con-
cluded that dermatomal SEPs have limited
usefulness, because the range of normal val-
ues is broad and the sensitivity and specificity
of the test are poor.50 Also, dermatomal SEP
studies are time-consuming and difficult to
interpret because the responses can be small
and poorly formed.
Although SEPs may have limited useful-
ness in the diagnosis of lumbosacral radicu-
lopathies, dermatomal SEPs may be helpful
in the evaluation of lumbosacral spinal
stenosis. Dermatomal SEPs are abnormal be-
cause of compression of the cauda equina at
several levels in its long course through the
spinal canal. Bilateral dermatomal stimula-
tion of L3, L4, L5, and SI may show absent,
prolonged, or low-amplitude responses.51-52
Disorders of the Central

Nervous System
Abnormalities of SEPs have been described
in many diseases of the CNS. However, SEPs
are most helpful in detecting CNS lesions
when the results of the clinical examination
are normal or equivocal. Diseases of myelin Figure 16-10. MRI of the cervical spine of the same pa-
tient as in Figure 16-9. Note cervical spondylosis with
tend to produce prominent changes in la- myelomalacia and spinal cord atrophy at C3-C4. The
tency, whereas diseases of axons preferen- patient had a history of previous C3-C4 anterior cervi-
tially affect the amplitude of central poten- cal fusion.
lotic myelopathy recorded with a C6Sp-to-an- potential testing is also helpful if symptoms
terior cervical montage frequently show an are suggestive of myelopathy when no def-
abnormal N13 potential, even when sensory inite abnormalities are found on physical
examination findings are normal. It has examination. Median SEPs are abnormal in
been suggested that this is the result of de- about two-thirds of all patients with multi-
creased blood supply caused by compression ple sclerosis and in about half of patients
of the anterior spinal artery.56 The median who have no sensory symptoms or signs.
N13 potential is more likely to be abnormal The rates of abnormality are somewhat
when compression occurs at several spinal higher for lower limb SEPs, because of the
cord levels rather than at either the C4-5 or greater length of white matter traversed.
C5-6 level alone.57 The absence of the cer- The most frequent finding is prolongation
vical peak (N30) after tibial nerve stimula- of SEP interpeak latencies. Low-amplitude
tion is common in cervical myelopathy, but and dispersed scalp responses are also com-
it may also be absent in normal older per- mon. It may be difficult to determine whether
sons, making interpretation a problem. the early scalp responses are markedly pro-
Therefore, the most reliable measurement longed or are absent and only late potentials
is prolongation of the tibial N22-P38 in- are present. The N13 potential recorded
terpeak latency. after median nerve stimulation is often absent
Somatosensory evoked potential testing or attenuated, whereas the scalp response
repeated a year after successful surgical is still present. Lesions of the cervical cord
treatment shows persisting abnormalities, seen on MRI usually result in an abnormal
and the results are of little value in assessing SEP.60
the adequacy of the decompression.58 It is Absent or delayed scalp evoked potentials
unexpected that the severity of MRI abnor- after pudendal nerve stimulation are com-
malities, clinical examination findings, and mon in patients with bladder dysfunction
SEP abnormalities show no clear correlation. and in those with clinically probable multi-
However, severe MRI abnormalities usually ple sclerosis.61 Although the overall sensitiv-
are associated with abnormal SEPs. Con- ity of SEPs is lower than that of MRI in
versely, some patients with normal imaging multiple sclerosis, SEPs may be better for de-
studies may have abnormal SEPs.59 tecting spinal cord lesions that are below the
resolution of MRI.62 However, SEPs are not
a reliable method for monitoring disease
DEMYELINATING DISEASE progression or the effect of treatment. Mea-
Most frequently SEPs are used to evaluate surement of MRI lesions, including the num-
ber present, the occurrence of new lesions,
suspected multiple sclerosis (Fig. 16-11),
especially in documenting a second clini- and the volume of the lesions, and the pres-
ence of cerebral atrophy may be superior to
cally silent lesion, for example, in a patient
with optic neuritis. Somatosensory evoked SEP testing.
SYRINGOMYELIA
In syringomyelia, ulnar and median SEPs
usually are normal if there is dissociated sen-
sory loss and usually abnormal if all sensory
modalities are impaired. The absence of N13
suggests involvement of the central gray mat-
ter of the cervical cord.63 Because of con-
duction directly up the dorsal columns, PI4
and N20 may still be recorded when spinal
Figure 16-11. Right median Somatosensory evoked po- N13 is absent. Tibial SEPs most frequently
tentials in a patient with multiple sclerosis. N13 is ab- are abnormal, showing an absence of the
sent. The N9-N20 interpeak latency is quite prolonged
at 14.1 ms (normal, < 6.8 ms). The scalp response has neck response (N30), delayed low-amplitude
a low amplitude and is dispersed. These findings are or absent scalp potentials, or prolongation
consistent with demyelination. of the N22-P38 latency.44
SPINAL CORD TUMORS

Various types of spinal cord tumor com-
monly have associated SEP abnormalities.
An absent or reduced N13 response indi-
cates involvement of dorsal horn gray mat-
ter and is correlated with disturbed pain and
temperature sensation and reduced reflexes
in the upper limbs. P14 abnormalities are
correlated with impaired joint and touch
Figure 16-12. Left median somatosensory evoked po-
sensation. Involvement of the lumbar cord tentials in a 17-year-old man with Friedreich's ataxia.
is associated with an abnormal N22 poten- The N13-N20 interpeak latency is prolonged at 11.3 ms
tial. Prolongation of interpeak latencies is (normal, < 6.7 ms). The N9-N13 interpeak latency is
also common.64 Somatosensory evoked po- normal.
tentials may be normal in slow-growing as-
trocytomas that infiltrate but do not destroy
sensory pathways. brain stem or the thalamocortical radiation.
Somatosensory evoked potentials are normal
in the Wallenberg and Weber syndromes,
MISCELLANEOUS SPINAL because the medial lemniscus is unaffected.
CORD LESIONS Central conduction times in hereditary mo-
Markedly prolonged conduction times are tor and sensory neuropathy type I are usu-
seen in patients with adrenomyeloneuropa- ally normal.41 In Friedreich's ataxia, SEPs
thy and are caused by involvement of distal typically show absent or low-amplitude N9-
and proximal peripheral nerves and the N13 potentials of normal or near normal la-
CNS.65 Slowing is found also in the majority tency, whereas the latency of cortical re-
of adrenoleukodystrophy carriers.66 In vita- sponses is increased, often with broadening
min Big deficiency, changes in central con- of N2070 (Figs. 16-12 and 16-13). Abnormal
duction usually appear before peripheral median and tibial SEPs are also seen in pa-
slowing, suggesting that the central process tients with other autosomal dominant cere-
of a dorsal root ganglion cell is affected ear- bellar ataxias.71*72 In Huntington's disease,
lier than the distal process.67 Early treatment the amplitude of the cortical SEP progres-
lessens the abnormalities of SEPs. The util- sively decreases, perhaps reflecting degen-
ity of SEPs for predicting neurologic recov- eration of thalamic neurons. However, this
ery in traumatic spinal cord injury has been change is not helpful in diagnosis of the dis-
reviewed recently.68 They do not seem to be ease in an individual patient. In thalamo-
more effective than clinical examination for cortical lesions with no sensory loss, SEPs are
predicting outcome, but in an unresponsive normal, but if the sensory loss is complete,
or uncooperative patient, they are useful in
determining whether a spinal cord injury is
present. In acute transverse myelitis, tibia!
SEPs are abnormal in about 75% of patients
and median SEPs are usually normal. Assess-
ment of muscle strength and electromyo-
graphic findings of denervation are most
useful for predicting outcome, although in
the first 2 weeks, SEP testing may be used in-
stead of electromyography.69
BRAIN STEM AND SUPRATENTORIAL

LESIONS Figure 16-13. Tibial scalp potentials, in same patient as
in Figure 16-12, are low amplitude, and die P38 latency
Central conduction times may be prolonged is approximately 52.2 ms. The N22-P38 interpeak la-
with lesions of the lemniscal pathway in the tency is quite prolonged at 28 ms (normal, < 20 ms).
Figure 16-14. Upper and lower extremity somatosensory evoked potentials in a patient with stimulus-sensitive
myoclonus, illustrating the characteristic pattern of very high-amplitude cortical potentials recorded from the scalp
electrodes.
the patient will have no N20 or subsequent motor and sensory nerves75'76 (Fig. 16-15).
potentials. Patients with minimal cortical With tibial nerve stimulation, changes occur
sensory loss may have asymmetrical cortical in the amplitude and field distribution of the
potentials. Evoked potentials in patients with early cortical components, but not in pa-
hysterical sensory loss should be normal. Me- tients with progressive muscular atrophy.77
dian SEPs are generally normal in Alzhei- Severe abnormalities in patients with sus-
mer's disease but are frequently abnormal in pected motor neuron disease should raise
those with multi-infarct dementia. suspicion of other conditions that may
mimic amyotrophic lateral sclerosis, for ex-
ample, cervical spondylosis. An exception to
MYOCLONUS this generalization is X-linked spinobulbar
Myoclonus occasionally is associated with muscular atrophy (Kennedy's syndrome), in
very high-amplitude cortical evoked re- which the scalp evoked potentials after me-
sponses (Fig. 16-14). The giant SEPs are sep- dian and tibial nerve stimulation are poorly
arable into three types that reflect hyperex- formed or absent, whereas the peripheral
citability in the afferent or efferent system and spinal potentials are relatively pre-
(or both) of the somatosensory cortex.73 served.78
The amplitude may be increased up to 10
times normal in patients with cortical reflex
myoclonus, including progressive myo-
clonus epilepsy and epilepsia partialis con-
tinua.74 The cortical SEPs are not exagger-
ated in patients with myoclonus of brain
stem or spinal cord origin.
MOTOR NEURON DISEASE

Patients with amyotrophic lateral sclerosis
may have minor abnormalities of central
conduction, delay, or absence of cortical re-
Figure 16-15. Left median somatosensory evoked po-
sponses and increased mean values of all potentials in a 57-year-old man with amyotrophic lateral
tential latencies with median nerve stimula- sclerosis. There is mild slowing of central conduction
tion and stimulation of lower extremity (N13-N20 interpeak latency).
PROGNOSIS IN COMA plitude recover gradually, with differences

persisting between patients who make a
Head Injury good recovery and those who remain dis-
The major factors that determine outcome abled. Somatosensory evoked potentials are
after head trauma include the (I) location, superior to EEG in determining prognosis
extent, and type of brain injury; (2) in- of severe head injury, by allowing prediction
tracranial pressure; and (3) functional re- of favorable and unfavorable outcomes in a
versibility of the injury. Clinical examination much larger number of patients.82'83
findings in patients admitted to the hospital A battery of auditory evoked potentials,
in coma following head trauma usually can visual evoked potentials, and SEPs is also an
predict which ones are likely to have a very accurate and reliable prognostic indicator
poor recovery and those likely to have a very after severe head injury, with an overall ac-
good recovery, but the findings are less reli- curacy of approximately 91%.84~86 Patients
able for the rest. The patient's initial degree are assigned to multimodality evoked po-
of neurologic impairment, as assessed with tential (MMEP) groups according to the
the Glasgow Coma Scale, and the likelihood most abnormal study obtained in any mo-
of a poor outcome are generally correlated. dality. The scale ranges from grade 1 (nor-
The strongest predictive signs are those com- mal) to grade 4 (absence of activity). As the
patible with brain death, although even severity of MMEP abnormality increases, so
these may be erroneous in the presence of does mortality. Of patients with mildly ab-
drug intoxication, hypothermia, hypoten- normal MMEP scores, 81% have a return to
sion, or hypoxemia. An assessment of the normal life or only moderate disability, and
neurologic status of a comatose patient 76% of those with severely abnormal MMEP
based on clinical judgment is difficult and scores have a poor outcome. A good out-
inaccurate, particularly if multiple neuroin- come is realized by 76% of patients with a
tensive measures have been used, including grade 1 MMEP, 61% with grade 2, 35% with
hyperventilation, muscle relaxation, high- grade 3, and 0% with grade 4. Overall, 87%
dose barbiturate therapy, and hypothermia. of patients with a grade 1 MMEP have a
In an individual case, it is difficult to know good-to-moderate outcome at 1 year.
with certainty whether the brain dysfunction
of a comatose patient is irreversible and Anoxia
whether aggressive treatment will restore
function.79 Somatosensory evoked potentials Outcome has been examined in a series of
are helpful in defining the functional re- patients with anoxic coma caused by car-
versibility of cerebral injuries of any cause.80 diopulmonary arrest or severe hypotension
In 78% of patients, SEPs can predict out- whose prognoses were uncertain on the ba-
come within 3.5 days.81 Surviving patients with sis of clinical findings on day I.87 Patients
persistent asymmetries of cerebral evoked with an obviously good or bad prognosis clin-
potentials remain hemiplegic, whereas those ically were excluded. All 18 patients with ab-
with absence of evoked potentials over both sent -or low-amplitude responses had no re-
cerebral hemispheres die. In most patients covery. It was found that some patients with
who have a good recovery, conduction times initially malignant EEGs and normal SEPs
return to normal before day 10, but in those may recover and should be supported until
who remain disabled or who die, conduction the prognosis is more definitive.
times are persistently abnormal or scalp po-
tentials are absent. When cortical evoked po- Infants and Children
tentials are absent, it generally is futile to use
heroic measures such as barbiturate coma or Somatosensory evoked potential results sim-
decompressive craniectomy to control re- ilar to those in adults have been found in
fractory increased intracranial pressure. children comatose from hypoxic-ischemic
However, SEPs are not perfect indicators encephalopathy, head injury, or other con-
of recovery. A few patients with poor recov- ditions.88 In a study of 127 children who
ery can have normal SEPs. Serial studies have were comatose because of severe head in-
shown that both conduction time and am- jury, all 32 who had an absence of brain stem
auditory evoked responses and SEPs died. Of comatose but not brain dead. In contrast, au-
children with normal evoked potential stud- ditory brain stem evoked responses reflect
ies, 78% had a good prognosis.89 Somato- pontine and midbrain function rather than
sensory evoked potentials recorded in the medullary function and can fail to detect re-
first week after admission correlate highly maining brain stem function.95 For similar
with outcome assessed 1 and 5 years after se- reasons, the use of a frontal-to-nasopharyn-
vere brain injury.90 Somatosensory evoked geal (Fz-PgZ) montage to record PI4 is also
potentials are also accurate prognostic tools advocated as a clear sign of involvement of
for newborns with asphyxia. The absence of the craniocervical junction in brain-dead
scalp potentials is a very poor prognostic patients.96'97
sign, and delayed latencies are associated
with deficits in most patients. Somatosensory SOMATOSENSORY EVOKED
evoked potentials also have long-term pre- POTENTIALS RECORDED IN
dictive value for deficits that become appar- THE INTENSIVE CARE UNIT
ent at school age.91
Recording SEPs in the intensive care unit
Cerebral Hemorrhage and Stroke presents problems, usually in the form of
high-amplitude 60 Hz artifact, not encoun-
The combination of SEPs and brain stem au- tered in the outpatient laboratory. Sugges-
ditory evoked response is useful in assessing tions for reducing the artifact include shut-
the prognosis of patients with subarachnoid ting off all nonessential electric equipment
or hypertensive hemorrhage and cerebral such as lights, cardiac monitor, cooling blan-
infarction.92'93 ket, feeding pumps, and blood warmers to
ascertain if one of these is causing the arti-
SOMATOSENSORY EVOKED fact. The impedance of the electrodes
POTENTIAL FINDINGS IN should be checked and, if necessary, the
BRAIN DEATH electrodes reapplied or new electrodes sub-
stituted. If artifact is present in all channels,
All patients with brain death have bilateral replace the ground.
loss of median SEP N20 components, but In an unconscious patient or in one un-
cervical N13 potentials can still be elicited. der anesthesia, subcutaneous needle elec-
The presence of SEP N13 is helpful because trodes may be applied; this not only saves
it establishes that the input signal has time but also reduces artifact. If muscle ar-
reached the CNS. However, this finding does tifact is a problem in a patient who had a
not prove brain death, because rare patients head injury and is decerebrate or decorti-
who have severe bilateral supratentorial le- cate and on a respirator, a single dose of a
sions, drug intoxication, or severe cerebral neuromuscular blocking agent can be given
edema but no clear clinical signs of brain safely. If a paralyzing agent cannot be given,
death may also have a loss of the N20 po- increasing the gain to 20 fJiV or 50 /u,V per
tential.94 Most patients with brain death do division on channels recording from the el-
not have brain stem auditory evoked re- bow, Erb's point, or neck may eliminate
sponses, including wave I. Occasionally, the blocking of the amplifier and still result in
interpretation of brain stem auditory evoked recognizable peaks, which can be amplified
responses is uncertain, because their ab- after the recording has been completed.
sence could also be caused by preexisting Patients with severe head injury often re-
deafness or damage to peripheral auditory quire neurosurgical procedures for evacua-
pathways from a temporal bone fracture. tion of intracerebral hematomas and fre-
Recent research has suggested that N18 is quently have intracranial pressure monitors
a useful indicator of brain stem function. placed. The location of the incisions and
This potential has the advantage of being drains may require placement of recording
generated by the cuneate nucleus in caudal electrodes slightly anterior or posterior to
medulla, close to the respiratory center. N18 the usual locations. Generally, this does not
is almost always lost in brain death and pre- result in a marked change in the morphol-
served in recordings from patients who are ogy of the cortical evoked potentials, and
they are still diagnostically useful. The rate 8. Emerson RG, Seyal M, Pedley TA. Somatosensory
of stimulation may also have an effect in evoked potentials following median nerve stimula-
tion. I. The cervical components. Brain 107:169-
some patients with head injury. Recording 182, 1984.
at 5 per second occasionally causes a reflex- 9. Jeanmonod D, Sindou M, Mauguiere F. Three
ive increase in decerebrate posturing, which transverse dipolar generators in the human cervi-
can be eliminated by reducing the rate of cal and lumbo-sacral dorsal horn: evidence from di-
rect intraoperative recordings on the spinal cord
stimulation to 1 or 2 per second and by us- surface. Electroencephalogr Clin Neurophysiol
ing the lowest voltage possible to elicit a mus- 74:236-240, 1989.
cle twitch. 10. Mauguiere F. Anatomic origin of the cervical N13
potential evoked by upper extremity stimulation.
J Clin Neurophysiol 17:236-245, 2000.
SUMMARY 11. Lesser RP, Lueders H, Hahn J, Klem G. Early so-
matosensory potentials evoked by median nerve
stimulation: intraoperative monitoring. Neurology
Somatosensory evoked potentials recorded 31:1519-1523, 1981.
with surface electrodes represent volume- 12. Sonoo M, Shimpo T, Genba K, Kunimoto M, Man-
conducted activity arising from myelinated nen T. Posterior cervical N13 in median nerve SEP
has two components. Electroencephalogr Clin Neu-
peripheral and central axons, synapses in rophysiol 77:28-38, 1990.
central gray matter, and changes in the size 13. Zanette G, Tinazzi M, Manganotti P, Bonato C, Polo
and shape of the volume conductor. They A. Two distinct cervical N13 potentials are evoked
provide an objective measure of function in by ulnar nerve stimulation. Electroencephalogr
large-diameter myelinated sensory afferents Clin Neurophysiol 96:114-120, 1995.
14. Tomberg C, Desmedt JE, Ozaki I, Noel P. Na-
peripherally and in proprioceptive pathways sopharyngeal recordings of somatosensory evoked
centrally. Changes in amplitude and latency potentials document the medullary origin of the
can be used to localize lesions in the nervous N18 far-field. Electroencephalogr Clin Neurophys-
system, to identify objectively abnormalities iol 80:496-503, 1991.
15. Restuccia D. Anatomic origin of P13 and P14 scalp
in patients with few sensory manifestations far-field potentials. J Clin Neurophysiol 17:246-257,
or none at all, and to monitor function over 2000.
time. 16. Urasaki E, Uematsu S, Lesser RP. Short latency so-
matosensory evoked potentials recorded around
the human upper brain-stem. Electroencephalogr
Clin Neurophysiol 88:92-104, 1993.
REFERENCES 17. Noel P, Ozaki I, DesmedtJE. Origin of N18 and
PI 4 far-fields of median nerve somatosensory
1. Dumitru D, Jewett DL. Far-field potentials. Muscle evoked potentials studied in patients with a brain-
Nerve 16:237-254, 1993. stem lesion. Electroencephalogr Clin Neurophysiol
2. KimuraJ, Mitsudome A, Yamada T, Dickins QS. Sta- 98:167-170, 1996.
tionary peaks from a moving source in far-field 18. Sonoo M, Kobayashi M, Genba-Shimizu K, Mannen
recording. Electroencephalogr Clin Neurophysiol T, Shimizu T. Detailed analysis of the latencies of
58:351-361, 1984. median nerve somatosensory evoked potential
3. KimuraJ, Ishida T, Suzuki S, Kudo Y, Matsuoka H, components, 1: selection of the best standard pa-
Yamada T. Far-field recording of the junctional po- rameters and the establishment of normal values.
tential generated by median nerve volleys at the Electroencephalogr Clin Neurophysiol 100:319-
wrist. Neurology 36:1451-1457, 1986. 331, 1996.
4. Lee EK, Seyal M. Generators of short latency hu- 19. Sonoo M. Anatomic origin and clinical application
man somatosensory-evoked potentials recorded of the widespread N18 potential in median nerve
over the spine and scalp. J Clin Neurophysiol somatosensory evoked potentials. J Clin Neuro-
15:227-234, 1998. physiol 17:258-268, 2000.
5. Kakigi R, Shibasaki H, Tanaka K, et al. CO2 laser- 20. Emori T, Yamada T, Seki Y, et al. Recovery func-
induced pain-related somatosensory evoked poten- tions of fast frequency potentials in the initial neg-
tials in peripheral neuropathies: correlation be- ative wave of median SEP. Electroencephalogr Clin
tween electrophysiological and histopathological Neurophysiol 78:116-123, 1991.
findings. Muscle Nerve 14:441-450, 1991. 21. Vanderzant CW, Beydoun AA, Domer PA, Hood
6. Kakigi R, Watanabe S, Yamasaki H. Pain-related so- TW, Abou-Khalil BW. Polarity reversal of N20 and
matosensory evoked potentials. J Clin Neurophys- P23 somatosensory evoked potentials between scalp
iol 17:295-308, 2000. and depth recordings. Electroencephalogr Clin
7. DesmedtJE, Cheron G. Prevertebral (oesophageal) Neurophysiol 78:234-239, 1991.
recording of subcortical somatosensory evoked po- 22. Yamada T, Muroga T, Kimura J. Tourniquet-in-
tentials in man: the spinal PI3 component and the duced ischemia and somatosensory evoked poten-
dual nature of the spinal generators. Electroen- tials. Neurology 31:1524-1529, 1981.
cephalogr Clin Neurophysiol 52:257-275, 1981. 23. Valeriani M, Restuccia D, di Lazzaro V, et al. Giant
central N20-P22 with normal area 3b N20-P20: an somatosensory conduction time from 10 to 79 years.
argument in favour of an area 3a generator of early Electroencephalogr Clin Neurophysiol 54:49-54,
median nerve cortical SEPs? Electroencephalogr 1982.
Clin Neurophysiol 104:60-67, 1997. 39. Strenge H, Hedderich J. Age-dependent changes in
24. Allison T, McCarthy G, Wood CC, Jones SJ. Poten- central somatosensory conduction time. Eur Neu-
tials evoked in human and monkey cerebral cortex rol 21:270-276, 1982.
by stimulation of the median nerve. A review of 40. Yiannikas C. Short-latency somatosensory evoked
scalp and intracranial recordings. Brain 114:2465- potentials in peripheral nerve lesions, plexopathies,
2503, 1991. radiculopathies, and spinal cord trauma. In Chi-
25. Valeriani M, Le Pera D, Tonali P. Characterizing appa KH (ed). Evoked Potentials in Clinical Medi-
somatosensory evoked potential sources with dipole cine, 2nd ed. Raven Press, New York, 1990, pp 439-
models: advantages and limitations. Muscle Nerve 469.
24:325-339, 2001. 41. Aramideh M, Hoogendijk JE, Aalfs CM, et al. So-
26. Buchner H, Waberski TD, Fuchs M, Wischmann matosensory evoked potentials, sensory nerve po-
HA, Wagner M, Drenckhahn R. Comparison of re- tentials and sensory nerve conduction in hereditary
alistically shaped boundary-element and spherical motor and sensory neuropathy type I. J Neurol
head models in source localization of early so- 239:277-283, 1992.
matosensory evoked potentials. Brain Topogr 42. Comi G. Evoked potentials in diabetes mellitus.
8:137-143, 1995. Clin Neurosci 4:374-379, 1997.
27. Seyal M, Gabor AJ. The human posterior tibial so- 43. Yiannikas C, Walsh JC. Somatosensory evoked re-
matosensory evoked potential: synapse dependent sponses in the diagnosis of thoracic oudet syn-
and synapse independent spinal components. Elec- drome. J Neurol Neurosurg Psychiatry 46:234-240,
troencephalogr Clin Neurophysiol 62:323-331, 1983.
1985. 44. Veilleux M, Stevens JC. Syringomyelia: electrophys-
28. Halonen JP, Jones SJ, Edgar MA, Ransford AO. iologic aspects. Muscle Nerve 10:449-458, 1987.
Conduction properties of epidurally recorded 45. Komanetsky RM, Novak CB, Mackinnon SE, Russo
spinal cord potentials following lower limb stimu- MH, Padberg AM, Louis S. Somatosensory evoked
lation in man. Electroencephalogr Clin Neuro- potentials fail to diagnose thoracic outlet syn-
physiol 74:161-174, 1989. drome. J Hand Surg [Am] 21:662-666, 1996.
29. Seyal M, Kraft LW, Gabor AJ. Cervical synapse-de- 46. Cakmur R, Idiman F, Akalin E, Gene A, Yener GG,
pendent somatosensory evoked potential following Ozturk V. Dermatomal and mixed nerve so-
posterior tibial nerve stimulation. Neurology 37: matosensory evoked potentials in the diagnosis of
1417-1421, 1987. neurogenic thoracic outlet syndrome. Electroen-
30. Yamada T. Neuroanatomic substrates of lower ex- cephalogr Clin Neurophysiol 108:423-434, 1998.
tremity somatosensory evoked potentials. J Clin 47. Aminoff MJ, Goodin DS, Parry GJ, Barbara NM,
Neurophysiol 17:269-279, 2000. Weinstein PR; Rosenblum ML. Electrophysiologic
31. Baumgartner U, Vogel H, Ellrich J, Gawehn J, evaluation of lumbosacral radiculopathies: electro-
Stoeter P, Treede RD. Brain electrical source anal- myography, late responses, and somatosensory
ysis of primary cortical components of the tibial evoked potentials. Neurology 35:1514-1518, 1985.
nerve somatosensory evoked potential using re- 48. Rodriquez AA, Karris L, Rodriquez AA, Lane D. So-
gional sources. Electroencephalogr Clin Neuro- matosensory evoked potentials from dermatomal
physiol 108:588-599, 1998. stimulation as an indicator of L5 and SI radicu-
32. Aminoff MJ, Eisen AA. AAEM minimonograph 19: lopathy. Arch Phys Med Rehabil 68:366-368, 1987.
Somatosensory evoked potentials. Muscle Nerve 49. Walk D, Fisher MA, Doundoulakis SH, Hemmati M.
21:277-290, 1998. Somatosensory evoked potentials in the evaluation
33. Klausner AP, Batra AK. Pudendal nerve so- of lumbosacral radiculopathy. Neurology 42:1197-
matosensory evoked potentials in patients with void- 1202, 1992.
ing and/or erectile dysfunction: correlating test re- 50. Dumitru D, Dreyfuss P. Dermatomal/segmental
sults with clinical findings. J Urol 156:1425-1427, somatosensory evoked potential evaluation of
1996. L5/S1 um'lateral/unilevel radiculopathies. Muscle
34. Slimp JC, Rubner DE, Snowden ML, Stolov WC. Nerve 19:442-449, 1996.
Dermatomal somatosensory evoked potentials: cer- 51. Kraft GH. A physiological approach to the evalua-
vical, thoracic, and lumbosacral levels. Electroen- tion of lumbosacral spinal stenosis. Phys Med Re-
cephalogr Clin Neurophysiol 84:55-70, 1992. habil Clin N Am 9:381-389, 1998.
35. Leandri M, Schizzi R, Favale E. Blink reflex far fields 52. Snowden ML, Haselkorn JK, Kraft GH, et al. Der-
mimicking putative cortical trigeminal evoked po- matomal somatosensory evoked potentials in the di-
tentials. Electroencephalogr Clin Neurophysiol 93: agnosis of lumbosacral spinal stenosis: comparison
240-242, 1994. widi imaging studies. Muscle Nerve 15:1036-1044,
36. Guidelines for somatosensory evoked potentials. 1992.
Muscle Nerve 22 (Suppl 8):S123-S138, 1999. 53. Veilleux M, Daube JR. The value of ulnar so-
37. Sonoo M, Hagiwara H, Motoyoshi Y, Shimizu T. Pre- matosensory evoked potentials (SEPs) in cervical
served widespread N18 and progressive loss of myelopathy. Electroencephalogr Clin Neurophysiol
P13/14 of median nerve SEPs in a patient with 68:415-423, 1987.
unilateral medial medullary syndrome. Electroen- 54. Yrannikas C, Shahani BT, Young RR. Short-latency
cephalogr Clin Neurophysiol 100:488-492, 1996. somatosensory-evoked potentials from radial, me-
38. Hume AL, Cant BR, Shaw NA, Cowan JC. Central dian, ulnar, and peroneal nerve stimulation in the
assessment of cervical spondylosis. Comparison logic recovery in traumatic cervical spinal cord in-
with conventional electromyography. Arch Neurol jury. Arch Phys Med Rehabil 79:1456-1466, 1998.
43:1264-1271, 1986. 69. KaliltaJ, Misra UK, Mandal SK. Prognostic predic-
55. Yu YL, Jones SJ. Somatosensory evoked potentials tors of acute transverse myelitis. Acta Neurol Scand
in cervical spondylosis. Correlation of median, ul- 98:60-63, 1998.
nar and posterior tibial nerve responses with clini- 70. Jones SJ, Baraitser M, Halliday AM. Peripheral and
cal and radiological findings. Brain 108:273-300, central somatosensory nerve conduction defects in
1985. Friedreich's ataxia. J Neurol Neurosurg Psychiatry
56. Restuccia D, Di Lazzaro V, Valeriani M, Tonali P, 43:495-503, 1980.
Mauguiere F. Segmental dysfunction of the cervi- 71. Abele M, Burk K, Andres F, et al. Autosomal dom-
cal cord revealed by abnormalities of the spinal N13 inant cerebellar ataxia type I. Nerve conduction
potential in cervical spondylotic myelopathy. Neu- and evoked potential studies in families with SCA1,
rology 42:1054-1063, 1992. SCA2 and SCA3. Brain 120:2141-2148, 1997.
57. Kaneko K, Kawai S, Taguchi T, Fuchigami Y, Ito T, 72. Perretti A, Santoro L, Lanzillo B, et al. Autosomal
Morita H. Correlation between spinal cord com- dominant cerebellar ataxia type I: multimodal elec-
pression and abnormal patterns of median nerve trophysiological study and comparison between
somatosensory evoked potentials in compressive SCA1 and SCA2 patients. J Neurol Sci 142:45-53,
cervical myelopathy: comparison of surface and 1996.
epidurally recorded responses. J Neurol Sci 73. Shibasaki H, Yamashita Y, Neshige R, Tobimatsu S,
158:193-202, 1998. Fukui R. Pathogenesis of giant somatosensory
58. de Noordhout AM, Myressiotis S, Delvaux V, Born evoked potentials in progressive myoclonic epilepsy.
JD, Delwaide PJ. Motor and somatosensory evoked Brain 108:225-240, 1985.
potentials in cervical spondylotic myelopathy. Elec- 74. Kakigi R, Shibasaki H. Generator mechanisms of
troencephalogr Clin Neurophysiol 108:24-31, giant somatosensory evoked potentials in cortical
1998. reflex myoclonus. Brain 110:1359-1373, 1987.
59. Berthier E, Turjman F, Mauguiere F. Diagnostic 75. Georgesco M, Salerno A, Camu W. Somatosensory
utility of somatosensory evoked potentials (SEPs) in evoked potentials elicited by stimulation of lower-
presurgical assessment of cervical spondylotic limb nerves in amyotrophic lateral sclerosis. Electro-
myelopathy. Neurophysiol Clin 26:300-310, 1996. encephalogr Clin Neurophysiol 104:333-342,1997.
60. Turano G, Jones SJ, Miller DH, Du Boulay GH, 76. Zakrzewska-Pniewska B, Gasik R, Kostera-Pruszczyk
Eakigi R, McDonald WI. Correlation of SEP ab- A, Emeryk-Szajewska B. Reconsiderations about the
normalities with brain and cervical cord MRI in abnormalities of somatosensory evoked potentials
multiple sclerosis. Brain 114:663-681, 1991. in motor neuron disease. Electromyogr Clin Neu-
61. Sau G, Siracusano S, Aiello I, et al. The usefulness rophysiol 39:107-112, 1999.
of the somatosensory evoked potentials of the pu- 77. Zanette G, Tinazzi M, Polo A, Rizzuto N. Motor neu-
dendal nerve in diagnosis of probable multiple scle- ron disease with pyramidal tract dysfunction in-
rosis. Spinal Cord 37:258-263, 1999. volves the cortical generators of the early so-
62. Guerit JM, Monje Argiles A. The sensitivity of mul- matosensory evoked potential to tibial nerve
timodal evoked potentials in multiple sclerosis. A stimulation. Neurology 47:932-938, 1996.
comparison with magnetic resonance imaging and 78. Polo A, Teatini F, D'Anna S, et al. Sensory involve-
cerebrospinal fluid analysis. Electroencephalogr ment in X-linked spino-bulbar muscular atrophy
Clin Neurophysiol 70:230-238, 1988. (Kennedy's syndrome): an electrophysiological
63. Urasaki E, Wada S, Kadoya C, Matsuzaki H, Yokata study. J Neurol 243:388-392, 1996.
A, Matsuoka S. Absence of spinal N13-P13 and nor- 79. de Weerd AW, Groeneveld C. The use of evoked
mal scalp far-field PI 4 in a patient with sy- potentials in the management of patients with se-
ringomyelia. Electroencephalogr Clin Neurophys- vere cerebral trauma. Acta Neurol Scand 72:489-
iol 71:400-404, 1988. 494, 1985.
64. Restuccia D, Di Lazzaro V, Valeriani M, Colosimo 80. Carter BG, Butt W. Review of the use of so-
C, Tonali P. Spinal responses to median and tibial matosensory evoked potentials in the prediction of
nerve stimulation and magnetic resonance imaging outcome after severe brain injury. Grit Care Med
in intramedullary cord lesions. Neurology 29:178-186, 2001.
46:1706-1714, 1996. 81. Hume AL, Cant BR. Central somatosensory con-
65. Kaplan PW, Tusa RJ, Rignani J, Moser HW. So- duction after head injury. Ann Neurol 10:411-419,
matosensory evoked potentials in adrenomyelo- 1981.
neuropathy. Neurology 48:1662-1667, 1997. 82. Hutchinson DO, Frith RW, Shaw NA, Judson JA,
66. Restuccia D, Di Lazzaro V, Valeriani M, et al. Neu- Cant BR. A comparison between electroen-
rophysiological abnormalities in adrenoleukodys- cephalography and somatosensory evoked poten-
trophy carriers. Evidence of different degrees of tials for outcome prediction following severe head
central nervous system involvement. Brain 120: injury. Electroencephalogr Clin Neurophysiol
1139-1148, 1997. 78:228-233, 1991.
67. Jones SJ, Yu YL, Rudge P, et al. Central and pe- 83. Moulton RJ, Brown JI, Konasiewicz SJ. Monitoring
ripheral SEP defects in neurologically symptomatic severe head injury: a comparison of EEG and so-
and asymptomatic subjects with low vitamin B12 lev- matosensory evoked potentials. Can J Neurol Sci
els. J Neurol Sci 82:55-65, 1987. 25:S7-S11, 1998.
68. Kirshblum SC, O'Connor KG. Predicting neuro- 84. Greenberg RP, Newlon PG, Hyatt MS, Narayan RK,
Becker DP. Prognostic implications of early multi- predictors of outcome in neonatal intensive care
modality evoked potentials in severely head-injured unit survivors: review of the literature. Pediatr Neu-
patients. A prospective study. J Neurosurg 55:227- rol 14:189-195, 1996.
236, 1981. 92. Facco E, Behr AU, Munari M, et al. Auditory and
85. Narayan RK, Greenberg RP, Miller JD, et al. Im- somatosensory evoked potentials in coma following
proved confidence of outcome prediction in severe spontaneous cerebral hemorrhage: early prognosis
head injury. A comparative analysis of the clinical and outcome. Electroencephalogr Clin Neuro-
examination, multimodality evoked potentials, CT physiol 107:332-338, 1998.
scanning, and intracranial pressure. J Neurosurg 93. Haupt WF, Pawllik G. Contribution of initial me-
54:751-762, 1981. dian-nerve somatosensory evoked potentials and
86. Newlon PG, Greenberg RP, Hyatt MS, Enas GG, brain stem auditory evoked potentials to prediction
Becker DP. The dynamics of neuronal dysfunction of clinical outcome in cerebrovascular critical care
and recovery following severe head injury assessed patients: a statistical evaluation. J Clin Neurophys-
with serial multimodality evoked potentials. J Neu- iol 15:154-158, 1998.
rosurg 57:168-177, 1982. 94. Schwarz S, Schwab S, Aschoff A, Hacke W. Favor-
87. Chen R, Bolton CF, Young B. Prediction of out- able recovery from bilateral loss of somatosensory
come in patients with anoxic coma: a clinical and evoked potentials. Grit Care Med 27:182-187,1999.
electrophysiologic study. Grit Care Med 24:672- 95. Sonoo M, Tsai-Shozawa Y, Aoki M, et al. N18 in me-
678, 1996. dian somatosensory evoked potentials: a new indi-
88. Ruiz-Lopez MJ, Martinez de Azagra A, Serrano A, cator of medullary function useful for the diagno-
Casado-Flores J. Brain death and evoked potentials sis of brain death. J Neurol Neurosurg Psychiatry
in pediatric patients. Grit Care Med 27:412-416, 67:374-378, 1999.
1999. 96. Roncucci P, Lepori P, Mok MS, Bayat A, Logi F,
89. Butinar D, Gostisa A. Brain stem auditory evoked Marino A. Nasopharyngeal electrode recording of
potentials and somatosensory evoked potentials in somatosensory evoked potentials as an indicator in
prediction of posttraumatic coma iri children. brain death. Anaesth Intensive Care 27:20-25,
Pflugers Arch 431:R289-R290, 1996. 1999.
90. Carter BG, Taylor A, Butt W. Severe brain injury in 97. Wagner W. Scalp, earlobe and nasopharyngeal
children: long-term outcome and its prediction us- recordings of the median nerve somatosensory
ing somatosensory evoked potentials (SEPs). In- evoked P14 potential in coma and brain death. De-
tensive Care Med 25:722-728, 1999. tailed latency and amplitude analysis in 181 pa-
91. Majnemer A, Rosenblatt B. Evoked potentials as tients. Brain 119:1507-1521, 1996.
Chapter 17
BRAIN STEM AUDITORY EVOKED
POTENTIALS IN CENTRAL
DISORDERS
John N. Caviness
AUDITORY ANATOMY AND PHYSIOLOGY Intrinsic Brain Stem Lesions

AUDITORY EVOKED POTENTIALS IN Coma and Brain Death
NORMAL SUBJECTS SUMMARY
METHODS
USE OF BRAIN STEM AUDITORY
EVOKED POTENTIALS IN CLINICAL
PROBLEM SOLVING
Acoustic Neuroma
Demyelinating Disease
Brain stem auditory evoked potentials (BAEPs) tex positive waveforms, numbered I through
are electrophysiologic studies that usually V, are invariably present in normal subjects
have abnormal results in patients with le- after peripheral auditory stimulation. The
sions involving the auditory portion of cra- common BAEP alterations in patients with
nial nerve VIII (CN VIII), the auditory path- brain stem lesions include increased I-to-V
ways in the brain stem, or both. Brain stem interpeak latency, a low-amplitude or absent
auditory evoked potentials may be per- wave V, and an absence of all waveforms.
formed in awake and cooperative patients or Results of BAEP studies are useful not only
those with altered mental status (for exam- in evaluating hearing and specific neuro-
ple, sedation, general anesthesia, or coma). logic disorders but also in evaluating the
The rationale for these studies in patients significance of certain symptoms and signs.
with neurologic disease is the close correla- Brain stem auditory evoked potentials may
tion between specific auditory waveforms be useful in evaluating patients with sus-
and structures in the brain stem. pected acoustic neuroma, multiple sclerosis,
Brain stem auditory evoked potentials are or brain stem glioma. The studies also pro-
usually performed with a click stimulus that vide prognostically important information
is delivered to each ear and that activates the about comatose patients. For BAEPs to as-
peripheral and central auditory pathways. sess peripheral auditory function, the
Auditory stimuli cause sequential activity in method must be altered.
CN VIII, the cochlear nucleus, the superior This chapter reviews the method, inter-
olivary nucleus, the lateral lemniscus, and pretation, and clinical applicability of BAEPs
the inferior colliculus. Five prominent ver- in patients with neurologic disease.
204
Brain Stem Auditory Evoked Potentials in Central Disorders 205
AUDITORY ANATOMY AUDITORY EVOKED

AND PHYSIOLOGY POTENTIALS IN NORMAL
SUBJECTS
Knowledge of the auditory system is essen-
tial to understand the structures that are se- Stimulation of the peripheral auditory ap-
quentially activated during BAEPs.1 The au- paratus in normal subjects may produce
ditory system begins with the peripheral seven vertex positive waveforms, labeled
auditory apparatus. The cochlea and the spi- I-VII (note the use of Roman numerals)
ral ganglion must be activated (with monau- (Fig. 17-1).1 Waves VI and VII are variably
ral stimulation) before the central auditory present and so are not useful clinically. Con-
pathways can be assessed. The initial central ventional audiometric earphones are used to
structure that is activated is the auditory por- deliver a click, that is, an electric square
tion of CN VIII, which enters the brain stem wave. The stimulus that optimally activates
at the pontomedullary junction. Sequential the central auditory system is maximal to the
activation involves the cochlear nucleus in click threshold for each ear. Monaural stim-
caudal pons, the superior olivary complex in ulation is used, with the contralateral ear
caudal to mid-pons, the lateral lemniscus in masked by white noise. Because binaural
mid-pons, and the inferior colliculus in cau- stimulation may fail to reveal abnormality in
dal midbrain. In normal subjects, hearing is a patient with a unilateral auditory lesion, its
associated with bilateral activation of the au- use should be avoided. Usually, the pre-
ditory pathway, maximal contralateral to the ferred stimulus for waveform recognition is
ear stimulated. Brain stem auditory evoked 65-70 dB above the click hearing threshold,
potentials may be associated with activation the click sensation level The optimal stimulus
of the brain stem pathways that are involved repetition rate for identifying waveforms is
with sound localization rather than with approximately 10/second. Electrodes are
hearing. placed on each earlobe (Al and A2) and at
Figure 17-1. Normal brain stem auditory evoked potentials in a 26-year-old man. I-V, characteristic vertex positive
waveforms. (From Daube JR, Regan TJ, Sandok BA, Westmoreland BF. Medical Neurosciences, 2nd ed. Little, Brown
and Company, Boston, 1986, p 361. By permission of Mayo Foundation.)
the vertex (CZ) to record the auditory wave-

forms (Fig. 17-1). Mastoid electrodes are
not used routinely because of increased
muscle artifact.
The specific generators of all the wave-
forms have not been clarified; however, the
anatomical regions that are activated are
known. Importantly, the proposed sites of
waveform generation are often based on lim-
ited data from patients with brain stem le-
sions.1"3 It is not known whether BAEPs are
generated by nuclei or tracts or a combina-
tion of the two. Wave I represents the distal
action potential of CN VIII and appears as a
negative potential at the ipsilateral ear elec-
trode. If wave I is absent, central auditory
conduction cannot be assessed reliably. Sup-
plementary electrodes, for example, a nee-
dle electrode in the external auditory canal,
may help register this wave if it cannot be
recorded with a conventional earlobe elec-
trode. Wave II may be generated by either
the ipsilateral proximal CN VIII or the
cochlear nucleus. Wave III is likely related to
activation of the ipsilateral superior olivary
nucleus. Wave IV is produced by activation
of the nucleus or axons of the lateral lem-
niscus. Wave V appears to result from activa-
tion of the inferior colliculus. Waves VI and
W/are presumed to be generated by the me-
dial geniculate body and the thalamocorti-
cal pathways, respectively.
Physiologic variables in normal subjects Figure 17-2. Effect of age (weeks) on wave I and wave
that may alter BAEPs include age, sex, and V absolute latencies in an infant of 28-week gestation.
auditory acuity.4 Stimulus repetition rate, in- IPL, interpeak latency; Stim AS, stimulation of left ear.
tensity, and polarity also may affect BAEPs. (From Stockard JE, Westmoreland BF. Technical con-
The age of the patient may affect waveform siderations in the recording and interpretation of the
brain stem auditory evoked potiential for neonatal neu-
morphology and latency. Brain stem audi- rologic diagnosis. Am J EEC Technol 21:31-54, 1981.
tory evoked potentials can be recorded even By permission of the American Society of Electroneu-
in premature infants, but the absolute and rodiagnostic Technologists.)
interpeak latencies are more prolonged
than in older patients (Fig. 17-2) .5 By 2 years
of age, the latencies are about the same as sessed. The absence of waves II-Vin patients
the normal values of adult subjects. Persons with a normal wave I or an increased wave
older than 60 years have a statistically sig- I-V interpeak latency cannot be explained
nificant increase in BAEP latencies com- on the basis of hearing loss alone. Common
pared with those of younger subjects. The BAEP findings in patients with hearing loss
BAEP interpeak latencies are significantly are a delay in the absolute latencies of all
shorter in women than in men. waveforms, absence of only wave I with de-
For patients with hearing loss, a higher layed waves II-V, or absence of all waveforms.2
stimulus intensity is required to activate the The stimulus variables used may affect
central auditory pathways. Significant pe- substantially the interpretation of the stud-
ripheral auditory dysfunction may not allow ies. Stimulation rates greater than 10 per sec-
brain stem auditory conduction to be as- ond may be associated with a significant in-
crease in absolute and interpeak latencies (1) increase stimulus intensity, (2) change
and a decrease in waveform amplitude. De- the click polarity to condensation, (3) slow
creasing stimulus intensity affects the BAEPs the stimulus rate, (4) use an external canal
much like hearing loss does. Stimulus in- supplementary electrode, or (5) decrease
tensities less than 65-70 dB above the sen- muscle artifact. Sedation with chloral hy-
sation level increase absolute and interpeak drate or diazepam given orally may be used
latencies, decrease waveform amplitude, and if patients are unable to relax or if excessive
alter waveform morphology. The polarity of muscle artifact is present. Never hesitate to
the click produces movement of the ear- take extra time to perform these maneuvers.
phone diaphragm away from the tympanic The presence of a discrete wave I and the
membrane (rarefaction) or toward the tym- demonstration of reproducibility for all
panic membrane (condensation). The former waves I to V are critical for rational inter-
polarity may be preferred because of an in- pretation of the study.
crease in wave I amplitude. Occasionally, Recent advances in BAEPs have focused
condensation may produce a more obvious on improved methodology. The three main
wave I. A mixture of the two polarities (al- areas of emphasis have been (1) increasing
ternating) is not used routinely because of al- wave definition, (2) decreasing recording
terations in waveform morphology and in- time, and (3) increasing the objectivity of de-
terpeak latencies. tecting and identifying waves. Methods that
use nonconventional averaging formulas,
steady state responses, improved signal-to-
noise ratio calculations, and template corre-
METHODS lation analysis have been developed.6 These
newer techniques may have application for
In patients with suspected neurologic dis- studies involving patients whose cooperation
ease, a BAEP study begins with an assessment is minimal, for example, newborns and chil-
of peripheral auditory function. A bedside dren. Equipment in which some of these
test of auditory acuity should be conducted, techniques are incorporated probably will
for example, by using a wristwatch at a fixed become more available commercially.
distance from the patient's ears. The ratio-
nale for the study and the proposed method
should be explained to the patient. The time USE OF BRAIN STEM
required to perform BAEPs depends on sev-
eral clinical factors, but the study usually can AUDITORY EVOKED
be completed in 30-45 minutes. The opti- POTENTIALS IN CLINICAL
mal filter bandpass is 100-3000 Hz, and the PROBLEM SOLVING
bandpass should be held constant for clini-
cal BAEP studies. Averaging is performed on Brain stem auditory evoked potential vari-
10 ms of data after auditory stimulation (Fig. ables evaluated in patients with suspected
17-1). The channel derivations include ip- neurologic disease include measurement of
silateral ear to vertex and contralateral ear absolute waveform latencies and interpeak
to vertex. In some circumstances, an ipsilat- latencies (I-III, III-V, and I-V) and deter-
eral ear to contralateral ear derivation may mination of wave V/I amplitude ratio.7 Nor-
assist in identifying wave I. Wave I is identi- mative data obtained by similar methods,
fied in the ipsilateral ear derivation as a neg- preferably within the same laboratory,
ative peak (near-fieldpotential) at the ear and should be available to determine latency and
as a positive peak (far-field potential) at the amplitude criteria for abnormal BAEPs.
vertex (Fig. 17-1). The contralateral ear Right and left ear evoked potential studies
channel may be useful for distinguishing should be compared only by using identical
wave IV from wave V. At least two averages stimulus variables.
of 2000-4000 responses are obtained from The use of the BAEPs provides an objec-
each ear. Additional trials may be necessary tive physiologic measure that complements
to recognize waveforms. If wave I is not iden- the findings of the clinical history and ex-
tified, the following maneuvers may help: amination and neuroimaging. Categories of
clinical problems for which BAEPs can be tivity if the brain stem lesion does not involve
used include the central auditory pathways. Clinical stud-
1. Confirmation of brain stem abnormal- ies have indicated that BAEPs are comple-
ity if the symptoms and signs of brain mentary to magnetic resonance imaging
stem dysfunction are equivocal (MRI) studies for certain central auditory ab-
2. Confirmation of brain stem abnormal- normalities, for example, acoustic neuromas.2
ity in patients known to have diffuse Brain stem auditory evoked potential ab-
or multifocal central nervous system normalities have been observed in several
(CNS) disease specific neurologic disorders, including
3. Screening for brain stem dysfunction cerebellopontine angle tumors, demyelinat-
in patients who have symptoms that ing disease, brain stem tumors, brain stem
usually refer to a peripheral cranial infarcts and hemorrhages, coma, and leuko-
nerve or sensory organ, for example, dystrophies.4 The sensitivity and specificity
hearing loss, dizziness, peripheral fa- of BAEPs for these neurologic diseases have
cial weakness, diplopia, and peripheral been determined. The most common neu-
jaw weakness. rologic indication for BAEPs is evaluation
The proposed advantages of BAEPs, com- for suspected acoustic neuroma or multiple
pared with neuroimaging studies, are lower sclerosis. The rationale for BAEP studies in
cost, less discomfort for patients, and a these suspected disorders is to demonstrate
shorter waiting time. Also, sequential BAEP an electrophysiologic alteration indicative of
studies are easier to perform. A disadvantage a CNS abnormality and to provide informa-
of BAEPs is the lack of specificity and sensi- tion about the anatomical localization of the
Rgure 17-3. Brain stem auditory

evoked potentials in a patient with
multiple sclerosis are abnormal be-
cause of prolonged I—V interpeak la-
tency. The patient had no symptoms
or signs of brain stem disease, and
the neurologic examination find-
ings were unremarkable.
Figure 17-4. Preoperative and postoperative brain stem auditory evoked potentials in a 74-year-old woman with a
right acoustic neuroma associated with brain stem compression. The tumor was resected. No response was observed
after stimulation of the right (Rt) ear, either preoperatively or postoperatively. Postoperatively, stimulation of the
left (Lt) ear showed significant shortening of the III-V interpeak latency related to resection of the lesion. (From
StockardJJ, Sharbrough FW. Unique contributions of short-latency auditory and somatosensory evoked potentials
to neurologic diagnosis. Prog Clin Neurophysiol 7:231-263, 1980. By permission of S. Karger, AG.)
lesion. Brain stem auditory evoked poten-

tials may be useful as a prognostic indicator
in patients with coma. Also, BAEPs have
been used to monitor response to therapy.9
The most common BAEP abnormality in
patients with CNS disease is a prolonged I-V
interpeak latency (Fig. 17-3). Other alter-
ations include the absence of all waveforms,
a decreased V/I amplitude ratio, and preser-
vation of wave I with poorly formed waves
II-V (Figs. 17-4 to 17-6). The I-III and
III-V interpeak latencies may be useful in
determining the anatomical localization of
auditory dysfunction. In patients with a pro-
longed I-III interpeak latency and a normal
III-V interpeak latency, the auditory dys-
function is assumed to be located between
the distal part of CN VIII (near the cochlea)
and the superior olivary nucleus, ipsilateral
Figure 17-5. Brain stem auditory evoked potentials in
to the ear stimulated. In patients with a pro- a 39-year-old woman with multiple sclerosis show the
longed III-V interpeak latency and a normal combined abnormalities of prolonged I-V interpeak la-
I-III interpeak latency, the auditory con- tency and decreased V/I amplitude ratio.
Figure 17-6. Brain stem auditory evoked potentials in a 30-year-old woman with multiple sclerosis reveal only
wave I bilaterally. Waves II-V are absent. Stim. A.S., stimulation of left ear. Stim. A.D., stimulation of right ear.
duction defect likely is located between the

superior olivary nucleus and the inferior
colliculus, ipsilateral to the ear stimulated
(Fig. 17-7).
Acoustic Neuroma
Brain stem auditory evoked potentials are a
reliable indicator of the presence of cere-
bellopontine angle tumors affecting CN VIII
(see Fig. 17-4).10 Auditory conduction ab-
normalities almost invariably are found in
patients with acoustic neuroma, even in
those who are asymptomatic.11 Brain stem
auditory evoked potentials may be abnormal
when the findings of other audiometric stud-
ies and even neuroimaging studies fail to dis-
close an alteration. According to Chiappa,2
"BAEPs are the most sensitive screening test
when an acoustic neuroma is suspected."
The characteristic BAEP changes are pro-
longed I-V and I-III interpeak latencies ip-
silateral to the tumor. The absolute or asym-
metrical prolongation of the I-III interpeak
latency may be the most sensitive BAEP vari-
able. Patients with autosomal dominant neu-
rofibromatosis in whom bilateral acoustic
neuromas develop may have normal BAEPs
if the tumors are asymptomatic, small, and
confined to the intracanalicular region. Nor-
mal BAEP results in a patient with symptoms Figure 17-7. Brain stem auditory evoked potentials
suggestive of acoustic neuroma, such as from a normal subject and a patient with a brain stem
glioma. The latter study shows prolonged III-V in-
dizziness and hearing loss, argue strongly terpeak latency, indicating a central auditory conduc-
against the diagnosis. Magnetic resonance tion defect between the caudal pons and caudal mid-
imaging has a low diagnostic yield in patients brain.
with normal BAEPs and suspected acoustic tials may implicate brain stem disease in
neuroma. Chiappa2 has suggested that neu- patients with vague, ill-defined, nonspecific
roimaging studies are unnecessary in most symptoms. Potentially, BAEPs may be used
patients with suspected acoustic neuroma to monitor the response to treatment of pa-
and normal BAEPs (if the studies are per- tients with multiple sclerosis. Ultimately,
formed correctly). Other cerebellopontine multiple sclerosis is a clinical diagnosis, and
angle tumors, for example, meningiomas, the results of BAEP studies must be inter-
may not produce BAEP abnormalities until preted carefully in conjunction with the rest
the tumor is large and involves CN VIII. of the neurologic evaluation.
Thus, the diagnostic yield of BAEPs for these
tumors early in the course of disease is low,
and neuroimaging is a more important neu- Intrinsic Brain Stem Lesions
rodiagnostic technique.
Brain stem auditory evoked potentials are
abnormal in most patients with brain stem
Demyelinating Disease tumors, for example, pontine glioma (Figs.
17-7 and 17-8).4 The findings are similar to
The frequency of BAEP abnormalities in pa- those in patients with other central auditory
tients with suspected multiple sclerosis is re- conduction defects. Brain stem tumors are
lated directly to the likelihood the person often associated with bilateral BAEP abnor-
has the disorder and the presence of clini- malities (maximal ipsilateral to the lesion).
cal evidence for brain stem disease.12 Ab- Extra-axial tumors may not be associated
normal BAEP results are common in pa- with BAEP abnormalities unless there is di-
tients with clinically definite multiple rect compression on and disruption of the
sclerosis (Figs. 17-3, 17-5, and 17-6). In a brain stem. Ependymomas of the fourth ven-
study of 60 patients with clinically definite tricle and cerebellar tumors may also be as-
multiple sclerosis and symptoms or signs of sociated with central auditory conduction
brain stem lesions, 34 (57%) had abnormal defects.
BAEPs, and of 33 patients with brain stem The results of BAEP studies are variable
disease and possible multiple sclerosis, 7 in brain stem strokes, that is, infarcts and
(21%) had central auditory conduction de- hemorrhages (Fig. 17-9). Although abnor-
fects.2 Importantly, BAEPs may be abnormal mal BAEPs are found in these patients,
in patients with suspected multiple sclerosis BAEPs are normal if the stroke spares the
who do not have evidence of brain stem le-
sions. Of patients without symptoms or signs
of brain stem disease related to multiple scle-
rosis, 20%-50% may have abnormal BAEPs.
Brain stem auditory evoked potential re-
sults in patients with demyelinating disease
include a unilateral or bilateral prolonged
I-V interpeak latency and a decreased V/I
amplitude ratio (Figs. 17-3,17-5, and 17-6) .4
Characteristically, patients with bilateral
central auditory conduction defects do not
have auditory symptoms or abnormal click
thresholds. The usefulness of BAEPs in these
cases includes identifying unsuspected brain
stem lesions in patients with an anatomically
unrelated disorder such as optic neuritis.
Brain stem auditory evoked potentials may Figure 17-8. Brain stem auditory evoked potentials in
also provide confirmatory evidence of a CNS a 66-year-old woman before and after (p) radiation ther-
apy for a brain stem tumor. Note the significant short-
alteration when neurologic evaluation does ening of III-V interpeak latency after treatment. (From
not suggest the diagnosis of demyelinating Stockard et al.7 By permission of Churchill Living-
disease. Brain stem auditory evoked poten- stone.)
ence of wave I and the absence of waves II

to V bilaterally. Brain stem auditory evoked
potentials are useful for monitoring auditory
pathway integrity in an unconscious patient
during a neurosurgical procedure.13
SUMMARY
Brain stem auditory evoked potentials are
performed primarily in patients with sus-
pected neurologic disorders to determine
whether there is evidence of a CNS lesion.
These sensory-evoked potential studies are
highly sensitive to auditory conduction de-
fects, but the findings are not pathologically
specific. Brain stem auditory evoked poten-
Figure 17-9. Top, Brain stem auditory evoked potentials,
in a normal subject and, bottom, in a patient who died
tials provide data that are highly repro-
of brain stem hemorrhage. Note the absence of waves ducible and objective and lend themselves
III-V in the latter study. to sequential studies for comparison. Brain
stem auditory evoked potentials are nonin-
vasive and can be performed not only in the
clinical neurophysiology laboratory but also
brain stem auditory pathways, for example, in a hospital room or the intensive care unit.
infarction of the posterior inferior cerebel- Patient cooperation is not critical, because
lar artery (lateral medullary syndrome). sedation is permitted. Important factors that
Brain stem transient ischemic attacks usually need to be considered for accurate inter-
do not produce abnormal BAEPs.2 pretation of BAEPs include the patient's age,
sex, and auditory acuity. The diagnostic yield
of BAEPs has been confirmed in patients
Coma and Brain Death with acoustic neuromas or intra-axial brain
stem lesions involving the auditory pathways.
Brain stem auditory evoked potentials may Brain stem auditory evoked potential stud-
be normal or abnormal in a comatose pa- ies appear to be complementary to structural
tient, depending on the underlying cause neuroimaging studies, for example, MRI.
and the presence of a brain stem lesion.
Brain stem auditory evoked potentials may
provide prognostically important informa- REFERENCES
tion about outcome.8 Patients in whom
BAEPs are absent (if peripheral auditory dys- 1. Moller AR. Auditory neurophysiology. J Clin Neu-
function can be excluded) are unlikely to rophysiol 11:284-308, 1994.
survive. Lesions confined to the cerebral 2. Chiappa KH. Evoked Potentials in Clinical Medi-
cine, 3rd ed. Lippincott-Raven, Philadelphia, 1997,
hemispheres are not associated with abnor- pp 31-130.
mal BAEPs unless the upper brain stem is 3. Nuwer MR. Fundamentals of evoked potentials and
functionally disrupted. Brain stem auditory common clinical applications today. Electroen-
evoked potentials may be useful in assessing cephalogr Clin Neurophysiol 106:142-148, 1998.
the integrity of brain stem structures when 4. Markand ON. Brain stem auditory evoked poten-
tials. J Clin Neurophysiol 11:319-342, 1994.
the findings on neurologic examination are 5. Picton TW, Taylor MJ, Durieux-Smith A. Brain stem
unreliable, for example, after treatment with auditory evoked potentials in pediatrics. In Aminoff
a high dose of barbiturates or with the pa- MJ (ed): Electrodiagnosis in Clinical Neurology,
tient under general anesthesia. The brain- 3rd ed. Churchill Livingstone, New York, 1992,
pp 537-569.
dead person invariably has abnormal BAEPs, 6. Hall JW III, Rupp KA. Auditory brain stem re-
with the characteristic findings of the bilat- sponse: recent developments in recording and
eral absence of all waveforms or the pres- analysis. Adv Otorhinolaryngol 53:21-45, 1997.
7. Stockard JJ, Pope-Stockard JE, Sharbrough FW. schwannomas (acoustic neuromas): the value of
Brain stem auditory evoked potentials in neurology: neurophysiology for evaluation and prediction of
methodology, interpretation, and clinical applica- auditory function in 420 cases. Neurosurgery 40:
tion. In Aminoff MJ (ed). Electrodiagnosis in Clin- 919-929, 1997.
ical Neurology, 3rd ed. Churchill Livingstone, New 11. Tandon OP. Average evoked potentials—clinical
York, 1992, pp 503-536. applications of short latency responses. Indian J
8. Guerit JM. Evoked potentials: a safe brain-death Physiol Pharmacol 42:172-188, 1998.
confirmatory tool? Eur J Med 1:233-243, 1992. 12. Drislane FW. Use of evoked potentials in the diag-
9. Nuwer MR, Packwood JW, Myers LW, Ellison GW. nosis and follow-up of multiple sclerosis. Clin Neu-
Evoked potentials predict the clinical changes in a rosci 2:196-201, 1994.
multiple sclerosis drug study. Neurology 37:1754- 13. Kumar A, Bhattacharya A, Makhija N. Evoked po-
1761, 1987. tential monitoring in anaesthesia and analgesia.
10. Matthies C, Samii M. Management of vestibular Anaesthesia 55:225-241, 2000.
Chapter 18
AUDIOGRAM, ACOUSTIC REFLEXES,
AND EVOKED OTOACOUSTIC
EMISSIONS
Wayne 0. Olsen
AUDIOGRAM EVOKED OTOACOUSTIC EMISSIONS

Speech Recognition Thresholds Neonatal Screening
Word Recognition Pseudohypacusis
ACOUSTIC REFLEX SUMMARY
Cranial Nerve VIII Versus Cochlear Findings
Cranial Nerve VII
Brain Stem
Hearing tests performed with pure-tone and (CN VIII) lesions and to evaluate a portion
speech stimuli are used to assess hearing of CN VII.
function in patients for whom there is con- Evoked otoacoustic emissions require no
cern about hearing or balance. These tests voluntary behavioral response, assess pre-
can help determine whether there are re- neural function in the cochlea, and are use-
lated balance and hearing problems, can ful screening tests for hearing in newborn
document difficulties in communication at- infants and for documenting pseudohypa-
tributable to hearing disorders, and can help cusis, that is, feigned or exaggerated hear-
establish a diagnosis. On the basis of the pat- ing loss.
terns of results from these tests, hearing loss Pure-tone and speech audiometric testing
can be categorized as follows: constitute the basic hearing evaluation of
1. Conductive—abnormality of the ear persons who have a balance or hearing prob-
canal, tympanic membrane, or middle lem or both. Acoustic reflex tests often are
ear ossicles, or a combination of these administered to patients who have tinnitus,
2. Sensorineural—disorder of the co- unilateral or asymmetric sensorineural hear-
chlea or cranial nerve (CN) VIII ing loss, vestibular disorders, or facial paral-
3. Mixed—a combination of conductive ysis (or a combination of these) to help de-
and sensorineural disorders termine whether the lesion involves the end
Acoustic reflex tests assess contraction of organ or the acoustic nerve portion of CN
the stapedius muscle and require no volun- VIII or to help define the site of involvement
tary behavioral response from the patient. of CN VII. Evoked otoacoustic emissions are
They are used primarily to help differenti- used to test cochlear function in newborn
ate sensory (cochlear) lesions from neural infants, young children undergoing hearing
214
Audiogram, Acoustic Reflexes, and Evoked Otoacoustic Emis s 215
evaluations, and older children or adults sus- sponse on hearing the designated signals.
pected of feigning or exaggerating hearing The responses of the patient are plotted on
loss during routine pure-tone and speech a standardized chart called an audiogram and
testing. compared with internationally established
reference levels for normal hearing.
A conventional audiogram format is
shown in Figure 18-1, with intensity in deci-
AUDIOGRAM bels of hearing level (HL) (the American Na-
tional Standards Institute1) on the ordinate
Basic audiologic tests use pure tones deliv- and frequency (125-8000 Hz) on the ab-
ered by standard earphones and bone vi- scissa. The "0 decibel HL line" is an inter-
brators to assess thresholds of hearing (just nationally accepted reference representing
barely audible) for air-conducted and bone- the average hearing sensitivity for young
conducted stimuli. These tests are akin to people with normal hearing. The normal
the Schwabach and Rinne tuning fork tests. range indicated on the right side of the au-
However, the presentation of electronically diogram extends to 25 dB HL, because per-
generated signals through standard trans- sons with hearing thresholds in this range,
ducers allows testing over a greater fre- at least for frequencies of 500-4000 Hz, gen-
quency and intensity range and with far erally do not report difficulty hearing and
greater precision than is possible with tun- understanding conversational speech in
ing forks. Children as young as 6 months can quiet surroundings. People whose hearing
be conditioned to respond to pure-tone thresholds are
stimuli at or near threshold levels. School- 1. in the 26-45 dB HL range—have mild
age and older children and adults need only hearing loss and difficulty hearing soft
to be instructed to provide a behavioral re- or distant speech
Figure 18-1. Audiogram showing pure-tone and speech test results for sensorineural hearing loss (right ear) and
conductive hearing loss (left ear). Degree of hearing loss on right ordinate. ANSI, American National Standards
Institute.
2. in the 46-65 dB HL range—have dif- level of 25 dB, that is, 25 dB above the speech
ficulty hearing speech at normal con- recognition thresholds. The score of 76% re-
versational levels and are considered to veals less than perfect understanding of the
have moderate hearing loss test items through the right ear. The score
3. in the 66-85 dB HL range—have severe for the left ear is perfect (100%), as expected
hearing loss, indicating difficulty hear- for conductive hearing loss. These results
ing even loud speech demonstrate that speech loud enough to be
4. greater than 85 dB HL—have pro- heard easily was understood perfectly by the
found hearing loss left ear (conductive hearing loss) but not by the
The term deaf is reserved for people in group right ear (sensorineural hearing loss). Scores of
4, and hearing impaired and hard of hearing 90% to 100% indicate that people should
are used for those in groups 1-3. have little difficulty understanding speech
The threshold data shown in Figure 18-1 loud enough to be heard easily in quiet en-
reveal mild hearing loss in the left ear vironments. Correct responses to 70% to
(marked by "X" in Fig. 18—1), and, in the 88% of the monosyllables in a given list sug-
right ear (marked by "O"), hearing sensitiv- gest occasional difficulty; 60% to 68%, defi-
ity that ranges from normal for the lower fre- nite difficulty; 40% to 58%, marked diffi-
quencies (125 and 250 Hz) to mild hearing culty; and less than 40%, extreme difficulty
loss for the middle frequencies (500—3000 in understanding speech. Persons who have
Hz), and moderate to severe hearing loss for CN VIII lesions often have considerable dif-
the higher frequencies. The hearing loss for ficulty comprehending speech in the af-
the right ear is sensorineural, as shown by the fected ear.2
interweaving air conduction and bone con-
duction thresholds (marked by "O" and "[,"
respectively, in Fig. 18-1). The hearing loss ACOUSTIC REFLEX
in the left ear is conductive, as indicated by
the separation of the air-conduction and Intense stimulation of either ear causes con-
bone-conduction thresholds (marked by "X" traction of the stapedius muscle in both ears.
and "]," respectively, in Fig. 18-1). This protective response, called the acoustic
reflex, is measured easily and quickly with a
device called an immittance unit (Fig. 18—2).
Speech Recognition Thresholds The system microphone monitors the level
of a low-frequency tone (226 Hz) main-
The audiogram in Figure 18-1 also shows tained in the space between the probe tip
speech recognition thresholds that are based sealed in the ear canal and the tympanic
on correct responses to 50% of spondaic membrane while intense tones (500, 1000,
words (spondees are words with two syllables 2000, or 4000 Hz) are presented to the op-
of equal stress when spoken, such as "air- posite ear or, in some situations, to the same
plane" and "baseball") at the hearing levels ear as the 226 Hz probe tone. Contraction
indicated (40 dB HL). These values indicate of the stapedius muscle, which is attached to
mild hearing loss for speech, as do pure-tone the neck of the stapes, stiffens the middle
thresholds in the 500-2000 Hz frequency ear system, thereby altering the level of the
range. 226 Hz tone maintained between the probe
tip and the tympanic membrane. This
change is the acoustic reflex response mea-
Word Recognition sured by the immittance unit.
Measurement of acoustic reflexes requires
The ability to understand lists of monosylla- an intact tympanic membrane, mobile mid-
bles (one-syllable words such as "thin," dle ear ossicles (no conductive hearing loss),
"sack," and "vote," presented singly) at hearing adequate to allow sufficient stimu-
clearly audible levels is shown by the per- lation of the ear with at least one of the
centage correct word recognition score. In above-mentioned tones, intact CN VII, CN
the example shown in Figure 18-1, lists of VIII reflex arc in the brain stem, and
monosyllables were presented at a sensation stapedius muscle attachment to the stapes.
Audiogram, Acoustic Reflexes, and Evoked Otoacoustic Emissions 217
Figure 18-2. Block diagram of immittance unit showing setting for eliciting contralateral acoustic reflexes (stimu-
lus presented through earphone). HL, hearing level.
Because of the complexity of this system, var- ity (Table 18-1). The sensitivity and speci-
ious response patterns emerge (Table 18-1). ficity of acoustic reflex and reflex decay tests
In conjunction with the case history and are 85%; that is, this test combination cor-
other audiologic test results, acoustic reflex rectly identifies lesions of CN VIII and cor-
testing provides valuable diagnostic infor- rectly rules out such lesions 85% of the time.
mation.
Cranial Nerve VII

Cranial Nerve VIII Versus
Cochlear Findings Measurement of stapedius muscle contrac-
tion on the same side as facial paralysis re-
The absence of acoustic reflexes or response veals that the CN VII lesion is distal to the
only to very intense tones in an ear with sen- branch that innervates the stapedius muscle.
sorineural hearing loss no worse than severe The absence of a reflex response on the
in degree makes one suspect neural (CN same side as facial paralysis indicates that the
VIII) involvement on the side of the stimu- involvement of CN VII is medial to the
lated ear.3 Similarly, acoustic reflex decay, that stapedius branch of the nerve (Table 18-1).
is, diminished amplitude of the acoustic re-
flex response to less than half within 5 sec-
onds to a 500 Hz or 1000 Hz tone delivered Brain Stem
10 dB above acoustic reflex threshold, sug-
gests a lesion of CN VIII. Elicitation of the The absence of acoustic reflexes witii con-
acoustic reflexes by normal levels of stimu- tralateral stimulation (for example, stimu-
lation and the absence of reflex decay indi- lating the right ear and measuring the
cate that the middle and inner ears are nor- acoustic reflex in the left ear and vice versa)
mal or, in the case of sensorineural hearing but their occurrence with ipsilateral stimu-
loss, indicate sensory (cochlear) abnormal- lation (that is, stimulating and measuring
Table 18-1. Audiogram and Acoustic Reflex Findings for Various Conditions
ACOUSTIC
AUDIOGRAM REFLEXES
Normal Conductive Sensorineural Normal Abnormal
Condition Hearing Hearing Loss Hearing Loss Response Response
Normal X X
Cerumen plug X X
Thickened tympanic X X
membrane
Perforated tympanic X X
membrane
Otitis media fluid X X
Ossicular discontinuity X X
Otosclerosis stapes fixation X X
Sensorineural loss
Cochlea X X
CN VIII X X
Facial paralysis
Medial to stapedial X X
branch CN VII
Peripheral to stapedial X X
branch CN VII
CN, cranial nerve.
Modified from Keating LW, Olsen WO. Practical considerations and applications of middle-ear impedance mea-
surements. In Rose DE (ed). Audiological Assessment, 2nd ed. Prentice-Hall, Englewood Cliffs, New Jersey, 1978,
pp 336-367. By permission of the author.
the response in the same ear) indicates a simultaneously. Displays of normal TEOAEs
brain stem lesion that interrupts the cross- and DPOAEs are shown in Figures 18-4 and
ing acoustic reflex tracts (Fig. 18—3). 18—5, respectively. The different segments in
Figure 18-4 show the waveform, stability,
level, and spectrum of the click stimulus as
well as the waveform, reproducibility, level,
EVOKED OTOACOUSTIC and spectrum of the response from the
EMISSIONS cochlea, signal-to-noise ratio of the re-
sponse, the noise level in the ear canal, test
Evoked otoacoustic emissions (EOAEs) reflect time, and other variables. The graph in Fig-
the response of electromotile activity within ure 18-5 shows the amplitude of the distor-
the cochlea in response to external sound tion products generated within the cochlea
stimuli.4 This miniscule activity can be mea- (line graph near center on left) in response
sured in the ear canal with a sensitive mi- to two tones presented simultaneously, the
crophone sealed in the ear canal. The out- noise level in the ear canal, frequency sepa-
put of the microphone is averaged to reduce ration, level of the stimulus tones, and test
the inherent physiologic and environmental time.
noise in the ear canal. Transient evoked oto- Robust EOAEs, such as those shown in Fig-
acoustic emissions (TEOAEs) are measure- ures 18-4 and 18-5, indicate good function
ments of the active cochlear response to of the cochlear outer hair cells and gener-
clicks. Distortion product otoacoustic emissions ally are associated with normal hearing sen-
(DPOAEs) reflect the interaction within the sitivity, 25 dB HL or better for frequencies
cochlea to stimulation with two pure tones of 1000-6000 Hz. Low-frequency physiologic
Figure 18-3. Contralateral and ipsilateral acoustic reflex arcs. Crossing tracts for contralateral reflexes are shaded.
N, cranial nerve. (From Wiley TL, Block MG. Acoustic and nonacoustic reflex patterns in audiologic diagnosis. In
Silman S [ed]. The Acoustic Reflex: Basic Principles and Clinical Applications. Academic Press, Orlando, Florida,
1984, pp 387-411. By permission of the publisher.)
noise in the ear canal, and occasionally low- 4000 Hz range provides important informa-
frequency environmental noise, precludes tion, because it suggests normal hearing sen-
measurement of Otoacoustic emissions for sitivity for the frequencies most important
frequencies less than 1000 Hz.5 Neverthe- for hearing and understanding speech.
less, observation of EOAEs for the 1000- Evoked Otoacoustic emissions rarely are ob-
Kgure 18-4. Display of transient evoked Otoacoustic emission (TEOAE) measurement showing test variables and
response.
Figure 18-5. Display of distortion product otoacoustic emission (DPOAE) showing test variables and response.
served at a given test frequency when sen- otoacoustic emissions often are used to help
sorineural hearing loss is 30 dB HL or confirm behavioral observations made dur-
greater at that frequency or when conduc- ing an audiologic evaluation of a young
tive hearing loss blocks transmission of the child.
low-level otoacoustic emissions from the
cochlea back to the microphone in the ear
canal. Pseudohypacusis
For patients who are not cooperative or who
Neonatal Screening are suspected of feigning or exaggerating a
hearing loss greater than 30 dB HL, EOAE
Because EOAE tests can be completed tests provide a fast and objective assessment
quickly and do not require a voluntary re- of the peripheral portion of the auditory sys-
sponse and because elicitation of a response tem. If behavioral responses indicate a hear-
suggests normal middle ear and cochlear ing loss that is mild or greater in degree for
function, they have been used as hearing frequencies between 1000 and 6000 Hz (or
screening tests in neonatal intensive care no behavioral responses are obtained), but
units and well-baby nurseries. The absence EOAEs are observed at various test frequen-
of a response raises the possibility of a sig- cies, there is reason to suspect pseudohypa-
nificant hearing loss that warrants further cusis and the need for additional testing to
audiologic and medical evaluation. Evoked establish hearing sensitivity more precisely.
SUMMARY screening tests for hearing in infants and in

patients suspected to have pseudohypacusis,
Audiologic testing in the form of pure-tone that is, feigned or exaggerated hearing loss.
air-conduction and bone-conduction audio-
grams provides diagnostic information
about the type of hearing loss (conductive, REFERENCES
sensorineural, or mixed) and the degree of
hearing loss and attendant communication 1. American National Standards Institute. American
difficulties. The addition of speech tests that National Standard Specification for Audiometers
use specific types of speech stimuli directly (ANSI S3.6-1996). The Society, New York, 1996.
2. Van DijkJE, Duijndam J, Graamans K. Acoustic neu-
assesses the patient's ability to hear and to roma: deterioration of speech discrimination related
understand speech. Acoustic reflex and re- to thresholds in pure-tone audiometry. Acta Oto-
flex decay tests are used to evaluate the in- laryngol 120:627-632, 2000.
tegrity of a complicated neural network in- 3. Bauch CD, Olsen WO, Harner SG. Preoperative and
postoperative auditory brain-stem response results
volving not only the auditory tracts to and for patients with eighth-nerve tumors. Arch Oto-
through the brain stem but also decussating laryngol Head Neck Surg 116:1026-1029, 1990.
pathways in the brain stem and the course 4. Robinette MS, Glattke TJ (eds). Otoacoustic Emis-
of CN VII to the innervation of the stapedius sions: Clinical Applications, 2nd ed. Thieme, New
muscle. Evoked Otoacoustic emission tests York, 2002.
5. Headley GM, Campbell DE, Gravel JS. Effect of
provide an objective measurement of the pe- neonatal test environment on recording transient-
ripheral hearing system from the external evoked Otoacoustic emissions. Pediatrics 105:1279-
ear through the cochlea and are useful 1285, 2000.
Chapter 19
BRAIN STEM AUDITORY EVOKED
POTENTIALS IN PERIPHERAL
ACOUSTIC DISORDERS
STIMULI ELECTRODES
INTERPRETATION APPLICATIONS
Absolute Latencies SUMMARY
Interaural Latency Differences
Interpeak Intervals
The otoneurologic assessment of patients is usually 30 minutes or less and seldom re-
who, because of complaints of dizziness, quires sedation.
hearing loss, or tinnitus, have a suspected
retrocochlear abnormality often includes
brain stem auditory evoked potential (BAEP) STIMULI
(also called brain stem auditory evoked re-
sponse [BAER]) testing. Because conductive, Clicks are the most effective stimuli in
hearing losses resulting from abnormalities of BAEP assessment; their short duration, 50-
the ear canal, tympanic membrane, or mid- 100 microseconds, and abrupt onset dis-
dle ear can affect BAEP waveform morphol- perse acoustic energy and provide good syn-
ogy and latencies, as can sensorineural disor- chronization of neural discharges across a
ders (lesions of the cochlea or cranial nerve broad frequency range. However, the im-
[CN] VIII), the potential effects of hearing portance of high-frequency hearing sensitiv-
losses must be considered in the interpreta- ity is accentuated by the spectral character-
tion of BAEP findings. istics of the earphone and by the response
Referral for otoneurologic BAEP evalua- characteristics of the ear canal and middle
tion depends on the patient's symptoms. ear, resulting in greater excitation in the
Typical symptoms include unilateral hearing 2000-4000 Hz range. Because this region is
loss, tinnitus, dizziness, unsteadiness, and fa- stimulated maximally by clicks, routine pure-
cial weakness. The BAEP evaluation records tone assessment is recommended before
neuroelectric potentials from CN VIII and BAEP evaluation. Hearing losses, particu-
ascending brain stem pathways in response larly in the 2000-4000 Hz frequency range
to click stimuli. Typically, surface electrodes can affect BAEP results.1 Behavioral thresh-
are placed at or near the vertex and the ears olds for clicks are not an adequate screen for
to record the responses. The evaluation time hearing, because the click's spectral spread
222
BAEPs in Peripheral Acoustic Disorders 223
of energy can yield relatively good thresh- ten indistinguishable from those of patients
olds despite significant hearing loss in the with lesions of CN VIII.
2000-4000 Hz range.
Interaural Latency Differences

INTERPRETATION
Another measure, interaural latency differ-
Three basic measurements are often made ences, compares wave-V latencies at the two
in the typical evaluation of BAEP waveforms: ears of the patient. The advantage of this
absolute latencies, wave-V interaural latency measure is that the patient serves as his or
differences, and interpeak intervals. Brain her own control. Normal variability for in-
stem auditory evoked potential waveform teraural latency differences is 0.2, 0.3, or 0.4
amplitude may be an unreliable criterion for ms. Larger wave-V latency differences be-
clinical testing because of marked variations tween ears are considered indicative of CN
among normal subjects.1'2 VIII involvement. However, the degree of
hearing loss in the 2000-4000 Hz range also
can influence the validity of such compar-
Absolute Latencies isons.3 When wave-V latency differences be-
tween ears exceed the predetermined crite-
Absolute latencies of the BAEP waves may be rion, the examiner must determine what
influenced by peripheral, that is, conductive influence, if any, the hearing loss has on the
or cochlear, hearing loss. Conductive hear- results. Adjustments in wave-V latency based
ing loss reduces the effective stimulus reach- on various levels of high-frequency hearing
ing the cochlea and causes absolute latency loss have been advocated, but the applica-
delays dependent on the degree of conduction of these corrections is often misleading
tive impairment. On average, a 0.4-ms shift and confusing.
can be expected for each 10 dB of conduc-
tive hearing loss.
For cochlear disorders, the hearing Interpeak Intervals
thresholds for 2000, 3000, and 4000 Hz are
important. As hearing thresholds for these Interpeak intervals reflect the time interval
frequencies become poorer, the latencies of from one neural generator to another. The
waves I, III, and V increase systematically. primary interpeak intervals (I-III, III-V, and
Presumably, this is because of the time I-V) separate a delayed wave-V absolute la-
delay of the traveling wave in the cochlea tency into its peripheral (I-III) and central
reaching more responsive apical (lower fre- (III-V) components. Normal I-III and III-V
quency) areas and also because of the de- intervals are each approximately 2 ms, which
crease in effective intensity stimulating the provide an overall I-V interval of approxi-
defective cochlea. These factors, and the fact mately 4 ms. An advantage of measuring the
that the cochlea produces more synchro- interpeak intervals is that they usually are
nous responses at its high-frequency basal not affected by moderate-to-severe levels of
end, lead to latencies that depend on the in- cochlear or conductive hearing loss. A pro-
tegrity of high-frequency hearing. Clinical longed I-V interval (longer than 4.54 ms)
experience has shown that when the average suggests a retrocochlear lesion, whereas con-
hearing thresholds for 2000, 3000, and 4000 ductive and cochlear hearing losses usually
Hz are equal to or greater than a 60 dB hear- have normal intervals (4.54 ms or less). The
ing loss, at least 60% of the absolute laten- main disadvantage to using the interpeak in-
cies for waves I, III, and V are abnormal. tervals is that wave I cannot always be iden-
Knowledge of these tendencies is important tified if peripheral hearing loss is mild, mod-
in the interpretation of absolute wave-V la- erate, or severe. In these cases, the examiner
tencies for patients with a suspected CN VIII must rely on the absolute latencies or in-
abnormality. Wave latency delays for patients teraural latency difference measures for
with similar cochlear hearing losses are of- interpretation. When all three measures—
fortably, and prevents collapse of the ear

canals. Its disadvantage is that it can be used
only once.
APPLICATIONS
Absolute latencies of waves I, III, and V have
been compared between patients with CN
VIII tumors (tumor group) and patients with
cochlear hearing loss (nontumor group)
matched for pure-tone audiometric config-
urations. Mean wave-I absolute latencies are
Figure 19-1. Etymotic ER-3A transducer, sound tube, usually similar between the tumor and non-
ear canal electrode (TIPtrode), and electrode lead.
(Nicolet Biomedical Instruments, Inc., Madison, WI). tumor groups, but mean latencies for waves
III and V are prolonged by as much as 1 ms
for the tumor group. The range of latencies
absolute latency, interaural latency differ- for waves I, III, and V is also considerably
ence, and interpeak intervals—are used col- larger for patients in the tumor group than
lectively, the sensitivity of BAEP for CN VIII for patients in the nontumor group who
nerve lesions is more than 90%,4~7 and the have a similar degree of cochlear hearing
specificity for cochlear hearing loss is nearly loss.
90%.4 Interaural latency differences that exceed
the 0.4-ms criterion identify more than 90%
of the patients with CN VIII tumors. If the
ELECTRODES criterion is decreased to 0.3 ms, the rate of
tumor detection increases only slightly and
Although conventional mastoid or earlobe the number of patients with cochlear hear-
electrodes usually allow recording of waves ing loss that exceeds the 0.3-ms criterion is
I, III, and V from patients with normal hear- substantial. The 0.4-ms criterion for inter-
ing sensitivity, wave I is difficult or impossi- aural latency differences appears to be a rea-
ble to identify in patients with mild, moder- sonable compromise.3
ate, or severe cochlear hearing loss. An ear Interpeak intervals have also been com-
canal electrode can enhance wave-I ampli- pared between tumor and nontumor groups
tude. The electrode is a disposable, soft,
foam plug wrapped in a thin layer of con-
ducting foil. It couples to a transducer
through flexible silicon tubing. Such an elec-
trode serves dual roles as a recording elec-
trode and a stimulus delivery system (Fig.
19-1).
In direct comparison with mastoid elec-
trodes, the ear canal electrode improves
wave-I amplitude by nearly 100% for patients
with normal hearing and 41%-127% for
those widi mild-to-severe hearing losses. In
a large sample of hearing-impaired patients,
wave I was identified easily 96% of the time
with the ear canal electrode, compared with
70% of the time with mastoid electrodes.8
The primary advantages of the ear canal Figure 19-2. Brain stem auditory evoked potential
electrode are that wave-I amplitude is im- recordings showing, A, normal waveforms (I, III, V)
proved for all degrees of hearing loss and and, B and C, abnormal waveforms of patients with a
the foam material is compressible, fits com- CN VIII tumor.
BAEPs in Peripheral Acoustic Disorders 225
Figure 19-3. Percentage of abnormal (delayed or absent) brain stem auditory evoked potentials (BAEPs) for 75 pa-
tients without tumor (Nontumor) and 75 patients with CN VIII tumor (Tumor) matched for hearing loss. I, III, V,
BAEP waves; I-III, III-V, I-V, interpeak intervals; ILD, wave V interaural latency difference. (From Bauch et al.4 By
permission of the American Speech-Language-Hearing Association.)
matched for hearing loss. The mean I-III in- these same criteria was 88% (false-positive
terval for the tumor group exceeds that of rate of 12%). Absolute latency measures for
the nontumor group by approximately 0.6 waves III and V are also sensitive for retro-
ms, whereas the mean III-V interval is simi- cochlear disorders, but they have an unac-
lar for both groups. The mean overall I-V ceptably high false-positive rate (25% and
interval is larger by nearly a whole millisec- 37%, respectively) because of the influence
ond for the tumor group. Only rarely does of cochlear hearing loss.
the I-V interpeak interval for patients in the Tumor size influences the sensitivity of tra-
nontumor group exceed 4.54 ms. ditional BAEP latency measurement for pa-
Brain stem auditory evoked potential tients in the tumor and nontumor groups
waveforms for a person with normal hearing (Fig. 19-4). In a study that compared tumor
(A) and for patients with a tumor of CN VIII size, five BAEP indices had a sensitivity of
(B, C) are shown in Figure 19-2. The nor- 100% if the tumor was larger than 2 cm.
mal tracing (A) is well defined and depicts However, if the tumor was 1 cm or smaller,
waves I, III, and V at the appropriate laten- the best sensitivity was 82%.4
cies. The lower tracings show abnormal I-III
and I-V interpeak intervals (B) or the ab-
sence of waves following wave I (C). SUMMARY
Various BAEP latency indices and their
sensitivity and specificity for CN VIII lesions Hearing sensitivity in the 2000-4000 Hz
are shown in Figure 19-3. These results were range is important to BAEP assessment. Ab-
obtained from 75 patients with confirmed solute latencies and interaural latency dif-
CN VIII tumors who were matched audio- ferences are often affected by increasing de-
metrically with 75 patients with cochlear grees of hearing loss in this frequency range,
hearing loss.4 The highest sensitivity for this whereas interpeak intervals are relatively
group of CN VIII tumors was 92% when us- stable measures, even for patients with
ing abnormal wave-V interaural latency dif- moderate-to-severe degrees of peripheral
ference (greater than 0.4 ms) or abnormal hearing loss. However, the reduction in am-
I-V interpeak interval. The specificity with plitude or the absence of a measurable wave
Figure 19-4. Percentage of abnormal (delayed or absent) brain stem auditory evoked potentials (BAEPs) for 75 pa-
tients with CN VIII tumor as a function of tumor size. I, III, V, BAEP waves; I-III, III-V, I-V, interpeak intervals;
ILD, wave V interaural latency difference. (From Bauch et al.4 By permission of the American Speech-Language-
Hearing Association.)
1 associated with peripheral hearing losses velopment of wave amplitude. Hear Res 68:35-41,
often makes it difficult or impossible to mea- 1993.
sure I-III or I-V intervals. Overall sensitivity 3. Bauch CD, Olsen WO. Wave V interaural latency dif-
ferences as a function of asymmetry in 2,000-4,000
of BAEP is 92% for patients with a CN VIII Hz hearing sensitivity. Am J Otol 10:389-392, 1989.
tumor. The false-positive rate for patients 4. Bauch CD, Olsen WO, Pool AF. ABR indices: sensi-
with cochlear hearing loss is 12%. Tumor tivity, specificity, and tumor size. Am J Audiol
size influences BAEP test results: the sensi- 5:97-104, 1996.
5. El-Kashlan HK, Eisenmann D, Kileny PR. Auditory
tivity is 100% for CN VIII tumors larger than brain stem response in small acoustic neuromas. Ear
2 cm, but it is only 82% for CN VIII tumors Hear 21:257-262, 2000.
1 cm or smaller. 6. HaapaniemiJJ, Laurikainen ET, Johansson R, Rhine
T, Varpula M. Audiovestibular findings and location
of an acoustic neuroma. Eur Arch Otorhinolaryngol
257:237-241, 2000.
7. Schmidt RJ, Sataloff RT, Newman J, Spiegel JR, My-
REFERENCES ers DL. The sensitivity of auditory brain stem re-
sponse testing for the diagnosis of acoustic neuro-
mas. Arch Otolaryngol Head Neck Surg 127:19-22,
1. Hall JW. Handbook of Auditory Evoked Responses. 2001.
Allyn and Bacon, Boston, 1992. 8. Bauch CD, Olsen WO. Comparison of ABR ampli-
2. Jiang ZD, Zhang L, Wu YY, Liu XY. Brain stem au- tudes with TIPtrode and mastoid electrodes. Ear
ditory evoked responses from birth to adulthood: de- Hear 11:463-467, 1990.
Chapter 20
VISUAL EVOKED POTENTIALS
John N. Caviness
VISUAL SYSTEM ANATOMY AND USE OF VISUAL EVOKED POTENTIALS IN

PHYSIOLOGY CLINICAL PROBLEM SOLVING
VISUAL EVOKED POTENTIALS IN LOCALIZATION OF VISUAL SYSTEM
NORMAL SUBJECTS LESIONS
METHODS Prechiasmatic and Chiasmatic Lesions
GENERAL PRINCIPLES OF Retrochiasmatic Lesions
INTERPRETATION OF VISUAL EVOKED SUMMARY
POTENTIALS
Visual evoked potentials (VEPs) are highly neuritis. Full-field visual stimulation usually
sensitive to lesions of the optic nerve and an- does not demonstrate abnormality in pa-
terior chiasm but relatively insensitive to tients with unilateral retrochiasmatic lesions.
ophthalmologic disorders. Visual evoked po- The results of VEP studies in cases of bilat-
tentials are noninvasive studies that allow a eral optic nerve lesions may be indistin-
quantitative determination of visual func- guishable from those of bilateral retrochias-
tion. They usually are performed in cooper- matic lesions. Alterations in test methods
ative patients with good visual acuity. The can assist with delineating retrochiasmatic
method has to be altered for assessment in lesions.
unconscious patients. Importantly, these This chapter reviews the method, inter-
electrophysiologic studies are sensitive but pretation, and clinical applicability of tran-
not pathologically specific. sient full-field pattern-reversal VEPs in pa-
Visual evoked potentials generally are per- tients with neurologic disease.
formed by using a shift of a checkerboard
pattern without changing luminance. Mon-
ocular visual stimulation is always preferred. VISUAL SYSTEM ANATOMY
In normal subjects, the visual stimulus AND PHYSIOLOGY
evokes a prominent waveform with positive
polarity in the posterior head region at a The visual system functions at several levels,
mean latency of approximately 100 ms. This beginning with the retina and terminating
potential, the PI or PI00 wave, is generated in several regions of the cerebral cortex.
by striate and peristriate occipital cortex af- Each eye projects to both occipital lobes
ter visual stimulation. The most common through the decussation of the axons from
transient VEP abnormality in patients with the nasal half of each retina. Important
anterior visual pathway lesions is prolonged structures involved in visual conduction in-
latency of the PI wave. clude the macula, optic nerve, optic chiasm,
Monocular Pi-wave alterations in latency optic tract, lateral geniculate body in the
have a higher diagnostic yield than physical thalamus, and thalamocortical pathways.
examination findings in patients with optic The macula at the posterior pole of the
227
retina is specialized for high-acuity central

vision. The primary visual system projects to
striate and peristriate areas of the occipital
cortex (Brodmann areas 17, 18, and 19).
The occipital cortex projects to the midtem-
poral cortex and the posterior parietal cor-
tex. Cells in the visual cortex are most sen-
sitive to movement and to edges. The retina
topographically transmits visual information
to the occipital cortex. The macula projects
to the occipital poles, and more peripheral
regions of the retina project to medial cal-
carine cortex. Different features of a visual
stimulus activate specific neurons in the vi-
sual system. For example, certain neuronal
groups in the retina and lateral geniculate
body are involved primarily in detecting vi-
sual motion or color. Neurons in the visual
cortex also appear to demonstrate these
unique electrophysiologic properties.
VISUAL EVOKED POTENTIALS Figure 20-1. Full-field visual evoked potential after stim-
ulation of the right eye of a 25-year-old woman. PI wave-
IN NORMAL SUBJECTS form is maximal in the posterior head region (Oz elec-
trode); PI amplitude and latency are normal. (From
A normal transient VEP to a pattern-reversal Stockard et al.12 By permission of the American Society
of Electroneurodiagnostic Technologists.)
checkerboard is a positive midoccipital peak
that occurs at a mean latency of 100 ms (Fig.
20-1 ).2 The waveform consists of three sep-
arate phases: an initial negative deflection
(Nl or JV75), a prominent positive deflection tients with decreased visual acuity. A de-
(PI or PI 00), and a later negative deflection crease in check size to 10' in people with
(N2 or N145). The numbers used for the normal acuity is associated with an increase
waveform designation refer to the approxi- in the amplitude and latency of the PI wave.
mate latency (in milliseconds) in the control The fovea is most sensitive to smaller checks
population. The amplitude and latency of and makes the largest contribution to the PI
the Nl and N2 waveforms are too variable amplitude.6 However, smaller check sizes are
in normal subjects to be useful in interpret- inherently sensitive to ophthalmologic dis-
ing VEPs in patients with neurologic disorders, including poor visual acuity, and are
eases. Because sedation and anesthesia abol- not used routinely in performing VEPs. A
ish VEPs, these studies are not useful for normal response to hemifield stimulation
intraoperative neurophysiologic monitor- shows paradoxical localization. This refers to
ing. Flash VEPs can be used in such situa- an ipsilateral distribution of the PI wave to
tions, but the information gathered is more the hemifield being stimulated.
qualitative than quantitative.3'4 Visual acuity, pupillary size, age, and sex
The size of the checks used in the checker- may alter the PI waveform in normal sub-
board pattern may affect the amplitude and jects.3 In the absence of an alteration in lu-
latency of the PI wave.5 The checks are mea- minance, visual acuity must be decreased to
sured by the degree of visual angle (minutes 20/200 for the PI latency to be abnormal
of arc [']). Checks that are between 28' and (Fig. 20-2). PI is not prolonged with visual
31' are associated with stimulation of the acuity of 20/200 if large checks, for exam-
central retina and are usually satisfactory.3 ple, greater than 35', are used. Therefore,
Larger check sizes may be necessary in pa- subtle changes in visual acuity, for example,
Visual Evoked Potentials 229
METHODS
Transient VEP studies generally are per-
formed with a shift of a checkerboard pattern
(black and white). No change in luminance
(total light output) occurs with this form of
stimulation.7 Studies that change luminance,
for example, pattern-flash or strobe light,
produce more variable results in normal sub-
jects and are not as sensitive for detecting ab-
normalities in visual conduction. Flash VEPs
may be appropriate for intraoperative moni-
toring of visual function.4 Monocular testing
is preferred because binocular stimulation
may mask a unilateral visual conduction ab-
normality. The patient is seated a fixed dis-
tance, 70-100 cm, from the screen (usually a
television monitor) and is asked to focus on
the center of the screen. In certain situations,
the technician may need to verify that the pa-
tient is looking at the screen. The patient
should not be sedated before being exam-
ined. Electrodes are placed at Cz, Oz, Al, and
Fz. Other acceptable electrode positions in-
Figure 20-2. Full-field visual evoked potential obtained
with and without a contact lens. Top, PI amplitude and clude midoccipital (5 cm above the inion),
latency are normal. Bottom, With reduction of visual acu- right and left occipital (5 cm lateral to the
ity to 20/400, PI morphology is distorted and latency midoccipital electrode), and midfrontal (12
is prolonged. cm above the nasion).
With full-field stimulation, the PI wave-
form is maximal in the midoccipital region
but may be well recorded between the inion
20/40, do not explain significant prolonga- and the vertex of the head. The checker-
tions of PI latency (Fig. 20-2). Patients who board pattern is reversed (black to white to
have an asymmetry in pupillary diameter black) at a rate of once or twice per second.
may have interocular differences in PI la- An increase of the stimulus rate to 4 per sec-
tency. Therefore, patients should not have ond or greater may prolong the PI latency.
their pupils dilated before undergoing VEP Steady-state VEPs obtained with stimulus
studies. A miotic pupil may reduce lumi- rates of 8-10/second are technically more
nance and prolong the latency and decrease difficult to perform and are not commonly
the amplitude of PI. Age is a significant vari- used to evaluate patients with suspected neu-
able in determining normal PI latency. The rologic disease.3 The low-frequency filter
latency increases with age and may be more may range between 0.2 and 1.0 Hz; the high-
marked in those older than 60 years. Also, frequency filter should be 200-300 Hz. A
women have a shorter PI latency than men. sweep length of 200-250 ms is used, and
This factor has to be considered when de- 100-200 responses are averaged. Increasing
ciding on normative data for PI latency. the number of responses may produce a
Also, if a normal subject chooses not to look more favorable signal-to-noise ratio, but the
directly at the screen used for visual stimu- subject may find it difficult to maintain fix-
lation, the PI waveform may be distorted. Pa- ation for a longer time. At least two trials
tient cooperation may be extremely impor- should be performed before the PI latency
tant in a person with psychiatric disease and is identified. The trials should be repro-
in very young and old people. Patients with ducible. The American Electroencephalo-
oculomotor disorders may not be able to vol- graphic Society guidelines recommend a
untarily fixate on the screen. check size of approximately 30'.8
The procedure for performing a VEP test ered abnormal if the asymmetry cannot be
should be explained to the patient before explained by technical factors.6 However,
the study is initiated. Also, the physician who amplitude abnormalities usually occur with
requested the study should explain to the pa- latency criteria for an abnormal VEP. Cer-
tient the rationale for the examination. Vi- tain lesions in the visual pathway may distort
sual acuity and pupillary size should be de- amplitude more than latency. In reporting
termined in each eye of the patient before VEP studies, the anatomical localization (or
the study is performed. If appropriate, the the lack thereof) of the lesion in the visual
patient should wear his or her eyeglasses or pathway and the lack of specificity must be
contact lenses for the study. Mydriatic drops emphasized.
should not be used before the procedure.
USE OF VISUAL EVOKED

GENERAL PRINCIPLES OF POTENTIALS IN CLINICAL
INTERPRETATION OF VISUAL PROBLEM SOLVING
EVOKED POTENTIALS
Visual evoked potential studies provide an
The interpretation of VEPs in patients with objective physiologic measure that comple-
suspected neurologic disease begins with the ments the results obtained for the clinical
identification of the amplitude and latency history and examination and from neu-
of the PI wave. The results of VEP studies in roimaging. Categories of clinical problems
normal subjects should be available in the to which VEP studies can be applied include
laboratory to determine whether an absolute the following:
PI latency and the interocular difference in 1. Confirmation of a visual system abnor-
latency are abnormal (Table 20-1). Each mality in the presence of current equiv-
evoked potential laboratory preferably ocal visual symptoms and signs.
should have its own normative data. An ac- 2. Confirmation of a visual system abnor-
ceptable alternative is to use published normality in the presence of known or sus-
mal values obtained at a reference labora- pected diffuse or multifocal central
tory. Before VEP studies are performed, nervous system disease.
however, at least 20 normal subjects should 3. Confirmation of a visual system abnor-
be examined with methods similar to those mality when functional recovery has
of the reference laboratory. PI latencies and occurred after a past visual system in-
interocular differences in latencies greater sult. The classic example is finding a PI
than the mean plus three standard devia- latency delay after a patient has recov-
tions are often used to identify abnormal ered from an episode of optic neuritis
studies. Absolute amplitude determinations in the past.
are not particularly useful when interpreting 4. Producing evidence for the nature of
a VEP study. An interocular difference in the pathologic process.9 Demyelinating
amplitude greater than two may be consid- disease (for example, multiple sclero-
sis) usually produces significant PI la-
tency delays, with relative preservation
of amplitude. Compressive or ischemic
Table 20-1. Normative PI Latency lesions often show amplitude loss, with
Values Used at Mayo Clinic relative preservation of latency. Visual
evoked potential changes in degenera-
LATENCY (ms) tive disease are more nonspecific, and
small changes in latency and amplitude
Age, year Females Males are seen.
Less than 60 < 115 < 120 5. Localization of visual system lesions
60 or older < 120 < 125
(this is considered below under Local-
ization of Visual System Lesions).
LOCALIZATION OF VISUAL
SYSTEM LESIONS
Prechiasmatic and Chiasmatic
Lesions
Transient full-field VEPs are highly sensitive
to anterior visual conduction lesions. Uni-
lateral PI abnormalities indicate a visual
conduction defect anterior to the optic chi-
asm (Fig. 20-3). An abnormal interocular
difference in PI latency when both PI val-
ues are normal suggests an optic nerve le-
sion on the side of the increased value. Bi- Figure 20-3. Full-field visual evoked potential in a 42-
lateral increased PI latency values can be year-old man with multiple sclerosis and optic neuritis
found with bilateral optic nerve lesions, a in the left eye. PI latency and amplitude are normal
chiasmatic lesion, or bilateral retrochias- with stimulation of the unaffected right eye. Absolute
matic lesions (Fig. 20-4). However, if the in- PI latency is prolonged and the interocular difference
is abnormal with stimulation of the left eye. Note that
terocular difference is abnormal when both the amplitude in the lower tracing is preserved. This
PI latencies are prolonged, bilateral retro- study suggests an anterior conduction defect on the left.
chiasmatic lesions are less likely. It should be (From Benarroch EE, Westmoreland BF, Daube JR,
emphasized that VEP abnormalities are non- Reagan TJ, Sandok BA. Medical Neurosciences: An
Approach to Anatomy, Pathology, and Physiology by
specific. However, the most common neu- Systems and Levels, 4th ed. Lippincott Williams &
rologic disease associated with a unilateral Wilkins, Philadelphia, 1999, p 592. By permission of
PI abnormality is demyelinating disease. An Mayo Foundation.)
Figure 20-4. PI latencies are prolonged bilaterally, maximal on the left, in patient who subsequently was shown to
have demyelinating disease. (From Stockard et al.12 By permission of the American Society of Electroneurodiag-
nostic Technologists.)
alteration of VEPs may be identified in pa- latency prolongation (Fig. 20-3 and Fig.
tients with retrobulbar neuritis who charac- 20-5). This may be shown by an abnormal-
teristically have no abnormality on physical ity in the absolute PI latency or with a pro-
examination. The sensitivity of VEPs may be longed interocular difference. The ampli-
superior to that of magnetic resonance imag- tude of the PI wave may be normal even
ing (MRI) in patients with optic nerve le- when the latency is markedly prolonged. Vir
sions and demyelinating disease.2 tually all patients with clinically demon-
Visual evoked potentials should be used to strated optic neuritis have unilateral or bi-
complement other neurodiagnostic studies lateral abnormalities in VEPs. Rarely, in
and should be correlated with the clinical acute optic neuritis with severe alteration in
presentation before the diagnosis of de- visual acuity, a PI wave is not recorded.
myelinating disease is made. As mentioned Tumors compressing the optic nerve and
above, the results of electrophysiologic stud- optic chiasm or occurring within the optic
ies remain abnormal even several years after nerve may be associated with a unilateral PI
the optic neuritis has resolved. Visual evoked alteration.4'10 PI latency may be prolonged;
potentials may be useful diagnostically in however, more commonly, the amplitude is
demonstrating a lesion in the optic nerve in decreased disproportionately to the change
patients with suspected multiple sclerosis in latency. The morphology of the VEP may
who have disease localized to the cerebral be markedly distorted, and occasionally the
hemispheres or spinal cord. The incidence PI wave may not be recorded. Neoplasms as-
of abnormalities of VEPs depends on the de- sociated with optic nerve compression in-
gree of clinical confidence that a patient has clude optic nerve gliomas, meningiomas,
multiple sclerosis.10 Patients with clinically craniopharyngiomas, and pituitary tumors.
definite or probable multiple sclerosis are more Giant aneurysms may produce a similar op-
likely to have abnormal VEPs than those with tic nerve lesion. Improvement in PI wave-
possible disease. The most common VEP in pa- forms is variable in patients examined after
tients with optic neuritis is an ipsilateral Pi- surgical extirpation of the tumor.
Figure 20-5. Full-field visual evoked potential in a normal subject (left) and in a patient with an anterior visual con-
duction defect (right). Note that the PI latency is prolonged in the patient, with preservation of PI amplitude.
Other anterior visual pathway lesions that terior visual conduction defects.3 Partial-
may be associated with an abnormality in field studies are not commonly performed
full-field VEPs include anterior ischemic op- and require a modified method. They re-
tic neuropathy, toxic (drug-induced) ambly- quire the additional placement of lateral
opia, glaucoma, and Leber's optic atrophy.7 temporal electrodes (Fig. 20-6). (See Chi-
The results of the electrophysiologic studies appa3 for a more complete discussion of the
must be correlated with the clinical presen- method and interpretation of partial-field
tation to confirm these diagnoses. studies.) The clinical applicability of partial-
field VEPs is uncertain because of develop-
ments in quantitative visual perimetry and
Retrochiasmatic Lesions neuroimaging.
Patients with cortical blindness associated
The recording of full-field VEPs from the with various pathologic processes have been
midoccipital region usually does not show studied with transient VEPs.12 Importantly,
any PI abnormality in patients with unilat- full-field VEPs have been reported to be nor-
eral posterior visual conduction defects. Bi- mal in patients with blindness and with neu-
lateral PI abnormalities are seen in retro- roimaging and pathologic changes confined
chiasmatic lesions, but this VEP result is to the visual cortex.5 The sensitivity of VEPs
nonlocalizing and nonspecific. Magnetic in patients with cortical blindness depends
resonance imaging increasingly has been on the anatomy of the cortical lesion and the
shown to be more useful than full-field tran- method of the study. Lesions involving only
sient VEPs in evaluating patients with Brodmann area 17 (bilaterally) may be as-
retrochiasmatic lesions.11 Full-field VEPs sociated with visual loss and normal VEPs.
may be normal even in patients with abnor- The use of smaller check sizes is important
mal neuroimaging findings retrochiasmati- to identify changes in VEPs. Patients evalu-
cally or visual field defects or both. The diated with normal size checks, for example,
agnostic yield of VEPs is increased with 27', may have normal VEPs, but checks less
partial-field stimulation in patients with pos- than 20' usually reveal an alteration.5 Nor-
Figure 20-6. Partial-field visual evoked potential after stimulation of the right hemifield. A PI waveform maximal
on the right is present in the posterior head region. (From Stockard et al.12 By permission of the American Soci-
ety of Electroneurodiagnostic Technologists.)
mal VEPs obtained with large checks in pa- quire minimal patient cooperation. The
tients with suspected cortical blindness waiting period at most institutions is also
should not be considered evidence for func- considerably shorter for performing VEPs
tional visual loss. A normal PI latency and than for neuroimaging studies.
amplitude in a blind person are highly un-
usual except for those with visual cortex dis-
ease. Normal findings on a VEP study virtu- REFERENCES
ally exclude an optic nerve or anterior
chiasm lesion as the cause of visual loss.6 As 1. Celesia GG, Peachey NS, Brigell M, DeMarco PJJr.
noted above, with small checks, a significant Visual evoked potentials: recent advances. Elec-
percentage of patients with retrochiasmatic troencephalogr Clin Neurophysiol Suppl 46:3-14,
1996.
lesions have changes in VEPs. However, in 2. Aminoff MJ, Goodin DS. Visual evoked potentials.
most patients with cortical blindness, the J Clin Neurophysiol 11:493-499, 1994.
neuroimaging findings indicate the anatomy 3. Chiappa KH (ed). Evoked Potentials in Clinical
and pathology of the lesion. Medicine, 3rd ed. Lippincott-Raven, Philadelphia,
1997.
4. Jones NS. Visual evoked potentials in endoscopic
and anterior skull base surgery: a review. J Laryn-
SUMMARY gol Otol 111:513-516, 1997.
5. Celesia GG. Visual evoked potentials and elec-
An important question that remains is, troretinograms. In Niedermeyer E, Lopes da Silva
F (eds). Electroencephalography: Basic Principles,
"What, if any, is the role of VEPs in evaluat- Clinical Applications, and Related Fields, 3rd ed.
ing patients with neurologic disease in an era Williams & Wilkins, Baltimore, 1993, pp 911-936.
of advanced neuroimaging techniques?" 6. Epstein CM. Visual evoked potentials. In Daly DD,
Magnetic resonance imaging is clearly supe- Pedley TA (eds). Current Practice of Clinical Elec-
rior in sensitivity and specificity to VEPs in troencephalography, 2nd ed. Raven Press, New
York, 1990, pp 593-623.
detecting retrochiasmatic lesions. However, 7. Tandon OP. Average evoked potentials—clinical
in patients with lesions involving the optic applications of short latency responses. Indian J
nerve and anterior chiasm, VEPs have sev- Physiol Pharmacol 42:172-188, 1998.
eral important advantages: (1) VEPs are ob- 8. American Electroencephalographic Society. Guide-
line nine: guidelines on evoked potentials. J Clin
jective and reproducible and may demon- Neurophysiol 11:40-73, 1994.
strate a functional abnormality that is not 9. Nuwer MR. Fundamentals of evoked potentials and
evident on physical examination or with neu- common clinical applications today. Electroen-
roimaging studies, (2) VEP abnormalities cephalogr Clin Neurophysiol 106:142-148, 1998.
may persist over time even when there is clin- 10. Ng YT, North KN. Visual-evoked potentials in the
assessment of optic gliomas. Pediatr Neurol 24:44-
ical resolution of visual symptoms, (3) VEPs 48, 2001.
may be a more reliable indicator of disease 11. Drislane FW. Use of evoked potentials in the diag-
than MRI (MRI may reveal abnormalities nosis and follow-up of multiple sclerosis. Clin Neu-
that do not represent a pathologic process, rosci 2:196-201, 1994.
12. Stockard JJ, Hughes JF, Sharbrough FW. Visual
such as alterations of the white matter of the evoked potentials to electronic pattern reversal: la-
cerebral hemisphere), and (4} VEP studies tency variations with gender, age, and technical fac-
are less expensive than MRI studies and re- tors. Am J EEC Technol 19:171-204, 1979.
SECTION 2
Electrophysiologic
Assessment of
Neural Function
Part C
Motor Pathways
Weakness, fatigue, loss of strength, and loss of power are among the major
symptoms of neurologic disease that can be assessed with neurophysiologic
testing. Strength and movement are under the control of the motor system,
which includes the central mechanisms for integrating motor activity and
the output pathways. Reflexes and other central motor control systems are
discussed in Part E of this section. The electrophysiologic assessment of pe-
ripheral motor pathways is reviewed in this part and Part D. As with the sen-
sory pathways, the most direct assessment of the motor pathways can be ob-
tained with stimulation along the motor pathway and measurement of the
response evoked by the stimulation. These measurements can include the
threshold for activation, the conduction time or velocity (or both) between
the points of stimulation and recording, and the size and shape of the evoked
response.
Compound muscle action potentials recorded directly from a muscle are
measured for each assessment of the motor pathways whether activated cen
trally or peripherally. The method of application, the strength, and the type
of stimulus vary with the site along the motor pathway being stimulated.
Stimulation at the cortical level requires high-intensity electric or magnetic
stimuli to produce useful responses. Deep-lying motor nerves, such as the
spinal nerves, may require needle electrodes for stimulation. Surface elec-
trical stimulation is adequate for stimulation of most peripheral motor
nerves.
The recording of compound muscle action potentials described in Chap-
ter 21 assesses motor nerve function in peripheral neuromuscular disorders.
Repetitive activation of compound muscle action potentials, described in
Chapter 22, assesses the function of the neuromuscular junction. Central
stimulation of motor pathways at the spinal cord or cortical level evokes com-
pound muscle action potentials, called motor evoked potentials, is described in
Chapter 23. The distinction between the terms compound muscle action po-
236 Motor Pathways
tential and motor evoked potential is made on the basis of the site of stimula-
tion. Stimulation of motor nerve fibers anywhere along their course after
they leave the spinal cord produces a response in the muscle called a com-
pound muscle action potential. Stimulation along the motor pathways in the
spinal cord or at the cortical level produces an identical muscle response
called a motor evoked potential.
The use of motor evoked potentials for monitoring central motor func-
tion during surgery has been expanded recently. Compound muscle action
potentials continue to be the mainstay for providing insight into peripheral
neuromuscular disease involving motor fibers.
Chapter 21
COMPOUND MUSCLE ACTION
POTENTIALS
Jasper R. Daube
CLINICAL APPLICATION COMPOUND MUSCLE ACTION

RECORDING COMPOUND MUSCLE POTENTIAL CHANGES IN DISEASE
ACTION POTENTIALS Pathophysiology
Type of Recording Electrode Mechanisms of Conduction in Myelinated
Location of Recording Electrode Fibers
STIMULATION Mechanisms of Slow Conduction in Disease
Type of Stimulating Electrode FINDINGS IN PERIPHERAL NERVE
Position of Stimulating Electrode DISORDERS
COMPOUND MUSCLE ACTION Findings in Focal Lesions
POTENTIAL MEASUREMENTS Liability to Pressure Palsies
Normal Values in Compound Muscle Action Evaluation of Focal Neuropathies
Potential Recordings Findings in Diffuse Peripheral Nerve Damage
F WAVES Peripheral Neuropathies
Methods Focal Neuropathies
PHYSIOLOGIC VARIABLES SUMMARY
A compound muscle action potential (CMAP) is are not altered until the disease is either se-
the action potential recorded from muscle vere or late in its course, when marked at-
when stimulation anywhere along the motor rophy and loss of muscle tissue occur. Com-
pathway is sufficient to activate some or all pound muscle action potential assessment
the muscle fibers in that muscle. The CMAP for disease of the neuromuscular junction is
is the summated activity of the synchro- discussed in Chapter 22 (repetitive stimula-
nously activated muscle fibers in the muscle tion studies). Compound muscle action po-
innervated by the axons and motor units rep- tentials are also recorded with motor evoked
resented in that muscle. Therefore, a CMAP potentials to assess central motor pathways
provides a physiologic assessment of (7) the (Chapter 23). The major application of
descending motor axons in the pathway be- CMAP recording is in motor nerve conduc-
low the level of stimulation, (2) the neuro- tion studies.
muscular junction, and (3) the muscle fibers Motor nerve conduction studies and
activated by the stimulus. Because disease of CMAP recordings are equivalent. The rest of
the axons, neuromuscular junctions, or mus- this chapter focuses on several aspects of
cle fibers can alter the CMAP, CMAP record- CMAP recording as part of motor nerve con-
ing can be used to assess disease at each of duction studies and their application. The
these locations. Compound muscle action chapter begins with a review of the tech-
potential recordings are least useful for as- niques of stimulation and recording, includ-
sessing muscle disease because the potentials ing technical problems. The next section dis-
237
238 Motor Pathways
cusses modifications of the techniques of traced. This is particularly useful for Martin-
stimulation and recording to obtain F-wave Gruber anastomosis (median to ulnar) in
latencies and is followed by a general dis- the forearm, Riech-Cannieux anastomosis
cussion of the approach to selecting motor (ulnar to median) in the hand, the deep ac-
nerve conduction studies and CMAP record- cessory branch of the superficial peroneal
ing for different clinical entities. nerve in the leg, and crossed innervation af-
ter reinnervation.
A less common application of CMAP
CLINICAL APPLICATION recording is to identify and to measure tran-
sient loss of function in primary muscle dis-
Recording of CMAPs in motor nerve con- ease such as periodic paralysis. The record-
duction studies is used for several purposes ings also can be used to study abnormal
in assessing neuromuscular disease. Com- reflex responses in upper motor neuron le-
pound muscle action potentials are particu- sions. In selected patients who have primary
larly useful in providing objective measure- muscle disease, a study of the mechanical
ments of the extent and type of weakness. If twitch and its relationship to electric events
the weakness is caused by a peripheral neu- may be useful as part of a CMAP recording.
romuscular disease, motor nerve conduction
studies can identify and localize the sites of
damage, whether from compression, isch-
emia, or other focal lesion. These studies can RECORDING COMPOUND
also characterize the type of abnormality as MUSCLE ACTION POTENTIALS
a conduction block with neurapraxia or as
slowing of conduction at a localized area. Type of Recording Electrode
They can identify the changes associated
with wallerian degeneration and regenera- The electrodes that record CMAPs can al-
tion in the motor nerve. Measurement of ter the size, shape, and, to a lesser extent,
CMAPs can assist in distinguishing periph- the latency of the response. Large surface
eral nerve disease from lower motor neuron electrodes, small surface electrodes, subcu-
disease, neuromuscular junction disease, taneous electrodes, and intramuscular elec-
and myopathies. Nerve conduction studies trodes each have advantages and disadvan-
can also assist when the weakness may be tages. Most commonly, a 5-10-mm surface
caused by hysteria, malingering, or upper electrode is used because of the ease of ap-
motor neuron disease. In these situations, plication and the availability of well-defined
the CMAP is normal. normal values. Normal values for CMAP re-
Compound muscle action potential recording depend in part on the type of elec-
cordings can go beyond confirming the pres- trode used for recording; therefore, these
ence of disease and the definition of sever- values are most reliable when the electrodes
ity by identifying disease that may not be are identical. Larger electrodes provide a
apparent clinically. For example, in patients better depiction of CMAPs obtained from
with clinical evidence of a mononeuropathy, large muscles or multiple muscles with a
CMAP recording may show signs of multiple common innervation; they also have better
mononeuropathies or widespread periph- reproducibility.2 Generally, the use of these
eral nerve damage that may not be apparent electrodes is limited to laboratories in which
clinically. In patients with inherited neu- normal values have been developed specifi-
ropathies, motor conduction studies can cally for them. Also, in some laboratories,
identify the process early in the disease or large electrodes are used to measure the
when there is mild involvement and no clin- number of motor units in a muscle (Chap-
ical evidence of neuropathy. Motor conduc- ter 27).
tion studies also can identify disease early in Subcutaneous needle electrodes have the
its evolution, for example, diabetes mellitus, advantage of being placed closer to muscle
when a mild peripheral neuropathy may not tissue; therefore, they sometimes record
yet be apparent clinically.1 In patients with higher amplitude CMAPs. For some muscles,
an atypical distribution of deficits, the pres- these electrodes are easier to apply. The dis-
ence of anomalous innervation can be advantages are those of any invasive tech-
Compound Muscle Action Potentials 239
Figure 21-1. Summation of compound

muscle action potentials (CMAPs) re-
corded from thenar muscles with stim-
ulation of the median and ulnar nerves.
Rows 1 and 3 show the CMAPs obtained
with isolated stimulation of each of the
two nerves. Note the initial positivity
with ulnar stimulation that results from
recording CMAPs from ulnar-inner-
vated muscles at a distance from the
thenar recording electrode. Row 2
shows the effect of simultaneous stimu-
lation of both median and ulnar nerves,
with summation of the potentials re-
corded in rows 1 and 3.
nique, including greater discomfort for the the muscle, the potential is a well-defined,
patient and the (extremely low) risk of in- large negative waveform. When recording
fection. The subcutaneous needle-recording the CMAP with electrodes that are either off
electrode occasionally has higher imped- the muscle or at some distance from the
ance, resulting in greater noise, shock arti- muscle that generates the CMAP, the po-
fact, or both. Intramuscular needle or wire tential is predominantly or initially positive
electrodes have the advantage of recording in polarity and much smaller, with a signifi-
from well-defined small areas of muscle, cantly slower rise time to the negative peak,
thereby better isolating the CMAP of indi- as shown in Figure 21-1. The presence of an
vidual muscles. Intramuscular recordings initial positivity on a CMAP is evidence that
are able to record small potentials, particu- the active electrode is not over the end plate
larly of deep muscles that may not be record- region of the muscle generating the CMAP
able on the skin surface. However, the con- and may be entirely off the muscle. The
figuration of the potential varies markedly slope or rate of rise of the positive-to-nega-
with the precise location of the recording tive peak of the CMAP is a rough gauge of
electrode, which may shift during the move- the distance between the active recording
ment produced by the stimulation. Thus, electrode and the muscle generating the
amplitude and area measurements with in- CMAP.
tramuscular recordings are not sufficiently The size and configuration of the CMAP
reliable to be useful clinically. Latencies may vary with placement of the active recording
be difficult to measure because of irregular electrode over a muscle, as shown in Figure
initiation of the CMAP. 21-2.3 The active electrode is optimally
placed over the end plate region, where
there is an initial negativity with a sharp in-
Location of Recording Electrode flection and maximal amplitude. If multiple
muscles are activated or the end plate region
Compound muscle action potential record- is not well localized, as in some disease, the
ing is made with an active electrode and a CMAPs are more complex. Large electrodes
reference electrode whose locations are crit- will average the differences between elec-
ical for the size, shape, and latency of the trode locations to reduce the variation that
CMAP. Normal values must be determined can occur with different placements of small
with specific recording electrode locations. electrodes.
The amplitude and area of the CMAP de- Location of the reference electrode,
crease with the distance of the active elec- sometimes referred to as the inactive, G2, or
trode from the muscle.2 A CMAP can be terminal-two electrode, also has an effect on the
recorded with the active electrode far from amplitude and configuration of the CMAP
the muscle, but it is maximal directly over and, thus, must be the same as the location
the muscle generating it. When recording a used to obtain the normal values. Maximal
CMAP with the active electrode directly over amplitude generally is obtained with the ref-
240 Motor Pathways
Figure 21-2; Amplitude and con-

figuration changes in ulnar (up-
per) and median (lower) com-
pound muscle action potentials
(CMAPs) in two normal subjects.
Compound muscle action poten-
tials were recorded from small
electrodes in multiple locations
in 8 X 9 grids over the thenar
and hypothenar muscles to show
the variation in size and shape
with electrode location. Note the
double peaks, marked changes in
potential over short distances,
and the differences between sub-
jects. These variations make
CMAPs highly susceptible to
small differences in electrode
placement, especially with small
recording electrodes. Note also
the difference in pattern of dis-
tribution between normal sub-
jects. (From van Dijk et al.3 By
permission of John Wiley &
Sons.)
erence electrode over the tendon of the nerve and in parallel with it evoke the most
muscle being recorded, optimally at the reproducible responses with the lowest stim-
junction of the tendon with the muscle. Be- ulus intensity. Different types of stimulating
cause muscles vary in size, the reference elec- electrodes have different advantages and dis-
trode should not be at a fixed distance from advantages, which must be understood to se-
the active electrode. Compound muscle ac- lect the optimal electrode for each motor
tion potentials recorded with fixed inter- nerve tested. The commonly used handheld
electrode distances often have a shorter du- surface stimulator allows the electrodes to be
ration, lower amplitude, and smaller area, moved easily in search of the nerve. Smooth,
and, occasionally, a different configuration; rounded, 5-10-cm electrodes mounted on
they should not be used.4 the end of curved, removable stimulating
poles permit rapid change of the anode and
cathode positions. This stimulator is most
convenient for stimulating nerves that may
STIMULATION require pressing on the overlying skin so the
electrode is closer to the nerve and for ro-
Compound muscle action potential record- tating the position of the anode to reduce
ing requires stimulating a nerve somewhere shock artifact. This type of electrode is opti-
along its length. Stimuli can be applied in mal for standard motor nerve conduction
several ways. The stimulation technique used studies. When stimuli have to be applied for
to activate a nerve affects the values ob- longer periods, as in testing periodic paral-
tained. This must be kept in mind when se- ysis and measuring motor unit number es-
lecting a stimulation technique.5 timates, flat disk electrodes taped on the
skin over the nerve or a pair of electrodes
mounted in a bar with the electrode pro-
Type of Stimulating Electrode truding from the bar allows more stable po-
sitioning of the electrode.
Stimulation usually is electric, applied Needle electrodes, 1-2 cm, that are en-
through a cathode (negative) and an anode tirely uninsulated or a longer needle that is
(positive) that may vary in size and shape. insulated except for 1-2 mm at its tip can
Surface electrodes that are placed over the also be used to stimulate motor fibers. These
electrodes are particularly useful for stimu- less than 2 mA are adequate for activating
lating deep nerves such as the median and all the motor axons. An anode at some dis-
ulnar nerves in the forearm and the tibial tance from the nerve, either on the surface
and sciatic nerves in the leg. The disadvan- or elsewhere in the tissue, may be used with
tage of needle electrodes is that they are the needle cathode near the nerve. A distant
more difficult to move when attempting to anode can result in a somewhat higher
find the optimal location for nerve stimula- threshold for activation, a greater risk of cur-
tion. rent spreading to the surrounding nerves,
Magnetic stimulation can activate some and a less accurate site of stimulation. These
but not all peripheral nerves and is seldom disadvantages are generally outweighed by
used for neuromuscular electrodiagnosis. the advantage of not having to probe the tis-
Because the site of stimulation cannot be de- sue with the anode to find the optimal lo-
fined as precisely with magnetic stimulation cation near the nerve. The invasive nature
as with electric stimulation, the measure- of needle stimulation and the time it takes
ment of distances and conduction velocity is to achieve optimal location of the stimulat-
less reliable. The advantage of magnetic ing electrode have made it less accepted
stimulation is that it is sometimes easier to than surface stimulation.
obtain and record potentials from deep The ideal position of surface stimulating
nerves with less discomfort to the patient. electrodes is along the length of the nerve,
with the cathode closest to the recording
electrode. The anode and cathode must be
Position of Stimulating Electrode farther apart than for needle electrode stim-
ulation. If the anode and cathode are too
Depolarization of motor axons occurs at the close, current flow passes directly between
cathode. The anode hyperpolarizes the them without entering the tissue to the
nerve and may block conduction of an ac- depth of the nerve. Thus, activation of all
tion potential through the area of hyperpo- motor axons may not occur despite the use
larization. Activation of a motor axon re- of high voltage and the passage of a large
quires areas of both depolarization and current. For most motor nerves, a distance
hyperpolarization along the length of the of 3-5 cm between the anode and cathode
axon, with current flow through the axon be- is sufficient for adequate current to pene-
tween the two locations.5 Therefore, the op- trate the tissues to the depth of the motor
timal position of stimulating electrodes is for axons. This increases the likelihood of stim-
the cathode to be as close as possible to the ulating surrounding nerves, which must be
nerve between the anode and the recording kept in mind as a potential technical prob-
site, so that the activated action potential lem. For nerves that are very deep in the tis-
does not traverse the area of hyperpolariza- sue, a greater distance between the anode
tion at the anode. The optimal location of and cathode may be necessary, increasing
the anode is longitudinally along the course the risk of inadvertent stimulation of other
of the axon away from the recording elec- nerves and muscles. The anode may also be
trode. Ideally, the anode and cathode are ad- placed perpendicularly to the course of the
jacent to the nerve and only a few millime- nerve and laterally from it. The anode may
ters apart, so that all current flow is directed need to be on the opposite side of the limb,
through the nerve being tested and not for example, to activate the tibial nerve in
into surrounding muscle, another nerve, or an obese patient. A perpendicular location
other tissue. requires a higher current intensity to obtain
Needle electrodes can be placed immedi- depolarization, increasing the possibility that
ately adjacent to the nerve, but this may re- adjacent nerves will be stimulated. A lateral
quire considerable probing in the tissue. position is used for the anode only when this
The optimal location of a needle electrode position is necessary for other purposes.
can be obtained by repeated stimulation to The most common need for the lateral po-
identify the region of minimum threshold. sition is when the stimulating and recording
When the anode and cathode are both im- electrodes are placed so close that a promi-
mediately adjacent to the nerve, stimuli of nent shock artifact occurs in the recording.
242 Motor Pathways
The shock artifact occurs because the cur- by total current flow, which is a function of
rent flow from the stimulating electrode the intensity of the applied voltage, the re-
spreads through the tissue directly to the sistance to current flow, and the duration of
recording electrode and charges the capac- the stimulus. Pulses of 0.1-0.2 ms are usually
itance of the intervening tissue, which then adequate for stimulation of motor nerves,
discharges over 2-20 ms, with a waveform su- but longer durations of up to 1 ms may be
perimposed on the CMAP. This occurs es- necessary for deep or diseased nerves.
pecially with stimulation of the tibial nerve
at the ankle, the sural nerve at the ankle, and
the facial nerve at the angle of the mandible.
In these situations, it may become necessary COMPOUND MUSCLE ACTION
to locate the anode perpendicularly to the POTENTIAL MEASUREMENTS
nerve as the anode is rotated to find a posi-
tion of minimal shock artifact. The CMAP recorded over any muscle in re-
The location of most nerves can be iden- sponse to the stimulation of the muscle's
tified reasonably well from anatomical land- peripheral nerve is sometimes called the
marks for each nerve. However, it must al- M wave (Mfor motor). It is characterized by
ways be remembered that the exact location several specific measurements. The most
of a nerve can vary significantly among nor- valuable measurement is the size of the
mal subjects. The most striking example is CMAP, measured either as the amplitude or
the peroneal nerve at the ankle; its position area of the CMAP. Both of these variables
can vary from 0.5 cm to 4 cm lateral to the reflect the total number of muscle fibers that
tibia. Therefore, when attempting to stimu- contribute to the potential. In most labora-
late a motor nerve, the nerve must be local- tories, amplitude is measured from the base-
ized to minimize stimulus intensity for less- line to the peak, although the measurement
ened patient discomfort and to decrease the can be made from the positive peak to the
likelihood of current spread to other nerves. negative peak. The duration of the response
Placing the stimulating electrodes at the lo- of the CMAP is a function of the synchrony
cation judged to be over the nerve and then of firing of the muscle fibers contributing to
obtaining an initial low-amplitude CMAP the potential. A loss of synchrony results in
best accomplishes this. The stimulating elec- longer duration and lower amplitude. Thus,
trode is then moved medially or laterally per- the area of the CMAP is related most directly
pendicularly to the nerve without changing to the number of muscle fibers or motor
the stimulus intensity. If the subsequent units that contribute to the CMAP.
CMAPs have increasing amplitude, the elec- The latency of the CMAP from the time
trode is being moved closer to the nerve. of stimulation is best measured to the onset
However, if the amplitude decreases, the of the initial negativity. The latency defines
electrode is being moved away from the the time it takes the action potential to travel
nerve. The electrode continues to be moved from the stimulation site to the recording
until the maximal amplitude is obtained site and depends mainly on the conduction
with the original stimulus intensity. The time in the peripheral axons. A small
voltage is then increased until the CMAP amount of time is needed to traverse the
does not increase further with a 25%-30% neuromuscular junction. If the electrodes
increase in applied voltage or current. are not over the end plates, latency also in-
In normal subjects, these techniques allow cludes the time for conduction along the
supramaximal or full amplitude CMAPs to muscle fiber to the recording electrode. In
be obtained with stimulus intensities less this case, the CMAP initially is positive rather
than 20 mA (100 V) in the arm and less than than negative, with the elapsed time to reach
40 mA (200 V) in the leg. In obese subjects the end plate being the latency of the initial
or in cases of particularly deep nerves and positive deflection (Figs. 21-1 and 21-2).
in patients with peripheral nerve disease, a Initial positive deflections may also be
greater intensity of current may be needed caused by the recording of a CMAP of a dis-
to activate motor nerves. The intensity of a tant muscle, for example, a contribution
stimulus applied to a motor nerve is defined from the anterior compartment muscles
with stimulation of the peroneal nerve at the Dividing the difference in CMAP latencies
knee when recording from the extensor dig- by the distance between the two points mea-
itorum brevis. This initial positivity should sures the conduction velocity of the nerve
not be measured. fibers (Fig. 21-3). Because the latency mea-
Distal latency is the onset of a CMAP at surements are made to the initial negativity,
the most distal site of stimulation and is best the conduction velocity measurement is that
measured as an absolute value. Attempts of the fastest conducting fibers. Paired stim-
have been made to correct for slowing in the ulation techniques, in which the action po-
nerve terminal and at the neuromuscular tentials in the fast conducting fibers are
junction, a measurement called residual la- obliterated by collision, have been used to
tency. This method has been reported to be measure conduction velocity in slower con-
of value in diagnosing early carpal tunnel ducting axons. However, the additional clin-
syndrome. Residual latency is calculated with ical data provided by paired stimulation are
the formula: not sufficiently useful clinically to make it a
standard procedure.6
RL = DML - (Distal Distance/Conduction Velocity) Compound muscle action potential size is
also measured as part of repetitive stimula-
where RL is residual latency and DML is dis- tion in response to exercise or drugs used
tal motor latency. in disorders such as periodic paralysis and
Latency measurements should be made myasthenia gravis (see Chapter 22).
from the point at which the negative portion Several potential sources of error must be
of the action potential is initiated, as defined kept in mind in measuring CMAPs during
by the nerve action potentials seen with stim- nerve conduction studies. The most com-
ulation of the distal site. The reproducibil- mon one is incorrect measurement of the
ity of latency measurements can be en- distance between the two points of stimula-
hanced by automated measurement at a tion, which may be caused by (1) distortion
fixed voltage above baseline (200 /jiV/cm is of the skin when applying the stimulating
often recommended) .5 electrodes or when making the measure-
The difference in CMAP latency with stim- ment, (2) nonstandard position of the body
ulation at two points along a nerve is a func- during the measurement, such as having the
tion of the distance between the two points elbow extended rather than flexed during
and the rate of conduction of the action po- ulnar nerve conduction studies, (3} erro-
tentials in that nerve between the two points. neous polarity of the stimulating electrode,
Figure 21-3. Calculation of conduction velocity from latency and distance measurements on standard nerve con-
duction studies. (From DaubeJR. Nerve conduction studies. In AminofFMJ [ed]. Electrodiagnosis in Clinical Neu
rology, 3rd ed. Churchill Livingstone, New York, 1992, p 289. By permission of WB Saunders Company.)
244 Motor Pathways
and (4) simultaneous stimulation of adjacent of stimulation. When a CMAP is recorded at

nerves (Fig. 21-1). Sources of error in la- two sites of stimulation, it should be very sim-
tency measurements include (/) failure to ilar at both sites unless disease or anomalous
note the sweep speed correctly, (2) a poorly innervation is present. If the two responses
defined shock artifact that interferes with are not similar, technical or physiologic er-
the take-off of the CMAP, (3) incorrect elec- rors must be excluded before the difference
trode location, resulting in an initial posi- is attributed to localized disease. Technical
tivity or a poorly defined onset of the nega- errors can be caused by submaximal stimu-
tive CMAP (Fig. 21-4), and (4) failure to lation at one location or excessive stimula-
select the same point on the inflection of the tion with activation of an adjacent nerve in
CMAP for the measurement at two points another location (Fig. 21-1). Also, excessive
Figure 21-4. The size and configuration of compound muscle action potentials (CMAPs) evoked by ulnar nerve
stimulation vary with location of the hypothenar recording electrodes. Top, Location of the active recording elec-
trode with the reference electrode on the fifth digit. Bottom, The corresponding CMAPs. (From Carpendale MTF.
Conduction Time in the Terminal Portion of the Motor Fibers of the Ulnar, Median and Peroneal Nerves in Healthy
Subjects and in Patients With Neuropathy. Thesis, Mayo Graduate School of Medicine [University of Minnesota],
Rochester, 1956. By permission of Mayo Foundation.)
stimulation at one site may shorten the la-

tency because of current spread along the
nerve.
Normal Values in Compound

Muscle Action Potential
Recordings
For clinical reports, values in CMAP record-
ings shpuld include the actual value mea-
Figure 21-5. Compound muscle action potential
sured and the appropriate associated normal recorded from abductor hallucis muscle with tibial
values. Ideally, normal values are those col- nerve stimulation at, A, low and, B, high amplification.
lected in the same laboratory using the same F waves are depicted much better for measurement at
techniques, including careful attention to higher amplification. (From Daube JR. Nerve conduc-
tion studies. In Aminoff MJ [ed]. Electrodiagnosis in
sources of error.7 This is not always possible, Clinical Neurology, 3rd ed. Churchill Livingstone, New
and values collected in large studies can York, 1992, p 293. By permission of WB Saunders
serve if ;the techniques are similar to those Company.)
of the user.8'9 First, normal values are cor-
rected as needed for the physiologic vari-
ables described below. The resulting normal which motor units are activated varies from
values have been presented in various ways, stimulus to stimulus. Therefore, F waves are
but often without a full understanding of much lower in amplitude than the directly
their complexity and possible errors result- evoked CMAP. F-wave latencies vary with
ing from their skewed distribution.10'11 Be- each stimulus, because axons with different
cause no single value can truly identify a con- conduction velocities are activated from
duction value as normal or abnormal, it is stimulus to stimulus (Fig. 21-5). As the site
best to use normal deviates or percentile val- of stimulation is moved proximally on a
ues. Normal deviates define the value's extent limb, F-wave latency decreases (because the
of deviation from normal, and percentiles de- distance the action potential travels de-
fine the proportion of a normal population creases) and the M-wave latency increases
that has this value.12 until the two potentials merge, usually with
stimulation at the elbow or just proximal to
it (Fig. 21-6). The F-wave latency varies with
F waves are small CMAPs recorded from the

muscle fibers of a single motor unit or a
small number of motor units activated by
antidromic-action potentials that travel cen-
trally along motor axons to anterior horn
cells. Consequently, the latency of an F wave
includes the time required for the action po-
tential to travel antidromically from the site
of stimulation to the spinal cord and the
time to travel orthodromically from the
spinal cord to the muscle. Because F waves
travel over long segments of nerve, they are Figure 21-6. Compound muscle action potential
among the most sensitive measures of dif- (CMAP) recorded from hypothenar muscles with ulnar
fuse nerve disease.13 nerve stimulation at, A, wrist and, B, elbow. The CMAP
latency increases, whereas the F-wave latency decreases
In most muscles, only a small proportion with more proximal stimulation. With proximal stimu-
of the motor units is activated antidromically lation, F waves summate with the late component of the
by any one supramaximal stimulus, and CMAP.
246 Motor Pathways
Rgure 21-7. Variation of F-wave latency with distance in normal subjects for, A, arm conduction studies and, B, leg
conduction studies.
the distance from the spinal cord to the site the values plotted as a histogram that gives
of stimulation, with the distance to the mus- the dispersion (chrono-dispersiori) of the F la-
cle, and with the conduction velocity of the tencies, but this is time-consuming and adds
motor fibers14 (Fig. 21-7). little additional value clinically17 (Fig. 21-8).
Different laboratories use different distance
measurements. Normal values must be re-
corded using the same techniques. In the
Methods Mayo electromyographic (EMG) laboratory
we have found that arm measurements made
Stimulation applied to the median, ulnar, from the site of stimulation at the wrist (cath-
tibial, or peroneal nerve at the wrist or anode) to the sternoclavicular joint and leg
kle evokes an F wave that is separated clearly measurements from the cathode to the
from the M wave. The cathode should be xiphoid process are most useful.
proximal to the anode, and the stimulus In measuring F-wave latencies, it is partic-
should be supramaximal to ensure anti- ularly important to pay attention to poten-
dromic activation of all the axons.15 A series tial errors.18 A poorly relaxed muscle may
of stimuli is applied until a minimum of produce deflections throughout the sweep,
10 F waves has been obtained.16 Too few F making it difficult to identify F waves (Fig.
waves will result in an inadequate sample for 21-9). Late components or satellite poten-
reliable measurement of the variables. In tials of a dispersed compound action po-
some nerves, particularly the peroneal, F tential may be identified incorrectly as F
waves may be too infrequent for an ade- waves. Satellite potentials can be recognized
quate number to be obtained for reliable by their constant location and configuration,
measurements. in contrast to the variable F waves. Also, the
Recording electrodes for F waves are latency of satellite potentials increases with
placed over the muscle in the standard lo- more proximal stimulation, whereas F-wave
cations used for motor nerve conduction latencies decrease.
studies so that F-wave recordings can be Other late responses, A waves, can resem-
made immediately after standard nerve con- ble F waves, but they are more persistent at
duction studies have been completed. a single stimulus intensity. The two types of
Higher amplification is needed than for A waves are axon reflexes and indirect dis-
standard nerve conduction studies; gains of charges.19 Both decrease in latency with
200 or 500 mV/cm are usually adequate. more proximal stimulation (Fig. 21-10,
The longer latencies of F waves require small arrow). Indirect discharges are the iden-
slower sweep speeds than needed for stan- tical backfiring activation at a proximal lo-
dard nerve conduction studies. The latency cation on an axon that can be blocked by
is measured to the earliest reproducible po- paired stimuli, as can F waves (Fig. 21-11).
tential in the series recorded. The latency of Axon reflexes are potentials that invade a prox-
each of the F waves can be measured and imal branch of an axon and can become
Figure 21-8. Calculated values for ulnar F-wave latency based on recordings from 96 normal subjects. All values are
derived from the F-wave latencies shown in A, by the calculations shown with each histogram. B, Central latency es-
timates the time from elbow stimulation to return of the F-wave response to the elbow location. C, Conduction ve-
locity is the velocity of the F waves over the length of the nerve from the wrist to the spinal cord. D, Estimated latency
for the F wave is based on peripheral conduction and the distance from the wrist to the sternal notch. E, Latency
difference compares estimated latency with measured F-wave latency. Proximal slowing alone results in positive dif-
ferences; distal slowing alone results in negative differences.
more or less frequent with a change in stim-

ulus intensity. Thus, the criteria for identi-
fying F waves are responses that are variable
in latency, amplitude, and configuration but
occur grouped within a consistent range of
latencies. The latency increases with more
distal sites of stimulation and has a well-
defined onset. F-wave latency measurements
are made to the initial deflection from the
baseline (positive or negative) of 10 repro-
ducible responses of shortest latency. F-wave
latencies obtained in normal subjects 18-88
years old are listed in Table 21-1. The F-wave
latency should not be more than 4 ms longer
than the estimated F-wave latency (see dis-
cussion below).
Several methods have been suggested for
assessing F waves, including comparing the figure 21-9. F-wave recordings made, A, with the mus-
latency with normal values corrected for age cle at rest and, JB, with muscle contraction. Reliable mea-
and distance, calculating the conduction ve- surements are not possible with poor relaxation.
248 Motor Pathways
can be calculated on the basis of the distance

and conduction velocity in the distal seg-
ment with the formula (Fig. 21-8.D):
Fest = (2 X Distance/Conduction Velocity)
+ Distal Latency
F-wave latencies should be within the nor-
mal range of F-wave estimates. If they are
shorter, proximal conduction is faster than
distal conduction. If they are longer, proxi-
mal conduction is slower than distal con-
duction. Thus, F waves can distinguish dif-
Figure 21-10. Superimposed abductor hallucis com-
pound muscle action potentials recorded in response fuse peripheral nerve disorders from those
to tibial nerve stimulation. Late responses are variable that are primarily distal and those that are
latency and amplitude F waves (thick arrow) and stable primarily proximal.
axon reflexes (thin arrow). Both have shorter latency
with more proximal stimulation. (From Daube JR.
Nerve conduction studies. In Aminoff MJ [ed]. Elec-
trodiagnosis in Clinical Neurology, 3rd ed. Churchill
Livingstone, New York, 1992, p 294. By permission of
WB Saunders Company.)
locity in the central segment, and calculat-

ing a central latency and comparing it with
an estimated latency based on known con-
duction velocity (Fig. 21-8). The most con-
venient and readily applied method is to
compare F-wave latency with normal values
corrected for distance. F waves can be
elicited by stimulating any nerve, but they
are more prominent in some nerves, for ex-
ample, in the tibial nerve when recording
from foot muscles (hence, F for foot).
The rate of stimulation does not affect the
F waves, but minimal muscle contraction
may enhance them. However, such contrac-
tion can make it more difficult to recognize
F waves. Contraction and relaxation of mus-
cles in another limb or in the jaw may en-
hance F waves without obscuring them. F
waves should be recorded with only supra-
maximal stimulation; otherwise, they may be Figure 21-11. Diagram showing the effects of an-
confused with H reflexes. F waves are seen tidromic action potentials on two forms of A wave (late
best with stimulation at distal sites. The de- responses). Top, A single axon with proximal hyperex-
crease in latency with more proximal stimu- citability is reactivated by an antidromic action poten-
tial to give a late response; the reactivated late response
lation is an important test to ensure that the is blocked by a paired stimulus. Bottom, Theoretical out-
responses are late responses. come if the late responses were activated ephaptically.
Because F-wave latency varies with dis- None of the patients with these late responses showed
tance, it depends on limb length. Measure- the pattern of ephaptic activation, e, point of ephaptic
activation; O, point of action potential backfiring; S',
ments of limb length should be made as de- single stimulus; S", second of a pair of stimuli. (From
scribed for each nerve whenever F waves are Magistris and Roth.19 By permission of Elsevier Scien-
recorded. An estimated F-wave latency (Fest) tific Publishers.)
Table 21-1. F-Wavc Latency in Normal Subjects 18-88 Years Old

Contralateral
Mean (ms) Range (ms) Distance (cm) Difference
Ulnar/ADM 26.6 21-32 50-76 0-3
Median/APB 26.4 22-31 57-73 0-3
Tibial/AHB 48.6 41-57 106-125 0-4
Peroneal/EDB 47.4 38-57 102-128 0-4
ADM, abductor digiti minimi; AHB, abductor hallucis brevis; APB, abductor pollicis bre-
vis; EDB, extensor digitorum brevis.
PHYSIOLOGIC VARIABLES tween 20 and 30 years of age, and by age 80,

it is approximately 10 m/second slower.21
In normal subjects, CMAPs vary with several Conduction velocities are slower in children
factors, which need to be controlled. The younger than 3 years and in people older
temperature of the limb is the most signifi- than 65 years. Compound muscle action po-
cant factor; temperature decrease produces tential recordings show no significant dif-
a 2.0 m/second slowing per degree centi- ferences between men and women. Height
grade and increases both amplitude and and body size, for example, finger circum-
area.20 Temperature is a greater cause of ference for median sensory values, are also
variation in measurements of conduction ve- contributing factors to normal values. Con-
locity than errors in measurement of latency duction velocity slows with axonal length.
or distance. Between 22°C and 38°G, con- The effect is particularly noteworthy in per-
duction velocity is related approximately lin- sons taller than 6 feet in whom normal val-
early to the temperature, increasing about ues are significantly slower than in shorter
50% when the temperature is increased subjects.22 Ideally, the normal value for a pa-
10°C (Qio = 1.5). Thus, a nerve with a con- tient should be adjusted for temperature,
duction velocity of 60 m/second at 36°C age, and height.
conducts at 40 m/second at 26°C (that is, a The range of normal values is wide, mak-
decrease of 2 m/second per degree centi- ing the measurement of a single value less
grade). The change per degree centigrade reliable in identifying mild disease. For ex-
is proportionally less for nerves that have a ample, the range of normal peroneal/ex-
lower conduction velocity. Calculations can tensor digitorum brevis CMAP amplitudes
be made to correct for a cool limb, but it is from 2 to 12 mV means that a patient who
more reliable and effective to warm a cool has a baseline amplitude of 10 mV may lose
limb. Immersing the limb in a water bath at 80% of the response before the value is out-
40°C for 5 minutes is best. The temperature side the normal range. The range of normal
of the arm measured on the surface over the values for conduction velocity, latency, and
hand should be at least 31 °C; the tempera- F-wave latency is narrower and, thus, some-
ture of the leg measured on the surface over what better for identifying mild changes in
the lateral malleolus should be at least 29°C. disease. However, in each case, percentile or
More distal sites have lower temperatures. normal deviation measurements are better
Compound muscle action potential mea- for detecting mild disease.12 Combinations
surements in patients should always be per- of variables may improve the recognition
formed in the same temperature range in scores.23
which the normal values were determined. There are also significant differences in
Age must also be considered in deter- the normal values for amplitudes and rates
mining the significance of prolonged laten- of conduction between different nerves, par-
cies, slow conduction velocities, and low ticularly between the upper and lower ex-
amplitudes of compound action potentials. tremities. In unilateral disorders, compar-
Conduction velocity slows progressively be- isons of values obtained in the affected limb
250 Motor Pathways
with those obtained in the opposite limb that excites the subsequent node. Conduc-
can be helpful, but there may be large side- tion velocity is determined by the time re-
to-side differences in normals (amplitude, quired for one node to excite the next. Thus,
20%-70%; latency, 30%-40%; conduction if the distance between two nodes (intern-
velocity, 20%-30%; F-wave latency, 10%).24 odal distance) is 1 mm and the nodal con-
Therefore, the significance of any value is duction time is 20 microseconds, the con-
best evaluated by comparing it with values duction velocity is 50 m/second. The time
obtained in the same nerve in a limb at the required for one node to excite the next
same temperature of subjects of the same node is determined by several factors:
age who participated in the study in which 1. The faster the rate of rise of the action
the normal values were determined, using potential at node 1, the more rapidly
the same methods and making a percentile node 2 will be activated.
comparison. 2. The smaller the amount of current re-
The reproducibility of CMAP recordings quired to neutralize the charge held by
also must be considered both in identifying the membrane capacitance of node 2
abnormality and in comparing values over and to depolarize the nodal membrane
time when a patient's condition is being to threshold, the more rapidly an ac-
monitored. These range from 5% for F-wave tion potential will appear at node 2.
latencies in the arm to 15% for CMAP am- 3. The more current that is lost in neu-
plitudes in the foot.5 Reliable motor nerve tralizing the charge across the axonal
conduction studies require vigilance in rec- membrane in the internode and by
ognizing the many pitfalls possible.25 leakage through the myelin, the longer
it will take to activate the next node.
4. The higher the resistance to current
COMPOUND MUSCLE ACTION flow in the axoplasm from node 1 to
POTENTIAL CHANGES node 2, the longer it will take to acti-
vate node 2.
IN DISEASE
Pathophysiology
Mechanisms of Slow Conduction
The techniques routinely used to study in Disease
nerve conduction test large diameter affer-
ent fibers and alpha motor fibers. The Paranodal demyelination increases the ca-
compound action potential recorded from pacitance of the internodal membrane.
peripheral nerves is the action potential that More current is needed to neutralize the
results from activation of large myelinated charge across the internodal membrane and
fibers. Even the nerve action potential from less is available to discharge node 2. Thus, it
a mixed nerve is predominantly from large takes longer to initiate an action potential at
afferent fibers. Components resulting from node 2. Segmental demyelination results in
activation of small myelinated (delta) fibers a more profound increase in capacitance
and C fibers cannot be identified. Special and decrease of resistance across the inter-
techniques of measuring distribution of con- nodal membrane. In large diameter fibers,
duction in the activated axons have not been conduction may be blocked. In smaller di-
generally accepted.26"28 ameter fibers, conduction may become con-
tinuous, as in unmyelinated fibers, instead
of saltatory.29
Mechanisms of Conduction in A decrease in the diameter of fibers oc-
Myelinated Fibers curs with axonal atrophy or compression. It
has been observed in a compressed zone and
Conduction in myelinated fibers is saltatory. for 1—2 cm proximal to it. Decreased diam-
The action potential jumps from one node eter increases the resistance to flow of cur-
of Ranvier to the next, with the action po- rent from node 1 to node 2. Reduction in
tential of one node providing the current current flow increases the time required to
excite node 2. Simultaneous reduction in induced or absent responses are the result of
ternodal membrane capacitance because of total conduction block, wallerian degenera-
reduced membrane area does not compen- tion (after axonal disruption), or axonal de-
sate for the higher resistance. generation, as in dying-back neuropathies.
With the loss of large-diameter fibers be- Large-diameter myelinated fibers are the
cause of conduction block or degeneration, nerve fibers most sensitive to damage by lo-
conduction measurements reflect conduc- calized pressure. The largest ones are the af-
tion in smaller diameter, more slowly con- ferent fibers that mediate touch-pressure, vi-
ducting fibers instead of the slowing of con- bration, and proprioception. In a mixed
duction in larger fibers. This may account nerve, these fibers generally have larger di-
for the slow conduction observed in seg- ameters than alpha motor fibers, as evi-
ments proximal to a focal lesion instead of denced by their 10%-15% faster conduction
reflecting an extension of the lesion proxi- velocity. In a chronic compression lesion,
mal to the site of compression. measurement of conduction velocity in the
With decreased myelin thickness, particu- sensory fibers often demonstrates an abnor-
larly during remyelination, the number of mality before it is evident in motor fibers.
myelin lamellae is small in proportion to Conduction block is identifiable most
fiber diameter. The capacitance and con- clearly in individual axons at a site where the
ductance of the internodal membrane are action potential cannot be transmitted to the
high, the loss of current through the in- next segment. No response occurs with stim-
ternode is more than normal, and a longer ulation proximal to the block, and a full re-
time is required to excite the next node. sponse is seen with stimulation distal to the
Other possible factors in the slowing of con- block. Thus, conduction block in a whole
duction are altered characteristics of the nerve may be total, in which no axons trans-
nodal membrane, which affect the genera- mit potentials across the site of damage, or
tion of the action potential. No such factor partial, in which only a proportion transmit
has been identified in focal lesions. potentials across the block. In conduction
Other effects of demyelination should also block associated with a localized mononeu-
be kept in mind. Demyelination increases ropathy, the CMAP area (or amplitude) ob-
the refractory period, decreases the ability of tained with stimulating just proximal to the
the fiber to conduct impulses at high fre- site of the block is decreased compared with
quency, and increases the susceptibility to that just distal to the block. Conduction block
blocking of conduction with increasing generally means there are intact axons that
temperature. are unable to transmit potentials across a lo-
cal area of damage. However, an acute in-
jury to a nerve that destroys all axons will
have the appearance of a conduction block
FINDINGS IN PERIPHERAL for a few days until wallerian degeneration
NERVE DISORDERS occurs.30 Because many other factors may re-
sult in changes that have the appearance of
The only electrophysiologic findings in pe- conduction block, explicit criteria are re-
ripheral nerve disorders are conduction quired for identifying conduction block.31'32
slowing, conduction block, and reduced Slowing in some of the axons with disper-
CMAPs or their absence. Each may have a sion of the CMAP decreases the amplitude,
focal or a diffuse distribution. Conduction but it increases the duration and area.
slowing may be seen as prolonged distal la- Therefore, amplitude is less reliable in rec-
tencies, slow conduction velocity, or pro- ognizing conduction block. Because of nor-
longed F-wave latencies. Segmental demyeli- mal dispersion over longer segments of
nation and the narrowing of axons both slow nerve, stimulation over short segments is
conduction. Conduction block can result from more reliable for identifying conduction
a metabolic alteration in the axonal mem- block.33 Dispersion of a CMAP can also be
brane, such as local anesthetic block, or associated with decreased area because of
structural changes in the myelin, such as phase cancellation, that is, the summation of
telescoping or segmental demyelination. Re- positive and negative components of action
252 Motor Pathways
Table 21-2. Duration of Deficit after Peripheral Nerve Injury

Injury Duration of Deficit
Conduction block (amplitude change)

Metabolic Seconds to minutes
Myelin loss Days to weeks
Axonal distortion Weeks to months
Axonal disruption (fibrillation potentials)

Few axons No deficit
Many axons Weeks to months
All axons Months to years
From Daube JR. Nerve conduction studies. In Aminoff MJ (ed). Electrodiagnosis in
Clinical Neurology, 3rd ed. Churchill Livingstone, New York, 1992, p 308. By permission
of WB Saunders Company.
potentials from different axons.6 Thus, con- may persist for weeks or months.35 Some
duction block is more difficult to identify patterns of abnormality are summarized in
with low-amplitude potentials. Conduction Table 21-3. Compound muscle action po-
block increases with temperature in a dam- tential measurements cannot predict how
aged nerve, providing another important long these abnormalities will last.
reason for monitoring temperature. 4
In contrast, conduction slowing is seen as
prolonged latency. Although conduction Findings in Focal Lesions
block and slowing may occur together, they
often occur independently. Conduction A low-amplitude CMAP, slow conduction, or
block is more common in rapidly develop- conduction block characterizes localized pe-
ing disorders, and conduction slowing is ripheral nerve damage. The amplitude of
more common in chronic disorders. Loss of the CMAP may be low at all sites of stimula-
strength is quantified most accurately by re- tion if wallerian degeneration has occurred.
duction in CMAPs to stimulation at a prox- In conduction block, some axons are unable
imal site. Compound muscle action poten- to transmit action potentials through the
tial amplitude and area changes can help damaged segment but are functioning distal
in categorizing nerve damage into broad to it. If all axons are blocked, no response
groups. For instance, in traumatic injuries of is obtained proximal to the site of the lesion.
a nerve, there is usually conduction block A block in conduction must be distinguished
(with an area or amplitude change) or ax- from slowing in conduction, which may re-
onal disruption (with low amplitude at all semble it. In conduction block, there is an
stimulation points) or some combination of abrupt decrease in CMAP amplitude and
the two. The clinical deficit caused by either area over a short segment. Slowing of nerve
conduction block or axonal disruption may conduction in some axons is associated with
have variable duration (Table 21-2). The re- a gradual decrease in amplitude as stimula-
sults of nerve conduction studies are a function is moved proximally, because of dis-
tion of the underlying pathologic change, persion of the CMAP. The area of the evoked
not the duration of the disorder. No single response remains constant with conduction
change in nerve conduction studies is typi- slowing unless phase cancellation also occurs.
cal of the clinical phenomenon of neur- The CMAP findings in focal nerve dam-
apraxia, in which there is transient weakness age evolve over time with the restoration of
without atrophy. If neurapraxia is caused by function. Local membrane changes, para-
a metabolic alteration, it lasts only a few min- nodal demyelination, segmental demyeli-
utes; however, if it is caused by axonal dis- nation, and axonal telescoping are all fol-
tortion with telescoping of internodes, it lowed by local repair within days to weeks
Table 21-3. Patterns of Abnormality in Nerve Conduction Studies of Peripheral Neuromuscular Disorders
MOTOR NERVE STUDIES SENSORY NERVE STUDIES
ACTION POTENTIAL ACTION POTENTIAL

Conduction F-Wave Conduction
Disorder Amplitude Duration Velocity Latency Amplitude Duration Velocity
Axonal neuropathy 4 Normal >70% Mild t 44 Normal > 70%

Demyelinating neuropathy 4 Proximal t Proximal <50% T 4 t Proximal <50%
Mononeuropathy 1 t 4 T 44 t 4
Regenerated nerve 4 t 4 t 4 4 4
Motor neuron disease 44 Normal >70% Mild t Normal Normal Normal
Neuromuscular transmission defect (4) Normal Normal Normal Normal Normal Normal
Myopathy (4) Normal Normal Normal Normal Normal Normal
Symbols: |, increase; 4 , decrease; 4 4 > greater decrease; ( 4 ) > occasional decrease.
From DaubeJR. Nerve conduction studies. In Aminoff MJ (ed). Electrodiagnosis in Clinical Neurology, 3rd ed. Churchill Livingstone, New York, 1992, p 309. By permission
254 Motor Pathways
Figure 21-12. Left, Ulnar and, right, median compound muscle action potential amplitude and conduction velocity
changes over 100 days after localized compression neuropathy. The upper arm conduction block is seen as the
marked difference in response amplitude with elbow and upper arm stimulation. Note the normal conduction v
locities. (Courtesy of Dr. E. H. Lambert, Mayo Clinic.)
(Fig. 21-12). Wallerian degeneration of than collateral sprouting. The evolution of

some axons in a nerve is followed by collat- the nerve conduction changes is oudined in
eral sprouting of surrounding axons, result- Table 21-4.
ing in reinnervation of the muscle and the
restoration of CMAPs within a few weeks, de-
pending on the amount of axonal loss. Ax- Liability to Pressure Palsies
onal sprouting within the nerve and the rein-
nervation of the muscle after the loss of all In patients with hereditary neuropathy with
the axons are much slower and less complete liability to pressure palsies, even slight trac-
Table 21-4. Changes in Nerve Conduction Variables after Focal Nerve Injury*
CHANGES
Variable Acute (< 7 days) Subacute (weeks) Progressive Residual*
Compound muscle Normal Low if severe Low if severe Low if severe
action potential
Motor conduction Normal < 30% slow if < 30% slow if (< 30% slow if
velocity severe severe severe)
Motor distal Normal < 30% long if < 30% long if (< 30% long if
latency severe severe severe)
F wave Prolonged or Absent or Absent or (Prolonged or
absent prolonged prolonged absent)
*Parentheses indicate changes that occur sometimes, but not always, at that stage.
tion or compression of a nerve may cause stimulation than between other sites. The
motor and sensory disturbances. Further- anatomical location of this point is measured
more, the nerves of clinically unaffected rel- from a fixed landmark (for example, the
atives also may have EMG and histologic ab- medial epicondyle). In this way, a conduc-
normalities. These patients typically have tion block can be localized precisely along
evidence of an underlying sensory and dis- the nerve. Stimulation with near-nerve nee-
tal motor neuropathy.37 An increased inci- dle electrodes can be used for inching in
dence of pressure palsies has been observed nerves deep in the tissue, for example, the
among patients with diabetes mellitus.38 median nerve in the forearm.
Slowing of conduction in Guillain-Barre
syndrome (acute inflammatory polyradicu-
loneuropathy) is often most marked at sites Findings in Diffuse Peripheral
commonly affected by pressure lesions (for Nerve Damage
example, the median nerve at the wrist, the
ulnar nerve at the elbow, and the peroneal Nerve conduction studies can differentiate a
nerve at the knee). Other conditions, in- disorder of primarily axonal loss from a dis-
cluding renal failure, alcoholism, and mal- order of primarily demyelinating nerves. A
nutrition, have been reported to increase disorder associated primarily widi axonal
susceptibility to focal compression lesions. destruction, as in axonal dystrophies and
dying-back neuropathies, is associated pre-
dominantly with low-amplitude CMAPs at all
Evaluation of Focal Neuropathies sites of stimulation, with no more than about
30% slowing in conduction velocity. Seg-
If a nerve is conducting slowly, it is impor- mental demyelination is associated with pro-
tant to identify whether the abnormality is nounced slowing of conduction, usually to
localized or diffuse. Latencies and ampli- less than 50% of normal, and with a pro-
tudes obtained with stimulation (or record- gressive decrease in the CMAP amplitude at
ing) over short distances provide the best lo- proximal stimulation sites (Fig. 21-13).
calization of focal nerve damage.39 If there Slowing of conduction also occurs in se-
is slowing of conduction velocity over any vere, chronic axonal disorders with axonal
length of nerve (for example, the median narrowing.
and ulnar nerves in the forearm or the tib-
ial and peroneal nerves in the leg), the sever-
ity of slowing must be compared with that of Peripheral Neuropathies
other nerves in the patient. If the slowing is
out of proportion to the slowing elsewhere The severity of peripheral nerve damage can
or the decrease in amplitude is more than be well defined by nerve conduction studies,
the normal for that nerve, a localized ab- but the pathologic alteration cannot be pre-
normality must be sought. Stimulating prox- dicted for mixed patterns or mild changes.
imally and distally to the suspected area of Diabetes mellitus, the most common cause
local abnormality (for example, knee or el- of peripheral neuropathy, has a mixed pat-
bow) can identify localized lesions. If con- tern of abnormalities. In this condition,
duction block or slowing is found between there commonly is a mild generalized distal
two points of stimulation, die method of neuropathy caused by multiple small addi-
inching should also be used. Inching begins tive lesions along the nerves, often with
with supramaximal stimulation in the nor- features of a demyelinating neuropathy.41
mal segment just distal to the area of ab- There also may be mononeuropathies of the
normality.40 The point of stimulation is median nerve at the wrist, the ulnar nerve at
noted, and stimulation is reapplied at 2-cm the elbow, or the peroneal nerve at the knee,
intervals proximally along the nerve. The re- with localized slowing or conduction block
sponses are superimposed and compared superimposed on a generalized decrease in
along the length of the nerve. A localized the CMAP amplitude and mild generalized
area of abnormality is indicated by a greater slowing of conduction. F-wave latencies and
decrease in amplitude or a greater increase sural sensory nerve action potential ampli-
in latency between two adjacent points of tudes are the most sensitive motor and sen-
256 Motor Pathways
Figure 21-13. Hypothenar compound muscle action potential recorded with ulnar nerve stimulation in a patient
with generalized peripheral neuropathy. Velocity is slow throughout, with gradual dispersion of the compound mus-
cle action potential to produce lower amplitudes with proximal stimulation. J,, stimulation. (Courtesy of Dr. E. H.
Lambert, Mayo Clinic.)
sory nerve conduction variables, respectively, prognosis; others with low-amplitude CMAPs
for identifying neuropathy.13'42 Needle ex- and prominent fibrillation potentials have
amination usually shows only mild changes severe axonal destruction and a poor prog-
distally. However, some patients with dia- nosis for rapid recovery.
betes mellitus have a lumbosacral polyradicu- Although many patients with neuropathy
lopathy, with diffuse fibrillation potentials in have mixed findings on nerve conduction
lumbar paraspinal muscles. This pattern may studies, some patients may have a predomi-
be associated with prolongation of F-wave nantly axonal or segmental demyelinating
latencies caused by proximal slowing of neuropathy (Table 21-5).
conduction.
Guillain-Barre syndrome, predominantly AXONAL NEUROPATHIES
a demyelinating disorder, has a spectrum of
electrophysiologic changes.43 Nerve con- Axonal neuropathies primarily affect the
duction studies may show no abnormalities axon and produce either diffuse degenera-
or the abnormalities may be limited to prox- tion or dying-back of the distal portion of
imal slowing with prolongation of the F-wave the axon. Axonal damage is particularly
latency or H-reflex latency. Distal recording common with toxic and metabolic disorders.
with stimulation at proximal sites, such as a The major change on nerve conduction
spinal nerve or brachial plexus, also may studies is a decrease in the amplitude of the
show abnormalities.44 More commonly, how CMAP or the compound nerve action po-
ever, Guillain-Barre syndrome is associated tential (or both). This decrease is propor-
with prolonged distal latencies of a mild-to- tional to the severity of the disease. Some ax-
moderate degree, dispersion of CMAPs with onal neuropathies, such as those associated
proximal stimulation, and symmetrical or with vitamin Big deficiency, carcinoma, and
asymmetrical slowing of conduction veloci- Friedreich's ataxia, chiefly affect sensory
ties. A-waves are particularly prominent in fibers; others, such as lead neuropathies,
Guillain-Barre syndrome and are often the have a greater effect on motor fibers. Sen-
earliest sign of the disease on conduction sory axons commonly are involved earlier
studies.43'45 Facial nerves or other cranial and more severely than motor axons. Occa-
nerves may be involved, with abnormalities sionally, sensory potentials can be low am-
seen on blink reflex testing or facial nerve plitude and associated with only mild sen-
stimulation. Patients with mild nerve con- sory symptoms. In contrast to the change in
duction abnormalities and only mild amplitude, there usually is little change in
changes on needle examination have a good latency or conduction velocity in axonal
Table 21-5. Patterns of neuropathies; conduction in individual ax-

Electromyographic Abnormality in ons generally is normal until the axon has
Conditions Associated with Peripheral degenerated. Therefore, normal conduction
Neuropathy velocities should not be considered evidence
against the presence of neuropathy. Often,
Predominant EMG/NCS Changes of Axonal the only finding in a case of axonal neu-
Degeneration ropathy is fibrillation potentials on needle
Diabetes mellitus (some patients) examination of distal muscles, especially in-
Guillain-Barre syndrome (some patients)
trinsic foot muscles.
If many large axons are lost because of ax-
Toxic—vincristine, acrylamide, others
onal neuropathy, conduction velocity may
Alcohol be decreased but not to less than 70% of nor-
Uremia mal. Axonal neuropathies typically affect the
Acute intermittent porphyria longer axons earlier and are first identified
Collagen-vascular diseases in the lower extremities. Nerves that are
Carcinoma more susceptible to local trauma because of
Amyloidosis their superficial location are also more sen-
sitive to axonal damage; typically, axonal
Predominant EMG/NCS Changes of Segmental neuropathies are manifested first as per-
Demyelination oneal neuropathies with low-amplitude or
Diabetes mellitus (some patients) absent responses, while other motor nerves
Guillain-Barre syndrome (some patients) remain intact. Axonal neuropathies may be
Dejerine-Sottas disease associated with a change in the refractory pe-
Diphtheritic neuropathy riod of the nerve and with a relative resis-
Chronic inflammatory neuropathy tance to ischemia. Amyotrophic lateral scle-
Refsum disease rosis, a disease of anterior horn cells, shows
Leukodystrophies changes on motor conduction studies typi-
Neuropathy with monoclonal protein cal of axonal neuropathy46 (Table 21-6) and
is notable for large, repeater^ waves resulting
EMG, electromyographic; NCS, nerve conduction from collateral sprouting of axons.47
studies.
From Daube JR. Nerve conduction studies. In
Aminoff MJ (ed). Electrodiagnosis in Clinical Neurol- SEGMENTAL DEMYELINATING
ogy, 3rd ed. Churchill Livingstone, New York, 1992, NEUROPATHIES
p 312. By permission of WB Saunders Company.
Segmental demyelinating neuropathies are usu-
ally subacute inflammatory disorders, such
as Guillain-Barre syndrome, chronic inflam-
matory demyelinating polyradiculoneuropa-
Table 21-6. Changes in Nerve Conduction Variables in Amyotrophic Lateral Sclerosis

CHANGES
Variable Subacute/Active Chronic Active Inactive or Residual
Compound muscle action Normal unless Low if severe Low if severe

potential severe
Motor conduction Normal < 30% slow if < 30% slow if
velocity/F wave severe severe
Motor distal latency Normal < 30% long if < 30% long if severe
severe
Repetitive stimulation Decrement in some Decrement in some Normal
258 Motor Pathways
thy (CIDP), and diphtheritic neuropathies. asymmetrical neuropathies of the peroneal,

However, similar patterns may be seen in the ulnar, or median nerves at the knee, elbow,
inherited chronic hypertrophic neuropa- or wrist, respectively. The refractory period
thies, for example, Dejerine-Sottas disease in demyelinating neuropathies is decreased,
and hereditary motor sensory neuropathy. often to the extent that repetitive stimula-
Demyelinating neuropathies typically are as- tion at rates as low as 5 Hz causes a decre-
sociated with prolonged latencies and a pro- ment. The decrement usually does not ap-
nounced slowing of conduction, often in the pear until rates of 10 or 20 Hz are used.
range of 10—20 m/second.48 Commonly, the Maximal voluntary contraction may increase
amplitude is relatively preserved on distal the conduction block and weakness because
stimulation but is decreased proximally be- of the axonal membrane changes in CIDP.50
cause of dispersion of the CMAP on proxi- Electrodiagnostic criteria proposed for the
mal stimulation. In some hereditary disor- identification of CIDP are not as sensitive as
ders, such as Dejerine-Sottas disease, the histologic criteria in identifying patients
velocity may be only a few meters per sec- who may respond to immune suppressive
ond. The electrophysiologic findings also therapy.51'52
vary among the inherited neuropathies, as One form of demyelinating neuropathy is
shown in Table 21—7.49 Acquired demyeli- multifocal motor conduction block, which may
nating neuropathies commonly affect sites superficially resemble amyotrophic lateral
of nerve compression early and produce sclerosis. Although it may have other fea-
tures of a generalized demyelinating neu-
ropathy, the classic finding is that of con-
Table 21-7. Electrophysiologic duction block, especially in the median
Findings of Inherited Demyelinating nerve in the forearm. The conduction block
can increase with activity53 and is hyperex-
Neuropathies citable with fasciculation potentials. The
Inherited Disorders with Uniform Conduction conduction block may persist for years.54
Slowing At times, the pattern of abnormality in de-
Charcot-Marie-Tooth 1A myelinating neuropathies helps differentiate
an acquired process from a hereditary one.55
Charcot-Marie-Tooth IB
An acquired demyelinating neuropathy has
Dejerine-Sottas disease
scattered areas of slowing, with some areas
Metachromatic leukodystrophy being much more abnormal than others;
Cockayne's disease hereditary disorders generally have a sym-
Krabbe's disease metrical pattern. Acquired demyelinating
disorders often show more dispersion with
Inherited Disorders with Multifocal Conduction
Slowing proximal stimulation than hereditary disor-
ders do. This distinction is not always reli-
Hereditary neuropathy with liability to
pressure palsies able because some patients who have a
hereditary demyelinating neuropathy with a
Charcot-Marie-Tooth X
low-amplitude CMAP may have pronounced
Adrenomyeloneuropathy
dispersion at proximal sites of stimulation.
Pelizaeus-Merzbacher disease
Refsum's disease
Inherited Disorders with Incompletely Characterized Focal Neuropathies
Electrophysiology
PM22 point mutations On nerve conduction studies, changes in
mononeuropathies vary with the rapidity of
PO point mutations
development, the duration of damage, and
Adult-onset leukodystrophies the severity of damage as well as with the un-
Merosin deficiency derlying disorder.56 With a chronic com-
EGR 2 mutations pressive lesion, localized narrowing or para-
From Lewis et al.49 By permission of John Wiley & nodal or internodal demyelination produces
Sons. localized slowing of conduction. Narrowing
Table 21-8. Compound Action Potential Amplitude after

Peripheral Nerve Injury*
AMPLITUDE
Injury 0-5 Days After 5 Days Recovery
Conduction block
Proximal stimulation Low Low Increases
Distal stimulation Normal Normal Normal
Axonal disruption
Distal stimulation Normal Low Increases
*Supramaximal stimulation.
of axons distal to a chronic compression evoked response can be used as an approx-

slows conduction along the entire length of imation of the number of intact, viable ax-
the nerve. Telescoping of axons with intus- ons (Table 21-8).
susception of one internode into another The evolution of electrophysiologic changes
distorts and obliterates the nodes of Ranvier after peripheral nerve injury is also seen on
and thus blocks conduction. Moderate seg- needle examination and is an aid in charac-
mental demyelination and local metabolic terizing mononeuropathies (Table 21-9).
alterations also are often associated with con- Therefore, an adequate assessment of a pe-
duction block. The segment of nerve with ripheral nerve injury should include both
disruption of the axons distal to an acute le- needle examination and nerve conduction
sion may continue to function normally for studies. The significance of changes with
as long as 5 days; then, as the axons undergo time after injury is outlined in Table 21-10.
wallerian degeneration, conduction ceases The sequence of changes shows that nerve
and the amplitude of the evoked response conduction studies can be important in as-
diminishes and finally disappears. One week sessing localized nerve injury within the first
after an acute injury, the amplitude of the few days after injury.
Table 21-9. Findings on Needle Examination after Peripheral Nerve Injury

0-15 Days After 15 Days Recovery
Conduction block
Fibrillation potentials None None None
Motor unit potentials | Recruitment 4 Recruitment | Recruitm
Axonal disruption
Fibrillation potentials None Present Reduced
Motor unit potentials 4 Recruitment 4 Recruitment Nascent
I , decrease.
From Daube JR. Nerve conduction studies. In Aminoff MJ (ed). Electrodiagnosis in Clinical Neurology, 3rd ed.
Churchill Livingstone, New York, 1992, p 314. By permission of WB Saunders Company.
260 Motor Pathways
Table 21-10. Electromyographic Interpretations after Peripheral

Nerve Injury
Finding Interpretation
0-5 days
Motor unit potentials present Nerve intact, functioning axons
Fibrillations present Old lesion
Low compound action potential Old lesion
5-15 days
Compound action potential distal only Conduction block
Low compound action potential Amount of axonal disruption
Motor unit potentials present Nerve intact
After 15 days
Compound action potential distal only Conduction block
Fibrillation potentials Amount of axonal disruption
Distribution of damage
Recovery
Increasing compound action potential Block clearing
Increasing number of motor unit Block clearing
potentials
Decreasing number of fibrillation Reinnervation
potentials
"Nascent" motor unit potentials Reinnervation
MEDIAN NEUROPATHIES values but is more reliable when compared

with the latency in ulnar sensory fibers over
The most common focal mononeuropathy the same distance. Improved sensitivity,
is carpal tunnel syndrome, in which the specificity, and reliability have been re-
median nerve is compressed in the space ported with a combined sensory index (CSI).
formed by the wrist bones and the carpal The CSI adds the difference between nor-
ligament. Many different approaches have mal and recorded values for median and ra-
been suggested for the electrodiagnosis of dial to the thumb, median and ulnar to the
this condition.57"59 An automated device for ring finger, and median and ulnar pal-
rapid electrodiagnosis is being evaluated.60 mar.23'61 Moderate nerve compression de-
Early or mild compression of the median creases the amplitude of the sensory nerve
nerve in the carpal tunnel may not show any action potential and prolongs the latency to
electrophysiologic abnormalities, especially a greater extent. Severe median neuropathy
of CMAPs. However, more than 90% of at the wrist also increases the distal motor la-
symptomatic patients have localized slowing tency to the thenar muscles and decreases
of conduction in sensory fibers. The sensory the thenar CMAP (Fig. 21-14). A decrease
latency through the carpal tunnel is the most in the CMAP is often associated with mild
sensitive single measurement for identifying slowing of motor conduction velocity in the
the earliest abnormality. This so-called pal- forearm and fibrillation potentials in the
mar latency may be compared with normal thenar muscles. Carpal tunnel syndrome of
Figure 21-14. Right and left median nerve conduction studies in carpal tunnel syndrome. Distal latency is pro-
longed on the right. (From Department of Neurology, Mayo Clinic and Mayo Foundation for Medical Education
and Research: Clinical Examinations in Neurology, 6th ed. Mosby-Year Book, St Louis, 1991, p 434. By permission
of Mayo Foundation.)
moderate severity is often associated with sponse may be evoked inadvertently by high-
anomalous innervation of the thenar mus- voltage stimulation of the median nerve and
cles, with the amplitude of the response be- recorded as an apparent median sensory po-
ing higher on elbow stimulation than on tential. Occasionally, patients have sensory
wrist stimulation.62 branches that innervate one or more fingers,
Many patients with carpal tunnel syn- which are anatomically separated from the
drome have bilateral abnormalities on nerve motor fibers and relatively spared. Also, the
conduction studies, even though the symp- severity of compression may not be the same
toms may be unilateral. Therefore, the con- for all the fascicles of the median nerve,
duction in the opposite extremity should be which would result in greater slowing in the
measured if a median neuropathy at the axons to some digital nerves than to others.
wrist is identified. Several other considera- This variation in involvement is the likely
tions must be kept in mind when testing for reason for the added value of the CSI. A me-
carpal tunnel syndrome. A few patients have dian neuropathy may be an early finding in
a normal sensory response and a prolonged patients with more diffuse neuropathies. To
distal motor latency. Chronic neurogenic at- exclude this possibility, it is necessary to as-
rophy from a proximal lesion, such as dam- sess other nerves.
age to a spinal nerve or anterior horn cells, Median neuropathies in the forearm are
can result in distal motor slowing and a nor- much less common and only rarely show
mal sensory response. A radial sensory re- abnormality on nerve conduction studies,
262 Motor Pathways
other than slightly low-amplitude sensory or the slowing in the peripheral segments. Ul-
motor responses (or both).63 Anterior in- nar neuropathies are commonly bilateral; if
terosseous neuropathy and pronator syn- an ulnar neuropathy is evident on one side,
drome are usually manifested electrophysio- the opposite extremity should also be tested.
logically by fibrillation potentials in the
appropriate muscles. Infrequently, patients PERONEAL NEUROPATHIES
have localized slowing of conduction in the
damaged segment of nerve. Neuropathy of the peroneal nerve at the
head of the fibula is another common focal
ULNAR NEUROPATHIES lesion. Peroneal neuropathy of recent onset
caused by compression typically is associated
Findings in ulnar neuropathy vary with the with a conduction block that can be local-
severity and location of the lesion.64 In most ized precisely using the inching procedure
patients, the abnormality is at the elbow (Fig. to identify the area where the evoked re-
21—15). Various methods have been sug- sponse decreases (Fig. 21-16). Conduction
gested for electrodiagnostic evaluation of ul- across this segment usually is not slowed, al-
nar neuropathy.59 As in carpal tunnel syn- though in lesions of longer standing the
drome, sensory fibers are more likely to be slowing becomes prominent.69
damaged than motor fibers, so that the sen- Nerve conduction studies of the superfi-
sory nerve action potential is commonly lost cial peroneal nerve may be of value in dif-
early. In some patients, focal slowing can be ferentiating a peroneal neuropathy from an
demonstrated in ulnar sensory fibers across L5 radiculopathy with some motor slowing.
the elbow. Thus, direct measurement of the If the superficial peroneal sensory nerve is
orthodromic compound nerve action po- normal, it is more likely that the damage is
tential may be an efficient and accurate at the root level rather than the peripheral
method for recognizing mild ulnar neu- level.70 Patients with a moderately severe
ropathy.64 Motor involvement often occurs peroneal neuropathy often do not have an
later and may involve different fascicles se- evoked response from the extensor digito-
lectively; thus, motor recordings from the rum brevis. Recordings from the anterior
first dorsal interosseous as well as the hy- tibial muscle with stimulation at the head of
pothenar muscles may increase the sensitiv- the fibula and the knee may still demon-
ity.65 Although there may be slowing of strate a block or slowing of conduction in
conduction to the flexor carpi ulnaris, this the nerve. Anomalous innervation of the ex-
muscle usually shows little or no change on tensor digitorum brevis muscle by a deep ac-
nerve conduction studies and needle exam- cessory branch of the superficial peroneal
ination. Measurements are more accurate nerve may make it more difficult to recog-
with the arm extended laterally and the el- nize a peroneal neuropathy. In apparent
bow flexed to 45°, because of better access peroneal neuropathies without localized
to and measurement of the ulnar nerve at slowing of conduction, the short head of the
the elbow.66 The most common localizing biceps femoris muscle should be tested for
finding in ulnar neuropathy of recent onset fibrillation potentials to exclude a sciatic
is conduction block at the elbow that can be nerve lesion. Sciatic nerve lesions may pre-
localized precisely by using the inching sent with only peroneal deficit and may re-
procedure67 (Fig. 21-15). This conduction quire deep needle electrode stimulation for
block may be associated with local slowing. sciatic nerve conduction studies.
Chronic ulnar neuropathy usually results in
slowing of conduction. If there has been sig- OTHER MONONEUROPATHIES
nificant axonal loss with low-amplitude mo-
tor responses, the associated axonal atrophy Most other neuropathies are traumatic in
results in slowing in the forearm. Occasion- origin and may be localized by motor con-
ally, a lesion proximal to the elbow requires duction studies.71 Neuropathies of the ra-
that stimulation be applied in the upper arm dial, tibial, or phrenic nerves may similarly
as well.68 In ulnar and median neuropathies, be localized by nerve conduction studies but
F-wave latency is prolonged proportional to technically are more difficult.72'73 Evalua-
Figure 21-15. A, Compound muscle action potential

recordings with right ulnar neuropathy at the elbow. B,
The inching procedure across the elbow demonstrates
localized slowing and conduction block at the elbow.
(Courtesy of Dr. E. H. Lambert, Mayo Clinic.)
tion of most other nerves is not aided by slowing that occurs with a long-standing
nerve conduction studies because these neu- proximal lesion.74'75 In facial neuropathies
ropathies do not show localized slowing. Re- like Bell's palsy, stimulation cannot be ap-
ports of localized slowing of such nerves gen- plied proximal to the site of the lesion. The
erally have not taken into account the distal usual findings in Bell's palsy with neur-
264 Motor Pathways
Rgure 21-16. Right and left peroneal nerve conduction studies with compression peroneal neuropathy at the fibula.
A, Sites of stimulation and recording. B, Compound muscle action potentials. Note the decrease in amplitude (con-
duction block) with stimulation at the fibula on the right. Ant. Tibial M., anterior tibial muscle; Ext. Dig. Br., ex-
tensor digitorum brevis muscle. (Courtesy of Dr. E. H. Lambert, Mayo Clinic.)
apraxia are normal amplitudes and laten- is reduced or absent. In patients with slowly
cies; in axonal degeneration, the amplitude evolving or compressive lesions of the
of the evoked response is decreased in pro- plexus, such as tumors, a localized slowing
portion to the axonal destruction.76 Blink re- of conduction of motor fibers and occa-
flexes can be used to measure conduction sionally conduction block may be identified
across the involved segment, but they are on stimulation at the supraclavicular or nerve
commonly absent in Bell's palsy. Conduc- root level. Neuralgic amyotrophy (Parsonage-
tion studies can help differentiate hemifacial Turner syndrome) has been reported to
spasm from other facial movements by show proximal conduction block with root
demonstrating ephaptic activation of lower stimulation.77'78 Thoracic outlet syndrome,
facial muscles during periods of spasm, which has been reported to show abnor-
called the lateral spread response. malities on nerve conduction studies, is usu-
Most brachial plexus lesions are trau- ally a vascular syndrome with a change, if
matic, and nerve conduction studies are of any, only in sensory potential amplitudes
limited value. Generally, the amplitude of and little or no slowing of nerve conduction.
the CMAP is reduced and sensory responses
are absent in the distribution of the dam- RADICULOPATHIES
aged fibers. In patients with lower trunk le-
sions, the ulnar sensory response is absent, Cervical and lumbosacral radiculopathies
and in those with upper trunk lesions, the are not usually associated with changes in
median sensory response of the index finger motor nerve conduction studies; however, if
there is sufficient destruction of axons and or localized slowing of conduction. Con-

wallerian degeneration in the distribution of duction block is a change in size of the
the nerve being tested, the amplitude of the CMAP when stimulating at two points near
CMAP may be decreased.79 For example, in each other along the nerve. Both conduc-
an L5 radiculopathy with weakness, the re- tion block and slowing of conduction rep-
sponse of the extensor digitorum brevis mus- resent pathophysiologic changes in the
cle on peroneal nerve stimulation is often of nerve, which can sometimes be predicted by
low amplitude or absent. In the presence of the changes found on nerve conduction
atrophy and a low CMAP, there may be mild studies. These changes can be helpful in
slowing of conduction in the motor axons defining prognosis for improvement after
innervating the atrophic muscle. In a few pa- nerve damage.
tients with lumbosacral radiculopathy, mea-
surements of F-wave or H-reflex latencies in
mild lesions have been valuable in identify- REFERENCES
ing proximal slowing of conduction.79 Be-
1. Dyck PJ, Karnes JL, O'Brien PC, Litchy WJ, Low PA,
cause most lesions of the spinal nerve and Melton LJ. The Rochester Diabetic Neuropathy
nerve root are proximal to the dorsal root Study: reassessment of tests and criteria for diag-
ganglion, the sensory potentials usually are nosis and staged severity. Neurology 42:1164—1170,
normal, even in the distribution of a sensory 1992.
2. Tjon ATAM, Lemkes HH, van der Kamp-Huyts AJ,
deficit. This phenomenon is valuable in van DijkJG. Large electrodes improve nerve con-
identifying avulsion of a nerve root in which duction repeatability in controls as well as in pa-
there is total anesthesia and loss of motor tients with diabetic neuropathy. Muscle Nerve
function, with normal sensory potentials. No 19:689-695, 1996.
evidence has been found of double crush, a 3. van DijkJG, van Benton I, Kramer CG, Stegeman
DF. CMAP amplitude cartography of muscles in-
peripheral mononeuropathy related to a nervated by the median, ulnar, peroneal, and tib-
radiculopathy.80 ial nerves. Muscle Nerve 22:378-389, 1999.
4. Jonas D, Bischoff C, Conrad B. Influence of differ-
ent types of surface electrodes on amplitude, area
SUMMARY and duration of the compound muscle action po-
tential. Clin Neurophysiol 110:2171-2175, 1999.
5. Falck B, Stalberg E. Motor nerve conduction stud-
Compound muscle action potentials are ies: measurement principles and interpretation of
among the most helpful recordings in the findings. J Clin Neurophysiol 12:254-279, 1995.
electrophysiologic assessment of peripheral 6. Kimura J. Kugelberg lecture: principles and pitfalls
of nerve conduction studies. Electroencephalogr
neuromuscular disease. Compound muscle Clin Neurophysiol 106:470-476, 1998.
action potentials are the recordings made 7. Bril V, Ellison R, Ngo M, Bergstrom B, Raynard D,
for all motor conduction studies, both of the Gin H. Electrophysiological monitoring in clinical
directly recorded M wave used for periph- trials. Roche Neuropathy Study Group. Muscle
eral conduction and the F-wave late re- Nerve 21:1368-1373, 1998.
8. Buschbacher RM. Median nerve motor conduction
sponse used for testing proximal conduc- to the abductor pollicis brevis. Am J Phys Med Re-
tion. Reliable CMAP recordings require the hab 78 (Suppl):Sl-S31, 1999.
use of standard stimulating and recording 9. Daube JR. Compound muscle action potentials. In
electrode types and locations and standard DaubeJR (ed). Clinical Neurophysiology. FA Davis
Company, Philadelphia, 1996, pp 199-234.
measurement criteria.81 The sensitivity and 10. Rivner MH. Statistical errors and their effect on
specificity of motor conduction studies de- electrodiagnostic medicine. Muscle Nerve 17:811-
pend on comparing the results obtained in 814, 1994.
a patient with the normal values obtained 11. Wang SH, Robinson LR. Considerations in refer-
by using exactly the same methods. The ence values for nerve conduction studies. Phys Med
Rehabil Clin N Am 9:907-923, 1998.
normal values of motor conduction studies 12. Dyck PJ, O'Brien PC, Litchy WJ, Harper CM, Daube
vary with physiologic factors such as age and JR., Dyck PJ. Use of percentiles and normal deviates
temperature, which must be controlled and to express nerve conduction and other test abnor-
adjusted. malites. Muscle Nerve 24:307-310, 2001.
13. Andersen H, Stalberg E, Falck B. F-wave latency, the
Motor nerve conduction studies with most sensitive nerve conduction parameter in pa-
CMAPs localize focal lesions in a nerve by tients with diabetes mellitus. Muscle Nerve 20:1296-
identifying either localized conduction block 1302, 1997.
266 Motor Pathways
14. Guiloff RJ, Modarres-Sadeghi H. Preferential gen- huber ME, Zierz S. Length dependence of variables
eration of recurrent responses by groups of motor associated with temporal dispersion in human mo-
neurons in man. Conventional and single unit F tor nerves. Muscle Nerve 24:527-533, 2001.
wave studies. Brain 114:1771-1801, 1991. 34. Franssen H, Wieneke GH, Wokke JH. The influence
15. Young MS, Triggs WJ. Effect of stimulator orienta- of temperature on conduction block. Muscle Nerve
tion on F-wave persistence. Muscle Nerve 21:1324- 22:166-173, 1999.
1326, 1998. 35. Ochoa J, Fowler TJ, Gilliatt RW. Anatomical
16. NobregaJA, Manzano GM, Novo NF, Monteagudo changes in peripheral nerves compressed by a
PT; Sample size and the study of F waves. Muscle pneumatic tourniquet. J Anat 113:433-455, 1972.
Nerve 22:1275-1278, 1999. 36. Brown WF, Yates SK. Percutaneous localization of
17. Weber F. The diagnostic sensitivity of different F conduction abnormalities in human entrapment
wave parameters. J Neurol Neurosurg Psychiatry 65: neuropathies. Can J Neurol Sci 9:391-400, 1982.
535-540, 1998. 37. Andersson PB, Yuen E, Parko K, So YT. Electro-
18. Rivner MH. The contemporary role of F-wave stud- diagnostic features of hereditary neuropathy with
ies. F-wave studies: limitations. Muscle Nerve 21: liability to pressure palsies. Neurology 54:40-44,
1101-1104, 1998. 2000.
19. Magistris MR, Roth G. Motor axon reflex and in- 38. Mulder DW, Lambert EH, Bastron JA, Sprague RG.
direct double discharge: ephaptic transmission? A The neuropathies associated with diabetes mellitus:
reappraisal. Electroencephalogr Clin Neurophysiol a clinical and electromyographic study of 103 un-
85:124-130, 1992. selected diabetic patients. Neurology 11:275-284,
20. Dioszeghy P, Stalberg E. Changes in motor and sen- 1961.
sory nerve conduction parameters with tempera- 39. van DijkJG, Meulstee J, Zwarts MJ, Spaans F. What
ture in normal and diseased nerve. Electroen- is the best way to assess focal slowing of the ulnar
cephalogr Clin Neurophysiol 85:229-235, 1992. nerve? Clin Neurophysiol 112:286-293, 2001.
21. Garcia A, Calleja J, Antolin FM, Berciano J. Pe- 40. Campbell WW. The value of inching techniques
ripheral motor and sensory nerve conduction stud- in the diagnosis of focal nerve lesions. Inching is
ies in normal infants and children. Clin Neuro- a useful technique. Muscle Nerve 21:1554-1556,
physiol 111:513-520, 2000. 1998.
22. Stetson DS, AlbersJW, Silverstein BA, Wolfe RA. Ef- 41. Wilson JR, StittsworthJD, Kadir A, Fisher MA. Con-
fects of age, sex, and anthropometric factors on duction velocity versus amplitude analysis: evidence
nerve conduction measures. Muscle Nerve 15: for demyelination in diabetic neuropathy. Muscle
1095-1104, 1992. Nerve 21:1228-1230, 1998.
23. Robinson LR, Micklesen PJ, Wang L. Strategies for 42. ShinJB, Seong YJ, Lee HJ, Kim SH, Suk H, Lee YJ.
analyzing nerve conduction data: superiority of a The usefulness of minimal F-wave latency and
summary index over single tests. Muscle Nerve sural/radial amplitude ratio in diabetic polyneu-
21:1166-1171, 1998. ropathy. Yonsei Med J 41:393-397, 2000.
24. Bromberg MB,Jaros L. Symmetry of normal motor 43. Roth G, Magistris MR. Indirect discharges as an
and sensory nerve conduction measurements. Mus- early nerve conduction abnormality in the Guillain-
cle Nerve 21:498-503, 1998. Barre syndrome. Eur Neurol 42:83-89, 1999.
25. Krarup C. Pitfalls in electrodiagnosis. J Neurol 44. Menkes DL, Hood DC, Ballesteros RA, Williams DA.
246:1115-1126, 1999. Root stimulation improves the detection of ac-
26. Wells MD, Gozani SN. A method to improve the es- quired demyelinating polyneuropathies. Muscle
timation of conduction velocity distributions over a Nerve 21:298-308, 1998.
short segment of nerve. IEEE Trans Biomed Eng 45. Rowin J, Meriggioli MN. Electrodiagnostic signifi-
46:1107-1120, 1999. cance of supramaximally stimulated A-waves. Mus-
27. Tu Y, Honda S, Tomita Y. Estimation of the con- cle Nerve 23:1117-1120, 2000.
duction velocity distribution of peripheral nerve 46. Feinberg DM, Preston DC, Shefner JM, Logigian
trunks. Front Med Biol Eng 9:189-197, 1999. EL. Amplitude-dependent slowing of conduction in
28. Schulte-Mattler WJ, Jakob M, Zierz S. Assessment of amyotrophic lateral sclerosis and polyneuropathy.
temporal dispersion in motor nerves with normal Muscle Nerve 22:937-940, 1999.
conduction velocity. Clin Neurophysiol 110:740- 47. Ibrahim IK, el-Abd MA. Giant repeater F-wave in
747, 1999. patients with anterior horn cell disorders. Role of
29. Bostock H, Sears TA. Continuous conduction in motor unit size. AmJ Phys Med Rehabil 76:281-287,
demyelinated mammalian nerve fibers. Nature 1997.
263:786-787, 1976. 48. Gutmann L, Fakadej A, RiggsJE. Evolution of nerve
30. McCluskey L, Feinberg D, Cantor C, Bird S. conduction abnormalities in children with domi-
"Pseudo-conduction block" in vasculitic neuropa- nant hypertrophic neuropathy of the Charcot-
thy. Muscle Nerve 22:1361-1366, 1999. Marie-Tooth type. Muscle Nerve 6:515-519, 1983.
31. Olney RK. Consensus criteria for the diagnosis of 49. Lewis RA, Sumner AJ, Shy ME. Electrophysiologi-
partial conduction block. Muscle Nerve Suppl cal features of inherited demyelinating neuropa-
8:S225^S229, 1999. thies: a reappraisal in the era of molecular diagno-
32. Pfeiffer G, Wicklein EM, Wittig K. Sensitivity and sis. Muscle Nerve 23:1472-1487, 2000.
specificity of different conduction block criteria. 50. Cappelen-Smith C, Kuwabara S, Lin CS, Mogyoros
Clin Neurophysiol 111:1388-1394, 2000. I, Burke D. Activity-dependent hyperpolarization
33. Schulte-Mattler WJ, Muller T, Georgiadis D, Korn- and conduction block in chronic inflammatory de-
myelinating polyneuropathy. Ann Neurol 48:826- nerve conduction studies in patients with ulnar
832, 2000. neuropathy at the elbow. Arch Phys Med Rehabil
51. Report from an Ad Hoc Subcommittee of the 79:87-89, 1998.
American Academy of Neurology AIDS Task Force: 66. KincaidJC, Phillips LH, Daube JR. The evaluation
Research criteria for diagnosis of chronic inflam- of suspected ulnar neuropathy at the elbow. Nor-
matory demyelinating polyneuropathy (CIDP). mal conduction study values. Arch Neurol 43:44—
Neurology 41:617-618, 1991. 47, 1986.
52. Haq RU, Fries TJ, Pendlebury WW, Kenny MJ, Bad- 67. Herrmann DN, Preston DC, Mclntosh KA, Logigian
ger GJ, Tandan R. Chronic inflammatory demyeli- EL. Localization of ulnar neuropathy with conduc-
nating polyradiculoneuropathy: a study of pro- tion block across the elbow. Muscle Nerve 24:698-
posed electrodiagnostic and histologic criteria. 700, 2001.
Arch Neurol 57:1745-1750, 2000. 68. Ochiai N, Honmo J, Tsujino A, Nisiura Y. Electro-
53. Kaji R, Bostock H, Kohara N, Murase N, Kimura J, diagnosis in entrapment neuropathy by the arcade
Shibasaki H. Activity-dependent conduction block of Struthers. Clin Orthop 12:129-135, 2000.
in multifocal motor neuropathy. Brain 123:1602- 69. Singh N, Behse F, Buchthal F. Electrophysical study
1611, 2000. of peroneal palsy. J Neurol Neurosurg Psychiatry
54. Taylor BV, Wright RA, Harper CM, Dyck PJ. Nat- 37:1202-1213, 1974.
ural history of 46 patients with multifocal motor 70. Levin KH. L5 radiculopathy with reduced superfi-
neuropathy with conduction block. Muscle Nerve cial peroneal sensory responses: intraspinal and ex-
23:900-908, 2000. traspinal causes. Muscle Nerve 21:3-7, 1998.
55. Lewis RA, Sumner AJ. The electrodiagnostic dis- 71. Robinson LR. Traumatic injury to peripheral
tinctions between chronic familial and acquired nerves. Muscle Nerve 23:863-873, 2000.
demyelinative neuropathies. Neurology 32:592- 72. Trojaborg W. Rate of recovery in motor and sen-
596, 1982. sory fibres of the radial nerve: clinical and electro-
56. Fowler TJ, Danta G, Gilliatt RW. Recovery of nerve physiological aspects. J Neurol Neurosurg Psychia-
conduction after a pneumatic tourniquet: observa- try 33:625-638, 1970.
tions on the hind-limb of the baboon. J Neurol Neu- 73. Cruz-Martinez A, Armijo A, Fermoso A, Moraleda
rosurg Psychiatry 35:638-647, 1972. S, Mate I, Marin M. Phrenic nerve conduction study
57. Girlanda P, Quartarone A, Sinicropi S, et al. Elec- in demyelinating neuropathies and open-heart sur-
trophysiological studies in mild idiopathic carpal gery. Clin Neurophysiol 111:821-825, 2000.
tunnel syndrome. Electroencephalogr Clin Neuro- 74. Post M. Diagnosis and treatment of suprascapular
physiol 109:44-49, 1998. nerve entrapment. Clin Orthop 368:92-100, 1999.
58. Stevens JC. AAEM minimonograph #26: The elec- 75. Uludag B, Ertekin C, Turman AB, Demir D, Kiyli-
trodiagnosis of carpal tunnel syndrome. American oglu N. Proximal and distal motor nerve conduc-
Association of Electrodiagnostic Medicine. Muscle tion in obturator and femoral nerves. Arch Phys
Nerve 20:1477-1486, 1997. Med Rehabil 81:1166-1170, 2000.
59. American Association of Electrodiagnostic Medi- 76. Olsen PZ. Prediction of recovery in Bell's palsy.
cine: Practice parameter for electrodiagnostic stud- Acta Neurol Scand Suppl 61:1-121, 1975.
ies in carpal tunnel syndrome: summary statement. 77. Lo YL, Mills KR. Motor root conduction in neural-
Muscle Nerve Suppl 8:S141-S143, 1999. gic amyotrophy: evidence of proximal conduction
60. Stalberg E, Stalberg S, Karlsson L. Automatic carpal block. J Neurol Neurosurg Psychiatry 66:586-590,
tunnel syndrome tester. Clin Neurophysiol 111: 1999.
826-832, 2000. 78. Watson BV, Nicolle MW, Brown JD. Conduction
61. Lew HL, Wang L, Robinson LR Test-retest relia- block in neuralgic amyotrophy. Muscle Nerve
bility of combined sensory index: implications for 24:559-563, 2001.
diagnosing carpal tunnel syndrome. Muscle Nerve 79. Wilbourn AJ, Aminoff MJ. AAEM minimonograph
23:1261-1264, 2000. 32: The electrodiagnostic examination in patients
62. van Dijk JG, Bouma PA. Recognition of the Martin- with radiculopathies. American Association of Elec-
Gruber anastomosis. Muscle Nerve 20:887-889, trodiagnostic Medicine. Muscle Nerve 21:1612-
1997. 1631, 1998.
63. Buchthal F, Rosenfalck A, Trojaborg W. Electro- 80. Richardson JK, Forman GM, Riley B. An electro-
physiological findings in entrapment of the median physiological exploration of the double crush hy-
nerve at wrist and elbow. J Neurol Neurosurg Psy- pothesis. Muscle Nerve 22:71-77, 1999.
chiatry 37:340-360, 1974. 81. Stalberg E, Falck B, Gilai A, Jabre J, Sonoo M, Tod-
64. Merlevede K, Theys P, van HeesJ. Diagnosis of ul- nem K. Standards for quantification of EMG and
nar neuropathy: a new approach. Muscle Nerve neurography. The International Federation of Clin-
23:478-481, 2000. ical Neurophysiology. Electroencephalogr Clin
65. Kothari MJ, Heistand M, Rutkove SB. Three ulnar Neurophysiol Suppl 52:213-220, 1999.
Chapter 22
ASSESSING THE NEUROMUSCULAR
JUNCTION WITH REPETITIVE
STIMULATION STUDIES
ANATOMY AND PHYSIOLOGY OF THE SELECTION OF NERVE-MUSCLE

NEUROMUSCULAR JUNCTION COMBINATIONS
TECHNIQUE CLINICAL CORRELATIONS
CRITERIA OF ABNORMALITY SUMMARY
RAPID RATES OF STIMULATION
Repetitive stimulation is a clinical neurophysi- cular junction as it applies to repetitive stim-

ologic technique designed and used to eval- ulation, a detailed discussion of the tech-
uate the function of the neuromuscular nique involved, criteria used to classify the
junction. The neuromuscular junction is the results as normal or abnormal, and the
anatomical site where the motor nerve axon patterns of abnormalities seen in various
connects functionally with a striated (volun- diseases.
tary) muscle fiber. The function of the neu-
romuscular junction is disturbed and the test
results are abnormal in a group of rare ANATOMY AND PHYSIOLOGY
diseases that includes myasthenia gravis, OF THE NEUROMUSCULAR
Lambert-Eaton myasthenic syndrome, botu- JUNCTION
lism, and several rare congenital myasthenic
syndromes. Repetitive stimulation is often Knowledge of the anatomy and function of
important in the detection, clarification, and the neuromuscular junction is important
follow-up of these unusual diseases. More in understanding the indications for, tech-
commonly, this test is useful in excluding niques of, and results of repetitive stimula-
these rare disorders in patients with the tion.1
common symptoms of fatigue, vague weak- Each muscle fiber is innervated by a mo-
ness, diplopia, ptosis, and malaise or with the tor neuron. The myelinated axon of the mo-
finding of weakness of uncertain origin. tor neuron divides into numerous branches
Repetitive stimulation testing demands an (collaterals), each of which loses its myelin
unusual degree of experience, attention to sheath near the muscle fiber and joins the
detail, and technical expertise to avoid mis- muscle fiber midway along its length. As the
leading false-positive and false-negative re- axonal branch nears the muscle fiber, it ex-
sults. pands into a presynaptic terminal bouton
This chapter includes a brief review of the that lies within a depression in the muscle
anatomy and physiology of the neuromus- cell membrane. The muscle cell membrane
268
Repetitive Stimulation Studies 269
Figure 22-1. Functional anatomy of the neuromuscular junction.
(postsynaptic membrane) beneath the nerve acetylcholine joins with the acetylcholine
terminal is specialized, specifically, the postreceptor and produces a small depolariza-
synaptic membrane is highly folded, the tion of the muscle membrane in the area
Junctional folds. The presynaptic neural and around the neuromuscular junction. This lo-
postsynaptic muscle membrane specializa- cal change is called a miniature end plate po-
tions constitute the neuromuscular junction, tential (MEPP).
that is, the synapse between nerve and mus- When an action potential reaches the
cle (Fig. 22-1). nerve terminal, it causes the release of a large
The presynaptic nerve terminal has spe- number of vesicles (quanta) of acetylcholine
cialized anatomical and metabolic features in a short time. The acetylcholine binds with
for the formation, storage, release, and re- the postsynaptic receptors to produce a de-
uptake of acetylcholine. Acetylcholine is repolarization that is much larger than an
quired for chemical synaptic transmission MEPP. This larger depolarization is termed
(Fig. 22-1) and is stored in synaptic vesicles an end plate potential (EPP). An EPP is large
that release their contents into the synaptic enough to depolarize the muscle membrane
cleft under appropriate conditions. The around the end plate sufficiently to generate
amount of acetylcholine contained in a sin- a muscle fiber action potential, which then
gle vesicle is called a quantum. spreads in an all-or-none fashion over the
The postsynaptic membrane contains membrane of the entire muscle fiber.
acetylcholine receptor protein molecules Through excitation-contraction coupling,
concentrated on the crest of the Junctional the action potential causes contraction of the
folds (Fig. 22-1). When acetylcholine binds muscle fiber. In normal persons, the EPP is
to the postsynaptic acetylcholine receptor much greater in amplitude than necessary
protein molecules, it causes a configura- for the muscle cell membrane to reach
tional change in the acetylcholine receptor threshold, and each EPP generates contrac-
protein molecule. This opens a pore or tion of the muscle fiber. The size of the EPP
channel in the membrane and results in de- is determined by the amount of acetyl-
polarization of the muscle cell membrane. choline in each vesicle, the number of vesi-
Randomly, presynaptic vesicles containing cles released, and the number and function
acetylcholine join the presynaptic mem- of the acetylcholine receptors stimulated.
brane and release their quanta! contents The amount of acetylcholine released at
into the neuromuscular junction. The the neuromuscular junction varies under
270 Motor Pathways
Figure 22-2. The effects of repetitive stimulation at slow and fast rates on the release of acetylcholine (ACh), end
plate potential, individual muscle fiber action potentials, and compound muscle action potential in normal sub-
jects, patients with myasthenia gravis, and patients with myasthenic syndrome.
different conditions. The mechanisms in- the EPP with repetitive firing have no effect.
volved in the release of acetylcholine by an Each nerve action potential produces con-
action potential are such that if another ac- traction of the muscle fiber.
tion potential occurs within 200 ms after the However, in disorders in which the size of
first one, the amount of acetylcholine re- the EPP is decreased to the point the am-
leased is greater with the second action po- plitude falls just above or just below thresh-
tential (Fig. 22-2). If a second action po- old, minor physiologic fluctuations in the
tential arrives later than 200 rns after the amplitude of the EPP may assume major im-
first, less acetylcholine is released with the portance. If the EPP is marginally above
second action potential. threshold, repetitive stimulation at slow rates
Thus, if a nerve is stimulated repetitively, results in a lower amplitude EPP that may
the amount of acetylcholine released varies not reach threshold and neuromuscular
depending on the rate of stimulation. At fast transmission may fail, causing a decrease in
rates of repetition, that is, more than 10/ the number of muscle fibers that contract.
second (short interval between successive If the EPP is just below threshold for gener-
stimuli), the amount of acetylcholine re- ating an action potential, repetitive stimula-
leased increases or is potentiated. After a se- tion at rapid rates may result in an increased
ries of stimuli (called "tetanic stimulation"), release of acetylcholine. An increased re-
the potentiation of acetylcholine release may lease of acetylcholine produces increased
persist for 30-60 seconds. With slow rates of EPP amplitude that may exceed threshold
repetitive stimulation, that is, less than and result in an increase (increment) or fa-
5/second (long interval between stimuli), cilitation of neuromuscular transmission,
the amount of acetylcholine released is less with an increment in the number of muscle
with each of the first four stimuli. This de- fibers responding.
crease is much greater for 2-5 minutes after
a period of exercise or after repetitive, or
tetanic, stimulation. TECHNIQUE
In normal subjects, the amplitude of the
EPP is so much greater than required to Technique is very important in repetitive
reach threshold, called the safety factor, that stimulation because, on the one hand, poor
small decreases or increases in amplitude of technique can result in "abnormal" find-
ings in patients with normal neuromuscu- movement. Immobilization devices may in-
lar transmission, leading to an erroneous clude clamps, boards, straps, sheets, and tow-
diagnosis of a disorder of neuromuscular els (Fig. 22-3). Basically, any device that
transmission.2"4 On the other hand, poor limits shifts between the electrodes and
technique may result in "normal" findings recording or stimulating sites without harm-
in patients with abnormal neuromuscular ing the patient or recording setup could be
transmission, producing a falsely negative useful.
result and a missed diagnosis. The stimulation technique is based on the
The basic techniques required are those usual techniques of supramaximal stimula-
used for routine motor nerve conduction tion with the smallest stimulus possible in
studies (Chapter 21). These basic tech- terms of duration and intensity. Therefore,
niques must be mastered before repetitive the stimulating cathode and anode must be
stimulation is attempted. Attention to these as close to the nerve as possible. The longer
basic technical details results in reliable and the duration and the greater the intensity of
rapid testing. Ignoring these technical de- the stimulus, the more uncomfortable the
tails can produce unreliable results, requir- patient will be, the more movement that will
ing repetition of the procedure and causing occur, the more stimulus artifact that will be
great expenditure of time and unnecessary created, and the greater the chance of stim-
patient discomfort. ulating unwanted nearby nerves. For these
Recording electrode placement is the reasons, the nerve must be carefully local-
same as for routine motor nerve conduction ized. In repetitive stimulation, the stimulus
studies, with the active, or Gl, electrode over site must be stable. This is a challenge be-
the end plate or motor point, and the G2 cause movements produced by the stimula-
electrode over the tendon. Repetitive stimu- tion and the reaction of patients tend to
lation requires that the Gl electrode be po- cause movement of the stimulating elec-
sitioned so that the initial upward, or nega- trode in relation to the nerve. Stimulation
tive, deflection is very sharp and there is no of the nerve distally limits the contraction of
initial downward (positivity at Gl) deflec- unwanted muscles and results in less artifact.
tion. The recording electrodes and wires Stimulation with a near-nerve needle elec-
must be attached securely so that no move- trode results in shorter duration, lower in-
ment or loosening occurs during the study. tensity, and more localized stimulation. This
The ground electrode should be positioned may produce less stimulus artifact, less pa-
to minimize stimulus artifact. tient discomfort, and more reliable results.
Patient cooperation is critical for a tech-
nically satisfactory study. The patient has to
understand the purpose and importance of
the study. Patient cooperation results in a
shorter, less painful study and produces
more reliable results. Instruct patients about
the purpose of the study and the importance
of remaining as relaxed as possible. The pa-
tient should let the movement produced by
the stimulation occur but avoid the con-
traction of other muscles, especially between
stimuli.
Immobilization is very important in repet-
itive stimulation studies. Movement results
in broken connectors and wires, loose elec-
trodes, alterations in the relationship be-
tween the recording electrode and the mus- Figure 22-3. Example of a simple immobilization de-
cles, introduction of unwanted activity from vice for ulnar motor repetitive stimulation studies. Two
neighboring muscles, and shifts in the loca- boards are connected by adjustable clamps and padded
for patient comfort. The fingers are immobilized with
tion of the stimulator and probably many a Velcro strap. The hand is immobilized in this fash-
more effects. Some form of physical restraint ion after the recording electrodes have been attached
should be used when possible to minimize securely.
272 Motor Pathways
The stimulus must be strong enough to

excite all the motor axons, and the intensity
must be increased 10%-25% above the level
that produced a maximal response. This
supramaximal stimulus cancels the effect of
small degrees of movement of the stimulat-
ing electrode away from the nerve. Watch
the response carefully for any change in am-
plitude or configuration. If any such change
is noted, the adequacy of the stimulus in-
tensity should be checked immediately.
The stimulus rate and the number of stim-
uli vary depending on the clinical problem.
In most situations, slow rates of stimulation
of 2/second, with an interstimulus interval Figure 22-4. Example of a repetitive stimulation study
in a patient with myasthenia gravis, with stimulation of
of 500 ms, will maximize any potential the ulnar nerve and recording of the compound mus-
decrement. The greatest decrease in acetyl- cle action potential from the abductor digit! minimi
choline release at slow rates of stimulation (ADM) muscle of the hand. A train of four stimuli were
occurs during the first four stimuli. For given at 2/second (s) with the muscle rested on three
occasions, separated by 30 seconds of rest for the three
these reasons, the best standard approach is baseline studies (Bl, B2, B3). Next, the muscle was ex-
four or five stimuli at 2 Hz. The slower the ercised voluntarily for 1 minute (m). The train of four
rate and the fewer the number of stimuli stimuli at 2 Hz was repeated 3 seconds, 30 seconds, 1
given, the better the patient is able to tol- minute, 2 minutes, 3 minutes, and 4 minutes after ex-
erate the procedure. ercise (AE). Top, Histogram of the amplitudes of the
four responses of each train. This histogram is a good
The train of four stimuli should be re- example of the pattern of abnormality that can be ex-
peated, with at least 15—30 seconds of rest pected in disorders of neuromuscular transmission. In
between trains (Fig. 22-4). The trains are re- each train of four, the greatest decrement is between
peated to check for reproducible ampli- the first and second response, with less decrement be-
tween the second and third and third and fourth re-
tudes, areas, and configurations as well as sponses. Immediately after exercise, the decrement is
the stability of the baseline, the presence of less and there is some postactivation facilitation of the
stimulus artifact, patient relaxation or move- amplitude of the compound muscle action potential at
ment, and the stability of the recording and baseline. At 4 minutes after exercise (AE 4 m), the
stimulating electrodes. If any abnormalities decrement is greater than it was at baseline. Middle, The
four responses to 2/second stimulation 4 minutes after
are found, it is wise to consider the problem exercise are displayed in the x-shifted fashion. Bottom,
to be a technical one and to proceed with a Numerical display of the amplitudes and areas of each
systematic checklist to eliminate artifact. of the responses (Pot [potential] 1-4) and the per-
After all have been checked, three repro- centage decrements (deer) in amplitude (amp) and
area 4 minutes after exercise.
ducible and technically satisfactory sets of
four stimuli at 2 Hz with 15-30 seconds be-
tween sets should be obtained as a baseline. tential is potentiated for 30-60 seconds. Dur-
Depending on the clinical problem and ing this period of postactivation, or post-
the results of the baseline 2 Hz repetitive tetanic, potentiation, the amplitude of the
stimulation, a decision must be made about EPP is increased, and the amplitude of the
the usefulness of further testing of neuro- evoked potential may be markedly increased
muscular transmission with repetitive stimu- in the myasthenic syndrome or botulism. In
lation after exercise or tetanic stimulation myasthenia gravis, the decrement at baseline
(Fig. 22-5). In general, exercise is done for may be decreased or repaired during this pe-
a brief period (10 seconds) or an interme- riod. This phenomenon is called postactiva-
diate period (1 minute). tion facilitation, or repair (Fig. 22—4 and 22-5).
Brief (10 seconds) periods of exercise in If there is no decrement or only a very
a cooperative patient have almost the same questionable decrement on baseline testing
effect as rapid stimulation at 20-50 Hz for and the purpose of the electrophysiologic
10 seconds but are not nearly as uncom- examination is to unmask a borderline or
fortable. After 10 seconds of exercise, the re- mild defect of neuromuscular transmission,
lease of acetylcholine with each action po- the patient should be exercised for 1 minute.
Figure 22-5. Examples of supramaximal repetitive stimulation at 3 Hz of the ulnar nerve at the wrist while record-
ing over the hypothenar muscle at rest and at 3 seconds, 2 minutes, and 10 minutes after exercise. Each waveform
consists of three compound muscle action potentials superimposed in a normal subject, in a patient with myas-
thenia gravis, and in another patient with myasthenic syndrome. In the patient who has myasthenia gravis, the de-
crease in amplitude from the first to third response recovers after 10 seconds of exercise and is accentuated 2 min-
utes after exercise. In a patient with the Lambert-Eaton myasthenic syndrome, the compound muscle action potential
(CMAP) at rest is very low in amplitude; there is a decrement that is not appreciable at this sensitivity. After brief
exercise, there is a transient facilitation in amplitude of the CMAP. (From Lambert EH, Rooke ED, Eaton LM,
Hodgson CH. Myasthenic syndrome occasionally associated with bronchial neoplasm: neurophysiologic studies. In
Viets HR [ed]. Myasthenia Gravis [The Second International Symposium Proceedings]. Charles C Thomas, Pub-
lisher, Springfield, IL, 1961, pp 362-410. By permission of the publisher.)
The 1-minute period of exercise should be The display of the results varies with the
performed with 20 seconds of exercise, 2-5 machine, hardware, software, and display de-
seconds of rest, exercise for 20 seconds, 2-5 vices (Fig. 22-6). In general, the sensitivity
seconds of rest, and exercise for 20 seconds should be adjusted to display the potentials
to simulate prolonged stimulation. Two-five as large as possible without overflowing or
minutes after exercise, the amount of acetyl- blocking. The sweep speed should be fast
choline released with each stimulus should enough to spread the potential out so that
be minimal, providing the greatest chance it can be analyzed visually and slow enough
for detecting any defect of neuromuscular that the entire potential is displayed, in-
transmission (Fig. 22-5). Usually, after 1 cluding any late components. The repetitive
minute of exercise, four stimuli are given at potentials should be displayed in an x-shifted
2 Hz immediately after exercise, and at 30, fashion. This means that the onset of the
60,120,180, and 240 seconds after exercise. sweep for each successive stimulus is shifted
As emphasized above, any change in ampli- to the right on the horizontal, or x-axis,
tude, configuration, or area should be con- or delayed so that every potential can be
sidered a technical problem, and technical analyzed individually. Thus, if there are
factors, including strength of stimulation, changes, it is possible to determine which
should be checked. potential in the sequence of four changed
274 Motor Pathways
nerve conduction studies are complete. Mea

surements of amplitude can be made either
by hand or with manual or automated mark-
ers. Measurements of area are quickly and
reliably made by computer-driven digital ma-
chines provided that the markers are placed
accurately. Such computer-based digital ma-
chines can quickly measure changes in am-
plitude and area between the first and sub-
sequent responses.
Decreases in amplitude or area are ex-
pressed as a percentage decrement as de-
rived from the formula (see below).
Note that no change from baseline is a fa-
Figure 22-6. A, Super-imposition and, B, staggering of cilitation of 100% and a doubling in ampli-
four compound muscle action potentials (CMAPs). The tude is a facilitation of 200% (Fig. 22-7).
potentials evoked by repetitive stimulation can be dis-
played in a staggered, or X-shifted, fashion to allow in- A small increment in amplitude (< 40%)
spection of each potential and determination of the may be seen in normal subjects. The incre-
sequence of any changes. Superimposition of the ment is usually accompanied by a decrease
potentials allows easier visual identification of small in duration and little or no change in area.
decrements.
An increased synchronization of the firing
of motor units rather than an increase in the
number of units or the amplitude of the re-
and what the order of change was. Super- sponse of individual fibers best explains this.
imposition of successive stimuli may allow Abnormal results that are not caused by
closer inspection and detection of changes disease may be obtained in several situations.
in amplitude, area, or configuration but 1. There may be movement of the stimu-
does not allow the determination of the se- lating electrode in relation to the
quence of changes. It is necessary to print nerve, producing a random variation
the results immediately or to store them for in amplitude or, less frequently, a se-
review and printing later. quential decrement. This is more com-
Mainly, the amplitude, duration, and area mon just after exercise and is more
of the waveform of the compound muscle likely to occur when the stimulus is not
action potential should be measured (Fig. supramaximal. Submaximal stimula-
22-4). This assumes that the other standard tion is suggested by a loss of amplitude
/Amplitude of \ /Amplitude of
I First Response L/ ~ \ Subsequent Response
Percentage Decrement = -* — j—; X 100
Amplitude of First Response
or
_ / 1 — Amplitude of Last Response
Percentage Decrement = X 100
Increases in amplitude or area are expressed as percentage increments or facilitation, as

in the following:
/Amplitude of \ /Amplitude of \
\Last Response / ~~ 1N First Response/
L
Percentage Increment = -* -f—-—' „ X 100
Facilitated Amplitude
Facilitation = X 100
Baseline Amplitude
Figure 22-7. 1-10, Examples of facilitation of compound muscle action potentials of the hypothenar muscle dur-
ing 40/second-50/second repetitive nerve stimulation in 10 patients with Lambert-Eaton myasthenic syndrome.
The potentials are shown at slow sweep speeds. The facilitation ranges from 160%-1,400%. Note the variety of pat-
terns of facilitation that occur. In some recordings, a constant amplitude shock artifact is seen with each compound
muscle action potential (1, 2, 3, 5, 7, and 9). (Courtesy of Dr. E. H. Lambert.)
of the initial response in the train from

the baseline responses (Fig. 22-8).
2. Movement of the recording electrode
relative to the underlying muscle pro-
duces a change in configuration and
amplitude that is usually random but
occasionally may be in a decremental
or incremental pattern. Movement may
produce a shift in the baseline or visi-
ble muscle activity between the stimuli.
3. Technical factors should be suspected
if (a) the results are not reproducible;
(b) the pattern of decrement, incre-
ment, postexercise potentiation, or ex-
haustion is unusual; (c) there are base-
line shifts or changes in configuration;
or (d) there is evidence of muscle ac-
tivity or movement between stimuli.
Figure 22-8. Technical problems such as poor relax-
If technical factors cannot be excluded ation or movement during repetitive stimulation can
and are suspected to be the cause of the ab- produce apparent decrements in normal subjects, as
normality, the study should be considered seen in these examples.
276 Motor Pathways
technically inadequate and not diagnostic. 3. The changes induced by exercise with
This is preferable to making a serious diag- potentiation and exhaustion should be
nosis on the basis of questionable data. Re- compatible with what is seen in the dis-
peating the study at a later date or under ease under question (Fig. 22-5).
other circumstances may be helpful. 4. Edrophonium (Tensilon) or neostig-
False-negative results or normal results in mine (Prostigmin) should decrease or
a patient with a disorder of neuromuscular correct the abnormalities seen in myas-
transmission can be caused by low tempera- thenia gravis.
tures, which will mask a mild defect of neu-
romuscular transmission. The patient must
be warm before and during the study. This RAPID RATES OF STIMULATION
implies that temperatures are monitored
and the hand skin temperature is greater Rapid rates of stimulation of 10-50 Hz are
than 32°C and the foot skin temperature is helpful in some disorders such as the Lam-
greater than 30°C. If the temperatures are bert-Eaton myasthenic syndrome and botu-
cool, the patient should be warmed before lism in which there may be a marked in-
further studies are performed. crement with rapid rates of stimulation
False-negative results may occur if treat- (Fig. 22-7). Rapid repetitive stimulation is
ment with anticholinesterases, immunosup- painful. However, brief exercise with volun-
pressants, intravenous immunoglobulin, or tary, strong contraction produces the same
plasma exchange is successful. The test effect without pain, although some patients
should be conducted when the patient is are unable to produce a strong contraction
most symptomatic—usually late in the day because of extreme weakness, lack of un-
when fatigued. When possible, treatment derstanding, or inability to cooperate. If a
with acetylcholinesterase inhibitors such as disorder is suspected such as the Lambert-
pyridostigmine (Mestinon) or neostigmine Eaton myasthenic syndrome, rapid rates of
should be discontinued for at least 4-6 hours stimulation are necessary. Explain the details
before the test. The risk to the patient from of the test to the patient before proceeding.
discontinuing treatment must be weighed Stress the importance of remaining as still
against the importance of the test. and relaxed as possible so that the results of
Patients for whom the diagnosis is in ques- the test will be reliable. Before beginning,
tion probably should be tested when the ef- be careful to check the machine and elec-
fects of treatments such as plasma exchange, trode setup to exclude any technical errors
intravenous gamma globulin, and cortico- that might make the results uninterpretable.
steroids are minimal. If a large facilitation is expected, the gain or
sensitivity should be set so a much larger re-
sponse can be recorded without blocking.
CRITERIA OF ABNORMALITY Stimulate at 20-50 Hz for 2-10 seconds, de-
pending on the situation. An increment
In normal subjects, no decrement should oc- greater than 40% in adults or 20% in infants
cur with 2 Hz stimulation. In fact, technical after 3 seconds of stimulation is considered
problems often result in small decrements. abnormal.
A conservative criterion of abnormality is to
require a decrement of at least 10% in two
different muscle/nerve preparations that SELECTION OF NERVE-MUSCLE
meet the following criteria: COMBINATIONS
1. Reproducible results should be ob-
tained on repeated testing. Most of the diseases that are studied tend
2. The pattern of decrement or the shape to affect certain muscles more than others.
of the envelope of the responses should Proximal muscles are usually more involved
be like that seen in neuromuscular dis- than distal muscles. A basic set of rules can
orders. In myasthenia gravis the decre- be developed for selecting the nerve-muscle
ment is usually greatest between the preparations to study in the search for de-
first and second response in the train fects of neuromuscular transmission (Table
of four (Fig. 22-4). 22-1):
Table 22-1. Specific Nerve-Muscle Combinations
Nerve Muscle Stimulation Site Advantages Disadvantages Immobilization
Ulnar Abductor digiti Wrist Reliably immobilized, Distal muscle may be spared Velcro strap, clamp
quinti manus well tolerated
Median Abductor pollicis Wrist Well tolerated Distal muscle may be spared, Thumb restrained
brevis difficult to immobilize with towel
Musculocutaneous Biceps Lower edge axilla Proximal muscle Unstable stimulus, difficult to Board
immobilize, painful
Axillary Deltoid Supraclavicular Proximal muscle Unstable stimulus, difficult to Large Velcro strap
immobilize, painful or sheet
Spinal accessory Trapezius Posterior border upper Proximal muscle, Difficult to immobilize Strap or hands under
sternocleidomastoid well tolerated chair
Facial Nasalis Between mastoid and Proximal muscle Painful, unstable stimulation, None
tragus shock artifact, cannot
immobilize. Masseter
contraction
Peroneal Anterior tibial Knee Leg muscle Distal muscle may be spared Board
Femoral Rectus femoris Femoral triangle Proximal leg muscle Painful, difficult to immobilize Restrained manually
278 Motor Pathways
1. Because the clinically weakest muscles myasthenia gravis associated with the use of
are most likely to show a defect on repet- the drug D-penicillamine.
itive stimulation, those nerve-muscle Lambert-Eaton myasthenic syndrome is a
combinations should be checked. Even rare entity associated with systemic malig-
unusual combinations such as stimula- nancies such as small cell carcinoma of the
tion of the phrenic nerve while record- lung in about half the cases.7 This syndrome
ing from the diaphragm are feasible.5 is an autoimmune disorder resulting from
2. In most cases, proximal muscles are an antibody that alters the function of the
more likely to show abnormalities. voltage-gated calcium channels in the axon
3. A clinically uninvolved muscle is un- terminal of the neuromuscular junction.
likely to reveal abnormal results. The result is decreased release of acetyl-
4. Testing distal muscles is technically eas- choline with each action potential. The
ier and more reliable and causes less amount of acetylcholine released increases
discomfort to the patient. rapidly with rapid rates of activation or stim-
5. Generally, it is probably wise to begin ulation. The weakness is more generalized
with a distal muscle in the most affected than in myasthenia gravis but involves prox-
extremity and then move more proxi- imal muscles more than distal muscles and
mally if indicated. leg muscles more than arm muscles. The bul-
bar muscles are not involved as prominently
as they are in myasthenia gravis. Some pa-
tients have autonomic symptoms such as dry
CLINICAL CORRELATIONS mouth, impotence, and constipation. Base-
line nerve conduction studies usually dem-
Myasthenia gravis is the classic disease of the onstrate low-amplitude compound muscle
neuromuscular junction.6 It usually is the reaction potentials at rest. Repetitive stimula-
sult of an autoimmune-mediated attack on tion at 2 Hz produces decrements similar to,
the acetylcholine receptor on the muscle but often more prominent than, those seen
(postsynaptic) cell membrane. This results in myasthenia gravis. Brief (10 second) ex-
in fewer functional receptors and fluctuat- ercise or rapid stimulation at 50 Hz produces
ing, fatiguable weakness involving proximal a marked increment or facilitation of the am-
muscles more than distal ones, particularly plitude of the compound muscle action po-
the bulbar muscles and often the extraocu- tentials to 2-20 times the baseline amplitude
lar muscles. Experimentally, the amplitudes (Figs. 22-5 and 22-7). This effect is transient
of MEPPs and EPPs are low (Fig. 22-2). The and must be looked for in a well-rested,
resting compound muscle action potential is warm muscle immediately after brief exer-
normal or low normal except in the more cise. After 60-120 seconds, the amplitude re-
severe cases. With repetitive stimulation at 2 turns to baseline and the decrement returns
Hz, there is a decrement, with the greatest and may be more prominent 3-4 minutes
decrease between the first and second re- after exercise.
sponse and lesser decreases after that (Fig. Botulism, another rare disorder, is caused
22-4 and 22-5). By the fifth response, the by exposure to one of the seven types of
decrement levels off. After 10 seconds of ex- toxin produced by Clostridium botulinum.8~11
ercise, the decrement is partially or com- In adults, the toxin is ingested in inade-
pletely repaired and there may be an incre- quately preserved or prepared food. In in-
ment in the CMAP amplitude of the fants, botulism is caused by ingestion of
compound muscle action potential. In se- spores that germinate into bacteria that pro-
vere cases in which the amplitude of the duce botulinum toxin in the gut. From 12
compound muscle action potential at base- to 36 hours after the toxin is ingested,
line is low, there may be a marked increment blurred vision, dysarthria, dysphagia, dry
in this amplitude after brief exercise, like mouth, dyspnea, and generalized weakness
that usually associated with Lambert-Eaton develop. The toxin markedly decreases the
myasthenic syndrome.4 At 2-4 minutes after number of quanta of acetylcholine released
1 minute of exercise, the decrement is larger by an action potential, thus reducing the am-
than at rest. Similar findings are seen in plitude of the EPP. Clinically, results of
Figure 22-9. Example of repetitive stimulation study of the ulnar nerve while recording from hypothenar muscles
in a patient with myotonia. There is a very small reproducible decrement at baseline. The decrement and the am-
plitude of the compound muscle response are decreased after 10 seconds of exercise, but 3 minutes after exercise
the amplitude and the decrement have increased again.
nerve conduction studies are normal except that is more prominent as the rate of stim-
for low-amplitude compound muscle action ulation is increased. The decrement occurs
potentials. Rapid repetitive stimulation at 50 at lower stimulation frequencies in myotonia
Hz or brief exercise produces an increment congenita than in myotonic dystrophy. In
in most cases (in 62% of adults and 92% of paramyotonia, the decrement may progress
infants). This may not be found in very se- to electrical silence with cooling of the mus-
vere cases; when found, it is in the range of cle. In myotonic disorders, the decrement is
30%-200%. There is only a small decrement decreased after exercise and the amplitude
at slow rates of repetitive stimulation in 56% of the compound muscle action potential is
of the cases in infants and in only 8% of the reduced for several minutes (Fig. 22-9). It
adult cases. Postexercise exhaustion is not may be possible to separate myotonic dys-
seen. trophy, which demonstrates such a decrease
Congenital myasthenia is the general term in amplitude after exercise, from proximal
applied to a group of rare inherited disor- myotonic myopathy (PROMM), which does
ders of neuromuscular transmission caused not have such a decrease.15
by various structural or functional alter- Periodic paralysis may reveal a gradual re-
ations of the neuromuscular junction, such duction in amplitude of the compound mus-
as decreased filling of the synaptic vesicles cle action potential for 20-30 minutes after
with acetylcholine, decreased number of exercising the muscle for 3-5 minutes (Fig.
synaptic vesicles available for release, ab- 22-10).16 There usually is no decrement at
sence of one form of acetylcholinesterase, slow rates of repetitive stimulation.
decreased number of acetylcholine recep- In disorders of the motor neuron such as
tors on the junctional folds, or kinetic de- amyotrophic lateral sclerosis, a decrement at
fects in the acetylcholine receptor that slow rates of stimulation may be found that
increase or decrease the response to acetyl- is less prominent after brief exercise and in-
choline.12"14 Results of repetitive stimula- creased after prolonged exercise. Such ab-
tion are abnormal, but the pattern of abnormalities are seen more frequently when
normality varies among the different the disease is rapidly progressive.
disorders. Special technique is required in Disordered neuromuscular transmission
patients with a defect of acetylcholine resyn- is rarely found in peripheral neuropathies
thesis or packaging, formerly named famil- and inflammatory myopathies.17
ial infantile myasthenia. Routine repetitive
stimulation may be normal, and prolonged
exercise or prolonged repetitive stimulation SUMMARY
at 10 Hz for 5-10 minutes may be required
to demonstrate the abnormalities. The electrophysiologic technique of repeti-
In myotonic disorders, repetitive stimula- tive stimulation is an important component
tion may produce a small decrement at rest of the evaluation of patients with fatigue,
280 Motor Pathways
Figure 22-10. Results of the prolonged exercise test in a normal subject (top) and a patient with hypokalemic pe-
riodic paralysis (bottom). The subject is studied by recording the compound muscle action potential amplitude with
single stimuli at 1-minute intervals while the patient is at rest, during 3-5 minutes of exercise, and for 20-40 min-
utes after exercise. In hypokalemic periodic paralysis, the amplitude increases 39% during exercise and then de-
creases dramatically by 62% after exercise.
weakness, ptosis, diplopia, dysphagia, and 2. Keesey JC. AAEE Minimonograph #33: Electrodi-
dysarthria. The technique requires a knowl- agnostic approach to defects of neuromuscular
transmission. Muscle Nerve 12;613-626, 1989.
edge of the physiology and pathophysiology 3. Sanders DB. Electrophysiologic study of disorders
of neuromuscular transmission and the ba- of neuromuscular transmission. In Aminoff MJ
sics of nerve conduction studies for proper (ed). Electrodiagnosis in Clinical Neurology, 3rd
application and interpretation. Errors in ed. Churchill Livingstone, New York, 1992, pp
327-354.
technique can produce falsely positive or 4. Harper C Jr. Electrodiagnosis of endplate disease.
negative results. The test must be individu- In Engel AG (ed). Myasthenia Gravis and Myas-
alized for each case, with the proper selec- thenic Disorders. Oxford University Press, New
tion of nerve and muscle combinations, rates York, 1999, pp 65-84.
of stimulation, and types of exercise. The ex- 5. Zifko UA, Nicolle MW, Grisold W, Bolton CF.
Repetitive phrenic nerve stimulation in myasthenia
aminer must be aware of the varied abnor- gravis. Neurology 53:1083-1087, 1999.
malities that occur with different disease 6. Sanders DB (ed). Myasthenia gravis and myasthenic
entities. syndromes. Neurol Clin 12:231-442, May 1994.
7. Newsom-Davis J, Lang B. The Lambert-Eaton myas-
thenic syndrome. In Engel AG (ed). Myasthenia
Gravis and Myasthenic Disorders. Oxford Univer-
sity Press, New York, 1999, pp 205-228.
REFERENCES 8. PickettJB III. AAEE case report #16: Botulism. Mus-
cle Nerve 11:1201-1205, 1988.
1. Engel AG. Anatomy and molecular architecture of 9. Sheth RD, Lotz BP, Hecox KE, Waclawik Aj. In-
the neuromuscular junction. In Engel AG (ed). fantile botulism: pitfalls in electrodiagnosis. J Child
Myasthenia Gravis and Myasthenic Disorders. Ox- Neurol 14:156-158, 1999.
ford University Press, New York, 1999, pp 3-39. 10. Chaudhry V, Crawford TO. Stimulation single-fiber
EMG in infant botulism. Muscle Nerve 22:1698- 14. Vedanarayanan W. Congenital myasthenic syn-
1703, 1999. dromes. Neurologist 6:186-196, 2000.
11. Hatanaka T, Owa K, Yasunaga M, et al. Electro- 15. Sander HW, Scelsa SN, Conigliari MF, Chokroverty
physiological studies of a child with presumed bot- S. The short exercise test is normal in proximal my-
ulism. Childs Nerv Syst 16:84-86, 2000. otonic myopathy. Clin Neurophysiol 111:362-366,
12. Engel AG, Ohno K, Sine SM. Congenital myas- 2000.
thenic syndromes. In Engel AG (ed), Myasthenia 16. McManis PG, Lambert EH, Daube JR. The exercise
Gravis and Myasthenic Disorders. Oxford Univer- test in periodic paralysis (abstract). Muscle Nerve
sity Press, New York, 1999, pp 251-297. 7:579, 1984.
13. Jones HR, Bolton CF, Harper CM Jr. Pediatric Clin- 17. Rubin DI, Hermann RC. Electrophysiologic find-
ical Electromyography. Lippincott-Raven, Philadel- ings in amyloid myopathy. Muscle Nerve 22:355-
phia, 1996, pp 353-385. 359, 1999.
Chapter 23
MOTOR EVOKED POTENTIALS
Jasper R. Daube
MOTOR EVOKED POTENTIAL Bony Spine Disease

STIMULATION Stroke and Trauma
Technical Aspects of Electric Stimulation Peripheral Nerve Disease
Technical Aspects of Magnetic Stimulation Other Disorders
MOTOR EVOKED POTENTIAL Complications
RECORDING SUMMARY
MOTOR EVOKED POTENTIAL VARIABLES
APPLICATIONS
Demyelinating Disease
The term motor evoked potentials (MEPs) has cations. The clinical value of MEPs has been
been used to describe potentials from mus- tested in many disorders,1 but their impor-
cle or nerve elicited by several methods of tant clinical applications parallel the appli-
stimulation. Strictly, stimulating a peripheral cation of somatosensory evoked potentials.
nerve and recording the muscle action po- There is a high correlation between clinical
tential is a form of MEP but is referred to findings and MEPs, with additional subclin-
as a compound muscle action potential (CMAP) ical abnormalities found in 10%-26% of pa-
(see Chapter 21). However, CMAPs recorded tients.2 The major value of both MEPs and
from muscles after stimulation of motor somatosensory evoked potentials has been in
structures in the central nervous system are 1) identifying slowing of impulse conduction
referred to as motor evoked potentials. Motor in demyelinating diseases, particularly mul-
evoked potentials may be elicited by stimu- tiple sclerosis, and 2) monitoring central
lating the spinal cord or the cerebral hemi- nervous system pathways intraoperatively.
spheres with electric or magnetic stimuli. The former is the more common clinical di-
For many years, surgeons have identified the agnostic application. Intraoperative moni-
motor cortex intraoperatively with localized, toring with MEPs has become widely used as
direct, electric stimulation of the exposed the reliability, validity, and range of appli-
cerebral cortex which elicits MEPs. Clinical cation of the method have become known
and experimental studies of the efficacy and (see Chapter 43).
safety of transcranial (across the skull) electric
and magnetic MEPs have been performed
extensively in other countries, including
Canada. Clinical applications of transcranial MOTOR EVOKED POTENTIAL
MEPs are now permitted in all countries ex- STIMULATION
cept the United States, where the method is
still considered experimental. Theoretically, MEPs can be obtained by
Motor evoked potential testing involves stimulating anywhere along motor pathways.
stimulating and recording from various lo- For clinical purposes, MEPs are elicited with
282
Motor Evoked Potentials 283
stimulation of either the cerebral cortex or likelihood of activating action potentials de-
the cervical spinal cord.3 At each location, scending in the corticospinal tracts to ante-
motor pathways can be activated by electric rior horn cells (motor neurons). Anterior
or magnetic stimulation. Although the tech- horn cells are activated by descending im-
nical aspects of electric and magnetic stim- pulses in the motor pathways occurring at 3-
ulation differ, the physiology of MEP activa- to 5-ms intervals. Activation of the anterior
tion is similar for both forms of stimulation. horn motor neurons is more likely when
Direct electric activation of the motor multiple depolarizing action potentials ar-
pathways at the level of the cerebral cortex rive at the neuron within a few milliseconds
has been used in experimental animals for of one another. Microelectrode recordings
many years to study the motor pathways. Pen- of anterior horn cell depolarization in re-
field conducted the first extensive study of sponse to cortical stimulation show an initial
stimulation of the motor cortex in humans large wave referred to as a direct or D wave,
more than 50 years ago during surgical pro- followed by a series of three to eight smaller
cedures for epilepsy.* He noted that the re- depolarizations occurring at 3- to 5-ms in-
sponses were attenuated substantially by tervals and referred to as indirect or I waves.
anesthesia and so conducted many of his op- The number of I waves increases with repet-
erations with patients under local or light itive stimulation.
anesthesia. He also recognized that the cere- For diagnostic purposes, a series of 3-10
bral cortex was activated more readily with stimuli to the brain would be expected to
rapid repetitive stimulation, ranging from produce optimal activation of MEPs. Newly
20 to 50 Hz and that optimal activation oc- designed transcranial, electric and mag-
curred with the anode, rather than the cath- netic, stimulators provide repetitive stimula-
ode, placed over the motor area. Thus, stim- tion that enhances MEPs. Comparable en-
ulation of the motor cortex differs from that hancement of MEPs also can be obtained by
of peripheral nerves or motor fiber tracts in low-level voluntary activation, which partially
the spinal cord in both the polarity of stim- depolarizes the cell body and axon hillock
ulation and the rate of repetitive stimula- before the stimulus is applied (Fig. 23-1).
tion. On electromyographic (EMG) recordings
Direct stimulation of the cerebral cortex from the muscles used to record MEPs, 15%
may activate the dendrites, cell bodies, or the activation has been found to give optimal fa-
axon hillocks of the motor neurons in the cilitation of MEPs. Other methods have been
precentral gyrus. Motor evoked potentials suggested to facilitate MEPs in disorders in
depend on activation at the axon hillocks in which the patient is unable to voluntarily
the depths of the cerebral cortex. Motor contract the muscle.1'5'6
evoked potential activation is obtained by
electric stimulation with the anode over the
region of the neuron and the cathode at a
distal site. This orientation enhances hyper- Technical Aspects of
polarization at the cortical surface and de- Electric Stimulation
polarization in the deep layers of the corti-
cal gray matter. With magnetic stimulation, In some cases of intraoperative MEP moni-
an intense current in an external coil in- toring, electrodes are placed directly on the
duces local depolarizing electric currents dura mater over the cerebral cortex or spinal
that flow through the neuron and axon cord (see Chapter 43). However, when MEPs
hillock; this depolarization initiates de- are used as a diagnostic test and in most in-
scending action potentials in the corti- stances when they are used for intraopera-
cospinal pathway. tive monitoring, transcranial electric stimu-
Motor evoked potential augmentation lation of the motor cortex through the intact
with repetitive stimulation occurs at the cor- bony skull or percutaneous stimulation of
tical and spinal cord levels. Repetitive acti- the spinal cord is used. The high resistance
vation of cortical motor neurons drives the of bone to electric current requires much
membrane potentials of these cells closer to higher applied voltages to drive the few mil-
threshold for firing, thereby increasing the liamps of current to the cerebral cortex
284 Motor Pathways
Figure 23-1. Enhancement of forearm muscle, near threshold motor evoked potentials by "thinking to move" the
wrist joint. Four or five potentials are superimposed. (From Rossini and Rossi.1 By permission of Elsevier Science
Ireland.)
needed for cortical stimulation. Transcra- a flat electrode placed directly on the dura
nial electric stimulation in normal adults mater in the surgical field or a needle elec-
often requires stimuli of 1000-1500 V. The trode placed either in the interspinous liga-
stimuli are applied in short pulses, with du- ment or over the vertebral lamina. Anode
rations of less than half a millisecond. The placement for activation of the spinal cord
intensity of the stimulus produces consider- may be epidural, laminar, esophageal, sub-
able contraction of the cranial musculature. cutaneous, or on the surface of the skin at
Some patients report it to be uncomfortable, some distance. Stimulation of the cervical
but most tolerate it without difficulty. The spinal cord is enhanced by repetitive stimu-
pulse duration and configuration of the elec- lation at intervals of 3—5 ms. In surgical set-
tric stimulus appear to be less important for tings, activation of motor pathways in the
activation at the cortical level than they are cervical spinal cord is easier than with corti-
for peripheral stimulation. Because of the cal stimulation because anesthetics have less
high voltages required, MEPs typically are effect on the spinal cord than on the cere-
obtained with special equipment that gen- bral cortex. However, volatile anesthetics do
erates the high voltages without causing risk suppress transsynaptic activity in the anterior
to the patient or staff. horn cells, thereby partially reducing MEPs
Electric activation of the motor pathways with spinal cord stimulation, but to a lesser
in the spinal cord occurs by depolarization extent than with cortical stimulation.
of the axons at the cathode, with hyperpo-
larization at the anode. Therefore, activation
of a spinal cord pathway is similar to that of
a peripheral nerve. Electric stimulation at Technical Aspects of
the spinal cord level relies on depolarization Magnetic Stimulation
of the motor axons in the corticospinal tract.
This can be obtained percutaneously with Activation of the motor pathways with mag-
stimuli greater than 1000 V, similar to the netic stimulation generally is more difficult
electric stimuli used in transcranial stimula- and less successful than it is with electric
tion. However, in the operating room, axons stimulation. However, magnetic stimulation
in the spinal cord can be activated with lower is tolerated better than electric stimulation
intensity stimuli because the cathode can be by patients who are awake, because the pa-
tient does not feel a local electric shock sen- The selection of the muscles for recording
sation. Activation of either the axons in the is determined by the clinical problem. Mo-
corticospinal tract at the level of the cervical tor evoked potentials elicited by stimulation
cord or the neurons in the motor cortex de- of the cerebral cortex or the cervical spinal
pends on the same mechanism. In both cord may be recorded from any of the limb
cases, an intense pulse of current in an ex- muscles.
ternal coil induces a magnetic field that tra- Typically, the potentials are recorded with
verses the skin and bone without effect. A surface electrodes, but subcutaneous elec-
sudden change in the magnetic field induces trodes, intramuscular wires, or other elec-
a local electric current in brain tissue that trodes can be used. The standard gains, fil-
can cause neuronal or axonal depolariza- ter settings, and sweep speeds that are used
tion; this depolarization initiates descending for peripheral motor conduction studies are
action potentials in the corticospinal path- usually satisfactory, although in some in-
way. Magnetic stimulation differs from elec- stances of intraoperative MEP monitoring
tric stimulation in that no current flow oc- the amplitudes are small enough to require
curs in the superficial levels of the skin, averaging.
muscle, or bone. All current flow is induced Commonly, surface electrodes placed over
intracranially by the rapid change in current the major muscle groups, including the ex-
in the stimulating coil. An increase or de- tensors and flexors of the ankle, knee, wrist,
crease of current or a change in direction of and elbow, provide satisfactory measures of
current always establishes a magnetic field central motor conduction deficits. Typically,
around the wire through which current is the ankle extensor and flexor muscles and
flowing. In transcranial magnetic stimula- the quadriceps muscles are tested in the
tion, the electrically induced pulse of the lower extremity, and the intrinsic hand mus-
magnetic field induces current flow in the cles, forearm extensors, and arm flexors are
underlying tissue.7 When this current flow tested in the upper extremities. The thenar
reaches threshold, it activates either a nearby muscle in the hand and the anterior tibial
neuronal cell body or motor axon. These in muscle in the leg are the sites most fre-
turn generate action potentials that travel to quently used and for which the most defin-
the anterior horn cells to elicit MEPs. itive normal data are available. Attempts at
With optimal recording and no interfer- recording MEPs from muscles innervated by
ence, MEPs can initiate supramaximal or cranial nerves have been limited to record-
nearly supramaximal CMAPs. However, in ings from facial muscles, with stimulation of
many settings, magnetic stimulation is not the facial nerve.
able to evoke supramaximal CMAPs. At any
given site, current flow is greater and acti-
vation is obtained more readily and more ef-
fectively with electric than with magnetic MOTOR EVOKED POTENTIAL
stimulation. VARIABLES
Motor evoked potentials can be measured
with the same variables used for measuring
MOTOR EVOKED POTENTIAL peripherally evoked CMAPs, that is, latency
RECORDING and amplitude. However, MEPs are more vari-
able in size and configuration than CMAPs
Stimulation of the motor pathways at either evoked with peripheral stimulation. They
the cortical or spinal cord level activates can vary with the attention and level of re-
multiple descending pathways and, thus, laxation of the patient. Thus, amplitude
produces widespread muscle contraction. measures are of little clinical diagnostic
Recording electrodes can be placed over value. However, major changes in amplitude
multiple muscles or muscles of particular can be used to detect motor pathway dam-
clinical concern. Selection of the location of age during surgical monitoring.
the recording electrode defines which com- The latency of responses from the time of
ponents of the motor pathways are tested. stimulation to recording is the most com-
286 Motor Pathways
mon measurement. The most direct and re- the techniques has not been defined suffi-
liable method of measuring MEP latency is ciently to allow adoption. Also, normal data
by direct measurements that are compared have been obtained in only a few laborato-
with those of normal subjects matched for ries (and for specific stimulation and record-
age, height, and sex; however, according to ing methods that have not been widely
some reports, this has not been sufficiently adopted).9 Furthermore, variability of the
sensitive for identifying mild involvement by response—with and without voluntary mus-
disease. cle contraction—has raised concern about
Because MEPs traverse the peripheral seg- the reliability of the responses for clinical in-
ment of the motor pathway before being terpretation.10 Although some reports in-
recorded, several methods have been devel- clude normative data, most do not. Conse-
oped to distinguish conduction over the cen- quently, generally accepted values for MEP
tral portion of the motor pathways from that amplitude and latencies do not exist. For
over the peripheral portion. Central conduc- clinical studies, amplitude measures gener-
tion time is a measure designed to better iden- ally are not used because of the marked dif-
tify central disorders.8 Because MEPs are de- ferences among normal subjects. Latency
signed primarily to detect disease of the measures can show marked slowing or dis-
central motor pathway, central conduction persion of the response. Such a finding is
time is of particular interest. It is calculated strong evidence for a demyelinating process,
by subtracting the time needed for the sig- as in multiple sclerosis. Finally, although
nal to travel over the peripheral segment MEPs have been shown to have clinical value
from the spinal cord to the muscle from the in patients with multiple sclerosis, magnetic
total latency of the MEP from the site of stim- resonance imaging and other laboratory
ulation to the muscle. The latency of the pe- testing have been found to be more specific.
ripheral segment may be obtained by evok- Motor evoked potentials have been tested
ing an MEP at the level of the spine and and reported on in many neurologic disor-
measuring latency at that point. Peripheral ders, including cerebral infarcts, Parkinson's
latency can also be obtained indirectly from disease, other movement disorders, motor
F-wave measurements made during standard neuron disease, cervical spondylosis, and less
nerve conduction studies. Because the F common disorders.1'11'12
wave traverses the peripheral motor pathway
from the anterior horn cell to the muscle, F-
wave latency minus the distal latency at the Demyelinating Disease
site of stimulation divided by 2 gives the time
from the spinal cord to the distal site of stim- In clinical diagnostic neurology, the most
ulation (see Chapter 21). The distal latency common use of MEPs is the search for evi-
is added to this to obtain the spinal latency dence of demyelinating disease.13 Motor
for the MEP. The spinal latency subtracted evoked potentials show marked slowing and
from the latency obtained with cortical stim- dispersion in the presence of demyelinating
ulation gives the central conduction time. disease in the spinal cord or brain stem,
Thresholds for activation of an MEP have particularly in transverse myelitis14'15 (Fig.
been used widely in physiologic studies of 23—2). Such changes may be present even in
cortical function, but they have not proved the absence of any major or clear-cut neu-
of value in clinical assessment. rologic deficit. Therefore, MEPs can be used
to identify subclinical lesions in patients widi
multiple sclerosis and help define the sig-
nificance of fatigue.16 Many reports have
APPLICATIONS shown the sensitivity of MEPs in detecting
and characterizing slowing of conduction in
Motor evoked potentials are not widely used the motor pathways that may not produce a
clinically because of several restricting fac- measurable motor deficit. Virtually all pa-
tors. The equipment needed for either mag- tients with multiple sclerosis who have weak-
netic or transcranial electric stimulation has ness in a limb also have abnormal MEPs in
not been approved for clinical use in the that limb. Fifty percent of patients with mul-
United States, mainly because the safety of tiple sclerosis who do not have a neurologic
Figure 23-2. Hand motor evoked potentials (MEPs) in a normal subject (left) and in a patient with documented
multiple sclerosis (right). Function and clinical testing were normal in the hand and arm. ADM, abductor digiti
minimi of the hand. Black triangles, MEP latencies. (From Hess CW, Mills KR, Murray NM, Schriefer TN. Magnetic
brain stimulation: central motor conduction studies in multiple sclerosis. Ann Neurol 22:744—752, 1987. By per-
mission of the American Neurological Association.)
deficit in a limb nonetheless have abnormal to evaluate different stroke populations, but
findings on MEP testing. no clear value has been defined to warrant
the general application of MEP testing in
stroke patients.19'20 Motor evoked potential
Bony Spine Disease testing has been applied also in studies of re-
covery after severe traumatic injury to the
Motor evoked potentials are frequently ab- brain or spinal cord and in acute brain stem
normal in patients with cervical or lumbar lesions and may be helpful in predicting
sjpondylosis.17 The changes may be a pro- outcome.21"23
Icjngation of latency in that segment of the
recording or a change in configuration be-
tween sites of recording. In cervical spondy- Peripheral Nerve Disease
losis, abnormality on MEP testing may be
caused by bony compression of the spinal Peripheral segments of the motor pathways
cord or the spinal nerves in the interverte- are included in standard MEP latency and
bral foramina. The abnormalities are in- amplitude measurements; therefore, MEPs
creased latency and decreased amplitude.18 are often abnormal in cases of peripheral
Generally, a change in latency is more neuropathy.24 Technical problems and the
prominent than a reduction in amplitude. lack of reliability associated with MEP test-
Amplitude reduction is particularly difficult ing make it less satisfactory than standard
to assess in MEPs, because it is difficult to nerve conduction studies for assessing pe-
obtain a supramaximal response in many ripheral neuropathies. However, in some pa-
normal subjects. Assessment of amplitude tients with primarily proximal involvement,
depends on the ability to evoke a supra- as in acute or chronic demyelinating poly-
maximal response. This difficulty is particu- radiculopathy, MEP testing may demon-
larly true for the lower extremities, and it strate slowing of conduction at the level of
may also be true of testing the upper ex- the spinal nerve and nerve root when it is
tremities of elderly subjects. not readily apparent on standard nerve con-
duction studies.24
Stroke and Trauma
Other Disorders
Motor evoked potentials have been studied
as a predictor of outcome in stroke.1 Many Case reports or reports of a small number of
variations in MEP methods have been used patients with other disorders who have had
288 Motor Pathways
MEP testing are available, but their clinical REFERENCES

significance is difficult to determine. In amy-
otrophic lateral sclerosis, MEP testing is re- 1. Rossini PM, Rossi S. Clinical applications of motor
ported to show abnormalities of latency and evoked potentials. Electroencephalogr Clin Neuro-
threshold, but these are difficult to distin- physiol 106:180-194, 1998.
guish from the changes that would be ex- 2. Di Lazzaro V, Oliviero A, Profice P, et al. The di-
agnostic value of motor evoked potentials. Clin
pected solely from loss of amplitude and loss Neurophysiol 110:1297-1307, 1999.
of facilitation and, thus, MEP testing is un- 3. Rossini PM. Methodological and physiological as-
likely to add to the standard EMG evaluation pects of motor evoked potentials. Electroencepha-
despite some reported abnormalities.25'26 logr Clin Neurophysiol Suppl 41:124-133, 1990.
Motor evoked potential testing also has been 4. Penfield W, Jasper H. Epilepsy and the Functional
Anatomy of the Human Brain. Little, Brown,
applied extensively in analyzing the under- Boston, 1954.
lying pathophysiology of such disorders as 5. Magistris MR, Rosier KM, Truffert A, Myers JP.
Parkinson's disease, chorea, and myoclonus. Transcranial stimulation excites virtually all motor
Motor evoked potentials have helped define neurons supplying the target muscle. A demon-
stration and a method improving the study of mo-
the localization of motor cortex in humans.27 tor evoked potentials. Brain 121:437-450, 1998.
Attempts to apply measurements of thresh- 6. Han TR, Kim JH, Lim JY. Optimization of facilita-
old to MEPs have not been successful in dis- tion related to threshold in transcranial magnetic
tinguishing specific clinical disorders. stimulation. Clin Neurophysiol 112:593-599, 2001.
7. Counter SA, Borg E. Analysis of the coil generated
impulse noise in extracranial magnetic stimulation.
Electroencephalogr Clin Neurophysiol 85:280-288,
Complications 1992.
8. Glaus D. Central motor conduction: method and
Reports of complications of MEP testing are normal results. Muscle Nerve 13:1125-1232, 1990.
9. Eisen A, Siejka S, Schulzer M, Calne D. Age-de-
few. Most studies that have specifically as- pendent decline in motor evoked potential (MEP)
sessed safety have not found side effects,28 amplitude: with a comment on changes in Parkin-
although a magnetic pulse applied near the son's disease. Electroencephalogr Clin Neurophys-
ear has been reported to damage the hear- iol 81:209-215, 1991.
ing mechanism. Few studies have assessed 10. Triggs WJ, Kiers L, Cros D, FangJ, Chiappa KH. Fa-
cilitation of magnetic motor evoked potentials dur-
MEPs in children.29 Motor evoked potential ing the cortical stimulation silent period. Neurol-
thresholds are altered in epilepsy and by ogy 43:2615-2620, 1993.
antiepileptic drugs, but seizures caused by 11. Murray NM. Magnetic stimulation of cortex: clini-
diagnostic MEP testing have not been re- cal applications. J Clin Neurophysiol 8:66-76, 1991.
12. PillaiJJ, Markind S, Streletz LJ, Field HL, Herbison
ported,30'31 although the initiation of an G. Motor evoked potentials in psychogenic paraly-
epileptic discharge after repeated MEP test- sis. Neurology 42:935-936, 1992.
ing in an epileptic patient has been de- 13. Jones SM, Streletz LJ, Raab VE, Knobler RL, Lublin
scribed. No evidence has been found that FD. Lower extremity motor evoked potentials in
MEP testing causes increased frequency of multiple sclerosis. Arch Neurol 48:944-948, 1991.
14. Ho KH, Lee M, Nithi K, Palace J, Mills K. Changes
seizures in patients with epilepsy. in motor evoked potentials to short-interval paired
transcranial magnetic stimuli in multiple sclerosis.
SUMMARY 15. Kalita J, Misra UK. Neurophysiological studies in
acute transverse myelitis. J Neurol 247:943-948,2000.
16. Colombo B, Martinelli Boneschi F, Rossi P, et al.
Motor evoked potential testing is difficult to MRI and motor evoked potential findings in
perform reliably, but it is a direct method of nondisabled multiple sclerosis patients with and
assessing descending motor pathways. Both without symptoms of fatigue. J Neurol 247:506-509,
electric and magnetic stimulation can elicit 2000.
17. Travlos A, Pant B, Eisen A. Transcranial magnetic
MEPs, but in the United States this method stimulation for detection of preclinical cervical
can only be applied clinically to peripheral spondylotic myelopathy. Arch Phys Med Rehabil
nerves. In Europe, MEP testing has proved 73:442-446, 1992.
most useful in identifying subclinical de- 18. Tavy DL, Franssen H, Keunen RW, Wattendorff AR,
Hekster RE, Van Huffelen AC. Motor and somato-
myelination in multiple sclerosis and in sensory evoked potentials in asymptomatic spondy-
monitoring motor pathways intraopera- lotic cord compression. Muscle Nerve 22:628-634,
tively. 1999.
19. Pennisi G, Rapisarda G, Bella R, et al. Absence of Electroencephalogr Clin Neurophysiol 101:32-41,
response to early transcranial magnetic stimulation 1996.
in ischemic stroke patients: prognostic value for 26. Truffert A, Rosier KM, Magistris MR. Amyotrophic
hand motor recovery. Stroke 30:2666-2670, 1999. lateral sclerosis versus cervical spondylotic myelopa-
20. Feys H, Van Hees J, Bruyninckx F, Mercelis R, De thy: a study using transcranial magnetic stimulation
Weerdt W. Value of somatosensory and motor with recordings from the trapezius and limb mus-
evoked potentials in predicting arm recovery after cles. Clin Neurophysiol 111:1031-1038, 2000.
a stroke. J Neurol Neurosurg Psychiatry 68:323—331, 27. Abbruzzese G, Marchese R, Trompetto C. Sensory
2000. and motor evoked potentials in multiple system at-
21. Moosavi SH, Ellaway PH, Catley M, Stokes MJ, rophy: a comparative study with Parkinson's dis-
Haque N. Corticospinal function in severe brain in- ease. Mov Disord 12:315-321, 1997.
jury assessed using magnetic stimulation of the mo- 28. Jahanshahi M, Ridding MC, Limousin P, et al.
tor cortex in man. J Neurol Sci 164:179-186, 1999. Rapid rate transcranial magnetic stimulation—a
22. Curt A, Keck ME, Dietz V. Functional outcome fol- safety study. Electroencephalogr Clin Neurophysiol
lowing spinal cord injury: significance of motor- 105:422-429, 1997.
evoked potentials and ASIA scores. Arch Phys Med 29. Nezu A, Kimura S, Uehara S, Kobayashi T, Tanaka
Rehabil 79:81-86, 1998. M, Saito K Magnetic stimulation of motor cortex
23. Schwarz S, Hacke W, Schwab S. Magnetic evoked in children: maturity of corticospinal pathway and
potentials in neurocritical care patients with acute problem of clinical application. Brain Dev 19:176-
brain stem lesions. J Neurol Sci 172:30-37, 2000. 180, 1997.
24. Takada H, Ravnborg M. Magnetically evoked mo- 30. Caramia MD, Gigli G, lani C, et al. Distinguishing
tor potentials in demyelinating and axonal polyneu- forms of generalized epilepsy using magnetic brain
ropathy: a comparative study. Eur J Neurol 7:63-69, stimulation. Electroencephalogr Clin Neurophysiol
2000. 98:14-19, 1996.
25. Kohara N, Kaji R, Kojima Y, et al. Abnormal ex- 31. Nezu A, Kimura S, Ohtsuki N, Tanaka M. Trans-
citability of the corticospinal pathway in patients cranial magnetic stimulation in benign childhood
with amyotrophic lateral sclerosis: a single motor epilepsy with centre-temporal spikes. Brain Dev
unit study using transcranial magnetic stimulation. 19:134-137, 1997.
SECTION 2
Electrophysiologic
Assessment of
Neural Function
Part D
Assessing the
Motor Unit
Contraction of somatic muscles underlies all human movement. Therefore,

motor pathways that control movement must act on muscle. Motor units
form the connection between motor pathways in the central nervous system
and the muscles. For this reason, the motor neuron has been called the
final common pathway. Lower motor neurons are the connection between the
central motor pathways and the muscles.
Each motor unit consists of a cell body in the central nervous system, an
axon in the peripheral nervous system, and all the muscle fibers innervated
by that axon. The cell body is that of a motor neuron located in a brain
stem cranial nerve motor nucleus or the anterior horn of the spinal cord.
The peripheral axon is a myelinated fiber that travels in a cranial nerve or
peripheral nerve to the muscle, where the axon branches into multiple nerve
terminals. The number of terminal branches determines the innervation ra-
tio (the number of muscle fibers in a motor unit). Innervation ratios are
as small as 50 (in extraocular muscles and other small muscles requiring
fine control) and as large as 2000 (in large, powerful muscles such as the
gastrocnemius).
Most peripheral neuromuscular diseases involve one or more components
of a motor unit. Primary myopathies and disorders such as myasthenia gravis
affect the muscles and neuromuscular junctions. Mononeuropathies, pe-
ripheral neuropathies, and radiculopathies involve the peripheral axons,
whereas motor neuron disease, such as amyotrophic lateral sclerosis, involves
the anterior horn cells.
Nerve conduction studies measure alterations in function of the periph-
eral motor axons, but only indirectly measure the extent of axonal de-
struction or anterior horn cell loss. The electrophysiologic assessments de-
292 Assessing the Motor Unit
scribed in the three chapters in this section 'complement nerve conduction

studies in defining the character, severity, and distribution of neuromuscu-
lar disease. To distinguish among primary muscle disease, disorders of the
neuromuscular junction, and neurogenic disorders often depends on nee-
dle electromyography (Chapter 24). In addition, needle electromyography
helps characterize the defects of neuromuscular transmission and the num-
ber of functioning axons or anterior horn cells. These two aspects of neu-
romuscular disease can be quantified more precisely with the special tech-
niques of single fiber electromyography and motor unit number estimate
(MUNE); in this edition, the discussion of these techniques has been ex-
panded in Chapters 25 and 26. Expanded experience with the methods of
MUNE and their reproducibility and application has allowed better quan-
tification of the number of motor units in a muscle or in a group of mus-
cles. Thus, MUNE can provide critical information about the severity and
progression of a neurogenic process (Chapter 27).
Chapter 24
ASSESSING THE MOTOR UNIT WITH
NEEDLE ELECTROMYOGRAPHY
Jasper R. Daube
KNOWLEDGE BASE OF NEEDLE ABNORMAL ELECTRIC ACTIVITY

ELECTROMYOGRAPHY Insertional Activity
CONDUCTING THE NEEDLE Fibrillation Potentials
EXAMINATION Myotonic Discharges
Preparing the Patient Complex Repetitive Discharges
Selecting the Muscles to Test Fasciculation Potentials
NEEDLE EXAMINATION TECHNIQUES Myokymic Discharges
Electrode Types Neurotonic Discharges (Neuromyotonia)
Needle Insertion Cramp Potentials
Needle Movement Short-Duration Motor Unit Potentials
Data Collection Long-Duration Motor Unit Potentials
Special Problems Polyphasic Motor Unit Potentials
ELECTROMYOGRAPHIC ANALYSIS Mixed Patterns: Long-Duration and Short-
Pattern Recognition Duration Motor Unit Potentials
Semiquantitative Electromyography Motor Unit Potential Variation
ORIGIN OF ELECTROMYOGRAPHIC Doublets (Multiplets)
POTENTIALS Abnormal Recruitment
NORMAL ELECTROMYOGRAPHIC Disorders of Central Control
ACTIVITY Synkinesis
Pattern of Firing Patterns of Abnormality
Spontaneous Activity SUMMARY
Voluntary Activity
RECRUITMENT
SIGNAL ANALYSIS
Measurement of Motor Unit Potentials
Recording Technique
Needle electromyography (EMG) is a major gist. The skills of needle EMG are a combi-
diagnostic tool for identifying and charac- nation of auditory pattern recognition and
terizing disorders of the motor unit, includ- semiquantitation, and they require time and
ing anterior horn cells, peripheral nerves, diligence to acquire. Although these skills
neuromuscular junctions, and muscles. Nee- can be acquired through experience, they
dle EMG requires a unique combination of are learned more readily if the person un-
knowledge and skills. The knowledge is fa- derstands them and is determined to learn
miliar to practicing clinicians but requires a them. Many facets of needle EMG depend
detail beyond that of the general neurolo- on the unique analysis and processing capa-
293
bilities of the human brain and have pre- physiology, diseases, techniques, electricity,
cluded the replacement of skilled physician and patient interaction. Conducting a suc-
electromyographers by automated equip- cessful needle examination requires attend-
ment. These skills cannot be learned only ing to several specific issues in a series of
from reading about them; they are learned steps:
best in an apprenticeship. Successful, high- • Clinical evaluation
quality, reproducible EMG results depend • Preparing the patient
on the skills of a clinician practiced in as- • Selecting the muscles to test
sessing clinical problems and on the unique • Special considerations
art and skills of an expert electromyogra- • Performing the examination
pher in analyzing electric signals recorded
from muscle.1'2
CONDUCTING THE NEEDLE
EXAMINATION
KNOWLEDGE BASE OF NEEDLE
ELECTROMYOGRAPHY Preparing the Patient
An initial step in learning EMG is to under- Before the study is performed, most patients
stand the range of information that EMG have received information about the needle
can provide to extend the clinical evaluation examination and may have a few questions.
of patients with suspected neuromuscular It is still helpful to explain briefly that the
diseases.1 Any one or more of the following needle will be inserted into several muscles
may be provided. The expert clinical elec- and will cause some discomfort. The patient
tromyographer knows how EMG can provide will appreciate knowing approximately how
each of these and keeps each one in mind long the study will take and how many mus-
when performing and reporting the results cles will be examined. Before each needle is
of EMG testing. inserted, indicate to the patient the approx-
• Confirm the diagnosis imate location and alert him or her to an
• Exclude other disease imminent "stick." Wipe the skin over each
• Identify unrecognized (subclinical) dis- puncture site with alcohol before the needle
ease is inserted.
• Localize abnormalities
• Define severity
• Define the pathophysiologic mecha- Selecting the Muscles to Test
nism by pathology or disease
• Define evolution, stage, and prognosis The groups of muscles to be tested are se-
For a clinical electromyographer to pro- lected initially on the basis of the clinical hy-
vide most efficiendy the clinical information potheses, for example, proximal muscles for
needed requires extending the hypothesis myopathy, single limb muscles for radicu-
generation and testing commonly used in lopathies, and widely distributed muscles for
clinical neurology. The electromyographer motor neuron disease. Ideally, the individ-
reviews all the clinical data, confirms and ob- ual muscles selected for examination should
tains additional history as needed, and ex- be superficial, easily palpated, and readily
amines the patient to define the clinical identified. They should be located away
deficit. On the basis of this information, the from major blood vessels, nerve trunks, and
electromyographer generates a set of hy- viscera. Select muscles whose testing will
potheses of possible causes of the clinical cause the least discomfort for the patient.
problem. These determine which electro- For example, testing the thenar or small foot
physiologic tests need to be performed, both muscles often makes patients more uncom-
EMG and nerve conduction studies. fortable than testing other muscles. Hence,
The specific knowledge base required for thenar and small foot muscles should be
performing EMG is broad and includes tested only when the information is not avail-
knowledge of anatomy, physiology, patho- able from other muscles. Because the ap-
Assessing the Motor Unit with Needle Electromyography 295
pearance of motor unit potentials can vary

greatly among different muscles, the mus-
cles selected should be familiar to the ex-
aminer, both how to test the muscle and the
range of normal findings.
NEEDLE EXAMINATION
TECHNIQUES
The ability to record normal and abnormal
electric activity from muscle depends consid-
erably on an electromyographer's knowledge
of and skill with a needle recording elec-
trode. In addition, careful attention to sev-
Figure 24-1. Electrode types used in recording elec-
eral special problems—skin infection or tromyographic signals. A, The standard concentric and,
other condition, bleeding disorder, cardiac D, monopolar electrodes are the ones commonly used
valvular disease, obesity—presented by a few for clinical diagnostic recordings. B, Bipolar concentric
patients must be considered before a needle and, C, single fiber electromyographic needle elec-
trodes are used to isolate potentials.
examination is performed.
Electrode Types surrounding muscle. Although these elec-

trodes were expensive, inexpensive dispos-
Various types of electrodes have been used able models are now available. The common
to record the electric activity of normal and sizes available are 25 mm (26 gauge), 50 mm
diseased muscles. Surface electrodes can de- (26 gauge), and 75 mm (20 gauge). The nee-
pict the extent of EMG activity and measure dle is a detachable electrode connected to
the conduction velocity in muscle fibers, but the preamplifier by a cable. Because of the
the electric signals are distorted by inter- narrow gauge, electrodes are particularly
vening skin, subcutaneous tissue, and other delicate and need to be handled carefully.
muscles. Recent studies with high-density They are most fragile at the junction of the
recordings have shown some unique physi- shaft and hub and may bend or break at this
ologic findings that suggest surface record- location. This electrode type has several ad-
ings may have a role in the future.3 Needle vantages: (1) its ability to record EMG activ-
electrodes inserted into the muscle can de- ity with a minimum of interference from
pict the electric signals accurately, but de- surrounding muscles, (2) its fixed-size
pending on needle type, these electrodes recording surface, (3) the absence of a sep-
record from different numbers of muscle arate reference electrode, and (4) the ex-
fibers and from muscle fibers in different tensively defined quantitation of the sizes of
locations4 (Fig. 24-1). Standard concentric normal motor unit potentials for various
and monopolar electrodes are the ones most ages and muscles.
commonly used for diagnostic clinical EMG.
MONOPOLAR ELECTRODES
STANDARD CONCENTRIC
ELECTRODES A Teflon-coated fine needle electrode, usu-
ally made of stainless steel, can have a very
A bare, 24- to 26-gauge hollow needle with fine gauge and an extremely sharp point.
a fine wire down the center is beveled at the Monopolar electrodes consist of a solid 22-
tip to expose an active, oval recording sur- gauge to 30-gauge needle with a bare tip ap-
face of 125 fj,m by 580 fj,m. The electrode proximately 500 )Ltm in diameter. These elec-
is referenced to the shaft of the needle, trodes record essentially the same activity
thereby canceling unwanted activity from as recorded with standard concentric elec-
trodes, but motor unit potentials are slightly a needle electrode referenced to a surface
longer in duration and have a higher am- electrode. The macroelectrode records from
plitude.5 Monopolar electrodes are pre- a large number of muscle fibers of multiple
ferred by some electromyographers because motor units in a cylinder along the shaft of
they are less expensive and, at times, less un- the needle. This recording summates the ac-
comfortable for patients. tivity of many motor unit potentials, which
cannot be differentiated from one another.7
SINGLE FIBER ELECTRODES
The potential from a single motor unit is iso-
lated with the help of simultaneous record-
Recordings made with electrodes with small ing of potentials from single muscle fibers
(25 /Ltm) recording surfaces referenced to with a 25 pirn-diameter electrode halfway
the shaft of the needle, with filtering of the along the shaft of the macroelectrode on a
low-frequency components, focus on a small second channel. The second channel is used
number of muscle fibers in the immediate to identify the firing pattern of a single mo-
vicinity of the electrode (see Chapter 26). tor unit. The electric activity recorded from
Single fiber EMG needles record from small the macroelectrode at the time of the firing
areas of muscle and cannot be used to char- of a single fiber potential on the small elec-
acterize the size of motor unit potentials. trode is averaged over multiple discharges.
This method has been used primarily in This results in an averaged potential from all
studying disorders of neuromuscular trans- muscle fibers along the macroelectrode,
mission because it can detect variation in which are innervated by the same motor unit
motor units (jitter between single fiber po- as the single muscle fiber. Thus, the aver-
tentials) not seen with other needle elec- aged potential gives an estimate of the ac-
trodes. Single fiber EMG can also be used to tivity in a larger portion of the muscle fibers
quantify the density of muscle fibers in a mo- of the motor unit. Occasionally, macroelec-
tor unit (fiber density}, a measurement closely trode recordings are able to identify changes
related to the percentage of motor unit po- in the whole motor unit that are not appar-
tentials that are polyphasic and the number ent with smaller electrodes.8
of turns on the motor unit potential.
NEEDLE ELECTRODE
BIPOLAR CONCENTRIC
CHARACTERISTICS
ELECTRODES
The selection depends on a number of pa-
A larger area of muscle is sampled with a tient-related and examiner considerations.
bipolar concentric electrode, in which two
Needle electrodes must be sterile, sharp, and
recording surfaces (80 ptm by 320 /im) are
straight. The recording surface must be ab-
side-by-side and insulated from one another solutely clean. A thin, poorly conducting
in the beveled open end of the needle (Fig.
film on the electrode surface can cause a low-
24-1). This electrode has not been used for voltage, irregular, positive waveform, pop-
standard clinical diagnostic studies; thus, ping artifact that can be mistaken for end
standard values for the size and shape of mo- plate noise or positive waves. Electric im-
tor unit potentials obtained with this elec-
pedance should be checked if a break or
trode have not been defined. This type of short is suspected (correct impedance at
electrode has been used to record the firing 60 Hz is 5-20 Mil).
patterns of single motor units at high levels
of force because it is able to isolate individ-
ual motor unit potentials better than other
electrodes. Needle Insertion
MACROELECTRODES
The muscle to be tested should be palpated
during intermittent contraction to localize
A larger needle electrode is the macro nee- its borders. The skin is made taut and pulled
dle, or macroelectrode.6 A macroelectrode re- a short distance over the muscle to reduce
cording is made from 15 mm of the shaft of bleeding. The needle electrode should be
Assessing the Motor Unit with Needle Electrbmyography 297
held firmly in the fingers and inserted tivates a few motor units (low to moderate
smoothly and quickly through the skin into effort). Selective activation of the muscle of
the subcutaneous tissue or superficial layers interest may be needed to determine needle
of the muscle. This minimizes the force nec- position when examining deep muscles,
essary to achieve penetration and distracts muscles that are difficult to palpate, or small
the patient before the skin is punctured. muscles. The steps in testing contracting
muscle include the following:
• Withdraw the needle to a subcutaneous
Needle Movement position before asking for muscle con-
traction.
The muscle is examined by moving the nee- • Position the limb and muscle and initi-
dle along a straight line into the muscle in ate contraction before moving the nee-
short steps (0.5-1 mm). Large movements dle into the muscle. Advance the nee-
are more painful.9 The pace of needle move- dle until encountering motor unit
ment should not be rushed. A brief pause (1 potentials with a rapid rise time and a
second or longer) between each step move- sharp clicking sound.
ment is needed to listen and watch for slow • Position the joint across which the mus-
abnormal activity. The needle is advanced in cle acts to limit the activity of synergis-
5-30 such steps depending on muscle di- tic and adjacent muscles.
ameter. After the diameter of the muscle has • Ask the patient to perform a movement
been traversed, the needle is withdrawn that requires activation only of the mus-
from the muscle—but not from the skin— cle being examined.
and reinserted from a different angle at the • Palpate the contracting muscle as a
same location. Two to four such passes guide to needle movement.
through the muscle are made until an ade-
quate number of sites in the muscle have
been examined. Special Problems
Small muscles are best tested with an oblique
Data Collection needle course through the muscle to
lengthen the needle's path. Deep muscles
The muscle should be examined at multiple and obese patients require a 75, 90, or 120
sites both at rest and during contraction. mm needle. Some muscles, such as the deep
paraspinal muscles, may be difficult to reach,
even in average-sized patients, without a long
RESTING MUSCLE needle. Needles up to 120 mm long should
The resting muscle is tested for spontaneous be available.
activity at a gain of 50 ju,V/cm. When the nee-
dle is well within the muscle, it should not PAIN CONTROL
be moved for several seconds so the exam-
iner can listen for fasciculations. It is not Most patients are able to tolerate the dis-
always easy to obtain muscle relaxation. In comfort of the needle examination without
tense patients or during a painful examina- difficulty, but a few need a special approach.
tion, relaxation can be enhanced by care- Pain minimization requires attention to all
fully positioning the patient, providing ade- interactions with the patient, in particular
quate support and passive manipulation of the techniques of the needle examination
the limb, activating an antagonist, distract- itself.
ing the patient with conversation, and pro-
viding reassurance.
ELECTROMYOGRAPHIC
CONTRACTING MUSCLE ANALYSIS
Contracting muscle is best examined with It would be ideal to have formal, quantita-
the muscle at a level of contraction that active measures of each of the variables of the
potentials that are assessed during needle repetitive discharges, that are unlike any spe-
EMG, as are available for nerve conduction cific one heard before but similar enough
studies. The limitations of current EMG that it can be recognized as a complex repeti-
equipment and the time required to active discharge.
complish such measurements preclude this The skill of pattern recognition allows the
for routine EMG. In fact, a skilled elec- auditory system to do much more than a
tromyographer can achieve close to the ideal computer or visual representation can do in
by applying the well-defined techniques of recognizing and categorizing the wide range
pattern recognition and semiquantitative of EMG signals that occur in normal and dis-
EMG, as described in the following sections. eased muscle.
Pattern Recognition Semiquantitative

Electromyography
A major component of the EMG is auditory
pattern recognition, a skill that most persons Semiquantitative EMG allows a variable of an
have that allows them to recognize the voice EMG potential to be estimated instead of be-
of a friend and to recognize and name the ing measured formally. Success with semi-
enormous range of sounds in the environ- quantitative EMG depends on taking the
ment. Only a limited number of automated time to learn the methods and then apply-
systems have been able to make these dis- ing them consistently on each recording un-
tinctions.10 Auditory pattern recognition, til the techniques are mastered. When that
like visual pattern recognition, is so intrinsic occurs, the time taken is far less than that of
to cortical function that once learned it oc- quantitative EMG. Once learned, semiquan-
curs essentially instantaneously. The skills of titative EMG includes the following steps
auditory pattern recognition form the basis for both spontaneous and voluntary EMG
of learning the major distinct patterns of fir- activity:
ing of EMG discharges: • One-five recurrent potentials from a
• Semi-rhythmic—recurring in orderly, single area of muscle are recorded and
but not precise, intervals displayed.
• Regular—recurring at precisely defined • The number of potentials and their in-
intervals that may be identical, may be dividual rates of firing are determined.
changing slowly or rapidly, or may be • The variables of motor unit potentials
changing in a linear or exponential (rise time, duration, amplitude, phases,
manner turns, stability) are determined for each
• Irregular—recurring in random inter- of the potentials.
vals with no predictability • With no change in activation, the steps
• Burst—groups of discharges firing at above are repeated in additional areas
one interval in the burst, with the burst of muscle (0.5-mm movements).
recurring at slower intervals • Recordings in different areas are re-
Another level of pattern recognition is the peated until a minimum of 30 poten-
recognition of new sounds or images that are tials has been recorded. The findings at
different from those learned in the past but each location are averaged mentally.
similar enough to be categorized, for exam- • Selected, typical samples of the poten-
ple, seeing a different breed of dog or hear- tials in the muscle are photographed for
ing a new bird song. Although that particu- confirmation and record keeping.
lar breed of dog had not been seen before • The recruitment and variables of motor
or that particular bird song heard before, unit potentials are best recorded for the
each can be identified as a dog or a bird muscle before proceeding to test the
song. After the essence of dogs or bird songs next muscle.
has been learned, new sounds or images in After these steps have been mastered, the
the same category can be recognized. This electromyographer should be able to make
skill allows electromyographers to recognize each of the measures with more than 90%
and categorize EMG signals, like complex accuracy.
ORIGIN OF that are not innervated by an axon have an

ELECTROMYOGRAPHIC unstable muscle fiber membrane potential
and fire individually without external stimu-
POTENTIALS lation, usually with a regular rhythm. These
are fibrillation potentials.
The electric activity of motor units recorded
with a needle electrode in a muscle is de-
rived from the action potentials of the mus- NORMAL ELECTROMYOGRAPHIC
cle fibers that are firing singly or in groups
near the electrode.11'12 The muscle fiber and
ACTIVITY
motor unit potentials have a triphasic con-
figuration. If the motor unit potential is
Pattern of Firing
recorded from a region of a muscle fiber that
Motor unit potentials under voluntary con-
is unable to generate a negative potential
trol normally fire with a semirhythmic pat-
(for example, if the membrane of the mus-
tern at a relatively constant frequency, al-
cle fiber is damaged), the potential will be
though this frequency continuously changes
recorded as a large positivity followed by a
as the voluntary activation increases or de-
long low negativity.13 For example, poten-
creases. The rate of firing is used as a gauge
tials recorded from damaged areas of fibril-
of the intensity of excitation of the anterior
lating muscle fibers are recorded as positive
horn cell by the central nervous system. Mo-
waves (Fig. 24-2). The amplitude of the ex-
tor units in various muscles initiate firing
ternally recorded action potential and the
rates at 5-8/second and, with effort, gradu-
rate of rise of the positive—negative inflec-
ally increase up to 20-40/second. The fol-
tion (rise time) fall off exponentially in pro-
lowing are definitions of motor unit poten-
portion to the distance the recording elec-
tial firing:
trode is from the muscle fibers.14 Normally,
• Slowfiring—individualmotor unit po-
the size and shape of the potential are con-
tentials slower than 10/second
stant each time it occurs. The size of single
• Poor activation—slow firing despite
fiber action potentials is related directly to
effort
the diameter of the muscle fiber and can be
• Rapidfiring—individualmotor unit po-
used clinically to judge the duration of
tentials faster than 12/second
denervation.1^
• Reduced recruitment (reduced num-
Action potentials in individual muscle
ber)—rapid firing with few motor unit
fibers may occur spontaneously or they may
potentials (poor recruitment, slow recruit-
be initiated by external excitation. Normally,
ment)
muscle fibers are under neural control and .
• Rapid recruitment (rapid firing)
fire only in response to an end plate poten-
tial that reaches threshold. Therefore, the • Doublets (triplets)—the firing of a mo-
tor unit potential twice (or three times)
rate and pattern of firing of all muscle fibers
in rapid succession (interval less than
in the motor unit are under neural control.
50 ms), but with each pair firing in the
These are motor unit potentials. Muscle fibers
usual pattern for motor unit potentials
(under voluntary control at 3—20/sec-
ond)
Spontaneous Activity
Normal muscle fibers show no spontaneous
electric activity outside the end plate region.
In the end plate region, spontaneous minia-
Figure 24-2. Simultaneous recording of a single mus- ture end plate potentials occur randomly.
cle fiber potential from one needle electrode that me-
chanically initiates the potentials and from another These may be recorded with needle elec-
electrode, a few millimeters distant, recording from the trodes as many monophasic negative waves
same muscle fiber. that have amplitudes less than 10 ju,V and du-
recorded as rapid, irregularly firing, positive

waves. Normal muscle fibers that are not in-
nervated, for example, muscle fibers in tis-
sue culture, generate rhythmic spontaneous
action potentials. These potentials are asso-
ciated with regular muscle fiber twitches
called fibrillations.
Voluntary Activity
Figure 24-3. Top, Spike form of normal spontaneous All voluntary muscle activity is mediated by
activity in the end plate region. Middle, End plate noise lower motor neurons and the muscle fibers
is seen as an irregularity in the baseline at a low am-
plification in addition to the spike form. Bottom, End they innervate (motor units) and is recorded
plate spikes and noise are seen as small triphasic po- electrically as motor unit potentials. The mo-
tentials at a fast sweep speed and high amplification. tor unit potential is the sum of the poten-
tials of the muscle fibers innervated by a sin-
gle anterior horn cell. The muscle fibers in
rations of 1-3 ms or less. Individual poten- the region of the needle electrode discharge
tials occur irregularly but usually cannot be in near synchrony; the motor unit potential
distinguished. This activity is usually seen as has a more complex configuration of higher
an irregular baseline called end plate noise, it amplitude and longer duration than a sin-
has a typical "seashell sound" (Fig. 24-3). gle fiber action potential. Motor unit po-
The action potentials of some muscle tentials may be characterized by their firing
fibers may be recorded in the end plate re- pattern and by their appearance. Only a
gion as brief spike discharges called end plate small proportion of the fibers in a motor
spikes.1^ They have a rapid irregular pattern, unit are near the electrode; those at a dis-
with an initial negative deflection. End plate tance contribute little to the motor unit po-
spikes sound like "sputtering fat in a frying tential. Thus, the appearance of a motor unit
pan." They are caused by mechanical acti- potential from one motor unit varies with
vation of a nerve terminal with secondary dis- electrode position. No single motor unit po-
charge of a muscle fiber. End plate spikes of- tential characterizes a motor unit but rather
ten have interspike intervals of less than a multiplicity of them recorded from differ-
50 ms. If the needle electrode has damaged ent sites (Fig. 24-4). Motor unit potentials
the muscle fibers, end plate spikes will be are under voluntary control and have a char-
Figure 24-4. Normal motor unit potentials in the biceps muscle showing the variety of configurations that can be
seen.
acteristic semirhythmic firing pattern in patients with amyotrophic lateral sclerosis.

which the rate is continuously changing by The normal onset frequency in the biceps
small amounts. The firing pattern of the mo- muscle ranged from 6 to 8 Hz, with the re-
tor unit potential is assessed in terms of rate cruitment frequency of the second motor
and recruitment. unit at 7-12 Hz. In patients with amyo-
Clinical EMG judgments about the num- trophic lateral sclerosis, the onset frequency
ber of motor units compare the rate of fir- was from 8 to 20 Hz, with recruitment fre-
ing of single units with the total number of quencies of 12-50 Hz. These studies pro-
motor units. The determination of the rate vided quantitative measures of motor unit
of firing is one of the more difficult steps to number estimate. Formal quantitative mea-
make in standard EMG, because of difficulty sures can provide evidence of the reliability
in obtaining sufficient control by the patient of the clinical methods; however, they are so
of motor units to isolate one or two units. time consuming and complex that they have
When it is possible, the rate of firing of the not been applied clinically (see Chapter 25).
motor unit initially activated is measured at Further studies and technical developments
the time the second unit begins to fire. In may eventually allow recruitment analysis to
most muscles, this occurs at 6-8 Hz. If the provide more accurate estimates of the
number of motor units in a muscle is de- number of motor units in a muscle.
creased, the second motor unit does not be- Recruitment is the initiation of the firing of
gin to fire until the first has reached a higher additional motor units as the rate of dis-
rate of firing. Or, less commonly, the first charge of the active motor unit potential in-
unit begins firing at a higher rate than nor- creases. Normally, recruitment of additional
mal (more than 10 Hz). For most normal motor unit potentials occurs at low levels of
muscles, there will be two motor unit po- effort and at slow rates of firing (Fig. 24-5).
tentials firing if one of them is firing at 10 Recruitment can be characterized by recruit-
Hz, three at 15 Hz, and four at 20 Hz. The ment frequency, which is the frequency of fir-
ratio of the number of units firing to the rate ing of a unit when the next unit is recruited
of firing can provide a rough gauge of the (that is, begins to discharge). This is a func-
number of motor units. If the ratio is greater tion of the number of units capable of fir-
than 5, there is virtually always some de- ing and is 7-16 Hz for motor units in a
crease in the number of motor units. Thus, normal muscle during mild contraction.19
firing rate of motor unit potentials is an im- Recruitment frequencies vary in different
portant measure of the loss of axons and mo- muscles and for different types of motor
tor neurons.17 This semiquantitative method units. Recruitment may also be character-
of determining reduced recruitment pro- ized by the ratio of the rate of firing of the
vides a more accurate and reproducible es- individual motor units to the number that
timate of the number of motor units than are active. Normally, this ratio averages less
full interference pattern analysis. than 5 (for example, 3 units firing at about
Although recruitment analysis is reason- 15 Hz each). If the ratio is greater than 5
ably reproducible and clinically reliable, it is (for example, 2 units firing at 16 Hz), a loss
usually a subjective judgment made by elec- of motor units is indicated. Motor unit fir-
tromyographers on the basis of experience. ing can also be characterized in terms of the
It requires taking into account differences in number of motor unit potentials firing in re-
recruitment in different areas of individual lation to the force being exerted.
muscles and the even greater differences
among different muscles. Automated meth-
ods for formally quantitating the recruit- RECRUITMENT
ment pattern have been developed.18 In au-
tomated studies, individual motor unit The number of motor units in a muscle may
potentials were isolated in human muscles be considered in two ways. The first is the
under voluntary control in an experimental total number of motor units that could be
setting. The interpotential interval (the in- fired if the anterior horn cell pool received
verse of frequency of firing) was determined adequate input. The second is the actual
for a population of normal subjects and for number of motor units that are activated
Figure 24-5. Motor unit potential firing under voluntary control showing minimal reduction in recruitment in an extensor carpi radialis muscle with normal
strength. Top, Two motor units (A and B) initially fire at 5 and 6 Hz. Middle, With increased voluntary effort, firing rate of A and B increases to 8-9 Hz, with re-
cruitment of a third unit (C). Bottom, With greater effort, the rates increase to 10-11 Hz, with no additional nearby units recruited. Only a small, distant unit
begins firing at 7 Hz (D). (From DaubeJR. Electrodiagnostic studies in amyotrophic lateral sclerosis and other motor neuron disorders. Muscle Nerve 23:1488-1502,
2000. By permission of TohnJWiley & Sons.)
when a patient attempts a voluntary con- should not be used to describe the con-
traction. The first of these is more pertinent dition of patients in whom relatively few
in assessing the presence or absence of dis- motor unit potentials fire because of
ease involving the lower motor neuron and pain, strong muscles, upper motor neu-
is demonstrated in recruitment. The second ron lesions, or poor cooperation. Al-
is quite variable and changes with the pa- though few potentials are fired, they
tient's cooperation, the strength of the mus- fire slowly with a normal pattern of
cle, pain, and the presence or absence of recruitment.
disease of the upper motor neuron. • Incomplete activation—a normal re-
The recruitment of motor unit potentials cruitment pattern and normal recruit-
as reported in the EMG report should esti- ment frequency, but with relatively few
mate the total number of motor units avail- motor potentials firing. These poten-
able for activation and not the actual num- tials fire slowly, but recruitment of
ber that the patient fires. The total number additional potentials is normal. This oc-
available can best be assessed by the pattern curs with upper motor neuron disor-
of recruitment of motor unit potentials with ders, poor cooperation by the patient,
increasing voluntary effort. In normal mus- pain, excessively strong muscle, or two-
cle, as the effort exerted increases, the fre- joint muscles, such as the gastrocne-
quency of each motor unit potential that is mius. It is not evidence of lower motor
firing increases.20 As this occurs, additional neuron disease.
motor units begin to fire (to be recruited). • Increased number in proportion to
In the presence of lower motor neuron dis- force—the occurrence of large num-
ease, with loss of either the anterior horn bers of motor unit potentials with nor-
cell or peripheral motor axon, recruitment mal recruitment frequency and normal
frequency increases, that is, motor unit po- patterns of recruitment, but with mini-
tentials fire more rapidly before additional mal effort. This must be graded in pro-
motor units are recruited. Conversely, the portion to the force exerted, because
rate of firing of those motor unit potentials the patterns of firing are entirely nor-
already firing will be unduly fast for the num- mal. It is the only estimate described
ber of motor unit potentials that have been that requires consideration of the force
activated. Because there may be selective loss exerted by the muscle. It is evidence of
of higher threshold motor units, recruit- disease involving the muscle directly.
ment analysis should include levels of effort A motor unit potential is also character-
associated with firing rates in the range of ized by its appearance, including duration
15 Hz. However, full interference patterns (onset from the baseline to the final return
are not necessary. By using the pattern of re- to baseline), number of phases (baseline
cruitment, it is possible to assess the nor- crossings plus one), amplitude (peak to
mality of the number of motor units in the peak), turns (potential reversals), area (un-
muscle with mild, moderate, or maximal ef- der the curve), and rate of rise of the fast
fort on the part of the patient. component, that is, rise time (Fig. 24-6).
• Normal—the pattern of recruitment is Each of these characteristics has multiple
normal for that muscle, with an ade- determinants. Technical factors have a ma-
quate number of motor unit potentials jor influence on the appearance of motor
being recruited for the frequency of fir- unit potentials. These factors include the
ing present. If maximal effort can be ob- type of needle electrode, the area of exposed
tained, a full interference pattern is surface of the active leads, the characteris-
seen, but individual motor unit firing tics of the metal recording surfaces, and the
rates of 15 Hz are sufficient for recruit- electric characteristics of the cables, pream-
ment analysis. plifier, and amplifier.
• Reduced recruitment—a higher re- The appearance of motor unit potentials
cruitment frequency or a smaller num- also changes with several normal physiologic
ber of motor unit potentials recruited variables, including the subject's age, the
for firing rates in tile range of 15 Hz muscle being studied, the location of the
than expected for that muscle. This needle in the muscle, the manner of activa-
Figure 24-6. Schema of motor unit potential showing characteristics that can be measured.
tion of the potentials (minimal voluntary locity of the nerve terminals; the diameter
contraction, maximal voluntary contraction, of the muscle fibers; and the relative loca-
reflex activation, or electric stimulation), tion of the end plates on the muscle fibers.
and the temperature of the muscle.21'22 If The firing characteristics of the motor unit
these technical and physiologic factors are depend on the amount of overlap with other
controlled, the normal anatomical and his- motor units, the number of motor units in
tologic features of the motor unit and any the muscle (or per given area), the differ-
pathologic changes that may affect these fea- ential response to sources of activation
tures will determine the characteristics of (monosynaptic, local spinal cord, higher
the motor unit potentials. The anatomical centers), and the rates and patterns of dis-
and histologic features include inneruation charge of the anterior horn cell.
ratio (number of muscle fibers in the motor The configuration of a motor unit poten-
unit), fiber density (number of muscle fibers tial may be monophasic, biphasic, or tripha-
per given cross-sectional area), the distance sic or it may have multiple phases. The con-
of the needle tip from the muscle fibers and figuration depends on the synchrony of
from the end plate region, and the direction firing of the muscle fibers in the region of
of the axis of the muscle fiber. The capaci- the electrode. Usually, only a small propor-
tance and resistance of the tissue between tion of motor unit potentials have more than
the electrode and the discharging muscle four phases; those that do are called polypha-
fibers depend On the amount of connective sic potentials. The percentage varies with the
tissue, blood vessels, and fat. The character- muscle being tested and the age of the pa-
istics of the action potentials generated by tient. A late spike, distinct from the main po-
individual muscle fibers depend on muscle tential, that is time locked to the main po-
fiber membrane resistance and capacitance, tential may occur; this is called a satellite
intracellular and extracellular ionic concen- potential (Fig. 24-6). The satellite potential
trations, muscle fiber diameter, and con- is generated by a muscle fiber in a motor
duction velocity. The synchrony of firing of unit that has a long nerve terminal, narrow
the muscle fibers in a motor unit depends diameter, or distant end plate region. If a
on the length, diameter, and conduction ve- motor unit potential is recorded from dam-
aged muscle fibers or from the end of the most exclusively to motor unit activity, but
muscle fibers, it may have the configuration not to spontaneous activity. These methods
of a positive wave with low, long, late nega- are not widely used clinically.
tivity.
The rise time is the duration of the rapid
positive-negative inflection and is a function Measurement of Motor
of the distance of the muscle fibers from the Unit Potentials
electrode. It is less than 500 microseconds if
the electrode is near muscle fibers in the ac- The variables of a motor unit potential may
tive motor unit. The duration of the motor be assessed subjectively, but this requires ex-
unit potential is the time from the initial de- perience and knowledge of the normal vari-
flection away from baseline to the final re- ations from muscle to muscle and with age.
turn to baseline. It varies with the muscle, It is more reliable to make a quantitative
muscle temperature, and the patient's age. measurement of the characteristics of the
The amplitude of the potential is the maxi- potentials. This is needed in questionable
mal peak-to-peak amplitude of the potential cases to increase the certainty of a diagno-
and varies with the size and density of the sis. Objective measurement may be a neces-
muscle fibers in the region of the recording sity in recognizing mild myopathies.
electrode and with their synchrony of firing. Measurements may be made in two ways:
It also differs with the muscle, muscle tem- by isolation and measurement of a single
perature, and the patient's age. Decreased motor unit potential and by interference
muscle temperature produces higher ampli- pattern analysis. Because of the number and
tude, longer duration potentials. Variability variety of normal motor unit potentials, both
of a motor unit potential is any change in its of these methods require multiple measure-
configuration, or amplitude, or both in the ments and a statistical description of the
absence of movement of the recording elec- results obtained from different areas of a
trode as the motor unit fires repetitively. muscle.
Normally, there is no such variation. The classic method of measuring a motor
The polarity of all potentials recorded unit potential is to isolate and record at least
on needle EMG depends on recording the 20 single potentials and then manually mea-
potential with the active (Gl amplifier in- sure the duration, number of phases, and
put) electrode. If a motor unit potential is amplitude. These measurements must be
recorded with the shaft of a standard con- compared with the values recorded from the
centric electrode or with the reference of a same muscle in normal subjects of the same
monopolar electrode, it will be displayed as age. This method provides no quantitative
an inverted triphasic potential (apparently assessment of recruitment and makes the
negative-positive—negative). measurements only at minimal-to-moderate
levels of contraction. Recently developed
digital EMG machines have automated the
measurements.
SIGNAL ANALYSIS Interference pattern analysis summates
the effect of recruitment with the duration
The spontaneous activity and motor unit po- and amplitude of the potentials and records
tentials recorded from muscle can be dis- the number of turns and total amplitude
played and analyzed in several different of the electric activity during a fixed time
ways.23"25 The usual method in clinical stud- with an automatic counting device.26 This
ies is to display and measure isolated poten- method varies with patient effort, which
tials (described below under Measurement must be accounted for in measurements.
of Motor Unit Potentials); however, other Recording EMG by isolation of single motor
approaches that analyze the entire sequence unit potentials and by interference pattern
of waveforms in an interference pattern provides reliable estimates of the electric ac-
when multiple motor units are firing have tivity in a muscle. The results of these two
also been used. Such analyses are applied al- methods correlate well with each other and
with muscle histology. Neither method has ations. Oscilloscope sweep speeds of 5-10 ms
been shown to be superior; they may com- per centimeter are best for characterizing
plement one another. the appearance of motor units, but slower
speeds of 50 or 100 ms per centimeter are
needed to characterize firing patterns. Am-
Recording Technique plification settings of 50 />tV/cm and 200
fjiV/cm are most useful for examining spon-
Needle electrode recording must sample taneous and voluntary activity, respectively.
three types of activity: insertional activity, Filter settings of approximately 30 Hz and
spontaneous activity, and voluntary activity. 10,000 Hz or more should be used for rou-
Because needle electrodes primarily record tine studies. Measurements of the duration
activity from a small area in a muscle, the of motor unit potentials should be made
electrode must be moved in the muscle to with a gain of 100 /JiV/cm at a sweep speed
record each type of activity in multiple ar- of 5 ms/cm (10 ms/cm if long duration) and
eas. Adequate control during needle ma- with a low filter at 2-3 Hz. The convention
nipulation can only be obtained manually of displaying negative potentials at the active
with small advances of the needle. With the electrode as upward deflections is now gen-
examiner's hand resting on the patient, the erally accepted in clinical EMG.
needle should be held firmly and steadily in Insertional activity is the electric response
the hand without release. Small movements of the muscle to the mechanical damage by
are less painful than large movements with a small movement of the needle (Fig. 24-7).
insertional bursts longer than 500 ms. Evaluation of insertional activity requires a
Each electromyographer develops a pref- pause of 0.5-1 second or more to see any
erence for how to display the electric activ- repetitive potentials that may be activated.
ity; however, certain variables should be fa- Insertional activity may be increased, de-
miliar to all examiners because of their creased, or show specific wave forms, such as
common use and advantages in certain situ- myotonic discharges.
Figure 24-7. Top, Normal short burst of insertional activity with needle movement. Bottom, Increased insertional ac-
tivity with a train of repetitive firing potentials after the insertional burst.
Spontaneous activity occurs with the needle ings are combinations of different specific
and muscle both at rest but does not include types of abnormal electric wave forms de-
motor unit potentials occurring because of scribed in the following sections. The clini-
upper motor neuron activity, as in poor re- cal electromyographer must recognize spe-
laxation, tremor, or spasticity (these can be cific discharges and know what diseases are
distinguished by their firing patterns). Spon- associated with them. In most cases, a spe-
taneous activity is assessed during the pauses cific discharge may be associated with several
between needle movements. It decreases different diseases. The following discussion
when muscle temperature is low. A search describes the types of abnormal electric ac-
for spontaneous activity requires adequate tivity recorded with a needle electrode and
relaxation of the muscle, which may be dif- the diseases associated with them. Neuro-
ficult to obtain in some patients. Relaxation muscular diseases may show abnormal spon-
is achieved readily by gende manipulation taneous discharges or abnormal voluntary
or passive positioning of the extremity. Oc- motor unit potentials or both. Abnormal
casionally, it can be obtained by activation spontaneous activity includes fibrillation po-
of an antagonist, for example, by flexing the tentials, fasciculation potentials, myotonic
patient's neck when testing cervical para- discharges, complex repetitive discharges,
spinal muscles. myokymic discharges, cramps, and neu-
The recruitment and appearance of mo- romyotonic discharges. Motor unit poten-
tor unit potentials are examined during vol- tials may have an abnormal duration, be
untary activity. Several different motor unit polyphasic, or vary in size. The recruitment
potentials (a minimum of 20) in different pattern of the potentials may be altered or
areas of the muscle must be assessed. They there may be abnormal patterns of activa-
are studied most efficiently by having the pa- tion, as in tremor and synkinesis.
tient maintain a minimal voluntary contrac- All normal and abnormal EMG discharges
tion while the needle is advanced through are recognized most accurately and reliably
different areas of the muscle. It is difficult by auditory pattern recognition. The trained
to identify individual potentials when there ear of an electromyographer can define the
is tremor or strong contractions, and at- discharge frequency, rise time, duration,
tempts should be made to minimize these and number of turns/phases of EMG po-
conditions. Motor unit potentials cannot be tentials. This skill is called semiquantitation.
measured reliably during a strong voluntary
contraction, which normally produces a
dense pattern of multiple superimposed po-
tentials called an interference pattern. Less Insertional Activity
dense patterns may occur if there is a loss of
motor units, poor effort, or an upper motor Insertional activity may be prolonged (Fig.
lesion or if the muscle is powerful. The lat- 24-7) or reduced from the brief burst that
ter three conditions can be distinguished occurs in normal subjects. Prolonged inser-
from a loss of motor units only by estimates tional activity occurs in two types of normal
of firing rates. The combination of rapid fir- variants and in denervated muscle and my-
ing rates of individual potentials and few mo- otonic discharges. The normal variants are
tor units occurs only with a loss of motor recognized by their widespread distribution.
units. One type is composed of short trains of reg-
ularly firing positive waves. Sometimes, this
type is familial and may be a subclinical myo-
tonia. The second type has short recurrent
ABNORMAL ELECTRIC bursts of irregularly firing potentials. It oc-
ACTIVITY curs most often in muscular persons, espe-
cially in their calf muscles. Reduced inser-
Neuromuscular diseases are best described tional activity occurs in periodic paralysis
by a combination of clinical findings, histo- (during paralysis) and with replacement of
logic changes, and the pattern of abnormal muscle by connective tissue or fat in my-
findings on needle EMG. Needle EMG find- opathies and neurogenic disorders.
Figure 24-8. Fibrillation potentials. A, Spike form. B, Positive waveform. C, Development of a positive waveform
from a spike form (serial photographs taken after insertion of needle electrode).
Fibrillation Potentials
Fibrillation potentials are the action potentials Table 24-1. Diseases with Fibrillation
of single muscle fibers that are twitching
spontaneously in the absence of innervation. Potentials
These potentials typically fire in a regular Lower motor neuron diseases
pattern at rates of 0.5—15 per second (Fig.
Anterior horn cell diseases
24-8). Infrequently, they may be intermit-
tent or irregular, but if so, the interspike in- Polyradiculopathies
terval is longer than 70 ms. Fibrillation po- Radiculopathies
tentials have one of two forms, either a brief Plexopathies
spike or a positive wave.27'28 If fibrillation po- Peripheral neuropathies, especially axonal
tentials occur as brief spikes (spike form), they Mononeuropathies
are triphasic or biphasic, 1-5 milliseconds in
duration, and 20-200 /u,V in amplitude, with Neuromuscular junction diseases
an initial positivity (unless recorded at the Myasthenia gravis
site of origin). If fibrillation potentials occur Botulinum intoxication
as positive waves (positive waveform), they are Muscle diseases
of long duration and biphasic, with an ini-
Myositis
tial sharp positivity followed by a long-dura-
tion negative phase. Amplitudes are from 20 Duchenne dystrophy
to 200 ^tV and durations from 10 to 30 mil- Myotonic dystrophy
liseconds. Amplitude is proportional to mus- Myotubular myopathy
cle fiber diameter and decreases with mus- Late-onset rod myopathy
cle atrophy.29 The positive waveforms are Toxic myopathy
muscle fiber action potentials recorded from Hyperkalemic periodic paralysis
an injured portion of the muscle fiber. The Acid maltase deficiency
spike form and the positive waveform are Rhabdomyolysis
both recognized as fibrillation potentials by Trichinosis
their slow, regular firing pattern, which is Muscle trauma
like the "ticking of a clock."
Figure 24-9. Fibrillation potentials in denervated muscle. Grades of activity: 1+, fibrillation potentials persistent in
at least two areas; 2+, moderate number of persistent fibrillation potentials in three or more areas; 3+, large num-
ber of persistent discharges in all areas; 4+, profuse, widespread, persistent discharges that fill the baseline.
Rarely, fibrillation potentials are observed neously in a prolonged fashion after exter-
to transform from a spike to a positive wave- nal excitation. They are less readily elicited
form or vice versa; even less frequently, two in a muscle that has just been active. The po-
fibrillation potentials are time-locked. ° Any tentials wax and wane in amplitude and fre-
muscle fiber that is not innervated can be quency because of an abnormality in the
expected to fibrillate; thus, many neuro- membrane of the muscle fiber. Myotonic dis-
genic and myopathic disorders show fibril- charges are regular in rhythm, but they vary
lation potentials (Table 24-1). These po- in frequency between 40 and 100/second,
tentials may occur in muscle fibers that (1) which makes them sound like a "dive-
have lost their innervation, (2) have been bomber." Myotonic discharges occur as brief
sectioned transversely or divided longitudi- spikes or positive waveforms, depending on
nally, (3) are regenerating, and (4) have never the relation of the recording electrode to the
been innervated. The density of fibrillation muscle fiber. When initiated by insertion of
potentials is a rough estimate of the number the needle, myotonic potentials have the
of denervated muscle fibers (Fig. 24-9). configuration of a positive wave, with an
Other forms of electric activity may be mis- initial sharp positivity followed by a long-
taken for fibrillation potentials. These in- duration negative component. These are ac-
clude the spontaneous activity in the region tion potentials recorded from an injured
of the end plate (end plate noise and end area of the fiber. Both amplitude and fre-
plate spikes), short-duration motor unit po- quency may increase or decrease as the dis-
tentials, and motor unit potentials with a charge continues (Fig. 24-10). Myotonic
positive configuration. All of them are best discharges that occur after a voluntary
distinguished from fibrillation potentials by contraction are brief, biphasic or triphasic,
their firing patterns.
Myotonic Discharges
Myotonic discharges are the action potentials
of muscle fibers that are firing sponta- Flgure 24-10. Myotonic discharge.
Table 24-2. Muscle Diseases with and cessation. During the discharge, they
Myotonic Discharges may have abrupt changes in their configu-
ration. The frequency is uniform, ranging
Myotonic dystrophy from as slow as 3 to up to 40/second (Fig.
Myotonia congenita 24-11). Although their form is variable, it
Paramyotonia typically is polyphasic, with 3-10 spike com-
Hyperkalemic periodic paralysis ponents with amplitudes from 50 to 500 fiV
and durations of up to 50 ms. Complex
Polymyositis
repetitive discharges sound like "a motor
Acid maltase deficiency boat that misfires." They occur in several
chronic disorders, both myopathic and neu-
rogenic (Table 24-3).
Complex repetitive discharges may be
initially positive spikes of 20-300 /iV that re- confused with other repetitive discharges,
semble the spikes of fibrillation potentials. such as myokymic discharges, cramps, neu-
They wax and wane, similar to mechanically romyotonia, tremor, and synkinesis. How-
induced myotonic discharges. This afterdis- ever, each of these has a characteristic pat-
charge corresponds to the clinically evident tern of firing best recognized by its sound
poor relaxation. and distinct from that of complex repetitive
In several disorders, myotonic discharges discharges.
may occur with or without clinical myotonia
(Table 24-2). Rarely, similar discharges may
occur with fibrillation potentials in chronic Fasciculation Potentials
denervating disorders and with some drugs
(for example, 20,25-diazocholesterol, tripar- Fasciculation potentials are randomly dis-
anol, 2,4-dichlorophenoxyacetic acid, and charging action potentials of a group of mus-
clofibrate). cle fibers innervated by an anterior horn cell
(Fig. 24-12). The rates of discharge of an
individual potential may vary from a few per
Complex Repetitive Discharges second to fewer than 1/minute. The sum of
all fasciculations in a muscle may reach
Complex repetitive discharges, referred to pre- 500/minute. These potentials maybe of any
viously as bizarre repetitive (or high-frequency) size and shape, depending on the character
potentials or pseudomyotonic discharges, are the of the motor unit from which they arise and
action potentials of groups of muscle fibers their relation with the recording electrode,
discharging spontaneously in near syn- and they may have the appearance of nor-
chrony. Standard and single fiber EMG mal or abnormal motor unit potentials. They
recordings suggest that they are the result of can be identified only by their firing pattern.
ephaptic activation of groups of adjacent The discharges may arise from any portion
muscle fibers.31 Complex repetitive dis- of the lower motor neuron but usually from
charges are characterized by abrupt onset spontaneous firing of the nerve terminal.
Figure 24-11. Two examples of complex repetitive discharges recurring at 30-40 per second.
Table 24-3. Disorders Associated with Table 24-4. Common Sources of

Complex Repetitive Discharges Fasciculation Potentials
Motor neuron and peripheral nerve disease Normal
Poliomyelitis Benign (fatigue)
Amyotrophic lateral sclerosis Benign with cramps
Spinal muscular atrophy
Metabolic disorders
Chronic radiculopathies
Tetany
Charcot-Marie-Tooth disease
Thyrotoxicosis
Chronic neuropathies
Anticholinesterase medication
Myopathies
Lower motor neuron diseases
Polymyositis
Amyotrophic lateral sclerosis
Duchenne dystrophy
Root compression
Limb-girdle dystrophy
Peripheral neuropathy
Myxedema
Creutzfeldtjakob disease
Schwartz-Jampel syndrome
The random occurrence sounds like "large tractions. Such contraction fasciculations
raindrops on a roof." must be differentiated from true fascicula-
Fasciculation potentials may occur in nor- tions by the pattern of firing.
mal persons and in many diseases. They are
especially common in chronic neurogenic
disorders but have been found in all neuro- Myokymic Discharges
muscular disorders (Table 24-4). It has not
been shown clearly that they occur more of- Spontaneous muscle potentials associated
ten in patients with myopathy than in nor- with the fine, worm-like quivering of facial
mal persons. Fasciculations usually occur in myokymia are called myokymic discharges.
an overworked muscle, especially if there is They have the appearance of normal motor
underlying neurogenic disease. unit potentials that fire with a fixed pattern
No reliable method exists for distinguish- and rhythm and occur in bursts of 2-10 po-
ing between benign fasciculations and those tentials that fire at 40-60 Hz. The bursts re-
associated with specific diseases. However, in cur at regular intervals of 0.1-10 seconds
normal persons, fasciculations occur more (Fig. 24-13). The firing pattern of any one
rapidly, on the average, and are more sta- potential is unrelated to the myokymic dis-
ble.32 Patients who have large motor units charge pattern of other potentials and is un-
caused by chronic neurogenic diseases may affected by voluntary activity. Myokymic dis-
have visible twitching during voluntary con- charges sound like "marching soldiers."
Some forms of clinical myokymia—espe-
cially the form in the syndrome of continu-
ous muscle fiber activity (Isaac's syn-
drome)—are associated with neurotonic
discharges and not with myokymic dis-
charges. Although discharges that have reg-
ular patterns of recurrence but fire at dif-
ferent rates or with a regularly changing rate
of discharge may have similar mechanisms,
they are better classified with the broad
group of iterative discharges. Some investi-
gators consider iterative discharges and
Figure 24-12. Fasciculation potentials recurring in an myokymic discharges to be forms of fascicu-
irregular pattern. A, Slow sweep speed, continuous. B, lation because they arise in the lower motor
Fast sweep speed, raster. neuron or axon. It is best to separate these
Figure 24-13. Examples of recurrent bursts of myokymic discharges at a slow (left) and fast (right) sweep speed. Fir-
ing rate is 20-30/second within bursts, with variable recurrence rates of the bursts.
discharges from fasciculation potentials be- quencies of 100—300 Hz are called neurotonic
cause of their distinct patterns and different discharges, or neuromyotonia (Figs. 24—14 and
clinical significance. Diseases associated with 24-15). These potentials may decrease in
myokymic discharges are listed in Table amplitude because of the inability of muscle
24-5. fibers to maintain discharges at rates greater
than 100 Hz. The discharges may be con-
tinuous for long intervals or recur in bursts.
Neurotonic Discharges They are unaffected by voluntary activity and
(Neuromyotonia) are commonly seen in neurogenic disorders
(Table 24-6),33
Motor unit potentials that are associated Neuromyotonia occurring with tetany may
with some forms of continuous muscle fiber be distinguished by its precipitation by or
activity (Isaac's syndrome) and fire at fre- augmentation with ischemia. Neurotonic
discharges also occur intraoperatively with
the mechanical irritation of cranial or pe-
ripheral nerves and, thus, are valuable in
Table 24-5. Disorders Associated with alerting surgeons to possible nerve damage
Myokymic Discharges (Fig. 24-16).
Facial muscles
Multiple sclerosis Cramp Potentials
Brain stem neoplasm
Polyradiculopathy The potentials associated with a muscle
Facial palsy cramp do not have a specific name but can
Extremity muscles be distinguished from other spontaneous ac-
tivity by their firing pattern. Individual po-
Radiation plexopathy
tentials are not distinctive and resemble mo-
Chronic nerve compression (for example, tor unit potentials. They fire rapidly at 40-60
carpal tunnel syndrome)
Hz, usually with abrupt onset and cessation;
figure 24-14. Examples of neuromyotonic (neurotonic) discharges in Isaac's syndrome.
however, during their discharge, they may the cramp develops and then stop firing as
fire irregularly in a sputtering fashion, es- the cramp subsides. Cramps are a common
pecially just before termination (Fig. 24-17). phenomenon in normal persons and usually
Typically, an increasing number of poten- occur when a muscle is activated strongly in
tials that fire at similar rates are recruited as a shortened position (Table 24-7).
Figure 24-15. Two examples of neuromyotonic discharges in spinal muscular atrophy firing at over 200/second.
Table 24-6. Disorders Associated with reduced recruitment or normal amplitude

Neuromyotonic Discharges (or both). They may be as short in duration
as a fibrillation potential if only a single mus-
Isaac's syndrome (autoimmune) (see Fig. 24-14) cle fiber is in the recording area of the nee-
Anticholinesterase poisoning dle electrode. Short-duration motor unit po-
Tetany tentials occur in diseases in which there is
Intraoperative nerve irritation (see Fig. 24-16) (1) physiologic or anatomical loss of muscle
fibers from the motor unit or (2) atrophy
Chronic spinal muscular atrophy (see Fig.
of component muscle fibers (Table 24-8).
24-15)
They are most common in primary muscle
Hereditary motor neuropathy
diseases.34 Some myopathies, such as meta-
bolic and endocrine disorders, show few
short-duration motor unit potentials or
none. Technical errors (for example, incor-
Short-Duration Motor rect filter settings, electrical short in the
Unit Potentials recording electrode or connecting cables,
and reduced recording surface area) that
Single potentials that are outside the normal may produce similar findings must be ex-
range or groups of motor unit potentials that cluded to identify true short-duration motor
have a mean duration less than the normal unit potentials.
range for the same muscle in a patient of the Short-duration motor unit potentials may
same age are called short-duration motor unit be intermingled with long-duration poten-
potentials. Commonly, these potentials also tials in rapidly progressing neurogenic atro-
have low amplitude and show rapid recruit- phies such as amyotrophic lateral sclerosis
ment with minimal effort, but they may have and Werdnig-Hoffmann disease.
Figure 24-16. Neurotonic discharges in facial muscles during acoustic neuroma surgery. The times of recordings
were at 1:10 PM, 1:50 PM, and 3:10 PM.
Figure 24-17. Muscle cramp with motor unit potentials firing at 30-50/second.
Long-Duration Motor firing of fibers in a motor unit (Table 24-9).

Unit Potentials Motor unit potential configuration is an im-
portant tool in distinguishing muscle from
Individual motor unit potentials that are out- nerve disease,35 but long-duration motor
side the normal range or groups of motor unit potentials are a common finding in in-
unit potentials that have a mean duration clusion body myositis and in some cases of
greater than the normal range for the same chronic polymyositis. It needs to be remem-
muscle in a patient of the same age are called bered that the size of motor unit potentials
long-duration motor unit potentials (Fig. 24- is dependent on the level of activation, with
18). They generally have high amplitude and larger motor unit potentials becoming active
show poor recruitment, but they may have at a stronger force.36
normal or low amplitude. Motor unit po-
tentials recorded from damaged muscle
fibers are preponderantly positive and have Polyphasic Motor Unit Potentials
a long late negativity, which is a recording
artifact that should not be measured. Long- A polyphasic motor unit potential has five or
duration motor unit potentials occur in dis- more phases. (Phase is defined as the area of
eases in which there is increased fiber den-
sity in a motor unit, an increased number of Table 24-8. Disorders Associated with
fibers in a motor unit, or loss of synchronous Short-Duration Motor Unit Potentials
Myasthenia gravis
Table 24-7. Disorders Associated with Myasthenic syndrome
Cramp Discharges Botulinum intoxication
Early reinnervation after nerve damage
Salt depletion Late-stage neurogenic atrophy
Chronic neurogenic atrophy Muscular dystrophies (all forms)
Benign nocturnal cramps Periodic paralysis
Myxedema Polymyositis
Pregnancy Toxic myopathies
Uremia (dialysis) Congenital myopathies (not all forms)
Figure 24-18. Single long-duration voluntary motor unit potential displayed on a free-running and triggered sweep.
Semirhythmic firing rate of 9/second without recruitment of other potentials is abnormal for this muscle.
a potential on either side of the baseline; in size and shape (nascent motor unit poten-
it is equal to the number of baseline cross- tial). If a motor unit potential is large, an
ings plus one.) Polyphasic potentials may increase in fiber density produces more
be of any duration. Some may have late, turns without extra phases, a complex motor
satellite components, sometimes called unit potential. Polyphasic motor unit po-
linked potentials or parasites, that give the to- tentials may occur in any of the myopathies
tal unit a long duration37 (Figs. 24-19 and or the neurogenic atrophies listed above
24-20). Such long-duration motor unit and are identified as abnormal by an in-
potentials with satellites may occur in my- crease in the proportion of polyphasic po-
opathies or neurogenic diseases. The indi- tentials over that which is normal for that
vidual components of a polyphasic poten- muscle in a patient of the same age. These
tial are action potentials recorded from potentials must be differentiated from dou-
single muscle fibers. Thus, they are com- blets, multiplets, tremor, and motor unit
monly seen in myopathies in which there is potentials that vary in configuration.
fiber regeneration and increased fiber den-
sity. Regeneration of axons in neurogenic
diseases can also produce low-amplitude Mixed Patterns: Long-Duration
polyphasic motor unit potentials that vary and Short-Duration Motor
Unit Potentials
Occasionally, patients have a combination
Table 24-9. Disorders Associated with of the abnormalities described for short,
Long-Duration Motor Unit Potentials long, and polyphasic motor unit potentials,
but instead of having the usual pattern of
Motor neuron diseases (all types) an excess of either long-duration or short-
Axonal neuropathies with collateral sprouting duration potentials, both types occur. The
Chronic radiculopathies quantitative distribution becomes broad
Chronic mononeuropathies rather than shifting to long or short. Rarely,
the distribution of durations may be bi-
Residual of a neuropathy (polyneuropathy or
mononeuropathy) modal (Fig. 24-21). These combinations
commonly occur in chronic myositis or
Chronic myositis
rapidly progressing motor neuron disease.
Figure 24-19. Relative average durations and amplitudes of some electric potentials observed in electromyography
of human muscle.
Motor Unit Potential Variation

As motor unit potentials fire repetitively un-
der voluntary control, they normally have
the same amplitude, duration, and configu-
ration. Fluctuation in time of any of these
variables during repeated discharge of a mo-
tor unit potential is abnormal and is usually
caused by blocking the discharge of action
potentials of the individual muscle fibers in
the motor unit. The disorders in which mo-
tor unit potentials fluctuate from moment to
Figure 24-20. Long-duration polyphasic motor unit po- moment (Fig. 24-22) are listed in Table
tential with satellite potential. 24-10. In disorders of muscle membrane,
Figure 24-21. Patient with inclusion body myositis. Quantitation of motor unit potentials shows a bimodal distri-
bution.
such as myotonia, there may be a slower pro- Doublets (Multiplets)

gressive decrease or increase in a motor unit
potential (Fig. 24-23). In myasthenia gravis Motor units under voluntary control nor-
or in cases of active reinnervation, the am- mally discharge as single potentials in a
plitude initially may decline, but in the myas- semirhythmic fashion. In some disorders,
thenic syndrome, it may increase (Fig. they fire two or more times at short intervals
24-24). of 10-30 ms (Table 24-11) (Fig. 24-25).
Figure 24-22. Top, Normal voluntary motor unit potentials. Middle and bottom, Motor unit instability in myasthenia
gravis.
Table 24-10. Disorders Associated number firing; this is referred to as reduced

with Rapid Motor Unit Potential recruitment. It may be found in any disease
Variation (Unstable Motor Unit process that destroys or blocks conduction
Potential) in the axons innervating the muscle or de-
stroys a sufficient proportion of the muscle
Myasthenia gravis so that whole motor units are lost. This pat-
tern occurs in association with all neuro-
Myasthenic syndrome
genic disorders and may be the only finding
Botulinum intoxication
in neurapraxia in which the sole abnormal-
Myositis ity is a localized axonal conduction block or
Muscle trauma in cases of acute axonal loss in which fibril-
. Reinnervation after nerve injury lation has not yet developed. In myopathies,
Rapidly progressing neurogenic atrophy (for however, more motor units are activated
example, amyotrophic lateral sclerosis) than would be expected for the force ex-
erted in disorders in which the force that a
single motor unit can generate is decreased.
This is called rapid recruitment. The recruit-
These are called doublets, triplets, or multiplets. ment frequency and rate of firing in relation
The bursts of two or more recur in a semi- to number are normal with rapid recruit-
rhythmic pattern under voluntary control. ment.
They are often increased by ischemia.
Disorders of Central Control

Abnormal Recruitment
One pattern of motor unit firing that is
In normal muscle, increasing voluntary ef- caused by disorders of the central nervous
fort causes an increase in the rate of firing system must be recognized because it may
of individual motor unit potentials and ini- resemble the changes seen with lower mo-
tiates the discharge of additional motor unit tor neuron disease. In muscle tremor, which
potentials. The relation of the rate of firing may not be apparent clinically, motor unit
of individual potentials to the number of potentials fire in groups but not in a fixed
potentials firing is constant for a particular relation. The potentials of these motor units
muscle and is called the recruitment pattern. If are superimposed and may resemble poly-
there is a loss of motor unit potentials in any phasic, complex, or long-duration motor
disease process, the rate of firing of individ- unit potentials (Fig. 24-26). They are rec-
ual potentials will be disproportionate to the ognized by their rhythmic pattern and their
Figure 24-23. Myotonia congenita with spontaneous, slow change in motor unit potential size over 1 minute.
Figure 24-24. Motor unit potential variation with gradual increase in amplitude in the Lambert-Eaton myasthenic
syndrome.
changing appearance. Minimal activation, cation of a specific disease. The combina-

with slightly increasing and decreasing ef- tions of particular forms of spontaneous ac-
fort, often allows single motor unit poten- tivity and changes in motor unit potentials
tials to be resolved and characterized. Mo- in neuromuscular diseases are too varied to
tor unit potential firing patterns in stiff-man be included in this review, but some general
syndrome, rigidity, and spasticity resemble comments about patterns of abnormality of
normal patterns, but with loss of voluntary motor unit potentials can be made. Motor
control. In upper motor neuron weakness, unit potential changes have been divided
patients cannot maintain motor unit firing. broadly into neuropathic and myopathic.38
(SSAPs—small, short, abundant potentials—
is an alternative term for myopathic.) The
Synkinesis concept that motor unit potential changes
must be either one or the other of these two
The aberrant regeneration of axons after types is incorrect and can lead to misinter-
nerve injury may result in two different mus- pretations.
cles being innervated by the same axon. In Each of the variables—recruitment, dura-
such cases, voluntary potentials may be mis- tion, amplitude, configuration, and stabil-
taken for spontaneous activity. Examples in- ity—of motor unit potentials may be altered
clude potentials in facial muscles in associa- separately or in combination with one or
tion with blinking and potentials in shoulder more of the others in different disorders.39
girdle muscles in association with respira- Each must be judged individually, quantified
tion. if necessary, and compared with normal val-
ues. The result should then be interpreted
on the basis of known pathophysiologic
Patterns of Abnormality mechanisms or by common association with
known disorders. Recruitment, duration,
The types of needle EMG abnormalities de- and stability are the important features of
scribed above may occur in different com- motor unit potentials in determining the un-
binations. Only through knowledge of these
combinations can reliable interpretations be
made. No single finding allows the identifi-
Table 24-11. Disorders Associated

with Doublets
Hyperventilation
Tetany
Motor neuron disease (infrequent)
Other metabolic diseases Figure 24-25. Voluntary motor unit potentials. A, Dou-
blets. B, Multiplets.
SUMMARY
Figure 24-26. Superimposed motor unit potentials with
tremor that resemble polyphasic potentials. Virtually all primary neuromuscular diseases
result in changes in the electric activity
recorded from muscle fibers. These changes
can best be depicted using fine needle elec-
derlying pathologic factors. With these three trodes inserted into the muscle to record
criteria, it is possible to distinguish most pat- spontaneous and voluntary EMG. Thus,
terns of motor unit potential abnormality. EMG can be used to distinguish among
Each pattern of abnormality changes with lower motor neuron, peripheral nerve, neu-
the severity and duration of the disease. romuscular junction, and muscle disease
Careful attention to the independent with great sensitivity and some specificity.
changes of the variables of motor unit po- The sensitivity is usually greater than clini-
tentials can allow an electromyographer to cal measures; specificity in identifying the
comment on the severity, duration, and cause of the disease often requires muscle
prognosis of a disease.40 Because of the var- biopsy or other clinical measures. Aldiough
ious patterns that may be found, a descrip- EMG is somewhat uncomfortable for pa-
tion of the abnormalities should always in- tients because needles need to be inserted
clude comments about each of the variables. into die muscles, it generally is well tolerated
The findings then can be interpreted most by patients and provides a rapid, efficient
reliably by listing the disorders that may be means of testing the motor unit.
seen with the pattern of abnormality found The application of techniques of clinical
(Table 24-12). neurophysiology in the evaluation of pe-
Table 24-12. Patterns of Abnormality Seen with Needle Electromyography

Recruitment Appearance Variation Disorders
Normal Normal No Normal
Some endocrine and metaboli
myopathies
Yes Myasthenia gravis
Myasthenic syndrome
Short-duration, polyphasic No Primary myopathies
Yes Severe myasthenia
Botulinum intoxication
Reinnervation (neurogenic or
myositis)
Mixed short-duration and Yes or no Chronic myositis and inclusior
long-duration body myositis
Rapidly progressing neurogeni
disorder, such as amyotroph
lateral sclerosis and Werdnig
Hoffmann disease
Poor Normal No Acute neurogenic lesion
Long-duration, polyphasic No Chronic neurogenic atrophy
Yes Progressing neurogenic atroph
Short-duration, polyphasic No Severe myopathy (end-stage,
neurogenic atrophy)
Yes Early reinnervation after sever*
nerve damage
ripheral neuromuscular disorders relies 7. Finsterer J, Fuglsang-Frederiksen A. Concentric

heavily on needle EMG. It was the first of the needle EMG versus macro EMG I. Relation in
healthy subjects. Clin Neurophysiol 111:1211-1215,
electrophysiologic techniques to be applied 2000.
in this way, and it has remained a mainstay 8. Gan R,JabreJF. The spectrum of concentric macro
of electrodiagnosis. The collection of data EMG correlations. Part II. Patients with diseases of
with needle EMG and the interpretation for muscle and nerve. Muscle Nerve 15:1085-1088,
clinical purposes require a firm understand- 1992.
9. Strommen JA, Daube JR. Determinants of pain in
ing of the physiology of muscle fibers and needle electromyography. Clin Neurophysiol
motor units. It is heavily dependent on con- 112:1414-1418, 2001.
trolling various technical factors, mastering 10. Schulte-Mattler WJ, Georgiadis D, Zierz S. Dis-
the skills of data collection, and understand- charge patterns of spontaneous activity and motor
units on concentric needle electromyography. Mus-
ing the changes that occur with the many dis- cle Nerve 24:123-126, 2001.
orders that may affect peripheral nerves, neu- 11. Stalberg E, Karlsson L. Simulation of the normal
romuscular junctions, and muscle. concentric needle electromyogram by using a mus-
The essence of quality needle EMG rests cle model. Clin Neurophysiol 112:464-471, 2001.
12. Stalberg E, Karlsson L. Simulation of EMG in patho-
with the ability to isolate, recognize, and in- logical situations. Clin Neurophysiol 112:869-878,
terpret the wide range of specific waveforms 2001.
and their variation that occur in normal and 13. Lorente de No R. A Study of Nerve Physiology. Stud-
diseased muscle. The nature and meaning of ies From the Rockefeller Institute for Medical Re-
fibrillation potentials and the alteration in search. Vol. 132. The Rockefeller Institute for Med-
ical Research, New York, 1947, pp 466-470.
appearance and firing pattern of motor unit 14. Okajima Y, Tomita Y, Ushijima R, Chino N. Motor
potentials in each muscle tested are the data unit sound in needle electromyography: assessing
on which a clinical interpretation is based. normal and neuropathic units. Muscle Nerve
The extent and distribution of these abnor- 23:1076-1083, 2000.
15. Jiang GL, Zhang LY, Shen LY, Xu JG, Gu YD. Fib-
malities allow conclusions to be drawn about rillation potential amplitude to quantitatively assess
the type and severity of disease, its duration denervation muscle atrophy. Neuromuscul Disord
or stage of evolution, and the likely anatom- 10:85-91, 2000.
ical location of the pathologic process. 16. Dumitru D, KingJC, Stegeman DF. Endplate spike
morphology: a clinical and simulation study. Arch
Phys Med Rehabil 79:634-640, 1998.
17. Schulte-Mattler WJ, Georgiadis D, Tietze K, Zierz S.
REFERENCES Relation between maximum discharge rates on
electromyography and motor unit number esti-
1. Caruso G, Eisen A, Stalberg E, et al. Clinical EMG mates. Muscle Nerve 23:231-238, 2000.
and glossary of terms most commonly used by clin- 18. Sun TY, Chen JJ, Lin TS. Analysis of motor unit fir-
ical electromyographers. The International Fed- ing patterns in patients with central or peripheral
eration of Clinical Neurophysiology. Electroen- lesions using singular-value decomposition. Muscle
cephalogr Clin Neurophysiol Suppl 52:189-198, Nerve 23:1057-1068, 2000.
1999. 19. Gunreben G, Schulte-Mattler W. Evaluation of mo-
2. Stalberg E, Falck B, Gilai A, JabreJ, Sonoo M, Tod- tor unit firing rates by standard concentric needle
nem K. Standards for quantification of EMG and electromyography. Electromyogr Clin Neurophys-
neurography. The International Federation of Clin- iol 32:103-111, 1992.
ical Neurophysiology. Electroencephalogr Clin 20. Conwit RA, Tracy B, Cowl A, et al. Firing rate anal-
Neurophysiol Suppl 52:213-220, 1999. ysis using decomposition-enhanced spike triggered
3. Drost G, Blok JH, Stegeman DF, van Dijk JP, van averaging in the quadriceps femoris. Muscle Nerve
Engelen BG, Zwarts MJ. Propagation disturbance of 21:1338-1340, 1998.
motor unit action potentials during transient pare- 21. Bischoff C, MachetanzJ, Conrad B. Is there an age-
sis in generalized myotonia: a high-density surface dependent continuous increase in the duration
EMG study. Brain 124:352-360, 2001. of the motor unit action potential? Electroen-
4. Dumitru D, KingJC, Nandedkar SD. Motor unit ac- cephalogr Clin Neurophysiol 81:304-311, 1991.
tion potential duration recorded by monopolar and 22. Buchthal F, Pinelli P, Rosenfalck P. Action poten-
concentric needle electrodes. Physiologic implica- tial parameters in normal human muscle and their
tions. Am J Phys Med Rehabil 76:488-493, 1997. physiological determinants. Acta Physiol Scand
5. Chan RC, Hsu TC. Quantitative comparison of mo- 32:219-229, 1954.
tor unit potential parameters between monopolar 23. Fuglsang-Frederiksen A. Quantitative electromyog-
and concentric needles. Muscle Nerve 14:1028- raphy. I. Comparison of different methods. Elec-
1032, 1991. ' tromyogr Clin Neurophysiol 27:327-333, 1987.
6. Stalberg E. Macro EMG, a new recording tech- 24. Fuglsang-Frederiksen A. Quantitative electromyog-
nique. J Neurol Neurosurg Psychiatry 43:475-482, raphy. II. Modifications of the turns analysis. Elec-
1980. tromyogr Clin Neurophysiol 27:335-338, 1987.
25. Fuglsang-Frederiksen A, Ronager J. EMG power 33. Hart IK. Acquired neuromyotonia: a new autoanti-
spectrum, turns-amplitude analysis and motor unit body-mediated neuronal potassium channelopathy.
potential duration in neuromuscular disorders. J Am J Med Sci 319:209-216, 2000.
Neurol Sci 97:81-91, 1990. 34. Liguori R, Fuglsang-Frederiksen A, Nix W, Fawcett
26. Pfeiffer G, Kunze K. Turn and phase counts of in- PR, Andersen K. Electromyography in myopathy.
dividual motor unit potentials: correlation and re- Neurophysiol Clin 27:200-203, 1997.
liability. Electroencephalogr Clin Neurophysiol 35. Pfeiffer G. The diagnostic power of motor unit po-
85:161-165, 1992. tential analysis: an objective bayesian approach.
27. Dumitru D. Configuration of normal and abnormal Muscle Nerve 22:584-591, 1999.
non-volitional single muscle fiber discharges. Clin 36. Akaboshi K, Masakado Y, Chino N. Quantitative
Neurophysiol 111:1400-1410, 2000. EMG and motor unit recruitment threshold using
28. Dumitru D, King JC. Hybrid fibrillation potentials a concentric needle with quadrifilar electrode.
and positive sharp waves. Muscle Nerve 23:1234- Muscle Nerve 23:361-367, 2000.
1242, 2000. 37. Lang AH, Partanen VS. "Satellite potentials" and
29. Kraft GH. Fibrillation potential amplitude and mus- the duration of motor unit potentials in normal,
cle atrophy following peripheral nerve injury. Mus- neuropathic and myopathic muscles. J Neurol Sci
cle Nerve 13:814-821, 1990. 27:513-524, 1976.
30. Nandedkar SD, Barkhaus PE, Sanders DB, Stalberg 38. Buchthal F. Electrophysiological signs of myopathy
EV. Some observations on fibrillations and positive as related with muscle biopsy. Acta Neurol (Napoli)
sharp waves. Muscle Nerve 23:888-894, 2000. 32:1-29, 1977.
31. Roth G. Repetitive discharge due to self-ephaptic 39. Sonoo M, Stalberg E. The ability of MUP parame-
excitation of a motor unit. Electroencephalogr Clin ters to discriminate between normal and neurogenic
Neurophysiol 93:1-6, 1994. MUPs in concentric EMG: analysis of the MUP
32. de Carvalho M, Swash M. Fasciculation potentials: "thickness" and the proposal of "size index." Electro-
a study of amyotrophic lateral sclerosis and other encephalogr Clin Neurophysiol 89:291-303, 1993.
neurogenic disorders. Muscle Nerve 21:336-344, 40. Johnston SC. Prognostication matters. Muscle
1998. Nerve 23:839-842, 2000.
Chapter 25
QUANTITATIVE
ELECTROMYOGRAPHY
CHARACTERISTICS OF THE MOTOR MANUAL ANALYSIS OF SINGLE MOTOR

UNIT POTENTIAL UNIT POTENTIALS
CHARACTERISTICS OF THE RECORDING COMPUTER-ASSISTED QUANTITATIVE
EQUIPMENT ANALYSIS OF MOTOR UNIT
PROPERTIES OF MOTOR UNIT POTENTIALS
POTENTIALS WITH STANDARD AUTOMATED METHODS OF ANALYSIS OF
ELECTRODES MOTOR UNIT POTENTIALS
MOTOR UNIT POTENTIAL COMPLEXITY DECOMPOSITION OF THE INTERFERENCE
(PHASES, TURNS, CROSSINGS) PATTERN INTO CONSTITUENT
PROPERTIES OF MOTOR UNIT MOTOR UNIT POTENTIALS
POTENTIALS MEASURABLE ONLY TURNS AND AMPLITUDE ANALYSIS OF
WITH SPECIAL ELECTRODES THE INTERFERENCE PATTERN
PROPERTIES OF INTERFERENCE POWER-SPECTRUM ANALYSIS
PATTERN SUMMARY
The results of the needle examination, or long-duration motor unit potentials (MUPs)
electromyography (EMG), are interpreted in in a few muscles. Also, in some diseases the
most laboratories by the subjective method EMG findings are mixed, with both large and
or semiquantitative EMG (see Chapter 24). small MUPs. In these cases, more objective
This method is fast and efficient if the elec- measurements of MUP variables might be
tromyographer is well-trained and experi- more precise and resolve the uncertainties
enced. Inexperienced examiners are more the examiner may have.
prone to mistakes.1"3 Examiner bias is easily These problems have led to efforts to ap-
introduced in a subjective analysis. Some ply quantitative analysis to the results of
cases that are encountered regularly in the electrophysiologic tests. The quantitative ap-
practice of EMG are not adequately defined proach is to measure variables of the electric
by the usual methods of semiquantitative activity of muscle fibers and motor units as
analysis. The difficult cases usually are those precisely as possible and to record the nu-
of mild or equivocal disease in which the ex- merical values derived from these measure-
aminer, after the usual subjective analysis of ments. The numerical data can be graphed
the data, is in doubt about whether the find- and subjected to statistical and computer
ings are minimally abnormal or normal. Bor- analysis in many different ways. Data from
derline results are obtained most frequently groups of normal subjects can then be de-
from patients who may have mild myopathy rived and compared with data from patients
or possible radiculopathy and questionably with disease.
324
Quantitative Electromyography 325
The potential advantages of qiiantitation ence pattern. This chapter considers only
are numerous. Quantitation should allow a quantitative analysis of the findings on nee-
clear distinction between patients with neu- dle EMG.
romuscular disease and those without neu-
romuscular disease, and patients with my-
opathies should be distinguishable from
those with neurogenic disorders or defects CHARACTERISTICS OF THE
of neuromuscular transmission. Quantita- MOTOR UNIT POTENTIAL
tion should allow a distinction to be made
among degrees of severity of disease in dif- The MUP, sometimes referred to as the mo-
ferent patients with the same disease or in a tor unit action potential, is the sum of the po-
single patient at different times during the tentials of muscle fibers innervated by a sin-
course of the disease and its treatment. The gle anterior horn cell, or lower motor
technique should eliminate examiner bias. neuron, that are within the recording area
Quantitative EMG should produce repro- of the electrode. The characteristics of the
ducible results when performed by the same MUP are the result of several complex fac-
examiner at different times and by different tors, particularly the characteristics of the
examiners in the same or different institu- motor unit. The number of motor units in
tions. Comparisons of the results in differ- the muscle varies from muscle to muscle. Es-
ent laboratories should be more reliable, timates range from 100—500 motor units in
and reporting of normal and abnormal re- a normal muscle in a human extremity.4 Dif-
sults in the literature should be more ferent types of lower motor neurons (types
amenable to statistical analysis and critical I and II) make up the 100-500 motor units.
analysis by others in the field. An ideal In general, type I lower motor neurons are
method of quantitation of the EMG should activated at low levels offeree, have a smaller
be less time consuming, relatively easy to per- cell body and motor axon, and have a
form, improve accuracy and reliability, be in- smaller MUP. Type II lower motor neurons
expensive, and be applicable to all sites and are larger, have a larger axon and a larger
muscles. The technique should sample a MUP, and are activated at higher levels of
large number of motor units. Ideally, such force.5'6
techniques should examine EMG activity in The number of muscle fibers innervated
different ways and allow the extraction of by each lower motor neuron is termed the
data that has not been available with previ- innervation ratio, and it varies from neuron
ous methods. to neuron. In the extraocular muscles, there
In practice, quantitation often has disad- are 9 muscle fibers per motor unit, com-
vantages. These include increased cost, need pared with 1900 muscle fibers per motor
for special equipment, and extra time to per- unit in the gastrocnemius muscle.7 The mus-
form the examination. The introduction of cle fibers of one motor unit are distributed
new techniques requires the establishment over 5-10 mm of the cross-sectional diame-
of new normative databases. Also, the tech- ter of the muscle. This is called the motor unit
niques must be shown to be superior to the territory. The number of muscle fibers per
best method available for difficult cases given cross-sectional area in the motor unit
with subtle abnormalities. Nerve conduction territory is the fiber density. The motor unit
studies, evoked potentials, and needle ex- territory of one motor neuron contains mus-
amination studies have all been quantified. cle fibers from other motor units.
Spontaneous activity and voluntary activity The synchrony of firing of the muscle
on needle EMG are the most difficult to fibers of a motor unit influences the MUP
quantify because of the complexity and characteristics. The synchrony of firing is de-
rapid recurrence of the waveforms. The determined by the (1) length, diameter, and
velopment of analog-to-digital conversion of conduction velocity of the motor nerve ter-
needle examination data and the availability minals, (2) the location and function of the
of low-cost, fast microcomputers are leading neuromuscular junctions, (3) the diameter,
to the development of automated tech- conduction velocity, and membrane charac-
niques of evaluating MUPs and the interfer- teristics of the muscle fibers, (4) tempera-
ture, and (5) arrangement of muscle fibers CHARACTERISTICS OF THE

in the motor unit. RECORDING EQUIPMENT
In most muscles, a muscle fiber has only
one neuromuscular junction, or end plate. The characteristics of the recording elec-
Exceptions are the extraocular muscles and trode determine many of the properties of
the extensor digitorum communis muscle in the recorded activity. Several different types
the forearm.8 The location of the neuro- of recording electrodes are available (Fig.
muscular junction is termed the end plate 25-1). The concentric and monopolar nee-
zone, and it varies from muscle to muscle. In dle electrodes are used most commonly. The
the biceps brachii muscle, the end plate standard concentric needle electrode is con-
zone is an irregular V-shaped band 5 mm structed with a wire approximately 150 ;nm
wide, but in the deltoid, it forms an irregu- in diameter inserted down the center of a
lar sinusoidal pattern across the muscle. In hollow cannula. The wire is separated from
the anterior tibial muscle, the end plate zone the cannula by an insulator. The tip of the
is at the periphery and is cone-shaped. electrode is ground to a 15-degree angle,
The characteristics of the action potentials producing an exposed surface of the center
generated by individual muscle fibers are fun- wire that is a 150 /xm by 580 /mi ellipse. The
damental to die characteristics of the MUP. 0.07 mm recording surface of the central
Muscle cell diameter and conduction veloc- wire provides a very stable recording surface.
ity (1.5—6.5 m/second)9 affect the muscle- The activity at the exposed tip of the central
fiber action potentials that summate to form wire is connected to the Gl input of the dif-
the MUP. The amount and characteristics of ferential amplifier, and the recording sur-
the tissue between the measuring electrode face of the cannula is connected to the G2
and the discharging muscle fibers, including input. The recording area is smaller and di-
the amount of connective tissue, blood ves- rectional compared with the monopolar
sels, and fat, affect the MUP. The tissue be- electrode. Motor unit potential amplitudes
tween the source of the MUP and the record- are smaller and the duration shorter than
ing electrode acts as a high-frequency filter. those measured with a monopolar electrode.
It diminishes high-frequency signals and rel- Because the recording territory is smaller,
atively enhances low-frequency activity. fewer fibrillations, fasciculations, and com-
Figure 25-1. Various types of needle electrodes used in quantitative electromyography.

plex repetitive discharges are recorded. for patients. Because of their larger pick-up
There is less recording noise and the record- area, monopolar electrodes may detect
ing electrode surface area is more constant more fibrillation potentials, fasciculation,
than that of a monopolar needle. Motor unit and other spontaneous activity.
potentials recorded from deeper in the mus- Single fiber EMG electrodes are con-
cle appear larger than those recorded su- structed of a very thin wire in a hollow can-
perficially. The examination with a concen- nula. The wire is insulated from the cannula
tric needle electrode may be slightly more and exposed through a hole in the shaft of
painful than with a monopolar electrode. the cannula. The recording surface is small
The so-called monopolar electrode is a (25 ^m). The small recording area in com-
needle insulated except for the 0.56-0.80 bination with higher low-frequency filtering
mm2 conical tip that serves as the recording (500 Hz) allows individual muscle fibers of
surface and is connected to the Gl input of the motor unit to be recorded. This is dis-
the differential amplifier of the EMG ma- cussed in detail in Chapter 26.
chine. The G2 electrode is a surface elec- The filter settings of the recording device
trode. The recording surface is larger than can alter the appearance of the MUP. Elec-
that of the concentric needle electrode, but tromyographic signals are contaminated by
the active electrode recording area may be low-frequency activity from motion and sur-
more variable if the insulation is peeled rounding distant MUP activity. This is par-
back. Monopolar electrodes have a larger ticularly prominent with low-frequency filter
pick-up area that is multidirectional. Motor settings of 2 Hz. Increasing the low-fre-
unit potentials recorded with a monopolar quency filter setting to 30 Hz eliminates
needle electrode have longer durations and much of the noise, and settings of 500 Hz
higher amplitudes than the same potentials eliminate almost all of it. However, increas-
recorded with a concentric needle elec- ing the low-frequency filter from 2 to 500 Hz
trode. The recordings also contain more alters drastically the variables of the MUPs.
noise from distant motor units. Frequently, Increasing the amount of low-frequency fil-
the recording is contaminated by activity tering may add extra components to the
from the surface G2 electrode. Monopolar MUP such as the terminal negative afterpo-
electrodes are less expensive and less painful tential (Fig. 25-2). Stalberg et al.7 noted that
Figure 25-2. Effects of varying low-frequency filter setting on recorded motor unit potential (MUP) morphology.
The waveforms on the left are from a single MUP recorded with a low-frequency filter of 2 Hz. The top tracing is
an average of the triggered superimposition of 5 recurrences of the same MUP in the lower half. The waveforms
on the right are the same MUP recorded at the same position with a low-frequency filter setting of 30 Hz. Note
changes in configuration, particularly the introduction of a new phase, the negative afterpotential.
Figure 25-3. Effects of different

sensitivities on measured duration
of the same motor unit potential.
At higher sensitivities, measured
duration is longer.
the negative afterwave that follows the return component of the MUP. Duration is mea-
to baseline of the positive afterwave is an ar- sured in milliseconds and includes the main
tifact generated by the capacitance of the spike and the initial and terminal parts (Fig.
low linear frequency filter. Stalberg and as- 25-4). The deviation of the initial and ter-
sociates considered it an artifact that can be minal components from the baseline is of-
minimized with a low linear frequency filter ten very gradual and difficult to define. In
setting of 2 Hz and should be ignored in distal leg muscles, MUP duration, ampli-
measurements of duration. tude, and number of turns increase with age.
The sensitivity settings of the amplifier In proximal muscles and distal arm muscles,
have a marked effect on measurements of MUP variables are affected less significantly
MUP variables, particularly measurements of by age. The greatest change in duration oc-
duration (Fig. 25-3). The greater the sensi- curs after age 60 years. Duration also in-
tivity, the longer the apparent duration of creases by 5%-10% per degree centigrade
the MUPs. This secondarily affects other decrease in temperature.
measures, such as area, thickness, and size The duration of the MUP reflects the ac-
index. If the sensitivity is too great, the MUP tivity of muscle fibers of the motor unit that
overloads the amplifier and the amplitude are within 2.5 mm of the recording electrode
cannot be measured. If the sensitivity is set (Fig. 25-5). If the muscle fiber density or the
too low, the MUP may not be detected. territory of the motor unit changes, the du-
Amplifier characteristics such as input im- ration of the MUP will change. In general,
pedance, inherent noise level, amplifier re- an increase in muscle fiber density or terri-
covery time, analog-to-digital sampling rate, tory of the motor unit will increase duration.
signal-to-noise ratio, and common mode re- A decrease in muscle fiber density or terri-
jection affect the characteristics of the tory of the motor unit will decrease duration.
recorded EMG activity. Such variables limit Motor unit potential duration is increased if
the reliability and reproducibility of data ob- there is increased variation in the diameter,
tained from different equipment.10 length, or conduction velocity of the nerve
terminal or muscle fiber. Increased distance
between the recording electrode and neu-
PROPERTIES OF MOTOR UNIT romuscular junctions also increases dura-
POTENTIALS WITH STANDARD tion. The tissue between the muscle fibers
ELECTRODES and recording electrodes function as a high-
frequency filter with capacitance and resis-
The duration of the MUP is the time between tance. Because slower frequency activity is
the starting point and end point of the slow transmitted farther through connective tis-
Quantitative Electromyogr 329
figure 25-4. Commonly measured variables of the motor unit potential. See text for explanations.
sue, muscle fibers at the periphery of the tor unit) within 1.5-2 mm of the core of the
pick-up area of the recording electrode pri- concentric needle electrode contributes to
marily generate low-frequency, slow initial the area of the MUP measured.11 The am-
and terminal components of the MUP. plitude is measured from the maximual pos-
The area of the MUP is the space under itive peak to the maximal negative peak.
the curve of the waveform (Figs. 25-3 and Amplitude is determined by the action po-
25-4). It reflects the amount of functioning tentials of the 2-12 muscle fibers, sometimes
muscle near the electrode better than does only 1 or 2, of the motor unit within 500 fim
the amplitude or other variables. Digital of the electrode (Fig. 25-4). Although the
analysis allows accurate measurement of the diameter of the muscle fibers and the fiber
area. Simulation studies suggest that the ac- density affect amplitude, it is more depen-
tivity of 15-20 muscle fibers (of a single mo- dent on the proximity of the electrode to the
Figure 25-5. Relative size of the terri-

tory of muscle fibers of a single motor
unit compared with the pick-up area of
a concentric needle electrode. The
amplitude of the motor unit potential
is determined by muscle fibers of the
motor unit within 500 /u,m of the
recording surface. The area is deter-
mined by muscle fibers within 1.5-
2 mm and duration by muscle fibers
of 2.5 mm. (Data from Nandedkar
etal.11)
Figure 25-6. Illustration of very

gradual return to baseline of the
terminal component of a motor
unit potential. The end is deter-
mined more easily when the
baseline is displayed, as in this
illustration.
muscle fibers than the duration or area is. frequency filter setting of 2 Hz and should
Thickness is a MUP property derived by di- be ignored in measurements of duration. If
viding the area by the amplitude. Some in- the terminal negative afterpotential is in-
vestigators have found that for the detection cluded in measurements, duration will be ar-
of myopathies, decreased thickness is the tificially prolonged.
most sensitive variable of motor unit mor-
phology.11
Rise time is the duration of the rising phase
of the spike from the positive peak to the MOTOR UNIT POTENTIAL
negative peak (Fig. 25-4). Rise time, or rise COMPLEXITY (PHASES,
rate, is the best indicator of how close the TURNS, CROSSINGS)
source of the potential is to the recording
electrode. Most commonly, the overall shape of the
Spike duration is the time from the first pos- MUP waveform is triphasic, with an initial
itive peak to the last positive peak of the slow downward deflection followed by a
main component of the MUP. The spike is large, rapid, upward deflection or spike, and
the sum of the action potentials of usually then a slow downward deflection called the
fewer than 15 (1-8) muscle fibers closest to terminal component (Fig. 25-4). If the record-
the recording electrode.6 The spike may ing electrode is close to the end plate zone,
have linked potentials or satellites separate the initial slow component is lost and the po-
from the main spike. The terminal component tential is biphasic, beginning with the up-
of the MUP is measured from the end of the ward spike component. Monophasic MUPs,
main spike to the return to baseline (Fig. with only a single upward or negative phase,
25-4). The terminal component is gener- are less common. An MUP may have the con-
ated by action potentials traveling away from figuration of a "positive" wave, with an ini-
the recording electrode along muscle fibers. tial large downward "positive" component
Marking the end of the slow return of the followed by a low, long, late "negativity."
terminal component to the baseline is the Such "positive" MUPs are recorded from
most inconsistent measurement in semi- damaged muscle fibers, the ends of the
quantitative and quantitative EMG (Fig. fibers or the tendon, or the cannula (G2
25-6). The imprecision in marking the end electrode) of the concentric needle elec-
of the terminal component results in major trode. Such positive potentials should be ex-
variations in quantitative analysis of MUP cluded from analysis of MUP properties.
morphology. Large amounts of low-fre- The number of components of the wave-
quency filtering (10-30 Hz) may add an ex- form above or below the baseline are con-
tra component to the MUP, called the ter- sidered phases (Fig. 25-4). If there are more
minal negative afterpotential, that follows the than four phases, the MUP is considered
return to baseline of the positive afterwave polyphasic. The number of phases can be de-
(Fig. 25-2). It is an artifact generated by termined by counting the portions of the
the capacitance of the low linear frequency waveform above or below the baseline or by
filter and can be minimized with a low- determining the number of baseline cross-
Figure 25-7. Example of a single large-amplitude, long-duration motor unit potential firing at very rapid rates and
randomly varying in amplitude from one spike to the next.
ings and adding one. Normal muscles may

have 5%-15% polyphasic motor units. An in-
crease in the percentage of polyphasic mo-
tor units is a very sensitive but very nonspe-
cific indicator of neuromuscular disease.
A turn is a peak in the waveform of the
MUP (Fig. 25-4). The number of turns is
determined by counting the number of pos-
itive and negative peaks separated from the
preceding potential by some arbitrary
amount, usually 50 fjN or more. If the MUP
contains more than five turns, it is termed
complex, or serrated. Satellite potentials are late
components of the MUP, separated from the
main component by a segment of flat base-
line (Fig. 25-4). Many different terms have
been used for satellite potentials, including
coupling discharges, parasites, and linked poten-
tials. They usually follow the main compo-
nent, but they may precede it. Such satellite
potentials are typically excluded from mea-
surements of MUP duration. Satellite po-
tentials can be seen in l%-3% of MUPs of
normal muscles. Lang and Partanen12 found
these potentials in 10% of normal muscles,
12% of cases of neuropathies, 60% of cases
of old poliomyelitis, and 45% of cases of
myopathies.
Motor unit potential variation is a general
term that refers to a change in amplitude or
configuration of the MUP with successive
discharges (Fig. 25-7). Jiggle refers to varia-
tion in the position of phases or turns in the
MUP relative to each other from one dis-
charge of the MUP to another (Fig. 25-8).13
These two terms reflect an instability of con- Figure 25-8. Multiple firings of a single large-amplitude,
long-duration, polyphasic complex motor unit poten-
duction along nerve terminals, across the tial. Some of the components block (second trace
neuromuscular junction, or along muscle down) and others jiggle back and forth relative to each
fibers. Although classically thought of as a other.
feature of disorders of the neuromuscular The activity at the single fiber electrode trig-
junction, MUP variation may be less specific. gers the sweep, and the MUP of the motor
Firing rate and recruitment are important unit is recorded at the concentric needle.
factors for measurement in quantitative The concentric needle is then advanced
analysis, and they are discussed in detail in through the muscle until no activity is
Chapter 24. recorded from the motor unit. Next, the
The evaluation of motor unit variables is concentric needle is connected to a stepper
subject to bias even with quantitative EMG. motor that withdraws the concentric needle
If a trigger line and delay are used, the sin- through the muscle in small steps. The MUP
gle trigger tends to bias the examination to- is recorded and averaged at each site and
ward larger potentials. An excessive number the needle is withdrawn another step and
of units firing at an excessive force level bi- the process is repeated until the concentric
ases the examination toward the larger units. needle is withdrawn from the territory of the
Because MUPs appear larger when the nee- motor unit under study. Scanning EMG
dle is deeper in the muscle, the depth of nee- demonstrates that the shape of the MUP
dle insertion can skew the results. The slow varies considerably within the motor unit ter-
initial and terminal components of the MUP ritory. There may be one, two, or more dis-
must be marked with special care. The neg- tinct areas of activity, sometimes occurring
ative afterpotential should be excluded from with different latencies. These have been
measurements. called motor unit fractions, and they probably
are generated by groups of muscle fibers,
each innervated by one major intramuscular
PROPERTIES OF MOTOR UNIT axonal branch. Scanning EMG allows mea-
surement of the size of the motor unit and
POTENTIALS MEASURABLE the density and distribution of muscle fibers
ONLY WITH SPECIAL within the motor unit territory.
ELECTRODES Macro-EMG was developed by Stalberg16
in 1980 to record the activity of most of the
The size of the motor unit territory was mea- muscle fibers of a single motor unit. The
sured by Buchthal et al.,14 with a multilead concentric needle records activity from 2 to
electrode that had recording sites along the 3 mm of the 5-10 mm diameter of the limb
length of a needle electrode, in normal sub- motor units. A macro-EMG needle is a mod-
jects and in patients with diseases. The ter- ified single fiber EMG needle, with the can-
ritory of motor units in limb muscles was cir- nula insulated except for the terminal 15
cular in cross section, with a diameter of mm (Fig. 25-1). This 15 mm length serves
5.1 mm in the biceps brachii and 10 mm in as the recording surface for the macro-EMG.
the rectus femoris. Motor unit territory di- In most limb muscles, the muscle fibers of a
ameter and fiber density tended to be re- single motor unit are scattered over a cross-
duced in myopathies. sectional area of 5—10 mm. The 15 mm
The distribution of electric activity of recording area of the macro-EMG needle
the motor unit can be mapped in a cross- thus covers the territory of most motor units
sectional plane with the scanning EMG tech- in limb muscles. There is also a 25 jum di-
nique.15 This technique uses a standard con- ameter wire electrode exposed on the shaft
centric needle electrode and a standard sin- of the terminal part of the cannula 7.5 mm
gle fiber needle electrode. The activity from from the tip that records single muscle-fiber
the two needles is recorded on different activity. Recordings are made on two chan-
channels of the EMG machine with differ- nels. The first channel records activity from
ent filter and sensitivity settings. The single the 15 mm bare shaft (Gl electrode) and a
fiber needle electrode is inserted into the surface electrode (G2). The second channel
muscle and manipulated to a point where records single fiber activity from the 25 /am
the activity of one muscle fiber of a motor wire electrode (Gl) and the shaft (G2). The
unit is recorded. The concentric needle elec- needle is inserted into the muscle and ma-
trode is inserted at a nearby point, perpen- nipulated during minimal levels of contrac-
dicular to the course of the muscle fibers. tion to a position at which the action po-
tential of a single muscle fiber is recorded creasing the sweep speed. Other special
from channel 2. The activity in channel 2 recording techniques can measure muscle-
acts as a trigger, and the activity from chan- fiber conduction velocity, contraction time,
nel 1 is then recorded and averaged over twitch time and tension, and the effects of
60-80 ms. Next, the needle is moved to a dif- fatigue, but these are not considered in this
ferent site in the muscle and the process is chapter.
repeated. Usually, 20 different potentials are
recorded from 20 different sites. Normal val-
ues have been established.17 PROPERTIES OF INTERFERENCE
In neurogenic disorders characterized by PATTERN
denervation and reinnervation, the ampli-
tudes of the macro-EMG potentials generally At minimal levels of muscle contraction
are increased. In disorders of muscle or my- producing low levels of force, individual
opathies, the amplitudes are often low, par- MUPs can be identified and measured (Fig.
ticularly in subacute myopathies. In chronic 25-9). As the level of force produced in-
or long-standing myopathies, the amplitudes creases to the maximal voluntary contraction
may be increased. force, the number and firing rate of the ac-
Fiber density, blocking, and jitter are tive motor units increase. At the same time,
characteristics of the MUPs optimally evalu- the size of the individual MUPs increases. As
ated with single fiber EMG. Similar mea- the number of MUPs increases, it becomes
surements can be made with a standard con- difficult and ultimately impossible to iden-
centric needle electrode by increasing the tify individual potentials with standard elec-
low-frequency filter setting to 500 Hz and in- trodes. The activity, then, is described as an
Figure 25-9. The top tracing shows a partial interference pattern. The bottom tracing, an exploded view of a short
segment of the top tracing, illustrates some of the properties measured in quantitation of the interference pattern.
interference pattern in which individual MUPs cial equipment. Analysis can be performed
are superimposed on others, resulting in a at fixed forces of 2-5 kg weights or at forces
complex mix of summation and cancellation varying from minimal contraction to maxi-
of the activity of different MUPs.18 Auto- mal contraction force. After the measure-
mated techniques have been developed to ments have been made at one site, the nee-
try to quantify the interference pattern. dle is moved to another distinct site and
The advantages of interference pattern another measurement is made. Ideally, 30
analysis are that it incorporates signals from sites are measured through three or more
more of the muscle under study and samples skin insertions.
activity from motor units that are activated Many different properties of the interfer-
at higher force levels, usually type II motor ence pattern have been measured. The num-
units. The technique samples a much larger ber of turns is measured frequently (Fig.
amount of electric activity. Many of these 25-9). Turns represent a change in signal di-
techniques are available on commercial rection of at least 50 fjiV. Turns indirectly re-
EMG machines. The evaluations are often flect the number of active MUPs, the pro-
fast. Most of the techniques are applicable to portion of polyphasic MUPs, and the MUP
all muscles, and many can be applied to un- firing rate. A turn may reflect a peak in an
cooperative patients, such as small children. MUP, an interaction between overlapping
The disadvantages of interference pattern MUPs, or noise. Baseline or zero crossings
analysis are numerous. The variables mea- are the number of voltage crossings of the
sured cannot be related in a simple and di- baseline per unit time. The time in mil-
rect manner to the properties of the con- liseconds between turns or peaks can be
stituent MUPs. The effects of summation measured as time intervals, or Tis (Fig. 25-9).
and cancellation of superimposed MUP ac- Measurement of the number of short time
tivity are complex and difficult to under- intervals or comparing the number of turns
stand. Large-amplitude potentials obscure with short time intervals of 0-1.5 ms to those
activity from small MUPs, and a few long- with longer time intervals of 1.5-5 ms and
duration polyphasic MUPs may give results 5-20 ms seems to have considerable clinical
similar to many short-duration MUPs. Also, usefulness.20"22
interference pattern analysis systems do not Amplitude is measured as the potential
measure activity less than 50-100 fjV. Tech- difference between successive turns (Fig.
nical factors such as system noise, filters, and 25—9). Cumulative amplitude is the total am-
analog-to-digital conversion rates can inter- plitude of turns over a certain time. Divid-
fere. Normal values are available for a lim- ing the cumulative amplitude for a fixed
ited number of muscles, but their usefulness time interval by the number of turns during
in mild or borderline cases has not been that same interval defines mean amplitude.
evaluated carefully. The problem of enter- The data have also been expressed as a ra-
ing data that are not well-analyzed by the ex- tio called the turns/amplitude ratio, which is
aminer into a program that carries out com- derived from the number of turns for a cer-
plex analyses and provides data difficult to tain time interval divided by the mean am-
check by other means introduces uncer- plitude for that same interval. Others have
tainty. measured the maximal value of the turns/
The basic techniques of interference pat- amplitude ratio for all the sites tested and
tern analysis are straightforward. A needle called this the peak ratio. Peak ratio appears
electrode or surface recording electrode to be a useful measurement for distinguish-
can be used, but the surface electrode tech- ing between normal subjects and patients
niques are not clearly useful.19 The needle with neuromuscular disease.
electrode is inserted into an area of the mus-
cle where MUPs have short rise times. In-
terference pattern analysis is performed at MANUAL ANALYSIS OF SINGLE
different force levels in different systems. MOTOR UNIT POTENTIALS
The test can be performed at variable or
fixed percentages of maximal voluntary con- Historically, quantitative EMG began with
traction (usually 30%), but this requires spe- the measurement of MUP characteristics by
manual analysis of tracings or photographs COMPUTER-ASSISTED

of MUPs.23"27 This technique required min- QUANTITATIVE ANALYSIS OF
imal activation of preferably only one but at
most three MUPs at a time. This limited the MOTOR UNIT POTENTIALS
evaluation to activity of type 1 muscle fibers
and MUPs. The needle electrode was posi- The introduction of triggering and delay
tioned to obtain fast rise times. Photographs techniques permitted the sampling of one
were taken of several recurrences of the MUP over and over, even if other motor
same potential and measurements were units were active.28 This allowed more rapid
made. The amplitude was recorded as the collection of data, with an improved signal-
maximal deflection on the screen possible to to-noise ratio. Digitized signals allowed com-
prevent the loss of low-amplitude compo- puter averaging and storage of waveforms
nents and to prevent blocking of the largest and the measurement of variables, such as
component of the MUP. Motor unit poten- area and thickness of the motor unit, not
tials less than 50 />tV in amplitude were ex- readily available with the previous tech-
cluded. The low-frequency filters were set at niques. Such analyses are not directly com-
2 Hz and high-frequency filters at 10,000 Hz. parable to the results of the manual method
From 20 to 40 different recording sites per of Buchthal et al.23~25>27 and Petersen and
muscle were evaluated, with the sites sepa- Kugelberg.26 The newer technique gives a
rated by at least 3 mm. The technique was sampling bias favoring larger MUPs.
accurate but time-consuming. Normal values A commonly used technique for automated
have been published for a large number of analysis of single MUPs with computer-aided
muscles. These data have been widely used methods is the quantitative EMG (QEMG)
for many years as the reference standard. program (Nicolet) (Fig. 25-10). To use the
Figure 25-10. Computer-assisted quantitation of motor unit potentials. The different waveforms represent differ
ent motor unit potentials. The measurements are summarized in the lower right corner (see text).
original Buchthal normative data, 2 Hz-10 two trigger lines, with sampling of the
kHz filter settings, degree of minimal acti- smaller MUPs. The MUP fires repetitively,
vation, and standard sensitivity settings and multiple traces are superimposed (5-15
should be employed. Ideally, MUPs should work best). At least 20 MUPs are collected
have rise times less than 500 microseconds, from different positions in the muscle. Dif-
although many experts think that a rise time ferent positions should be sampled to avoid
of 1000 microseconds is acceptable. The ex- recording repeatedly from the same MUP
aminer must be able to hold the needle im- and to increase the likelihood of finding ab-
mobile during data collection. The patient normalities localized to one part of the
must be able to activate only one to three muscle.
MUPs at a time and to maintain a steady fir- An important issue with these systems is
ing rate. The trigger must be adjusted to iso- how often the examiner must correct the
late individual units. Some bias is intro- measurements made by the automated sys-
duced by triggering with the largest unit, tem (Fig. 25-11). The examiner must always
causing the selection of large-amplitude check the automated markers. Marking the
long-duration MUPs. This can be avoided by gradual onset and termination of the MUP
using minimal activation and a dual trigger is frequently subject to variability and error
line to select the smaller units as well. The (Fig. 25-6). The duration marked is usually
dual trigger uses two lines that can be ad- greater at higher sensitivities, so the sensi-
justed together to select peaks between the tivity should be close to the 100 /iV/division
Figure 25-11. Failure of the computer to mark correctly the duration of a motor unit potential because of a mod-
erate level of background noise.
that is generally accepted when marking du- DECOMPOSITION OF THE

ration (Fig. 25-3). Background noise must INTERFERENCE PATTERN
be minimized. Relaxation of neighboring INTO CONSTITUENT MOTOR
muscles is particularly important.
The analysis generates a report of the UNIT POTENTIALS
properties of all the MUPs recorded and a
separate listing of the properties of the sim- A promising technique is the decomposition
ple and complex MUPs. The properties in- of the interference pattern into component
clude the duration, amplitude, turns, and MUPs,29 which extracts individual MUPs
percentage of polyphasic MUPs and their from an interference pattern. The technique
mean, median, mode, variation, minimal- decomposes or breaks the interference pat-
maximal values, standard deviations, and tern into constituent MUPs so that the mor-
confidence intervals. phologic features and firing pattern of the
MUPs can be determined. Such a technique
could gather large amounts of data quickly,
sample multiple sites and MUPs that are re-
AUTOMATED METHODS OF cruited later at greater force (type II MUP),
ANALYSIS OF MOTOR and study MUP recruitment and firing
UNIT POTENTIALS patterns.
Dorfman and McGill30 described a pro-
Automated methods that identify MUPs gram called Automatic Decomposition EMG, or
and measure them with minimal human ADEMG. The technique uses standard EMG
decision-making have been developed. needle electrodes and evaluates a single-
However, the examiner should be able to channel interference pattern at steady iso-
examine and to edit the raw data. The sys- metric contraction with forces up to 30% of
tem developed by Stalberg et al.28 auto- maximal voluntary contraction. The analysis
matically detects events that may be MUPs. is performed in nearly real time (less than 1
It segments the record into epochs with minute). The program can extract up to 15
and without MUPs usually on the basis of simultaneously active MUPs at a single site
an amplitude threshold. Presumptive (practically, 4-8). Later recruited MUPs have
MUPs are then sorted and classified by different characteristics from the initially re-
comparing their wave shapes sequentially cruited MUPs. The mean MUP amplitudes
(template matching) using predetermined are significantly greater and increase with in-
match criteria. An MUP is accepted as an creasing force. The mean MUP duration is
MUP when an arbitrary number of recur- shorter and declines with increasing force.
rences (2-10) are identified. The recur- The number of turns increases with increas-
rences may be averaged to produce a po- ing force. The mean firing rates are linearly
tential less affected by random noise. The related to contractile force. The mean am-
waveforms that do not recur are considered plitude, duration, and number of turns in-
to represent either noise or superimposi- crease with increasing age, but mean MUP
tion of more than one MUP and are not firing rates decrease with age. The results are
accepted. The properties of the identified presented in numeric and graphic format.
MUP are then measured and accumulated. The results cannot be compared with the
Several different MUPs can be collected at manual quantitative normative data. Auto-
each site. Graphs and reports of the data matic Decomposition EMG is limited in that
can be viewed or printed out. Presumably, MUPs smaller than 100 /iV, unstable MUPs,
a sample of MUPs can be collected quickly. and MUPs widi slow rise times cannot be
The examiner should review the individual recorded. Normative data are available for
MUPs accepted and their markings to en- only a few muscles. Studies demonstrating
sure they are accurate. The collection of the clinical usefulness of ADEMG are few and
data requires a quiet background and acti- small.
vation of only a few MUPs at a time. This LeFever and De Luca31 used multichan-
analysis is time-consuming. nel (three channels) recording from a spe-
cial needle with four recording sites to the deltoid, biceps brachii, first dorsal in-
perform precision decomposition. Precision de- terosseous, vastus lateralis, and anterior tib-
composition is a hybrid visual-computer de- ial muscles. A good correlation has been
composition scheme based on template demonstrated between examinations per-
matching and firing statistics for MUP recog- formed by different examiners, repeat ex-
nition. With this program, it is possible to aminations, and side-to-side comparisons.
identify MUPs at strong levels of contraction The multi-MUAP analysis program takes a
and to evaluate firing patterns and recruit- different approach to defining the limits of
ment. Akaboshi et al.6 have recently used a normal by using outliers, based on the as-
similar technique to decompose the inter- sumption that abnormalities may be limited
ference pattern at forces of up to 50% of to a few MUPs. Such changes may be lost
maximal voluntary contraction. They were when averaged in with other MUPs. The out-
able to determine the firing rates and re- lier limits were determined by determining
cruitment frequencies of the motor units as the third largest and third smallest value of
well as to isolate the units for morphologic a given variable in normal subjects. The
analysis. This technique is too slow for rou- highest and lowest values of these limits for
tine clinical work. the whole control group were chosen as the
Bischoff et al.32 developed a technique of extreme outlier limits. The only outlier limit
multi-motor unit action potential analysis found to change with age was the amplitude
(multi-MUAP analysis) that uses standard of MUP in the anterior tibial muscle but not
concentric needle or monopolar needle other muscles. None of the normal muscles
electrodes. The electrode is inserted into the had more than two values outside the de-
middle part of the muscle at different depths fined limits, using the above criteria.
at three skin insertions. No attempt is made Multi-MUAP analysis detected abnormali-
to position the electrode to record maximal ties in 25 of 31 cases of neuropathy.32 The
amplitude, but the electrode is positioned so size index, amplitude, and duration were the
that at least some MUPs are "sharp" or crisp most frequently abnormal variables. The
and have a short rise time. Force is varied method detected abnormalities in 6 of 8
from slight to moderate muscle contrac- cases of myopathy, with amplitude abnor-
tions, but no special equipment is required malities more common than duration. Out-
to measure force. The baseline should be liers were as sensitive as mean values in
clearly discernible between signals. This neuropathies and more sensitive than
analysis reportedly requires fewer than 4-8 mean values in myopathies. Bischoff et al.32
minutes per muscle. Short segments (5 sec- pointed out that determining mean values
onds) of activity are recorded. At least 20 may miss mild abnormalities of a few MUPs.
MUPs are recorded from each muscle. Mo- An increased number of outliers that indi-
tor unit potentials are classified on the basis cate abnormality can be found after evalu-
of shape variables by a multiple template ating only a few MUPs, making it unneces-
matching technique. A minimum of five re- sary to evaluate 20 or more units and, thus,
currences of each MUP is averaged. From saving time.
two to five MUPs can be recorded at one site. There appear to be advantages of multi-
Motor unit potentials must be larger than 50 MUAP analysis. It allows sampling of a large
jLtV in amplitude and meet a rise criteria of number of MUPs in a short time, is repro-
less than 30 juV/0.1 ms. Duration markers ducible, and allows MUP sampling at levels
can be corrected manually and require man- of contraction greater than threshold. It
ual correction of 25% of the recorded MUP. would seem to reduce examiner bias. In
From 5% to 15% of MUPs need to be re- this author's experience, the analysis takes
jected because of noise. longer than 5 minutes per muscle when edit-
The program automatically measures am- ing time is included.
plitude, duration, spike duration, thickness, Recently, Fang et al.33 introduced a de-
phases, and turns and calculates mean val- composition program that measures the sim-
ues and standard deviations. Reference val- ilarities of waveforms using wavelet domain
ues for different age groups are available for analysis techniques. Although the technique
is too slow for routine clinical use, it illus- patients with myopathies or neuropathies.
trates some of the new methods of signal They used standard concentric needle elec-
analysis that are becoming available.34 trodes inserted at three different sites (prox-
imal, medial, and distal) in the muscle and
recorded activity at a total of 10 different
recording sites at least 5 mm apart. The force
TURNS AND AMPLITUDE of contraction gradually increased from 0 to
ANALYSIS OF THE maximum over 10 seconds and then the pa-
INTERFERENCE PATTERN tient rested 1 or 2 minutes between each
contraction. Of the patients with myopathy,
Probably the best-studied and most widely 92% had abnormally high peak ratio values;
used method of automatic interference pat- the number of time intervals of short dura-
tern analysis was the turns analysis technique tion was increased in 84%. All patients were
developed by Willison.18 Initially, the num- classified correctly by using the peak ratio
ber of turns and the mean amplitude of the and time intervals. In comparison, QEMG
turns were measured from photographs of was diagnostic of myopathy in 72% of cases.
the interference pattern. Later, Fitch and Of patients with neurogenic disorders, 86%
Willison35 developed an electronic analyzer had low peak ratio values. Short time inter-
that could extract the data on-line. It vals were reduced in 48%, and long time in-
counted the number of turns in a given time. tervals were increased. In all patients, the
The amplitude was measured between the diagnosis was made correctly using the com-
turns. This was done at constant forces of 2 bination of peak ratio and time intervals,
or 5 kg. It was found in myopathies that the whereas QEMG was diagnostic in 95%. The
turns per second increased and amplitude technique appears to be objective, fast, and
per turn decreased. In neuronal or axonal reliable, but it takes at least 20 minutes per
loss and reinnervation, amplitude per turn muscle.
increased and turns per second decreased. Stalberg et al.36 and Sanders et al.37 de-
Fuglsang-Frederiksen18 obtained more picted interference pattern analysis graphi-
consistent results using a fractional contrac- cally in a scatter plot (Fig. 25-12) without
tile force of 30% of the maximal voluntary careful control of force. With a steady con-
force and found it necessary to record from traction for 1 second and rest for a few sec-
multiple sites (10 for each muscle). The onds between epochs, force varied from
most sensitive variable was the ratio of the slight to maximal. Standard concentric or
number of turns and the mean turn ampli- monopolar needles were used. The needle
tude in myopathies and the decrease in the was moved to a place in the muscle where a
number of turns in neuropathic disorders. "spiky" pattern was obtained. The filters
Interference pattern analysis did not replace were set at a low linear frequency of 3.2 Hz
evaluation of single MUPs but increased the and a high linear frequency of 8 kHz. The
sensitivity and specificity. sensitivity was varied between 200 and 1000
Hirose et al.3 evaluated the interference /^V/division to allow adequate display of the
pattern at maximal contraction and found activity without blocking. Twenty epochs
that the frequency of short-duration turns of were recorded in each muscle. Turns per
low amplitude was most sensitive for differ- second were plotted against mean amplitude
entiating myopathies from neuropathies. per turn. In normal muscles, the data points
Liguori et al.21 measured mean amplitude fall within a so-called normal cloud. In my-
(cumulative amplitude per time divided by opathies, the data points fall below the nor-
number of turns per time), maximum ratio mal cloud, because of excessive turns and
(peak ratio, which is the maximal value for low amplitude. In neuropathic disorders,
the ratio of the number of turns per time to the data points fall above (lie normal cloud,
the mean amplitude), and the incidence of because of increased amplitude and a low
different time intervals between peaks (0- to turn count. Nirkko et al.22 prospectively eval-
1.5-ms intervals, 1.5- to 5-ms intervals, and 5- uated 239 patients referred for quantitative
to 20-ms intervals) in normal subjects and analysis of MUPs. They found that interfer-
Figure 25-12. Quantitative analysis of the interference pattern (upper) using measurements of time intervals (lower
left), mean amplitude (middle), and the "cloud" (lower right).
ence pattern analysis with the Stalberg the characteristics of the recording elec-
technique was more sensitive and specific trodes and the characteristics of the instru-
than quantitative measurement of MUPs at ments. The shape is generally an inverted
minimal effort (QEMG) or semiquantitative "U," with a broad peak or plateau between
EMG in the detection of myopathies or 100 and 500 Hz, and the power falls to 0 at
neuropathies. approximately 800-2000 Hz. The lower fre-
quency components (10-50 Hz) of the
power spectrum tend to reflect the firing
POWER-SPECTRUM ANALYSIS rates of MUPs. The higher frequency com-
ponents are more related to MUP mor-
Another less common method is the appli- phology.
cation of fast Fourier transforms to the in- In myopathies, the interference pattern
terference pattern to obtain the power of the power spectrum contains relatively more
interference pattern at different frequen- power in the higher frequencies, probably
cies. This can be calculated and displayed because of the short mean duration of the
across the whole frequency spectrum or MUPs. Frequency peaks are above 400 Hz.
given at specific frequencies.18'38 The inter- In neuropathies, the interference pattern
ference pattern can be described mathe- power spectrum contains relatively more
matically as a sum of sine waves of different power in the lower frequencies, probably
frequencies using fast Fourier transforma- because of the long mean duration of the
tion. The interference pattern can be char- MUPs. Ronager et al.38 examined power
acterized according to the density of power spectrum analysis in the biceps brachii of
it contains at various frequencies, the power normal subjects and patients with myop-
spectrum.39 The precise shape of the inter- athies and neurogenic disorders. They de-
ference pattern power spectrum depends on termined the mean power frequency; the
power at 140 Hz, 1400 Hz, 2800 Hz, and The development of innovative methods
4200 Hz relative to the total power; and the of signal analysis, including wavelet domain
high/low ratio (1400/140). At a force of analysis, and the use of artificial intelligence
30% of maximal voluntary contraction, techniques, including neural networks that
power spectrum analysis identified 55% of can learn, suggest that improved methods of
patients with myopathies and 64% of those analysis of EMG signals will become avail-
with neurogenic diseases. In myopathies, the able. These new approaches in combination
relative power at 1400 Hz was increased in with the availability of faster and cheaper mi-
50%. In neurogenic disorders, there was a crocomputers will probably lead to rapid ad-
decrease in relative power at 1400 Hz and vances in quantitative EMG in the next few
decrease in the high/low ratio in 55% of years.
patients.
Frequency analysis of the EMG signal has
been used as a tool for measuring muscle fa- REFERENCES
tigue.40 The mean and median frequency of
the power-density spectrum decay with fa- 1. Hirose K, Sobue I. Quantitative electromyogra-
tigue as the result of changes in muscle fiber phy—a method by computer analysis. Electromyogr
conduction velocity and motor unit activa- Clin Neurophysiol 12:421-429, 1972.
tion. The technique is limited by poorly de- 2. Hirose K, Uono M, Sobue I. Quantitative elec-
fined normal values and wide ranges of val- tromyography. Comparison between manual values
and computer ones on normal subjects. Elec-
ues in normal subjects. tromyogr Clin Neurophysiol 14:315-320, 1974.
3. Hirose K, Uono M, Sobue I. Quantitative elec-
tromyography. Difference between myopathic find-
ings and neuropathic ones. Electromyogr Clin Neu-
rophysiol 15:431-449, 1975.
SUMMARY 4. McComas AJ. Motor unit estimation: anxieties and
achievements. Muscle Nerve 18:369-379, 1995.
The quantitative evaluation of individual 5. Conwit RA, Stashuk D, Tracy B, McHugh M, Brown
MUPs, decomposition of the interference WF, Metter EJ. The relationship of motor unit size,
pattern, interference pattern analysis of firing rate and force. Clin Neurophysiol 110:1270-
1275, 1999.
turns and amplitude, and frequency and 6. Akaboshi K, Masakado Y, Chino N. Quantitative
power spectrum analysis of the interference EMG and motor unit recruitment threshold using
pattern are all useful techniques. They eval- a concentric needle with quadrifilar electrode. Mus-
uate different features of the activity of the cle Nerve 23:361-367, 2000.
7. Stalberg E, Nandedkar SD, Sanders DB, Falck B.
muscle, and the results probably reflect dif- Quantitative motor unit potential analysis. J Clin
ferent properties of the function and dys- Neurophysiol 13:401-422, 1996.
function of the motor unit. One method is 8. Trontelj JV, Stalberg EV. Multiple innervation of
not necessarily superior to the others. It is muscle fibers in myasthenia gravis. Muscle Nerve
likely that the more methods used, the closer 18:224-228, 1995.
9. Stalberg E. Propagation velocity in human muscle
the examiner will come to understanding fibers in situ. Acta Physiol Scand Suppl 287:1-112,
the problem. 1966.
The limiting variable is time. In a busy 10. Bromberg MB, Smith AG, Bauerle J. A comparison
EMG practice, it is not possible to use all of two commercial quantitative electromyographic
algorithms with manual analysis. Muscle Nerve
these techniques in one patient. It is the au- 22:1244-1248, 1999.
thor's practice to use semiquantitative EMG 11. Nandedkar SD, Barkhaus PE, Sanders DB, Stalberg
analysis first because it can be performed on EV. Analysis of amplitude and area of concentric
any muscle and on a number of muscles needle EMG motor unit action potentials. Elec-
quickly. The question is usually answered troencephalogr Clin Neurophysiol 69:561-567,
1988.
with this technique. If doubt still remains, 12. Lang AH, Partanen VS. "Satellite potentials" and
quantitation of the MUP in the most suspect the duration of motor unit potentials in normal,
muscle is performed with the QEMG pro- neuropathic and myopathic muscles. J Neurol Sci
gram or the multi-MUAP analysis program. 27:513-524, 1976.
13. Stalberg EV, Sonoo M. Assessment of variability in
Interference pattern analysis with the Stal- the shape of the motor unit action potential, the
berg cloud, interpotential intervals, and the "jiggle," at consecutive discharges. Muscle Nerve
peak ratio can provide supportive evidence. 17:1135-1144, 1994.
14. Buchthal F, Guld C, Rosenfalck F. Multielectrode tentials at different ages. Arch Neurol (Chicago)
study of the territory of a motor unit. Acta Physiol 6:366-373, 1962.
Scand 39:83-104, 1957. 28. Stalberg E, Andreassen S, Falck B, Lang H, Rosen-
15. Stalberg E, Dioszeghy P. Scanning EMG in normal falck A, Trqjaborg W. Quantitative analysis of
muscle and in neuromuscular disorders. Electro- individual motor unit potentials: a proposition for
encephalogr Clin Neurophysiol 81:403-416, 1991. standardized terminology and criteria for mea-
16. Stalberg E. Macro EMG, a new recording tech- surement. J Clin Neurophysiol 3:313-348, 1986.
nique1. J Neurol Neurosurg Psychiatry 43:475-482, 29. Daube JR. Quantitative EMG in nerve-muscle dis-
1980. orders. In Stalberg E, Young R (eds). Clinical Neuro-
17. Stalberg E, Fawcett PR. Macro EMG in healthy sub- physiology. Butterworths, London, 1981, pp 33-65.
jects of different ages. J Neurol Neurosurg Psychi- 30. Dorfman LJ, McGill KG. AAEE minimonograph
atry 45:870-878, 1982. #29: Automatic quantitative electromyography. Mus-
18. Fuglsang-Frederiksen A. The utility of interference cle Nerve 11:804-818, 1988.
pattern analysis. Muscle Nerve 23:18-36, 2000. 31. LeFever RS, De Luca CJ. A procedure for decom-
19. Zwarts MJ, Drost G, Stegeman DF. Recent progress posing the myoelectric signal into its constituent ac-
in the diagnostic use of surface EMG for neuro- tion potentials—Part I: technique, theory, and im-
logical diseases. J Electromyogr Kinesiol 10:287- plementation. IEEE Trans Biomed Eng 29:149-157,
291, 2000. 1982.
20. Liguori R, Dahl K, Fuglsang-Frederiksen A. Turns- 32. Bischoff C, Stalberg E, Falck B, Eeg-Olofsson KE.
amplitude analysis of the electromyographic re- Reference values of motor unit action potentials ob-
cruitment pattern disregarding force measure- tained with multi-MUAP analysis. Muscle Nerve
ment. I. Method and reference values in healthy 17:842-851, 1994.
subjects. Muscle Nerve 15:1314-1318, 1992. 33. FangJ, Agarwal GC, Shahani BT. Decomposition of
21. Liguori R, Dahl K, Fuglsang-Frederiksen A, Tro- multiunit electromyographic signals. IEEE Trans
jaborg W. Turns-amplitude analysis of the elec- Biomed Eng 46:685-697, 1999.
tromyographic recruitment pattern disregarding 34. Pattichis CS, Schizas CN, Middleton LT. Neural net-
force measurement. II. Findings in patients with work models in EMG diagnosis. IEEE Trans Bio-
neuromuscular disorders. Muscle Nerve 15:1319- med Eng 42:486-496, 1995.
1324, 1992. 35. Fitch P, Willison RG. Automatic measurement of
22. Nirkko AC, Rosier KM, Hess CW. Sensitivity and the hum^n electromyogram. J Physiol (Lond)
specificity of needle electromyography: a prospec- 178:28P-29P, 1965.
tive study comparing automated interference pat- 36. Stalberg E, Chu J, Bril V, Nandedkar S, Stalberg S,
tern analysis widi single motor unit potential anal- Ericsson M. Automatic analysis of the EMG inter-
ysis. Electroencephalogr Clin Neurophysiol 97:1- ference pattern. Electroencephalogr Clin Neuro-
10, 1995. physiol 56:672-681, 1983.
23. Buchthal F, Guld C, Rosenfalck P. Action potential 37. Sanders DB, Stalberg EV, Nandedkar SD. Analysis
parameters in normal human muscle and their de- of the electromyographic interference pattern. J
pendence on physical variables. Acta Physiol Scand Clin Neurophysiol 13:385-400, 1996.
32:200-218, 1954. 38. Ronager J, Christensen H, Fuglsang-Frederiksen A.
24. Buchthal F, Pinelli P, Rosenfalck P. Action poten- Power spectrum analysis of the EMG pattern in nor-
tial parameters in normal human muscle and their mal and diseased muscles. J Neurol Sci 94:283-294,
physiological determinants. Acta Physiol Scand 1989.
'32;219-229, 1954. 39. Kamen G, Caldwell GE. Physiology and interpreta-
25. Buchthal F, Rosenfalck P. Action potential param- tion of the electromyogram. J Clin Neurophysiol
eters in different human muscles. Acta Psychiatr 13:366-384, 1996.
Neurol Scand 30:125-131, 1955. 40. Krivickas LS, Taylor A, Maniar RM, Mascha E, Reis-
26. Petersen I, Kugelberg E. Duration and form of ac- man SS. Is spectral analysis of the surface elec-
tion potential in normal human muscle. J Neurol tromyographic signal a clinically useful tool for
Neurosurg Psychiatry 12:124-128, 1949. evaluation of skeletal muscle fatigue? J Clin Neu-
27. Sacco G, Buchthal F, Rosenfalck P. Motor unit po- rophysiol 15:138-145, 1998.
Chapter 26
SINGLE FIBER
ELECTROMYOGRAPHY
TECHNIQUE Split Fiber or Ephaptic Activation

Hardware Neurogenic Blocking
Software Pitfalls Unique to Stimulated Single Fiber
Method of Activation Electromyography
Measurement CLINICAL APPLICATIONS OF SINGLE
PITFALLS OF SINGLE FIBER FIBER ELECTROMYOGRAPHY
ELECTROMYOGRAPHY Primary Disorders of Neuromuscular
General Transmission
Unstable Trigger Primary Disorders of Muscle
False Trigger Primary Neurogenic Disorders
Incorrect Measurement Position SUMMARY
Damaged Fiber
Single fiber electromyography (SFEMG) is a se- as the results of nerve conduction studies
lective method for evaluation of individual and concentric needle electromyography
components of the motor unit. The SFEMG (EMG).
variables that have been shown to be useful Single fiber electromyography is the most
clinically include fiber density, jitter, and selective technique available in clinical neu-
blocking. Fiber density is increased non- rophysiology to study the motor unit. It is
specifically in various neurogenic and myo- more selective than concentric needle EMG,
pathic disorders, but it can be used in con- which in turn is more selective than surface
junction with jitter and blocking to make recordings. Selectivity refers to the ability to
inferences about the time course and pro- resolve individual generators of electrical ac-
gression of disorders. Jitter and blocking tivity within a volume conductor. In EMG,
provide a quantitative and sensitive measure the generator is the action potential of a
of the efficiency of neuromUscular trans- single muscle fiber. Selectivity depends on
mission. This is particularly useful for iden- three main factors:
tifying abnormalities early in the course of 1. The size of the electrode in relation to
disease when standard electrodiagnostic the size of the action potential genera-
studies are abnormal. Although SFEMG is tor
highly sensitive, it is nonspecific, because 2. The filtering characteristics of the con-
denervation and reinnervation associated ducting medium
with muscle or nerve disease also impair neu- 3. The filter settings used in the record-
romuscular transmission. Thus, SFEMG re- ing process
sults should be interpreted carefully in the Surface electrodes record summated ac-
context of clinical and laboratory data as well tivity from many different motor units. A
343
compound muscle action potential (CMAP) precise recordings from individual muscle
is recorded when motor units discharge syn- fibers, which can be measured and quanti-
chronously after supramaximal electrical tated accurately. The standard SFEMG nee-
nerve stimulation. Assuming the electrodes dle has the recording electrode located
are large enough, the amplitude and area of along the shaft of the cannula 3 mm proxi-
the CMAP reflect the summated activity of mal to the nonbeveled side of the tip. This
the entire muscle. When the muscle is acti- minimizes the chance of recording activity
vated voluntarily, surface electrodes record from muscle fibers that are damaged by the
activity from a large number of motor units. tip and further enhances selectivity by
These types of recordings provide informa- recording activity directly in front of the
tion about firing patterns of large motor unit recording electrode.
groups but do not permit selective record- Single fiber electromyography can be per-
ing of individual motor unit potentials. The formed with minimal voluntary activation or
lack of selectivity results from the large size electric stimulation. The single muscle fiber
of the electrode relative to the size of indi- action potential recorded during SFEMG is
vidual muscle fibers and the tendency of the typically biphasic, with an initial positive
intervening tissue of the volume conductor phase followed by a major negative spike.
to act as a high-frequency filter. When the electrode is close to the muscle
The concentric needle EMG electrode, fiber (rise time < 500 microseconds), the
with a recording surface of 150 X 580 /xm, amplitude ranges from 500 //-V to 10 mV with
is able to record selectively from individual a duration of 1-1.5 ms. The amplitude varies
motor unit potentials containing an average greatly with minor changes in distance be-
of several hundred muscle fibers. Because of cause of the small size of the recording elec-
the high-frequency filter effect of muscle tis- trode. The power spectrum of the single
sue, most of the motor unit potential wave- fiber action potential ranges from 100 to
form is generated from 10 to 20 muscle 5000 Hz, with a peak from 1 to 2 kHz. Four
fibers within several millimeters of the elec- types of measurement can be made during
trode. The selectivity of concentric needle SFEMG:
EMG is limited by the large electrode size 1. Fiber density reflects the packing density
(150 X 580 /mi) relative to the diameter of of muscle fibers within the recording
a single muscle fiber (25-100 /tin). This can area of the single fiber electrode. It
be overcome to some extent by increasing correlates with the degree of motor
the low-frequency filter to 500 Hz, which at- unit potential polyphasia recorded in
tenuates low-frequency activity from distant concentric needle EMG. Fiber density
muscle fibers. This narrows the recording is increased in neurogenic and myo-
area of the concentric needle EMG elec- pathic disease.
trode to 500 to 1000 fJLm and, in many cases, 2. Jitter measures the latency variability of
allows recording of potentials from single muscle fiber action potentials within
muscle fibers or potentials summated from the same motor unit. This typically re-
two or three fibers. Thus, a reasonable esti- flects the variability in rise time of the
mation of jitter is obtained using a concen- end plate potential, thereby providing
tric needle EMG electrode by increasing the a sensitive indicator of a mild defect of
low-frequency filter to 500 Hz. neuromuscular transmission. Jitter is
The SFEMG electrode has a circular re- increased in disorders associated with
cording surface with a diameter of 25 /Am, denervation and reinnervation as well
which is about the same as the diameter of as primary neuromuscular junction
individual muscle fibers. When used in con- diseases.
junction with a 500 Hz low-frequency filter, 3. Blocking measures the intermittent loss
the effective recording distance is limited to of a regularly firing muscle fiber action
200 /am. The combination of small electrode potential within a motor unit. This typ-
size, low-pass filter characteristics of muscle ically reflects the failure of the end
tissue, and use of a 500 Hz low-frequency fil- plate potential to reach threshold in
ter provide the selectivity required to make disorders of neuromuscular transmis-
Single Fiber Electromyography 345
sion, but it also can occur in neuro- gle fiber potential. Most standard clinical
genic disorders when the impulse is EMG preamplifiers and amplifiers are suffi-
blocked along a terminal branch of the cient for SFEMG recording as long as the
motor axon. Blocking is present in low-frequency filter can be set to 500 Hz.
moderate to severe disorders of neuro-
muscular transmission, in disorders as-
sociated with denervation and reinner- TRIGGER, DELAY, AND DISPLAY
vation of muscle, and in neuropathies An amplitude trigger placed on the negative
associated with impulse blocking in the phase of the single fiber action potential ini-
nerve terminal. tiates the sweep, and an analog or digital de-
4. Duration measures the time between lay allows the triggered potential and all po-
muscle fiber action potentials within tentials that are time-locked to the triggered
recording distance of the electrode. potential to be displayed. The trigger can be
This reflects differences in conduction used with voluntary, reflex, or electric acti-
time along the terminal axonal branch vation. The potential should be displayed at
and muscle fiber. Duration correlates a sweep speed of 0.2-1 ms and a sensitivity
with the duration of motor unit poten- of 20 fjiV-2 mVfor accurate resolution of sin-
tials recorded in concentric needle gle fiber potentials and measurement of
EMG and is increased in neurogenic fiber density, jitter, blocking, and duration.
disease and in some chronic myop- When analog equipment and manual mea-
athies. surement are used, a counter and filming
system that provide printouts of five consec-
utive groups of 10 superimposed sweeps are
TECHNIQUE required to measure and to quantitate jitter.
Hardware
Software
NEEDLE ELECTRODE
For standard SFEMG recordings, the needle The majority of commercially available sys-
consists of a stainless steel shaft (0.5 mm di- tems used to perform SFEMG are digital.
ameter), with a single platinum wire down Digital conversion of the signal affords the
the center, opening onto a side port oppo- advantages of flexible display, automated
site the beveled edge, 3 mm proximal to the measurement and calculation, storage, and
electrode tip. The active recording surface reanalysis of data. Each system has features
is circular and 25 ^tm in diameter. The shaft that differ with regard to ease of use, degree
of the needle serves as the reference elec- of automation, display, and price. Key fea-
trode. Single fiber electromyographic elec- tures that are essential include automated jit-
trodes are expensive and, thus, are sterilized ter (mean consecutive difference) measure-
and reused. The electrode should be in- ment of a minimum of 50 consecutive
spected under a dissecting microscope after sweeps, storage capacity for at least 30 pairs,
being used every 5-10 times and sharpened and the ability to review and reanalyze all the
as needed. Electrolyte treatment may be re- raw data.
quired if single fiber amplitudes are low or
noise is excessive.
Method of Activation
AMPLIFIER VOLUNTARY SINGLE FIBER
ELECTROMYOGRAPHY
The high impedance of the SFEMG elec-
trode requires that the recording amplifier Muscle fiber action potentials are isolated
have high input impedance and low input with minimal voluntary muscle contraction.
capacitance, which maintains the frequency Approximately 60% of the time, a single po-
response and prevents distortion of the sin- tential is recorded. For the other 40%, from
one to five potentials that are time-locked to duration) are used to activate a small num-
the triggered potential are recorded on the ber of muscle fibers. The current and posi-
same sweep. The level of activation should tion of the SFEMG recording electrode are
be adjusted to maintain the triggered po- adjusted until a single muscle action poten-
tential at a firing rate of 10-15 Hz. When ac- tial with a rise time less than 500 microsec-
tivation is too vigorous, various technical onds and time-locked to the stimulus is
problems can arise, including overestima- recorded. Jitter is measured as the latency
tion of jitter (caused by unstable trigger and variability of the muscle fiber action poten-
variation in amplitude of measured potential in relation to the stimulus. When in-
tial) and false blocking (caused by alterna- creased jitter or blocking is observed, the
tion of the trigger between a time-locked stimulus is increased slightly. If the blocking
and a single potential). The position of the disappears or the jitter lessens, the abnor-
needle is adjusted to maximize the rise time malities likely were caused by slight variation
of the triggered and time-locked potentials. in current strength above and below the
Minor rotational movements of the needle threshold of activation. This is a technical
help reduce noise from distant potentials problem unique to stimulated SFEMG. Re-
and separate time-locked potentials that are duced excitability resulting in latency pro-
fused with the triggered potential. longation, increased jitter, and blocking can
also be seen with prolonged stimulation at
STIMULATED SINGLE FIBER rates of stimulation greater than 20 Hz. This
ELECTROMYOGRAPHY can be avoided by keeping stimulation rates
less than 20 Hz except for brief 5- to 10-
Muscle fiber action potentials are recorded second intervals of stimulation at higher
after electric stimulation in the study of dis- rates. Very low jitter (mean consecutive dif-
orders of neuromuscular transmission or in- ference < 10 microseconds) is related to
vestigation of certain reflexes (F waves, H re- direct muscle stimulation and should be
flex, blink reflex, etc.). Electric stimulation ignored.
is useful when patients are unable to coop-
erate with voluntary SFEMG or when the ef-
fect of a change in firing rate on jitter and
blocking needs to be quantitated. Stimu-
Measurement
lated SFEMG has several disadvantages. FIBER DENSITY
Fiber density cannot be measured accu-
rately, and careful attention to technical Fiber density is defined as the average num-
problems is necessary to ensure accurate and ber or density of muscle fiber action poten-
reliable measurement of jitter and blocking. tials within the recording area of the single
Different normal values are used for jitter in fiber needle electrode. The needle is ad-
stimulated and voluntary SFEMG, because justed until a single fiber action potential
jitter reflects neuromuscular transmission with an adequate rise time (< 500 mi-
from a single end plate in stimulated SFEMG croseconds) is isolated. The number of
and from two end plates in voluntary SFEMG muscle fiber action potentials with an am-
(that is, both the triggered and measured plitude greater than 200 /iV that are time-
potential). Finally, electric stimulation tends locked to the triggered potential are
to selectively activate large diameter axons counted. This procedure is repeated for 30
with low activation thresholds, unlike volun- separate triggered potentials to obtain an av-
tary SFEMG in which smaller diameter ax- erage fiber density for the entire muscle. In
ons are recruited initially and recorded normal subjects, a single potential is isolated
preferentially. 60% of the time, two potentials 35%, and
Electric stimulation can be applied to a three or more potentials 5% of the time. Av-
branch of the nerve located outside the mus- erage fiber density ranges from 1.3 to 1.8 in
cle or to an intramuscular motor branch. A normal persons younger than 70 years. Fiber
monopolar needle is used as a cathode with density reflects the density of muscle fibers
another needle or surface electrode as the in one motor unit within the recording area
anode. Very small currents (1-10 mA, 0.5 ms and corresponds most directly to the num-
her of turns seen on standard concentric tential interval is less than 1 ms. In normal
needle EMG. This feature of the motor unit subjects, there are small variations in the size
potential is sometimes called complexity, be- of the end plate potential caused by varia-
cause each of these turns typically represents tions in the number of quanta of acetyl-
a separate fiber that contributes to the mo- choline released from the nerve terminal. A
tor unit potential. Fiber density has a less smaller end plate potential has a slower rise
direct relationship to the percentage of time and reaches threshold later than a
polyphasic motor unit potentials, because in larger end plate potential, so that the time
a polyphasic potential a phase may include from the action potential in the nerve ter-
more than one turn. Satellite potentials seen minal to the action potential in the muscle
in standard recordings also are recorded as fiber varies by as much as 50 microseconds.
separate single fiber potentials in SFEMG. The presence of this variation is evidence of
Fiber density is increased in disorders that a synapse between the activation site and
produce denervation and reinnervation. recording site. Two single fiber potentials
Thus, increased fiber density is observed in with little or no jitter are either time-locked
most motor neuron diseases and peripheral by ephaptic (electric) activation of each
neuropathies. The finding of increased fiber other or are recorded from a single muscle
density is particularly striking and out of pro- fiber that has been split or otherwise dis-
portion to other changes on SFEMG early torted. The amplitude and the rise time of
in the course of reinnervation, when differ- the end plate potential are a direct reflec-
ences in conduction along regenerating tion of the safety margin of neuromuscular
nerve terminals cause marked asynchrony of transmission. Any disorder of neuromuscu-
firing of newly reinnervated muscle fibers lar transmission that decreases the safety
and dispersion of single muscle fiber action margin will increase jitter. This includes
potentials. Chronic disorders with minimal primary disorders of neuromuscular trans-
active denervation but considerable com- mission (for example, myasthenia gravis,
pensated reinnervation have increased fiber Lambert-Eaton myasthenic syndrome, and
density, with only minimal increase in jitter botulism) or secondary disorders that re-
and blocking. In contrast, subacute progres- duce the safety margin by producing dener-
sive disorders associated with ongoing rein- vation and reinnervation with immature
nervation have a marked increase in jitter nerve terminals (for example, amyotrophic
and blocking and only a mild increase in lateral sclerosis, some peripheral neurop-
fiber density. Fiber density is also increased athies, and some myopathies).
in myopathies that are associated with fiber Jitter can be measured as the standard de-
splitting, degeneration, and regeneration. viation of the interpotential interval, but be-
Therefore, fiber density can be used to quan- cause of the occasional occurrence of a grad-
titate the severity and time course of some ual change in the mean interpotential
neuromuscular disorders, but it cannot dis- interval over time, it is more reliable to use
tinguish between neurogenic and myopathic the mean consecutive difference (MCD):
disorders.
JITTER
In voluntary SFEMG, jitter is defined as vari-
ation in the interpotential interval between where IPI is interpotential interval, and N is
two single muscle fiber action potentials the number of intervals measured. For max-
recorded simultaneously from a single mo- imal reliability, the MCD should be calcu-
tor unit. Jitter typically results from variation lated from 50 or more consecutive interpo-
in the rise time and amplitude of the end tential intervals. The MCD is calculated for
plate potential at the neuromuscular junc- each pair and then the mean of 20 pairs is
tion. Jitter can also result from variability of calculated and compared with normal values
conduction along the muscle membrane, for the muscle and the age of the patient.
but these factors produce negligible jitter at When the firing interval is variable or
regular firing rates and when the interpo- the interpotential interval is large (that is,
> 1 ms), the variation in the conduction ve- munis, the most commonly recorded limb
locity of the muscle fiber can affect jitter. muscle, is between 16 and 34 microseconds
This is described by the velocity recovery func- (upper limit of normal for a single pair is 55
tion, which demonstrates that when the pre- microseconds) for persons younger than 60
ceding discharge interval is short the con- years. For persons older than 60 years, the
duction velocity of the muscle fiber is faster upper limit for MCD increases to 43 mi-
on the subsequent discharge. If there is vari- croseconds, and it is normal to have up to
ability in the discharge frequency and the in- two pairs with an MCD greater than 55 mi-
terpotential interval is long (that is, if the croseconds. The jitter is smaller in facial
velocity recovery function is beginning to af- muscles than in limb muscles. Facial muscles
fect jitter), then the mean sorted difference are also less susceptible to local trauma,
(MSD) is a better representation of jitter which can increase jitter indefinitely. The
than the MCD. The MSD is calculated by first MCD for the frontalis muscle is 23-31 mi-
sorting the potentials in ascending order of croseconds in normal subjects younger than
size of the interpotential interval and calcu- 60 years and 23-35 microseconds for those
lating the MCD. If the MCD:MSD ratio is 60 years and older. Similar normal values
greater than 1.25, the variability of the dis- have been defined for these and other mus-
charge frequency is affecting jitter and the cles.1 Normal MCD values for stimulated
MSD should be used. SFEMG are approximately 80% of the value
The best point to measure jitter is on the obtained with voluntary SFEMG of the same
steep rising phase of the potential close to muscle in the same age group.
the baseline crossing. Movement of the trig- In normal muscle, jitter is not identifiable
gered potential (amplitude jitter) or contam- on standard concentric or monopolar nee-
ination of the baseline with other potentials dle EMG. However, if a motor unit potential
or one potential riding upon another will aris recorded with a standard concentric nee-
tificially increase jitter. dle electrode at a low-frequency filter of 500
In stimulated SFEMG, jitter is measured Hz and a sweep speed of 100 or 200 jLts/cm,
as the MCD of the interpotential difference jitter can be identified. Quantitative mea-
between the stimulus artifact and the single surements of jitter with a standard concen-
muscle fiber potential. The MCD reflects the tric needle electrode are somewhat larger
jitter from a single end plate rather than the than those recorded with a single fiber elec-
pair of end plates measured in voluntary trode, making the study less specific in the
SFEMG. The velocity recovery function of detection of mild defects of neuromuscular
muscle does not affect jitter measurement transmission.
during stimulated SFEMG because there is Because jitter is the result of fluctuations
no random variation in discharge frequency. in the amplitude of the end plate potential,
Several software applications are available any disorder that decreases the end plate po-
to automate the measurement of jitter and tential produces increased jitter. This occurs
blocking. Each program provides a graphic in disorders of neuromuscular transmission,
display of the calculated MCD for 50-100 such as myasthenia gravis, but also in disor-
consecutive sweeps and the ability to store, ders with ongoing reinnervation or regen-
review, and reanalyze each individual sweep eration of muscle fibers, such as amy-
collected in order to ensure accuracy of the otrophic lateral sclerosis and polymyositis.
data collected. Manual calculation of MCD Thus, abnormalities of jitter are not diag-
requires a counter that captures and displays nostic of a specific disease of the neuro-
50 consecutive sweeps in five groups of 10 muscular junction but must be considered
superimposed images. The variation of the in relation to findings obtained with stan-
interpotential difference can be measured dard electrophysiologic recordings. Abnor-
directly from each of the five groups and the malities of jitter can occur without clinical
MCD calculated directly with a conversion weakness in the muscle.
factor. Jitter is a function of the variation in
Normal values for MCD vary with age and synaptic potential size; therefore, it is pres-
the muscle. In the Mayo EMG laboratory, the ent in recordings that include other synapses.
normal jitter in the extensor digitorum com- F waves are a result of antidromic activation
of the anterior horn without a central the motor unit potential. Some authors have
synapse, so that F-wave jitter is approximately referred to the instability of the motor unit
the same magnitude as with voluntary mo- potential related to neuromuscular junction
tor unit potentials. In contrast, H reflexes, disease as jiggle.
which include a synapse in the spinal cord Blocking is observed in disorders of neu-
in addition to the neuromuscular junction, romuscular transmission such as myasthenia
have normal jitter of two to three times that gravis, and, when present, it provides evi-
of voluntary motor unit potentials. Other dence of a defect severe enough to produce
more complex reflex phenomena, such as weakness either at rest or with exertion. As
the blink and flexion reflexes, have corre- with jitter, however, blocking can occur in
spondingly larger amounts of jitter. other disorders in which neuromuscular
transmission may be impaired, such as amy-
otrophic lateral sclerosis, polymyositis, and
BLOCKING ongoing reinnervation. For blocking to be
In a normal muscle, the end plate potential considered evidence of a disorder such as
always reaches threshold and initiates a sin- myasthenia gravis, it should be found in the
gle fiber action potential. Therefore, when absence of other electrophysiologic signs of
multiple single fiber potentials are found, neurogenic or myopathic disease.
they occur with each discharge of the motor
unit potential. Blocking occurs when an end DURATION
plate potential does not reach threshold or
when conduction fails in the nerve terminal, The interval between the first and last po-
resulting in a loss of a single fiber potential tential of multiple single fiber potentials
during one or more discharges of the motor recorded from a motor unit has been mea-
unit. Blocking is measured as the percent- sured as duration. Duration is the total time
age of discharges of a motor unit in which a from the first to the last potential averaged
single fiber potential is missing. A motor unit for all multiple potentials recorded. An al-
in which a single fiber potential did not fire ternative measure is mean interspike interval,
half the time would have 50% blocking. which divides the duration by the number of
Blocking for a particular muscle is expressed single fiber potentials in each discharge.
either as the percentage of 20 fiber pairs that Mean interspike interval and duration in-
show any blocking or as the total percentage crease when the activation of individual sin-
of discharges in the 20 pairs that displayed gle fiber potentials is dispersed in time.
blocking. For example, if 20 potential pairs Factors that may contribute to dispersion in-
each discharge 50 times and blocking occurs clude reduced synchrony of firing, anatom-
a total of 20 times in two of the pairs, 10% ical dispersion of end plates along muscle
of the pairs show blocking and 2% of all dis- fibers, and differences in conduction along
charges have blocking. the terminal axon or muscle fiber. Duration
Normal elderly subjects display occasional is used less frequently than other SFEMG
blocking in some muscles. In fact, in a study measurements, because similar information
of 20 pairs of single fiber potentials, if a sin- is obtained from the duration of motor unit
gle pair exceeds the limit of normal jitter potentials and the examination for satellite
or displays blocking (or both), many elec- potentials during standard concentric or
tromyographers do not use those changes monopolar needle EMG.
alone to interpret the study as abnormal.
Blocking begins to occur when the jitter in
a pair has increased to 80-100 microsec- PITFALLS OF SINGLE FIBER
onds. In disorders of neuromuscular trans- ELECTROMYOGRAPHY
mission, amplitude decrement of a CMAP
with 2 Hz repetitive stimulation is caused by General
blocking, as is moment-to-moment variation
observed on standard concentric or mono- Most errors in SFEMG result from needle
polar needle EMG. Electrode movement can movement, excessive activation, or variabil-
also produce variation of the amplitude of ity in the firing rate of the muscle action po-
tential. The selectivity that affords the single

fiber electrode the ability to focus the
recording on 1 or 2 muscle fibers also ren-
ders SFEMG extremely sensitive to small
movements of the needle electrode. Minor
movement along the long axis as well as an-
gulation or rotation of the cannula often
produces marked changes in the amplitude
or configuration of the muscle fiber action
potential. Variation in amplitude and con-
figuration leads to errors in jitter and block-
ing measurements. The ability to hold the
electrode motionless is the greatest techni-
cal skill required to produce reliable SFEMG
recordings. The second great technical chal-
lenge in voluntary SFEMG is to maintain
minimal activation at relatively stable firing
rates. Excessive noise produced by nearby
muscle fiber action potentials may distort
the baseline or the single fiber potentials of Figure 26-1. Superimposition of sweeps during single
fiber electromyography recordings. Poor Superimposi-
interest, which artificially increases jitter. tion traces in Pair 2 demonstrates an unstable triggered
Changes in firing rates affect the velocity re- potential.
covery function of the muscle fiber, which
also increases jitter.
tude, and configuration. When this occurs,
the trigger may "jump" from one potential
Unstable Trigger to the other. If one of the potentials is a
double—that is, it has a second potential time-
Because all measurements are taken in ref- locked to the first—and the other potential
erence to the triggered potential, stability of is a single, alternation of the trigger between
this potential forms the cornerstone of vol- the two potentials gives the false impression
untary SFEMG. Gross instability results in in- of blocking (Fig. 26-2).
termittent loss of the triggered potential and
interruption of the SFEMG recording. When
this occurs occasionally, the sweep can be Incorrect Measurement Position
deleted from analysis, but frequent loss of
the triggered potential makes it impossible Jitter is best measured from the inflection
to record jitter or blocking accurately. Even point of the second potential, which typically
minor fluctuation of the amplitude of the occurs as the rise time of the potential
triggered potential can add as much as 10 crosses the baseline. In voluntary SFEMG,
microseconds to the MCD. Superimposition this position requires minimal activation, be-
of consecutive sweeps is an effective way to cause any noise that moves the baseline will
check for stability of the triggered potential affect the measurement of jitter. The farther
(Fig. 26-1). Electrode movement, excessive the measurement point is moved away from
activation, or variation in firing rates can the inflection point, the more jitter is af-
produce an unstable trigger. fected by changes in amplitude of the po-
tential produced by needle movement.
Some automatic algorithms measure jitter
False Trigger from the peak rather than the rising phase
of the potential (Fig. 26-3). This method is
Occasionally in voluntary SFEMG, two mus- also useful when multiple potentials are su-
cle fiber action potentials close to the elec- perimposed on one another. The accuracy
trode have similar firing thresholds, ampli- of the peak jitter method depends on the
ability of the examiner and the software to

clearly define stable peaks of the waveforms.
Damaged Fiber
Recording from a damaged fiber often pro-
duces a broad positivity, followed by a nega-
tive phase of shorter duration and lower am-
plitude. The negative phase sometimes has
the appearance of a second potential, which
usually has low jitter with the parent poten-
tial (Fig. 26—4). This has been referred to as
a false double and should not be included in
jitter calculations.
Split Fiber or Ephaptic Activation

Jitter less than 10 microseconds generally
indicates the absence of a neuromuscular
junction between the two recorded poten-
tials (Fig. 26-5). This occurs when the po-
Figure 26-2. False trigger. The trigger alternates be- tentials are generated from the same muscle
tween two potentials of similar size and shape. Poten- fiber that has split into two sections or when
tial #1 is a single potential, whereas potential #2 is time-
locked to a second potential (that is, part of a double). one fiber is activated by an adjacent fiber by
Alternation of the trigger between potentials #1 and #2 ephaptic transmission. In stimulated SFEMG,
gives the false impression of blocking. low jitter occurs when direct muscle stimu-
Figure 26-3. Jitter measurement. Left, Jitter measurement from the peak of the potential. Right, Jitter measurement
from the negative slope of the potential.
Figure 26-4. Damaged fiber. Jitter from the second

component should not be included in calculations of
mean consecutive difference.
lation occurs. These values should be elimi-

nated from jitter calculations.
Neurogenic Blocking
There are two mechanisms by which block-
ing occurs in neurogenic disorders. The
first, which occurs at the neuromuscular
junction, is caused by immaturity of recently
reinnervated end plates. The second, which
occurs within the terminal branch of the mo-
tor axon, is caused by a prolonged refractory
period related to branching and demyelina- Figure 26-6. Neurogenic blocking. The second and
tion of the nerve terminals. Axonal blocking third potentials block concomitantly.
can be recognized when more than one
muscle fiber action potential disappears or
blocks simultaneously during either volun-
tary or stimulated SFEMG (Fig. 26-6). This frequency on jitter and blocking. A disad-
phenomenon has also been referred to as vantage of stimulated SFEMG is the inability
concomitant blocking. to measure fiber density. Other problems
are technical in nature and must be ac-
counted for to ensure accuracy of jitter and
Pitfalls Unique to Stimulated blocking measurements. The first technical
Single Fiber Electromyography pitfall is related to direct muscle stimulation.
This is fairly easy to recognize because of
Stimulated SFEMG is useful when evaluating the resulting small jitter. A more subtle but
poorly cooperative patients or when it is im- equally important technical problem is re-
portant to measure the effect of discharge lated to a false increase in jitter and block-
ing that occurs when the stimulus intensity
is close to the axonal threshold of the single
fiber potential being examined. Increasing
the stimulus intensity slightly identifies this
problem, which should eliminate blocking
and reduce jitter if the problem is technical.
The axonal threshold may also increase with
prolonged stimulation or stimulation at
rates in excess of 20 Hz. Thus, the effect of
small increases in stimulus intensity should
Figure 26-5. Low jitter indicates recording from a split always be determined before jitter and
fiber or one activated by ephaptic transmission. MCD, blocking are measured during stimulated
mean consecutive difference. SFEMG.
CLINICAL APPLICATIONS firing rate and muscle activity and decreased

OF SINGLE FIBER with rest or edrophonium. Frequent block-
ing in patients with severe myasthenia gravis
ELECTROMYOGRAPHY made SFEMG recordings difficult to per-
form.
Primary Disorders The work of Stalberg et al.3 was confirmed
of Neuromuscular and amplified by other investigators.2'7'8
Transmission Even though the patient populations stud-
ied by these authors differed, all of them
The sensitivity of SFEMG allows an abnor- found that 77% to 95% of patients with myas-
mality of neuromuscular transmission to be thenia gravis have abnormal SFEMG find-
recognized in the absence of clinical weak- ings, with increased jitter or blocking (or
ness or abnormality revealed by other phys- both). Patients with generalized myasthenia
iologic tests.2 The clinical usefulness of gravis have a higher proportion (98% to
SFEMG in identifying and quantitating de- 100%) of abnormal jitter, and if there is
fects of neuromuscular transmission has weakness caused by myasthenia gravis in the
been demonstrated repeatedly.3"5 Although muscle being tested, all authors agreed that
the method is more complicated and time- the jitter in that muscle is abnormal.
consuming than repetitive stimulation stud- The frequency of SFEMG abnormalities in
ies, it can be learned readily and applied patients with ocular myasthenia gravis was
with a minimum of specialized equipment. less consistent among the authors men-
However, in uncooperative or tremulous tioned above, ranging from 20% to 70% ab-
patients, reliable voluntary SFEMG can be normal in the extensor digitorum commu-
time-consuming, so the selection of patients nis muscle and from 60% to 100% in
and muscles for SFEMG requires considera- proximal and facial muscles.9'10 The pres-
tion. ence of increased jitter in the extensor dig-
itorum communis muscle in patients with
AUTOIMMUNE MYASTHENIA GRAVIS clinical disease restricted to ocular muscles
increases the risk of the development of
The abnormalities found on SFEMG in pa- clinically generalized myasthenia.11
tients with myasthenia gravis were demon- Slight increases in fiber density have been
strated clearly in the early studies of Ekstedt reported for up to 25% of patients with
and Stalberg and are reviewed in the text- myasthenia gravis; however, this usually is
book by Stalberg and Trontelj,6 Single Fiber not found, and if it is present, it is minimal
Electromyography. Both jitter and blocking are and of little clinical significance. Single fiber
increased in proportion to the severity of electromyography is often abnormal when
clinical involvement, with greater abnormal- myasthenia gravis is in clinical remission
ity in weaker muscles. However, in a given (66% to 83%) and in patients with thymoma
muscle, and even among the end plates of a and no clinical weakness.
single motor unit, there is marked variation Single fiber electromyography findings
in the amount of jitter in different fiber have been compared with those of other di-
pairs. In a single muscle, some fiber pairs agnostic studies. All the comparisons have
may be entirely normal and others may be shown a much greater frequency of abnor-
grossly abnormal, with frequent blocking. mality on SFEMG than on repetitive stimu-
Most important among the features that lation.12 All the studies found a slightly
Stalberg, Ekstedt, and Broman3 noted was higher percentage of abnormality for
the presence of abnormal jitter even in clin- SFEMG than for acetylcholine receptor an-
ically normal muscles. In their early experi- tibody levels. However, the studies of both
ence with 70 patients with myasthenia gravis, Stalberg et al.4 and Kelly et al.8 demon-
jitter was always abnormal. They concluded strated that SFEMG and acetylcholine re-
that if jitter was normal in the presence of ceptor antibodies are complementary tests:
weakness, the weakness was not caused by one test identifies abnormality in some pa-
myasthenia gravis. They also noted that jit- tients in whom the results of the other test
ter was increased with motor unit potential are normal.
Single fiber electromyography has been potentials are studied for the variation in size
used in the Mayo Clinic EMG Laboratory or shape that can provide clear evidence of
since 1976 as a useful adjunct to standard a defect of neuromuscular transmission. The
EMG and nerve conduction studies in pa- presence of abnormal motor unit potentials
tients thought to have a defect of neuro- or motor unit potential variation on stan-
muscular transmission. Patients with other dard needle EMG makes SFEMG unneces-
disorders have been evaluated with SFEMG sary unless further quantitation is needed.
only as part of special studies and not rou- Therefore, in our practice, SFEMG is lim-
tinely. A large number of patients are eval- ited to patients with normal findings on stan-
uated, primarily for diagnostic purposes, in dard studies. Single fiber electromyography
the Mayo Clinic laboratory, and the efficient studies begin on the extensor digitorum
evaluation of individual patients becomes a communis muscle, because it is easy to ex-
major determinant in the selection of stud- amine and is well defined and age-controlled
ies. Single fiber electromyography is applied normal values are available. Twenty or more
regularly only to patients in whom defects of fiber pairs are measured for MCD, blocking,
neuromuscular transmission are suspected. and mean duration of the total interpoten-
Furthermore, it is limited to patients who tial interval. This muscle is tested even with
have no evidence of disease on other elec- no symptoms or signs of weakness in the ex-
trophysiologic testing. tremities, because a high proportion of them
Each patient in whom myasthenia gravis show abnormalities. If the extensor digito-
or a similar disorder is suspected first un- rum communis muscle is normal, SFEMG is
dergoes repetitive stimulation of distal and performed on other muscles selected on the
proximal muscles before and after exercise. basis of clinical weakness. The frontalis mus-
Repetitive stimulation is a reliable, readily cle is commonly used for patients with ocu-
performed method of obtaining a quantita- lar or bulbar symptoms.13
tive measure of a defect of neuromuscular
transmission. The presence of a repro- CONGENITAL MYASTHENIC
ducible decrement with slow rates of stimu- SYNDROMES
lation, which is partially repaired after exer-
cise and is enhanced late after exercise Use of voluntary SFEMG to study patients
provides enough evidence of abnormality with congenital myasthenic syndromes is of-
that SFEMG is not required unless further ten limited by the difficulty in obtaining co-
quantitation of the defect is needed. More- operation in young patients. Stimulated
over, most patients who have a decrement SFEMG provides a good alternative in these
with repetitive stimulation have marked circumstances. There are few reports of the
SFEMG abnormalities that make measure- use of stimulated SFEMG in congenital myas-
ment of the jitter and blocking more diffi- thenic syndromes. We have used this tech-
cult and less reliable. nique in selected patients but have yet to
In addition to repetitive stimulation, each demonstrate that it provides additional in-
patient undergoes standard needle EMG of formation to repetitive stimulation and stan-
the proximal and distal muscles, including dard concentric needle EMG. Determina-
clinically weak muscles, paraspinal muscles, tion of the mechanism of the neuromuscular
and bulbar muscles. The needle examina- transmission defect in congenital myas-
tion can provide evidence of other disease thenic syndromes requires a combination of
that may be associated with abnormal microelectrode and morphological studies
SFEMG, such as myopathies and anterior performed on an intercostal or anconeus
horn cell disease. These diseases must be muscle biopsy specimen.14
identified before SFEMG is performed, to
save the time and effort of performing LAMBERT-EATON MYASTHENIC
SFEMG and to ensure that they are not mis- SYNDROME AND BOTULISM
taken for myasthenia gravis. Also, standard
needle EMG can sample rapidly several mus- Presynaptic disorders of neuromuscular
cles on which it may be difficult to perform transmission have also been studied with
SFEMG. In these muscles, single motor unit single fiber techniques. Abnormalities, both
with standard EMG and SFEMG, are readily information to the concentric needle exam-
seen in disorders such as the Lambert-Eaton ination about the degree and timing of as-
myasthenic syndrome and botulism.15'16 In- sociated reinnervation and may detect early
creased jitter and blocking that decrease at mild abnormalities not recognized on stan-
higher innervation rates are characteristic of dard EMG. The usefulness of SFEMG for se-
presynaptic disorders.17'18 rial study of myopathies during treatment
Postsynaptic disorders typically demon- trials and for investigation of membrane ab-
strate increased jitter and blocking at higher normalities has not been determined.
discharge rates.19 With the possible excep-
tion of the effect of firing rate on jitter and
blocking, SFEMG has not been shown to add Primary Neurogenic Disorders
substantial information to repetitive stimu-
lation and standard concentric needle EMG Because reinnervating nerve terminals have
in patients with either Lambert-Eaton myas- immature neuromuscular junctions, SFEMG
thenic syndrome or botulism. demonstrates abnormalities. The specific
type and degree of abnormality depend on
tile magnitude and rate of progression of the
Primary Disorders of Muscle neurogenic process. Abnormalities on stan-
dard nerve conduction studies and EMG dif-
Most muscle disorders that cause abnormal- ferentiate these from disorders of the neu-
ities on SFEMG do so by causing muscle ne- romuscular junction.
crosis, with secondary regeneration and
reinnervation. However, muscle fiber mem- PERIPHERAL NEUROPATHY
brane abnormalities that cause electrophys-
iologic failure can also lead to changes. De- In studies of severed nerves, increased fiber
creased fiber density and blocking without density is the first sign of reinnervation. In-
much increase in jitter were observed dur- creases in fiber density are seen as early as 3
ing a paralytic attack in a patient with hy- to 4 weeks after nerve injury, usually before
pokalemic periodic paralysis.20 changes can be detected on muscle biopsy.
Muscular dystrophies and inflammatory Fiber density increases rapidly for the first 3
myopathies are the muscle diseases that months after injury and slowly thereafter. In-
show the most marked abnormalities on creased jitter and blocking are seen for 3 to
SFEMG, but abnormalities have been noted 6 months after the injury but rarely longer
in metabolic and congenital myopathies as than that. Most clinical neuropathic disease
well.21'22 Foote et al.2* found markedly in- presents with more complex findings, be-
creased jitter, blocking, duration of the sin- cause the process is progressive rather than
gle fiber complex, and fiber density in 18 a single insult and disease may affect the abil-
patients with inflammatory myopathy. A ity to reinnervate.
significant negative correlation existed be- Thiele and Stalberg24 reported SFEMG
tween motor unit potential duration (mea- findings in 54 patients with polyneuropathy
sured with concentric needle electrode) and associated with uremia, diabetes mellitus, or
the degree of jitter and blocking. This was alcohol abuse. Findings of increased fiber
believed to provide evidence that the myo- density, jitter, and blocking were seen in al-
pathic process was involved in the genesis coholic patients who had only mildly slowed
of both electrophysiologic abnormalities. nerve conduction velocities but evidence of
Duchenne muscular dystrophy causes in- denervation on concentric needle examina-
creased fiber density and a markedly in- tion. Patients with diabetic or uremic
creased duration as well as moderately in- polyneuropathies had slower nerve conduc-
creased jitter and blocking. Limb girdle tion velocities, relatively normal concentric
muscular dystrophy demonstrates quantita- needle examinations, and mild abnormali-
tively less marked changes. ties on SFEMG, that is, mildly increased
Single fiber electromyography is not an ef- jitter without blocking and normal fiber
fective method for primary diagnosis of my- density and duration. Single fiber electro-
opathies. However, it adds complementary myography corroborated standard EMG and
pathologic data about the neuropathies of ease. Patients with chronic nonprogressive
these patients. Similar findings have been clinical conditions demonstrated complex,
demonstrated by other investigators in criti- but stable, motor unit action potentials and
cal illness neuropathy25 and length-dependent increased fiber density. Patients with recent
diabetic polyneuropathy.26 Single fiber elec- clinical progression demonstrated more
tromyography may also help detect conduc- blocking.
tion block in focal neuropathy.27 Daube and Mulder31 reported mildly in-
creased fiber density and more markedly in-
ANTERIOR HORN CELL DISORDERS
creased jitter and blocking in 31 unselected
patients with amyotrophic lateral sclerosis.
Stalberg et al.28 reported the SFEMG find- The patient's age, clinical severity, CMAP
ings in 21 patients with anterior horn cell amplitude, and presence of a decrement to
disease and in 3 with syringomyelia. All pa- slow repetitive stimulation were valuable in
tients had increased fiber density. The in- predicting longevity. However, SFEMG find-
crease was greatest in anterior horn cell dis- ings did not add to the prognostic accuracy.
orders that were slowly progressive (fiber Single fiber study of spontaneously recorded
density, 5.4). The increase (fiber density, fasciculations in patients with amyotrophic
3.3) was less in rapidly progressive amyo- lateral sclerosis has documented increased
trophic lateral sclerosis. Increased jitter and jitter and blocking in those discharges.
blocking were observed in all these condi- In neurogenic disorders of all types, the
tions: the largest increase was in amyo- abnormalities on SFEMG complement those
trophic lateral sclerosis, and the increase was seen during conventional needle EMG. The
less in the spinal muscular atrophies and single fiber profile can delineate the rate of
syringomyelia. In the chronic conditions, progression and longevity of the disease pro-
the complexes (particularly the initial part) cess. Whether SFEMG will become clinically
were more stable. Duration of single fiber useful in establishing prognosis in certain
potentials varied considerably. However, the disease states or in serially following the dis-
longest durations were seen in the more ease during treatment trials or for evaluat-
chronic conditions. The authors concluded ing reinnervation potential is not known.
that the dual findings of moderately in-
creased fiber density and unstable com-
plexes of varying configuration represent a SUMMARY
rapidly progressive process with active rein-
nervation, such as amyotrophic lateral scle- Single fiber electromyography is a highly se-
rosis. Markedly increased fiber density and lective technique that permits recording of
relatively stable complexes (particularly of individual components of the motor unit.
the initial part) indicate a slowly progressive The selectivity of SFEMG depends on the use
disease or burned-out process with long- of a low-frequency filter of 500 Hz and a
standing reinnervation. The combination of small electrode size. Voluntary activation or
markedly increased fiber density and unsta- electric stimulation is used to activate the
ble potentials was believed to reflect reacti- muscle fiber in SFEMG. Voluntary activa-
vation of a long-standing process.29 tion allows measurement of fiber density,
Schwartz et al.30 reported similar conclu- whereas jitter and blocking are recorded
sions in 10 patients with long-standing sy- with both voluntary and stimulated SFEMG.
ringomyelia. Single fiber electromyography Single fiber electromyography is techni-
abnormalities (and clinical changes) were cally demanding and requires specialized
maximum in muscles innervated by spinal recording equipment. A variety of modern
segments C8 and Tl. In the first dorsal in- digital equipment is available that assists in
terosseous muscle, mean fiber density was the collection, display, analysis, reporting,
4.1, with 21% of potential pairs demonstrat- and archiving of SFEMG data.
ing increased jitter and 7% demonstrating Single fiber electromyography is the most
blocking. The distribution of abnormalities, sensitive clinical electrophysiologic tech-
rather than the type, differentiated these pa- nique for the detection of a defect of neu-
tients from those with anterior horn cell dis- romuscular transmission. The findings are
not specific or diagnostic for individual dis- lecular genetic and patch-clamp studies. Ann N Y
eases. Single fiber electromyography findings Acad Sci 841:140-156, 1998.
15. Chaudhry V, Crawford TO. Stimulation single-fiber
also are abnormal in disorders associated EMG in infant botulism. Muscle Nerve 22:1698-
with denervation and reinnervation of mus- 1703, 1999.
cle, such as certain myopathies, motor neu- 16. Padua L, Aprile I, Monaco ML, et al. Neurophysi-
ron diseases, and peripheral neuropathies. ological assessment in the diagnosis of botulism:
usefulness of single-fiber EMG. Muscle Nerve 22:
Correlation of fiber density and jitter/ 1388-1392, 1999.
blocking analysis may help to determine dis- 17. Trontelj JV, Stalberg E, Mihelin M, Khuraibet A. Jit-
ease chronicity and rate of progression. ter of the stimulated motor axon. Muscle Nerve
15:449-454, 1992.
18. Mandler RN, Maselli RA. Stimulated single-fiber
electromyography in wound botulism. Muscle
REFERENCES Nerve 19:1171-1173, 1996.
19. Lin TS, Chiu HC. Motor end-plate jitter in myas-
1. Bromberg MB, Scott DM. Single fiber EMG refer- thenia gravis at different firing rates. J Clin Neuro-
ence values: reformatted in tabular form. AD HOC physiol 15:262-267, 1998.
Committee of the AAEM Single Fiber Special In- 20. De Grandis D, Fiaschi A, Tomelleri G, Orrico D.
terest Group. Muscle Nerve 17:820-821, 1994. Hypokalemic periodic paralysis. A single fiber elec-
2. Sanders DB, Howard JF Jr, Johns TR. Single-fiber tromyographic study. J Neurol Sci 37:107-112,
electromyography in myasthenia gravis. Neurology 1978.
29:68-76, 1979. 21. Ukachoke C, Ashby P, Basinski A, Sharpe JA. Use-
3. Stalberg E, Ekstedt J, Broman A. The electromyo- fulness of single fiber EMG for distinguishing neu-
graphic jitter in normal human muscles. Electro- romuscular from other causes of ocular muscle
encephalogr Clin Neurophysiol 31:429-438, 1971. weakness. CanJ Neurol Sci 21:125-128, 1994.
4. Stalberg E, Trontelj JV, Schwartz MS. Single- 22. Bertorini TE, Stalberg E, Yuson CP, Engel WK. Sin-
muscle-fiber recording of the jitter phenomenon in gle-fiber electromyography in neuromuscular dis-
patients with myasthenia gravis and in members of orders: correlation of muscle histochemistry, single-
their families. Ann N Y Acad Sci 274:189-202,1976. fiber electromyography, and clinical findings.
5. Murga L, Sanchez F, Menendez C, Castilla JM. Di- Muscle Nerve 17:345-353, 1994.
agnostic yield of stimulation and voluntary single- 23. Foote RA, O'Fallon WM, Daube JR. A comparison
fiber electromyography in myasthenia gravis. Mus- of single fiber and routine EMG in normal subjects
cle Nerve 21:1081-1083, 1998. and patients with inflammatory myopathy. Bull Los
6. Stalberg E, Trontelj J (eds). Single Fiber Elec- Angeles Neurol Soc 43:95-103, 1978.
tromyography: Studies in Healthy and Diseased 24. Thiele B, Stalberg E. Single fibre EMG findings in
Muscle, 2nd ed. Raven Press, New York, 1994. polyneuropathies of different aetiology. J Neurol
7. Konishi T, Nishitani H, Matsubara F, Ohta M. Myas- Neurosurg Psychiatry 38:881-887, 1975.
thenia gravis: relation between jitter in single-fiber 25. SchwarzJ, PlanckJ, BriegelJ, Straube A. Single-fiber
EMG and antibody to acetylcholine receptor. Neu- electromyography, nerve conduction studies, and
rology 31:386-392, 1981. conventional electromyography in patients with
8. Kelly JJ Jr, Daube JR, Lennon VA, Howard FM Jr, critical-illness polyneuropathy: evidence for a lesion
Younge BR. The laboratory diagnosis of mild myas- of terminal motor axons. Muscle Nerve 20:696-701,
thenia gravis. Ann Neurol 12:238-242, 1982. 1997.
9. Padua L, Stalberg E, LoMonaco M, Evoli A, Batoc- 26. Bril V, Werb MR, Greene DA, Sima AA. Single-fiber
chi A, Tonali P. SFEMG in ocular myasthenia gravis electromyography in diabetic peripheral polyneu-
diagnosis. Clin Neurophysiol 111:1203-1207, 2000. ropathy. Muscle Nerve 19:2-9, 1996.
10. Valls-Canals J, Montero J, Pradas J. Stimulated sin- 27. Padua L, Aprile I, D'Amico P, et al. A useful elec-
gle fiber EMG of the frontalis muscle in the diag- trophysiological test for diagnosis of minimal con-
nosis of ocular myasthenia. Muscle Nerve 23:779- duction block. Clin Neurophysiol 112:1041-1048,
783, 2000. 2001.
11. Weinberg DH, Rizzo JF III, Hayes MT, Kneeland 28. Stalberg E, Schwartz MS, Trontelj JV. Single fibre
MD, Kelly JJ Jr. Ocular myasthenia gravis: predic- electromyography in various processes affecting the
tive value of single-fiber electromyography. Muscle anterior horn cell. J Neurol Sci 24:403-415, 1975.
Nerve 22:1222-1227, 1999. 29. Rodriquez AA, Agre JC, Franke TM. Electromyo-
12. Sonoo M, Uesugi H, Mochizuki A, Hatanaka Y, graphic and neuromuscular variables in unstable
Shimizu T. Single fiber EMG and repetitive nerve postpolio subjects, stable postpolio subjects, and
stimulation of the same extensor digitorum com- control subjects. Arch Phys Med Rehabil 78:986-
munis muscle in myasthenia gravis. Clin Neuro- 991, 1997.
physiol 112:300-303, 2001. 30. Schwartz MS, Stalberg E, Swash M. Pattern of seg-
13. Ukachoke C, Ashby P, Basinski A, Sharpe JA. Use- mental motor involvement in syringomyelia: a sin-
fulness of single fiber EMG for distinguishing neu- gle fibre EMG study. J Neurol Neurosurg Psychia-
romuscular from other causes of ocular muscle try 43:150-155, 1980.
weakness. CanJ Neurol Sci 21:125-128, 1994. 31. Daube JR, Mulder DW. Prognostic factors in amy-
14. Engel AG, Ohno K, Milone M, Sine SM. Congeni- otrophic lateral sclerosis (abstract). Muscle Nerve
tal myasthenic syndromes. New insights from mo- 5 (Suppl):S107, 1982.
Chapter 27
ESTIMATING THE NUMBER OF
MOTOR UNITS IN A MUSCLE
Jasper R. Daube
MOTOR UNIT NUMBER ESTIMATE BY Spike-Triggered Motor Unit Potential

STANDARD ELECTROMYOGRAPHY Averaging
Interference Pattern All-or-None Increments in the Compound
Recruitment Analysis Muscle Action Potential
MOTOR UNIT NUMBER ESTIMATE BY F-Wave Measurements
STANDARD NERVE CONDUCTION Statistical Measurements
STUDIES CLINICAL APPLICATIONS
QUANTITATIVE MOTOR UNIT NUMBER SUMMARY
ESTIMATE
Underlying Assumptions of Quantitative
Motor Unit Number Estimate
Peripheral neuromuscular diseases are di- ciculation. Physiologic disorders may cause
vided broadly into neurogenic and myopathic a loss of function (as in conduction block)
diseases, with diseases of the neuromuscular or no change in function (as in slowing of
junction often being included with myo- conduction). Degenerative or destructive
pathic diseases. Neurogenic and myopathic processes result in the loss of an entire mo-
diseases both result in the clinical problem tor neuron or peripheral axon. The loss of
of muscle weakness. The task of clinicians is, either motor axons or neurons and con-
first, to distinguish between these two broad duction block are the basis of the weakness
groups of diseases and, second, to identify found in most patients with a neurogenic dis-
the specific type. ease. The severity of the clinical deficit is re-
The critical difference between neuro- lated directly to the number of motor neu-
genic and myopathic diseases is that motor rons or axons (or both) that are lost or
neurons or their axons are involved in neu- blocked. Therefore, an important part of the
rogenic diseases. Damage to either the an- assessment of neuromuscular disease is to
terior horn cell or its peripheral axon is determine the number of functioning mo-
the essence of a neurogenic process, even tor units.1'2 It would be ideal to have an ac-
though the associated change in muscle may tual measure of the number of motor units,
be more apparent clinically. Neurogenic but current methods—physiologic, histo-
diseases may produce histologic changes or logic, clinical, and histopathologic—are not
physiologic changes without histopathologic able to provide it. Only electrophysiologic
correlates. An example of a physiologic ab- methods can be used to estimate the num-
normality is the irritability, with excessive ac- ber of motor units in a muscle. This chapter
tivity in the motor nerve, that produces fas- describes the methods that have been de-
358
Estimating the Number of Motor Units in a Muscle 359
veloped for estimating the number of motor Two distinctly different methods of
units and for determining the number that MUNE have been used: one uses needle
has been lost through neurogenic disease. electromyography (EMG) and the other
Some of the terms used in this chapter are uses motor nerve stimulation and surface
defined as follows: motor unit refers to a sin- recordings. They are described below.
gle anterior horn cell or brain stem motor
neuron, its peripheral axon (which travels in
a cranial or peripheral nerve), and each of MOTOR UNIT NUMBER
the muscle fibers the axon innervates. Num-
ber of motor units refers to the number of func-
ESTIMATE BY STANDARD
tioning motor neurons or functioning mo- ELECTROMYOGRAPHY
tor axons innervating a muscle or group of
muscles. A physiologic determination of the Standard diagnostic clinical EMG has always
number of motor units is called motor unit included a subjective estimate of the num-
number estimate (MUNE). ber of motor units in a muscle. Electromy-
Physiologic estimates of the number of ographers have used recruitment analysis or
motor units have been hampered by the interference pattern analysis (or both) with
lack of a standard determination of the voluntarily activated EMG recordings to
number of motor units in a muscle. At best, judge the number of motor units in a mus-
histopathologic and anatomical determina- cle.- 6 The EMG recorded with either surface
tions are very rough measures. Attempts or intramuscular electrodes during strong
have been made to measure the number of voluntary contractions summates the activity
motor units in human fetal and newborn tis- of the muscle fibers in the activated motor
sue.3 In both studies, individual muscles and units to produce an interference pattern.
the motor nerves innervating them were dis- The greater the number of motor units ac-
sected and counted. The number of large tivated, the greater the density of the EMG
myelinated axons in the motor nerve was pattern. The increased density with in-
counted and divided by two to estimate what creased effort is the result of an increase in
proportion of these fibers were motor rather the number of motor units activated and an
than sensory. This proportion was based on increase in the firing rate of individual mo-
animal studies in which degeneration of the tor units. The combination of firing rate and
peripheral sensory axons after section of the the number of motor units reduces the reli-
dorsal root indicated that approximately ability of density measures in determining
half of the axons in a motor nerve are sen- the number of motor units. Density mea-
sory.4 This proportion of large myelinated sures of the number of motor units are fur-
axons then served as the estimate for the ther beset by the problem of having to rely
number of motor units innervating the mus- on the effort of the patient for obtaining ac-
cle. These studies also counted the number tivation. Clearly, the effort of the patient al-
of muscle fibers in a muscle to determine ters the number of motor units activated.
the innervation ratio of muscle fibers to ax- Nonetheless, measures of density, whether
ons. Although the results of the two studies from subjective or automated methods, have
were similar, the values were sufficiently dif- been used to make judgments about the loss
ferent to preclude the designation of a true of motor units. When the loss is moderate
standard measurement of the number of to severe, these methods can identify a loss
motor units innervating individual human of motor units, for example, as in a reduced
muscles. However, these studies serve as a interference pattern or a single motor unit
baseline comparison for the physiologic firing pattern.
methods that have been developed for
MUNE. The absence of a standard makes di-
rect comparison of the values obtained by Interference Pattern
different methods an equally important part
of the assessment of the validity of individ- The methods of interference pattern analy-
ual methods of MUNE. sis can be classified into two approaches
(Chapter 25). The more common one is the motor units. However, determining the rate
measurement of the number of turns (that of firing is one of the more difficult steps in
is, reversals of potential per unit time) and standard EMG. The ratio of the number of
the amplitude of the spikes in the EMG pat- units firing to the rate of firing can provide
tern. The other method analyzes the fre- a rough gauge of the number of motor units.
quency components of the EMG pattern. This semiquantitative method of determin-
Both methods provide some measure of den- ing reduced recruitment provides a more
sity and, indirectly, of the number of motor accurate and reproducible estimate of the
units. Neither method is reliable enough to number of motor units than does interfer-
provide a quantitative measure of the num- ence pattern analysis. Although recruitment
ber of motor units. Alterations in motor unit analysis is reasonably reproducible and clin-
potentials with disease further reduce the re- ically reliable, it usually is a subjective judg-
liability of both methods. For example, in ment made by electromyographers on the
some disease processes, the potential gener- basis of experience. It requires taking into
ated by individual motor units becomes account differences in recruitment in dif-
more complex with multiple phases. These ferent areas of individual muscles and the
additional phases contribute to both the even greater differences among different
high-frequency components and the num- muscles. Automated methods for formally
ber of turns, but they do not reflect a change quantifying the recruitment pattern have
in the number of motor units. Thus, inter- been developed (see Chapter 25). These for-
ference pattern analysis is unsatisfactory for mal quantitative measures can provide evi-
MUNE. dence of the reliability of the clinical meth-
ods; however, they are time-consuming and
complex and have not found clinical appli-
Recruitment Analysis cation. None of these methods provides an
actual estimate of the number of motor units
The second method used by clinical elec- in a muscle.
tromyographers to judge the number of mo-
tor units in a muscle is recruitment analysis
(see Chapter 24). Recruitment refers to the
initiation of firing of additional motor units MOTOR UNIT NUMBER
as the effort of voluntary contraction in- ESTIMATE BY STANDARD
creases. The intrinsic anatomical and physi- NERVE CONDUCTION STUDIES
ologic properties of motor neurons result in
a fixed pattern of activation in response to Motor nerve conduction studies are an im-
voluntary effort. During voluntary activation, portant part of the electrophysiologic analy-
low-threshold motor neurons begin firing at sis of peripheral neuromuscular disease (see
rates of 5-7 Hz and increase their firing fre- Chapter 21). The amplitudes of compound
quency with increasing effort. As the effort muscle action potentials (CMAPs) are re-
increases, additional motor units are acti- lated directly to the number and size of mus-
vated, and they, in turn, increase their fre- cle fibers in a muscle group and indirectly
quency of firing with further increases in ef- to the number of motor units in the muscle
fort. The recruitment of motor units during group.7 If a disease is known to be neuro-
activation is a fixed relationship between the genic and acute, the amplitude of a CMAP
number of activated motor units and their is a rough estimate of the number of motor
firing rates. In recruitment analysis, the units. Its value is limited by two factors. First,
number of motor unit potentials activated at the amplitude is decreased in myopathies
any given level of effort is compared with the with loss of muscle fiber tissue. Second, the
rate of firing of individual motor units. This CMAP reduction that occurs with the de-
ratio provides an indirect measure of the struction of axons can be compensated for
number of motor units in the muscle. partially or fully by reinnervation from col-
Clinical EMG judgments about the num- lateral sprouting of intact axons. Estimates
ber of motor units compare the rate of fir- of the number of motor units made on the
ing of single units with the total number of basis of the amplitude of the CMAP are ham-
pered further by the wide range of normal activates all the muscles innervated by
amplitudes. For example, even in the case of that nerve distal to the point of stimu-
acute traumatic section of a peripheral nerve lation. Therefore, measurements of the
that disrupts half of the motor axons, the CMAP are the summation of activity
amplitude of the ulnar/hypothenar CMAP from multiple muscles. For example,
may decrease from 12 to 6 mV and still be the median/thenar CMAP is the sum-
within normal limits. In the clinical setting mation of the activity of the opponens
in which the baseline amplitude is not pollicis, abductor pollicis brevis, flexor
known, the CMAP is only a rough guide to pollicis brevis, and, to a lesser extent,
the number of motor units. Therefore, the the lumbrical muscles. The ulnar/hy-
amplitude of the CMAP cannot be used to pothenar CMAP is the summation of all
obtain a reliable MUNE. the other intrinsic muscles of the hand.
Thus, MUNE is more accurately an es-
timate of the number of motor units in
QUANTITATIVE MOTOR UNIT groups of muscles rather than in a sin-
NUMBER ESTIMATE gle muscle. Moreover, the SMUPs in
more distant muscles contribute less to
Needle EMG and motor nerve stimulation the CMAP than do SMUPs in more
methods can provide quantitative MUNE.8"13 proximal muscles.
Both the needle EMG and motor nerve stim- 2. Each of the MUNE methods assumes
ulation approaches to MUNE make basic as- that each SMUP has the same size each
sumptions about the electric characteristics time it is activated. Thus, defects of
of motor unit potentials. Four methods of neuromuscular transmission that result
making quantitative MUNE have been de- in varying sizes of the SMUP will cause
veloped. Three of them record CMAPs in re- inaccuracy.
sponse to nerve stimulation, and the fourth 3. Although the assumption that all mo-
one relies on needle EMG. Variations of tor axons are activated by supramaxi-
each of these four basic techniques continue mal stimulation is generally true, it may
to evolve to improve on the accuracy and re- not be true of disease involving high-
liability of MUNE. Each method measures threshold axons (for example, severe
the average size of the potentials generated demyelination and regenerated ax-
by single motor units—single motor unit po- ons). In these situations, the supra-
tentials (SMUPs)—and the size of the CMAP maximal CMAP may be difficult to
obtained with supramaximal stimulation of obtain.
a motor nerve. Motor unit number estimate 4. The most critical issue in determining
is obtained by dividing the supramaximal the size of individual motor unit po-
CMAP by the average size of the SMUP. The tentials is the adequacy of sampling of
techniques differ in how they obtain the av- all the SMUPs in the muscle.14 In pa-
erage size of the SMUP. tients with severe neurogenic disease,
it is possible to identify each motor unit
and to obtain a reliable, reproducible,
direct count of up to 10 motor units.
Underlying Assumptions of This is the practical maximum of such
Quantitative Motor Unit direct counts. With more than 10 mo-
Number Estimate tor units, none of the methods allow re-
liable measures of each motor unit, so
All the techniques rely on several underly- it is uncertain whether a true count has
ing assumptions that must be understood to been made. In these cases, it is neces-
appreciate fully the advantages and draw- sary to measure the size of a subset of
backs of quantitative MUNE. the total population of motor units and
1. Motor unit number estimates are as- then calculate an average size for that
sumed to be measurements from a sin- subgroup. If all the motor units are of
gle muscle, but supramaximal stimula- nearly identical size, sampling a subset
tion of any peripheral motor nerve of them to obtain an average size of the
SMUP is a valid approach. With greater First, the method assumes that all motor unit
variation in the size of SMUPs, partic- potentials can be recorded at the surface.
ularly if the size range is not a normal Studies by Brown and coworkers17 suggest
distribution, estimates of the true size that this is true in superficial muscles. Sec-
become less reliable. Each of the meth- ond, the technique assumes that voluntary
ods for measuring the size of SMUPs activation recruits the full range of sizes of
must address the adequacy of the sam- motor units. This likely is not true. Larger
pling to determine whether a reason- motor units probably are not activated with
able representation of the total popu- standard voluntary contraction. Despite
lation of motor unit sizes has been these issues, the values obtained with the
obtained. method are comparable to those expected
Each of these assumptions is a possible on the basis of animal studies and those ob-
source error in each MUNE method. None- tained with other methods of recording
theless, the ease of measuring the small MUNE. De Koning et al.18 have modified
number of potentials in chronic neurogenic the technique by using macro-EMG needles
disorders with moderate to marked loss of to record CMAPs. It is assumed that a
motor units minimizes these sources of er- macroneedle provides a better representa-
ror and makes MUNE highly reliable. tion of the full range of motor units, partic-
ularly those deeper in the muscle. Milner-
Brown and Brown19 used the technique of
microstimulation of nerve terminals in the
Spike-Triggered Motor Unit end plate region to activate motor units
Potential Averaging recorded with a needle electrode. This re-
duces the bias in the selection of motor unit
Spike-triggered averaging relies on the ability to sizes that occurs with voluntary activation.
isolate SMUPs by voluntary activation on Each of these methods gives comparable
needle EMG with a two-channel EMG ma- MUNE.
chine15-16 (Fig. 27-1). Intramuscular motor The spike-triggered averaging methods
unit potentials are recorded on one channel generally are more time-consuming and
with one of several electrodes, including sin- more complex to perform because of the
gle-fiber EMG, bipolar concentric, standard need for two channels of recording: a mo-
concentric, or fine-wire electrodes. Individ- tor unit potential triggering channel and a
ual motor unit potentials are isolated on the SMUP averaging channel. The accuracy of
first channel, usually by an amplitude trig- the methods depends on the ability of the
ger window that selects potentials on the ba- patient and electromyographer to activate,
sis of peak amplitudes. The surface-recorded identify, and trigger individual motor unit
size of SMUPs is measured on a second chan- potentials for a period long enough to allow
nel of the EMG machine that is triggered by the size of the SMUPs to be measured.
the needle-recorded motor unit potential on
the first channel. The activity is averaged on
the second channel from the same surface
electrodes used to record the CMAP. The All-or-None Increments in
technique requires the isolation of at least the Compound Muscle
10, and preferably 20, SMUPs whose spike- Action Potential
triggered average can be recorded on the
surface. The amplitude or area of the sur- All-or-none increment measurement, introduced
face-recorded potentials is used to calculate by McComas,20 was the first method used for
the average size of the SMUPs in the mus- quantitative MUNE. The method is decep-
cle. Motor unit number estimate is then cal- tively simple and provides the easiest and
culated by dividing the size (area or ampli- most direct and reliable method of obtain-
tude) of the supramaximal CMAP by the ing MUNE.21'22 It is based on the well-known
average size of the SMUPs. all-or-none characteristic of the activation of
Several of the assumptions made in using peripheral motor axons with electric stimu-
this technique are possible sources of error. lation. In the incremental method, the stim-
Figure 27-1. Spike-triggered averaging for motor unit number estimate. A, Compound muscle action potential.
B and C, Each line (MU\-MUg) shows the triggered motor unit (MU) potential recorded with a needle (solid lines)
and the averaged surface compound muscle action potential (dotted line). (From Brown WF, Strong MJ, Snow R.
Methods for estimating numbers of motor units in biceps-brachialis muscles and losses of motor units with aging.
Muscle Nerve 11:423-432, 1988. By permission of John Wiley & Sons.)
ulus current is finely controlled in very small old of the axons of both motor units. When
steps designed to allow isolated stimulation the threshold of one of the axons is reached,
of individual motor units in a progressive the axon is fully activated and the CMAP sud-
fashion (Fig. 27-2). For example, if a mus- denly changes from no response to the re-
cle contains only two motor units, the CMAP sponse of the SMUP. This SMUP would
consists of the SMUPs of these two poten- make up roughly half of the size of the
tials only. Incremental testing with slowly CMAP. When the threshold of the second
and gradually increasing current will show axon is reached, this axon also fires and the
no response to a stimulus below the thresh- maximal CMAP is obtained. Changing the
taining three SMUPs of slightly different size

but of nearly the same threshold that result
in seven rather than three apparent incre-
ments (Fig. 27-3). The threshold of all ax-
ons varies within a small range so that for
any given stimulus, an SMUP has a percent-
age likelihood of firing. Therefore, any one
of the three axons with nearly identical
thresholds might be activated for each stim-
ulus. An axon that is activated first in one
trial may be activated second or third in sub-
sequent trials. Thus, the sizes of potentials
that could be obtained when there are three
motor units of different sizes, A, B, and C,
are those generated by A alone, B alone, C
alone, by A and C, B and C, A and B, and by
A, B, and C together, and three SMUPs might
be recorded as three to seven steps.
Several modifications have been devel-
oped to minimize these errors: (1) use of au-
tomated computer measurement of the tem-
plates of different SMUPs to decrease the
likelihood of measuring alternation, (2) stim-
ulation at different points along the nerve
(multipoint stimulation) to isolate only a sin-
B
gle SMUP,12'23"25 (3) use of recording elec-
trodes of different sizes and shapes, (4) use
Figure 27-2. Incremental method of motor unit num- of an automated method of incrementing
ber estimate (MUNE). A, Ten response increments in
500 fjV give a MUNE of 100. B, Ten response incre- the stimulus size,26 and (5) microstimulation
ments in 2000 /iV give a MUNE of 25. (From Brown
and Feasby.41 By permission of Elsevier Science.)
stimulus current above and below these

thresholds produces stepwise activation of
two steps, that is, the first SMUP and then
the full CMAP. If there are three motor
units in the muscle, three steps would be
recorded, and similarly for a larger number
of motor units. With this technique, the size
of the SMUP is estimated from the incre-
mental change in the CMAP, with control of
the stimulus current and a progressively in-
creasing number of motor units. The more
of these distinct steps of the total CMAP that
can be measured, the more reliable MUNE
becomes with incremental measurements.
Incremental MUNE has potential sources
of error. First, there could be a selection bias Figure 27-3. Alternation of three single motor unit po-
for larger or smaller motor units. Second, tentials during F-wave recording to give seven F waves
the occurrence of alternation caused by sim- (A-G). (From Feasby TE, Brown WF. Variation of mo-
tor unit size in the human extensor digitorum brevis
ilar SMUP thresholds can result in an error and thenar muscles. J Neurol Neurosurg Psychiatry
in MUNE measurement. Alternation is best il- 37:916-926, 1974. By permission of the Journal of Neu-
lustrated by the example of a muscle con- rology, Neurosurgery, and Psychiatry.)
of single nerve terminals at the end plate mal stimulation and template matching may
97 40
9R
region/'' make the method useful clinically. Simmons
The McComas incremental CMAP tech- et al.31 have described an automated method
nique uses whatever average values are ob- of measuring only recurrent, identical F
tained for the size (amplitude or area) of the waves that are more likely to represent
SMUP and compares them with the supra- SMUPs. This method has not been used
maximal size of the CMAP. Normal values widely because of the difficulty with imple-
determined by several authors have shown menting it.
that the mean normal MUNE for me-
dian/thenar muscles is approximately 350
(range, 100-500) and for the peroneal/ Statistical Measurements
extensor digitorum brevis, the other well-
studied muscle, 200 (range, 50-300). The fourth method of estimating the size of
The incremental technique is unreliable the SMUP uses direct stimulation of the mo-
when the motor unit potentials become tor nerve, similar to the all-or-none incre-
quite small, as in severe myopathies, facial mental method, but it is conceptually dif-
muscles, or widi nascent motor unit poten- ferent.9'32'33 With the statistical method, no
tials. The inability to record the steps results attempt is made to identify the potentials as-
in underestimation of MUNE. Nonetheless, sociated with individual motor units. The
the method of incremental CMAP is direct method relies on the known relationship be-
enough that it should be learned by every tween the variance of multiple measures of
electromyographer. It can be applied to pa- step functions and the size of the individual
tients in whom loss of motor units allows the steps when the steps have a Poisson distri-
remaining axons to be stimulated selectively bution. Poisson statistics are used to calcu-
and recorded as incremental steps by stim- late the number of quanta released from a
ulus control. nerve terminal at the neuromuscular junc-
tion when the individual quanta are too
small to be distinguished, as in myasthenia
F-Wave Measurements gravis. In Poisson statistics, the sizes of a se-
ries of measurements are multiples of the
F waves have been suggested as a method for size of a single component. Therefore, a
determining the size of the SMUP.29'30 Poisson distribution has discrete values at
Supramaximal activation of all the motor which responses are found. A Poisson dis-
axons in a peripheral nerve is associated with tribution has decreasing numbers of re-
antidromic activation of some of the ante- sponses with higher values. In a Poisson
rior horn cells. The small proportion of an- distribution, the variance of series of mea-
terior horn cells activated antidromically surements is equal to the size of the indi-
produces small late potentials, .F waves. Re- vidual components that make up each mea-
peated supramaximal stimuli activate differ- surement. The statistical MUNE method
ent anterior horn cells and produce differ- measures the variance of the CMAP and
ent F waves. Recording a range of sizes of does not require identification of individual
F waves can be used to estimate the average components; it can be used when the sizes
size of the SMUP. This average size can be of SMUPs are too small to be isolated, which
divided into the supramaximal CMAP to ob- is often the case in normal muscles and my-
tain MUNE. Simmons et al.31 have shown opathies. Also, it can be used with high-am-
that the drawbacks of this method are simi- plitude CMAPs that require gains at which
lar to those described in the preceding sec- the SMUP cannot be isolated.
tion for the incremental method and that Although the statistical method has ad-
multiple motor units are activated more vantages, it has the potential sources of er-
commonly than single motor units (Fig. ror described above for the other methods.
27-3). These drawbacks result in overesti- These include the following assumptions:
mating the average size of the SMUPs and each motor unit has a similar size, it is the
in underestimating the MUNE. Automated same size each time it is activated, the sam-
correction of these drawbacks by submaxi- ples tested are unbiased, and all motor ax-
ons are activated. Two other issues must be appropriate nerve. An initial "scan" of the
considered for statistical MUNE. First, with responses of the nerve to 30 stimuli of equal
a larger number of SMUPs making up the increments between threshold and maximal
CMAP, the distributions shift gradually from CMAP identifies large increments that may
a Poisson to a normal distribution. This may result from a large SMUP (Fig. 27-4). Such
produce an error of up to 10% in the statis- large SMUPs do not need further statistical
tical MUNE. Second, because all measure- testing. A sequence of 30 or more submaxi-
ments are statistical, the results vary with mal stimuli is given at a fixed stimulus in-
each sample. Consequently, the number of tensity in selected regions of small incre-
samples must be increased to provide re- ments on the scan. The threshold of
producibility comparable to that of the other individual motor axons fluctuates so that at
methods. any given intensity the likelihood of firing
In the statistical method, recording elec- ranges from 0% to 100%, with a finite range
trodes are applied as they are for standard where the axon only fires some percentage
nerve conduction studies, with the stimulat- of the time (Fig. 27-5). This inherent vari-
ing electrode taped firmly in place over the ability of the threshold of individual axons
Figure 27-4. Motor unit number estimate scan from, A, normal subject and, B, patient with amyotrophic lateral
sclerosis. Between threshold and the maximal compound muscle action potential, 30 equal increments in stimulus
intensity were applied to the nerve. The elicited compound muscle action potentials are superimposed above the
histograms. The histograms depict the area of each of the 30 responses. In A, note the smooth curve with small in-
crements. In B, the increments are larger, with a particularly large increment just before the maximal compound
muscle action potential. The latter is caused by activation of a single large motor unit.
Figure 27-5. Statistical method of motor unit number estimate (MUNE) in a normal subject. A, Normal variation
in compound muscle action potentials (CMAPs) with 30 equal stimuli. B, Calculated single motor unit potentials
(SMUP) and MUNE for four groups of 30 stimuli (table on right) and two runs at different stimulus intensities (table
on left). Superimposed CMAPs from the 30 stimuli are shown on top left in A and B.
367
Figure 27-6. Reproducibility of motor unit number estimate (MUNE) (multipoint method). A, Subjects with de-
creased MUNE because of amyotrophic lateral sclerosis. B, Normal subjects. (From Felice.12 By permission of John
Wiley & Sons.)
causes intermittent firing of axons and con- nerve disease (carpal tunnel syndrome, lum-
tinuous variations in the size of the CMAP. bar radiculopathies, and peripheral neu-
Therefore, the problem of alternation with ropathies) and in motor neuron disease39"46
activation of different motor units described (Fig. 27-7). These methods are particularly
for the incremental and F-wave methods is helpful in chronic disorders in which collat-
not an issue with the statistical method. Be- eral sprouting results in a CMAP of normal
cause the method is a statistical measure- amplitude despite a loss of motor units. Se-
ment, a somewhat different result is ob- rial measurements of MUNE are helpful in
tained each time, and multiple trials are following the course of a disease, especially
needed to obtain the most accurate mea- in treatment trials or studies of the evolution
surement. Experimental testing with trials of of disease.47"50
more than 300 stimuli has shown that re-
peated measurement of groups of 30 until
the standard deviation of the repeated trials SUMMARY
is less than 10% provides a close estimate of
the number obtained with many more stim- The all-or-none incremental, the spike-
uli. Felice34 has shown that the reliability and triggered averaging, and the statistical meth-
reproducibility of MUNE is better than other ods of obtaining MUNE give similar values
measures of axonal loss (Fig. 27-6). Ongo- in different muscles of normal subjects. This
ing studies by different investigators have suggests that each of these methods can be
shown improvement in the reproducibility used whenever and wherever it is most fea-
and reliability of the statistical method of sible and that different methods may be ap-
MUNE.31'35-38 propriate in different settings. For example,
when each of the motor units can be iden-
tified with small increments of stimuli in a
CLINICAL APPLICATIONS severe neurogenic process, MUNE is defined
most rapidly and accurately by actually
Applications of the methods of MUNE to dis- counting the total number of increments.
eases have shown the expected decrease in When the number of motor units is too large
the number of motor units in peripheral to do this or their size is too small for them
Figure 27-7. Decrease in thenar motor unit number estimate (MUNE) (statistical method) in 16 patients with amyo-
trophic lateral sclerosis. Note the rapid decrease in MUNE over a few weeks in patients whose initial values were
normal, but the decrease was slower after the initial reduction in MUNE. F, female; M, male. Numeral following F
and M indicate age of patient.
to be identified accurately in the CMAP, the 5. Erim Z, De Luca CJ, Mineo K, Aoki T. Rank-ordered
statistical method or multipoint methods are regulation of motor units. Muscle Nerve 19:563-
573, 1996.
appropriate. In muscles in which CMAPs 6. Schulte-Mattler WJ, Georgiadis D, Tietze K, Zierz S.
cannot be obtained reliably, as in proximal Relation between maximum discharge rates on
muscles that are difficult to immobilize dur- electromyography and motor unit number esti-
ing stimulation of the motor nerve, the mates. Muscle Nerve 23:231-238, 2000.
7. Lateva ZC, McGill KG, Burgar CG. Anatomical and
spike-triggered averaging method is most ap- electrophysiological determinants of the human
propriate. The value of newer methods of thenar compound muscle action potential. Muscle
MUNE remains to be determined.51"55 Nerve 19:1457-1468, 1996.
8. Armon C, Brandstater ME, Peterson GW. Motor
unit number estimates and quantitative muscle
strength measurements of distal muscles in patients
REFERENCES with amyotrophic lateral sclerosis. Muscle Nerve
20:499-501, 1997.
1. Armon C, Brandstater ME. Motor unit number es- 9. Daube JR. Estimating the number of motor units in
timate-based rates of progression of ALS predict pa- a muscle. J Clin Neurophysiol 12:585-594, 1995.
tient survival. Muscle Nerve 22:1571-1575, 1999. 10. Doherty TJ, Brown WF. The estimated numbers and
2. Johnston SC. Prognostication matters. Muscle relative sizes of thenar motor units as selected by
Nerve 23:839-842, 2000. multiple point stimulation in young and older
3. Feinstein B, Lindegard B, Nyman E, Wohlfart G. adults. Muscle Nerve 16:355-366, 1993.
Morphologic studies of motor units in normal hu- 11. Doherty T, Simmons Z, O'Connell B, et al. Meth-
man muscles. Acta Anat Scand 23:127-142, 1955. ods for estimating the numbers of motor units in
4. Azzouz M, Leclerc N, Gurney M, Warter JM, human muscles. J Clin Neurophysiol 12:565-584,
Poindron P, Borg J. Progressive motor neuron im- 1995.
pairment in an animal model of familial amyotrophic 12. Felice KJ. Thenar motor unit number estimates us-
lateral sclerosis. Muscle Nerve 20:45-51, 1997. ing the multiple point stimulation technique: re-
producibility studies in ALS patients and normal Kothari MJ, Shefner JM. Reproducibility of motor
subjects. Muscle Nerve 18:1412-1416, 1995. unit number estimation in individual subjects. Mus-
13. Roos MR, Rice CL, Vandervoort AA. Age-related cle Nerve 24:467-473, 2001.
changes in motor unit function. Muscle Nerve 32. Olney RK, Lomen-Hoerth C. Motor unit number
20:679-690, 1997. estimation (MUNE): How may it contribute to the
14. Slawnych M, Laszlo C, Hershler C. Motor unit diagnosis of ALS? Amyotrophic Lateral Sclerosis &
number estimation: sample size considerations. Other Motor Neuron Disorders 1 (Suppl 2):S41-
Muscle Nerve 20:22-28, 1997. S44, 2000.
15. Bromberg MB. Motor unit estimation: repro- 33. Yuen EC, Olney RK Longitudinal study of fiber
ducibility of the spike-triggered averaging tech- density and motor unit number estimate in patients
nique in normal and ALS subjects. Muscle Nerve with amyotrophic lateral sclerosis. Neurology
16:466-471, 1993. 49:573-578, 1997.
16. Bromberg MB, Abrams JL. Sources of error in the 34. Felice KJ. A longitudinal study comparing thenar
spike-triggered averaging method of motor unit motor unit number estimates to other quantitative
number estimation (MUNE). Muscle Nerve 18: tests in patients with amyotrophic lateral sclerosis.
1139-1146, 1995. Muscle Nerve 20:179-185, 1997.
17. Brown WF, Doherty TJ, Chan M, Andres A, Provost 35. Aggarwal A, Nicholson G. Motor unit number esti-
SM. Human motor units in health and disease. mation in preclinical and amyotrophic lateral scle-
Muscle Nerve Suppl 9:S7-S18, 2000. rosis subjects (abstract). Muscle Nerve Suppl
18. de Koning P, Wieneke GH, van der Most van Spijk 7:S203, 1998.
D, van Huffelen AC, Gispen WH, Jennekens FG. Es- 36. Lomen-Hoerth C, Olney R. A comparison of mul-
timation of the number of motor units based on tipoint and statistical motor unit number estima-
macro-EMG. J Neurol Neurosurg Psychiatry 51: tion (abstract). Muscle Nerve 22:1304, 1999.
403-411, 1988. 37. Olney RK, Yuen EC, Engstrom JW. Statistical mo-
19. Milner-Brown HS, Brown WF. New methods of es- tor unit number estimation: reproducibility and
timating the number of motor units in a muscle. J sources of error in patients with amyotrophic lat-
Neurol Neurosurg Psychiatry 39:258-265, 1976. eral sclerosis. Muscle Nerve 23:193-197, 2000.
20. McComas AJ. Motor unit estimation: anxieties and 38. ShefnerJMJillapalli D, Bradshaw DY. Reducing in-
achievements. Muscle Nerve 18:369-379, 1995. tersubject variability in motor unit number estima-
21. Galea V, de Bruin H, Cavasin R, McComas AJ. The tion. Muscle Nerve 22:1457-1460, 1999.
numbers and relative sizes of motor units estimated 39. Bromberg MB, Forshew DA, Nau KL, Bromberg J,
by computer. Muscle Nerve 14:1123-1130, 1991. Simmons Z, Fries TJ. Motor unit number estima-
22. Stein RB, YangJF. Methods for estimating the num- tion, isometric strength, and electromyographic
ber of motor units in human muscles. Ann Neurol measures in amyotrophic lateral sclerosis. Muscle
28:487-495, 1990. Nerve 16:1213-1219, 1993.
23. Doherty TJ, Stashuk DW, Brown WF. Determinants 40. Bromberg MB, Larson WL. Relationships between
of mean motor unit size: impact on estimates of mo- motor-unit number estimates and isometric
tor unit number. Muscle Nerve 16:1326-1331,1993. strength in distal muscles in ALS/MND. J Neurol
24. Doherty TJ, Brown WF. A method for the longitu- Sci 139 (Suppl) :38-42, 1996.
dinal study of human thenar motor units. Muscle 41. Brown WF, Feasby TE. Estimates of functional motor
Nerve 17:1029-1036, 1994. axon loss in diabetics. J Neurol Sci 23:275-293,1974.
25. Wang FC, Delwaide PJ. Number and relative size of 42. Dantes M, McComas A. The extent and time course
thenar motor units estimated by an adapted multi- of motoneuron involvement in amyotrophic lateral
ple point stimulation method. Muscle Nerve 18: sclerosis. Muscle Nerve 14:416-421, 1991.
969-979, 1995. 43. Kuwabara S, Mizobuchi K, Ogawara K, Hattori T.
26. Aoyagi Y, Strohschein FJ, Ming Chan K. Use of the Dissociated small hand muscle involvement in amy-
collision technique to improve the accuracy of mo- otrophic lateral sclerosis detected by motor unit
tor unit number estimation. Clin Neurophysiol number estimates. Muscle Nerve 22:870-873, 1999.
111:1315-1319, 2000. 44. McComas AJ, Quartly C, Griggs RC. Early and late
27. Arasaki K, Tamaki M, Hosoya Y, Kudo N. Validity losses of motor units after poliomyelitis. Brain
of electromyograms and tension as a means of mo- 120:1415-1421, 1997.
tor unit number estimation. Muscle Nerve 20:552- 45. Smith BE, Stevens JC, Litchy WJ, et al. Longitudi-
560, 1997. nal electrodiagnostic studies in amyotrophic lateral
28. Arasaki K, Tamaki M. A loss of functional spinal al- sclerosis patients treated with recombinant human
pha motor neurons in amyotrophic lateral sclero- ciliary neurotrophic factor (abstract). Neurology 45
sis. Neurology 51:603-605, 1998. (Suppl 4):A448, 1995.
29. Felice KJ. Nerve conduction velocities of single 46. Krajewski KM, Lewis RA, Fuerst DR, et al. Neuro-
thenar motor axons based on the automated anal- logical dysfunction and axonal degeneration in
ysis of F waves in amyotrophic lateral sclerosis. Mus- Charcot-Marie-Tooth disease type 1A. Brain 123:
cle Nerve 21:756-761, 1998. 1516-1527, 2000.
30. Stashuk DW, Doherty TJ, Kassam A, Brown WF. Mo- 47. Chan KM, Stashuk DW, Brown WF. A longitudinal
tor unit number estimates based on the automated study of the pathophysiological changes in single
analysis of F-responses. Muscle Nerve 17:881-890, human thenar motor units in amyotrophic lateral
1994. sclerosis. Muscle Nerve 21:1714-1723, 1998.
31. Simmons Z, Epstein DK, Borg B, Mauger DT, 48. Gooch CL, Pleitez M, Harati Y. MUNE and single
motor unit tracking: experience in a clinical ALS 52. Roeleveld K, Sandberg A, Stalberg EV, Stegeman
trial. EEC Clin Neurophysiol 107 (Suppl 110):1001, DF. Motor unit size estimation of enlarged motor
1999. units with surface electromyography. Muscle Nerve
49. Gooch CL. Repetitive axonal stimulation of the 21:878-886, 1998.
same single motor unit: a longitudinal tracking 53. Shahani BT, Fang J, Dhand UK A new approach
study. Muscle Nerve 21:1537-1539, 1998. to motor unit estimation with surface EMG trig-
50. Kuwabara S, Ogawara K, Mizobuchi K, Mori M, Hat- gered averaging technique. Muscle Nerve 18:1088-
tori T. Mechanisms of early and late recovery in 1092, 1995.
acute motor axonal neuropathy. Muscle Nerve 54. Slawnych M, Laszlo C, Herschler C. Motor unit es-
24:288-291, 2001. timates obtained using the new "MUESA" method.
51. FangJ, Shahani BT, Graupe D. Motor unit number Muscle Nerve 19:626-636, 1996.
estimation by spatial-temporal summation of single 55. Sun TY, Lin TS, Chen JJ. Multielectrode surface
motor unit potentials. Muscle Nerve 20:461-468, EMG for noninvasive estimation of motor unit size.
1997. Muscle Nerve 22:1063-1070, 1999.
SECTION 2
Electrophysiologic
Assessment of
Neural Function
Part E
Reflexes and
Central Motor
Control
Motor function is controlled by a complex combination of central nervous

system circuits that involve all levels of the neuraxis. Local reflexes at the
level of the spinal cord or brain stem mediate and integrate local sensory
input and input from descending motor pathways to the motor unit. De-
scending motor pathways from the cerebral hemisphere to the spinal cord
include the rapidly conducting direct corticospinal pathways and several in-
direct pathways that arise in the spinal cord and brain stem. Descending mo-
tor activity in these pathways is directed and controlled by motor areas in
the cerebral cortex, basal ganglia, and cerebellum. The cerebellum and ba-
sal ganglia form feedback loops that extend through the thalamus to the
cerebral cortex and control motor activities.
Many of these pathways and functions can be monitored electrically, as
described in the chapters of this section. H reflexes (Chapter 28) and cra-
nial nerve reflexes (Chapter 29) are localized responses of the motor neu-
rons in the spinal cord and brain stem to localized sensory input. Both
groups of reflexes can be used to assess peripheral sensory and motor func-
tion as well as their central connections in the spinal cord and brain stem.
In contrast, long loop reflexes and the silent period depend more on the
descending motor pathways from the brain to die spinal cord (Chapter 30).
Therefore, these reflexes are useful primarily in elucidating central disor-
ders of motor function or neuronal excitability.
Multichannel surface electromyographic recordings from agonist and an-
tagonist muscles in the limbs and trunk can be used to characterize several
374 Reflexes and Central Motor Control
motor disorders on the basis of the patterns of activation and the timing of
activity in different muscles, either in one limb or longitudinally in the body
(Chapter 31). New knowledge has allowed improvement in the analysis and
classification of tremor.
Surface electromyographic recordings in posturography and electronys-
tagmography are also used in measuring the motor control of posture and
vestibular function. These measurements (Chapter 32) assess the long path-
ways that control motor function and their integration in the neuronal pools.
Posturography and electronystagmography are useful in evaluating many dis-
orders of both the vestibular pathways and motor control pathways. Their
applications have been expanded with new approaches to Dix-Hallpike, dy-
namic walking, and optokinetic rotary chair testing. These new approaches
have helped increase the application of the tests in Parkinson's and Alzhei-
mer's diseases and in vestibular rehabilitation.
Chapter 28
H REFLEXES
Kathryn A. Stolp-Smith
PHYSIOLOGIC BASIS CLINICAL APPLICATION

TECHNIQUE Proximal Conduction
Soleus and Gastrocnemius Techniques Central Nervous System Excitability
Flexor Carpi Radialis Technique SUMMARY
PEDIATRIC H REFLEXES
In 1918, Hoffmann1 recorded from the of die physiologic basis, sources of error,
soleus muscle a late response that occurred and clinical applications and limitations en-
with submaximal electric stimulation of the hances the usefulness of the H reflex.
tibial nerve, similar to die response recorded
as a result of a tendon tap. Magladery and
McDougal2 later termed this response the PHYSIOLOGIC BASIS
Hoffmann reflex, or H reflex. This reflex has
been studied widely and has been found use- Stimulation of the tibial nerve in the pop-
ful in assessing experimental and clinical as- liteal fossa at intensities less than needed to
pects of disorders of the central and pe- generate an early M response, a compound
ripheral nervous systems. The H reflex may muscle action potential, from the soleus
be obtained from other muscles, including muscle elicits the H reflex. Magladery and
the gastrocnemius and flexor carpi radialis. McDougal2 first demonstrated that die H re-
Lachman et al.3 showed that the H reflex flex is a monosynaptic reflex response pro-
can be used to evaluate proximal nerve duced by activation of a small proportion of
segments inaccessible to routine nerve con- soleus motor neurons (Fig. 28-1). Although
duction studies in cases of suspected radicu- similar to a tendon reflex, the H reflex by-
lopathy or plexopathy or sensory neuropa- passes the muscle spindle that initiates the
thy, especially if routine nerve conduction tendon reflex. Muscle spindle afferents have
studies did not detect an abnormality that is a lower threshold, especially to stimuli of
suspected clinically. These are the most prac- longer duration, than large soleus motor
tical and clinically applicable reasons for the axons.4 With gradually increasing stimulus
H reflex test. In the research literature, die intensity, die amplitude of the H reflex in-
H reflex usually is discussed as a test of mo- creases as more spindle afferents are acti-
tor neuron excitability to assess disorders vated. However, as stimulus intensity in-
causing spasticity or rigidity, such as myelop- creases further with motor axon activation,
athy, motor neuron disease, Parkinson's dis- the amplitude begins to decrease as more
ease, and cerebellar disorders. and more of the reflex volley is blocked in
Technically, die H reflex test is relatively the motor axons or neurons by antidromi-
simple to perform in a clinical neurophysi- cally conducted impulses. A stimulus of long
ology laboratory. A diorough understanding duration allows for more selective activation
375
lation to the spinal cord, and the time to de-

tect a compound muscle action potential by
the recording electrode.5'6
The amplitude of the H reflex, and hence
the ability to record the reflex, is extremely
variable. The value of measuring the ampli-
tude of the H reflex in clinical settings is de-
bated. Normal amplitude values of the tibial
H reflex vary from 0.1 to 7 mV; thus, ampli-
tude may not be a reliable measure for clin-
ical studies.7'8 The amplitude can vary with
recording electrode position, stimulus in-
tensity and duration, posture, age, and tem-
perature.9 Also, amplitude is sensitive to the
level of motor neuron excitability. Repeated
recordings of the H reflex may demonstrate
moment-to-moment amplitude variability by
Figure 28-1. Physiology of the H reflex: selective acti- as much as 1.5 mV.7 Activation of one group
vation of muscle spindle afferents and monosynaptic re- of motor units can inhibit motor units of
flex response of soleus motor axons. The F wave rep- other muscle groups, presumably through
resents recurrent discharge of motor neurons. recurrent inhibition. Thus, agonist contrac-
tion can increase the amplitude of the H re-
flex and antagonist contraction can decrease
of afferent axons, whereas a stimulus of short it.10 In addition, excitability can vary at dif-
duration increases the likelihood of motor ferent spinal cord segments. Studies of re-
axon activation.4 covery of the H reflex have demonstrated
The latency of the H reflex depends on that it is slower for the soleus H reflex than
several factors. These include the time to for the flexor carpi radialis H reflex, pre-
activate the primary spindle afferents, con- sumably because the motor neurons in the
duction velocity of the primary spindle af- sacral segments have different characteris-
ferents, central conduction delay, conduc- tics from those in the cervical segments.11
tion velocity of motor axons and terminal Factors that affect the H reflex and that
conduction delay, neuromuscular transmis- should be considered in performing this
sion delay, distance from the site of stimu- study are listed in Table 28-1. It is impor-
Table 28-1. Factors That Affect the Presence or Amplitude of the H Reflex1'3'7'12
Suppression Facilitation
Contraction of antagonist muscles Mild contraction of agonists

Strong contraction of agonists Passive stretch
Passive shortening Labyrinthine vestibular stimulation
Strong electric stimulus Bite, grasp
Sleep
Vibration
Drugs (nicotine, pentobarbital, diazepam)
Stimulation rate < 1/second
Tendon tap
Strong flexion/extension of neck muscles
Proximal ischemia
Spinal anesthesia
H Reflexes 377
Table 28-2. Comparison of the H Reflex and F Wave in Normal Subjects8'13'14

H Reflex FWave
Suppressed by supramaximal stimulation Maximal with supramaximal stimulation
Optimal with submaximal stimulation Infrequent with submaximal stimulation
Amplitude 50%-100% of M maximum Amplitude 10% of M maximum
Latency relatively constant Latency variable
Morphology constant Morphology variable
Recorded from forearm and leg Recorded from hand and foot muscles
tant to distinguish the H reflex from the F source of the reflex compound muscle ac-
wave, which has a similar latency. The F wave tion potential, recording over this muscle re-
is not a reflex but represents recurrent dis- sults in a larger waveform, with an initial
charge of several motor neurons, with alpha negative deflection, than recording over the
motor neurons and axons serving as both gastrocnemius muscle. The patient is placed
the afferent and efferent pathways. This pro- supine to avoid excessive lengthening or
duces a smaller amplitude wave of varying la- shortening of the soleus muscle and to allow
tency and morphology obtained with supra- adequate relaxation of the patient. The ac-
maximal stimulation. The F wave can be tive recording electrode is placed medially
recorded from many skeletal muscles (Table over the soleus muscle approximately mid-
28-2), and it can be distinguished from an way between the site of stimulation and the
H reflex by paired stimulation, which blocks medial malleolus. The tibial nerve is stimu-
the recurrent F wave but not the H reflex. lated immediately lateral to the popliteal
artery at the level of the popliteal crease. To
avoid stimulation of the peroneal nerve by
TECHNIQUE current spread, observe the resultant twitch
for gastrocnemius-soleus contraction with-
The H reflex can be recorded from many out associated contraction of the anterior
muscles in normal neonates and in adults tibial or peroneal muscle. The cathode is
with upper motor neuron lesions. In normal placed proximal to the anode. Square-wave
adults, it is obtained reliably from the soleus pulses of 0.5- to 1.0-millisecond duration,
and flexor carpi radialis muscles and has with an intensity of 25-30 V, are delivered at
been reported in the palmaris longus, flexor a rate of 0.5 Hz. Intensity should be in-
carpi ulnaris, anterior tibial, vastus medialis, creased by small increments from an initial
masseter, extensor digitorum communis, low level until a response larger than the pre-
and ulnar-innervated hand intrinsic mus- ceding M wave is elicited.
cles.8 The H reflex is elicited most readily in Gastrocnemius recordings are made with
the soleus muscle by stimulation of the tib- the patient supine, with the recording elec-
ial nerve. Although electric stimulation is trodes placed posteriorly between the two
used most commonly, an Achilles tendon tap heads of the gastrocnemius muscle. Stimu-
or quick Achilles tendon stretch may also be lation is applied in the same fashion as for
used to elicit the reflex.8'15 soleus recordings. Latencies are similar for
soleus and gastrocnemius recordings.
With gradually increasing stimulus inten-
Soleus and Gastrocnemius sity, the H reflex usually appears before the
Techniques M response or shortly thereafter. As the stim-
ulus voltage is increased, the amplitude of
The H reflex can be recorded readily with the H reflex reaches a maximum before a
two methods: recording over the soleus maximal M response is obtained. With fur-
muscle or over the gastrocnemius muscle. ther increase in voltage, the amplitude de-
Because the soleus muscle is the primary clines and eventually disappears as the M
wave reaches maximal amplitude at supra- 100% of the maximal amplitude of the M
maximal stimulus intensity4*9'16 (Fig. 28-2A). wave. If recorded over the gastrocnemius
As recorded from the soleus muscle, the H muscle, an initial positivity may be present.
reflex appears as a biphasic wave with a large Stimulus intensity should be increased and
negative deflection, which is normally 50%- decreased as needed until a reproducible
maximal H reflex is obtained.8'1'This tech-
nique is important in differentiating the H
reflex from the F wave (Table 28—2). Latency
is measured to the initial deflection from
baseline. In adults, the H reflex usually oc-
curs at 28-35 ms and varies with leg length
and age. Braddom and Johnson18 developed
a nomogram and formula that allow for age
and leg-length differences. However, the
side-to-side comparison of the latency of the
H reflex is probably more widely used than
the absolute value and should vary by no
more than 3 ,^8,17,18,20,22 (Table 28_3) The
reliability of amplitude measurements is not
sufficient for clinical use.
Flexor Carpi Radialis Technique

The H reflex also can be recorded from the
flexor carpi radialis muscle. According to
the method described by Jabre,21 the active
recording electrode is placed over the mid
portion of the flexor carpi radialis, approx-
imately 10 cm distal to the elbow crease, with
the reference electrode over the brachiora-
dialis. A flexor carpi radialis study usually is
performed with the patient seated rather
than supine to increase the likelihood of
recording a response.23 Normal values are
listed in Table 28-3.
PEDIATRIC H REFLEXES
Cai and Zhang24 demonstrated that the con-
duction velocities of the H reflex do not
reach adult values until age 6 years. Normal
absolute latencies of the H reflex are corre-
lated with body length until 1 year of age.25
In infants, H reflexes are readily recordable
from calf and hand muscles. Mitsudome et
al.26 showed that although the evolution of
motor velocities is the same for both pre-
figure 28—2. A, H reflex recorded from soleus muscle. mature and full-term infants, the evolution
S, H reflex recorded from soleus muscle with decreas- of the velocity of the H reflex may be slower
ing stimulus intensity (from top to bottom). Note max-
imal amplitude of the H reflex with submaximal stim- in premature than in full-term infants. This
ulation and initial negative waveform, allowing reliable should be considered when evaluating in-
latency measurement fants.
H Reflexes 379
Table 28-3. Normal Values for H Reflex Latency

LATENCY, ms
Side-to-Side
Reference Muscle Recorded Mean Range Comparison
Braddom and Johnson18 Gastrocnemius — 28-35 < 1.2
Kimura8 Soleus 29.5 27-35 < 1.4
Mayo Clinic Soleus — 25-35 <3
Shahani and Young19 Soleus — <34 <2
Ongerboer de Visser, Flexor carpi radialis 16.8 15-20 <0.85
Schimsheimer, and Hart20
Jabre21 Flexor carpi radialis 15.9 11-21 < 0.7
Buschbacher5 Gastrocnemius 30.3 30-38 <2
CLINICAL APPLICATION after successful renal transplantation.27 Pro-

longed latency also has been demonstrated
Proximal Conduction in Guillain-Barre syndrome, chronic inflam-
matory polyradiculoneuropathies, and hered-
The H reflex is used best as a measure of itary mixed motor and sensory neuropathy
proximal conduction. It commonly is used type I.28 Because the H reflex may be absent
in assessing SI radiculopathy. Braddom and in persons older than 60 years, a large num-
Johnson18 proposed that a side-to-side ber of whom also have lumbar stenosis, the
latency difference in the H reflex greater usefulness of the H reflex in studying neu-
than 1.2 ms indicates a discrete lesion of the ropathy in this age group is limited.12'29 The
SI nerve root, and it may be the initial find- latency of the H reflex in the flexor carpi ra-
ing in acute radiculopathy or in cases of dialis muscle has proved useful in assessing
radiculopathy in which the only needle slowing of conduction through the proximal
electromyographic findings are fibrillation median nerve fibers in radiation-induced
potentials in paraspinal muscles.17 The pro- plexopathy.20'28 This may be used to assess
longation or absence of the H reflex corre- C7 radiculopathy, comparable to assessing
lates with clinically reduced or absent ankle SI radiculopathy with recordings from the
jerks. Any lesion along the pathway of the H soleus muscle (Fig. 28-3).
reflex may prolong its latency. Moreover, if
the largest diameter axons are not affected
by a root lesion, the H reflex may not remain Central Nervous System
normal. Therefore, additional corroborat- Excitability
ing electrophysiologic evidence is needed
to support the diagnosis of SI radicu- If the H reflex can be recorded readily in
lopathy.12'23 adults from muscles other than the soleus or
The latency of the H reflex also may be flexor carpi radialis, the possibility of a cen-
prolonged in cases of peripheral neuropa- tral nervous system disorder must be con-
thy, because of slowing of peripheral or sidered. The H reflex may be used in two
proximal conduction. The H reflex may dis- ways to study the excitability of the motor
appear before the F wave in cases of pe- neuron pool.6'7'10'30 The first way compares
ripheral neuropathy.3 Prolonged latency of the ratio of maximal H reflex amplitude with
the H reflex has been demonstrated in dia- maximal M response amplitude (Hmax/
betic, alcoholic, nutritional, paraneoplastic, Mmax). This ratio may be 1 in normal sub-
cisplatin, and vasculitic neuropathies. In ure- jects and increased in disorders causing spas-
mic neuropathy, this prolongation decreases ticity, although it correlates poorly with the
ing peripheral nerve disorders and central

motor neuron excitability. In clinical prac-
tice, soleus recordings provide the optimum
response, with reliable latency measure-
ments. The H reflex is identified by its elic-
itation with submaximal stimulation and
suppression with supramaximal stimula-
tion—features that distinguish it from the
F wave.
REFERENCES
1. Hoffmann P. Ueber die Beziehungen der Schnen-
reflexe zur willkurlichen Bewegung und zum
Tonus. Z Biol 68:351-370, 1918.
2. MagladeryJW, McDougal DB Jr. Electrophysiologi-
cal studies of nerve and reflex activity in normal
man. I. Identification of certain reflexes in the elec-
tromyogram and the conduction velocity of pe-
ripheral nerve fibres. Bull Johns Hopkins Hosp
86:265-290, 1950.
3. Lachman T, Shahani BT, Young RR. Late responses
as aids to diagnosis in peripheral neuropathy. J Neu-
rol Neurosurg Psychiatry 43:156-162, 1980.
Figure 28-3. H reflex recorded from forearm flexors 4. Panizza M, Nilsson J, Hallett M. Optimal stimulus
duration for the H reflex. Muscle Nerve 12:576-
579, 1989.
degree of spasticity noted clinically. The 5. Buschbacher RM. Normal range for H-reflex
Hmax/Mmax ratio is normal in patients with recording from the calf muscles. Am J Phys Med
rigidity.30 Rehabil 78 (Suppl 6):S75-S79, 1999.
The second way is to produce H reflex re- 6. Fisher MA. AAEM Minimonograph #13: H reflexes
and F waves: physiology and clinical indications.
covery curves, first described by Magladery Muscle Nerve 15:1223-1233, 1992.
and McDougal.2 The H reflex is elicited be- 7. Funase K, Miles TS. Observations on the variability
fore, during, and after conditioning stimuli, of the H reflex in human soleus. Muscle Nerve
usually electric stimuli, of a range of inter- 22:341-346, 1999.
8. KimuraJ. Electrodiagnosis in Diseases of Nerve and
stimulus intervals. It is important to keep the Muscle: Principles and Practice, 2nd ed. FA Davis
latency and amplitude of the M response Company, Philadelphia, 1989, pp 359-361.
constant during these trials. These curves 9. Nishida T, Kompoliti A, Janssen I, Levin KF. H re-
also have been used to study spasticity. How- flex in S-l radiculopathy: latency versus amplitude
ever, the changes in the curves are rarely controversy revisited. Muscle Nerve 19:915-917,
1996.
striking. The curve is influenced greatly by 10. lies JF, Pardoe J. Changes in transmission in the
the position and comfort of the patient, the pathway of heteronymous spinal recurrent inhibi-
angle of lower extremity joints, relaxation, tion from soleus to quadriceps motor neurons dur-
and the positioning of the head. Construc- ing movement in man. Brain 122:1757-1764, 1999.
11. Rossi-Durand C, Jones KE, Adams S, Bawa P. Com-
tion of these curves is time-consuming, and parison of the depression of H-reflexes following
reproducibility is poor. The pathophysio- previous activation in upper and lower limb mus-
logic basis of these curves is poorly under- cles in human subjects. Exp Brain Res 126:117-127,
stood, and they are impractical.10'15 1999.
12. Wilbourne AJ, Aminoff MJ. Radiculopathies. In
Brown WF, Bolton CF (eds). Clinical Electromyog-
raphy, 2nd ed. Butterworth-Heinemann, Boston,
SUMMARY 1993, pp 177-209.
13. Liveson JA, Ma DM. Laboratory Reference for Clin-
The H reflex, a monosynaptic reflex usually ical Neurophysiology, FA Davis, Philadelphia, 1992,
pp 238-246.
evoked with tibial nerve stimulation, is sim- 14. Maryniak O, Yaworski R. H-reflex: optimum loca-
ple to elicit and can be a useful adjunct to tion of recording electrodes. Arch Phys Med Re-
routine nerve conduction studies in assess- habil 68:798-802, 1987.
H Reflexes 381
15. Rico RE, Jonkman EJ. Measurement of the Achilles 24. Cai F, Zhang J. Study of nerve conduction and late
tendon reflex for the diagnosis of lumbosacral root responses in normal Chinese infants, children, and
compression syndromes. J Neurol Neurosurg Psy- adults. J Child Neurol 12:13-18, 1997.
chiatry 45:791-795, 1982. 25. Jones HR Jr, Harmon RL, Harper CM Jr, Bolton
16. Nishida T, Levy CE, Lewit EJ, Janssen I. Compari- CF. An approach to pediatric electromyography. In
son of three methods for recording tibial H reflex: Jones HRJr, Bolton CF, Harper CMJr (eds). Pedi-
a clinical note. Am J Phys Med Rehabil 78:474-476, atric Clinical Electromyography. Lippincott-Raven
1999. Publishers, Philadelphia, 1996, pp 1-36.
17. DeLisa JA, Lee HJ, Baran EM, Lai K-S, Spielholz N. 26. Mitsudome A, Yasumoto S, Ogata H. Late responses
Manual of Nerve Conduction Velocity and Clinical in full-term newborn infants. In KimuraJ, Shibasaki
Neurophysiology, 3rd ed. Raven Press, New York, H (eds). Recent Advances in Clinical Neurophysi-
1994, pp 168-175. ology: Proceedings of the Xth International Con-
18. Braddom RI, Johnson EW. Standardization of H re- gress of EMG and Clinical Neurophysiology, Kyoto,
flex and diagnostic use in SI radiculopathy. Arch Japan, October 15-19, 1995. Elsevier, New York,
Phys Med Rehabil 55:161-166, 1974. 1996, pp 761-765.
19. Shahani BT, Young RR. Studies of reflex activity 27. Bolton CF. Metabolic neuropathy. In Brown WF,
from a clinical viewpoint. In Aminoff MJ (ed). Elec- Bolton CF (eds). Clinical Electromyography, 2nd
trodiagnosis in Clinical Neurology. Churchill Liv- ed. Butterworth-Heinemann, Boston, 1993, pp
ingstone, New York, 1980, pp 290-304. 561-598.
20. Ongerboer de Visser BW, Schimsheinler RJ, Hart 28. Schimsheimer RJ, Ongerboer de Visser BW, Kemp
AA. The H-reflex of the flexor carpi radialis mus- B, Bour LJ. The flexor carpi radialis H-reflex in
cle; a study in controls and radiation-induced bra- polyneuropathy: relations to conduction velocities
chial plexus lesions. J Neurol Neurosurg Psychiatry of the median nerve and the soleus H-reflex latency.
47:1098-1101, 1984. J Neurol Neurosurg Psychiatry 50:447-452, 1987.
21. Jabre JF. Surface recording of the H-reflex of the 29. Wilbourn AJ. Diabetic Neuropathies. In Brown WF,
flexor carpi radialis. Muscle Nerve 4:435-438,1981. Bolton CF (eds). Clinical Electromyography, 2nd
22. Falco FJ, Hennessey WJ, Goldberg G, Braddom RL. ed. Butterwordi-Heinemann, Boston, 1993, pp
H reflex latency in the healthy elderly. Muscle 477-515.
Nerve 17:161-167, 1994. 30. Leonard CT, Moritani T. H-reflex testing to deter-
23. Sabbahi MA, Khalil M. Segmental H-reflex studies mine the neural basis of movement disorders of
in upper and lower limbs of healthy subjects. Arch neurologically impaired individuals. Electromyogr
Phys Med Rehabil 71:216-222, 1990. Clin Neurophysiol 32:341-349, 1992.
Chapter 29
CRANIAL REFLEXES
Raymond G. Auger
J. Clarke Stevens
BLINK REFLEX MASSETER INHIBITORY REFLEX

Methods Methods
Neuroanatomy Indications
Indications for Use SUMMARY
JAW JERK (MASSETER REFLEX)
Neuroanatomy
Methods
Indications
The integrity of the trigeminal and facial than proprioceptive receptors, as previously
nerves and their central connections can be thought.
evaluated electrophysiologically by studying
the reflex activity mediated by these nerves.
Thus, they are of greatest value in assessing Methods
cranial neuropathies, but they also can pro-
vide useful information in some cases of The blink reflex is elicited by mechanical or
polyradiculoneuropathy, peripheral neurop- electric stimulation over the face with a
athy, and brain stem lesion. The reflexes dis- graded threshold, with the lowest threshold
cussed in this chapter are the electrically being around the eye. The reflex usually
evoked blink reflex, the jawjerk (or masseter is elicited by stimulating the supraorbital
reflex), and the masseter inhibitory reflex. branch of the trigeminal nerve at the supra-
orbital notch, while recording simultane-
ously from both the left and right orbicularis
BLINK REFLEX oculi muscles (Fig. 29—1). The stimuli are
applied irregularly at least 5 seconds apart
Overend1 observed that a blink response oc- so that habituation is minimized. The stim-
curs after a light tap on the forehead. Kugel- ulating current required to evoke the reflex
berg2 elicited the reflex with an electric is small and not painful. In normal subjects,
stimulus and demonstrated two distinct re- an early (Rl) response is obtained ipsilateral
sponses: an early well-synchronized response to the stimulus, and the late (R2) responses
occurring ipsilateral to the stimulus (Rl) are obtained bilaterally (Fig. 29-2). Later bi-
and poorly synchronized bilateral responses lateral (R3) responses are also seen in nor-
with a longer latency (R2). Rushworth3 dem- mal subjects. With slowly increasing stimula-
onstrated that the afferent limb of the reflex tion, the R2 responses usually are elicited
is the first division of the trigeminal nerve before Rl. If an early response cannot be ob-
and the efferent component is the facial tained, a paired stimulus can be used, with
nerve. Shahani4 showed that the reflex was an interstimulus interval of 5 ms to take ad-
mediated by cutaneous receptors rather vantage of a period of facilitation that may
382
Cranial Reflexes 383
last 1-9 ms after a conditioning stimulus

(Fig. 29-3). Confusing results may be ob-
tained if the stimulating electrode is too
close to the midline, because a contralateral
Rl response may be obtained by stimulation
of the opposite trigeminal nerve. Ampli-
tudes of die responses usually are not mea-
sured because differences in amplitude of
up to 40% occur in normal subjects.5
Normal values, in milliseconds, are as

follows:
Ipsilateral Rl < 13
Ipsilateral R2 < 41
Figure 29-1. Electrode placement for blink reflex stud-
Contralateral R2 < 44
ies. The supraorbital nerve is stimulated with the cath-
ode at the supraorbital notch. Responses are recorded Ipsilateral vs. contralateral differences, in
ipsilaterally and contralaterally. Gl, active electrode; milliseconds:
G2, reference electrode. (Modified from Auger RG.
Brain stem disorders and cranial neuropathies. In Rl < 1.2
Brown WF, Bolton CF [eds]. Clinical Electromyogra-
phy. Butterworths, Boston, 1987, pp 417-429. By per- R2 <8
Figure 29-2. Responses obtained with stimulation of the supraorbital nerve and simultaneously recording from both
left and right Orbicularis oculi muscles. Rl, first response; R2, second response.
is relayed centrally through an oligosynaptic

pathway involving the principal sensory nu-
cleus of the trigeminal nerve in the pons.8
The afferent fibers relaying the R2 response
descend into the medulla and synapse in the
spinal nucleus of the trigeminal nerve.
Through polysynaptic pathways that pass ip-
silaterally and contralaterally, the afferent
limb is connected with the nucleus of the fa-
cial nerve, the effect limb of the reflex.
Indications for Use

Figure 29-3. In this patient, the Rl response could not TRIGEMINAL NERVE LESIONS
be elicited with a single stimulus. A normal response
was obtained with paired stimuli, with an interstimulus In lesions of the trigeminal nerve, blink re-
interval of 5 ms. flex responses are usually delayed or absent
with stimulation of the involved side, but
they are normal with stimulation of the un-
The Rl latency in young children reaches affected side (Fig. 29-5). In severe lesions of
adult values by the age of 2 years. However, the trigeminal nerve, stimulation of the in-
the R2 responses are sometimes absent in volved side may not elicit a response. In pa-
children younger than 2 years and attain tients with trigeminal sensory neuropathy
adult values at age 5-6 years. Because of the
variability of R2 responses, they are less use-
ful in children younger than 6 years.6
The R2 response correlates with contrac-
tion of the orbicularis oculi muscle and has
the same latency as the corneal reflex. The
physiologic significance of the Rl response
is unknown.
In patients with suspected lesions of the
maxillary or mandibular division of the
trigeminal nerve, the blink reflex obtained
with supraorbital nerve stimulation may be
normal. In this situation, the infraorbital
branch can be stimulated to assess the func-
tion of the maxillary division. Stimulation of
the infraorbital nerve consistently elicits R2
responses in normal subjects, but the Rl re-
sponse is frequendy absent. Stimulation of
the mental nerve (the mandibular division) Figure 29-4. Neural pathways associated with the blink
also elicits R2 responses, but the Rl response reflex. For the Rl response, afferent fibers from the
is never observed.7 trigeminal nerve synapse with the principal sensory nu-
cleus in the pons (Vp). Connections are then made with
the facial nucleus (VII) via one or two interneurons.
For the R2 response, trigeminal nerve fibers synapse
Neuroanatomy with the spinal nucleus of the trigeminal nerve via a
multisynaptic pathway. Connections are then relayed to
The afferent limb of the reflex arc is the the facial nuclei bilaterally. (Modified from Ongerboer
ophthalmic division of the trigeminal nerve, de Visser VW, Kuypers HGJM. Late blink reflex changes
in lateral medullary lesions: an electrophysiological and
with the afferent nerve cell body in the neuro-anatomical study of Wallenberg's syndrome.
gasserian ganglion (Fig. 29-4). The efferent Brain 101:285-294, 1978. By permission of Oxford Uni-
limb is the facial nerve. The early response versity Press.)
In Bell's palsy, routine conduction studies

of the facial nerve evaluate the segment of
the nerve distal to the stylomastoid foramen,
which is clinically helpful only when degen-
eration begins distally. The blink reflex,
however, assesses the entire nerve, including
the intraosseous portion. All the patients
with Bell's palsy studied by Kimura, Giron,
and Young12 had a delayed or absent Rl re-
sponse on the affected side of the face. On
the weak side, the R2 response was also ab-
normal in all patients. Of 127 patients tested
serially, 100 had return of the previously ab-
sent Rl and R2 responses, whereas responses
with stimulation of the facial nerve at the
mastoid process remained relatively normal.
The Rl latency was increased by more than
2 ms initially, suggesting demyelination of fa-
cial nerve fibers. The latency of Rl decreased
during the second month and returned to
normal by the fourth month. These patients
generally had good return of function within
a few months. The other 27 patients had
Figure 29-5. Blink reflex studies in a patient with a le-
sion of the left trigeminal nerve. Rl max., upper limit smaller amplitudes with direct stimulation of
of normal for the onset of the Rl response; R2 max., the facial nerve, and the blink reflex re-
upper limit of normal for the onset of die R2 response. sponses did not return. These patients had
degeneration of the facial nerve and poorer
recovery. More recent studies have reached
associated with connective tissue diseases, similar conclusions—when the blink reflex
there may be abnormalities of Rl only or bi- is normal or Rl is only delayed, the prog-
lateral abnormalities of Rl and R2; it is un-
common for only R2 to be affected.9 The
blink reflex is also helpful in identifying le-
sions of the sensory root and gasserian gan-
glion, including tumors, vascular malforma-
tions, and postherpetic lesions. The blink
reflex is usually normal in idiopathic trigem-
inal neuralgia and atypical facial pain.10
Trichloroethylene, long known to produce
trigeminal neuropathy, is associated with a
prolonged mean latency of Rl in persons
with occupational exposure or a contami-
nated water supply.11
FACIAL NEPVE LESIONS

In lesions of the facial nerve, the Rl and ip-
silateral R2 responses are delayed or absent
with stimulation of the involved side, but
the contralateral R2 response is normal if
trigeminal nerve function is preserved (Fig.
Figure 29-6. Blink reflex studies in a patient with a le-
29-6). In severe lesions of the facial nerve, sion of die right facial nerve. Rl max., upper limit of
the Rl and R2 responses may be absent on normal for die onset of die Rl response; R2 max., up-
the involved side. per limit of normal for die onset of die R2 response.
nosis is excellent. Absence of the blink re-

flex has been associated with a poor prog-
nosis in 56% of cases.13'14
Delay or absence of only the Rl response
indicates dysfunction of either the trigemi-
nal or facial nerve or the corresponding cen-
tral connections in the brain stem. Clinical
findings, for example, numbness of the face,
may indicate which nerve is responsible. If
both nerves are affected, as they may be in
acoustic neuroma, it may be impossible to
provide a more precise interpretation. If the
R2 component is also affected, the examiner
usually can clarify which of the nerves is in-
volved. Abnormality of the R2 responses with
a normal Rl response suggests a lesion of Figure 29-7. Assessment of facial synkinesis in a patient
the spinal tract and nucleus of the trigemi- with left hemifacial spasm. A synkinetic response is pres-
ent in the left orbicularis oris with stimulation of the
nal nerve. left supraorbital nerve. On the right (normal) side,
synkinesis is not present. (From Auger.15 By permission
of Lippincott, Williams & Wilkins.)
ASSESSMENT OF FACIAL
SYNKINESIS
The location of the recording electrodes can cleus may be hyperactive, with spasms re-
be modified to allow for objective assessment sulting from "backfiring" that is similar to an
of facial synkinesis. This can be accom- exaggerated F response.17'18 These patients
plished by recording simultaneously over the also have synkinesis, which appears similar
orbicularis oculi and another muscle on the to that seen in aberrant regeneration, but,
same side of the face which is also supplied unlike it, the synkinetic response is variable
by the facial nerve, such as the mentalis or in early cases and may not be present be-
orbicularis oris. In normal subjects, stimula- tween spasms.15
tion of the supraorbital nerve produces re- Abnormal communication between axons
flex activation of the orbicularis oculi only. of the facial nerve, termed lateral spread,
In hemifacial spasm and aberrant regenera- can be demonstrated by stimulating the
tion of the facial nerve after injury, the sec- mandibular branch of the facial nerve while
ond muscle often has a synkinetic response15 recording from the orbicularis oculi or by
(Fig. 29-7). This synkinetic response is not stimulating the zygomatic branch while
present in other movement disorders in- recording from the mentalis muscle (Figs.
volving the face, for example, facial myoky- 29-8 and 29-9). Surgical microvascular de-
mia, blepharospasm, or Meige's syndrome. compression of the facial nerve generally re-
Hemifacial spasm is characterized by uni- sults in loss of lateral spread and synkinesis
lateral, involuntary twitching of periorbital and relief of hemifacial spasm.19'20
muscles and muscles of the lower portion of Rubin et al.21 described a patient in whom
the face. In most patients, the disorder prob- an unusual movement disorder affecting the
ably is associated with compression of the muscles of mastication developed after sur-
nerve by a vascular loop near the brain stem. gical removal of a tumor involving the
Rarely, a tumor, aneurysm, or vascular mal- trigeminal nerve. They used the blink reflex
formation may compress the nerve. Focal to demonstrate a synkinetic response in the
demyelination ensues, producing ectopic masseter muscle with stimulation of the
generation of action potentials or ephaptic supraorbital nerve, thereby confirming that
transmission to other axons near the site of the masseter in their patient was innervated
compression.16 Recent evidence suggests by aberrandy regenerated branches of the
that in hemifacial spasm the facial motor nu- facial nerve.
ronopathy, but frequently abnormal in those

with Sjogren's syndrome/sicca complex or
idiopathic sensory neuronopathy.24
POSTERIOR FOSSA LESIONS

In patients with cerebellopontine angle le-
sions, blink reflex responses may be delayed.
Previously, the blink reflex was used as a di-
agnostic test to screen for suspected tumors
in the cerebellopontine angle. However,
with the advent of sophisticated imaging
techniques, the blink reflex has no place in
the diagnosis of these lesions.
MULTIPLE SCLEROSIS
Before the advent of modern averaging and
imaging techniques, the blink reflex was
used in some laboratories to search for evi-
Figure 29-8. Electrode placement to assess the lateral
spread response in a patient with hemifacial spasm. Gl,
active electrode; G2, reference electrode; Stim., stimu-
late. (From Harper CM Jr. AAEM case report 21: Hemi-
facial spasm: preoperative diagnosis and intra-operative
management. Muscle Nerve 14:213-218, 1991. By per-
mission of Mayo Foundation.)
PERIPHERAL NEUROPATHY
Delayed responses to supraorbital nerve
stimulation are found in hereditary motor
and sensory neuropathy type I (Charcot-
Marie-Tooth disease).22 The delay is most
marked in the distal segment of the facial
nerve even though no facial weakness may
be demonstrated. In contrast, the delayed la-
tencies in Guillain-Barre syndrome are asso-
ciated with facial muscle weakness. The blink
reflexes also may be abnormal in other de-
myelinating neuropathies, including chronic
inflammatory demyelinating polyradiculo-
neuropathy (Fig. 29-10). In most axonal
sensorimotor peripheral neuropathies, in-
cluding diabetic neuropathy,23 the blink re-
flex is unaffected.
The blink reflex may be helpful in the Figure 29-9. Lateral spread response in a patient with
evaluation of patients who present clinically hemifacial spasm. With stimulation of the mandibular
with subacute sensory neuronopathy. In a branch of the facial nerve (top), a delayed response is
study of patients who were evaluated in our recorded from the ipsilateral Orbicularis oculi. With
stimulation of the zygomatic branch of the facial nerve
laboratory, the blink reflex was normal in (bottom), a delayed response is recorded from the ipsi-
patients with paraneoplastic sensory neu- lateral mentalis.
Figure 29-10. Blink reflex studies in a pa-

tient with chronic inflammatory demyeli-
nating polyradiculoneuropathy. The Rl
and R2 responses are significantly de-
layed. Rl max., upper limit of normal for
the onset of the Rl response; R2 max., up-
per limit of normal for the onset of the
R2 response.
dence of clinically silent lesions that would had symptoms suggesting medullary in-
provide further support for the diagnosis of volvement. As might be expected in patients
multiple sclerosis (Fig. 29-11). Currently, it with multiple sclerosis, the incidence of de-
is rarely used for this purpose, but it may be layed Rl responses increases with time, al-
helpful in documenting objective abnor- though there maybe improvement when the
malities in patients with vague facial pares- disease is in remission.
thesias of uncertain significance.
In 260 patients with multiple sclerosis ob- EXTRAPYRAMIDAL DISEASE
served over a 7-year period, Kimura25 found
that the Rl response was delayed on one or In normal subjects, habituation of the R2
both sides in 66% of those with definite mul- component of the reflex occurs with regu-
tiple sclerosis, in 56% of those with proba- larly applied electric stimuli, resulting in de-
ble multiple sclerosis, and in 29% of those layed or absent R2 responses. This is the ba-
with possible multiple sclerosis. Rl was ab- sis for the glabellar tap sign elicited during
normal in 78% of the patients with neuro- neurologic examination. In patients with
logic signs that suggested pontine involve- Parkinson's disease or other disease affect-
ment, in 57% of those with signs of disease ing the extrapyramidal system, habituation
of the medulla or midbrain, and in 40% of may be impaired.
those who had no brain stem signs. R2 was
less diagnostic than Rl in detecting brain CAVEATS AND PITFALLS
stem lesions and was most often abnormal
in those with pontine signs. When Rl was Although blink reflex pathways are confined
normal and R2 was delayed, the patients to the brain stem, lesions rostral to the brain
Figure 29-11. Blink reflex studies in a patient

with multiple sclerosis and trigeminal neural-
gia. The Rl and R2 responses are markedly de-
layed. Rl max., upper limit of normal for the
onset of the Rl response; R2 max., upper limit
of normal for the onset of the R2 response.
stem may affect latencies and thus diminish veyed via the motor root of the trigeminal
the localizing value of the reflex. In almost nerve to the mesencephalic nucleus in the
half of the patients with cerebral infarction midbrain. Axons from this nucleus synapse
associated with hemiparesis, the Rl response in the motor nucleus of the trigeminal nerve
may be delayed for up to 1 week, and both to activate the efferent limb of the reflex arc.
direct and consensual R2 responses may be This reflex is unique among stretch reflexes
absent or diminished for several weeks.26 in that the cell bodies of the afferent limb
Therefore, in the acute phase, whether fa- (that is, the mesencephalic nucleus) lie intra-
cial paresis is caused by a central or a pe- axially in the brain stem rather than in the
ripheral process cannot be determined gasserian ganglion, which is the brain stem
solely on the basis of prolonged Rl latency. counterpart of a dorsal root ganglion. The
These effects may be the result of removal afferent nerve cell bodies subserving all
of crossed cortical facilitation in the brain other stretch reflexes reside extra-axially in
stem.27 dorsal root ganglia.
The R2 responses are attenuated with
sleep and the use of sedative drugs.
Methods
Recording electrodes are taped over the
JAW JERK (MASSETER REFLEX) belly of the masseter muscle bilaterally, and
the reference electrodes are placed over the
Neuroanatomy zygoma (Fig. 29-12). The reflex hammer
contains a microswitch that triggers a sweep
The jaw jerk, or masseter reflex, is a monosyn- across the monitor upon contact with the ex-
aptic muscle stretch reflex elicited by a tap aminer's finger, which is held on the pa-
on the jaw. Afferent impulses from muscle tient's chin. The latency is measured to the
spindles in the masseter muscle are con- initial reproducible deflection from baseline
Figure 29-12. Electrode placement for study of the jaw

jerk (masseter reflex). (Modified from Auger AG. Brain
stem disorders and cranial neuropathies. In Brown WF,
Bolton CF [eds]. Clinical Electromyography. Butter-
worths, Boston, 1987, pp 417-429. By permission of the
publisher.)
(Fig. 29-13). The normal range of latencies Figure 29-14. Superimposed responses recorded from
is 6-10.5 ms. The maximal side-to-side dif- the masseter muscles following four successive taps on
ference in latency is 1.5 ms. Wide variation the chin with a reflex hammer in a patient with a left
in amplitude among normal subjects pre- acoustic neuroma. The response is normal on the right
and absent on the left.
cludes the use of amplitude measurements
in clinical studies. In normal subjects, par-
ticularly elderly and obese ones, the reflex
is sometimes difficult to record using surface division of the trigeminal nerve. If a patient's
electrodes. symptoms are in the distribution of this di-
vision, the blink reflex may be normal. In
this situation, the jaw jerk may provide ob-
Indications jective evidence of involvement of the
mandibular division (Fig. 29-14). The most
The main indication for using the jaw jerk common abnormality is the absence of the
is to assess the function of the mandibular jaw jerk rather than prolongation of its la-
tency.
In patients with inflammatory polygan-
glionopathies presenting with pure sensory
neuronopathies, the jaw jerk may be normal
even though the patient is otherwise are-
flexic, has no sensory responses, and may not
have a blink reflex.28 This likely occurs be-
cause the afferent nerve cell body involved
with the jaw jerk is in the mesencephalic nu-
cleus of the trigeminal nerve, which is in the
brain stem and protected by the blood-brain
barrier. Although neurons in the dorsal root
ganglia and gasserian ganglion are pro-
tected by the blood-nerve barrier, it is not as
effective as the blood-brain barrier.
Ongerboer De Visser and Goor29 studied
jaw jerk and masseter electromyograms in
patients with vascular or neoplastic disease
of the midbrain and pons. Mesencephalic
Figure 29-13. Superimposed responses recorded from lesions were associated with abnormal jaw
the masseter muscles following four successive taps on
the chin with a reflex hammer in a normal subject. The
reflexes and normal masseter electromyo-
reflex hammer contains a microswitch that, upon con- grams. In the group with pontine lesions,
tact, initiates a sweep across the monitor. both the masseter electromyograms and the
jaw jerk were often abnormal. An abnormal jawjerk (Fig. 29-12). A mechanical tap usu-
jaw jerk suggests midbrain disease, whereas ally produces only one silent period (SPl),
an abnormal Rl response, with or without which begins between 11 and 15 ms after the
an associated change in the jawjerk, suggests tap and lasts for 14-30 ms (Fig. 29-15). With
a rostral pontine lesion.30'31 The latency of an electric stimulus, two silent periods oc-
the jaw jerk is not influenced by supraten- cur: the first one (SPl) corresponds to the
torial or primary cerebellar disease.3^ silent period after a mechanical tap and the
second one (SP2) begins 30-60 ms after the
stimulus (Fig. 29-15).
MASSETER INHIBITORY
REFLEX
Indications
Methods
The MIR sometimes is useful in the evalua-
The masseter inhibitory reflex (MIR) is impor- tion of peripheral neuropathies.33'34 In se-
tant in the reflex control of biting and chew- vere demyelinating neuropathies in which
ing. It evolved as a mechanism to protect no response occurs to stimuli in the limbs,
intraoral structures against the powerful jaw- MIR can be used to assess conduction delay,
closing muscles. It is elicited by applying a because it can still be measured in neu-
mechanical or electric stimulus to the skin ropathies that are severe enough to abolish
and mucous membrane supplied by the even the blink reflex (Fig. 29-16). In these
maxillary or mandibular division of the
trigeminal nerve during voluntary contrac-
tion of the masseter muscle. The recording
electrodes are placed on the masseter mus-
cle bilaterally, in the same manner as in the
Figure 29-15. Superimposed responses recorded from Figure 29-16. Masseter inhibitory reflex following a tap
the masseter muscles following four successive taps on on the chin in a patient with chronic inflammatory de-
the chin (top) and four successive electric stimuli to the myelinating polyradiculoneuropathy (CIDP) compared
mental nerve (bottom) while the subject clenched this with that in normal subject. The onset of the first silent
teeth. SPl, first silent period; SP2, second silent period. period (SP1) is delayed in the patient with CIDP.
situations, the latency of the onset of MIR Although it is not discussed in this chapter,
may be severely prolonged, thereby provid- these studies are usually performed in com-
ing evidence for a demyelinating compo- bination with a needle electrode examina-
nent. The reflex is normal in axonal neu- tion of muscles innervated by the trigeminal
ropathies. and facial cranial nerves.
The MIR may be abolished in some forms
of sensory neuropathy involving predomi-
nantly intraoral sensory nerves, giving rise to REFERENCES
severe impairment in chewing and swallow-
ing.35 1. Overend W. Preliminary note on a new cranial re-
In some laboratories, MIR is used to assess flex. Lancet 1:619, 1896.
2. Kugelberg E. Facial reflexes. Brain 75:385-396,
function of the maxillary and mandibular di- 1952.
visions of the trigeminal nerve by applying 3. Rushworth G. Observations on blink reflexes. J Neu-
an electric stimulus to the infraorbital and rol Neurosurg Psychiatry 25:93-108, 1962.
mental nerves, respectively, during voluntary 4. Shahani B. The human blink reflex. J Neurol Neu-
rosurg Psychiatry 33:792-800, 1970.
contraction of the masseter. Abnormalities 5. Stoehr M, Petruch F. The orbicularis oculi reflex:
of conduction can be detected by compar- diagnostic significance of the reflex amplitude.
ing the latency of the onset of SP1 on the Electromyogr Clin Neurophysiol 18:217-224, 1978.
two sides.36 6. Anday EK, Cohen ME, Hoffman HS. The blink re-
The MIR also has been used to assess cen- flex: maturation and modification in the neonate.
Dev Med Child Neurol 32:142-150, 1990.
tral inhibition in some disease states. In tris- 7. Jaaskelainen SK. Blink reflex with stimulation of the
mus associated with tetanus, MIR may be mental nerve. Methodology, reference values, and
abolished. In the rare condition of hemi- some clinical vignettes. Acta Neurol Scand 91:477-
masticatory spasm, MIR is attenuated on the 482, 1995.
8. Trontelj MA, Trontelj JV. Reflex arc of the first com-
side of the spasm, implying impaired inhi- ponent of the human blink reflex: a single mo-
bition during the period of spasm.37'38 toneurone study. J Neurol Neurosurg Psychiatry
41:538-547, 1978.
9. Hagen NA, Stevens JC, Michet CJ Jr. Trigeminal
sensory neuropathy associated with connective tis-
SUMMARY sue diseases. Neurology 40:891-896, 1990.
10. Kimura J, Rodnitzky RL, Van Allen MW. Electrodi-
Although a certain level of expertise is nec- agnostic study of trigeminal nerve. Orbicularis oculi
essary, electrophysiologic study of cranial re- reflex and masseter reflex in trigeminal neuralgia,
flexes is not technically demanding, time- paratrigeminal syndrome, and other lesions of the
trigeminal nerve. Neurology 20:574-583, 1970.
consuming, or associated with substantial 11. Feldman RG, Niles C, Proctor SP, Jabre J. Blink re-
patient discomfort. The information ob- flex measurement of effects of trichloroethylene
tained may document objective abnormality exposure on the trigeminal nerve. Muscle Nerve
and assist with localization. Blink reflexes are 15:490-495, 1992.
12. Kimura J, Giron LT, Young SM. Electrophysiologi-
useful for studying the function of the cal study of Bell palsy: electrically elicited blink re-
trigeminal and facial nerves and their cen- flex in assessment of prognosis. Arch Otolaryngol
tral connections in the brain stem. When 102:140-143, 1976.
nerve conduction studies in the limbs sug- 13. Ghonim MR, Gavilan C. Blink reflex: prognostic
gest a demyelinating peripheral neuropathy, value in acute peripheral facial palsy. ORLJ Otorhi-
nolaryngol Relat Spec 52:75-79, 1990.
the blink reflex can provide information 14. Heath JP, Cull RE, Smith IM, Murray JA. The neu-
about involvement of proximal nerve seg- rophysiological investigation of Bell's palsy and the
ments. Patterns of involvement of the facial predictive value of the blink reflex. Clin Otolaryn-
and trigeminal nerves are often helpful in gol 13:85-92, 1988.
15. Auger RG. Hemifacial spasm: clinical and electro-
suggesting the type of neuropathy under in- physiologic observations. Neurology 29:1261-1272,
vestigation. The jaw jerk is useful in assess- 1979.
ing the mandibular division of the trigemi- 16. Nielsen VK. Pathophysiology of hemofacial spasm:
nal nerve, and it can aid in evaluating I. Ephaptic transmission and ectopic excitation.
patients with suspected sensory ganglion- Neurology 34:418-426, 1984.
17. Moller AR. Interaction between the blink reflex and
opathies. The MIR is sometimes helpful in the abnormal muscle response in patients with
evaluating patients with demyelinating neu- hemifacial spasm: results of intraoperative record-
ropathies and in assessing central inhibition. ings. J Neurol Sci 101:114-123, 1991.
18. Valls-Sole J, Tolosa ES. Blink reflex excitability cycle 29. Ongerboer De Visser BW, Goor C. Jaw reflexes and
in hemifacial spasm. Neurology 39:1061-1066,1989. masseter electromyograms in mesencephalic and
19. Auger RG, Piepgras DG, Laws ERJr, Miller RH. Mi- pontine lesions: an electrodiagnostic study. J Neu-
crovascular decompression of the facial nerve for rol Neurosurg Psychiatry 39:90-92, 1976.
hemifacial spasm: clinical and electrophysiologic 30. Hopf HC, Thomke F, Gutmann L. Midbrain vs.
observations. Neurology 31:346-350, 1981. pontine medial longitudinal fasciculus lesions: the
20. Moller AR, Jannetta PJ. Monitoring facial EMG re- utilization of masseter and blink reflexes. Muscle
sponses during microvascular decompression op- Nerve 14:326-330, 1991.
erations for hemifacial spasm. J Neurosurg 66:681- 31. Thomke F. Isolated cranial nerve palsies due to
685, 1987. brain stem lesions. Muscle Nerve 22:1168-1176,
21. Rubin DI, MatsumotoJY, Suarez GA, Auger RG. Fa- 1999.
cial trigeminal synkinesis associated with a trigemi- 32. Hopf HC, Hinrichs C, Stoeter P, Urban PP, Marx
nal schwannoma. Neurology 53:635-637, 1999. J, Thomke F. Masseter reflex latencies and ampli-
22. Kimura J. An evaluation of the facial and trigemi- tudes are not influenced by supratentorial and cere-
nal nerves in polyneuropathy: electrodiagnostic bellar lesions. Muscle Nerve 23:86-89, 2000.
study in Charcot-Marie-Tooth disease, Guillain- 33. Auger RG. Latency of onset of the masseter in-
Barre syndrome, and diabetic neuropathy. Neurol- hibitory reflex in peripheral neuropathies. Muscle
ogy 21:745-752, 1971. Nerve 19:910-911, 1996.
23. Kirk VH Jr, Litchy WJ, Karnes JL, Dyck PJ. Mea- 34. Cruccu G, Agostino R, Inghilleri M, Innocenti P,
surement of blink reflexes not useful in detection Romaniello A, Manfredi M. Mandibular nerve in-
or characterization of diabetic polyneuropathy (ab- volvement in diabetic polyneuropathy and chronic
stract). Muscle Nerve 14:910-911, 1991. inflammatory demyelinating polyneuropathy. Mus-
24. Auger RG, Windebank AJ, Lucchinetti CF, Chalk cle Nerve 21:1673-1679, 1998.
CH. Role of the blink reflex in the evaluation of 35. Auger RG, McManis PG. Trigeminal sensory neu-
sensory neuronopathy. Neurology 53:407-408, ropathy associated with decreased oral sensation
1999. and impairment of the masseter inhibitory reflex.
25. Kimura J. Electrically elicited blink reflex in diag- Neurology 40:759-763, 1990.
nosis of multiple sclerosis. Review of 260 patients 36. Ongerboer de Visser BW, Cruccu G. Neurophysio-
over a seven-year period. Brain 98:413-426, 1975. logic examination of the trigeminal, facial, hy-
26. Berardelli A, Accornero N, Cruccu G, Fabiano F, poglossal, and spinal accessory nerves in cranial
Guerrisi V, Manfredi M. The orbicularis oculi re- neuropathies and brain stem disorders. In Brown
sponse after hemispheral damage. J Neurol Neu- WF, Bolton CF (eds). Clinical Electromyography,
rosurg Psychiatry 46:837-843, 1983. 2nd ed. Butterworth-Heinemann, Boston, 1993,
27. Kimura J, Wilkinson JT, Damasio H, Adams HRJr, pp 61-92.
Shivapour E, Yamada T. Blink reflex in patients with 37. Auger RG, Litchy WJ, Cascino TL, Ahlskog JE.
hemispheric cerebrovascular accident (CVA). Blink Hemimasticatory spasm: clinical and electrophysio-
reflex in CVA. J Neurol Sci 67:15-28, 1985. logic observations. Neurology 42:2263-2266, 1992.
28. Auger RG. Role of the masseter reflex in the as- 38. Kim HJ, Jeon BS, Lee KW. Hemimasticatory spasm
sessment of subacute sensory neuropathy. Muscle associated with localized scleroderma and facial
Nerve 21:800-801, 1998. hemiatrophy. Arch Neurol 57:576-580, 2000.
Chapter 30
LONG LATENCY REFLEXES AND
THE SILENT PERIOD
Joseph Y. Matsumoto
LONG LATENCY REFLEXES THE SILENT PERIOD

Long Latency Reflexes to Stretch SUMMARY
Long Latency Reflexes to Mixed Nerve
Stimulation
Cutaneous Reflexes
The Flexor Reflex
Long latency reflexes and the silent period are make them important tools in the study of
normal phenomena recorded from muscle normal and abnormal motor control.1
that are highly dependent on central ner- Hammond2 discovered long latency re-
vous system connections. The clinical appli- flexes (LLR) while studying the EMG re-
cation of these phenomena is limited, but sponses evoked in the biceps muscle by
they can be helpful in selected situations. sudden stretch. Electromyographic activity
Some long latency reflexes are abnormal in appeared at 70 ms, later than the monosyn-
myoclonic disorders, and the silent period aptic stretch reflex and earlier than the vol-
shows abnormalities with hyperexcitable mo- untary reaction time of 113 ms. A command
tor neurons or peripheral nerve. to "resist" the stretch augmented these long
latency responses, and the instruction to "let
go" resulted in their virtual disappearance.
LONG LATENCY REFLEXES Marsden et al.3 studied similar stretch re-
flexes in the long flexor of the thumb dur-
There is no exact moment when the reflex ing movement and theorized that long la-
response to a stimulus ends and the volun- tency reflexes served to reinforce volition or
tary reaction begins. Rather, from the onset intent against unexpected perturbations.
of a monosynaptic stretch reflex to the time The reflexes were thought to emanate from
of the first conscious voluntary reaction, the a transcortical loop, with one arm of the loop
cortical influence over spinal and brain stem ascending to the sensorimotor cortex in the
reflex activity gradually increases. Unique dorsal column—medial lemniscus system
electromyographic (EMG) phenomena called and the other arm descending in the corti-
long latency reflexes arise during this transition cospinal tract.
period. Like spinal reflexes, long latency re- Several lines of evidence support the con-
flexes have predictable latencies, but their cept of a transcortical reflex loop. One, long
amplitudes are modulated profoundly by latency reflexes are delayed or absent in pa-
context and volition. These characteristics tients with lesions of the dorsal columns or
394
Long Latency Reflexes and the Silent Period 395
sensorimotor cortex. Cortical potentials pre- ponent, appears at 55-65 ms in the wrist. Oc-
cede long latency reflexes by 30-50 ms, and casionally, a later M3 component may be
the two events correlate in amplitude.4 Two, seen individually or it may merge with the
long latency reflexes occur bilaterally and M2 component.
nearly simultaneously in response to a uni- In Parkinson's disease, the M2 component
lateral stimulus in patients with congenital is enlarged in the wrist flexors. This abnor-
mirror movements.5 Three, patients with mality corresponds to the degree of the pa-
cortical reflex myoclonus have hyperex- tient's rigidity. In contrast, a decrease in or
citable long latency reflexes, which clearly absence of the M2 component distinguishes
are cortically mediated.6 However, the per- Huntington's disease.11 M2 may be pro-
sistence of long latency reflexes in spinal an- longed in dystonia. As noted above, lesions
imals forces one to consider other possible of the dorsal columns ipsilateral to the tested
explanations. For example, repetitive firing limbs or the contralateral sensorimotor cor-
of muscle spindles or transmission of sensory tex may delay or abolish M2.
influences by slowly conducting fibers could
explain the appearance of reflex activity at
long latencies. It is likely that the neural cir- Long Latency Reflexes to
cuits that generate long latency reflexes de- Mixed Nerve Stimulation
pend on the type of stimulus.7'8 Although
the physiologic basis of long latency reflexes Upton et al.12 discovered a late response,
is a matter of controversy, the weight of evi- termed V2, in response to electric stimula-
dence strongly suggests that "loops" involv- tion of a mixed nerve. Since that time, a se-
ing the motor cortex are involved in their ries of long latency reflexes in the hand mus-
generation.9 cles has been distinguished in response to
Long latency reflexes have been recorded median nerve stimulation.
by various techniques using stretch or dif- The median nerve is stimulated at the
ferent forms of electric stimulation. The pre- wrist, with the cathode proximal. Surface
cise character of the reflexes depends on the EMG electrodes are placed over the abduc-
testing protocol, and it is unlikely that all tor pollicis brevis. The stimulus duration is
long latency reflexes have an identical phys- set to 1 ms, and the intensity is increased to
iologic basis.10 the level that produces the first small twitch.
Initially, recordings are made with the mus-
cle at rest, and single shocks are given. Next,
Long Latency Reflexes to Stretch the patient maintains a moderate contrac-
tion, and stimuli are given at a rate of 1-3
All protocols for stretch reflex testing involve Hz. The signal must be rectified and then
a computer-controlled torque motor that averaged. A total of 100-500 stimuli provide
can be programmed to maintain a steady reproducible results.
load or to introduce rapid perturbations. In normal subjects with their hands at rest,
Generally, the torque is delivered through a no late response is seen after the F wave and
manipulandum that the subject holds. The before the voluntary reaction time. With
subject receives visual feedback about the contraction and averaging, a short latency
position of the manipulandum and attempts reflex develops at approximately 28 ms and
to hold it stationary against a low constant is identical with the H reflex. A late response
torque. The computer delivers random named the LLRII appears in all normal sub-
torque pulses, and the surface EMG signals jects at 50 ms, and in one-third of them, an
are recorded over the agonist and antago- additional LLRI may be recorded at 40 ms
nist of the joint that is stretched. The EMG or an LLR III at 75 ms.
signal is rectified and averaged. In myoclonic disorders, a late response
In a normal response, an Ml component that corresponds to a reflex myoclonic jerk
occurs at 30 ms and corresponds to the is recorded at 40-60 ms in the hand at rest.
monosynaptic stretch reflex. The M2 com- Sutton and Mayer13 named this the C reflex,
ponent, the most frequent long latency com- a distinct abnormality that may be seen in
cortical reflex myoclonus, reticular reflex nal excitatory wave, E2, appears at approxi-
myoclonus, hyperekplexia, Alzheimer's dis- mately 66 ms.20'21 II and E2 depend on the
ease, and other symptomatic myoclonic dis- motor cortex and are absent with contralat-
orders. The pattern of an increased LLR I eral cortical lesions.21 E2 may be delayed in
and LLR III with a normal LLR II may typ- multiple sclerosis. The depth of II is de-
ify Parkinson's disease. The LLR II is absent creased in patients with Parkinson's disease
in Huntington's disease but is unaffected in as compared with control subjects.22 E2 is in-
other choreatic movement disorders.11'14 Pa- creased in myoclonus associated with aki-
tients with focal dystonia may display in- netic rigid disorders, and the latency of this
creased LLR I or reduced LLR II.15 Possibly component may separate Parkinson's dis-
in a subgroup of patients with essential ease and multiple systems atrophy from cor-
tremor, the LLR I is increased.16 A delayed ticobasal degeneration.23
LLR II reflects slowing of central conduction
in patients with multiple sclerosis.17 LLR II
disappears after thalamic infarction, corre- The Flexor Reflex
sponding to the clinical deficit and abnor-
malities on somatosensory evoked potential Certain stimuli may trigger reflex with-
testing. 18 drawal, flexor reflexes. Generally, the adequate
stimulus must be cutaneous and noxious.
Such stimuli affect a polysynaptic network of
Cutaneous Reflexes spinal neurons, termed flexor reflex afferents,
that program patterned withdrawal behavior.
Long latency reflexes evoked by purely cu- Meinck et al.24 standardized a technique
taneous nerve stimulation demonstrate well- for eliciting flexor reflexes. The medial plan-
defined inhibitory as well as excitatory peri- tar nerve is stimulated in the ball of the foot
ods.19 Stimuli are delivered to the index with a train of five shocks by using a stimu-
finger by ring electrodes. The surface EMG lus duration at 0.1 ms and an interstimulus
is recorded with an electrode over the belly interval of 3-5 ms. The stimulus intensity is
of the first dorsal interosseous and a refer- adjusted to the motor threshold of the flexor
ence electrode over the radial styloid pro- hallucis brevis muscle, a level that will be per-
cess. The patient maintains a steady force at ceived as mildly painful. EMG is recorded
20% of maximum. Stimulus intensity is ad- from the anterior tibial muscle, rectified,
justed to four times the sensory threshold, and then averaged. A total of eight trains are
with a stimulus duration of 0.2 ms. With a averaged, with stimuli repeated every 1-3
stimulus rate of 3 Hz, 100-500 samples are seconds. With this protocol, the normal ac-
collected, full-wave rectified, and then aver- tivity is triphasic, with a large Fl response at
aged. 70 ms, a period of silence, and then a small
A first excitatory period, El, is present at F2 burst at approximately 150 ms.
approximately 40 ms, followed by inhibition, With spinal cord lesions, exaggerated
/}, at approximately 51 ms (Fig. 30-1). A fi- withdrawal corresponds to a large response
at or beyond the latency of F2. Very high-in-
tensity stimulation may shorten the latency
of this response to a value consistent with Fl.
In spasticity caused by a hemispheric lesion,
stimulation may trigger alternating clonic
bursts in the anterior tibial and gastrocne-
mius muscles.
THE SILENT PERIOD

Figure 30-1. The cutaneous reflex. The index finger
was stimulated electrically, and electromyographic ac-
tivity was recorded from the flexor digitorum indicis.
If a strong shock is delivered to the nerve of
Two phases of excitation, El and E2, and an interven- a muscle that is tonically contracting, a pe-
ing phase of inhibition, II, are seen in the record. riod of relative or absolute silence begins im-
Long Latency Reflexes and the Silent Period 397
ened silent period or its absence has been

reported in the case of a cervical cord tumor
that produced arm rigidity.30 Prolonged du-
ration of the silent period has been reported
in dystonia and Parkinson's disease. 1 The
silent period persists in patients with pure
sensory neuropathy and absence of sensory
nerve action potentials, raising the possibil-
ity that it provides an electrophysiologic as-
Figure 30-2. The silent period. After a supramaximal sessment of the integrity of smaller, slower
shock, thenar electromyographic activity is inhibited. conducting sensory fibers.32
The silence is interrupted by an M wave and H reflex.
SUMMARY
mediately and persists for about 100 ms (Fig. Long latency reflexes and the silent period
30-2). The depth of the silent period depends are EMG phenomena that reflect the com-
entirely on the intensity of the shock. With plex interplay of spinal, brain stem, and cor-
supramaximal shocks, which are commonly tical influences in motor control. These
used, the silence is generally complete ex- techniques have been applied to the study
cept for an intervening F wave. With lower of disorders of motor control such as Parkin-
stimulation intensities, the LLR I-III de- son's disease, Huntington's disease, and dys-
scribed above appear. tonia. Abnormalities of these reflexes may
Initially, Merton25 thought the silent pe- help to detect lesions of the central nervous
riod resulted from the muscle twitch and un- system in multiple sclerosis.
loading of muscle spindles induced by the
shock. This hypothesis became untenable
with the demonstration that the silent pe- REFERENCES
riod persists with stimulation of a cutaneous
nerve or a nonhomologous nerve or with 1. Johnson MT, Kipnis AN, Lee MC, Ebner TJ. Inde-
stimulation proximal to a nerve block—all pendent control of reflex and volitional EMG mod-
conditions in which twitch is absent.26 The ulation during sinusoidal pursuit tracking in hu-
silent period should be viewed as a multi- mans. Exp Brain Res 96:347-362, 1993.
2. Hammond PH. The influence of prior instruction
factorial phenomenon. With supramaximal to the subject on an apparently involuntary neuro-
stimulation, approximately the first 30 ms of muscular response (abstract). J Physiol (Lond)
silence results from the collision of impulses 132:17P-18P, 1956.
in the nerve trunk. The next period, up to 3. Marsden CD, Merton PA, Morton HB. Servo action
approximately 60 ms, may reflect activation in the human thumb. J Physiol (Lond) 257:1-44,
1976.
of recurrent collaterals of Renshaw cells. 4. Goodin DS, Aminoff MJ, Shih PY. Evidence that the
The final period of silence should be viewed long-latency stretch responses of the human wrist
as a long latency inhibitory reflex. Recent ev- extensor muscle involve a transcerebral pathway.
idence, including the study of the silent pe- Brain 113:1075-1091, 1990.
5. Capaday C, Forget R, Fraser R, Lamarre Y. Evidence
riods after cortical magnetic stimulation, for a contribution of the motor cortex to the long-
raises the possibility of spinal inhibition of latency stretch reflex of the human thumb. J Phys-
corticospinal inputs or of cortically mediated iol 440:243-255, 1991.
inhibitory reflexes.27'28 6. Hallett M, Chadwick D, Marsden CD. Cortical re-
Few normative data exist about the depth flex myoclonus. Neurology 29:1107-1125, 1979.
7. Palmer E, Ashby P. Evidence that a long latency
and duration of the silent period; thus, this stretch reflex in humans is transcortical. J Physiol
period is interpreted in an "all-or-nothing" (Lond) 449:429-440, 1992.
fashion. In states of hyperexcitability of the 8. Palmer E, Ashby P. The transcortical nature of the
distal nerve or muscle, the silent period may late reflex responses in human small hand muscle
to digital nerve stimulation. Exp Brain Res 91:320-
be absent because ectopic impulses arise dis- 326, 1992.
tal to the stimulus. In tetanus, the silent pe- 9. Matthews PB. The human stretch reflex and the mo-
riod may be abbreviated or absent.29 A short- tor cortex. Trends Neurosci 14:87-91, 1991.
10. Hallett M. Long-latency reflexes. In Quinn NP, Jen- and their central nervous pathways studied in man.
ner PG (eds). Disorders of Movement: Clinical, J Physiol (Lond) 333:405-419, 1982.
Pharmacological and Physiological Aspects. Acade- 22. Fuhr P, Zeffiro T, Hallett M. Cutaneous reflexes in
mic Press, London, 1989, pp 529-541. Parkinson's disease. Muscle Nerve 15:733-739,1992.
11. Noth J, Podoll K, Friedemann HH. Long-loop re- 23. Chen R, Ashby P, Lang AE. Stimulus-sensitive my-
flexes in small hand muscles studied in normal sub- oclonus in akinetic-rigid syndromes. Brain 115:
jects and in patients with Huntington's disease. 1875-1888, 1992.
Brain 108:65-80, 1985. 24. Meinck HM, Kuster S, Benecke R, Conrad B. The
12. Upton AR, McComas AJ, Sica RE. Potentiation of flexor reflex—influence of stimulus parameters on
"late" responses evoked in muscles during effort. the reflex response. Electroencephalogr Clin Neu-
J Neurol Neurosurg Psychiatry 34:699-711, 1971. rophysiol 61:287-298, 1985.
13. Sutton GG, Mayer RF. Focal reflex myoclonus. 25. Merton PA. The silent period in a muscle of the
J Neurol Neurosurg Psychiatry 37:207-217, 1974. human hand. J Physiol (Lond) 114:183-198, 1951.
14. Deuschl G, Strahl K, Schenck E, Lucking CH. The 26. Leis AA, Ross MA, Emori T, Matsue Y, Saito T. The
diagnostic significance of long-latency reflexes in silent period produced by electrical stimulation of
multiple sclerosis. Electroencephalogr Clin Neuro- mixed peripheral nerves. Muscle Nerve 14:1202-
physiol 70:56-61, 1988. 1208, 1991.
15. Naumann M, Reiners K. Long-latency reflexes of 27. Fuhr P, Agostino R, Hallett M. Spinal motor neu-
hand muscles in idiopathic focal dystonia and their ron excitability during the silent period after corti-
modification by botulinum toxin. Brain 120:409- cal stimulation. Electroencephalogr Clin Neuro-
416, 1997. physiol 81:257-262, 1991.
16. Deuschl G, Lucking CH, Schenck E. Essential 28. Leis AA, Stetkarova I, Beric A, Stokic DS. Spinal mo-
tremor: electrophysiological and pharmacological tor neuron excitability during the cutaneous silent
evidence for a subdivision. J Neurol Neurosurg Psy- period. Muscle Nerve 18:1464-1470, 1995.
chiatry 50:1435-1441, 1987. 29. Risk WS, Bosch EP, Kimura J, Cancilla PA, Fis-
17. Deuschl G, Lucking CH, Schenck E. Hand muscle chbeck KH, Layzer RB. Chronic tetanus: clinical re-
reflexes following electrical stimulation in chore- port and histochemistry of muscle. Muscle. Nerve
atic movement disorders. J Neurol Neurosurg Psy- 4:363-366, 1981.
chiatry 52:755-762, 1989. 30. Weinberg DH, Logigian EL, Kelly JJ Jr. Cervical as-
18. Chen CC, Chen JT, Wu ZA, Kao KP, Liao KK. Long trocytoma with arm rigidity: clinical and electro-
latency responses in pure sensory stroke due to thal- physiologic features. Neurology 38:1635-1637,
amic infarction. Acta Neurol Scand 98:41-48,1998. 1988.
19. Caccia MR, McComas AJ, Upton AR, Blogg T. Cu- 31. Pullman SL, Ford B, Elibol B, Uncini A, Su PC,
taneous reflexes in small muscles of the hand. J Fahn S. Cutaneous electromyographic silent period
Neurol Neurosurg Psychiatry 36:960-977, 1973. findings in brachial dystonia. Neurology 46:503-
20. Fuhr P, Friedli WG. Electrocutaneous reflexes in 508, 1996.
upper limbs—reliability and normal values in 32. Leis AA, Kofler M, Ross MA. The silent period in
adults. Eur Neurol 27:231-238, 1987. pure sensory neuronopathy. Muscle Nerve 15:
21. Jenner JR, Stephens JA. Cutaneous reflex responses 1345-1348, 1992.
Chapter 31
SURFACE ELECTROMYOGRAPHIC
STUDIES OF MOVEMENT
DISORDERS
James H. Bower
Joseph Y. Matsumoto
TECHNIQUES DYSTONIA
NORMAL PATTERNS Recording Techniques
TREMOR Abnormal Patterns
Recording Techniques TICS, CHOREA, AND ATHETOSIS
Abnormal Patterns SURFACE ELECTROMYOGRAPHIC STUDY
MYOCLONUS OF VOLUNTARY MOVEMENT
Recording Techniques DISORDERS
Abnormal Patterns SUMMARY
The study of movement disorders encom- cient in discerning the fine details of move-
passes abnormalities of motor control that ment, such as timing (Which body part
result in either too little or too much move- moved first?) and regularity (Is the move-
ment. At one pole are akinetic-rigid syn- ment tremulous or irregular?). At times, de-
dromes, such as Parkinson's disease, and at tails such as these are critical in diagnosis.
the other are involuntary movements, such Also, experimental studies in motor control
as tremor, myoclonus, dystonia, chorea, and demonstrate clearly that the brain, spinal
tics. Movement disorders stem from com- cord, and musculoskeletal system are able to
plex and poorly understood pathophysio- produce a specific movement with a large
logic processes that occur in the central number of different motor patterns. As a
nervous system. These processes are largely practical example, rapid elbow flexion may
inaccessible with even the most sophisticated result from either a brief, isolated contrac-
electrophysiologic techniques and neuro- tion of the biceps muscle or prolonged ac-
imaging procedures. Thus, the enduring tivity of the biceps and triceps muscles. In
tool in evaluating clinical movement disor- this example, identification of the underly-
ders is the trained human eye that, together ing motor pattern may distinguish myo-
with the clinical history, provides an accu- clonus from dystonia.
rate diagnosis in most cases. Surface electromyographic (EMG) studies
Clearly, reliance on diagnosis by visual in- provide a method for obtaining data that
spection is not always sufficient. Although complements and extends the clinical ex-
observation is excellent for perceiving the amination. Multichannel EMG studies map
overall pattern of movement, it is less profi- the temporal pattern of movement with a
399
Table 31-1. Movement Disorders That Surface Electromyography

May Help to Identify and Characterize
Dystonia Myorhythmia
Idiopathic torsion dystonia Opsoclonus-myoclonus
Tardive dystonia
Tremor
Post-traumatic dystonia
Essential tremor
Blepharospasm
Exaggerated physiologic tremor
Meige's syndrome
Extrapyramidal/parkinsonian tremor
Oromandibular dystonia
Cerebellar tremor
Cervical dystonia
Neuropathic tremor
Focal arm dystonia (writer's
cramp) Post-traumatic tremor
Leg dystonia Orthostatic tremor
Truncal dystonia Psychogenic tremor
Psychogenic dystonia Primary writing tremor
Chorea Other
Huntington's disease Paroxysmal dyskinesias
Sydenham's chorea Multiple sclerosis tonic spasms
Neuroacanthocytosis Diaphragmatic flutter
Abdominal wall dyskinesias
Myoclonus Tourette's syndrome
Essential myoclonus Tic disorders
Symptomatic myoclonus Hyperekplexia
Cortical myoclonus Stiffman syndrome
Subcortical myoclonus Painful legs/moving toes
Spinal myoclonus Hemifacial spasm
Propriospinal myoclonus Hemimasticatory spasm
Psychogenic myoclonus Ballismus
Palatal myoclonus
resolution measured in milliseconds. Fur- find disposable adhesive electrodes conve-

thermore, the surface EMG reflects not only nient because they can be applied rapidly
alpha motor neuron activity but also, by in- when multiple muscles are recorded. After
ference, specific abnormal central com- the skin has been cleansed and mildly
mands that underlie a movement disorder. abraded, the electrodes are placed 2—3 cm
Surface EMG studies are noninvasive and are apart over the motor point of the muscle and
within the capability of any EMG laboratory. oriented parallel to the course of the mus-
We find them particularly useful in the study cle fibers. The iliac crest provides a relatively
of involuntary movement disorders in which inactive site for the ground electrode.
the patterns of surface EMG activity are best A major technical limitation of surface
described (Table 31-1). EMG studies is the lack of selectivity. The
activity of a single muscle is never actually
recorded because adjacent muscles in-
TECHNIQUES evitably contribute "crosstalk" to the signal.
This effect is minimized by use of short in-
Surface EMG recording can be performed terelectrode distances and by recording
with any high-quality disk electrodes. We from relatively superficial and isolated mus-
Surface Electromyographic Studies of Movement Disorders 401
cles, such as the biceps, deltoid, quadriceps, through a hypodermic needle that is then
tibialis anterior, or first dorsal interosseus. withdrawn. After the electrodes are in posi-
At times, a group of muscles, such as the tion, they remain stable for many hours and
forearm flexors or extensors, are intention- resist displacement by even vigorous body
ally recorded. movement. When selective recording is
The quality of the surface EMG signal needed for only short recording periods,
must be assessed carefully before analysis. standard concentric or monopolar needle
This signal represents the interference pat- recording may be suitable. In any situation,
tern of multiple motor units with high fre- the improved selectivity of intramuscular
quencies filtered out by the intervening skin recording must be balanced against the
and subcutaneous tissue. Deep muscles, added discomfort to the patient.
such as the gluteus maximus or any muscle
in an obese person, may produce a signal
that is too degraded for analysis. The fre- NORMAL PATTERNS
quency spectrum of the signal contains
power throughout the range between 10 and Three normal patterns of surface EMG ac-
1000 Hz, with maximal power at approxi- tivity are recognized: reflex, tonic, and ballis-
mately 100 Hz. In practice, a low-frequency tic (Fig. 31-1). Reflex activity, such as the
filter cutoff of at least 30 Hz must be used monosynaptic tendon jerk, produces brief,
to eliminate the unwanted effects of move- synchronized discharges of alpha motor
ment artifact. A high-frequency filter set- neurons. The surface EMG appearance of
ting of 2000 Hz passes the important high- this discharge is a short (10-30 ms) burst of
frequency components of the signal. The activity in agonist and, at times, antagonist
amplification factor is set arbitrarily to dis- muscles. This reflex pattern is an involuntary
play a maximal voluntary contraction that response to stimulation. Indeed, most per-
fills the amplifier range without blocking. Af- sons are unable to generate voluntarily
ter the EMG signal has been collected, it may bursts that have this short duration. How-
be displayed as the raw interference pattern ever, voluntary movement produces two
or digitally processed to display a full-wave types of EMG patterns. When a person
rectified signal or smoothed EMG envelope. moves a limb slowly or holds it in a static pos-
For movement disorder studies, the impor- ture, a tonic pattern results. This pattern con-
tant measurements are the onset latencies sists of a continuous and steady EMG dis-
and burst durations. The amplitude of the charge, often with cocontraction of agonist
bursts is extremely variable and rarely useful and antagonist muscles. Cocontraction is a
in routine clinical studies. normal mechanism of motor control that in-
However, if a study demands highly selec- creases the stiffness, or resistance, across a
tive recording, intramuscular electrodes joint. By contrast, when one wills a very rapid
must be used. Electrodes fashioned from movement of a joint, a ballistic pattern devel-
fine wire are useful for this purpose. Pairs of ops. An initial agonist burst of 50- to 100-ms
wires are inserted into the selected muscle duration leads the pattern, followed by an
Figure 31-1. The three normal

surface electromyographic pat-
terns: A, reflex pattern; B, tripha-
sic pattern; and C, tonic pattern.
Upper trace, agonist muscle; lower
trace, antagonist muscle.
antagonist burst of 50- to 100-ms duration Recording Techniques

and a silent period in the agonist. A final ag-
onist burst completes the EMG activity. The Electrodes are placed over the agonist and
triphasic pattern appears to be fundamental antagonist pairs of muscles active in driving
to the motor control of ballistic limb move- the tremor. Often, forearm flexors and ex-
ments.1 The theory explaining this is that the tensors are most active and care must be
initial agonist burst scales the size of the bal- taken to eliminate crosstalk between the sig-
listic movement, the antagonist burst brakes nals by electrode positioning, because this
the limb, and the final agonist burst deter- technical error may lead to misinterpreta-
mines its destination. Movement disorders tion. Amplitude of the tremor bursts is as-
reflect alterations in the excitability or con- sessed in various limb positions. Tremors are
trol of one of these three patterns. grouped into those occurring at rest and
those occurring with action. Action tremors
are further subdivided into postural tremors
TREMOR (a body part maintains a position against
gravity), isometric tremors (muscle contraction
Tremor is an oscillation of a body part at a against a stationary object), simple kinetic
regular frequency. Normally, motor unit re- tremors (nontarget-directed voluntary move-
cruitment produces a random pattern of fir- ment), and intention tremor (at the termina-
ing, with motor unit contractions separated tion of a target-directed movement).3 An-
evenly in time. In these circumstances, other variable used in tremor analysis is the
twitch tensions fuse to produce a steady pattern of agonist-antagonist firing. One
force. In tremor, the firing of motor unit po- muscle may fire while the other is silent, in
tentials becomes synchronized, increasing an alternating pattern. In other tremors, the
the probability that two or more motor unit pair may fire simultaneously in a synchro-
potentials will fire in temporal proximity and nous pattern. Rarely, the EMG pattern may
create a jerky, unfused force. As synchro- shift from one pattern to another during the
nization increases, so does the amplitude of period of recording.
the tremor. The three sources of tremor are
mechanical, reflex, and a central oscillator.
All mechanical objects oscillate at a resonant
frequency related to their inertia and stiff-
Abnormal Patterns
ness. Because body parts are physical objects, EXAGGERATED PHYSIOLOGIC
they have a resonant frequency. In addition,
TREMOR
muscles are connected to the central ner-
vous system through peripheral nerve reflex Even normal subjects have a fine, often im-
loops, which oscillate at varying frequencies. perceptible tremor when holding part of the
Also, areas of the central nervous system os- body in a static posture. Physiologic tremor is
cillate spontaneously, possibly producing primarily a mechanical tremor, but in some
rhythmic motor activity in the related body persons a central oscillator may have a
part. The contribution of all three of these role.4'5 This tremor often has no surface
sources can contribute to the clinical phe- EMG correlate. Under circumstances that
nomenon of tremor. increase catecholaminergic secretion, such
The surface EMG of tremors records the as anxiety, fatigue, hypothermia, hypo-
grouping of motor unit potentials as discrete glycemia, thyrotoxicosis, or pheochromocy-
bursts of activity.2 Analysis of these EMG toma, physiologic tremor increases in am-
bursts helps to establish whether a move- plitude and becomes visible. In this state,
ment is truly tremulous. Disorders such as known as exaggerated physiologic tremor, un-
phasic dystonia may appear regular on visual derlying surface EMG activity emerges. This
inspection but be shown to be irregular may consist of a low-grade interference pat-
when measured on an EMG recording. Con- tern or synchronous bursts of 50- to 100-ms
versely, a very coarse tremor may appear so duration and a frequency of 8-12 Hz. This
jerky as to be myoclonic. pattern is often best recorded in distal up-
per extremity muscles, such as ringer exten- Static postures initially attenuate the burst
sors and flexors. amplitudes; however, when such postures
are held for 20 seconds or longer, the tremor
bursts may reappear. The frequency of the
ESSENTIAL TREMOR postural/kinetic component is usually the
Essential, or familial, tremor is the most com- same as the rest component, but sometimes
mon movement disorder. It is usually a bi- it can be substantially (> 1.5 Hz) higher.
lateral, symmetrical postural or kinetic This may be the case in patients who have
tremor involving the upper extremities. Ad- both essential tremor and parkinsonism.3
ditional or isolated head tremor may also oc-
cur.3 The voice can also be affected. Essen- CEREBELLAR TREMOR
tial tremor is thought to result from a central
oscillator (an often suggested but unproven Cerebellar, or intention, tremor occurs with tar-
candidate is the inferior olivary nucleus).5 get-directed limb movements and is most
The surface EMG shows strongly circum- prominent toward the termination of the
scribed bursts of activity that are synchro- movement. The tremor frequency is mainly
nous in agonist and antagonist muscles, with less than 5 Hz. Postural tremor may also be
a frequency of 4-12 Hz. Of interest, the fre- present. Serial dysmetria is often confused
quency of essential tremor declines with with intention tremor. In dysmetria, surface
age.6 When the tremor is severe, less promi- EMG studies indicate that the terminal
nent activity may be recorded, even with the movements are not the regular oscillations
limb at rest. of tremor but rather a series of inaccurate,
Sabra and Hallett7 have reported patients irregular ballistic movements.
with essential tremor who clearly had an al-
ternating EMG pattern. This group of pa-
HOLMES TREMOR
tients appeared to have a tremor that was
slower in frequency and poorly responsive to Holmes tremor (also known as rubral tremor,
propranolol. Milanov8 has reported that the midbrain tremor, thalamic tremor, or myorhyth-
alternating pattern is more common than mid) is a rest and intention tremor that oc-
previously believed. This dictates caution in casionally has an irregular presentation.
diagnosing tremor on the basis of a single Holmes tremor frequency is usually less than
EMG characteristic. The surface EMG find- 4.5 Hz and often associated with a recog-
ings in neuropathic tremor are identical to nized pathologic insult (for example, infarct
those in essential tremor. or demyelinating plaque), in which case a
delayed onset (weeks to 2 years) is common.3
PARKINSONIAN TREMOR
TASK- AND POSITION-SPECIFIC
Tremor commonly accompanies akinetic-
TREMOR
rigid syndromes such as Parkinson's disease.
Clinically, parkinsonian tremors appear Several tremors occur only with specific tasks
maximal at rest and attenuate with action. or positions. The classic example is primary
The dominant frequency is 4—7 Hz. Surface writing tremor. Although this tremor is pre-
EMG studies of extrapyramidal tremors dominant with writing, it often spills over
demonstrate an alternating pattern of con- into other activities such as eating or groom-
traction that is constant through the period ing. The surface EMG correlate consists of
of recording. Burst durations are typically in bursts of 100-ms duration that occur maxi-
the range of 50-100 ms. The frequency of mally in the pronator teres, supinator, or
the bursts varies little through the period of wrist flexors and extensors. The pattern may
the recording. Bursts in muscles throughout be synchronous or alternating. Taps to the
the body have the same frequency. The reg- forearm, particularly in a direction that pro-
ularity of the tremor increases as the disease duces supination of the forearm, may stim-
progresses, and this may help distinguish ulate bursts of tremor. Isolated voice tremor is
parkinsonian tremor from other tremors.9 another example. It is not known whether
these tremors represent a form of essential PALATAL TREMOR

tremor or dystonia or have another cause.3
Palatal tremor is a rhythmic, involuntary
movement that causes palatal elevation. At
ORTHOSTATIC TREMOR times, it may spread to involve tongue, neck,
(SHAKY LEGS SYNDROME) facial, and even limb muscles. Palatal EMG
activity is detected with surface electrodes
Heilman10 has described a distinctive tremor, linked to the mastoid processes or placed on
called orthostatic tremor, that occurs predom- the anterior neck. Bursts recur at a regular
inately in the elderly. Patients complain of frequency of approximately 2 Hz. Palatal
"quivering" in their legs shortly after they tremor can stem from a symptomatic lesion,
stand. This may cause a sense of instability with associated olivary hypertrophy, or it can
so severe that walking becomes difficult. The be essential in nature. Rhythmic movements
surface EMG pattern is distinctive and dis- of the levator veli palatini muscles cause the
plays high-amplitude 13-18 Hz tremor bursts symptomatic palatal tremors, but the essen-
(Fig. 31-2). Because of the rapid frequency, tial palatal tremors involve mainly the ten-
this diagnosis can be difficult to make in the sor veli palatini. Involvement of the latter
office. A surface EMG study is often the only muscle often causes an audible ear click to
way to make the definitive diagnosis. The the patient.
bursts are recorded in the legs and para-
spinal muscles, with the patient standing. FUNCTIONAL TREMOR
The agonist—antagonist relationship may
vary during the recording. At rest, the legs Occasionally, tremor may have a pattern that
are electrically silent. is atypical, and hysteria or malingering is sus-
Figure 31-2. Orthostatic tremor. Surface electromyographic activity recorded with the patient standing and recti-
fied for display. Tremor activity at approximately 15 Hz is recorded maximally from the leg muscles.
pected. Surface EMG can provide support- encephalopathies, such as Lafora body dis-
ive evidence by demonstrating a pattern that ease, Creutzfeldtjakob disease, Alzheimer's
does not correspond to any of those de- disease, metabolic encephalopathies, and
scribed above. Functional tremors tend to be anoxic encephalopathies.
confined to a single limb and seldom display Physiologic analysis represents yet another
a dominant frequency throughout a pro- approach to myoclonus. By using available
longed recording. Indeed, the tremor fre- experimental and clinical data, Halliday13 di-
quency and amplitude tend to vary widely vided myoclonus into three categories: (1)
with time, with change of position, or with pyramidal myoclonus, characterized by brief
distraction. However, diagnostic proof for a EMG bursts that follow a cortical discharge
functional or voluntary origin for a tremor with a fixed, short latency; (2) extrapyramidal
cannot be offered. myoclonus, thought to arise from deeper,
noncortical sites and characterized by long
EMG burst durations and the absence of a
MYOCLONUS clear preceding cortical discharge; and (3)
segmental myoclonus, viewed as a local phe-
Myoclonus is a sudden, brief shock-like in- nomenon arising from brain stem or spinal
voluntary movement caused by muscular cord lesions.
contraction (positive myoclonus) or inhibition Hallett14 cited contemporary clinical neu-
(negative myoclonus). Of the involuntary rophysiologic data to suggest a classification
movement disorders, it presents the greatest that divides myoclonus into epileptic and
challenge to diagnosis and classification. My- nonepileptic forms. By definition, epileptic my-
oclonus is best considered a general term for oclonus is a fragment of epilepsy. Its supposed
various motor phenomena that appear to be generator is a brief, focal, and synchronous
similar but have diverse pathologic and neuronal discharge similar to the paroxys-
physiologic causes. It may be approached mal depolarization shift recorded in experi-
through several classification schemes. The mental epilepsy. This concept is supported
first describes the distribution of the muscle by back-averaging studies that demonstrate
jerks. They may be focal (involving only a sin- brief and spontaneous electroencephalo-
gle limb), multifocal (affecting more than graphic (EEG) transients preceding my-
one body part in a random, independent oclonic jerks and by the extremely short sur-
fashion), generalized (involving all body parts face EMG burst durations that resemble a
simultaneously), or segmental (involving only single clonic burst of a seizure. The dis-
muscles of a given cranial or spinal seg- charges generating epileptic myoclonus are
ment) . usually cortical and may be focal or diffuse.
An etiologic classification categorizes the However, epileptic myoclonus may also orig-
immense number of disease states that may inate from spontaneous discharges in the
have myoclonus as a prominent symptom. In brain stem. Experimental models implicate
one classification, myoclonus is divided into the nucleus reticularis gigantocellularis in
four etiologic groups: (1) physiologic myoc- the medulla as a possible brain stem gener-
lonus includes motor phenomena such as ator; thus, the term reticular myoclonus is syn-
hiccups, hypnic jerks, and the startle re- onymous with brain stem myoclonus. Epileptic
sponse, which have the appearance of my- myoclonus may be reflexive in nature, indi-
oclonus but occur in normal subjects; (2) es- cating that the jerks are triggered by so-
sential myoclonus designates disease in which matosensory or acoustic stimulation at short
abnormal muscle jerks are the sole feature and reproducible reflex latencies.
of the illness; (3) epileptic myoclonus refers to Nonepileptic myoclonus refers to muscle jerks
myoclonus in the setting of epilepsy; and (4) that are presumed to be generated by more
symptomatic myoclonus represents all other dis- complex mechanisms of the basal ganglia,
ease states in which myoclonus occurs as a brain stem, or spinal cord. These mecha-
sign, often in the setting of encephalopathy. nisms involve circuits more widely distributed
Dementia, ataxia, and parkinsonism are in space and time than those of epileptic dis-
common associated features.12 In this group- charges. Thus, a discrete EEG correlate is not
ing are included the progressive myoclonic recorded in these disorders. In general, the
EMG burst durations are longer than 100 ms spontaneous and stimulus-sensitive muscle
and are never as consistendy short as those jerks confined mainly to a single limb. If the
seen in epileptic myoclonus. condition is severe, generalized jerks may oc-
The physiologic abnormalities found in cur. Focal tongue and palatal jerks may af-
myoclonus do not always correspond to an fect speech. In these patients, EEG studies
anatomical abnormality. A similar phenom- demonstrate that a focal area of hyperex-
enon occurs in epilepsy, for example, when citable cortex transiently depolarizes ap-
a temporal lobe sharp wave is recorded in proximately 10-30 ms before EMG dis-
the absence of any demonstrable pathologic charges. These discharges tend to be
condition. Nevertheless, physiologic studies localized to a group of contiguous muscles
offer important guidance in evaluation and and are 15-40 ms in duration. Muscles dis-
treatment of myoclonus. charge synchronously, and there is little vari-
ation in the discharge pattern of the jerks.
Monitoring of cranial muscles demonstrates
Recording Techniques a cranial-to-caudal progression of activation
involving the masseter muscle, facial mus-
Our approach is first to attempt to record cles, and sternocleidomastoid muscle. This
the surface EMG pattern associated widi pattern may be expressed over several mil-
spontaneous muscle jerks. The critical mea- liseconds and requires fast recording speeds.
surement is that of the burst duration, for In cortical reflex myoclonus, these same EMG
which we use the median value of 10—20 bursts may be elicited by electrical or tap-
jerks. Multiple-channel recording allows one ping stimuli, generally at a latency of 40—60
to determine the order of muscle activation ms, in the hand (Cresponse). Somatosensory
for each burst and also the agonist-antago- evoked potentials typically demonstrate en-
nist patterns across a joint. Capturing the larged amplitudes of the P25 and N33 peaks.
surface EMG bursts responsible for inter- The causes of focal cortical myoclonus in-
mittent muscle jerks is a challenge unless clude anoxia, tumors, trauma, and infec-
the movements occur at least once every few tions. One group of patients with focal cor-
minutes. By using a triggering device and the tical dysplasia has been described.
delay buffer available on most EMG ma- Patients with primary generalized epilepsy
chines, data acquisition can be set to begin manifest a different form of cortical my-
with EMG activity in the most active muscle oclonus termed primary generalized epileptic
and to center the data points around the on- myoclonus. The common manifestation of
set of the EMG trigger. this fragmentary form of epilepsy is random,
After demonstrating the areas of maximal spontaneous twitching of the fingers. This
EMG activity, we select the most active chan- movement disorder, called polyminimyoc-
nels for C reflex or other stimulation stud- lonus, is nonspecific, was observed first in pa-
ies. Somatosensory evoked potentials may tients with neuromuscular disease, such as
also demonstrate cortical neurons that are spinal muscular atrophy, and is the outward
hypersensitive to peripheral stimulation. Fi- manifestation of frequent fasciculation. In
nally, back-averaging may be used if epilep- the setting of epilepsy, however, it should be
tic myoclonus is suspected and no cortical recognized as a form of cortical myoclonus,
accompaniments are observed on routine oftentimes accompanied by larger trunk and
EEG. limb jerks. Electroencephalographic analysis
shows that bifrontal and diffuse cortical
events precede the jerks. At times, these
events are visualized on the raw EEG as spike
Abnormal Patterns or polyspike discharges. Rarely, a similar
movement disorder may be seen in focal cor-
CORTICAL MYOCLONUS
tical myoclonus. The surface EMG manifes-
Cortical myoclonus, the most common form of tations of polyminimyoclonus are random,
epileptic myoclonus, exists in many physio- 10- to 30-ms duration bursts in the small
logic forms. Focal cortical myoclonus is a frag- hand muscles. The term cortical tremor has
ment of focal epilepsy. Patients present with been applied to a phenomenon of regular
rhythmic discharges, usually of the fingers, in the hand. Somatosensory evoked poten-
that clinically resemble essential tremor.15 tials are normal. The most common cause of
However, electrophysiologic studies show this syndrome is anoxia. Other causes in-
giant somatosensory evoked potentials, en- clude uremia and brain stem encephalitis.
hanced long loop reflexes, and a premove- Identification of reticular reflex myoclonus
ment cortical spike, indicating that cortical may be important therapeutically, because
tremor is simply a variant of cortical reflex Chadwick and associates17 found that it re-
myoclonus. sponded to 5-hydroxytryptophan and clon-
Cortical myoclonus may be present in dif- azepam better than other physiologic types
ferent forms in several symptomatic my- of myoclonus. Cortical and reticular myoc-
oclonias. In Alzheimer's disease and lonus may coexist.
Creutzfeldtjakob disease, multifocal, short-
duration EMG jerks are observed, similar to
those in cortical reflex myoclonus. Typically, SEGMENTAL MYOCLONUS
touch, shock, or noise triggers reflex my- Myoclonus confined to muscles of several
oclonus. A focal, contralateral EEG transient root distributions has long been recognized
precedes the myoclonic jerks in both dis- as a sign of spinal cord lesions, such as arte-
eases; however, timing, duration, and distri- riovenous malformations, tumors, traumatic
bution of this event differentiate the two dis- injury, or infections.18 The EMG bursts have
eases. Asterixis is a cortically driven negative durations of 250-1000 ms, appear rhythmic
myoclonus seen in various metabolic en- at frequencies of 1-3 Hz, and often persist
cephalopathies. The surface EMG demon- into sleep. A possible variant of this condi-
strates that voluntary tonic contraction is tion is dancing umbilicus syndrome, or belly-
interrupted by periods of silence that corre- dancers' dyskinesia. As these colorful terms
spond to jerk-like lapses of tone. Silent pe- suggest, the segmental myoclonic move-
riod-locked averaging clearly demonstrates ments are confined to the abdominal wall
that this inhibition is time-locked to a focal and cause distressing, continuous undula-
cortical discharge. tions. Because the involved muscles are
deep, needle recordings may be needed to
RETICULAR REFLEX MYOCLONUS uncover the cause of the symptoms.
Reticular reflex myoclonus provides the only
clear example yet described of epileptic PHYSIOLOGIC MYOCLONUS
brain stem myoclonus.16 The jerks predom- Hiccups
inantly affect proximal flexor muscles bilat-
erally and occur spontaneously 5-10 times Although hiccups are an everyday occur-
per minute. Joint stretch, noise, or touch rence, they may become a medical problem
may stimulate jerks. The EEG correlates are when prolonged and intractable. Davis19 de-
not time-locked to the myoclonus and may scribed the surface EMG characteristics of
occur before or after the EMG bursts. This individual hiccups. Electromyographic activ-
finding has been interpreted to reflect non- ity is recorded synchronously over the di-
specific activation of the cerebral cortex by aphragm and inspiratory muscles. The burst
ascending brain stem discharges. The EMG has a duration of 500 ms; however, inspira-
bursts are 10-30 ms in duration and are tory airflow is interrupted abruptly by glot-
widespread throughout the body. There is tic closure shortly after burst onset. In in-
considerable variability in the activation pat- tractable hiccups, the bursts may show a
tern of muscles from jerk to jerk. This so- periodicity of one every 2-10 seconds.
called central jitter is believed to represent
polysynaptic pathways in the brain stem neu- Hypnic Jerks
ronal generator. An ascending activation
pattern is recorded in the cranial muscles, A few myoclonic jerks, called hypnic jerks, are
suggesting a medullary generator of the ac- common during the initial descent from
tivity. The EMG bursts are linked to taps or drowsiness to sleep, and a vivid sensory ex-
shocks at short reflex latencies of 30-60 ms perience, such as the feeling of falling, of-
ten accompanies them. The jerks generally example, in the thumb, surface EMG stud-
are marked by a K complex visible on the ies record "overflow" of the typical triphasic
EEG, but little is known about their EMG EMG pattern from the distal limb into more
features. proximal muscles, such as the biceps and
triceps.
ESSENTIAL MYOCLONUS
STARTLE DISORDERS
Essential myoclonus denotes various disorders
in which myoclonus is the predominant fea- The startle reflex is a whole body jerk that
ture of the disease and encephalopathy or commonly occurs in response to sudden
seizures are absent. Most kindreds demon- unexpected noise or touch. Characteristic
strate autosomal dominant inheritance with EMG onset latencies to loud noise are well-
variable penetrance. Generalized myoclonus defined, with the orbicularis oculi invariably
is usually noted from early childhood. Stress, leading activation at 30-40 ms and the ster-
anxiety, and actions such as writing exacer- nocleidomastoid following at 55-85 ms.
bate the jerks. Many kindreds have demon- Limb muscles are less consistently active,
strated a dramatic response to alcohol. Some with the biceps activated at 85-100 ms and
family members may display associated leg muscles at 100-140 ms.21 Burst durations
tremor or focal dystonias. Under this rubric, range from 50 to 400 ms. The reflex habit-
various surface EMG patterns have been de- uates rapidly. As a normal phenomenon,
scribed in patients who otherwise were clin- startle represents another form of physio-
ically similar. The common forms are char- logic myoclonus.
acterized by 40- to 250-ms burst durations Exaggerated startle has numerous causes, in-
that occur asynchronously between muscles. cluding inflammatory brain stem lesions,
Distal upper extremity or neck muscles are anoxic injuries, psychiatric illnesses, and
affected most frequendy. There is no reflex drug intoxication. Hereditary hyperekplexia is
activation of the jerks by shocks or taps. Spo- an autosomal dominant condition charac-
radic cases of essential myoclonus have been terized by exaggerated startle to unexpected
reported. They are more clinically hetero- stimuli. A major form consisting of exagger-
geneous, and probably represent a mixed ated startle followed by prolonged stiffness
group of patients with undiscovered causes has been linked to a point mutation in the
or false family histories. gene encoding the a\ subunit of the glycine
Another pattern is seen in the syndrome of receptor.22 A minor form, consisting of ex-
myoclonic dystonia. This syndrome implies the aggerated startle alone, has no recognized
coexistence of myoclonic jerks with promi- cause. The audiogenic myoclonic jerks in hy-
nent dystonia. Frequently, torticollis is the perekplexia clearly correspond to the startle
accompanying dystonia. An autosomal dom- pattern described previously. However, the
inant or sporadic form has been described, startle reflex is increased in magnitude and
and improvement with alcohol is variable. is poorly habituating in this disorder.
The surface EMG often defines this syn- Diaphragmatic flutter may be a variant of
drome, showing prolonged synchronous palatal myoclonus. Patients with this disor-
bursts with durations of 50-250 ms blending der complain of involuntary abdominal
with prolonged dystonic activity that lasts for movements caused by bilateral diaphrag-
seconds. Antagonist-agonist cocontraction matic contractions at rates of 60-200 Hz.
is a consistent feature of the recordings.
Ballistic movement overflow myoclonus is an- AXIAL SPASMS OF PROPRIOSPINAL
other pattern. Hallett, Chadwick, and Mars- ORIGIN
den20 described patients with myoclonus af-
fecting all extremities present from birth. In this syndrome, generalized flexion jerks
Voluntary movement, especially rapid move- centered around the trunk and neck de-
ment, triggered large-amplitude jerks. The velop in middle life.23 The jerks are not trig-
patients reported no amelioration with al- gered by sensory stimuli. Surface EMG shows
cohol. With rapid voluntary movement, for initial activation of truncal muscles, espe-
Surface Electromyographic Studies of Movement Disorder 409
daily the rectus abdominis, with subsequent lus, and the activation pattern varies from
spread up and down spinal segments. The jerk to jerk.
EMG activity is bilateral, and the onset la-
tencies are consistent from jerk to jerk. Burst
durations are typically long but vary widely DYSTONIA
between 40 and 4000 ms. The estimated
conduction velocity up and down the cord Dystonia is a prolonged abnormal posture
is about 5 m/second, suggesting spread maintained by involuntary muscular con-
through slow propriospinal pathways. traction. It may be focal or generalized. The
most common focal dystonia is cervical dysto-
nia, or torticollis. Blepharospasm, oromandibu-
PERIODIC LIMB MOVEMENTS lar dystonia, and writers'or occupational cramps
OF SLEEP are other common focal dystonias. General-
Periodic jerks of the legs may interrupt sleep ized dystonia is usually a manifestation of
and cause insomnia or excessive daytime hereditary torsion dystonia. Generally, neu-
somnolence. Such periodic limb movements rophysiologic studies are most helpful in
of sleep commonly accompany restless legs evaluating the focal dystonias, often as a pre-
syndrome.24 They also may appear after lude to therapeutic injections of botulinum
spinal cord trauma or vascular injury, impli- toxin.
cating damage to descending inhibitory
pathways. During the transition between
drowsiness and light sleep, the movements Recording Techniques
begin their cyclic occurrence, with an aver-
age period lasting between 30 and 45 sec- The physiologic hallmark of dystonia is in-
onds. These movements resemble dystonia tense cocontraction of agonist and antago-
more than myoclonus; thus, the previous nist muscles, producing a marked increase
designation of nocturnal myoclonus has been in stiffness across the joint and abnormal
abandoned. The surface EMG pattern varies. posturing. Thus, muscles acting across the
Most often, the burst durations are longer postured joint should be studied to look for
than 500 ms. The earliest and most actively simultaneous interference patterns. Intra-
involved muscle is often the anterior tibial muscular electrodes often are needed to en-
muscle. Although the jerks may appear uni- sure selective recordings. Whereas cocon-
lateral, bilateral asynchronous EMG activa- traction is not specific for dystonia, it does
tion is the rule. rule out joint contractures or hysteria, in
which abnormal limb posture is unaccom-
panied by EMG activity. The EMG discharges
FUNCTIONAL MYOCLONUS may be tonic or occur in a pseudorhythmic
pattern called phasic dystonia. This pattern is
It is not uncommon for myoclonic-appear- distinguished from tremor by the lack of true
ing jerks to be the primary manifestation of rhythmicity, the variability of the burst du-
psychiatric illness such as hysteria. This di- rations, and the frequent intrusion of tonic
agnosis often is suspected clinically, but the dystonia.
neurophysiologic laboratory is asked to pro-
vide supportive evidence for the nonorganic
nature of the jerks. Thompson and cowork- Abnormal Patterns
ers25 have found several features of such
jerks on surface EMG recordings. Most im- TORTICOLLIS
portant, the jerks should be triggered by a
measured stimulus. When this is done, the Deuschl and associates26 have described the
onset latencies of the EMG bursts are vari- patterns of EMG discharge in spasmodic tor-
able and longer than the shortest voluntary ticollis. With rotational torticollis, the con-
reaction time in a normal subject. The jerks tralateral sternocleidomastoid and the ipsi-
often habituate rapidly to any given stimu- lateral splenius capitus are most often active.
In retrocollis, all posterior neck muscles are SPASTIC DYSPHONIA

active, and in laterocollis, the ipsilateral sple-
nius capitus and sternocleidomastoid mus- Spastic dysphonia may be of the adductor or
cles are active. We have found similar pat- abductor type. Patients with adductor spastic
terns, but variations in a particular pattern dysphonia present with a strained or tremu-
of muscle activity are common. For this rea- lous voice caused by dystonia of the thy-
son, we perform multichannel EMG record- roarytenoid muscles. These muscles can be
ing with intramuscular electrodes in all pa- recorded by needle examination performed
tients before injecting botulinum toxin percutaneously or by direct laryngoscopy.
(Fig. 31-3). Preinjection EMG studies may With either method, the participation of an
account for some degree of additional ther- experienced otorhinolaryngologist is ad-
apeutic benefit.27 vised. In contrast, abductor spastic dysphonia
is manifested as a whispering voice and the
dystonic activity is in the posterior cricoary-
OROMANDIBULAR DYSTONIA tenoid muscles, which are inaccessible to rou-
In oromandibular dystonia, patients may be un- tine examination.
able to eat or speak because of abnormal jaw
posturing. Jaw-opening dystonia frequently WRITERS' AND OCCUPATIONAL
reflects dystonic activity in the lateral ptery- CRAMPS
goid and digastric muscles. Jaw-closing dys-
tonia reflects activity in the temporalis, mas- In various circumstances, repetitive skilled
seter, and medial pterygoid muscles. motions may become complicated by painful
figure 31-3. Dystonia. Electromyographic activity recorded with intramuscular electrodes in a patient with spas-
modic torticollis. Both tonic and irregular phasic EMG bursts are present.
and disabling dystonia of the hand or wrist. longed. Finally, in patients with athetosis
There is no typical posture, and a combina- caused by cerebral palsy, excessive activity
tion of flexion and extension dystonia may occurs in muscles not normally involved in
occur. Intramuscular EMG recordings show the main action, and agonist and antagonist
the individual pattern of phasic or tonic muscles often cocontract.
spasms in multiple cocontracting forearm
muscles.
SUMMARY
TICS, CHOREA, AND Surface EMG recordings provide a simple
ATHETOSIS and noninvasive means of studying move-
ment disorders. These techniques are par-
Surface EMG recording is of limited useful- ticularly helpful in classifying involuntary
ness in the evaluation of tics, chorea, and movements such as tremor and myoclonus.
athetosis. Although these involuntary move-
ments are clinically distinct, the surface
EMG patterns are nonspecific and may ap- REFERENCES
pear similar in all three. Burst durations can
be 100-300 ms and can have reflex, tonic, 1. Hallett M, Shahani BT, Young RR. EMG analysis of
or ballistic patterns. stereotyped voluntary movements in man. J Neurol
Neurosurg Psychiatry 38:1154-1162, 1975.
2. Milanov I. Electromyographic differentiation of
tremors. Clin Neurophysiol 112:1626-1632, 2001.
SURFACE 3. Deuschl G, Bain P, Brin M. Consensus statement of
ELECTROMYOGRAPHIC the Movement Disorder Society on Tremor. Ad Hoc
Scientific Committee. Mov Disord 13 (Suppl 3):2-
STUDY OF VOLUNTARY 23, 1998.
MOVEMENT DISORDERS 4. Deuschl G, Krack P, Lauk M, Timmer J. Clinical
neurophysiology of tremor. J Clin Neurophysiol
Techniques are available for recording vol- 13:110-121, 1996.
5. Hallett M. Overview of human tremor physiology.
untary movement by surface EMG analysis. Mov Disord 13 (Suppl 3):43-48, 1998.
True quantitative evaluation in this realm re- 6. Elble RJ. Essential tremor frequency decreases with
quires additional equipment to record time. Neurology 55:1547-1551, 2000.
movement position or velocity. Qualitative 7. Sabra AF, Hallett M. Action tremor with alternat-
ing activity in antagonist muscles. Neurology
evaluation is helpful occasionally but must 34:151-156, 1984.
be interpreted with caution. Such evaluation 8. Milanov I. Clinical and electromyographic exami-
can be performed by having the patient nations of patients with essential tremor. Can J Neu-
make a short ballistic movement such as el- rol Sci 27:65-70, 2000.
bow flexion while recordings are made from 9. Vaillancourt DE, Slifkin AB, Newell KM. Regularity
of force tremor in Parkinson's disease. Clin Neu-
agonist and antagonist muscles. Under most rophysiol 112:1594-1603, 2001.
conditions when this movement is per- 10. Heilman KM. Orthostatic tremor. Arch Neurol
formed as fast as possible, the triphasic pat- 41:880-881, 1984.
tern appears and its configuration can be ex- 11. Fahn S, Marsden CD, Van Woert MH. Definition
and classification of myoclonus. Adv Neurol 43:1-5,
amined. Abnormalities have been reported 1986.
in various diseases. The ballistic movements 12. Caviness JN. Myoclonus. Mayo Clin Proc 71:679-
in Parkinson's disease are characterized by 688, 1996.
low amplitude of the initial agonist burst. 13. Halliday AM. The electrophysiological study of my-
This results in a small amplitude movement oclonus in man. Brain 90:241-284, 1967.
14. Hallett M. Myoclonus: relation to epilepsy. Epilep-
that the patient compensates for by making sia 26 (Suppl 1):S67-S77, 1985.
sequential small triphasic bursts. The move- 15. Ikeda A, Kakigi R, Funai N, Neshige R, Kuroda Y,
ments of cerebellar hypermetria have been Shibasaki H. Cortical tremor: a variant of cortical
ascribed to a delayed onset of the antagonist reflex myoclonus. Neurology 40:1561-1565, 1990.
16. Hallett M, Chadwick D, Adam J, Marsden CD. Retic-
burst28'29 or a normal onset but slow rise of ular reflex myoclonus: a physiological type of hu-
antagonist activity.30 With pyramidal lesions, man post-hypoxic myoclonus. J Neurol Neurosurg
the first agonist or antagonist burst is pro- Psychiatry 40:253-264, 1977.
17. Chadwick D, Hallett M, Harris R, Jenner P, movements in sleep (nocturnal myoclonus): rela-
Reynolds EH, Marsden CD. Clinical, biochemical, tion to sleep disorders. Ann Neurol 8:416-421,1980.
and physiological features distinguishing myoc- 25. Thompson PD, Colebatch JG, Brown P, et al: Vol-
lonus responsive to 5-hydroxytryptophan, trypto- untary stimulus-sensitive jerks and jumps mimick-
phan with a monoamine oxidase inhibitor, and ing myoclonus or pathological startle syndromes.
clonazepam. Brain 100:455-487, 1977. Mov Disord 7:257-262, 1992.
18. JankovicJ, Pardo R. Segmental myoclonus. Clinical 26. Deuschl G, Heinen F, Kleedorfer B, Wagner M,
and pharmacologic study. Arch Neurol 43:1025- Lucking CH, Poewe W. Clinical and polymyo-
1031, 1986. graphic investigation of spasmodic torticollis. J Neu-
19. Davis JN. An experimental study of hiccup. Brain rol 239:9-15, 1992.
93:851-872, 1970. 27. Van Gerpen JA, Matsumoto JY, Ahlskog JE,
20. Hallett M, Chadwick D, Marsden CD. Ballistic move- Maraganore DM, McManis PG. Utility of an EMG
ment overflow myoclonus: a form of essential my- mapping study in treating cervical dystonia. Muscle
oclonus. Brain 100:299-312, 1977. Nerve 23:1752-1756, 2000.
21. Wilkins DE, Hallett M, Wess MM. Audiogenic star- 28. Flament D, Hore J. Movement and electromyo-
de reflex of man and its relationship to starde syn- graphic disorders associated with cerebellar dys-
dromes. A review. Brain 109:561-573, 1986. metria. J Neurophysiol 55:1221-1233, 1986.
22. Tijssen MA, Shiang R, van Deutekom J, et al: Mo- 29. Gilman S, Bloedel J, Lechtenberg R. Disorders of
lecular genetic reevaluation of the Dutch hyper- the Cerebellum. Contemporary Neurology Series.
ekplexia family. Arch Neurol 52:578-582, 1995. Vol. 21. FA Davis Company, Philadelphia, 1981.
23. Brown P, Thompson PD, Rothwell JC, Day BL, 30. Man to M, Godaux E, JacquyJ, HildebrandJ. Cere-
Marsden CD. Axial myoclonus of propriospinal ori- bellar hypermetria associated with a selective de-
gin. Brain 114:197-214, 1991. crease in the rate of rise of antagonist activity. Ann
24. Coleman RM, Pollak CP, Weitzman ED. Periodic Neurol 39:271-274, 1996.
Chapter 32
VERTIGO AND BALANCE
Robert H. Brey
OFFICE PROCEDURES AND HISTORY Positional Testing Without Fixation

TAKING Caloric Irrigation
Descriptions of Symptoms OTOLITH ORGAN TESTING
Medications or Drugs SUPERIOR SEMICIRCULAR CANAL
PHYSICAL EXAMINATION OF THE DEHISCENCE
VESTIBULAR SYSTEM COMPUTERIZED ROTARY CHAIR TEST
Nystagmus (HARMONIC ACCELERATION
Vestibulospinal Reflexes TESTING)
LABORATORY EXAMINATION: COMPUTERIZED DYNAMIC
ELECTRONYSTAGMOGRAPHY TEST POSTUROGRAPHY
BATTERY Motor Control Test
Preparation for Testing Sensory Organization Test
Gaze Testing VESTIBULAR REHABILITATION
Oculomotor Testing SUMMARY
Balance is a complex mechanism that relies tigo (that is, the false sense of motion), nau-
primarily on input from the visual, soma- sea, and vomiting. These symptoms may
tosensory, and vestibular systems. Dizziness continue from seconds to days. The three
or disruption of balance occurs when the categories for causes of vertigo are (1) pe-
information sent to the central nervous sys ripheral, including the peripheral vestibu-
tem (CNS) by one of these systems conflicts lar sensory mechanisms and nerve fibers
with information provided by the other two leading to the brain stem; (2) CNS, includ-
systems. An example is the false sensation ing all the brain stem vestibular nuclei and
of movement experienced by a person sit- their connections with the visual and so-
ting in an automobile when a large vehicle matosensory systems and the cerebral cor-
parked alongside begins slowly to pull for- tex; and (3) systemic, including vascular
ward. The visual input to the person is con- problems.
sistent with the sensation of the car rolling The vestibular system has two types of sen-
backward; the reflexive action is to step on sory organs: the semicircular canals and the
the brakes. The instant it is realized that the otolith organs. These sensory organs contain
car did not come to a lurching stop, the hair cells that respond either to relative
CNS begins to resolve the sensory conflict fluid motion in the membranous labyrinth
by determining that it was the larger vehi- (caused by rotating the head) or to gravita-
cle that was moving and that the visual in- tional force exerted on calcium carbonate
formation was incorrect. If conflicting in- crystals, called otoconia (moving the head lin
formation is continually fed to the CNS early or tilting it). The anatomy and physi-
because of vestibular malfunction, the per- ology of the vestibular mechanism have been
son experiences the classic symptoms of ver- reviewed recently.1
413
OFFICE PROCEDURES AND ease (endolymphatic hydrops), labyrinthitis

HISTORY TAKING (viral or bacterial infection), vestibular neu-
ritis, perilymphatic fistula, and autoimmune
A complete history and description of the disease of the inner ear. Although tumors
patient's symptoms are critical for a work-up are another peripheral lesion that can affect
leading to the diagnosis of vertigo. The ar- the vestibular branch of cranial nerve (CN)
eas to be covered are (1) description of the VIII, those patients often have little or no
vertiginous symptoms, (2) their onset and symptoms of dizziness because the tumor
duration, (3) concomitant aural symptoms, grows so slowly that the CNS is able to com-
(4) visual symptoms, (5) general neurologic pensate physiologically for the slow changes
symptoms, (6) medical history, and (7) cur- caused by tumor growth. Head trauma and
rent and previous medications, including benign paroxysmal positional vertigo are not
use of drugs such as alcohol, nicotine, and disease processes, but they are related to me-
caffeine. chanical problems in the vestibular mecha-
nism that can cause vertigo.
Descriptions of Symptoms SYMPTOMS CAUSED BY

CENTRAL LESIONS
SYMPTOMS CAUSED BY
PERIPHERAL LESIONS Patients with CNS vestibular disorders most
often complain of a slow onset of symptoms
Patients who have an acute peripheral that worsen over time. These patients tend
vestibular attack usually can describe in de- to complain of unsteadiness, lightheaded-
tail what happened. Most patients experi- ness, nausea, and other neurologic symp-
ence a spinning sensation (either of the toms. Exceptions to this are patients who
environment or of themselves), nausea, vom- have an acute onset of symptoms because of
iting, disorientation or loss of control, and a stroke that damages the vestibular nuclei.
sweating, followed by extreme fatigue. Some Examples of CNS diseases that cause vestibu-
patients also describe a vertical roll or tilt of lar symptoms are multiple sclerosis, verte-
the whole environment. Others describe be- brobasilar insufficiency, migraine headache,
ing thrown to the floor or feeling that they compression of CN VIII by a vascular loop,
are being bounced against the walls. During cervical disk problems, head trauma, and
the acute attack, any movement of the head Arnold-Chiari malformation.
aggravates the symptoms. Such patients are
often hospitalized for 1 or 2 days, mainly be- SYSTEMIC SYMPTOMS
cause of concern that they are having a
stroke. Symptoms of patients with systemic vestibu-
Visual symptoms can help define a pa- lar problems tend to be similar to those of
tient's problems. With a unilateral periph- patients with CNS problems and can be
eral loss of vestibular function, blurring of caused by involvement of all three compo-
vision or retinal slip occurs during head nents: vestibular, visual, and somatosensory.
movement because the vestibular ocular reflex Examples of systemic problems are ortho-
(VOR) is unable to keep the visual image static problems, diabetes mellitus, hypogly-
centered on the macula of the retina, where cemia, hypothyroidism, drug effects, stress,
the image is sharpest. Bilateral loss of the pe- and allergies.
ripheral vestibular mechanism produces os-
cillopsia, in which the entire visual environ- CONCOMITANT AURAL SYMPTOMS
ment appears to oscillate as the patient's
head moves; this is because the VOR is un- The auditory and vestibular mechanisms
able to provide the necessary compensatory share the membranous labyrinth of the in-
eye movements to keep the image stable on ner ear and its fluid system. To determine
the retina. whether there may be a relationship be-
Diseases that can affect the peripheral tween aural and vestibular symptoms, it is
vestibular mechanism include Meniere's dis- critical to ask patients about tinnitus, aural
Vertigo and Balance 415
fullness, pressure in the ears or head, fluc- vertical), and congenital nystagmus. Congeni-
tuating hearing loss, or changes in the abil- tal nystagmus, which can incorporate both
ity to understand speech. For example, hy- pendular and jerk components, exhibits a
drops, or Meniere's disease, can cause inner null point where the patient's nystagmus is
ear pressure, tinnitus, vertigo, and a fluctu- diminished.
ating low-frequency sensorineural hearing The types of nystagmus commonly seen
loss. Acoustic tumors of the auditory nerve during vestibular attacks are horizontal, ver-
often cause tinnitus, hearing loss, and prob- tical, and torsional nystagmus. The corneo-
lems with understanding speech. Previous retinal potential (approximately 1 mV, with
ear surgery could be related to a perilym- the front of the eye positive relative to the
phatic fistula that could lead to hearing loss back of the eye) allows horizontal and verti-
and vertigo. cal eye movements to be measured with elec-
trodes placed around the eyes. Torsional
nystagmus can be recorded only with some
Medications or Drugs type of video monitor or magnetic search
coil placed over the lens of the eye, like a
Patients must be asked about any ototoxic contact lens. The traditional method for
medications they might have taken. Amino- plotting horizontal and vertical nystagmus
glycosides, salicylates, quinine, and chemo- over time is shown in Figure 32-1. For hor-
therapeutic agents such as cisplatin can izontal and vertical eye movements, pen up
directly affect and, in some cases, perma- defines the movements that are up or to the
nently damage the hearing and vestibular right and pen down defines those that are
mechanisms. down or to the left. The chart is read from
left to right. Nystagmus direction is defined
by the fast, or jerk, component, which is gen-
erated by the CNS. The slower component
PHYSICAL EXAMINATION OF is in the opposite direction and is generated
THE VESTIBULAR SYSTEM by the vestibular tracking mechanism. Gen-
erally, the slope (rise/run) of the slow com-
The symptoms of a patient examined during ponent is used to calculate, in degrees per
an acute attack of vertigo are different from second, the speed of the eye movement.
those revealed by a later examination. How- During the fast component, vision is in-
ever, both examinations can provide valu- terrupted. Thus, rather than sensing a
able information for making a diagnosis. jerking movement, one has a sensation of
During an acute peripheral vestibular attack, smooth rotation in one direction. This
nystagmus may be observed even when the means that if a person experiences a spin-
patient's eyes are open and fixed. However, ning sensation to the right, the fast compo-
at a later time, fixation may suppress the nys- nent of the nystagmus is to the right or in
tagmus. In this case, electro-oculographic the direction in which the head seems to be
examination (electric measurement of the turning. If the room or environment seems
corneoretinal potential) with the patient's to be spinning around the person to the
eyes closed is needed to observe the abnor- right, then the fast component of the nys-
mal eye movements. Frenzel lenses used with tagmus is to the left.
the patient in the dark with eyes open also
allow nystagmus to be observed. With cen-
tral vestibular lesions, nystagmus tends to be OBSERVATION OF EYE MOVEMENT
present or even more intense when the pa- Any abnormal eye movements detected dur-
tient's eyes are open and fixed. ing examination of a patient should be
noted. The term spontaneous nystagmus refers
to nystagmus present with the eyes open and
Nystagmus fixed in the primary position. In this case,
the examiner must look closely at the pa-
The many types of nystagmus include pen- tient's eyes to observe any small rhythmic
dular, rotatory (torsional), jerk (horizontal and movements.
Figure 32-1. Electro-oculographic recordings of horizontal and vertical nystagmus. A, Right-beating horizontal nys-
tagmus and up-beating vertical nystagmus. B, Left-beating horizontal nystagmus and down-beating vertical nystag-
mus. On the latter tracing is an example of how the velocity (slope, rise/run) of the slow component is measured
as 10° per second.
Gaze testing is performed with the patient's is referred to as spontaneous nystagmus with-
eyes in the primary gaze position and ± 30° out fixation (or with eyes closed), because it
from midline, both horizontally and verti- is present only when visual fixation is pre-
cally. It is helpful to observe a small vessel vented. The general rule is that nystagmus
on the side of the eye rather than to watch caused by a peripheral lesion is suppressed by vi-
the pupil. This allows detection of low-am- sual fixation, but nystagmus caused by a CNS
plitude (< 1°) horizontal, vertical, or tor- lesion is not.
sional nystagmus. Electro-oculography is ca- Any abnormal eye movements should be
pable of measuring only eye movements that documented on a chart similar to the one
are greater than 1°. If the center of the eye shown in Figure 32-2; lines and arrows are
is observed, torsional nystagmus may be used to indicate the direction and magni-
overlooked. Patients with either a CNS tude of eye movement. The concept of the
vestibular abnormality or an acute periph- Alexander law that nystagmus beats stronger
eral attack show abnormal nystagmus with when gaze is in the direction of the fast com-
their eyes open and fixed. However, if a pe- ponent of the nystagmus is illustrated in Fig-
ripheral lesion is not at an acute stage, the ure 32-2A. Alexander would define the re-
CNS likely suppresses any nystagmus. To sponse shown as a Brd-degree nystagmus (that
measure the eye movements in these cases, is, nystagmus present in all positions of gaze
Frenzel lenses, which prevent patients from and greater toward the unaffected ear). As
visually fixating, or electro-oculography the CNS compensates physiologically for the
must be used while the patient's eyes are deficit caused by the peripheral lesion, the
closed (or while the patients are in a dark nystagmus progresses to the 2nd degree (that
room). Nystagmus observed in this manner is, nystagmus with gaze away from the lesion
Figure 32-2. A-F, Method for doc-

umenting abnormal eye move-
ments. Arrows indicate the direc-
tion of the quick eye movement,
and the length of the line repre-
sents the subjective amplitude of
the movement. BPPV, benign
paroxysmal positioning vertigo;
CCW, counterclockwise; CNS, cen
tral nervous system; CW, clockwise
and midline) and finally to the 1st degree The Dix-Hallpike test is illustrated in Fig-
(that is, nystagmus only with gaze away from ure 32-3. The patient is seated initially with
the side of the lesion). the head turned 45° to the right or the left
and is then placed in the head-hanging right
POSITIONING-INDUCED NYSTAGMUS (or left) position with the eyes open so the
(BENIGN PAROXYSMAL examiner can observe any torsional nystag-
POSITIONING VERTIGO) mus. The classic response consists of a brief
delay and then a burst of torsional nystag-
A positioning maneuver to test for benign mus that lasts several seconds. This maneu-
paroxysmal positioning vertigo (BPPV) is dif- ver duplicates the patient's symptoms. The
ferent from static positional testing (dis- head-hanging right position produces coun-
cussed below in conjunction with the terclockwise nystagmus, whereas the head-
electronystagmography [ENG] test battery). hanging left position causes clockwise nys-
Dynamic positioning testing is called the Dix- tagmus (Fig. 32-2). When the patient sits
Hallpike test,2 or Nylen maneuver. The patient up, the nystagmus is reversed but less in-
is taken from a sitting position to a head- tense. The response weakens with repeated
hanging right (or left) position. trials.
Figure 32-3. Dix-Hallpike maneuver (Nylen maneuver) for the head-hanging right position. A, In the sitting posi-
tion, the patient turns the head 45° to the right with the eyes open and fixed. The patient is then put quickly into
the supine position. B, The patient in the head-hanging right position with eyes open and fixed. Observe for nys-
tagmus and dizziness for at least 30 seconds. The classic positive response is counterclockwise nystagmus with the
head down and turned to the right and clockwise nystagmus with the head down and turned to the left. The nys-
tagmus is reversed with sitting up but is less intense.
Early surgical treatment for BPPV was performed, because there are newer, less in-
neurectomy of the vestibular branch of CN vasive approaches. Other approaches that
VIII; this innervated the ampulla of the af- were used were exercises designed to habit-
fected posterior semicircular canal. Another uate the CNS.4
approach has been to occlude the posterior Epley5 postulated the existence of floating
semicircular canal to restrict endolymph particles in the posterior semicircular canal
flow.3 Currently, these procedures are rarely and proposed a physical maneuver, the
canalith repositioning procedure, to remove identify a past-pointing error because of the

diem. Several investigators have reported ex- many proprioceptive clues available to the
cellent success rates with this procedure.5"7 patient.
In 1991, Parnes and McClure3 discovered
these free-floating particles during surgery STATIC POSTURE TESTING
in the posterior semicircular canals of pa-
tients with BPPV. In the Romberg test, patients with vestibular
Semont et al. in 1988,8 reported on an- problems have difficulty standing with their
odier procedure, the liberatory maneuver, de- feet together and eyes closed. They tend to
signed to remove debris from the posterior sway toward the affected side, that is, with
semicircular canal. Herdman et al.9 gave a the slow component of nystagmus. In the
good description of this maneuver and made sharpened Romberg test, the patient is asked
some minor modifications. Results of the to stand with the feet aligned in the tandem
canalith repositioning procedure and the heel-to-toe position and with the eyes closed
liberatory maneuver indicated that approxi- and the arms folded against the chest. Most
mately 90% of the patients had either a cure normal subjects younger than 70 years
or significant relief from their symptoms of should be able to stand in this position for
BPPV. These two procedures are the treat- 30 seconds.
ment of choice for most patients with BPPV.
A diird approach involves rotating the pa- DYNAMIC WALKING TESTS
tient's head 360°, keeping the posterior
semicircular canal in an earth vertical plane In the Fukuda step test, a patient walks in place
by using a modified motorized circle bed or with the eyes closed and the arms held par-
rotating chair.10"12 allel, horizontal, and extending forward.
A variant of BPPV involves loose particles The examiner notes the angle of rotation.
in the horizontal semicircular canal.13'14 Again, the direction of rotation, as in the
Treatment for horizontal canal BPPV in- tests above, is that of the slow component of
volves a procedure in which a supine patient nystagmus. The tandem walking test is another
is rotated 360° to the right or left, in the di- often-used test for neurologic evaluation. If
rection away from the affected ear. a patient performs this test with his or her
Herdman15 has provided an excellent eyes open, it tests primarily cerebellar func-
flow-chart describing die various types of tion. However, a patient with an acute
BPPV and their treatments, including the vestibular lesion may also have difficulty per-
procedures described by Epley and Semont forming the test. For vestibular testing, the
and adaptation exercises by Brandt and patient is blindfolded or asked to walk in tan-
Daroff. dem fashion with the eyes closed and the
arms folded against the chest. The test be-
gins with the patient's feet in a tandem po-
Vestibulospinal Reflexes sition; the patient then takes 10 tandem
steps at a comfortable speed. The number
PAST-POINTING TEST of normal steps (without sidestepping) is
scored on three trials. Most normal subjects
In the past-pointing test, a patient is asked to are able to make a minimum of 10 normal
sit with the eyes closed and an arm extended, steps in the three trials. However, patients
pointing first above the head and then back with acute or chronic vestibular dysfunction
to an imaginary spot. The examiner's finger are unable to perform the test and fall, but
serves as a reference to determine how much the direction of fall is not related to the fast
deviation occurs. Barany found that point- or slow component of nystagmus.
ing tended to drift toward the ear with a pe-
ripheral lesion (that is, the direction of the HEAD-SHAKING NYSTAGMUS
slow component of nystagmus). However, he
indicated that this test could not be used in Hain, Fetter, and Zee17 reported diat spon-
isolation. Baloh and Honrubia16 pointed out taneous nystagmus develops in patients with
that the standard finger-to-nose test did not peripheral or central vestibular dysfunction
after 10 quick head shakes in the horizontal before undergoing vestibular testing. Pa-
plane. These authors suggested that the tients should also be counseled to refrain
spontaneous nystagmus develops because of from smoking tobacco because nicotine
asymmetrical velocity storage within the constricts blood vessels and, thus, impairs
central vestibulo-ocular pathways. In cases of the blood supply to the vestibular mecha-
unilateral peripheral vestibular lesions, the nism. As a stimulant, caffeine can adversely
slow component is toward the weaker side. affect the vestibular system. Also, patients
should avoid drinking alcohol for at least
24 hours before testing, because alcohol al-
DYNAMIC VISUAL ACUITY ters the chemical balance of the perilymph
A Snellen, or illegible E, eye chart can be and endolymph and induces nystagmus. Geo-
used to measure the VOR. The patient is tropic (beating toward the earth) positional
asked to sit, with head immobile, at the stan- nystagmus can be induced within l/% hour af-
dard distance from the chart and to read the ter alcohol ingestion and continue up to
lowest line possible. The head is then rotated 4 hours. Ageotropic (beating away from the
back and forth at 2 Hz. If visual acuity shifts earth) positional nystagmus can be observed
more than one line, the result is abnormal.16 from 5 to 24 hours after alcohol is ingested.
Bilateral vestibular weakness that causes os- The patient should not stop taking anti-
cillopsia produces abnormal test results. A convulsants or medications for vascular reg-
problem with this test is that patients tend ulation before testing. Patients with diabetes
to stop head movement momentarily to read should eat a light meal at least 2 hours be-
the visual symbol. Computer-generated dy- fore testing and should not avoid taking
namic visual acuity tests enhance this pro- their insulin. Leigh and Zee19 provide an in-
cedure by displaying the visual symbol only depth discussion of the effects of drugs on
when the head is moving at a specified ve- eye movement.
locity. This is accomplished with an angular Electrodes are placed at the outer canthus
acceleration device attached to the head to of each eye to measure the horizontal com-
measure head velocity.18 ponent of eye movement, and electrodes
Visual suppression of nystagmus can be placed above and below one eye are used to
measured by placing a patient in a chair that measure vertical movements. A common, or
can be rotated and asking him or her to ex- ground, electrode is placed on the forehead
tend a finger and fixate on it. Abnormal fix- just above the nose. If a patient is blind in
ation suppression (inability to suppress) is one eye, the electrodes must be placed
consistent with CNS dysfunction. around the good eye. If a patient is blind in
both eyes and has no corneoretinal poten-
tial, electro-oculography cannot be used to
measure eye movement. Other procedures
LABORATORY EXAMINATION: for measuring eye movements (for example,
ELECTRONYSTAGMOGRAPHY infrared systems, magnetic search coils, and
TEST BATTERY visual image processing) are beginning to be
used and will permit measurement of hori-
Preparation for Testing zontal, vertical, and torsional nystagmus.
As mentioned above, a carefully docu-

mented history should rule out any preex- Gaze Testing
isting condition (for example, congenital
nystagmus and use of medications or drugs) When gaze testing (described earlier under
that could influence test results. Many med- Observation of Eye Movement) is included
ications, such as vestibular suppressants, as part of an ENG test battery, electro-
sedatives, tranquilizers, antidepressants, and oculography can be used to measure and to
pain relievers, have side effects related to quantify the direction and velocity of the eye
dizziness. To avoid the adverse effects these movement. Gaze testing is performed by hav
medications can have on test results, the pa- ing the patient fixate on spots ± 30° from
tient should stop taking them 24-48 hours center in both the horizontal and the verti-
cal planes. Gaze greater than ± 30° may computerized pursuit tests usually cover a
cause end-point nystagmus, which is normal. range from 0.2 to 0.7 Hz. The cooperation
Remember, however, that visually the hu- of the patient is critical, and several trials
man eye is capable of picking up eye move- may be needed to ensure that the patient is
ments of less than 1°, but with electro- trying his or her best.
oculography the movements must be greater Abnormal test results are consistent with
than 1 ° to be detected and recorded. There CNS dysfunction. The most common ab-
fore, always inspect the patient's eyes during normality is cogwheel pursuit, in which the
gaze testing as part of the protocol. eyes are continually making saccadic move-
ments to catch up with the target. In another
type of abnormality, the visual pursuit keeps
Oculomotor Testing breaking up because the patient is unable to
follow the target consistently.
SACCADIC EYE MOVEMENT TESTING
During calibration for electro-oculography, OPTOKINETIC NYSTAGMUS
the patient looks back and forth ±10°,
which provides the examiner an opportunity Optokinetic nystagmus (OKN) is another test
to look for undershoot or overshoot during of smooth ocular pursuit or CNS function.
saccadic eye movements. If undershooting The stimulus is usually generated as a series
or overshooting occurs consistendy, it may of light and dark vertical bars or a rotating
be related to a CNS vestibular disorder. A sphere with an internal light source and
major improvement for evaluating the abil- many small holes that move from right to
ity of the ocular motor system to produce ac- left or left to right of the patient at 20°, 40°,
curate and timely saccadic eye movements is or 60° per second. Ideally, the entire visual
the computerized random saccade test. In field of the patient should be filled with
addition to allowing examiners to look at these stimuli in a darkened room. Less ac-
undershoot and overshoot, this system pro- ceptable alternatives are small hand-held ro-
duces random signals of varying degrees tating drums with black and white stripes or
over a range of ± 30° by turning on and off a series of LEDs that appear to move across
light-emitting diodes (LEDs) on a light bar. a light bar but are really tracking stimuli, not
The analog signals from the eye movements optokinetic stimuli. Abnormal findings in-
are digitized, and computer algorithms are clude asymmetry between right and left beat-
used to calculate the accuracy, latency, and ing or an inability of the patient to increase
velocity of the eye movements relative to the eye speed appropriately with increased stim-
stimulating signal. The values are compared ulus speed.
with those of normal subjects matched for Another application of OKN is the optoki-
age and sex. Abnormal test findings indicate netic after nystagmus test (OKAN). If a patient
a central vestibular abnormality. Perhaps the observes the OKN stimulus for 1 minute and
most significant abnormality is low eye ve- the room is then darkened (that is, visual fix
locity. Poor cooperation by the patient must ation is not possible), there should be little
be ruled out if latency and accuracy are ab- or no nystagmus. However, if the nystagmus
normal. persists, it is thought to indicate a CNS or
peripheral abnormality.
SMOOTH OCULAR PURSUIT

The smooth ocular pursuit test can be con- Positional Testing Without
ducted by having the patient hold the head Fixation
still and follow a pendulum with the eyes.
However, newer computerized systems pro- The purpose of positional testing without fixa-
duce a range of frequencies, using pendu- tion is to measure eye movement with the pa-
lar-like signals, by turning on and off adja- tient's head held in various static positions,
cent LEDs on the light bar. Ocular pursuit such as sitting, supine, supine head-right,
operates well up to a frequency of 1 Hz. The supine head-left, lateral right (no neck tor-
Figure 32-4. Head positions used in the static positional test. The patient's eyes remain closed, but the patient must
be alert.
sion), lateral left (no neck torsion), head (2) direction-changing nystagmus in two or
hanging, and supine 30° (Fig. 32-4). The more positions. Direction-changing can be
purpose for testing supine at 30° is to have geotropic, ageotropic, or direction-changing
a reference for caloric irrigation (see below). in a single head position.
Visual fixation and mental suppression There is also a continuum that ranges
must be avoided during positional testing. from persistent nystagmus (always present)
This is accomplished by testing in total dark- to intermittent nystagmus (occasional
ness with die patient's eyes open, or by us- beats). During testing, the examiner must
ing Frenzel lenses in a darkened room, or distinguish positioning nystagmus caused by
by having the patient keep his or her eyes the patient's moving into a new position
closed in semidarkness. The last procedure from positional nystagmus, which is present
is die one used most often. The patient must in the static head position. Recordings are
be mentally alert so as not to mentally sup- made for 30-60 seconds.
press die nystagmus. To do this, have the pa- Although the findings are nonlocalizing,
tient carry out a task that cannot be done there are some general rules. Chronic hy-
automatically, for example, counting back- pofunction usually results in nystagmus beat-
wards by 3s or naming states and cities. With ing toward the unaffected or less affected
mental alerting, subjects normally do not ear. With an irritative lesion, as in Meniere's
have nystagmus in any of the positions when disease, nystagmus can beat toward the dis-
their eyes are closed. eased ear. Observance of positional nystag-
Abnormal findings are categorized as (1) mus is also useful in monitoring physiologic
direction-fixed in one or more positions and compensation during die disease process,
because the nystagmus diminishes as com- The theory is that as the stimulus warms
pensation occurs. or cools the bone surrounding the lateral
Direction-changing nystagmus in a single semicircular canal, it induces a convection
head position (excluding positioning) is current in the endolymph that causes utricu-
usually a sign of CNS dysfunction and is re- lopetal or utriculofugal flow. The left lateral
ferred to as periodic alternating nystagmus, semicircular canal at rest and after warm
which changes direction about every 2 (44°C) and cool (30°C) caloric stimulation
minutes. is shown in Figure 32—5. Warm stimuli cause
the nystagmus to beat toward the stimulated
ear, and cool stimuli cause it to beat toward
Caloric Irrigation die nonstimulated ear, thus the acronym
"COWS" (cold opposite, warm same).
Caloric irrigation is performed with the pa- Caloric irrigation is the standard test for
tient's head in a 30° supine position (Fig. determining the laterality of the lesion. How-
32-4), which orients the lateral semicircu- ever, certain pitfalls must be avoided. The
lar canals vertically. The stimulus used is wa- patient must not be allowed to fixate visually
ter (± 7°C relative to normal body tem- but must remain alert mentally. Also, con-
perature) or air (± 13°C relative to normal genital nystagmus or any drugs or medica-
body temperature). The ear is irrigated tion that could influence the results must be
with water for 30-40 seconds or with air for ruled out.
60 seconds. Most examiners prefer the wa- The results in patients with perforated
ter stimulus because it transfers heat more tympanic membranes may be misleading. If
efficiently to the petrous bone surrounding the middle or external ear is wet because of
the vestibular mechanism. However, in infection and drainage, a warm air stimulus
cases of perforated tympanic membrane, initially cools rather than heats the bone and
air is the better stimulus, because it is not the nystagmus beats in the direction pro-
advisable to introduce water into the mid- duced by a cold stimulus. This could be mis-
dle ear. interpreted as a CNS abnormality.
Figure 32-5. A, Lateral semicircular canal at rest (that is, no caloric stimulation or fluid rotation). Kinocilium and
Stereocilia are vertical, producing a normal resting potential in vestibular nerve. B, Stimulation with warm water
(44°C) causes upward convection current (utriculopetal endolymph flow). Stereocilia bend toward kinocilium, de-
polarizing the dendrites at base of hair cell, thus increasing firing rate of the vestibular nerve. C, Stimulation with
cool water (30°C) induces downward convection current (utriculofugal endolymph flow), causing kinocilium to
bend toward Stereocilia, which hyperpolarizes dendrites at base of hair cell, thus decreasing firing rate of the vestibu-
lar nerve, (continued)
Unilateral weakness (UW) is determined ness, 49% weaker on the left side, is shown
by comparing the nystagmus generated on in Figure 32-6.
each side: Another way to analyze the response is to
measure the directional preponderance (DP):
where RW is the right warm-peak slow-com-

ponent velocity, RC is the right cool-peak The difference between the two directions
slow-component velocity, LW is the left must be at least 30% to be considered sig-
warm-peak slow-component velocity, and LC nificant. Directional preponderance is non-
is the left cool-peak slow-component veloc- localizing. It usually accompanies a direc-
ity. The result is a percentage difference be- tion-fixed positional nystagmus, because
tween the values for each ear. In most tes nystagmus sums algebraically.
ing laboratories, the percentage difference Bilateral weakness is another possible re-
must be 20%-25% to be clinically signifi- sult. If the sum of all four irrigations is less
cant. An example of a left peripheral weak- than 28° of peak nystagmus, it is considered
Figure 32-6. Responses to caloric stimuli in a patient with left unilateral vestibular weakness. A, Calculated response
of the nystagmus over time. The small boxes represent peak eye velocity values averaged for each of the four irri-
gations. These responses show a 49% left peripheral weakness and a 30% right-beating directional preponderance.
B, Raw data obtained during peak eye velocities. Note the weak responses obtained by stimulating left ear. SPV,
Slow phase velocity.
bilaterally weak. An example of this is shown maximal velocity. If the patient does not
in Figure 32-7. The total nystagmus gener- suppress the nystagmus by at least 30%-40%,
ated (in degrees) is 0 + 4 + 0 + 3 = 7 per the result is abnormal and indicative of a
second. CNS lesion.
Another test is to measure fixation sup- A summary of ENG test battery results is
pression shortly after the eyes reach their given in Table 32-1.20
Figure 32-7. Responses to caloric stimuli in patient with peripheral vestibular weakness bilaterally. A, Calculated re-
sponse of nystagmus over time. The small boxes represent peak eye velocity values averaged for each of the four
irrigations. These responses show a total of 7° of right-beating nystagmus, which is probably caused by the posi-
tional nystagmus present whenever eyes are closed. B, Raw data obtained during peak eye velocities. Note weak re-
sponses with stimulation of both ears. SPY, slow phase velocity.
426
Table 32-1. Electronystagmography Test Battery: Correlation of Abnormal

Findings and Suspected Site of Lesion*
Test Type of Abnormality Suspected Site of Lesion
Saccade Ipsilateral dysmetria Cerebellopontine angle
Bilateral dysmetria Cerebellum
Decreased velocity Throughout the central nervous
system, muscle weakness or
peripheral nerve palsy
Internuclear ophthalmoplegia Medial longitudinal fasciculus
Pursuit Break-up Brain stem or cerebrum
Saccadic Cerebellum
Gaze Direction-fixed and horizontal Peripheral vestibular
Direction-changing and vertical Brain stem
Up-beating Brain stem or cerebellum
Down-beating Cervicomedullary junction or
cerebellum
Rotary Vestibular nuclei/brain stem
Failure of fixation Less than 40% decrease Brain stem or cerebellum
suppression
Positional Direction-fixed Nonlocalizing or peripheral
Direction-changing Nonlocalizing or central
Dix-Hallpike Classic Peripheral vestibular—undermost ear
Nonclassic Nonlocalizing
Caloric Unilateral or bilateral weakness Peripheral vestibular
Directional preponderance Nonlocalizing
*Exceptions to the rule may occur.
From Cyr.20 By permission of Allyn & Bacon.
OTOLITH ORGAN TESTING traction. Sheykholeslami et al.21 reported

that the best placement for the electrodes is
It generally is accepted that ocular tilt is me- at the center of die muscle. A series of loud
diated in part by die utricle. Thus, a lesion stimuli is presented to one ear, and the myo-
in the utricle causes the patient's head to tilt genic responses of the muscle are averaged
in a compensatory manner to offset the tor- using surface electrodes. The result is a
sion caused by die bias difference between short-latency inhibitory response seen as a
die two utricles. Herdman15 described the P13-N23 evoked potential. This response is
subjective vertical or horizontal line tests (or ipsilateral and occurs in persons with pro-
both). Normal subjects are able to manipu- found sensorineural hearing loss but is ab-
late a horizontal or vertical line widiin ± 2 ° sent in patients who have had neurectomy
of true horizontal or vertical witiiout any vi- of die vestibular branch of CN VIII.
sual reference, as in a darkened room. Sub-
jects with unilateral lesions may be off by as
much as 15°. SUPERIOR SEMICIRCULAR
Herdman15 also described a method for CANAL DEHISCENCE
testing the saccule by using vestibular evoked
potentials. The saccule lies under die foot- Minor22 reported 17 cases of a dehiscence
plate of the stapes, and this proximity makes of the superior semicircular canal identified
die saccule responsive to loud clicks or low- widi high-resolution computed tomography.
frequency tone bursts. The test is conducted The patients had vertigo, oscillopsia, or both
widi the sternocleidomastoid muscle in con- when presented with intense sounds or stim-
uli that produced changes in middle ear or quency increases and approaches 0.64 Hz,
intracranial pressure. These stimuli pro- the phase difference approaches zero. With
duced torsional eye movements commensu- the patient rotating in the dark, the gain (ra-
rate with stimulating the affected canal. With tio of eye velocity to chair velocity) is low at
surgical plugging of the dehiscence, the pa- low frequencies and increases at higher fre-
tients' symptoms resolved. Brantberg et al.23 quencies. The relationships of phase, gain,
also reported that patients complained of and symmetry of chair velocity and eye ve-
"pulse-synchronous tinnitus and gaze insta- locity are shown in Figure 32-8. Normal
bility during periods of upper respiratory in- phase gain and symmetry are shown for a pa-
fections." Testing the saccule provides in- tient in Figure 32-9. The data from a patient
formation about the inferior branch of the with left peripheral vestibular weakness, as
vestibular nerve, whereas caloric testing and indicated by a 59% caloric difference be-
rotary chair testing measure the response in tween the two ears and a 22% right-beating
the superior branch. directional preponderance, are shown in
Figure 32-10. Note that the gain is normal
from 0.01 to 0.32 Hz (0.64 Hz was not
COMPUTERIZED ROTARY tested). Phase is abnormal or borderline ab-
CHAIR TEST (HARMONIC normal from 0.01 to 0.16 Hz. Asymmetry, al-
ACCELERATION TESTING) though within normal range, is slightly be-
low the line, indicating that right-beating is
Another approach for assessing the VOR is greater than left-beating. Often in such pa-
to place the patient in a rotary chair in tients, the gain recovers but phase can re-
a darkened or lighted room and electro- main abnormal, particularly if there is com-
oculographically measure the nystagmus as plete loss of function on one side.
the chair rotates back and forth. This is anal- Patients with total bilateral vestibular
ogous to the torsion swing chair test. The ad- weakness have poor gain at all frequencies
vantages of the rotary chair test are (1) eye and no response to caloric irrigation. When
movements can be quantified with or with- gain is below normal from 0.01 to 0.64 Hz,
out visual stimulation; (2) patients tolerate it phase and symmetry are meaningless be-
better than caloric irrigation; (5) angular ro- cause there is no eye velocity to use for
tation is a more consistent stimulus than comparison.
caloric irrigation, which produces gain, Another approach is the vestibular visual
phase, and symmetry information; (4) small ocular reflex (VVOR) test. The patient, with
children can be tested without difficulty; and eyes open, is rotated in a lighted room. Thus,
(5) it appears to be better for monitoring visual and vestibular clues are available. In
changes over time. Because rotary testing normal subjects, the test produces gain mea-
stimulates both lateral semicircular canals si- surements that approach 1.0 and phase mea-
multaneously, caloric irrigation is still the surements that approach 0°.
primary test for evaluating each ear inde- The step test provides the information nec-
pendently. essary to assess the time constant (that is,
The test used most often is the low-fre- the time it takes for nystagmus to decay to
quency rotary chair test, with the patient kept 37% of its maximum after stimulation has
in total darkness. It consists of accelerating stopped). The patient is quickly accelerated
and decelerating the chair from 0° to 50° from 0° to 60°, 100°, or 240° per second in
per second in a sinusoidal fashion from 0.01 0.5 second. The chair then continues to ro-
Hz to 0.64 Hz. Most systems use an infrared tate at 60°, 100°, or 240° per second for 1
camera to monitor the patient's eyes to en- minute. As the flow of the endolymphatic
sure that they are open. As the chair rotates, fluid in the semicircular canals approaches
the computer digitizes the analog signals the velocity of the head, the nystagmus de-
from the eyes and compares the eye move- cays. Because of the elasticity of the cupula,
ment with the chair rotation. The algorithms the response time for it to bend and to re-
compare the velocity, phase, and gain of the turn to its resting state is approximately 4-7
two signals. At low frequencies (for example, seconds. However, the nystagmus continues
0.01 Hz), normal eye velocity leads chair ve- for 10-30 seconds, which is attributed to cen-
locity by as much as 45°. As the chair fre- tral velocity storage. When the chair is
Figure 32-8. Measurement of phase (A),
gain (B), and symmetry (C) using a
computerized rotary chair. Sine waves
represent fast Fourier analysis of the ve-
locity of the chair and slow phase move-
ment of the eyes, as indicated.
Figure 32-9. Normal rotary chair test results for phase, gain, and symmetry obtained with patient rotating in the
dark. Results in shaded areas are abnormal.
429
Figure 32-10. Results of rotary chair test (conducted in darkness) in patient with right peripheral weakness. Phase
is abnormal, but gain and symmetry are normal. Shaded areas are abnormal.
stopped suddenly (0.5 second), the fluid matosensory information is the dominant in-
continues to move relative to the head but put, followed by visual and vestibular inputs.
in the opposite direction, thus generating The inputs from these three systems are in-
nystagmus in the opposite direction. The tegrated, analyzed, and incorporated into a
time constant during this period is also mea- complex network by the CNS for mainte-
sured. Abnormal findings can be caused by nance of balance. For many years, physicians
either CNS dysfunction or the inability of the have used subjective methods for assessing a
peripheral system to send the appropriate person's ability to maneuver and to maintain
information for integration. balance, with and without vision. Tests such
The consistent stimulus and reliable na- as the Romberg and tandem gait tests are
ture of rotary chair testing make it the best two examples.
choice for monitoring changes in the VOR Computerized dynamicposturography (GDP),
over time. Rotary chair testing is valuable a test for assessing balance, provides quanti-
when monitoring physiologic compensation tative information that can be used to mon-
or change in the vestibular mechanism in- itor the types of problems patients have with
duced by ototoxic medications. The rotary balance deficits. The GDP is designed to pro-
chair also provides the best environment for vide real-life experiences in a controlled
optokinetic testing, because most of the vi- laboratory environment so that the exam-
sual field can be filled with the moving vi- iner can evaluate the patient objectively and
sual stimuli. subjectively.
This test consists of two major compo-
nents, each containing subtests. The first
COMPUTERIZED DYNAMIC component is a test for motor control to
POSTUROGRAPHY maintain balance. The second component is
a test for measuring the patient's use of sen-
Balance is a complex function that requires sory information as it relates to maintaining
input from three major sensory systems. So- balance. On July 20, 1990, the American
Academy of Otolaryngology—Head and

Neck Surgery, Inc.24 ruled that dynamic pos-
turography is an acceptable test procedure
for assessing the balance system.
The patient's anterior-posterior sway is
monitored by measuring vertical force with
four strain gauges that are mounted under-
neath the two platforms on which the pa-
tient stands (one foot on each platform). A
fifth strain gauge, mounted perpendicular
to the other four, measures the shear force
that is generated as the hips are thrown for-
ward or backward to maintain balance in-
stead of rotating about the ankles. The plat-
form can move forward or backward to
produce perturbations of small, medium, or
large magnitude. It can also tilt up or down,
rotating at the axis below the patient's an-
kle. The visual surround can be made to sway
in the same anterior-posterior direction as
the patient. In addition to measuring the pa-
tient's anterior-posterior sway, the data can
be used to produce sway-referenced signals
to drive the visual surround or platform (or
both) to sway according to the input from
the patient. This forces the patient to ignore
or to compensate for the adverse stimula-
tion.
The results of the test are compared with
those from control subjects matched for age.
Much of the analysis is based on research figure 32-11. The cone of stability is 12.5° and is the
that indicates that humans have a cone of amount of sway against which anterior-posterior sway
is judged. (From EquiTest Systems. Installing EquiTest
stability of 12.5° for anterior-posterior sway1 version 4.03. NeuroCom International, Clackamas, OR.
(Fig. 32-11). Thus, a person who sways ap- By permission of NeuroCom International.)
proximately 12.5° is at the limits of stability
and is likely to fall. A fall is scored any time
a patient reaches out and touches the visual
surround to keep from falling or anytime he sponses of the gastrocnemius and anterior
or she moves the feet to keep the center of tibialis muscles. The symmetry and amount
gravity over the base of support. of force and weight distribution are also
measured and displayed with the MCT.
In the test for adaptation, toes up/down, the
Motor Control Test platform tilts up or down 8° to generate a
stimulus analogous to that of walking on un-
With the motor control test (MCT), the patient even surfaces. It is expected that patients will
is presented with forward and backward per- perform poorly on trial 1, but on trials 2-5,
turbations (varying from small to medium to they should adapt and perform normally.
large) of the platform. The computer algo-
rithms calculate the latency of the response
to each perturbation. This value is the la- Sensory Organization Test
tency of the long motor arc loop and serves
as a screening test for problems in that The sensory organization test consists of six dif-
system. The equipment is also capable of ferent conditions, with three trials possible
making electromyographic recordings with for each condition. These conditions, in-
surface electrodes to measure the actual re- cluding the support surface condition (fixed
tions 3 and 6. This has been demonstrated

in patients with Alzheimer's disease. How-
ever, patients with Parkinson's disease have
sensory deficits spread over a broader range
of conditions and not restricted to vision
only.25
One of the most effective uses for GDP is
testing patients who have functional abnor-
malities or at least functional overlays.26'27
Because patients must cooperate to com-
plete the test, they have ample opportunity
to exaggerate their responses. Of interest,
many patients with functional problems per-
Figure 32-12. The six sensory conditions of the sensory form relatively better on the difficult tests,
organization test. (From EquiTest Systems. Installing
EquiTest version 4.03. NeuroCom International,
such as conditions 5 and 6, and poorly on
Clackamas, OR. By permission of NeuroCom Interna- the easier tests, such as conditions 1 and 2.
tional.) Results such as these are physiologically
inconsistent.
or sway-referenced) and the visual condition

(fixed, eyes closed, or sway-referenced) are
summarized in Figure 32-12. Recall that
sway-referenced means that the visual sur-
round or platform is driven by the patient's
anterior-posterior sway and the patient must
either ignore or compensate for the inap-
propriate information. Test results for con-
ditions 1-6 for a normal subject are shown
in Figure 32-13. Scores approaching 100 in-
dicate little sway, whereas those near zero
represent a large amount of sway (relative to
the 12.5° of the cone of stability). If a pa-
tient touches the wall or moves the feet, the
trial is scored as a fall (0 points). Scores
falling into the shaded area are abnormal
(Fig. 32-13A). From these scores, a sensory
analysis is performed (Fig. 32-135). The
scores (ratios) are derived from comparisons
of the various conditions, as shown in Figure
32—14. This information provides a fairly
complete picture of a person's ability to
maintain stability.
Abnormal results typical of patients with
acute vestibular disorders are low scores on
conditions 5 and 6, with an abnormal vestibu- Figure 32-13. A, Sensory organization equilibrium
lar ratio. Patients with bilateral vestibular scores of a normal subject for the six test conditions
and the composite score. The latter is the mean of 14
hypofunction generally fall "like an inverted scores (the mean of conditions 1 and 2, and all 12 tr
pendulum" on conditions 5 and 6 because als of conditions 3-6). N/S, no score (trial was not run
they receive no information from the B, Normal sensory analysis scores. PREF, visual prefer-
vestibular system and their vision and so- ence; SOM, somatosensory; VEST, vestibular; VIS, vi-
sual. Results in shaded areas are abnormal. (From
matosensation are compromised. Patients EquiTest Systems. Installing EquiTest version 4.03.
who are unable to suppress inappropriate vi- NeuroCom International, Clackamas, OR. By permis-
sual information function poorly on condi- sion of NeuroCom International.)
Rgure 32-14. Summary of sensory analysis for six sensory conditions and the significance of their outcomes. See
Figure 32-13 for abbreviations. (From EquiTest Systems. Installing EquiTest version 4.03. NeuroCom International,
Clackamas, OR. By permission of NeuroCom International.)
VESTIBULAR REHABILITATION speed and improve the recovery of many

patients.
Computerized dynamic posturography also
provides quantitative information to help
document the need for and the results of SUMMARY
vestibular rehabilitation. Vestibular rehabil-
itation has been available for many years. Vertigo and imbalance are becoming serious
However, only recently have personalized problems, particularly as the population in-
evaluations and treatments been carefully creases in age. The patient's medical history
developed and shown to substantially im- is critical in determining whether the prob-
prove vestibulospinal compensation.28 These lem is peripheral, central, systemic, or some
programs consist of habituation exercises, combination of these. Aspects of a bedside
postural control exercises, and general con- or office examination for evaluating nystag-
ditioning.29 This relatively new area has be- mus or the status of the vestibular ocular re-
gun to provide a mechanism for helping to flex are described. The physician's informa-
treat the condition of patients who previ- tion can be supplemented by laboratory
ously were told to live with their balance examination, including the electronystag-
problem. We have learned that the vestibu- mography test, computerized rotary chair
lar mechanism is extremely plastic, that it test, a modern torsion swing chair test, and
can be shaped and modified. Appropriate computerized dynamic posturography.30 On
exercises prescribed by a physical therapist the basis of information gathered from all
trained in vestibular rehabilitation can help these tests, physicians can make a reasonable
diagnosis and develop a plan for providing 13. Lempert T, Tiel-Wilck K. A positional maneuver for
relief from the symptoms. Treatment may treatment of horizontal-canal benign positional ver-
tigo. Laryngoscope 106:476-478, 1996.
include surgery, medical management, or 14. McClure JA. Horizontal canal BPV. J Otolaryngol
monitoring the patient's progress. As part of 14:30-35, 1985.
medical management, vestibular rehabilita- 15. Herdman SJ. Vestibular Rehabilitation, 2nd ed. FA
tion can help patients to recover from their Davis, Philadelphia, 2000.
16. Baloh RW, Honrubia V. Clinical Neurophysiology
symptoms. of the Vestibular System, 2nd ed. FA Davis Com-
pany, 1990.
17. Hain TC, Fetter M, Zee DS. Head-shaking nystag-
mus in patients with unilateral peripheral vestibu-
REFERENCES lar lesions. Am J Otolaryngol 8:36-47, 1987.
18. Herdman SJ, Tusa RJ, Blatt P, Suzuki A, Venuto PJ,
1. Jacobson GP, Newman CW, KartushJM. Handbook Roberts D. Computerized dynamic visual acuity test
of Balance Function Testing. Mosby Year Book, St in the assessment of vestibular deficits. Am J Otol
Louis, 1993, p 24. 19:790-796, 1998.
2. Dix MR, Hallpike CS. The pathology, symptoma- 19. Leigh RJ, Zee DS. The Neurology of Eye Move-
tology and diagnosis of certain common disorders ments, 2nd ed. rev. FA Davis Company, Philadel-
of the vestibular system. Ann Otol Rhinol Laryngol phia, 1991.
61:987-1016, 1952. 20. Cyr DG. Vestibular system assessment. In Rintel-
3. Fames LS, McClure JA. Free-floating endolymph mann WF (ed). Hearing Assessment, 2nd ed. Pro-
particles: a new operative finding during posterior ed, Austin, Texas, 1991, p 777.
semicircular canal occlusion. Laryngoscope 102: 21. Sheykholeslami K, Murofushi T, Kaga K. The effect
988-992, 1992. of sternocleidomastoid electrode location on
4. Herdman SJ. Role of vestibular adaptation in vestibular evoked myogenic potential. Auris Nasus
vestibular rehabilitation. Otolaryngol Head Neck Larynx 28:41-43, 2001.
Surg 119:49-54, 1998. 22. Minor LB. Superior canal dehiscence syndrome.
5. Epley JM. The canalith repositioning procedure: Am J Otol 21:9-19, 2000.
for treatment of benign paroxysmal positional ver- 23. Brantberg K, Bergenius J, Mendel L, Witt H,
tigo. Otolaryngol Head Neck Surg 107:399-404, Tribukait A, Ygge J. Symptoms, findings and treat-
1992. ment in patients with dehiscence of the superior
6. Lynn S, Pool A, Rose D, Brey R, Suman V. Ran- semicircular canal. Acta Otolaryngol 121:68-75,
domized trial of the canalith repositioning proce- 2001.
dure. Otolaryngol Head Neck Surg 113:712-720, 24. Monsell EM, FurmanJM, Herdman SJ, Konrad HR,
1995. Shepard NT. Computerized dynamic platform pos-
7. Weider DJ, Ryder CJ, Stram JR. Benign paroxysmal turography. Otolaryngol Head Neck Surg 117:394-
positional vertigo: analysis of 44 cases treated by the 398, 1997.
canalith repositioning procedure of Epley. Am J 25. Chong RK, Horak FB, Frank J, Kaye J. Sensory or-
Otol 15:321-326, 1994. ganization for balance: specific deficits in Alzhei-
8. Semont A, Freyss G, Vitte E. Curing the BPPV with mer's but not in Parkinson's disease. J Gerontol A
a liberatory maneuver. Adv Otorhinolaryngol 42: Biol Sci Med Sci 54:M122-M128, 1999.
290-293, 1988. 26. Cevette MJ, Puetz B, Marion MS, Wertz ML,
9. Herdman SJ, Tusa RJ, Zee DS, Proctor LR, Mattox Muenter MD. A physiologic performance on dy-
DE. Single treatment approaches to benign parox- namic posturography. Otolaryngol Head Neck Surg
ysmal positional vertigo. Arch Otolaryngol Head 112:676-688, 1995.
Neck Surg 119:450-454, 1993. 27. Krempl GA, Dobie RA. Evaluation of posturogra-
10. Brey RH, Lynn S, Trine MB. Diagnosis and reme- phy in the detection of malingering subjects. Am J
diation of benign paroxysmal positioning vertigo. Otol 19:619-627, 1998.
Proceedings of the XXI Congress of the Neuroto- 28. Strupp M, Arbusow V, Maag KP, Gall C, Brandt T.
logical and Equilibriometric Society, Bad Kissingen, Vestibular exercises improve central vestibulospinal
Germany, March, 1994. compensation after vestibular neuritis. Neurology
11. FurmanJM, Cass SP, Briggs BC. Treatment of be- 51:838-844, 1998.
nign positional vertigo using heels-over-head rota- 29. Telian SA, Shepard NT. Update on vestibular re-
tion. Ann Otol Rhinol Laryngol 107:1046-1053, habilitation therapy. Otolaryngol Clin North Am
1998. 29:359-371, 1996.
12. Lempert T, Wolsley C, Davies R, Gresty MA, Bron- 30. Fife TD, Tusa RJ, Furman JM, et al. Assessment:
stein AM. Three hundred sixty-degree rotation of vestibular testing techniques in adults and children:
the posterior semicircular canal for treatment of be- Report of the Therapeutics and Technology As-
nign positional vertigo: a placebo-controlled trial. sessment Subcommittee of the American Academy
Neurology 49:729-733, 1997. of Neurology. Neurology 55:1431-1441, 2000.
SECTION 2
Electrophysiologic
Assessment of
Neural Function
Part F
Autonomic
Function
The autonomic nervous system regulates visceral function and the internal
environment of the body through its effects on the heart, gut and other in-
ternal organs, peripheral blood vessels, and sweat glands (Chapter 33).
Neural activity in the autonomic nervous system is difficult to record directly.
Although sympathetic nerve function in peripheral nerves can be recorded
with fine-tipped tungsten electrodes, this technique generally is difficult to
apply clinically. Therefore, the assessment of autonomic function depends
primarily on measuring the response of the autonomic nervous system to
external stimuli.
The measurements of sweating (Chapter 34), cardiovascular activity and pe-
ripheral blood flow (Chapters 35 and 37), and central autonomic-mediated
reflexes (Chapter 36) provide insight into the broad range of disorders that
affect the central and peripheral components of the autonomic nervous sys
tem—from the hypothalamus to the autonomic axons in the trunk and limbs.
Autonomic function is not measured as frequendy as it should be. With bet-
ter understanding of the importance of the measurements of autonomic
function and with the new tests of cardiovagal function, segmental sympa-
thetic reflexes, postural normotension, and power spectral analysis, the tests
and measurements of autonomic function will be of greater benefit in pa-
tient care.
Pain is mediated mainly through small nerve fibers, particularly in the au-
tonomic nervous system. Measurements of their function can help elucidate
the mechanisms underlying pain, especially peripheral pain. The emerging
modalities for assessment of pain pathways include quantitative sensory tests,
autonomic tests, microneurography, and laser-evoked potentials (Chapter
38). Direct recording of spontaneous electric activity in nerves by micro-
neurography is tedious but particularly helpful.
Chapter 33
CLINICAL PHYSIOLOGY OF THE
AUTONOMIC NERVOUS SYSTEM
SYMPTOMS AND DISEASES MUSCLE SYMPATHETIC ACTIVITY

GENERAL ORGANIZATION OF THE AUTONOMIC CONTROL OF HEART RATE
AUTONOMIC SYSTEM CARDIOVASCULAR REFLEXES
Visceral Afferents Arterial Baroreflexes
Visceral Efferents Cardiopulmonary Reflexes
SYMPATHETIC FUNCTION Venoarteriolar Reflexes
Functional Anatomy of the Sympathetic Er(£oreflexes
Outflow MAINTENANCE OF POSTURAL
Assessment of Sympathetic Function in NORMOTENSION
Humans SUMMARY
SYMPATHETIC INNERVATION OF THE
SKIN
SYMPTOMS AND DISEASES orders. Isolated autonomic failure may oc-

cur acutely or subacutely (for example, in
The autonomic nervous system consists of inflammatory or paraneoplastic pandys-
sympathetic (thoracolumbar) and parasympa- autonomia), or as a slowly progressive dis-
thetic (craniosacral) divisions. Both divisions ease (pure autonomic failure). Autonomic
contain general visceral afferent and effer- failure frequently occurs in small fiber neu-
ent fibers. The enteric nervous system, lo- ropathies (particularly in diabetes mellitus
cated in the wall of the gut, is considered a and amyloidosis), Guillain-Barre syndrome,
third division of the autonomic nervous sys and immune ganglioneuronopathies, such
tem. Disorders affecting the autonomic ner- as Sjogren's syndrome. Central neurode-
vous system may be manifested by failure or generative disorders associated with auto-
hyperactivity of one or many of the visceral nomic failure include multiple system atro-
effector organs. Manifestations of autonomic phy (MSA) and Parkinson's disease.
failure include orthostatic hypotension, heat Autonomic testing is indicated in patients
intolerance as a result of sudomotor failure, (1) with any of the aforementioned symp-
Homer's syndrome or Adie's pupil, gas- toms suggesting autonomic failure, (2) with
trointestinal dysmotility, hypotonic bladder, peripheral neuropathies (particularly small
or erectile failure. Generally, autonomic fiber neuropathies in which electromyo-
failure occurs as the only manifestation of graphic findings may be normal), or (3) with
disease or in association with peripheral neu- parkinsonism or cerebellar dysfunction in
ropathy or central degenerative motor dis- which MSA is suspected. Unlike the function
437
438 Autonomic Function
of somatic motor or sensory nerves, auto- Visceral Efferents

nomic nerve function is difficult to evaluate
precisely in humans. In general, evaluation Sympathetic and parasynipathetic auto-
of autonomic function has been restricted to nomic outflow systems produce responses
noninvasive recordings of heart rate, blood that are longer in latency and duration and
pressure, blood flow, or sweat production. more continuous and generalized than
The interpretation of the results of these those mediated by the somatic motor system.
tests may be difficult, because (1) the effec- This reflects important differences in the
tor organs react slowly to variations in neural functional organization of autonomic and
input, (2) the interactions of sympathetic somatic efferents. Autonomic output in-
and parasynipathetic outputs at a single tar- volves a two-neuron pathway that has at least
get level are complex, and (5) autonomic re- one synapse in an autonomic ganglion.4 The
sponses are affected by hormonal, local preganglionic axons are small myelinated
chemical, and mechanical influences. (type B) cholinergic fibers that emerge from
This chapter provides an overview of some neurons of the general visceral efferent col-
aspects of autonomic function that may help umn in the brain stem or spinal cord and
with interpreting the results of noninvasive pass through peripheral nerves to the auto-
autonomic tests commonly used clinically. nomic ganglia.5 Postganglionic autonomic
Therefore, the focus is on sudomotor and neurons send varicose, unmyelinated (type
cardiovascular functions. Gastrointestinal, C) axons to innervate peripheral organs. Au-
bladder, and sexual functions are not dis- tonomic terminals contain varicosities. The
cussed. primary postganglionic neurotransmitter at
most sympathetic neuroeffector junctions is
norepinephrine, which acts on different
GENERAL ORGANIZATION OF subtypes of a- and /3-adrenergic receptors.
THE AUTONOMIC SYSTEM In the sweat glands, sympathetic effects
are mediated by acetylcholine. The pri-
Visceral Afferents mary postganglionic neurotransmitter at all
parasynipathetic neuroeffector junctions is
Visceral receptors generally are slowly adapt- acetylcholine, which acts on different sub-
ing mechanoreceptors or chemoreceptors types of muscarinic receptors. Whereas the
that have a low level of spontaneous activity main consequence of denervation in striated
and are innervated by small myelinated and muscle is paralysis and atrophy, postgan-
unmyelinated fibers.1 Visceral afferent sig- glionic efferent denervation produces an ex-
nals may mediate local reflexes in periph- aggerated response of the target when it is
eral organs or provide collateral input to exposed to the neurotransmitter. This phe-
autonomic ganglion cells. However, most nomenon, called denervation supersensitivity,
visceral afferent fibers enter the central ner- is evidence of a lesion involving postgan-
vous system at a spinal or medullary level to glionic neurons.
initiate segmental or suprasegmental vis- Activity of most autonomic effectors is
cerovisceral reflexes or to relay visceral in- modulated by dual, continuous sympathetic
formation to higher centers.1"3 Spinal sym- and parasynipathetic controls.4'5 Sympa-
pathetic afferents mediate visceral pain and thetic control originates from preganglionic
initiate segmental and suprasegmental vis- neurons in segments T2-L1 of the spinal
cerovisceral reflexes. Visceral afferents that cord and predominates in blood vessels,
enter at the medullary level are in the vagus sweat glands, and cardiac muscle. Parasyni-
(cranial nerve [CN] X) and glossopharyn- pathetic outflow originates from pregan-
geal (CN IX) nerves and synapse in the nu- glionic neurons in nuclei of CNs III, VII, IX,
cleus of the tractus solitarius (NTS).2'3 This and X and in segments S2-S4 of the spinal
nucleus is critically involved in respiratory cord. It predominates in control of the sali-
and cardiovascular reflexes through its ex- vary glands, sinoatrial node, gastrointestinal
tensive connections with neurons in the so- tract, and bladder.4'5 Sympathetic-parasym-
called intermediate reticular zone of the pathetic interactions are not simply antago-
medulla.3'4 nistic but are functionally complementary.
Clinical Physiology of the Autonomic Nervous System 439
The sympathetic and parasympathetic sys- (somatic) nerve through the gray ramus
tems may interact at several levels, including communicans. These fibers provide vaso-
that of the central nervous system, auto- motor, sudomotor, and pilomotor input
nomic ganglia, neuroeffector junction, and to the extremities and trunk. Sympathetic
target organ.1"5 fibers are intermingled with somatic motor
and sensory fibers, and their distribution is
similar to that of the corresponding somatic
SYMPATHETIC FUNCTION nerve. Most sympathetic fibers are destined
for the hand and foot and are carried mainly
Functional Anatomy of the by the median, peroneal, and tibial nerves
Sympathetic Outflow and, to a lesser extent, the ulnar nerve.
Preganglionic sympathetic neurons have a SYMPATHETIC REFLEXES

slow, irregular, tonic activity that depends
mainly on multiple segmental afferent and Unlike somatic motor neurons, sympathetic
descending inputs.6'7 Preganglionic sympa- preganglionic neurons are not monosynap-
thetic neurons are organized into different tically driven by visceral, muscle, or cuta-
spinal sympathetic functional units that con- neous sensory input. These afferents con-
trol specific targets and are differentially in- verge on second-order neurons located in
fluenced by input from the hypothalamus laminae I, V, VII , and X of the spinal cord
and brain stem. Sympathetic functional that project to the sympathetic preganglioni
units include skin vasomotor, muscle vaso- neurons to initiate segmental somatosympa-
motor, visceromotor, pilomotor, and sudo- thetic and viscerosympathetic reflexes.6 Seg-
motor units.6'7 Thus, there are clear differ- mental reflex pathways arising from somatic
ences between the sympathetic outflow to or visceral afferents activate ipsilateral local
skin and to muscle, but the similarities in interneuronal networks in laminae I, V, and
sympathetic activity recorded simultaneously VII, which may directly activate or inhibit the
from different muscles at rest or from skin sympathetic preganglionic neurons. With
sympathetic nerves innervating the palms the exception of IA muscle spindle and IB
and the feet are remarkable.7"9 Golgi tendon organ afferents, activation of
Preganglionic sympathetic output has a all other groups of primary sensory fibers
segmental organization, but the segmental arising from skin (Aa/3, Ay, C), muscle
distribution of preganglionic fibers does not (groups II, III, IV), and viscera (groups Ay,
follow the dermatomal pattern of somatic C) can modulate preganglionic neuron ac-
nerves.6'8 For example, sudomotor func- tivity through segmental pathways.6'7 Seg-
tional units in segments Tl and T2 inner- mental sympathetic reflexes are segmentally
vate the head and neck, units in T3-T6 in- biased, predominantly uncrossed, and ex-
nervate the upper extremities and thoracic hibit ipsilateral, function-specific, recipro-
viscera, those in T7—Til innervate the ab- cal, and nonreciprocal patterns of response.
dominal viscera, and ones in T12-L2 inner- For example, nociceptive stimuli reflexly ac-
vate the lower extremities and pelvic and tivate segmental circuitry that generates ex-
perineal organs.8 Preganglionic sympathetic citation of vasoconstrictor outflow to skele-
axons exit through ventral roots and pass tal muscle and inhibition of vasoconstrictor
through the white ramus communicans of outflow to the skin.7 Unlike somatic reflexes
the corresponding spinal nerve to reach (for example, the flexor reflex), segmental
the paravertebral sympathetic chain and to sympathetic reflexes do not exhibit recipro-
synapse on neurons located in the paraver- cal contralateral response patterns, but they
tebral or prevertebral ganglia.8 Paraverte- may exhibit crossed, nonreciprocal re-
bral ganglia innervate all tissues and organs sponses. For example, during execution of
except those in the abdomen, pelvis, and the cold pressor test, decreases in cutaneous
perineum. Their postganglionic fibers des- blood flow in an arm exposed to ice cold wa-
tined for the trunk and limbs follow the ter are accompanied by cutaneous vasocon-
course of spinal nerves or blood vessels or striction in the contralateral forearm. All
both. Spinal fibers join the peripheral spinal segmental spinal reflexes are subject to
supraspinal modulation through several par- spontaneous skin sympathetic activity, but in
allel pathways arising in the hypothalamus, this case, the bursts are caused by simulta-
pons, and medulla and innervating the sym- neous activation of sudomotor and vasomo-
pathetic preganglionic neurons.2"4 tor impulses.8'9 Cutaneous arteries and veins
have a prominent noradrenergic innerva-
tion that regulates both nutritive and arte-
Assessment of Sympathetic riovenous skin blood flow.5'10 Nutritive skin
Function in Humans flow is carried by capillaries and is regulated
by sympathetic (a- and /3-adrenergic) and lo-
Sympathetic function in humans can be as- cal nonadrenergic mediators. Arteriovenous
sessed, directly or indirectly, with various skin blood flow is carried by low-resistance
noninvasive or invasive techniques. Indirect arteriovenous shunts, which receive abun-
methods include noninvasive tests of sudo- dant sympathetic vasoconstrictor input and
motor and cardiovascular function described have a key role in thermoregulation. >10 Skin
in Chapters 34-37; measurement of plasma vasoconstrictor neurons may coordinate
norepinephrine concentration in forearm their activity with vasomotor neurons in
veins with the subject supine and standing; other vascular beds to maintain cardiac out-
assessment of splanchnic and cerebral blood put, but they are not sensitive to baroreflex
flow using Doppler techniques; and assess- input.
ment of sympathetic innervation of the heart Skin blood flow is also controlled by so-
using radioisotope methods. In comparison, matosympathetic reflexes6'7 and three local
microneurographic technique allows direct axon reflexes: (1) axon flare response, (2)
recording of postganglionic sympathetic sudomotor axon reflex, and (3) venoarteri-
nerve activity in humans.8'9 Nerve record- olar reflex. The axon flare response is me-
ings are made with tungsten microelectrodes diated by nociceptive C-fiber terminals.11 Ac-
inserted percutaneously into a nerve, espe- tivation by noxious chemical or mechanical
cially the median, peroneal, or tibial nerve. stimuli produces antidromic release of neu-
This technique allows multiunit recordings ropeptides (substance P and others) that
of two different types of outflow: skin sym- cause skin vasodilatation directly and through
pathetic nerve activity and muscle sympa- stimulation of histamine release by mast
thetic nerve activity.8-9 cells. The sudomotor axon reflex is medi-
ated by sympathetic sudomotor C fibers. The
venoarteriolar reflex is mediated by sympa-
thetic vasomotor axons innervating small
SYMPATHETIC INNERVATION veins and arterioles. Skin vasomotor activity
OF THE SKIN has been studied clinically by using several
noninvasive methods for measuring skin
Sympathetic vasomotor, pilomotor, and su- blood flow, including plethysmography and
domotor innervation of skin effectors has laser Doppler flowmetry.11
primarily a thermoregulatory function.5'10 The eccrine sweat glands in humans have
Skin sympathetic activity is a mixture of su- a major role in thermoregulation. The seg-
domotor and vasoconstrictor impulses and mental pattern of distribution of sudomotor
may sometimes include pilomotor and va- fibers to the trunk and limbs is irregular and
sodilator impulses. The average conduction varies substantially among individuals and
velocity for skin sudomotor and vasomotor even between the right and left sides of an
fibers is 1.3 m/second and 0.8 m/second, re- individual.5 Sympathetic inputs to the sweat
spectively.8'9 The intensity of skin sympa- glands are mediated by acetylcholine, acting
thetic activity is determined mainly by envi- via M3 muscarinic receptors.
ronmental temperature and the emotional
state of the subject. Decreased or increased
environmental temperature can produce se- MUSCLE SYMPATHETIC
lective activation of the vasoconstrictor or su- ACTIVITY
domotor system, respectively, with suppres-
sion of activity in the other system. Emotional Muscle sympathetic activity is composed of
stimuli or inspiratory gasp also increases vasoconstrictor impulses that are strongly
modulated by arterial baroreceptors.7"9 quencies (approximately 0.15 Hz) are me-

Conduction velocity of the postganglionic C diated almost exclusively by the vagus
fibers has been estimated to be 0.7 m/sec- nerve.16"18 Vagal influences on the heart are
ond in the median nerve and 1.1 m/second associated directly with respiratory activity
in the peroneal nerve.8 At rest, there is a and are minimal during inspiration and
striking similarity between muscle sympa- maximal at the end of inspiration and in
thetic activity recorded in different extrem- early expiration. This is the basis of the res-
ities. However, this activity in the arm and piratory sinus arrhythmia.11'16'17 Spontaneous
leg can be dissociated during mental stress and baroreflex-induced firing of central
and during forearm ischemia after isometric cardiovagal neurons is inhibited during in-
exercise.12 Muscle sympathetic activity is spiration and is maximal during early expi-
important for buffering acute changes of ration.14 Increased tidal volume increases
blood pressure and decreases in response to respiratory sinus arrhythmia, whereas in-
baroreceptor influence.8'9'13 However, it has creased respiratory frequency decreases it.
much less importance for long-term control Heart rate variability is correlated inversely
of blood pressure.9 At rest, muscle sympa- with age in normal subjects at rest.11
thetic activity correlates positively with ante- Spontaneous lower frequency fluctuations
cubital venous plasma norepinephrine lev- (those less than 0.15 Hz) of heart rate are
els.14 Muscle sympathetic activity is also mediated by both the vagus and the sympa-
inhibited by cardiopulmonary receptors. thetic nerves and may be related to barore-
Respiratory cycle, changes of posture, or the flex activity, temperature regulation, and
Valsalva maneuver may affect muscle sym- other factors. Upright posture in humans
pathetic activity caused by changes in arter- dramatically increases sympathetic nerve
ial pressure. However, hypercapnia, hypoxia, activity as well as a large increase in low-
isometric handgrip, emotional stress, or the frequency heart rate power.16"18
cold pressor test increases muscle sympa-
thetic activity despite unchanged or in-
creased arterial pressure.8'9'14 CARDIOVASCULAR REFLEXES
Arterial Baroreflexes
AUTONOMIC CONTROL OF
HEART RATE In normal subjects, control of arterial pres-
sure depends primarily on the sympathetic
Heart rate depends on the effects of para- innervation of the blood vessels, particularly
sympathetic and sympathetic modulation of in the splanchnic bed.19"23 The number of
the intrinsic firing rate of the sinus node. splanchnic sympathetic preganglionic neu-
Parasympathetic control is provided by car- rons progressively decreases with age, and
diovagal neurons in the nucleus ambiguus orthostatic hypotension occurs after more
and dorsal motor nucleus in the medulla.5'15 than 50% of them have been lost.11 Changes
Sympathetic noradrenergic control derives in sympathetic outflow are regulated by ar-
mainly from the cervical and upper thoracic terial baroreceptors and chemoreceptors,
ganglia.5 In humans at rest, parasympathetic cardiopulmonary receptors, and receptors
tone predominates over the excitatory sym- in skeletal muscles (that is, ergorecep-
pathetic j8-adrenergic influence. Effects me- tors).19"24 In addition, these reflexes are
diated by the vagus nerve have a shorter la- modulated by central commands, particu-
tency and duration than those mediated by larly from the amygdala and the hypothala-
the sympathetic nerves. Heart rate has spon- mus.2'3 In humans, arterial pressure is regu-
taneous fluctuations, which reflect changing lated primarily by the carotid sinus and
levels of autonomic activity modulating si- aortic baroreceptors innervated by branches
nus-node discharge.16 of CNs IX and X, respectively, and by car-
Use of power spectral analysis of heart rate diopulmonary mechanoreceptors inner-
fluctuations allows a noninvasive quantita- vated by vagal and sympathetic afferents.
tive assessment of beat-to-beat modulation of The primary role of arterial baroreflexes is
neuronal activity affecting the heart.17"19 the rapid adjustment of arterial pressure
Fluctuations of heart rate at respiratory fre- around the existing mean arterial pres-
sure.19 23 The baroreceptor reflexes provide Venoarteriolar Reflexes

a negative feedback that buffers the mag-
nitude of arterial pressure oscillations The venoarteriolar reflex is a sympathetic
throughout the day. postganglionic C-fiber axon reflex, with re-
Baroreflexes induce short-term changes ceptors in small veins and effectors in mus-
in heart rate opposite in direction to the cle arterioles. Venous pooling activates re-
changes in arterial pressure, thus increasing ceptors in small veins of the skin, muscle,
heart rate variability. The carotid baroreflex and adipose tissue; the result is vasocon-
has been studied by applying negative and striction in the arterioles supplying these tis-
positive pressures to the neck, which in- sues. During limb dependency, this local re-
crease and decrease, respectively, carotid si- flex vasoconstriction may decrease blood
nus transmural pressure.25 The combined flow by 50%. The main function of this re-
influence of carotid and aortic barorecep- flex is to increase total peripheral resis-
tors in the control of heart rate has been tance.11
studied by measuring heart rate responses to
changes in arterial pressure induced by in-
travenous infusion of vasoconstrictor or va-
sodilator agents.25 Despite their major in- Ergoreflexes
fluence on heart rate, the buffering effects
of the carotid baroreflex depend predomi- Static muscle contraction increases heart
nantly on changes in total peripheral resis- rate and blood pressure. The mechanisms
tance.19"23 underlying these responses are thought to
Baroreflex control of regional circulation involve (1) reflexes initiated by the activation
is heterogeneous and largely affects resis- of chemosensitive endings of small myelin-
tance vessels in the splanchnic area.19"24 ated and unmyelinated afferent fibers by lo-
Sympathetic vasoconstriction in skeletal cal metabolites in the contracting muscle
muscle is also strongly modulated by the and (2) a central command that influences
baroreflex;8'9 this control is dynamic and descending autonomic pathways. Cardiovas-
more suitable for buffering short-term than cular responses to moderately intense static
long-term variations of arterial pressure.20'21 contraction may be produced primarily by
During exercise, carotid baroreceptor activ- the motor command, which is solely re-
ity is rapidly adjusted to a higher level; this sponsible for increased heart rate. At higher
allows increased arterial pressure to meet intensity, responses depend on both the mo-
the metabolic demands of the contracting tor command and muscle chemoreflexes.19"23
muscles.19
Cardiopulmonary Reflexes MAINTENANCE OF POSTURAL

NORMOTENSION
Cardiopulmonary receptors are innervated
by vagal and sympathetic myelinated and un- In healthy subjects, orthostatic stress such as
myelinated afferent fibers. Atrial receptors active standing, head-up tilt, or application
innervated by vagal myelinated fibers are ac- of lower body negative pressure results in
tivated by atrial distention or contraction pooling of venous blood in the capacitance
and initiate reflex tachycardia caused by se- vessels of the legs and abdomen. The bulk
lective increase of sympathetic outflow to the of venous pooling occurs within the first 10
sinus node. Cardiopulmonary receptors with seconds. In addition, central blood volume
unmyelinated vagal afferents, similar to ar- decreases following transcapillary filtration
terial baroreceptors, tonically inhibit vaso- of fluid into the interstitial space in the de-
motor activity. Unlike baroreceptors, car- pendent parts.26 These combined effects
diopulmonary receptors provide sustained lead to a decrease in venous return to the
rather than phasic control in sympathetic ac- heart, end-diastolic filling of the right ven-
tivity and vasomotor tone in the muscle and tricle, and stroke volume, resulting in ap-
have no major effect in controlling heart proximately a 20% decrease in cardiac out-
rate.19-23 put. The decrease in mean arterial pressure
is prevented by a compensatory vasocon- tor and vasoconstrictor impulses and is reg-

striction of the resistance and capacitance ulated mainly by environmental tempera-
vessels of the splanchnic, renal, and muscle ture and emotional influences. Muscle
vascular beds. The initial adjustments to or- sympathetic activity is composed of vasocon-
thostatic stress are mediated by baroreflexes strictor impulses that are strongly modulated
and cardiopulmonary reflexes; during pro- by arterial baroreceptors. Heart rate is con-
longed orthostatic stress, additional adjust- trolled by vagal parasympathetic and tho-
ments are mediated by humoral mechanisms, racic sympathetic inputs. Vagal influence on
including the arginine-vasopressin and renin- the heart rate is strongly modulated by res-
angiotensin-aldosterone systems.21'26 Carotid piration; it is more marked during expira-
sinus baroreceptors are of primary impor- tion and is absent during inspiration. This is
tance during standing, whereas cardiopul- the basis for the so-called respiratory sinus
monary receptors have a small role. When arrhythmia, which is an important index of
venous filling in the dependent parts in- vagal innervation of the heart. Power spec-
creases intravenous pressure to 25 mm Hg, tral analysis of heart rate fluctuations allows
it activates a local axon reflex, called the noninvasive assessment of beat-to-beat mod-
venoarteriolar reflex, which contributes to ulation of neuronal activity affecting the
adaptation to orthostatic stress. heart. Arterial baroreflex, cardiopulmonary
Patients with autonomic failure have dis- reflexes, venoarteriolar reflex, and ergore-
turbed neural reflex arterial vasoconstric- flexes control sympathetic and parasympa-
tion, and this is the primary mechanism of thetic influences on cardiovascular effectors.
orthostatic hypotension. The inability to in- The main regulatory mechanism that pre-
crease vascular resistance allows consider- vents orthostatic hypotension is reflex arter-
able venous pooling to occur in the skeletal ial vasoconstriction in the splanchnic, renal,
muscle, cutaneous, and splanchnic vascular and muscular beds triggered by a decrease
beds of the dependent parts. The abdomi- in transmural pressure at the level of carotid
nal compartment (splanchnic circulation) sinus baroreceptors.
and perhaps skin vasculature are the most
likely sites of venous pooling. The counter-
regulatory mechanism that reduces pooling
in the lower extremities is activation of the REFERENCES
skeletal muscle pump. Active muscle con-
traction increases intramuscular pressure, 1. Cervero F, Foreman RD. Sensory innervation of the
which opposes the hydrostatic forces and re- viscera. In Loewy AD, Spyer KM (eds). Central Reg
ulation of Autonomic Functions. Oxford University
duces venous pooling in the legs. Press, New York, 1990, pp 104-125.
2. Loewy AD. Anatomy of the autonomic nervous sys-
tem: an overview. In Loewy AD, Spyer KM (eds).
Central Regulation of Autonomic Functions. Ox-
SUMMARY ford University Press, New York, 1990, pp 3-16.
3. Dampney RA. Functional organization of central
pathways regulating the cardiovascular system.
The autonomic nervous system consists of Physiol Rev 74:323-364, 1994.
three divisions: sympathetic (thoracolum- 4. JanigW, Habler H-J. Organization of the autonomic
bar), parasympathetic (craniosacral), and nervous system: structure and function. In Vinken
PJ, Bruyn GW (eds). Handbook of Clinical Neu-
the enteric nervous system. The sympathetic rology. Vol. 74; series 30: The Autonomic Nervous
and parasympathetic autonomic outflow in- System. Part I. Normal Functions. Elsevier Science
volves a two-neuron pathway with a synapse Publishers, Amsterdam, 1999, pp 1-52.
in an autonomic ganglion. Preganglionic 5. Gibbins I. Peripheral autonomic nervous system. In
sympathetic neurons are organized into dif- Paxinos G (ed). The Human Nervous System. Aca-
demic Press, San Diego, 1990, pp 93-123.
ferent functional units that control specific 6. Cabot JB. Some principles of the spinal organiza-
targets and include skin vasomotor, muscle tion of the sympathetic preganglionic outflow. Prog
vasomotor, visceromotor, pilomotor, and Brain Res 107:29-42, 1996.
sudomotor units. Microneurographic tech- 7. Janig W. Spinal cord reflex organization of sympa-
thetic systems. Prog Brain Res 107:43-77, 1996.
niques allow recording of postganglionic 8. Wallin BG, Fagius J. Peripheral sympathetic neural
sympathetic nerve activity in humans. Skin activity in conscious humans. Annu Rev Physiol
sympathetic activity is a mixture of sudomo- 50:565-576, 1988.
9. Wallin BG, Elam M. Insights from intraneural system in clinical studies. Auton Neurosci 90:24-28,
recordings of sympathetic nerve traffic in humans. 2001.
News Physiol Sci 9:203-207, 1994. 18. Bernard! L, Porta C, Gabutti A, Spicuzza L, Sleight
10. Johnson JM. Nonthermoregulatory control of hu- P. Modulatory effects of respiration. Auton Neu-
man skin blood How. J Appl Physiol 61:1613-1622, rosci 90:47-56, 2001.
1986. 19. Mancia G, Grassi G, Ferrari A, Zanchetti A. Reflex
11. Low PA. Laboratory evaluation of autonomic func- cardiovascular regulation in humans. J Cardiovasc
tion. In Low PA (ed). Clinical Autonomic Disor- Pharmacol 7 (Suppl 3):S152-S159, 1985.
ders: Evaluation and Management, 2nd ed. Lip- 20. Abboud FM. Interaction of cardiovascular reflexes
pincott-Raven Publishers, Philadelphia, 1997, pp in humans. In Lown B, Malliani A, Prosdocimi M
179-208. (eds). Neural Mechanisms and Cardiovascular Dis-
12. Wallin BG, Victor RG, Mark AL. Sympathetic out- ease. Liviana Press, Padova, Italy, 1986, pp 73-84.
flow to resting muscles during static handgrip and 21. Cowley AW Jr. Long-term control of arterial blood
postcontraction muscle ischemia. Am J Physiol pressure. Physiol Rev 72:231-300, 1992.
256:H105-H110, 1989. 22. Joyner MJ, Shepherd JT. Autonomic regulation of
13. Somers VK, Dyken ME, Mark AL, Abboud FM. Sym- circulation. In Low PA (ed). Clinical Autonomic
pathetic-nerve activity during sleep in normal sub- Disorders: Evaluation and Management, 2nd ed.
jects. N Engl J Med 328:303-307, 1993. Lippincott-Raven Publishers, Philadelphia, 1997,
14. Eckberg DL, Nerhed C, Wallin BG. Respiratory pp 61-71.
modulation of muscle sympathetic and vagal car- 23. Wieling W, Wesseling KH. Importance of reflexes
diac outflow in man. J Physiol (Lond) 365:181-196, in circulatory adjustments to postural change. In
1985. Hainsworth R (ed). Cardiovascular Reflex Control
15. Spyer KM, Brooks PA, Izzo PN. Vagal preganglionic in Health and Disease. WB Saunders, London,
neurons supplying the heart. In Levy MN, Schwartz 1993, pp 35-65.
PJ (eds). Vagal Control of the Heart: Experimental 24. lellamo F. Neural mechanisms of cardiovascular
Basis and Clinical Implications. Futura Publishing regulation during exercise. Auton Neurosci 90:66-
Company, New York, 1994, pp 45-64. 75, 2001.
16. van Ravenswaaij-Arts CM, Kollee LA, Hopman JC, 25. Eckberg DL, Fritsch JM. How should human
Stoelinga GB, van Geijn HP. Heart rate variability. baroreflexes be tested? News Physiol Sci 8:7-12,
Ann Intern Med 118:436-447, 1993. 1993.
17. Baillard C, Goncalves P, Mangin L, Swynghedauw 26. Smit AA, Halliwill JR, Low PA, Wieling W. Patho-
B, Mansier P. Use of time frequency analysis to fol- physiological basis of orthostatic hypotension in au-
low transitory modulation of the cardiac autonomic tonomic failure. J Physiol (Lond) 519:1-10, 1999.
Chapter 34
QUANTITATIVE SUDOMOTOR AXON
REFLEX TEST AND RELATED TESTS
Phillip A. Low
LABORATORY EVALUATION OF Significance of the Test

AUTONOMIC FUNCTION Imprint Methods of Sweat Measurement
QUANTITATIVE SUDOMOTOR AXON Plastic and Silicone Imprints
REFLEX TEST SUMMARY
Normal Response
Abnormal Patterns
LABORATORY EVALUATION OF hidrosis4 and benign syncopes, need to be

AUTONOMIC FUNCTION differentiated from the more serious dysau-
tonomias. Distal small fiber neuropathy is
The quantitative sudomotor axon reflex test diagnosable with autonomic studies, and
(QSART) is a routine test of autonomic func- QSART is abnormal in 80% of cases,5'6 sug-
tion and a component of the autonomic re- gesting that most patients who have distal so-
flex screen (Table 34-1). A detailed de- matic C-fiber involvement also have auto-
scription of tests of autonomic function is nomic C-fiber impairment. This relationship
beyond the scope of this chapter but is con- is not invariable, and some patients with dis-
sidered elsewhere.1 Because autonomic tests tal small fiber neuropathy and abnormal
are affected substantially by many con- epidermal skin fibers will have a normal
founding variables, standardization,2 the QSART.7 Sympathetically maintained pain is
recognition of pitfalls,3 and patient prepa- associated with unilateral vasomotor and su-
ration (Table 34-2) are critically important. domotor abnormalities8'9 and is described in
The autonomic reflex screen noninvasively detail elsewhere.10
and quantitatively evaluates sudomotor, ad-
renergic, and cardiovagal functions. Sudo-
motor function is discussed in this chapter, QUANTITATIVE SUDOMOTOR
and adrenergic and cardiovagal functions AXON REFLEX TEST
are considered in Chapters 35 and 37.
The indications for autonomic testing are Many tests of sudomotor function are of his-
summarized in Table 34-3. The presence of torical and limited clinical and research in-
generalized autonomic failure, a major med- terest. For the modern clinical neurophysi-
ical problem, adversely affects prognosis. It ology laboratory, only the thermoregulatory
is important to quantify the deficits by sys- sweat test, QSART, and sweat imprint test
tem and by autonomic level. Certain benign (SIT) require consideration. The thermo-
disorders, such as chronic idiopathic an- regulatory sweat test is described in Chapter
445
Table 34-1. Autonomic Reflex Screen Table 34-3. Indications for Autonomic
Laboratory Evaluation
QSART distribution
Orthostatic blood pressure and heart rate Strong indications
response to tilt Suspicion of generalized autonomic failure
Heart rate response to deep breathing Diagnosis of benign autonomic disorders that
The Valsalva ratio mimic more serious dysautonomia
Beat-to-beat blood pressure to Valsalva Detection of distal small fiber neuropathy
maneuver, tilt, and deep breathing Diagnosis of autonomic neuropathy
QSART, quantitative sudomotor axon reflex test. Sympathetically maintained pain
Orthostatic intolerance, such as the postural
orthostatic tachycardia syndrome
Desirable indications
36. The QSART and the sweat imprint test Monitoring course of autonomic failure
are described in this chapter. Evaluation of response to therapy
Peripheral neuropathies
Syncope
Normal Response
Amyotrophic lateral sclerosis
The neural pathway consists of an axon re- Extrapyramidal and cerebellar degenerations
flex mediated by the postganglionic sympa- Research questions
thetic sudomotor axon. The axon terminal
is activated by iontophoresed acetylcholine.
The impulse travels antidromically, reaches
a branch point, and then travels ortho- The major pieces of equipment needed to
dromically in the branch to release acetylcho- perform QSART for each site are the su-
line from the nerve terminal. Acetylcholine dorometer, multicompartmental sweat cell,
traverses the neuroglandular junction and constant current generator, and some
binds to muscarinic receptors on eccrine method of displaying the sweat response. We
sweat glands to evoke the sweat response build our own sudorometer consoles; each
(Fig. 34-1). Alternative approaches are the consists of four sudorometers, which permits
use of acetylcholine gel instead of solution the dynamic recording of sweat output from
(improves contact) and the use of carba- four sites simultaneously. A sudorometer
chol11 instead of acetylcholine. The former modeled on the Mayo system is available
results in a larger response, because acetyl- commercially (WR Medical, Stillwater, MN).
cholinesterase breakdown does not occur. It is called QSWEAT and is highly sensitive
and robust. A major advantage of QSWEAT
is its use of room air (which it dehumidifies)
instead of reliance on nitrogen gas.
Table 34-2. Patient Preparation for The multicompartmental sweat cell (Fig.
34-2) is attached to the skin and permits the
Quantitative Sudomotor Axon iontophoresis of acetylcholine via the stimu-
Reflex Test lus compartment (C) and a constant current
generator. The axon-reflex-mediated sweat
No food for 3 hours before testing. The response is recorded from compartment A.
antecedent meal should be a light breakfast
or lunch without coffee or tea.
Quantitative sudomotor axon reflex test
responses are recorded from four standard
Treatment with anticholinergic and diuretic
agents should be stopped 48 hours and
sites, which are the distal forearm, proximal
preferably 4 days before the study. leg, distal leg, and proximal foot. These sites
The patient should be comfortable and pain-
were chosen because they provide a wide dis-
free (for example, bladder recently emptied). tribution of sites for four different periph-
The room should be warm and quiet. eral nerves. This distribution also provides
the means for detecting a length-dependent
Quantitative Sudomotor Axon Reflex Test and Related Tests 447
Figure 34-1. Left, Neural substrate of sudomotor axon reflex test (see text). Right, Representative response from a
36-year-old woman. Arrows indicate onset and cessation of iontophoresis. (From Low PA, Opfer-Gehrking TL,
Kihara M. In vivo studies on receptor pharmacology of the human eccrine sweat gland. Clin Auton Res 2:29-34,
1992. By permission of Lippincott, Williams & Willkins.)
neuropathy, which shows a progressive de- diabetic neuropathy. The coefficient of vari-
cline in sweat volume. Recently, some labo- ation is between 10% and 20%.
ratories have focused on the evaluation of Extensive control data are available. We
even more distal sites (such as the thumb) have normative QSART data from 357 nor-
and have reported that the ratio of volumes mal subjects 10—83 years old,14 evenly dis-
of distal to proximal sites provides an espe- tributed by age and sex. The data were ana-
cially sensitive index of length-dependent lyzed using stepwise regression analysis. For
denervation.12 The tests are sensitive and re- the forearm, proximal leg, distal leg, and
producible in controls13 and in patients with proximal foot sites, the regression coeffi-
Figure 34-2. Quantitative sudomotor axon reflex test. Left, Sweat cell and, right, response. The sweat response in
compartment A is evoked in response to iontophoresis of acetylcholine in compartment G Compartment B is an
air gap. Sweating causes a change in thermal mass of the nitrogen stream (D) that is sensed by the sudorometer
and displayed (right). (From Low PA. Sudomotor function and dysfunction. In Asbury AK, McKhann GM, McDonald
WI [eds]. Diseases of the Nervous System. WB Saunders Company, Philadelphia, 1986, pp 596-605. By permission
of Butterworth-Heinemann.)
cients are as follows, where bO is intercept,

b2 is sex, bl is age: forearm, 4.19072,
— 1.51975, bl = not significant (NS); proxi-
mal leg, 4.09765, -1.05836; -0.017960; dis
tal leg, 5.07821, -1.25166, -0.32466; prox-
imal foot, 3.81938, -1.05945, 0.015621. The
P value for all coefficients reaching statisti-
cal significance was P = 0.0001. For all sites,
sex had the largest effect, with the sweat vol
ume of women being 1.0-1.5 ju-L/cm2 lower
than that of men. The effect of age was sig
nificant for all three lower extremity sites but
not for the forearm.
Abnormal Patterns
Several abnormal QSART patterns are rec-
ognized. The response may be (1) normal,
(2) reduced, (3) absent, (4} excessive, (5)
persistent, or (6) associated with ultrashort
latencies. Short latencies are common with
patterns 4 and 5. Pattern 5 consists of a hung-
up sweat response that fails to turn off when
the stimulus ceases; it is often seen in painful
diabetic and other neuropathies and in
florid reflex sympathetic dystrophy. The ul- Figure 34-3. Quantitative sudomotor axon reflex test
trashort latency is caused by an augmented distribution in a patient with distal small fiber neu-
ropathy. Note reduction of sweat response in the distal
somatosympathetic response. The persistent leg and anhidrosis over the foot. (From Low PA,
sweat activity represents continuous firing of McLeodJG. Autonomic neuropathies. In Low PA [ed].
sympathetic sudomotor fibers. Patients with Clinical Autonomic Disorders: Evaluation and Man-
a distal small fiber neuropathy may show a agement, 2nd ed. Lippincott-Raven Publishers, Phila-
delphia, 1997, pp 463-486. By permission of Mayo
characteristic pattern with reduced re- Foundation.)
sponses confined to the distal leg and foot
(Fig. 34-3).
sympathetic axon. Damaged axons can fire

Significance of the Test spontaneously, and on stimulation, they may
have more persistent activity than normal
Normal test results indicate integrity of the post- axons.
ganglionic sympathetic sudomotor axon. The ab- The latency is sometimes markedly re-
sence of a response indicates failure of this duced in painful neuropathies with persis-
axon, providing that iontophoresis is suc- tent sweat activity. The mechanism of the
cessful and eccrine sweat glands are present. very short latency (10-30 seconds) is likely
Because the axonal segment mediating the augmented somatosympathetic reflexes, with
response is likely to be short, the test prob- a reduced threshold of polymodal C noci-
ably evaluates distal axonal function. ceptors—it often occurs with allodynia.
The presence of persistent sweat activity
occurs commonly in mild or painful neu- APPLICATIONS
ropathies. Because it is unlikely that contin-
uous activity occurs at the level of the sweat The QSART has been used to detect post-
gland, the mechanism of persistent sweat ac- ganglionic sudomotor failure in neurop-
tivity is likely to be repetitive firing of the athies,13'15 aging,16 Lambert-Eaton myas-
Quantitative Sudomotor Axon Reflex Test and Related Tests 449
thenic syndrome,17 amyotrophic lateral scle- in a solvent, which dries in approximately

rosis,18 and preganglionic neuropathies with 30 seconds. The sweat droplet forms an
presumed transsynaptic degeneration.19'20 imprint. Alternatives are silicone rubber
In patients with distal small fiber neuropa- monomers24 and colloidal graphite plastic.25
thy, it is the most sensitive noninvasive di- Harris et al.25 compared bromophenol-blue,
agnostic test available.5 It is an important colloidal graphite plastic, and silicone
part of the autonomic reflex screen, when monomer imprint methods and concluded
QSART distribution is taken in conjunction that the last was the most sensitive and re-
with cardiovascular heart rate tests and Fi- liable.
napres recordings of the Valsalva maneuver Kennedy et al.26 used a combination of pi-
and tilt.10 Recently, the test has been used locarpine iontophoresis, the Silastic imprint
to define the threshold and duration of the method, and imaging techniques to system-
local anhidrotic effect of botulinum toxin.11 atically study the sweat response in humans
In studies of sympathetically maintained and rodents. For human studies, a 1 % solu-
pain, QSART recordings are performed bition of pilocarpine is iontophoresed for 5
laterally and simultaneously for evidence of minutes over a 1 cm2 area of skin on the dor-
asymmetry of latency, volume, and mor- sum of the hand and foot. The imprint ma-
phology of the responses.8 terial (0.5 mL base and 4 droplets of accel-
Another application of QSART is detec- erator [Syringe Elasticon, Kerr Co]) is
tion of the site of the lesion. The QSART in spread thinly over the skin. The material sets
combination with the thermoregulatory within 3 minutes.
sweat test can be used to determine whether Counts of sweat gland density are made
a lesion is preganglionic or postganglionic. under a dissecting microscope and counting
A preganglionic site is deduced when an- grid. To estimate volume, the size of the
hidrosis on the thermoregulatory sweat test droplet can be determined with computer-
is associated with normal results on QSART, ized imaging analysis techniques. The mean
and the lesion is postganglionic when both density is 311 and 281 active glands/cm2 for
tests show anhidrosis. the hand and foot, respectively, without dif-
For comparisons among patients of differences for age or sex.27
ferent ages and sex, these confounding vari- This technique has been applied system-
ables can be managed by at least two ap- atically to patients with diabetic neuropa-
proaches. One, the data can be expressed as thy.26'27 Counts of active sweat glands were
normal deviates.8 Two, an autonomic scor- abnormal in the hands of 24% of patients
ing scale can be generated that corrects for with diabetic neuropathy and in the feet in
the effects of age and sex.21 On this scale, 56%. Of interest is that approximately one-
severity is scored from 0 (no deficit) to 3 third of patients with normal electrophysio-
(maximal deficit). logic tests have abnormal results on the
sweat imprint test.22'28
Imprint Methods of
Sweat Measurement
Several imprint methods have been used,
SUMMARY
but only the sweat imprint method is useful The application of noninvasive, sensitive,
clinically. The subject is reviewed else- quantitative, and dynamic tests of sudomo-
where.22 tor function enhances significantly our abil-
ity to quantitate one aspect of the autonomic
deficit. The QSART shows considerable
Plastic and Silicone Imprints promise in clinical applications such as bet-
ter definition of the course of neuropathy,
A colloidal graphite plastic method was in- its response to treatment, and further ex-
troduced in 1952.23 The plastic is dissolved ploration of sudomotor physiology.
REFERENCES 14. Low PA, Denq JC, Opfer-Gehrking TL, Dyck PJ,
O'Brien PC, Slezak JM. Effect of age and gender
on sudomotor and cardiovagal function and blood
1. Low PA. Laboratory evaluation of autonomic func- pressure response to tilt in normal subjects. Muscle
tion. In Low PA (ed). Clinical Autonomic Disor- Nerve 20:1561-1568, 1997.
ders: Evaluation and Management, 2nd ed. Lip- 15. Low PA, Zimmerman BR, Dyck PJ. Comparison of
pincott-Raven Publishers, Philadelphia, 1997, pp distal sympathetic with vagal function in diabetic
179-208. neuropathy. Muscle Nerve 9:592-596, 1986.
2. Low PA, Pfeifer MA. Standardization of clinical tests 16. Low PA, Opfer-Gehrking TL, Proper CJ, Zimmer-
for practice and clinical trials. In Low PA (ed). Clin- man I. The effect of aging on cardiac autonomic
ical Autonomic Disorders: Evaluation and Manage- and postganglionic sudomotor function. Muscle
ment. Little, Brown & Company, Boston, 1993, pp Nerve 13:152-157, 1990.
287-296. 17. McEvoy KM, Windebank AJ, Daube JR, Low PA. 3,4-
3. Low PA. Pitfalls in autonomic testing. In Low PA Diaminopyridine in the treatment of Lambert-
(ed). Clinical Autonomic Disorders: Evaluation and Eaton myasthenic syndrome. N Engl J Med
Management. Little, Brown & Company, Boston, 321:1567-1571, 1989.
1993, pp 355-365. 18. Litchy WJ, Low PA, Daube JR, Windebank AJ. Au-
4. Low PA, Fealey RD, Sheps SG, Su WP, Trautmann tonomic abnormalities in amyotrophic lateral scle-
JC, Kuntz NL. Chronic idiopathic anhidrosis. Ann rosis (abstract). Neurology 37 (Suppl 1):162, 1987.
Neurol 18:344-348, 1985. 19. Cohen J, Low P, Fealey R, Sheps S, Jiang NS. So-
5. Stewart JD, Low PA, Fealey RD. Distal small fiber matic and autonomic function in progressive auto-
neuropathy: results of tests of sweating and autonomic failure and multiple system atrophy. Ann
nomic cardiovascular reflexes. Muscle Nerve 15: Neurol 22:692-699, 1987.
661-665, 1992. 20. Sandroni P, Ahlskog JE, Fealey RD, Low PA. Auto-
6. Tobin K, Giuliani MJ, Lacomis D. Comparison of nomic involvement in extrapyramidal and cerebel-
different modalities for detection of small fiber neu- lar disorders. Clin Auton Res 1:147-155, 1991.
ropathy. Clin Neurophysiol 110:1909-1912, 1999. 21. Low PA. Composite autonomic scoring scale for lab-
7. Periquet MI, Novak V, Collins MP, et al. Painful sen- oratory quantification of generalized autonomic
sory neuropathy: prospective evaluation using skin failure. Mayo Clin Proc 68:748-752, 1993.
biopsy. Neurology 53:1641-1647, 1999. 22. Kennedy WR, Navarro X. Evaluation of sudomotor
8. Low PA, McManis PG. Autonomic reflex testing in function by sweat imprint methods. In Low PA (ed).
sympathetic reflex dystrophy (abstract). Neurology Clinical Autonomic Disorders: Evaluation and Man-
35 (Suppl 1):148A, 1985. agement. Little, Brown & Company, Boston, 1993,
9. Sandroni P, Low PA, Ferrer T, Opfer-Gehrking TL, pp 253-261.
Willner CL, Wilson PR. Complex regional pain syn- 23. Sutarman, Thomson ML. A new technique for enu-
drome I (CRPS I): prospective study and laboratory merating active sweat glands in man. J Physiol
evaluation. Clin J Pain 14:282-289, 1998. (Lond) 117:51P-52P, 1952.
10. Low PA. Laboratory evaluation of autonomic fail- 24. Sarkany I, Gaylarde P. A method for demonstration
ure. In Low PA (ed). Clinical Autonomic Disorders: of sweat gland activity. Br J Dermatol 80:601-605,
Evaluation and Management. Little, Brown & Com- 1968.
pany, Boston, 1993, pp 169-195. 25. Harris DR, Polk BF, Willis I. Evaluating sweat gland
11. Braune C, Erbguth F, Birklein F. Dose thresholds activity with imprint techniques. J Invest Dermatol
and duration of the local anhidrotic effect of bot- 58:78-84, 1972.
ulinum toxin injections: measured by sudometry. 26. Kennedy WR, Sakuta M, Sutherland D, Goetz FC.
BrJ Dermatol 144:111-117, 2001. Quantitation of the sweating deficiency in diabetes
12. Riedel A, Braune S, Kerum G, Schulte-Monting J, mellitus. Ann Neurol 15:482-488, 1984.
Lucking CH. Quantitative sudomotor axon reflex 27. Kennedy WR, Navarro X. Sympathetic sudomotor
test (QSART): a new approach for testing distal function in diabetic neuropathy. Arch Neurol
sites. Muscle Nerve 22:1257-1264, 1999. 46:1182-1186, 1989.
13. Low PA, Caskey PE, Tuck RR, Fealey RD, Dyck PJ. 28. Navarro X, Kennedy WR. Sweat gland reinnerva-
Quantitative sudomotor axon reflex test in normal tion by sudomotor regeneration after different
and neuropathic subjects. Ann Neurol 14:573-580, types of lesions and graft repairs. Exp Neurol
1983. 104:229-234, 1989.
Chapter 35
ADRENERGIC FUNCTION
Phillip A. Low
SKIN VASOMOTOR REFLEXES Denervation Supersensitivity

Beat-to-Beat Blood Pressure Response to Plasma Norepinephrine
the Valsalva Maneuver Sustained Handgrip
Beat-to-Beat Blood Pressure Response to Baroreflex Indices
Tilt-Up SUMMARY
Venoarteriolar Reflex
Standing up causes a transient decrease in skin vasoconstrictor reflexes.1 Studies are

pulse and blood pressure. Baroreceptors in performed usually on the toe or finger
the carotid sinus and aortic arch are acti- pads, because the sympathetic innervation in
vated, resulting in the unloading of barore- these sites is purely vasoconstrictor, whereas
flexes. The efferent portion of the barore- other sites, such as the forearm, have both
flex consists of a vagally mediated increase vasoconstrictor and vasodilator fibers.2
in heart rate and activation of sympathetic Skin blood flow (SBF) is measured with a
efferents. Preganglionic sympathetic fibers laser Doppler flowmeter or with plethys-
are cholinergic, and postganglionic fibers mography, and the vasoconstrictor response
are adrenergic. The preganglionic fiber is to an autonomic maneuver is determined.
short and ends in a paravertebral ganglion. Vasoconstriction can be induced by maneu-
In comparison, the postganglionic fiber tra- vers such as inspiratory gasp, response to
verses the entire length of a peripheral standing (for the finger), contralateral cold
nerve. Peripheral adrenergic function is im- stimulus, or the Valsalva maneuver. The re-
portant in the maintenance of postural nor- sponse can be expressed as a percentage of
motension. It may be impaired in peripheral vasoconstriction:
neuropathy and this may be manifested as
alterations in acral temperature, color, or Percentage of vasoconstriction =
volume. In spite of its importance, simple, Resting SBF - minimal SBF
accurate, and reproducible tests of periph- X 100
eral adrenergic function are generally un- Resting SBF
available. This chapter describes methods
used to determine peripheral adrenergic The pathways of these reflexes are com-
function and their value and shortcomings. plex. For instance, the response to standing
Microneurography is used occasionally to is mediated by the venoarteriolar reflex,3
evaluate adrenergic function, but it is inva- low-pressure and high-pressure barorecep-
sive and is not described here. tors,4'5 and, to a lesser extent, by increased
levels in epinephrine,6 norepinephrine,7
and renin.8 The advantage of these reflexes
SKIN VASOMOTOR REFLEXES is that they have different afferents but an
identical final efferent pathway. Thus, the
The integrity of postganglionic sympathetic relevant afferent pathway can be evaluated.
adrenergic fibers can be evaluated by testing However, this advantage is offset by a major
451
shortcoming of tests of skin adrenergic func- has also been validated by a study of control
tion, that is, the marked sensitivity of skin and three age-matched and sex-matched pa-
sympathetic fibers to emotional and tem- tient groups with graded adrenergic fail-
perature changes,9 which means there is ure.11 One group had generalized auto-
considerable ambient fluctuation1 of skin nomic failure, with an orthostatic decrease
blood flow. in systolic blood pressure during tilt of 30
Skin vasomotor reflexes have a coefficient mm Hg or greater. A second group had less
of variation greater than 20% and are best orthostatic decrease in blood pressure (less
regarded as semiquantitative tests. However, than 30 mm Hg but greater than 10 mm
the tests are useful in comparing vasocon- Hg), and a third group had well-docu-
strictor reflexes from identical sites simulta- mented peripheral autonomic failure (ab-
neously. sence of response on the quantitative sudo-
motor axon reflex test) but did not have
orthostatic hypotension. In contrast to con-
trols, all the patient groups, including group
Beat-to-Beat Blood Pressure 3, exhibited a significant reduction in phase
Response to the Valsalva II-L. An excessive decrease in blood pressure
Maneuver in phase II and absence of a phase IV over-
shoot were observed in the group with
The evaluation of adrenergic function in an florid orthostatic hypotension. Intermediate
autonomic laboratory includes measure- changes were seen in the group with bor-
ment of plasma levels of catecholamines derline orthostatic hypotension. The beat-to-
(supine and standing), the sustained hand- beat blood pressure changes during the Val-
grip test, and the response to pharmacologic salva maneuver, when coupled with blood
agents.10 These tests generally are insensitive pressure responses to tilt, provide a signifi-
(catecholamines), poorly reproducible in cantly better evaluation of adrenergic failure
the clinical laboratory setting, or too inva- than bedside recordings of blood pressure.
sive. In the Mayo Autonomic Laboratory, two Patients with peripheral adrenergic failure,
well-validated tests are used routinely: blood for example, those with neuropathy involv-
pressure and heart rate responses to head- ing autonomic C fibers, have an absence of
up tilt and beat-to-beat blood pressure re- phase II-L and a slight increase in phase II-
sponses to the Valsalva maneuver. These are E. Patients with more severe autonomic fail-
the focus of this chapter. ure, for example, widespread involvement of
Beat-to-beat blood pressure responses to limb and splanchnic adrenergic fibers, and
the Valsalva maneuver and tilt are recorded cardiovagal impairment have a large phase
simultaneously with the heart rate. Dynamic II-E, an absence of phase II-L, and a pre-
alterations during tilt and the Valsalva ma- served phase IV. If phase IV is absent, car-
neuver are particularly important in detect- diac adrenergic innervation has failed. A
ing adrenergic failure. The Valsalva maneu- caveat is that a small or, less often, absent
ver has four main phases: I, II, III, and IV, phase IV can occur in normal subjects if
with phase II subdivided into early (II-E) and phase II is modest or absent. The gradations
late (II-L) phases (Fig. 35-1, Table 35-1). of alterations of the Valsalva maneuver in dif-
The mechanisms of the Valsalva maneuver ferent types and degrees of autonomic fail-
are summarized briefly in Table 35-1. ure have been summarized on the basis of
The phases of the Valsalva maneuver can the experience in the Mayo Autonomic Lab-
be used to evaluate adrenergic function. oratory (Table 35-2).
This method of evaluating adrenergic func-
tion has been validated in two ways. First,
pharmacologic dissection studies have dem- Beat-to-Beat Blood Pressure
onstrated that phase II-L is primarily under Response to Tilt-Up
peripheral (a-) adrenergic control and is se-
lectively blocked by phentolamine, whereas Orthostatic blood pressure recordings to tilt
phase IV is completely blocked by propran- are recorded using beat-to-beat blood pres-
olol, indicating that it depends on (3- sure and verified with a sphygmomanome-
adrenoreceptors.11 Second, the technique ter cuff with the patient supine and follow-
Adrenergic Function 453
Figure 35-1. The four phases

(I-IV) of the Valsalva maneuver
in a control subject and a pa-
tient with autonomic neuropa-
thy. Note the compression of
pulse pressure and loss of late
phase II (II-L) and phase IV in
a patient with autonomic neu-
ropathy. BP, blood pressure;
II-E, early phase II.
ing tilt to 70°. It is important to perform the than after 1 minute; we routinely perform
upright tilt procedure at a standard time af- head-up tilt after a minimum of 30 minutes
ter the patient lies down, because the or- of recumbency. Beat-to-beat recordings of
thostatic reduction in blood pressure is systolic, mean, and diastolic blood pressure
greater after 20 minutes of preceding rest are displayed continuously on the computer
Table 35-1. Major Mechanisms of Phases of the Valsalva Maneuver

Phase Main Mechanism(s)
I & III Compression (I) and release from compression (III) of thoracic aorta
II-E T in venous return; partial compensation by vagus nerve
II-L t in peripheral arteriolar tone
IV t in cardiac sympathetic tone; sustained f in arteriolar tone
E, early; L, late.
t , increase.
Table 35-2. Changes in Phases of Valsalva Maneuver with Types of Autonomic

Failure
Condition II-E II-L IV Valsalva Ratio
Normal 5-15 mm i | to > baseline f to > baseline Normal

Vagal lesion ~ Normal ~ Normal Normal ~ Normal
Adrenergic failure
Peripheral = Normal 4 or absent == Normal = Normal
Generalized t Fall Absent Absent Reduced
1 , increase.
4 , decrease.
console, as is heart rate, which is derived indicate a system under stress), whereas the
from the electrocardiographic leads and gradual loss of variability indicates the fail-
monitor. Also, it is important to ensure that ure of compensation.
the arm is at heart level, because arm posi- A large autonomic database is available.
tion influences the measurement of blood The normal response varies by age and sex.
pressure. To facilitate comparison, a 10-point com-
During upright tilt, normal subjects have posite autonomic severity score (CASS) of
a transient decrease in systolic, mean, and autonomic function has been developed.13
diastolic blood pressure, followed by recov- The scheme allots 4 points for adrenergic
ery within 1 minute.12 The decrement is and 3 points each for sudomotor and car-
modest (less than 10 mm Hg, mean blood diovagal failure. Each score is normalized for
pressure). Patients with adrenergic failure the confounding effects of age and sex. Pa-
have a marked and progressive decrease in tients with a CASS score of 3 or less have only
blood pressure and pulse pressure. The mild autonomic failure, those with scores of
heart rate response typically is attenuated, 7-10 have severe failure, and those with
but in patients whose cardiac adrenergic in- scores between these two ranges have mod-
nervation is spared, the response is intact erate autonomic failure. The sensitivity and
and may be increased. Indices of mild ad- specificity of the method were assessed by
renergic impairment include excessive os- evaluating CASS in four groups of patients
cillations of blood pressure, an excessive de- with known degrees of autonomic failure: 18
crease (more than 50%) in pulse pressure, patients with multisystem atrophy, 20 with
a transient (first minute) decrease in systolic autonomic neuropathy, 20 with Parkinson's
blood pressure greater than 30 mm Hg, an disease, and 20 with peripheral neuropathy
excessive increment in heart rate (> 30 but no autonomic symptoms. The compos-
beats/min), and a failure of total systemic ite scores (mean ± SD) for these four
resistance to increase. Premonitory signs of groups were 8.5 ± 1.3, 8.6 ± 1.2, 1.5 ± 1.1,
syncope are a progressive decrease in blood and 1.7 ± 1.3, respectively. Patients with
pressure (especially diastolic blood pres- symptomatic autonomic failure had scores of
sure), total peripheral resistance, pulse pres- 5 or more, and those without symptomatic
sure, and loss of blood pressure (and heart autonomic failure had scores of 4 or less; no
rate) variability. Some of these indices are overlap existed among these groups.
expected abnormalities in a failure of arte-
riolar vasoconstriction (total peripheral re-
sistance and diastolic blood pressure). Some Venoarteriolar Reflex
are signs of increased vascular capacitance
(reduction in pulse pressure and excessive When venous transmural pressure is in-
increment in heart rate). Increased oscilla- creased by 25 mm Hg (for example, by low-
tions are indicative of intact compensatory ering the limb by 40 cm), reflex arteriolar
mechanisms (but are abnormal because they vasoconstriction occurs, reducing blood flow
constrictor response was greater in the con-

trol group than the diabetic group or other
neuropathy group, but the overlap among
the groups was marked. The venoarteriolar
reflex appeared to have lower specificity and
much lower sensitivity than other tests of au-
tonomic function. The sensitivity and speci-
ficity were considered inadequate for it to b
used as a clinical test of autonomic function.
Denervation Supersensitivity
An exaggerated pressor response to the
intra-arterial or intravenous application of
directly acting a-agonists (such as phenyl-
ephrine or norepinephrine), indicating den-
ervation supersensitivity, may occur when
there is widespread denervation of postgan-
Figure 35-2. The presumed pathway of the venoarteri- glionic sympathetic fibers.17"19 The mecha-
olar reflex. The receptor is the small vein, and the ef- nism of denervation supersensitivity is an in-
fector is the arteriole. The impulse passes antidromi-
cally along the postganglionic sympathetic sudomotor crease in receptor density, affinity, efficacy
axon to the branch point and then goes orthodromi- of receptor-effector coupling, or other
cally to activate the effector. The preganglionic and postreceptor events. These tests of adrener-
postganglionic axons emanating from the spinal cord gic denervation supersensitivity are too in-
are shown as thick and thin structures, respectively.
(From Moy et al.16 By permission of the American Acad- vasive for routine use and are insensitive.
emy of Neurology.) It would be preferable to measure acral
adrenergically mediated vasoconstriction in
response to the above-mentioned infusion.
However, recordings of vasoconstriction of
by 50%.3 This reflex, called the venoarterio- the muscle bed are indirect. Plethysmogra-
lar reflex, has receptors in small veins, and its phy can be performed, but recorded flow is
neural pathway appears to be that of the sym- contaminated by skin blood flow.
pathetic C fiber local axon reflex (Fig.
35-2).3 The function of the reflex has been
suggested to be to increase total peripheral Plasma Norepinephrine
resistance, compensating by up to 45% of
the orthostatic decrease in cardiac out- Plasma norepinephrine results from a spill-
put.3'14 It may also reduce the orthostatic over of norepinephrine from adrenergic
increase in tissue fluid by adjusting the postganglionic nerve terminals. The supine
precapillary-to-postcapillary resistance ratio. value is an index of net sympathetic activ-
The venoarteriolar reflex was measured in ity20"22 and is affected by the rate of norepi-
the feet of patients with diabetes mellitus nephrine secretion and clearance.23 Plasma
and was reported to be reduced in those with norepinephrine has been used to differenti-
diabetic neuropathy.15 The value of this test ate postganglionic from preganglionic fail-
is its theoretical ability to examine the status ure. In a disorder in which the lesion is pre-
of sympathetic vasoconstrictor fibers at the ganglionic, resting supine norepinephrine
postganglionic level. values are normal, but the response to stand-
Moy et al.16 studied the venoarteriolar re- ing is absent because of failure of activation.
flex, quantitative sudomotor axon reflex In a postganglionic lesion, the supine nor-
test, and heart rate responses to deep breath- epinephrine plasma values are decreased if
ing and the Valsalva maneuver in 40 control the lesion is widespread. The disadvantage
subjects, 49 diabetic subjects, and 29 subjects of the method is its low sensitivity, largely be-
with other neuropathies. The mean vaso- cause 80% of released norepinephrine is
taken up again by the neuron and only 20% in blood pressure. Recendy, there has been
enters venous blood. One method of en- considerable interest in baroreflex gain in
hancing clinical utility is to measure the in- response to spontaneous or induced changes
traneuronal metabolite dihydroxyphenylgly- in blood pressure. Commonly used maneu-
col (DHPG). vers include spectral analysis, the Valsalva
maneuver, and the sequence method.28
Sustained Handgrip
SUMMARY
Sustained muscle contraction causes an in-
crease in systolic and diastolic blood pres- For noninvasive evaluation of autonomic
sure and heart rate. The stimulus derives function, tests of peripheral adrenergic
from exercising muscle, the reaction is re- function are not as accurate or as repro-
flexive in nature, and the increase in blood ducible as tests of sudomotor and cardiova-
pressure is mediated by an increase in car- gal function. A noninvasive or minimally in-
diac output and peripheral resistance.24 The vasive method for evaluating muscle blood
test has been adapted as a simple test of sym- flow needs to be developed.
pathetic autonomic failure.25 Ewing et al.25
recommend 30% maximal contraction for
up to 5 minutes. Many patients have diffi- REFERENCES
culty sustaining the test for 5 minutes, but 3
minutes is sufficient. The test evaluates gen- 1. Low PA, Caskey PE, Tuck RR, Fealey RD, Dyck PJ.
eralized rather than peripheral adrenergic Quantitative sudomotor axon reflex test in normal
and neuropathic subjects. Ann Neurol 14:573-580,
function. 1983.
2. Roddie 1C, Shepherd JT, Whelan RF. A compari-
son of the heat elimination from the normal and
Baroreflex Indices nerve-blocked finger during body heating. J Phys-
iol Lond 138:445-448, 1957.
3. Henriksen O. Local sympathetic reflex mechanism
Baroreflex indices evaluate the heart period in regulation of blood flow in human subcutaneous
(reciprocal of heart rate) responses to in- adipose tissue. Acta Physiol Scand Suppl 450:1-48,
duced increases and decreases in blood pres- 1977.
sure. Phenylephrine or norepinephrine can 4. Sundlof G, Wallin BG. Human muscle nerve sym-
pathetic activity at rest. Relationship to blood pres-
be used to increase blood pressure, and tilt sure and age. J Physiol (Lond) 274:621-637, 1978.
or trinitroglycerin can be used to decrease 5. Zoller RP, Mark AL, Abboud FM, Schmid PG, Heis-
it. One approach, adapted from the method tad DD. The role of low pressure baroreceptors in
of Korner, is to determine the range and reflex vasoconstrictor responses in man. J Clin In-
mean gain of the heart period.17 These two vest 51:2967-2972, 1972.
6. Celander O. The range of control exercised by the
indices express the range of heart period to 'sympathico-adrenal system': a quantitative study on
a moderate pressor-hypotensive stress, and blood vessels and other smooth muscle effectors in
the mean rate of change in heart period in the cat. Acta Physiol Scand Suppl 116:1-132, 1954.
response to sudden changes in blood pres- 7. Vendsalu A. Studies on adrenaline and noradrena-
line in human plasma. Acta Physiol Scand 49 (Suppl
sure.17 Another related approach, described 173):1-123, 1960.
by Pickering et al.,26 relates beat-to-beat sys- 8. Oparil S, Vassaux C, Sanders CA, Haber E. Role of
tolic blood pressure to heart rate. An alter- renin in acute postural homeostasis. Circulation
native approach to stimulating barorecep- 41:89-95, 1970.
9. Burton AC, Taylor RM. A study of the adjustment
tors is to use a neck chamber whose pressure of peripheral vascular tone to the requirements of
can be increased or decreased.27 Finally, the the regulation of body temperature. Am J Physiol
heart period responses to the decrease and 129:565-577, 1940.
increase in blood pressure during the Val- 10. Low PA. Laboratory evaluation of autonomic func-
salva maneuver cart be related to changes in tion. In Low PA (ed). Clinical Autonomic Disorders:
Evaluation and Management, 2nd ed. Lippincott-
blood pressure. Baroreflex gain to an in- Raven Publishers, Philadelphia, 1997, pp 179-208.
crease or decrease in blood pressure evalu- 11. Sandroni P, Benarroch EE, Low PA. Pharmacolog-
ates primarily the vagal responses to changes ical dissection of components of the Valsalva ma-
neuver in adrenergic failure. J Appl Physiol 71: macologic distinction of different orthostatic hy-
1563-1567, 1991. potension syndromes. Neurology 31:1-7, 1981.
12. Low PA, Denq JC, Opfer-Gehrking TL, Dyck PJ, 21. Wallin BG, Sundlof G, Eriksson BM, Dominiak P,
O'Brien PC, Slezak JM. Effect of age and gender Grobecker H, Lindblad LE. Plasma noradrenaline
on sudomotor and cardiovagal function and blood correlates to sympathetic muscle nerve activity in nor-
pressure response to tilt in normal subjects. Muscle motensive man. Acta Physiol Scand 111:69-73,1981.
Nerve 20:1561-1568, 1997. 22. Ziegler MG, Lake CR, Kopin IJ. The sympadietic-
13. Low PA. Composite autonomic scoring scale for lab- nervous-system defect in primary orthostatic hy-
oratory quantification of generalized autonomic potension. N Engl J Med 296:293-297, 1977.
failure. Mayo Clin Proc 68:748-752, 1993. 23. Polinsky RJ, Goldstein DS, Brown RT, Reiser HR,
14. Henriksen O, Skagen K, Haxholdt O, Dyrberg V. Kopin IJ. Decreased sympathetic neuronal uptake
Contribution of local blood flow regulation mech- in idiopathic orthostatic hypotension. Ann Neurol
anisms to the maintenance of arterial pressure in 18:48-53, 1985.
upright position during epidural blockade. Acta 24. Coote JH, Hilton SM, Perez-Gonzalez JF. The reflex
Physiol Scand 118:271-280, 1983. nature of the pressor response to muscular exer-
15. Rayman G, Williams SA, Spencer PD, Smaje LH, cise. J Physiol (Lond) 215:789-804, 1971.
Wise PH, Tooke JE. Impaired microvascular hy- 25. Ewing DJ, Irving JB, Kerr F, Wildsmith JA, Clarke
peraemic response to minor skin trauma in type I BF. Cardiovascular responses to sustained handgrip
diabetes. Br Med J 292:1295-1298, 1986. in normal subjects and in patients with diabetes
16. Moy S, Opfer-Gehrking TL, Proper CJ, Low PA. The mellitus: a test of autonomic function. Clin Sci Mol
venoarteriolar reflex in diabetic and other neu- Med 46:295-306, 1974.
ropathies. Neurology 39:1490-1492, 1989. 26. Pickering TG, Gribbin B, Petersen ES, Cunning-
17. Low PA, Walsh JC, Huang CY, McLeodJG. The sym- ham DJ, Sleight P. Effects of autonomic blockade
pathetic nervous system in diabetic neuropathy. A on the baroreflex in man at rest and during exer-
clinical and pathological study. Brain 98:341-356, cise. Circ Res 30:177-185, 1972.
1975. 27. Eckberg DL, Cavanaugh MS, Mark AL, Abboud FM.
18. Moorhouse JA, Carter SA, Doupe J. Vascular re- A simplified neck suction device for activation of
sponses in diabetic peripheral neuropathy. Br Med carotid baroreceptors. J Lab Clin Med 85:167-173,
J 1:883-888, 1966. 1975.
19. Smith AA, Dancis J. Exaggerated response to in- 28. Dawson SL, Robinson TG, Youde JH, et al. The re-
fused norepinephrine in familial dysautonomia. N producibility of cardiac baroreceptor activity as-
Engl J Med 270:704-707, 1964. sessed noninvasively by spectral sequence tech-
20. Polinsky RJ, Kopin IJ, Ebert MH, Weise V. Phar- niques. Clin Auton Res 7:279-284, 1997.
Chapter 36
THERMOREGULATORY SWEAT TEST
Robert D. Fealey
ROLE OF THERMOREGULATORY SWEAT METHOD

TESTING: CLINICAL SYNDROMES AND NORMAL THERMOREGULATORY SWEAT
PROBLEMS EVALUATED DISTRIBUTIONS
Small Fiber Neuropathy REPORTING RESULTS
Diabetic Peripheral Neuropathy DIFFICULTIES AND PITFALLS IN
Severity of Autonomic Failure: INTERPRETATION
Extrapyramidal Syndromes SUMMARY
From a physiologic viewpoint, the thermoreg- feet syndrome. Although the results of nerve
ulatory sweat test consists of giving a con- conduction and electromyographic (EMG)
trolled heat stimulus to the body in a toler- studies and neurologic examination are of-
able fashion to produce a generalized sweat ten normal in these patients, findings on
response (that is, recruiting all areas of skin tests of skin autonomic innervation are usu-
capable of sweating). The test assesses the in- ally abnormal. The thermoregulatory sweat
tegrity of efferent (central and peripheral) test has been investigated in this syndrome,
sympathetic sudomotor pathways. Central and it shows impaired sweating in the distal
(preganglionic) structures tested include limbs of 72% of patients in whom the syn-
the hypothalamus, bulbospinal projections, drome had been diagnosed.2 The value of
intermediolateral cell column in the tho- the test in characterizing the distribution of
racic and upper lumbar spinal cord, and neuropathy has been emphasized recently.3
white rami. Peripheral (postganglionic) au-
tonomic structures tested include the sym-
pathetic chain ganglia and postganglionic
sudomotor axons and the sweat glands. Be- Diabetic Peripheral Neuropathy
cause the entire anterior body surface is
tested for both preganglionic and postgan- Diabetes mellitus produces distinct periph-
glionic lesions, the test is well suited for eral neurologic disorders, including length-
screening patients with certain clinical symp- dependent axonal neuropathy, painful
toms or for demonstrating autonomic in- truncal radiculopathy, and autonomic neu-
volvement in many disorders.1 ropathy.4-5 The ability to examine the entire
surface of the anterior body in detail and the
common involvement of sudomotor axons
ROLE OF THERMOREGULATORY in diabetes makes the thermoregulatory
sweat test particularly suited to the evalua-
SWEAT TESTING: CLINICAL tion of this disorder. The test examples in
SYNDROMES AND PROBLEMS Figure 36-1 are all from patients with dia-
EVALUATED betes, and the descriptions below refer to
this figure.
Small Fiber Neuropathy Peripheral neuropathy initially produces
distal sweat loss in the lower extremities, and
Peripheral neuropathies affecting small- as the neuropathy advances, the finger tips
diameter nerve fibers often cause burning and the lower anterior abdomen are affected
458
Thermoregulatory Sweat Test 459
many neurologic disorders. An important

use involves the evaluation of patients pre-
senting with an extrapyramidal syndrome.
The differential diagnosis often involves dif-
ferentiating multiple system atrophy, with
its severe autonomic failure, widespread an-
hidrosis, and poor prognosis, from Parkin-
son's disease, with mild or no autonomic in-
volvement on the thermoregulatory sweat
test,7'8 and better prognosis and response to
treatment. A patient with early multiple sys-
tem atrophy often exhibits a preganglionic,
segmental sweat deficit consistent with a cen-
tral lesion affecting the intermediolateral
cell columns of the spinal cord.1>9>1°
Other clinical disorders in which the ther-
moregulatory sweat test can provide diag-
nostic information are secondary autonomic
failure caused by neuropathy (primary sys-
temic amyloidosis,11 subacute autonomic,12'13
paraneoplastic disease,14 and leprosy1),
myelopathy,15 surgical sympathectomy, and
chronic idiopathic anhidrosis.16"19
Figure 36-1. a-e, Sweat distribution patterns: examples

from patients with diabetes mellitus. Black, areas of METHOD
sweating.
Thermoregulatory sweating is influenced by
the mean and local skin temperature as well
(Fig. 36-1, patient a). Painful truncal radicu- as by the central (blood or core) tempera-
lopathy has a distinct clinical presentation of ture.20 A maximal sweat response occurs
agonizing and, at times, lancinating pain as- when both central (oral) and mean skin tem-
sociated with cutaneous dysesthesia and a peratures are increased in a moderately hu-
characteristic thermoregulatory sweat test mid (about 35% relative humidity) environ-
pattern of patchy, asymmetric anhidrosis, pri- ment in which some degree of sweat
marily in the anterior distribution of one or evaporation can occur.1'20"22 Therefore,
several adjacent thoracic dermatomes, in the proper technique includes controlling the
distribution of the pain (Fig. 36-1, patient ambient air temperature and humidity as
c). Patterns relating to diabetic autonomic well as the patient's core and skin tempera-
nerve involvement include the autosympa- ture.
thectomy appearance of patient b in Figure Several techniques, including hot baths
36-1 or the severe global involvement ex- and infrared and incandescent heat lamps,
emplified by patient d. The percentage of have been used for the last 50 years to pro-
body surface anhidrosis correlates highly duce sweating, but the most satisfactory
with the degree of the symptoms and signs method is to use a cabinet in which the en-
of autonomic neuropathy.6 vironment is controlled and the entire body
(including the head) is heated. Guttmann23
described such a cabinet and demonstrated
Severity of Autonomic Failure: the usefulness of the thermoregulatory sweat
Extrapyramidal Syndromes test in the diagnosis and monitoring of
spinal cord and peripheral nerve lesions.
The thermoregulatory sweat test is helpful The thermoregulatory sweat test performed
in identifying autonomic involvement in in the Mayo Clinic Thermoregulatory Lab-
oratory is a modification of Guttmann's containing four baseboard heaters and a hu-

quinizarin sweat test1 and is described below. midifier and (2) overhead infrared heaters
The patient, covered only with a towel (s) that heat the skin and are carefully regulated
to maintain modesty, is placed in a supine by skin temperature feedback control. Arm-
position on a gurney and enclosed in the rests provide comfort for the patient as the
cabinet (Fig. 36-2A). The head end of the patient rests supine with palms down. Clear
cabinet is a clear vinyl curtain tented over Plexiglas windows that can be opened allow
the gurney; this arrangement allows the the technician to adjust the skin tempera-
head to be in the heated environment and ture probes if necessary and to observe the
still provide a clear view of the technician developing sweat distribution during the
and surroundings to minimize claustropho- test.
bia. Suitable variables for achieving a gen- Thermistor probes continuously measure
eralized sweat response within 60 minutes in- skin and oral temperatures during the test.
clude an ambient temperature of 48°C-50°C Sweating on the skin surface is visualized
and a relative humidity of 35%-40%, while best with an indicator powder that is applied
maintaining skin temperature between 39°C to the body before heating. A mixture of
and 40°C. The interior of the sweat cabinet alizarin red, cornstarch, and sodium car-
is shown in Figure 36-25. The major com- bonate in a 50:100:50 gram ratio, respec-
ponents are (1) the insulated cabinet walls tively, is suitable.1 It appears light orange on
Figure 36-2. Thermoregulatory sweat cabinet. A, Exterior, with temperature monitoring equipment. B, Interior.
nonsweating skin and turns dark purple on during the thermoregulatory sweat test was
sweating skin. Other indicators currently an increase of 1.2°C in 35-40 minutes.26 Be-
used include iodinated cornstarch24 and cause all subjects had sweat profusely at an
starch and iodine in solution.25 oral temperature of 38.0°C or less, we use
The average response of the oral temper- this temperature as a test end point. During
ature in 35 healthy control subjects (20-75 1998, the mean oral temperature increase in
years old) who achieved full-body sweating 767 patients was 1.5°C, with 38°C or a 1°C
Table 36-1. Sweat Test Procedure

1. Preheat and humidify the sweat cabinet by turning it on 45-60 minutes before the test. The set
points are 48°C air temperature and 37%-39% relative humidity. The overhead infrared heaters
can be set to 48°C to accelerate the preparation.
2. Meet the patient and briefly describe the test; have the patient undress.
3. Place towel (s) on the patient to cover as little skin as feasible to maintain modesty. Place
respirator mask on patient. Remove after powdering.
4. Apply the indicator powder (alizarin red, cornstarch, sodium carbonate).
5. Place elastic straps to hold the anterior thorax and thigh thermistor probes. Place the sterile
thermistor probe in its sponge holder, and place it in the patient's mouth between the gum and
cheek. Record baseline oral and skin temperatures. Check stability of oral temperature during
breathing and talking.
6. Turn down the overhead infrared heater control setting from 48°C (step 1) to 39°C. As quickly as
possible, open the vinyl curtain and place the patient and gurney in the cabinet; close the
curtain. Quickly place the feedback skin temperature probe next to the anterior thorax skin
monitoring probe under the elastic strap and directly under the infrared heater.
7. Connect the probes to the thermometer and allow 5 minutes for stabilization; begin timing the
sweat test.
8. Check the cabinet temperature and humidity every 5 minutes; operating conditions have to be
as follows to ensure an adequate heat stimulus:
Air temperature 45°C-50°C
Relative humidity 35%-40%
Skin temperature 39°C-40°C
9. Try to make all observations through the closed plastic doors; open the curtain if necessary to
wipe the patient's forehead.
10. If the patient sweats completely (fully saturating the powder on all skin areas) before the oral
temperature reaches 38°C, advise the physician and end the test. Otherwise, continue the test
until the oral temperature is 38°C or 1°C above the initial oral temperature (whichever is
greater). Do not exceed an oral temperature of 38.5°C or total heating time of 65 minutes.
11. At the end of the test, turn off the heat, remove the probes, and take the patient out of the
cabinet. Obtain digital photos, plot the pattern of sweating on the report form and call the
physician to inspect the sweat distribution, and prepare the report. Help the patient to shower.
Keep the laboratory neat by making sure the shower curtain is inside the stall and by vacuuming
any loose powder. Wash the skin probes in warm soap and water; wash off the oral probe and
put it in its pouch for sterilization.
12. Put fresh linen on the gurney for the next test.
13. Check to see that the patient has removed most of the powder in the shower, and help him or
her to dress if necessary. Place all used towels and linen in the laundry bag.
14. Enter the test data into the laboratory computer; draw the final body sweat distribution image
and calculate or measure the percentage of anhidrosis. Print out the report and put it in the
patient's medical record.
15. When performing a second sweat test on the same patient, ensure that the same end-point
temperature and increase in oral temperature are achieved to allow comparison of resulting
sweat distributions.
increase above baseline (whichever yielded

the higher temperature) as an end point.
These observations indicate that the often-
quoted 1°C temperature increase criterion
for an adequate thermoregulatoiy sweat
test27 is inadequate for patients with a low
(that is, less than 37°C) initial temperature.
If generalized sweating occurs at a lower
body temperature, the test can be ended be-
fore 38°C is reached.
For reasons of patient comfort and safety,
we do not increase oral temperature above
38.5°C or extend the heating period beyond
65 minutes. The current procedure followed
by technicians in the Mayo Clinic Ther-
moregulatory Laboratory is described in
Table 36-1.
NORMAL THERMOREGULATORY
SWEAT DISTRIBUTIONS
The normal variants in sweat distribution ob- Figure 36-3. Normal thermoregulatory sweating pat-
served in our laboratory in more than 35 terns, types 1-3. Black, areas of sweating. (From Fealey
et al.6 By permission of Mayo Foundation for Medical
healthy control subjects (16 women) from Education and Research.)
20 to 75 years old are shown in Figure 36-3.
Areas of "normal" anhidrosis may occur over
bony prominences (for example, the patella
and clavicle) and, with the subject supine, in matomal borders or spinal cord levels
the lateral calves and inner thighs. The prox- (Fig. 36-16).
imal extremities frequently show less sweat- 3. Focal sweat loss is confined to isolated
ing than the distal, but left-right symmetry dermatomes, peripheral nerve territo-
is always the rule. Males tend to show the ries, or small localized areas of skin
type 1 (heavy generalized sweating) pattern, (Fig. 36-1 c).
whereas females usually show the type 2 4. Global anhidrosis, by definition, occurs
(heavy generalized sweating but less in when more than 80% of the body sur-
proximal extremities) or type 3 (generalized face is involved (Fig. 36-1 d).
sweating but less in proximal arms and lower 5. Regional anhidrosis refers to large an-
body) pattern. Elderly men and women tend hidrotic areas (but less than 80%) that
to have types 2 and 3, and the lighter sweat- blend gradually into sweating areas and
ing areas may have a higher threshold of ac- that may or may not be contiguous; an-
tivation.28 hidrosis of the trunk alone and an-
We have described seven types of ther- hidrosis of the proximal parts of all
moregulatoiy sweat patterns or distributions four extremities are examples of this
that are used to report test results.6 pattern.
1. Distal anhidrosis is characterized by 6. Mixed patterns are combinations of any
sweat loss greatest in the fingers, the of the above in the same patient, for
legs below the knees, and the lower an- example, focal and distal patterns of
terior abdomen (Fig. 36-1 a). anhidrosis (Fig. 36—Ic).
2. Segmental anhidrosis involves large con- 7. A normal sweat distribution has no areas
tiguous zones of the body surface bor- of anhidrosis or minor areas of sweat
dered by areas of normal sweating; loss as observed in control subjects, as
these usually respect sympathetic der- previously described (Fig. 36-1 e).
REPORTING RESULTS [TST]%),1>6 as described below. When

printed on a color ink jet printer, the figure
Data about the name, identity number, clin- also provides a permanent record of the
ical problem of the patient, and the date of sweat distribution. The figure and camera im-
the thermoregulatory sweat test are indi- ages are reviewed by the reporting physician.
cated on the report (Fig. 36-4). A custom computer program is used to
The cabinet temperature and humidity count the number of anhidrotic pixels in the
ranges, the amount of time the patient was body image. This amount is divided by the
exposed to the heat stress, and the initial and total number of pixels constituting the an-
final oral temperatures are also indicated. terior body surface drawing and multiplied
The main part of the report includes a by 100 to produce the percentage of anhidro-
brief anatomical description of the results sis. Another accurate method for deriving
and states the clinical significance of the the TST% is to use planimetry (LASICO,
findings. The sweat distribution or pattern Model 1252 M; resolution = 0.005 cm2).
and the percentage of anhidrosis also appear The drawn body image is scrutinized, and
on the report. the regions of anhidrosis are oudined and
The anatomical figure is generated from a measured; individual areas are summated to
standard digital body image that is modified produce a total anterior body surface esti-
by the technician to look just like the color mate. The TST% provides for quantitation
digital camera images taken of the patient at of the test result and can be useful in clini-
the end of the test. This figure is used to cal- cal studies6""8 and in following the clinical
culate the percentage of anterior body sur- course of patients with autonomic disorders
face anhidrosis (thermoregulatory sweat test (Fig. 36-5).
THERMOREGULATORY SWEAT TEST

Clinic # Date 10/3/88
Name
Clinical Problem: Reports loss of sweating from the right arm and face Diagnosis 7030
Cabinet temperature range: 43.0 to 43.5 °C
Cabinet humidity range: 40 to 45 %
Time 28 min.
Oral temperature:
before test. . . 36.8 °C
after test.. . .38.0 °C
Body surface anhidrosis %. . . 24
Pattern of anhidrosis MIXED

RESULTS:
Sweating was completely absent in the T2 through T4
sympathetic dermatomes on the right. There was distal
sweat loss in the left hand and both feet as well.
Results indicate a pre or postganglionic segmental
lesion of the upper sympathetic chain on the right. The distal
loss suggests a mild peripheral neuropathy.
Physician Robert D. Fealey M.D.

Figure 36-4. Thermoregulatory sweat test report of a patient with a right sympathetic chain lesion (Pancoast tu-
mor) and a peripheral neuropathy. Black, areas of sweating.
not be taken for 48 hours before the test is

performed.
The application of skin lotions such as
moisturizing creams may produce a discol-
oration of alizarin-covered skin, making it
difficult to discern areas of anhidrosis. Con-
sequently, the use of lotions is prohibited on
the day of the test.
Anhidrosis in elderly patients may pre-
sent an interpretative challenge, because the
effect of aging on the autonomic nervous
system may be responsible for the regional
anhidrosis (most often affecting the lower
half of the body and proximal arms) that is
seen in some older women who have nor-
mal findings on neurologic examination but
report having dry skin for many years.
Whether this represents a variant of chronic
idiopathic anhidrosis18 or a loss of func-
tional sweat glands or reflects the loss of pre-
Figure 36-5. Serial thermoregulatory sweat tests on a ganglionic autonomic neurons known to
patient with subacute autonomic neuropathy, an im- occur with aging30 is unclear. Aging is asso-
mune-mediated ganglionopathy showing spontaneous ciated with decreasing postganglionic sym-
recovery of a substantial preganglionic deficit. A, 1997 pathetic sudomotor responses in the distal
test results. The percentage of anterior body surface an-
hidrosis (thermoregulatory sweat test %) was 94%, and
leg and foot.31 In our control subjects (19
the patient had severe constipation and colonic hypo- men, 16 women) 20-75 years old, there was
motility. B, 1998 test results. The gastrointestinal tract no significant regression of the percentage
dysautonomia had recovered spontaneously, and the of anhidrosis with age, although the mean
thermoregulatory sweat test % was only 4%. Black, ar- age was relatively young (52 years).6 Senile
eas of sweating. (Modified from Vernino et al.14 By per-
mission of the Massachusetts Medical Society.) atrophy of the skin and differences in sweat-
gland training between young and old may
also be a factor. Our current view is to in-
terpret the anhidrosis of the elderly as ab-
normal and deserving of additional evalua-
DIFFICULTIES AND PITFALLS tion to find a cause.
IN INTERPRETATION Potential problems with the thermoregu-
latory sweat test include the untidiness and
The interpreter must be aware of the nor- duration of the test, the possibility of skin
mal patterns of thermoregulatory sweating, heat injury, or skin irritation caused by the
including areas where anhidrosis may be indicator powder alizarin. Contact dermati-
seen normally, such as over bony promi- tis occurs only rarely (observed frequency,
nences or the lateral calves (Fig. 36-3). 5:1000 subjects) and is readily treated with
Patients who have worn pressure wraps oral and topical agents. A more common but
(that is, Ace bandages and abdominal not harmful problem is the persistence of
binders) within 12 hours before the test may purple discoloration of small areas of skin;
show anhidrosis in the areas that had been it may take several days for the color to wash
covered. This usually is recognized by the off. Because of repeated exposure to the in-
straight edges of the deficit. dicator powder, technicians in the Mayo
Severely dehydrated patients may sweat Clinic Thermoregulatory Laboratory wear
less overall,29 but they generally do not have masks, gloves, and goggles when applying
focal defects. Anticholinergic drugs, includ- the powder to minimize inhalation, oral in-
ing most tricyclic antidepressants, may ingestion, or contact with the eyes. Patients
hibit thermoregulatory sweating and should who are extremely claustrophobic, who in-
dicate a history of severe contact dermatitis, 8. Cohen J, Low P, Fealey R, Sheps S, Jiang NS. So-
or who are younger than 12 years are not matic and autonomic function in progressive auto-
nomic failure and multiple system atrophy. Ann
tested. Neurol 22:692-699, 1987.
9. Kihara M, SugenoyaJ, Takahashi A. The assessment
of sudomotor dysfunction in multiple system atro-
phy. Clin Auton Res 1:297-302, 1991.
SUMMARY 10. Fealey RD. Combined use of the Thermoregulatory
Sweat Test and the Quantitative Sudomotor Axon
The thermoregulatory sweat test assesses the Reflex Test clearly distinguishes autonomic failure
integrity of central and peripheral efferent due to multiple system atrophy from the dysau-
tonomia of Parkinson's disease (abstract). Neurol-
sympathetic sudomotor neural pathways. A ogy 56 (Suppl 3):A422, 2001.
controlled heat and humidity stimulus is 11. Wang AK, Fealey RD, Gehrking TL, Low PA. Auto-
given to produce a generalized sweating re- nomic failure in amyloidosis (abstract). Neurology
sponse in all skin areas capable of sweating. 52 (Suppl 2):A388, 1999.
Sweating is visualized by placing an indica- 12. Suarez GA, Fealey RD, Camilleri M, Low PA. Idio-
pathic autonomic neuropathy: clinical, neurophys-
tor powder on the skin beforehand. The en- iologic, and follow-up studies on 27 patients. Neu-
tire anterior body surface can be examined, rology 44:1675-1682, 1994.
and abnormalities can usually be detected at 13. Vernino S, AdamskiJ, Kryzer TJ, Fealey RD, Lennon
a glance. The clinical situations effectively VA. Neuronal nicotinic ACh receptor antibody in
subacute autonomic neuropathy and cancer-re-
evaluated with this test, the characteristic lated syndromes. Neurology 50:1806-1813, 1998.
normal and abnormal sweat distributions, 14. Vernino S, Low PA, Fealey RD, Stewart JD, Farru-
the methods to prepare a report of the test gia G, Lennon VA. Autoantibodies to ganglionic
results, including a technique to quantify the acetylcholine receptors in autoimmune autonomic
response known as the percentage of anhidro- neuropathies. N Engl J Med 343:847-855, 2000.
15. Fealey RD. Autonomic dysfunction in spinal cord
sis, are described herein. Important variables disease. In Critchley E, Eisen A (eds). Spinal Cord
of the heat stimulus, the patient's oral and Disease: Basic Science, Diagnosis and Management.
skin temperature response, and pitfalls in Springer-Verlag, London, 1997, pp 219-227.
the interpretation of the sweat test results are 16. Ando Y, Fujii S, Sakashita N, Uchino M, Ando M.
Acquired idiopathic generalized anhidrosis: clinical
described. manifestations and histochemical studies. J Neurol
Sci 132:80-83, 1995.
17. Faden AI, Chan P, Mendoza E. Progressive isolated
segmental anhidrosis. Arch Neurol 39:172-175,
REFERENCES 1982.
18. Low PA, Fealey RD, Sheps SG, Su WP, Trautmann
1. Fealey RD. Thermoregulatory sweat test. In Low PA JC, Kuntz NL. Chronic idiopathic anhidrosis. Ann
(ed). Clinical Autonomic Disorders: Evaluation and Neurol 18:344-348, 1985.
Management, 2nd ed. Lippincott-Raven Publishers, 19. Fealey RD. Thermoregulatory failure. In Vinken PJ,
Philadelphia, 1997, pp 245-257. Bruyn GW (eds). Handbook of Clinical Neurology.
2. Stewart JD, Low PA, Fealey RD. Distal small fiber Vol. 75, series 31. Autonomic Nervous System. Part
neuropathy: results of tests of sweating and auto- II. Dysfunctions. Elsevier Science Publishers, Ams-
nomic cardiovascular reflexes. Muscle Nerve 15: terdam, 2001, pp 53-84.
661-665, 1992. 20. McCaffrey TV, Wurster RD, Jacobs HK, Euler DE,
3. Dyck PJ, Grant LA, Fealey RD. Ten steps in charac- Geis GS. Role of skin temperature in the control of
terizing and diagnosing patients with peripheral sweating. J Appl Physiol 47:591-597, 1979.
neuropathy. Neurology 47:10-17, 1996. 21. SugenoyaJ, Ogawa T. Characteristics of central su-
4. Dyck JB, Dyck PJ. Diabetic polyneuropathy. In Dyck domotor mechanism estimated by frequency of
PJ, Thomas PK (eds). Diabetic Neuropathy, 2nd ed. sweat expulsions. JpnJ Physiol 35:783-794, 1985.
WB Saunders Company, Philadelphia, 1999, pp 22. Ogawa T, Asayama M. Quantitative analysis of the
255-278. local effect of skin temperature on sweating. Jpn J
5. Low PA, Fealey RD. Sudomotor neuropathy. In Physiol 36:417-422, 1986.
Dyck PJ, Thomas PK (eds). Diabetic Neuropathy, 23. Guttmann L. The management of the Quinizarin
2nd ed. WB Saunders Company, Philadelphia, Sweat Test (Q.S.T.). Postgrad Med J 23:353-366,
1999, pp 191-199. 1947.
6. Fealey RD, Low PA, Thomas JE. Thermoregulatory 24. Sato KT, Richardson A, Timm DE, Sato K. One-step
sweating abnormalities in diabetes mellitus. Mayo iodine starch method for direct visualization of
Clin Proc 64:617-628, 1989. sweating. Am J Med Sci 295:528-531, 1988.
7. Sandroni P, Ahlskog JE, Fealey RD, Low PA. Auto- 25. Khurana RK, Oculocephalic sympathetic dysfunc-
nomic involvement in extrapyramidal and cerebel- tion in posttraumatic headaches. Headache 35:
lar disorders. Clin Auton Res 1:147-155, 1991. 614-620, 1995.
26. Fealey RD. Much strain but no gain (in sweat out- in exercising humans. JAppl Physiol 51:1594-1600,
put, that is) (editorial). Clin Auton Res 11:225-226, 1981.
2001. 30. Low PA, Okazaki H, Dyck PJ. Splanchnic pregan-
27. Bannister R, Ardill L, Fentem P. Defective auto- glionic neurons in man. I. Morphometry of pre-
nomic control of blood vessels in idiopathic ortho- ganglionic cytons. Acta Neuropathol (Berl) 40:55-
static hypotension. Brain 90:725-746, 1967. 61, 1977.
28. Foster KG, Ellis FP, Dore C, Exton-Smith AN, 31. Low PA. The effect of aging on the autonomic ner-
Weiner JS. Sweat responses in the aged. Age Age- vous system. In Low PA (ed). Clinical Autonomic
ing 5:91-101, 1976. Disorders: Evaluation and Management, 2nd ed.
29. Fortney SM, Nadel ER, Wenger CB, BoveJR. Effect Lippincott-Raven Publishers, Philadelphia, 1997,
of blood volume on sweating rate and body fluids pp 161-175.
Chapter 37
CARDIOVAGAL AND OTHER
REFLEXES
Phillip A. Low
HEART RATE RESPONSE TO DEEP POWER SPECTRUM ANALYSIS

BREATHING THE VALSALVA MANEUVER
Physiologic Basis Normal Response and Physiologic Basis
Factors That Affect the Heart Rate Response Clinical Use of the Valsalva Maneuver As a
to Deep Breathing Test of Autonomic Function
Methods of Analysis Pitfalls of the Valsalva Maneuver
Reproducibility OTHER TESTS OF AUTONOMIC
Problems and Controversies FUNCTION
HEART RATE RESPONSE TO STANDING Pupil Cycle Time
OTHER TESTS OF CARDIOVAGAL SUMMARY
FUNCTION
Cardiovascular heart rate tests are a nonin- vascular heart rate tests, because the major
vasive, reliable, reproducible, and widely afferent and efferent pathways are both va-
used way to measure autonomic function in gal.6 The vagus nerve provides a beat-to-beat
human subjects. Although the physiologic control of the sinus node.7 This is mediated
basis of these autonomic maneuvers has by Ma type muscarinic cholinergic receptors
been known for a long time, it has been coupled to G-protein-activated inward rec-
mainly in the last two decades that these tests tifying potassium channels (GIRKs). In hu-
have been applied clinically, because of the mans, muscarinic receptor blockade with
enthusiastic promotion by Ewing and asso- atropine completely abolishes respiratory si-
ciates.1 Controversy exists about the defini- nus arrhythmia.
tion of an adequate battery.1'2 The pitfalls of The primary basis of respiratory sinus ar-
cardiovascular heart rate tests are under- rhythmia appears to be the interactions be-
appreciated.3"5 The following description tween the respiratory centers and cardioin-
focuses on the most useful tests. Their un- hibitory centers in the medulla, particularly
derlying physiology, clinical utility, and the nucleus ambiguus.8 Evidence for this is
shortcomings are described, followed by a cessation of vagal efferent activity during the
brief description of newer approaches. inspiratory phase of natural—but not artifi-
cial—ventilation, and the loss of respiratory
sinus arrhythmia in some patients with brain
HEART RATE RESPONSE TO stem lesions.9 Respiratory sinus arrhythmia
DEEP BREATHING is modulated by input from the lungs, heart,
and baroreceptors.10 Pulmonary stretch re-
Physiologic Basis ceptors that mediate the Hering-Breuer res-
piratory reflex modulate respiratory sinus
The heart rate response to deep breathing arrhythmia, although their role may be less
is probably the most reliable of the cardio- important in humans than in experimental
467
animals. Receptors from the right atrium ini- The sex of the subject does not greatly af-
tiate the vagally mediated Bainbridge reflex fect the heart rate response to deep breath-
and a venoatrial mechanoreceptor sympa- ing. No sex differences have been ob-
thetic reflex.8 Baroreflex sensitivity changes served.19'20 The amount of antecedent rest
throughout deep breathing, thus modulat- is not important in relaxed subjects. After
ing respiratory sinus arrhythmia.11 5 minutes of rest, another 25 minutes of
supine rest will not alter the response. No
significant differences have been found in
Factors That Affect the Heart Rate the response whether the test is performed
Response to Deep Breathing in the same subjects in the morning or in
the afternoon.22 One indirect effect of pro-
Numerous factors affect the heart rate re- longed hyperventilation is the reduction of
sponse to deep breathing. From the stand- Pcog, resulting in a depression in respiratory
point of clinical autonomic testing, the most sinus arrhythmia.
important of these are the effects of age and Basically, there have been two methods of
rate of forced respiration. A progressive de- respiratory cycling. The more commonly
crease in the response with increasing age used is to have the breathing of a subject
has been reported in all large studies.1'12"20 follow a pattern, usually a sine wave or an
We studied 376 control subjects of both oscillating bar, generated by a computer.
sexes, 10-83 years old, and reached a simi- The alternative is to instruct the subject to
lar conclusion.19'20 breathe in and out when instructed. The dif-
Maximal heart rate response to deep fering effect of the two approaches has not
breathing occurs at a breathing frequency of been studied systematically, but it is likely to
5-6 respirations per minute in normal sub- be minor.
jects;10'21"23 this observation forms the basis
for the standard test of deep breathing.2
However, the maximal heart rate response Methods of Analysis
to deep breathing is at lower frequencies in
patients with vagal neuropathy. Pragmati- The three methods of analysis generally used
cally, this variable is not a problem. Of are the heart rate range, the heart period range,
greater practical importance is the selection and the E:I ratio? The E:I ratio is the ratio of
for each subject of a respiratory rate in which the shortest RR interval during inspiration
the increase and decrease are additive in- to the longest RR interval during expiration.
stead of sub tractive.21 A clue to cancellation We prefer the heart rate range, because the
is the observation of a large first or second effect of resting heart rate on the range is
response, followed by smaller responses. smaller than its effect on heart period
There is a sympathetic modulation of the (Fig. 37-1). Weinberg and Pfeifer27 recom-
heart rate response to deep breathing that mended calculation of the circular mean re-
is inhibited by stress and enhanced by sultant, a method based on vector analysis,
/3-adrenergic blockade.24 Also, the response to eliminate the effects of trends in heart
is impaired during severe tachycardia, in rate over time and to attenuate the effect of
heart failure, and in deeply unconscious basal heart rate and ectopic beats in the cal-
patients.9'23'24 culated variability of heart rate.
The position of the subject has some ef-
fect on respiratory sinus arrhythmia. The re- Reproducibility
sponse is larger when the subject is supine
than when sitting or erect.22 The tests have been reproducible. Typical
A standardized deep-breathing protocol coefficients of variation have been 11%22
can be used, because depth of breathing and 8.9%.28
above a tidal volume of approximately 1.2 L
causes insignificant changes in the heart rate
response to deep breathing.25 Bennett and Problems and Controversies
associates22 and Eckberg26 found little or
no difference in the response for different The heart rate response to deep breathing
depths of respiration. is an indirect measure of cardiovagal func-
Cardiovagal and Other Reflexes 469
terference of phases. The reason subjects have

a decreased heart rate range less than 7 beats/
minute is because of negative interference.
HEART RATE RESPONSE

TO STANDING
The immediate heart rate response to stand-
ing can be recorded using an electrocardio-
gram (EGG) machine. In normal subjects,
tachycardia is maximal about the 15th beat
and relative bradycardia occurs around the
30th beat.30 The 30:15 ratio (RR interval at
beat 30)/(RR interval at beat 15) has been
recommended as an index of cardiovagal
function. Reflex tachycardia is thought to be
mediated by the vagus nerve, because the re-
sponse is abolished by atropine but not by
propranolol. A detailed evaluation of the
phases and mechanisms of the response to
standing has been reported.31 The initial
heart rate responses to standing consist of a
tachycardia at 3 seconds and at 12 seconds,
followed by a bradycardia at 20 seconds. The
initial cardioacceleration is an exercise re-
Figure 37-1. Heart rate response to deep breathing in flex, whereas the subsequent tachycardia
a normal subject (top) and in a patient with diabetic au-
tonomic neuropathy (bottom). In each panel, the three and bradycardia are baroreflex-mediated.
upper tracings correspond to systolic blood pressure
(SBP), mean blood pressure (MBP), and diastolic blood
pressure (DBF). The bottom tracing corresponds to
heart rate (HR). OTHER TESTS OF
CARDIOVAGAL FUNCTION
Potential tests of cardiovagal function are
tion. A reduced response indicates a lesion numerous and include heart rate response
anywhere in the complicated autonomic to lying down,32 squatting,33 coughing,34'35
nervous system, that is, in the afferent, cen- and facial immersion in cold water (the div-
tral processing unit, efferent, synapse, or ef- ing reflex) or application of cold compresses
fector apparatus. to the face (the cold face test) .36 After the sub-
To further complicate interpretation, a reject stands still for 3 minutes, squatting for
duced response does not unequivocally in- 1 minute produces a vagally mediated, at-
dicate cardiovagal failure. Heart rate usually ropine-sensitive lengthening of the RR in-
increases during inspiration and decreases terval, and standing from a squatting po-
during expiration, but even this observation sition elicits a sympathetically mediated
is not entirely correct. Both inspiration and shortening of the RR interval.33 Coughing
expiration are followed by an increase, then results in inspiration and an expiratory ef-
a decrease, in heart rate but at a different fort against a closed glottis, followed by an
rate of change, amplitude, time of appear- explosive expiration as the glottis suddenly
ance, and duration. Mehlsen and col- opens. The heart rate response consists of
leagues29 suggested that the reason the max- a cardioacceleration, which is maximal ap-
imal heart rate range in many subjects is 6 proximately 2-3 seconds after the last cough,
beats per minute is because they have well- and a return to the resting value in approx-
defined heart rate maxima with positive in- imately 12-14 seconds.34'35 The mechanism
is thought to be cholinergic and initially mal duration of recording is probably 5 min-

caused by muscle contraction followed by a utes of good quality recording.
baroreflex response to a decrease in blood
pressure.35 The diving reflex provoked by fa-
cial cooling consists of bradycardia, apnea,
and vasoconstriction. The bradycardia de- THE VALSALVA MANEUVER
pends on a trigeminal-cardiovagal reflex.36
The Valsalva maneuver is a global test of
reflex cardiovascular responses. It consists
of an abrupt transient increase in intratho-
POWER SPECTRUM ANALYSIS racic and intra-abdominal pressures induced
by blowing against a pneumatic resistance
Autonomic data generally are evaluated in while maintaining a predetermined pressure
the time domain, with a focus on the (straining)?9"42
changes in amplitude over time. The oscil-
lations also contain key information. Fre-
quency analysis focuses on the changes in
amplitude as a function of frequency.37 In Normal Response and
recordings of the heart period (the recipro- Physiologic Basis
cal of heart rate), oscillations at the respira-
tory frequency (typically approximately 0.25 Intra-arterial recordings of arterial pressure
Hz) are caused by parasympathetic function; and, more recently, noninvasive monitoring
hence, its power (amplitude) reflects para- of arterial pressure with a photoplethysmo-
sympathetic function. A slower frequency, graphic technique (Finapres) or tonometry
approximately 0.07-0.1 Hz, reflects the pe- (Colin Pilot or Colin 7000) have provided
riodicity of the baroreflex loop. Power at this important information about the hemody-
frequency reflects both sympathetic and namic changes during the Valsalva maneu-
parasympathetic function. Similar oscilla- ver in normal and pathologic conditions.43'44
tions occur in blood pressure recordings. The responses to the Valsalva maneuver
Several methods are available for evaluat- have been divided into four phases39"45 (Fig.
ing autonomic signals in the frequency do- 37-2). Phase I consists of a brisk increase in
main. Fast Fourier transform and autore- systolic and diastolic arterial pressure and a
gressive models are commonly used. Both of decrease in heart rate immediately after the
these require stationarity, a condition that is onset of the Valsalva strain and lasts ap-
difficult to satisfy with changing autonomic proximately 4 seconds. The increase in ar-
signals. An alternative approach is time- terial pressure during phase I reflects me-
frequency analysis, a method that resolves chanical factors and is not associated with an
signals in both the time and frequency do- increase in sympathetic activity. It persists in
mains simultaneously. patients with transections of the high cervi-
Head-up tilt results in attenuation of the cal spinal cord and in normal subjects after
respiratory frequency and augmentation of administration of ai-adrenergic blocking
the lower frequency. An advantage of fre- drugs.45 The slowing of the heart rate is re-
quency analysis is its ability to evaluate sym- flexive and mediated by increased parasym-
pathovagal balance. It can be expressed as pathetic efferent activity.46
the power in the low frequency in blood Phase II consists of a decrease (early phase
pressure (reflecting pure sympathetic func- II, IIe) and subsequent partial recovery (late
tion) over the respiratory frequency in heart phase II, Hi) of arterial pressure and con-
period (pure parasympathetic function).38 tinuous increase of heart rate during strain-
For clinical recordings of autonomic sig- ing. Continuous straining impedes venous
nals, it is essential that respiration be re- return to the heart and results in the dis-
corded or paced, because respiratory rhythms placement of large amounts of blood from
are entrained with respiration and if the sub- the thorax and abdomen to the limbs. The
ject breathes slowly, respiratory rhythms will decrease in venous return produces a re-
contaminate the lower frequencies. A mini- duction in left atrial and left ventricular di-
late phase II is abolished by ai-adrenergic

blockade with phentolamine.50 Blood pres-
sure during phase II decreases more if
compensatory tachycardia is prevented by
atropine and propranolol and if vasocon-
striction is prevented by ai-adrenergic block-
ade.50
Phase III consists of a sudden, brief (1-2
seconds) further decrease in arterial pres-
sure and increase in heart rate immediately
after the release of the straining. It is essen-
tially mechanical in nature.
Phase IV is characterized by increased sys-
tolic and diastolic arterial pressure above
control levels, called overshoot, accompanied
by bradycardia relative to the control level
of heart rate. In phase IV, venous return to
the heart, left ventricular stroke volume,
and cardiac output return toward normal,
whereas the arteriolar bed remains vasocon-
stricted because of the long time constant of
sympathetic responses.45 This combination
results in an overshoot of arterial pressure
above control values. Poststraining arterial
pressure increases are proportional to the
preceding increases in sympathetic nerve ac-
Figure 37-2. Changes in systolic blood pressure (SBP) tivity. The increase in arterial pressure dur-
and heart rate (HR) during the Valsalva maneuver in a ing phase IV can be prevented by j8-adren-
normal subject (top) and in patient with diabetic auto- ergic blockade.50 Increases in both cardiac
nomic failure (bottom). The normal beat-to-beat SBP output and total peripheral resistance are
recording shows the typical four phases (I, II, HI, IV)
of the Valsalva maneuver. The abnormal Valsalva ma-
important in producing the increase in ar-
neuver is characterized by a profound decrease in SBP terial pressure in phase IV.
in early phase II, absence of recovery in late phase II, Recent pharmacologic evidence indicates
and absence of SBP overshoot in phase IV. Note the at- that an increase in cardiac-output-mediated
tenuated HR responses during phases I and IV, result- cardioacceleration is more important than
ing in a reduced Valsalva ratio. EP, expiratory pressure.
vasoconstriction in producing arterial pres-
sure overshoot in phase IV. This overshoot
is abolished by /3-blockade with propranolol
but is maintained or even exaggerated dur-
mansions, left ventricular stroke volume, ing a-adrenergic blockade with phentol-
and cardiac output.39'40 This triggers reflex amine.40 The increase in arterial pressure
compensator)' tachycardia and vasoconstric- during phase IV stimulates the barorecep-
tion. The tachycardia during phase II results tors and results in reflex bradycardia caused
from a prominent early component of inhi- by increased parasympathetic activity, which
bition of cardiovagal output and is abolished is abolished with atropine.46'51 Sympathetic
with muscarinic blockade with atropine.47 inhibition after straining persists much
There is also a late contribution of increased longer than the increase in arterial pressure.
sympathetic cardioacceleratory output that The responses during the different phases
is blocked with propranolol. The progressive of the Valsalva maneuver depend on the vari-
recovery of arterial pressure during phase II able relationships between carotid and aor-
reflects a similarly progressive increase in to- tic baroreceptor inputs, and pressure tran-
tal peripheral resistance45 caused by in- sients lasting only seconds may reset the
creased sympathetic vasoconstrictor activ- relationships between the arterial pressure
ity.48'49 Increased arterial pressure during and the sympathetic or vagal responses.52
Clinical Use of the Valsalva Effects of Posture

Maneuver As a Test of In subjects in the supine position, changes
Autonomic Function in arterial pressure during phases II and IV
may be modest, because the large intratho-
TECHNIQUE racic blood volume may buffer the reduced
For testing the responses to the Valsalva ma- venous return during phase II. In the supine
neuver, care should be taken, as in any other position, some normal subjects may show a
autonomic test, to ensure that the patient is square-wave response similar to that of pa-
well hydrated and is not taking medications tients with congestive heart failure. The mag-
known to affect blood volume, cholinergic nitude of arterial pressure decrease during
function, or vasoreactivity. At our institution, phase II, subsequent systolic blood pressure
subjects are tested in the supine position and overshoot in phase IV, and the Valsalva ra-
asked to maintain a column of mercury of tio increase with a change to the sitting and,
40 mm Hg for 15 seconds through a bugle particularly, the standing position.55
with an air leak (to ensure an open glottis).
The responses are obtained in triplicate, and Effects of Test Duration
the largest response is accepted.43'50'53 In
most laboratories, only heart rate is moni- The duration of straining during the Val-
tored continuously. \Ve43>50>53 and others54'55 salva maneuver has different effects in va-
also monitor beat-to-beat arterial pressure gally and sympathetically mediated re-
with a noninvasive photoplethysmographic sponses, because of their different latencies
technique. The "normal" Valsalva response and time constants. When the Valsalva ma-
should be defined according to the tech- neuver is performed at low expiratory pres-
nique used in each laboratory, because sev- sures (20 mm Hg), the magnitude of the
eral technical variables affect the magnitude tachycardia in phase II is independent of the
of the response. The relationships between duration of the test, consistent with the short
arterial pressure and heart rate during phase latency of vagal responses. The maximal in-
II and phase IV of the maneuver have been crease in arterial pressure in late phases II
used to assess baroreflex sensitivity.56 and IV correlates with the duration of the
Valsalva maneuver. This may reflect the
TECHNICAL VARIABLES AFFECTING longer latency of sympathetic vasoconstric-
THE VALSALVA MANEUVER tor and cardioacceleratory responses.45 A
test duration of 10 seconds is effective, and
The cardiovascular changes during the Val- 15 seconds is a practical optimum and may
salva maneuver are determined mainly by be sufficient to assess sympathetically medi-
the magnitude of hemodynamic change dur- ated responses in a clinical setting.43
ing the forced expiratory effort, the time
course and efficiency of reflex cardiovagal
and sympathetic vasomotor and cardiomo- Effects of Expiratory Pressure
tor responses, and the modification of these
responses by interactions with respiratory The magnitude of most heart rate and arte-
mechanisms at both central and peripheral rial pressure responses during phases II and
levels. Accordingly, the responses to the Val- IV correlates with the magnitude of expira-
salva maneuver may be affected by (1) the tory pressure used during the Valsalva ma-
position of the subject during the maneuver, neuver. Maximal arterial pressure and heart
(2) the magnitude and duration of the rate responses are obtained with expiratory
straining, (3) the breathing pattern before pressures of 40-50 mm Hg.22'23'43'46'47'51
and after the maneuver, including depth
and phase of respiration preceding the Phase of Respiration
straining, and (4) the control of respiration
after release of the straining. Normative data In normal subjects, the magnitude of heart
on the phases of the Valsalva maneuver have rate responses during the Valsalva maneuver
been published recently.53 is significantly lower if the expiratory strain
is preceded by maximal inspiration instead neuver by only testing the Valsalva ra-
of tidal inspiration.46 tio is unreliable, because the magni-
tude and time course of the heart rate
THE VALSALVA RATIO response may be normal despite a re-
sponse of arterial pressure typical of
The Valsalva ratio is defined as the longest sympathetic failure.58
RR interval (pulse interval, in milliseconds) The integrity of reflex sympathetic re-
after the maneuver (phase IV) to the short- sponses cannot be inferred on the basis of
est RR interval during the maneuver (phase the Valsalva ratio. Both the magnitude of the
II). In clinical settings, the Valsalva maneu- decrease in mean arterial pressure in phase
ver commonly has been used to calculate the II and the overshoot of arterial pressure in
Valsalva ratio. The best of three responses is phase IV have been considered indices of va-
accepted. In more than 96% of control sub- somotor function.45'50 However, the magni-
jects, this ratio exceeds 1.5.22,23,46,47,51 jt js tude of the decrease in arterial pressure in
not affected by sex, but it decreases signifi- early phase II is also affected by the heart
cantly with age.57 rate responses, and the overshoot of arter-
ial pressure in phase IV may be more de-
PITFALLS OF THE VALSALVA RATIO pendent on cardiac output.50 In our labo-
ratory, the changes in arterial pressure
There is evidence that the Valsalva ratio in during early and late phases II and IV are
normal subjects depends mainly on cardio- used to assess sympathetic vasomotor func-
vagal function.23'46'47'51 However, the inter- tion.50'53 Late phase II is impaired in pa-
pretation of this ratio as a test of cardiovagal tients with or-adrenergic failure caused by
function without simultaneous recordings of dopamine-/3-hydroxylase deficiency.59
arterial pressure may be misleading for sev-
eral reasons.
1. The Valsalva ratio correlates better with Pitfalls of the Valsalva Maneuver
the heart rate response in phase II than
with the response in phase IV.43 There- There are several pitfalls in performance of
fore, if sufficient tachycardia is present the Valsalva maneuver.42
in phase II, the Valsalva ratio may be 1. The test requires patient cooperation
"normal" even in the absence of sig- and, thus, cannot be performed in pa-
nificant bradycardia in phase IV. This tients who are seriously ill or who have
may occur in patients with cardiovagal weak respiratory, facial, or oropharyn-
impairment but intact sympathetic in- geal muscles.
nervation.58 2. The maneuver should be avoided in pa-
2. The Valsalva ratio also correlates with tients with proliferative retinopathy,
the magnitude of arterial pressure because of the risk of intraocular hem-
overshoot in phase IV.43 The absence orrhage.
of bradycardia in phase IV, and thus an 3. Theoretically, the Valsalva maneuver
abnormal Valsalva ratio, may be caused can precipitate arrhythmias and angina
not only by vagal dysfunction but also and may cause syncope, particularly in
by the inability to increase arterial pres- elderly patients with impaired reflex
sure in phase IV. mechanisms that respond to the de-
3. Both the heart rate increase in phase crease in venous return.
II and the heart rate decrease in phase 4. Patients with congestive heart failure,
IV are affected critically by the magni- mitral stenosis, aortic stenosis, con-
tude of the decrease in venous return strictive pericarditis, or atrial septal de-
during the Valsalva maneuver, which fect may have an abnormal square-wave
depends on the position of the subject response of arterial pressure to the Val-
and the pooling and buffering effect of salva maneuver because of the increase
thoracic vessels.55 in pulmonary blood volume, which is
4. Assessing the integrity of the total capable of maintaining ventricular fill-
baroreflex arc during the Valsalva ma- ing during the Valsalva strain.
OTHER TESTS OF AUTONOMIC ing) , has been considered a test of cardiac

FUNCTION parasympathetic function. However, with-
out simultaneous recording of arterial pres-
Pupil Cycle Time sure, this may be misleading. An exagger-
ated decrease in arterial pressure during
Infrared recording of dark-adapted pupil di- phase II suggests sympathetic vasomotor
ameter has been suggested as a good quan- failure, whereas an absence of overshoot
titative measure of sympathetic function, during phase IV indicates the inability to in-
and pupil cycle time has been suggested as crease cardiac output and cardiac adrener-
an index of parasympathetic function.60"62 gic failure.
Pupil cycle time refers to the frequency of os-
cillations of the pupil in response to a light
stimulus. Thompson,60 who introduced the
test, has advised caution with performing it. REFERENCES
Pupil cycle time is also affected by lesions of
the optic nerve, sympathetic lesions, lesions 1. Ewing DJ, Martyn CN, Young RJ, Clarke BF. The
of the myoneural junction, narrow-angle value of cardiovascular autonomic function tests: 10
glaucoma, and the level of retinal illumina- years experience in diabetes. Diabetes Care 8:491-
498, 1985.
tion.60 Also, it is less accurate when pupillary 2. Low PA, Pfeifer MA. Standardization of autonomic
movements are weak.60 function. In Low PA (ed). Clinical Autonomic
Disorders: Evaluation and Management, 2nd ed.
Lippincott-Raven Publishers, Philadelphia, 1997,
pp 287-295.
SUMMARY 3. van Lieshout JJ, Wieling W, Wesseling KH, Rare-
maker JM. Pitfalls in the assessment of cardiovas-
cular reflexes in patients with sympathetic failure
Noninvasive cardiovascular tests are reliable but intact vagal control. Clin Sci 76:523-528, 1989.
and reproducible and are widely used to 4. Wieling W, van Lieshout JJ. The assessment of
evaluate autonomic function in human cardiovascular reflex activity: standardization is
subjects. The heart rate response to deep needed. Diabetologia 33:182-183, 1990.
5. Low PA. Pitfalls in autonomic testing. In Low PA
breathing is probably the most reliable test (ed). Clinical Autonomic Disorders: Evaluation and
for assessing the integrity of the vagal affer- Management, 2nd ed. Lippincott-Raven Publishers,
ent and efferent pathways to the heart. This Philadelphia, 1997, pp 391-401.
response is usually tested at a breathing fre- 6. Katona PG, Jih F. Respiratory sinus arrhythmia:
noninvasive measure of parasympathetic cardiac
quency of 5 or 6 respirations per minute and control. JAppl Physiol 39:801-805, 1975.
decreases linearly with age. The Valsalva ma- 7. Jalife J, Michaels DC. Neural control of sinoatrial
neuver consists of a forced expiratory effort pacemaker activity. In Levy MN, Schwartz PJ (eds).
against resistance and produces mechanical Vagal Control of die Heart: Experimental Basis and
(phases I and III) and reflex (phases II and Clinical Implications. Futura Publishing Company,
New York, 1994, pp 173-205.
IV) changes in arterial pressure and heart 8. Spyer KM. Central nervous control of the cardio-
rate. When performed under continuous ar- vascular system. In Mathias CJ, Bannister R (eds).
terial pressure monitoring with a noninva- Autonomic Failure: A Textbook of Clinical Disor-
sive (Finapres) technique, the Valsalva ma- ders of the Autonomic Nervous System, 4th ed. Ox-
ford University Press, New York, 1999, pp 45-55.
neuver provides valuable information about 9. Vallbona C, Cardus D, Spencer WA, Hoff HE. Pat-
the integrity of the cardiac parasympathetic, terns of sinus arrhythmia in patients with lesions of
cardiac sympathetic, and sympathetic vaso- the central nervous system. Am J Cardiol 16:379-
motor outputs. The responses to the Valsalva 389, 1965.
maneuver are affected by the position of the 10. Saul JP, Cohen RJ. Respiratory sinus arrhythmia. In
Levy MN, Schwartz PJ (eds). Vagal Control of the
subject and the magnitude and duration of Heart: Experimental Basis and Clinical Implica-
the expiratory effort. In general, it is per- tions. Futura Publishing Company, New York, 1994,
formed at an expiratory pressure of 40 mm pp 511-536.
Hg sustained for 15 seconds. The Valsalva ra- 11. Eckberg DL, Kifle YT, Roberts VL. Phase relation-
ship between normal human respiration and
tio, the relationship between the maximal baroreflex responsiveness. J Physiol (Lond) 304:
heart rate response during phase II (strain- 489-502, 1980.
ing) and phase IV (after release of the strain- 12. Bergstrom B, Lilja B, Rosberg K, Sundkvist G. Au-
tonomic nerve function tests. Reference values in 30. Ewing DJ, Campbell IW, Murray A, Neilson JM,
healthy subjects. Clin Physiol 6:523-528, 1986. Clarke BF. Immediate heart-rate response to stand-
13. Ingall TJ. Autonomic nervous system function in ing: simple test for autonomic neuropathy in dia-
ageing and in diseases of the peripheral nervous betes. Br Med J 1:145-147, 1978.
system. Thesis, University of Sydney, 1986. 31. Borst C, Wieling W, van Brederode JF, Hond A, de
14. Kaijser L, Sachs C. Autonomic cardiovascular re- Rijk LG, Dunning AJ. Mechanisms of initial heart
sponses in old age. Clin Physiol 5:347-357, 1985. rate response to postural change. Am J Physiol
15. Masaoka S, Lev-Ran A, Hill LR, Vakil G, Hon EH. 243:H676-H681, 1982.
Heart rate variability in diabetes: relationship to age 32. Bellavere F, Ewing DJ. Autonomic control of the
and duration of the disease. Diabetes Care 8:64-68, immediate heart rate response to lying down. Clin
1985. Sci 62:57-64, 1982.
16. O'Brien IA, O'Hare P, Corrall RJ. Heart rate vari- 33. Marfella R, Giugliano D, di Maro G, Acampora R,
ability in healthy subjects: effect of age and the de- Giunta R, D'Onofrio F. The squatting test. A use-
rivation of normal ranges for tests of autonomic ful tool to assess both parasympathetic and sympa-
function. Br Heart J 55:348-354, 1986. thetic involvement of the cardiovascular autonomic
17. Wieling W, van Brederode JF, de Rijk LG, Borst C, neuropathy in diabetes. Diabetes 43:607-612,1994.
Dunning AJ. Reflex control of heart rate in normal 34. Wei JY, Rowe JW, Kestenbaum AD, Ben-Haim S.
subjects in relation to age: a data base for cardiac Post-cough heart rate response: influence of age,
vagal neuropathy. Diabetologia 22:163-166, 1982. sex, and basal blood pressure. Am J Physiol
18. Clark CV, Mapstone R. Age-adjusted normal toler- 245:R18-R24, 1983.
ance limits for cardiovascular autonomic function 35. Cardone C, Bellavere F, Ferri M, Fedele D. Auto-
assessment in the elderly. Age Ageing 15:221-229, nomic mechanisms in the heart rate response to
1986. coughing. Clin Sci 72:55-60, 1987.
19. Low PA, Opfer-Gehrking TL, Proper CJ, Zimmer- 36. Khurana RK, Watabiki S, Hebel JR, Toro R, Nelson
man I. The effect of aging on cardiac autonomic E. Cold face test in the assessment of trigeminal-
and postganglionic sudomotor function. Muscle brain stem-vagal function in humans. Ann Neurol
Nerve 13:152-157, 1990. 7:144-149, 1980.
20. Low PA, Denq JC, Opfer-Gehrking TL, Dyck PJ, 37. Novak V, Novak P, Low PA. Time-frequency analy-
O'Brien PC, Slezak JM. Effect of age and gender sis of cardiovascular function and its clinical appli-
on sudomotor and cardiovagal function and blood cations. In Low PA (ed). Clinical Autonomic Dis-
pressure response to tilt in normal subjects. Muscle orders: Evaluation and Management, 2nd ed.
Nerve 20:1561-1568, 1997. Lippincott-Raven Publishers, 1997, pp 323-348.
21. Angelone A, Coulter NAJr. Respiratory sinus ar- 38. Novak V, Novak P, Opfer-Gehrking TL, Low PA.
rhythmia: a frequency dependent phenomenon. J Postural tachycardia syndrome: time frequency
Appl Physiol 19:479-482, 1964. mapping. J Auton Nerv Syst 61:313-320, 1996.
22. Bennett T, Fentem PH, Fitton D, Hamptom JR, 39. Booth RW, Ryan JM, Mellet HC, Swiss E, Neth E.
Hosking DJ, Riggott PA. Assessment of vagal con- Hemodynamic changes associated with the Valsalva
trol of the heart in diabetes. Measures of R-R in- maneuver in normal men and women. J Lab Clin
terval variation under different conditions. Br Med 59:275-285, 1962.
Heart J 39:25-28, 1977. 40. Brooker JZ, Alderman EL, Harrison DC. Alterations
23. Pfeifer MA, Cook D, BrodskyJ, et al. Quantitative in left ventricular volumes induced by Valsalva ma-
evaluation of cardiac parasympathetic activity in noeuvre. Br Heart J 36:713-718, 1974.
normal and diabetic man. Diabetes 31:339-345, 41. Sharpey-Schafer EP. Effects of Valsalva's manoeu-
1982. vre on the normal and failing circulation. BMJ
24. Coker R, Koziell A, Oliver C, Smith SE. Does the 1:693-695, 1955.
sympathetic nervous system influence sinus ar- 42. Nishimura RA, Tajik AJ. The Valsalva maneuver and
rhythmia in man? Evidence from combined auto- response revisited. Mayo Clin Proc 61:211-217,
nomic blockade. J Physiol (Lond) 356:459-464, 1986.
1984. 43. Benarroch EE, Opfer-Gehrking TL, Low PA. Use of
25. Freyschuss U, Melcher A. Sinus arrhythmia in man: the photoplethysmographic technique to analyze
influence of tidal volume and oesophageal pres- the Valsalva maneuver in normal man. Muscle
sure. Scand J Clin Lab Invest 35:487-496, 1975. Nerve 14:1165-1172, 1991.
26. Eckberg DL. Human sinus arrhythmia as an index 44. Imholz BP, van Montfrans GA, Settels JJ, van der
of vagal cardiac outflow. J Appl Physiol 54:961-966, Hoeven GM, Karemaker JM, Wieling W. Continu-
1983. ous non-invasive blood pressure monitoring: relia-
27. Weinberg CR, Pfeifer MA. An improved method for bility of Finapres device during the Valsalva ma-
measuring heart-rate variability: assessment of car- noeuvre. Cardiovasc Res 22:390-397, 1988.
diac autonomic function. Biometrics 40:855-861, 45. Korner PI, Tonkin AM, Uther JB. Reflex and me-
1984. chanical circulatory effects of graded Valsalva ma-
28. Smith SA. Reduced sinus arrhythmia in diabetic au- neuvers in normal man. J Appl Physiol 40:434-440,
tonomic neuropathy: diagnostic value of an age-re- 1976.
lated normal range. Br Med J 285:1599-1601,1982. 46. Eckberg DL. Parasympathetic cardiovascular con-
29. Mehlsen J, Pagh K, Nielsen JS, Sestoft L, Nielsen trol in human disease: a critical review of methods
SL. Heart rate response to breathing: dependency and results. Am J Physiol 239:H581-H593, 1980.
upon breathing pattern. Clin Physiol 7:115-124, 47. Leon DF, Shaver JA, Leonard JJ. Reflex heart rate
1987. control in man. Am Heart J 80:729-739, 1970.
48. Delius W, Hagbarth KE, Hongell A, Wallin BG. Ma- Wieling W. Assessment of cardiovascular reflexes:
noeuvres affecting sympathetic outflow in human influence of posture and period of preceding rest.
muscle nerves. Acta Physiol Scand 84:82-94, 1972. J Appl Physiol 68:147-153, 1990.
49. Wallin BG, Eckberg DL. Sympathetic transients 56. Kautzner J, Hartikainen JE, Camm AJ, Malik M. Ar-
caused by abrupt alterations of carotid barorecep- terial baroreflex sensitivity assessed from phase IV
tor activity in humans. Am J Physiol 242:H185- of the Valsalva maneuver. AmJ Cardiol 78:575-579,
H190, 1982. 1996.
50. Sandroni P, Benarroch EE, Low PA. Pharmacolog- 57. Low PA, Opfer-Gehrking TL, Proper CJ, Zimmer-
ical dissection of components of the Valsalva ma- man I. The effect of aging on cardiac autonomic
neuver in adrenergic failure. J Appl Physiol 71: and postganglionic sudomotor function. Muscle
1563-1567, 1991. Nerve 13:152-157, 1990.
51. Levin AB. A simple test of cardiac function based 58. Opfer-Gehrking TL, Low PA. Impaired respiratory
upon the heart rate changes induced by the Val- sinus arrhythmia with paradoxically normal Val-
salva maneuver. Am J Cardiol 18:90-99, 1966. salva ratio indicates combined cardiovagal and pe-
52. Smith ML, Beightol LA, Fritsch-Yelle JM, Ellenbo- ripheral adrenergic failure. Clin Auton Res
gen KA, Porter TR, Eckberg DL. Valsalva's maneu- 3:169-173, 1993.
ver revisited: a quantitative method yielding insights 59. Biaggioni I, Goldstein DS, Atkinson T, Robertson
into human autonomic control. Am J Physiol 271: D. Dopamine-beta-hydroxylase deficiency in hu-
H1240-H1249, 1996. mans. Neurology 40:370-373, 1990.
53. Denq JC, O'Brien PC, Low PA. Normative data on 60. Thompson HS. The pupil cycle time. J Clin Neu-
phases of the Valsalva maneuver. J Clin Neuro- roophthalmol 7:38-39, 1987.
physiol 15:535-540, 1998. 61. Martyn CN, Ewing DJ. Pupil cycle time: a simple
54. Parati G, Casadei R, Groppelli A, Di Rienzo M, Man- way of measuring an autonomic reflex. J Neurol
cia G. Comparison of finger and intra-arterial blood Neurosurg Psychiatry 49:771-774, 1986.
pressure monitoring at rest and during laboratory 62. Blumen SC, Feiler-Ofry V, Korczyn AD. The pupil
testing. Hypertension 13:647-655, 1989. cycle time in Horner's syndrome. J Clin Neu-
55. Ten Harkel AD, Van Lieshout JJ, Van Lieshout EJ, roophthalmol 6:232-235, 1986.
Chapter 38
ELECTROPHYSIOLOGY OF PAIN
Rose M. Dotson
QUANTITATIVE SENSORY TEST

AUTONOMIC TESTS
MICRONEUROGRAPHY
LASER EVOKED POTENTIALS
SUMMARY
The diagnosis and treatment of neuropathic that potentially could result in pain. They al-
pain are a challenge for physicians caring for low clinicians to assess neuropathic pain by
patients who have medical conditions caus- quantifying responses to neurophysiologic
ing this type of pain. The incompletely un- tests and using the results to develop a bet-
derstood complex mechanisms of neuro- ter understanding of the underlying patho-
pathic pain contribute to the difficulty. physiologic mechanisms of the pain.
Another major obstacle has been finding a This chapter reviews various neurophysio-
test to objectively demonstrate and quantify logic techniques used to study the function
a physiologic disruption in sensory function and dysfunction of the nociceptive system in
that may be the cause of the pain. For many humans, including quantitative sensory tests
years, those involved in the study of pain (QSTs), autonomic tests, microneurography
have attempted to quantify the pain experi- (MCNG), and laser evoked potentials (LEPs).
ence solely by bedside psychophysical evalu- The review outlines the physiologic basis of
ations and the subjective reports of patients. and methods for performing these tech-
Tests of large-fiber function, including vi- niques, some applications of the tests in pa-
brometry, sensory nerve conduction studies, tients with pain, and some of the potential
and routine somatosensory evoked poten- pitfalls in the use of these tests. Details of the
tials, have been used to document large techniques for performing autonomic tests
myelinated fiber sensory abnormalities that are given in Chapters 33 to 37.
might also involve small diameter axons of
the nociceptive system. The first step was to
develop neurophysiologic techniques to as- QUANTITATIVE SENSORY TEST
sess the function of the peripheral and cen-
tral components of pain pathways in normal The quantitative sensory test (QST) is useful
subjects. The next step was to refine the tech- for assessing myelinated and unmyelinated
niques and to gather normal data. These sensory fibers in the periphery and in noci-
techniques can be applied to patients with ceptive pathways in the central nervous sys-
altered sensory function, specifically those tem in persons with neurogenic pain.1 The
with neuropathic pain. The tests are objec- test uses controlled thermal stimulation of
tive or semiobjective physiologic measures the skin to allow examiners to reproducibly
that correlate with abnormal function in the quantify sensory nerve function of nocicep-
nociceptive system and document a lesion tive pathways. It relies on the subjective re-
477
port of the patient that he or she perceives The examiner can observe abnormal pat-
the stimulus. The QST quantifies sensory terns that correspond to pathophysiologic
nerve function by allowing the examiner to changes in the sensory nervous system, cen-
determine the intensity of stimulus required tral or peripheral. There are also clues to
for perception and, at times, to correlate this nonphysiologic or psychogenic abnormal
with the patient's report of the type and in- patterns with erratic results that are not re-
tensity of sensation. The test must use one producible.
stimulus type at a time, with precisely de- Quantitative sensory test systems provide
fined physical characteristics and intensity. a temperature stimulus by means of a con-
The apparatus must have a program that al- tact Peltier type thermode, with a surface
lows the examiner to deliver stimuli to the area of 3-13 cm2, that is applied to the skin
skin of the patient using appropriate algo- to warm or cool the area under the ther-
rithms. This ensures that the test provides an mode. A thermocouple is placed at the skin-
accurate, consistent, and semiobjective as- thermode interface to monitor the temper-
sessment of sensory function. ature of the stimulus. The thermode is set
The most useful and efficient algorithm initially at the adaptation or holding tem-
for determining pain thresholds—that is, perature of 30°C-34°C, within the range of
the intensity at which at least 50% of the temperature at which only a transient ther-
stimuli are perceived as painful—is the mal sensation is caused by placing the probe
method of limits. In addition, this algoridim al- on the skin.10 The examiner operates the ap-
lows the examiner to obtain reliable results paratus to cause temperature changes within
with delivery of the least number of poten- certain limits, usually 0°-50°C, that are set to
tially unpleasant stimuli. This is important prevent burning the patient's skin. The rate
because repetitive stimulation of a primary of temperature change should be standard-
afferent can produce sensitization. The ized to ensure the validity of normal data.
method of limits exposes the patient to a The findings from a study using psycho-
stimulus of changing intensity after he or she physical measures and MCNG to record C
has been instructed to signal the first per- nociceptor activity in response to noxious
ception of pain or temperature sensation, heat stimuli delivered with a QST apparatus
called the appearance threshold. Thresholds to the skin of normal human volunteers are
may also be approached from above thresh- shown in Figure 38-1. This study showed
old level, with the patient reporting the ces- that both the magnitude of pain perceived
sation of sensation, called the disappearance by the subjects and the frequency of C no-
threshold. This algorithm is as sensitive and ciceptor discharges in response to the stim-
reproducible as other less time efficient ulus increased with faster rates of tempera-
methods, for example, forced-choice testing, ture increase.11 Therefore, the examiner
which may give lower absolute values but do should conduct the QST using a consistent
not provide more accurate data.2"4 temperature ramp for gathering normal
Magnitude estimation, an algorithm that data and testing patients.
uses a visual analog scale to rate sensation The patient indicates when the first sen-
perceived when a suprathreshold stimulus is sation occurs in response to cooling or
applied, is useful in evaluating pain.5"7 Dyck warming the stimulus probe attached to the
and colleagues8 developed a method called skin. The sensation (stimulus perception or
the 4, 2, and 1 stepping algorithm that incor- pain) that occurs in response to a cold or
porates magnitude estimation with the ad- hot stimulus is indicated by the patient, who
dition of null stimuli randomly interposed either tells the examiner or presses a but-
between pyramidal and flat-topped pyrami- ton to halt the stimulus and to reset the
dal stimuli. The testing starts at an interme- thermode temperature to the holding level.
diate level of stimulus intensity, uses 25 steps The QST apparatus records the tempera-
to reach maximal intensity and uses die vi- ture of the stimulus at which the patient in-
sual analogue scale that allows patients to dicated perception of a change in temper-
rate the pain intensity. This method of test- ature or pain. The test apparatus includes
ing produces results that are well correlated standardized thermode size, stimulation
with the standard forced-choice algorithm.8-9 sites, rate of change in stimulus tempera-
Electrophysiology of Pain 479
figure 38-1. A-C, Discharge of a single C

nociceptor in response to heat stimulus
ramps between 32°C and 47°C, at three
rates of temperature rise. Note different
time bases for each of the three graphs.
Discharge rate recorded and peak magni-
tude estimate of evoked pain are given.
(From Yarnitsky et al.11 By permission of
The Physiological Society.)
ture, interstimulus intervals, and pretest present. The interpretation of results should
skin temperature. consider that cold hypalgesia may occur in
Tests are performed in normal volunteers normal subjects as well as in patients with
and in patients who have pain, to provide pain. Heat hypalgesia alone is rare and dif-
standardized normal values and valid com- ficult to document because of the setting of
parisons. In addition, comparison of the re- an upper temperature limit of 50° C to pre-
sults obtained on painful areas of skin with vent injury. The normal ranges of pain
the same site on the uninvolved side allows thresholds obtained using the method of
the examiner to determine whether unilat- limits are 44°C-47°C for a hot stimulus and
eral cold or heat hyperalgesia or allodynia is 9°C-12°C for a cold stimulus.12
The QST evaluates the entire peripheral

and central portions of the sensory system
that participate in the transmission and per-
ception of painful hot or cold stimuli. Ther-
mal stimulation directly activates the recep-
tor in the skin, and the receptor in turn
activates the axon innervating it. In the pe-
ripheral nervous system, the primary affer-
ent fibers that convey these messages to the
central nervous system are C and A8 noci-
ceptors. The central nociceptive system con-
sists of the spinothalamic tract, cerebral cor-
tex, antinociceptive areas of the brain stem,
and probably the hypothalamus, amygdala,
and limbic cortex.13
Decreased (hyperalgesia or allodynia) or
increased (hypalgesia) thresholds for the
perception of a cold or hot stimulus may oc-
cur in various combinations, as demon-
strated by Verdugo and Ochoa12 in patients
with somatosensory disorders (Figs. 38-2- Figure 38-2. Pure cold pain and/or heat pain hyperal-
38—4). In some patients with pain, the so- gesia. A, Normal pattern; B, pure cold pain hyperalge-
sia (30 patients); C, pure heat hyperalgesia (26 pa-
matosensory abnormality is hyperalgesia tients); D, combined cold pain and heat pain
selectively in response to a cold or a hot stim- hyperalgesia (5 patients). CP, cold pain; CS, cold sen-
ulus. Ochoa and Yarnitsky14 described pa- sation; HP, heat pain; WS, warm sensation. (From Ver-
tients with neuropathy who had cold skin, dugo and Ochoa.12 By permission of Oxford University
Press.)
Figure 38-3. Thermal hypesthesia combined with thermal pain hyperalgesia. A, Normal pattern; B, warm hypes-
thesia associated with cold pain hyperalgesia (6 patients); C, warm hypesthesia associated with heat pain hyperal-
gesia (1 patient); D, cold hypesthesia associated with cold pain hyperalgesia (4 patients); E, cold hypesthesia asso-
ciated with heat pain hyperalgesia (17 patients). CP, cold pain; CS, cold sensation; HP, heat pain; WS, warm sensation.
(From Verdugo and Ochoa.12 By permission of Oxford University Press.)
ened by decreasing skin temperature, but

die pain remains during compression-isch-
emia A-fiber nerve block. This experimental
finding indicates that C fibers mediate these
sensory changes of primary hyperalgesia
within the actual area of injury in patients
or in the area where capsaicin is applied di-
rectly to the skin of normal subjects.16 A
study in human subjects that used QST and
MCNG to identify sensitized C nociceptors
in the peripheral nerve innervating an area
of skin where the patient experienced spon-
taneous burning pain and heat hyperalgesia
showed that the neural discharge of the ab-
normal, sensitized C nociceptors correlated
with the perceived magnitude of pain as
measured by the visual analog scale17 (Figs.
38-5 and 38-6). The QST is useful in de-
termining the pattern and degree of abnor-
Figure 38-4. Thermal hypesthesia combined with ther- mality, in following the clinical course, and
mal pain hyperalgesia (continuation of Fig. 38-3). A, in documenting the response to treat-
Normal pattern; B, cold hypesthesia associated with cold
and heat pain hyperalgesia (5 patients); C, cold and
ment.15'18
warm hypesthesia associated with cold pain hyperalge- Quantitative sensory test measurements
sia (1 patient); A cold and warm hypesthesia associated can document and quantify hypesthesia, hy-
with cold and heat pain hyperalgesia (1 patient). CP, peresdiesia, hypalgesia, and hyperalgesia.
cold pain; CS, cold sensation; HP, heat pain; WS, warm Thus, the clinician has a reproducible mea-
sensation. (From Verdugo and Ochoa.1* By permission
of Oxford University Press.) surement of nerve dysfunction that can be
attributed to a particular type of primary af-
ferent nerve fiber that may be involved in
reduced ability to perceive cool stimuli, and pain production if the lesion is in the pe-
cold hyperalgesia. These patients had evi- riphery. However, the QST does not differ-
dence of small myelinated AS fiber neu- entiate central from peripheral dysfunction,
ropathy with sparing of unmyelinated C because the abnormalities found are not
fibers. specifically localized and may be located in
Many clinical reports indicate that pa- the nociceptive pathway at any level of the
tients with complex regional pain syndrome neuraxis. This test can be used to make
(CRPS), or reflex sympathetic dystrophy the initial assessment, to follow the clinical
(RSD), frequently report cold hyperalgesia course, and to determine the response to
on the QST or bedside examination. In fact, medications and other forms of interven-
a comprehensive QST evaluation of patients tion.19-20
with this clinical diagnosis by means of the
QST showed that they may have cold hyper-
algesia, heat hyperalgesia, or both. These ab- AUTONOMIC TESTS
normalities can be quantified and followed
clinically with the QST to show response to Patients with neurogenic pain and features
treatment.12'15 indicating involvement of the sympathetic
Heat hyperalgesia with spontaneous burn- nervous system, either sudomotor or vaso-
ing pain occurs in erythromelalgia or cap- motor, have a poorly understood pain symp-
saicin-treated skin. The QST pattern in these tom complex known as complex regional pain
instances is one of decreased threshold for syndrome (CRPS) or reflex sympathetic dystrophy
pain to hot, but not to cold, stimuli. The (RSD).21'22 Visual inspection and patients'
spontaneous burning pain and mechanical reports of alterations in sweating and skin
hyperalgesia in these patients can be less- temperature of involved body areas impli-
Figure 38-5. Correlation between neural discharge of an identified sensitized C polymodal nociceptor with recep-
tive field in symptomatic skin (upper trace) and simultaneous temporal profile of pain magnitude (lower trace) in re-
sponse to gentle stroking of the receptive field. (From Cline et al.17 By permission of Oxford University Press.)
cate the sympathetic nervous system. The within the same patient. These efferents may
multiple pathophysiologic mechanisms that be the passive or the reactive arc of a so-
may result in this clinical pattern are not matosympathetic reflex response to noxious
completely known. There are several possi- input. Also, denervated sympathetic end or-
bilities regarding the exact role of the sym- gans in case of nerve injury may cause the
pathetic efferents in this context, and this clinical symptoms or signs of autonomic ner-
may vary from patient to patient or even vous system involvement in neuropathic
pain. Roberts23 proposed that sympathetic
efferents normally have an active role in
helping to maintain low-threshold mechano-
receptor input to sensitized central noci-
ceptors. Another hypothesis is that nocicep-
tor activity is maintained by sympathetic
efferent activity, thereby causing sponta-
neous or stimulus-induced sympathetically
maintained pain. Indirect evidence supports
the idea that there is sympathetic activation
of nociceptors in humans with the clinical
features of CRPS.24 However, data from an-
imal research indicate that, in peripheral
nerve injury, sympathetic efferent activity
causes excitation of cutaneous nocicep-
tors.25-28
The quantitative sudomotor axon reflex
test (QSART), discussed in Chapter 34, is a
Figure 38-6. Response of an identified C polymodal sensitive test with approximately a 20% co-
unit in symptomatic skin (upper trace) compared with efficient of variation. Thus, this test gives re-
that from opposite nonpainful hand (lower trace). The producible results in normal subjects and in
mechanical stimulus (stroking with a blunt wooden patients with neuropathy. Furthermore,
stick) was applied to the receptive fields, at a time
marked by arrowheads. Note the prolonged after-dis- there is no significant side-to-side difference
charge for the C unit in symptomatic skin. (From Cline in normal subjects. This allows clinicians to
et al.17 By permission of Oxford University Press.) compare the results in patients to normative
values and to determine side-to-side differ- respond more vigorously to circulating cate-
ences in a patient with a unilateral painful cholamines. Thus, the skin on the involved
condition to obtain useful information side becomes cooler than that on the nor-
about sympathetic dysfunction or sympa- mal side. Maneuvers that usually result in re-
thetic involvement in neuropathic pain flex warming of the skin, such as warming
states that fit the definition of CRPS.29~31 another part of the body or sympathetic
Patterns of QSART response that occur in blockade of the affected area, do not cause
limbs affected by painful peripheral neu- warming of that area of skin.
ropathies and CRPS are excessive or persis- Patients with sensitized C nociceptors, eryth-
tent sweat responses with reduced latencies romelalgia, or topically applied capsaicin
or, in some cases, decreased sweat vol- secrete vasodilating substances antidromi-
umes.29"31 An abnormal QSART pattern in cally from active nociceptors, and as a result,
a patient complaining of pain provides an the skin is warm in the areas with pain and
objective measure indicating a pathophysio- hyperalgesia.16'35'36 This can be reproduced
logic change in the involved limb that may in normal human volunteers by performing
occur with, but is not necessarily a causative MCNG with intraneural microstimulation at
factor for, the neuropathic pain. intensities that produce a painful sensation.
Measurements of resting sweat output are Initially, the pain may cause vasoconstriction
helpful in conjunction with QSART to show that is readily apparent on infrared ther-
sudomotor abnormalities in patients with mography as cooling of the skin. Continued
the clinical features of CRPS. These patients activity in the primary nociceptors with mi-
tend to have increased resting sweat output crostimulation causes vasodilatation and
on the involved limb compared with the nor- warming of the skin that sympathetic reacti-
mal side.30'31 vation can override to produce cool skin.
The sympathetic skin response (SSR), dis- This may provide a pathophysiologic expla-
cussed in Chapter 35, evaluates sympathetic nation for the variability in the temperature
sudomotor function through somatosympa- of the painful area of skin compared with
thetic reflexes.32 Some reports have docu- that of normal skin in patients with neuro-
mented SSR abnormalities in sudomotor pathic pain.37
function in patients with CRPS.33'34 Because
of inherent difficulties in producing quan-
tifiable and reproducible data with the SSR MICRONEUROGRAPHY
technique, it is not a useful neurophysiologic
tool for the assessment of neuropathic pain. In MCNG, semimicroelectrodes with a tip di-
Skin temperature measurements with a ameter of 1 /u.g-15 ;iig are inserted percuta-
surface thermistor or infrared thermogra- neously into an accessible peripheral nerve
phy can compare multiple sites on the skin to record the activity in a single axon, in a
of the involved extremity with the corre- portion of a fascicle, or in an entire nerve
sponding areas on the asymptomatic ex- fascicle.38 Microneurography is useful for
tremity. Because patients with neuropathic uncovering the physiologic mechanisms of
pain or CRPS may have alterations in skin neuropathic pain.39 It is a time-consuming
temperature in conjunction with sensory test that requires a highly motivated and ob-
aberrations, these measurements permit ex- servant patient for successful acquisition of
aminers to document clinically useful ab- useful data. The electrode is connected via
normal temperature patterns or asymme- a preamplifier to an amplifier with attached
tries caused by various pathophysiologic audiomonitors and an oscilloscope to permit
mechanisms.30 Patients with lesions causing the examiner to monitor the neural activity
deafferentation pain and vasomotor dener- of a peripheral nerve innervating an involved
vation may have relatively warm skin on the area of skin. The recording of skin and mus-
involved side because of vasodilatation. cle sympathetic activity, A/3 low-threshold
Later in the course of the condition, dener- mechanoreceptors, and AS and C nocicep-
vation supersensitivity results in vasocon- tor afferent activity can provide pathophysi-
striction caused by up-regulation of adreno- ologic information about the mechanisms of
receptors on blood vessels that begin to different types of neuropathic pain.
As noted above, MCNG has documented skin of a patient with the clinical symptom
the occurrence of sensitized C nociceptors complex of CRPS. Our observations in sev-
in a patient with erythromelalgia-type pain.17 eral such patients are consistent with this
Torebjork and colleagues40 used MCNG to finding.
provide evidence that an injury or the ap-
plication of capsaicin to the skin causes cen-
tral sensitization in the area of secondary LASER EVOKED POTENTIALS
hyperalgesia outside the actual area of cap-
saicin injection or topical application. On- Laser evoked potentials (LEPs) provide a
going nociceptive input appears to help noninvasive, easily tolerated means of di-
maintain this sensitization.39 rectly assessing function of the central and
With MCNG, investigators have identified peripheral portions of the nociceptive sys-
three previously undescribed types of hu- tem.47 Carmon and colleagues48 first showed
man C nociceptors that respond only to me- that stimulation of normal human skin with
chanical, heat, or chemical stimuli.41 Some short-duration infrared COg laser pulses pro-
of these units were sensitized to heating or duced a near-field cerebral potential at the
mechanical stimuli after chemical stimula- vertex. Amplitudes of the cerebral response
tion with mustard oil, capsaicin, or tonic usually correlate well with the intensity of
pressure.42-43 These likely have a role in the perceived pain reported by patients in re-
primary hyperalgesia that occurs with chem- sponse to the stimulus and with the intensity
ical irritation or inflammation and in the sec- of the applied stimulus.49 Wu et al.50 recendy
ondary hyperalgesia caused by central sensi- reported on two patients with hyperalgesia
tization. (caused by central pain in one and to pe-
Animal experiments have shown that sym- ripheral neuropathic pain in the other) in
pathetic activation of primary afferents oc- whom the LEP responses were delayed, de-
curs with direct stimulation of sympathetic synchronized, and attenuated.
nerves.23 This was not found in MCNG stud- Heat-pain producing lasers, as opposed to
ies of human subjects and patients with the transcutaneous electrical stimulation of pe-
clinical features of CRPS in whom reflex ac- ripheral nerves traditionally used for so-
tivation of sympathetic efferents did not ac- matosensory evoked potentials, can induce
tivate low-threshold mechanoreceptors.44 In pain with minimal influence on other sen-
animals, the activity of sympathetic efferents sations (Fig. 38-7). Only minimal habitua-
results in neural activity in low-threshold tion, adaptation, or tissue damage tends to
mechanoreceptors in even the normal state; occur even with repeated applications of the
sympathetic efferents have a similar effect on laser stimulus. Although some laboratories
nociceptors only after nerve injury.26"28 Al- use intracutaneous electric shock to obtain
though patients with CRPS symptoms have pain somatosensory evoked potentials, most
allodynia on neurologic examination, activ- laboratories perform LEPs. The laser does
ity in single isolated low-threshold mechano- not contact the skin directly as it produces
receptors produced by intraneural micro- an invisible, inaudible, short-duration (20
stimulation at frequencies up to 30 Hz did ms) radiant-heat pulse. The very superficial
not cause pain.45 This suggests that tempo- layers of skin (20-50 jam) are able to absorb
ral and spatial summation may be necessary this pulse because of its long wavelength
for spontaneous or stimulus-induced pain (CO2 laser, 10.6 /am; and thulium YAG laser,
to occur with activation of low-threshold 1.8-2.01 /tm).48'51-54 This type of stimulus
mechanoreceptors in the presence of cen- produces a rapid increase in skin surface
tral nociceptor sensitization. temperature (50°C per second) and selec-
Microneurography may be used to unravel tively activates the smallest diameter nerve
the complex story of pain and the sympa- terminals of thinly myelinated A8 and un-
thetic nervous system in humans by direcdy myelinated C fibers.46'50'51 Laser stimulus in-
recording sympathetic efferent activity. With tensity is best characterized as stimulus en-
MCNG, Casale and Elam46 demonstrated ergy per unit area, and the average pain
normal activity in a sympathetic efferent threshold in young healthy adults is 10
fiber in a nerve innervating a painful area of mj/mm2.55
Figure 38-7. The measurement of pain-related cerebral potentials. Pain-inducing stimuli, such as intracutaneous
shock (lower right) or laser heat pulse, specifically activate nociceptive afferents, which conduct information in an-
terolateral and dorsal spinal tracts to the thalamus and, from there, to the cerebral cortex. Sensation is estimated
on an analogue scale, with values of 4 and more denoting increasing pain. Stimulus-induced brain potentials ap-
pear in the surface EEG and are visible after averaging more than 40 stimulus repetitions. The negativity (upward
deflection) at 150 ms (N150) after stimulus onset and the positivity at 250 ms (P250) are late components of the
evoked potential that reflect the painfulness of the stimulus applied. (From Bromm.51 By permission of Lippincott,
Williams & Wilkins.)
Laser evoked potentials have larger am- middle latency negative peak (Nl, Ml 70), a
plitudes than routine somatosensory evoked negative peak (N2) at a latency of 250 ± 20
potentials and require averaging of only ms (mean ± SD), and a positive peak (P2)
25-40 responses (Fig. 38-8). The compo- at 390 ± 30 ms. N2 is maximal at Cz, with
nents of LEPs include late and ultralate extension into the central leads, but P2 is
waveforms with a maximal amplitude over maximal at Cz and Pz. The level of attention,
the vertex at Cz (according to the Interna- arousal, and distraction influences these po-
tional 10—20 System). Laser activation of Ad tentials, especially Pz, and these factors must
fibers that have conduction velocities of be taken into account when performing the
4-30 m/second in humans38 causes first test.48'53'56
pain, with a latency of approximately 500 ms Activation of C fibers (conduction veloc-
corresponding with the late LEP. The typi- ity, 0.4-1.8 m/seconds in humans) results in
cal waveform obtained with stimulation of the ultralate components of LEPs. This re-
the skin on the dorsum of the hand has a sponse has a positive peak maximal at the
painful sensations when a repetitive 1 Hz

pinprick stimulus is applied to the skin.
Laser evoked potentials indicated the in-
volvement of Aft and A8 fibers in a patient
with polyneuropathy, muscle weakness, im-
paired sensation (cold, position, and vibra-
tory), absence of conventional tibial nerve
somatosensory evoked potentials, and large
myelinated fiber loss on sural nerve biopsy.53
In this patient, LEPs showed small late re-
sponses, evidence for impaired AS function,
and large ultralate responses, with a peak la-
tency of approximately 1600 ms, which is ev-
Figure 38-8. Late and ultralate laser evoked potentials
idence for preserved function of C fibers
(LEPs) in a healthy subject. Vertex vs. linked earlobes (Fig. 38-9).
(negativity upward). Stimulation of the back of the In a case of polyneuropathy in which the
hand elicited a late positivity at about 400 rns (A). Pref- nerve conduction distance between the
erential A-fiber block by pressure to the radial nerve at hand and foot was 0.8 m, the late and ul-
the wrist strongly attenuated the late LEP and an ul-
tralate potential appeared (B), indicating that the lat- tralate LEP responses corresponded to con-
ter was mediated by preserved C-fiber input. (From duction velocities of 16 m/second and
Treede et al.53 By permission of Lippincott, Williams & 1.2 m/second, respectively53 (Fig. 38-10).
Wilkins.) This study confirmed that A8 peripheral af-
ferents are responsible for transmission of
the late component and C-fiber activation
vertex and a latency of about 1400 ms. It is for the ultralate component. This patient
unreliable in recordings unless preferential had marked wind-up, despite hypalgesia in
A-fiber block suppresses the late compo- response to a single pinprick stimulus, and
nent.53 The ultralate wave is easily obtained there was unmasking of the ultralate com-
if the late component is absent because of ponent of the LEP. This unmasking appears
disease selectively affecting AS and not C to provide a cortical correlate for disinhibi-
fibers. tion of C-fiber responses to noxious heat that
Scalp topography and waveforms of the occurs in persons who display wind-up when
late and ultralate LEPs are similar, suggest- A fibers are impaired.53 Therefore, this tech-
ing that they have the same cerebral gener- nique can be useful in the evaluation of no-
ators.57 Spatiotemporal source analysis likely ciceptive pathways in general. Also, it can
indicates that N2 is generated by activity help document A5-fiber impairment with
mainly bilaterally in secondary somatosen- sparing of C-fiber function. The selective loss
sory cortex. A deep dipole in the midline of small unmyelinated fibers can only be
corresponding to the location of the ante- documented when A fibers are blocked or
rior cingulate gyrus is primarily responsible impaired, because activity in C fibers pro-
for the P2 component. Contralateral pri- duces the highly variable ultralate LEP re-
mary and secondary somatosensory cortex sponse.53
activity appears to be the generator of the
middle latency (Nl) component.
Laser evoked potentials are useful clini- SUMMARY
cally to evaluate objectively the peripheral
and central nociceptive pathways in patients Pain is a subjective experience in which the
with neuropathic pain and disturbances of patient's emotional state has a major role,
pain perception, such as hypalgesia, hyper- contributing to the challenge pain clinicians
algesia, allodynia, and spontaneous pain.58"60 have in quantifying and objectively evaluat-
Some of these patients have abnormal sum- ing this common complaint. The multiplic-
mation, or wind-up, consisting of the per- ity and complexity of the neural mechanisms
ception of continuous burning pain instead that produce chronic pain make the clini-
of the normal individual sharp-pricking cian's task even more challenging. During
subject H.6.. 67 years, polyneuropothy. laser EP
Figure 38-9. Late and ultralate laser evoked potentials in a 67-year-old man with polyneuropathy. Top traces:. Fol-
lowing stimulation of the right hand, a normal AS-fiber-related late potential was recorded. Bottom traces: Following
stimulation of the left foot, the late potential was markedly decreased in amplitude and a C-fiber-related ultralate
potential was documented. The heat-pain threshold for laser stimuli was unremarkable in both areas, but a pro-
nounced temporal summation occurred with stimulation of the foot. (From Treede et al.53 By permission of
Lippincott, Williams & Wilkins.)
Figure 38-10. Late and ultralate laser evoked potentials (LEPs) in a 25-year-old man with hereditary motor and sen-
sory neuropathy type I. The latency differences between hand and foot stimulation indicate that late LEPs are me-
diated by AS fibers and ultralate LEPs by C fibers. (From Treede et al.53 By permission of Lippincott, Williams &
Wilkins.)
487
the last 20 years, research in the assessment 7. Gruener G, Dyck PJ. Quantitative sensory testing:
of the nociceptive and autonomic systems methodology, applications, and future directions. J
has provided useful tools for the diagnosis, 8. Dyck PJ, O'Brien PC, Kosanke JL, Gillen DA, Karnes
treatment, and scientific investigation of JL. A 4, 2, and 1 stepping algorithm for quick and
neuropathic pain. Quantitative sensory tests accurate estimation of cutaneous sensation thresh-
provide an accurate, reproducible assess- old. Neurology 43:1508-1512, 1993.
9. Dyck PJ, Zimmerman I, Gillen DA, Johnson D,
ment of the sensory response to well-delin- Karnes JL, O'Brien PC. Cool, warm, and heat-pain
eated controlled stimuli for evaluating the detection thresholds: testing methods and infer-
function of small myelinated and unmyeli- ences about anatomic distribution of receptors.
nated fibers. These tests can be useful for Neurology 43:1500-1508, 1993.
documenting sensory disturbances that may 10. Darian-Smith I. Thermal sensibility. In Brookhart
JM, Mountcastle VB (eds). Handbook of Physiol-
occur in patients with neuropathy who ex- ogy: A Critical, Comprehensive Presentation of
perience hypalgesia, hyperalgesia, or allody- Physiological Knowledge and Concepts. Vol. 3: The
nia. The inclusion of autonomic tests in the Nervous System. American Physiological Society,
evaluation of some patients with neuro- Bethesda, 1984, pp 879-913.
11. Yarnitsky D, Simone DA, Dotson RM, Cline MA,
pathic pain may provide objective evidence OchoaJL. Single C nociceptor responses and psy-
of increased sympathetic tone, heightened chophysical parameters of evoked pain: effect of
somatosympathetic reflexes, or sympathetic rate of rise of heat stimuli in humans. J Physiol
denervation. This information may help the 450:581-592, 1992.
clinician to diagnose more accurately the 12. Verdugo R, OchoaJL. Quantitative somatosensory
thermotest. A key method for functional evaluation
particular type of pain syndrome the patient of small calibre afferent channels. Brain 115:893-
has so that the treatment may be better di- 913, 1992.
rected at the underlying mechanism. Mi- 13. Willis WD Jr. From nociceptor to cortical activity.
croneurography, primarily a research tool, is Adv Pain Res Ther 22:1-19, 1995.
14. OchoaJL, Yarnitsky D. The triple cold syndrome.
a powerful method for directly studying Cold hyperalgesia, cold hypoaesthesia and cold skin
primary afferent and sympathetic efferent in peripheral nerve disease. Brain 117:185-197,
neural activity in patients with pain caused 1994.
by lesions of the nervous system. Laser 15. Wahren LK, Torebjork E. Quantitative sensory tests
evoked potentials allow clinicians and rein patients with neuralgia 11 to 25 years after in-
jury. Pain 48:237-244, 1992.
searchers to investigate objectively function 16. Gulp WJ, Ochoa J, Cline M, Dotson R. Heat and
of the peripheral and central nociceptive mechanical hyperalgesia induced by capsaicin.
pathways in patients with neuropathic pain. Cross modality threshold modulation in human C
nociceptors. Brain 112:1317-1331, 1989.
17. Cline MA, Ochoa J, Torebjork HE. Chronic hyper-
algesia and skin warming caused by sensitized C no-
ciceptors. Brain 112:621-647, 1989.
REFERENCES 18. Wahren LK, Torebjork E, Nystrom B. Quantitative
sensory testing before and after regional guanethi-
1. Greenspan JD. Quantitataive assessment of neuro- dine block in patients with neuralgia in the hand.
pathic pain. Curr Pain Headache Rep 5:107-113, Pain 46:23-30, 1991.
2001. 19. Nygaard OP, Kloster R, Solberg T, Mellgren SI. Re-
2. Glaus D, Hilz MJ, Neundorfer B. Thermal discrim- covery of function in adjacent nerve roots after sur-
ination thresholds: a comparison of different meth- gery for lumbar disc herniation: use of quantitative
ods. Acta Neurol Scand 81:533-540, 1990. sensory testing in the exploration of different pop-
3. Dyck PJ, Karnes JL, Gillen DA, O'Brien PC, Zim- ulations of nerve fibers. J Spinal Disord 13:427-431,
merman IR, Johnson DM. Comparison of algo- 2000.
rithms of testing for use in automated evaluation of 20. Attal N, Gaude V, Brasseur L, et al. Intravenous li-
sensation. Neurology 40:1607-1613, 1990. docaine in central pain: a double-blind, placebo-
4. Peripheral Neuropathy Association. Quantitative controlled, psychophysical study. Neurology 54:
sensory testing: a consensus report from the Pe- 564-574, 2000.
ripheral Neuropathy Association. Neurology 43: 21. Dotson RM. Causalgia—reflex sympathetic dystro-
1050-1052, 1993. phy—sympathetically maintained pain: myth and
5. Price DD, McGrath PA, Rafii A, Buckingham B. The reality. Muscle Nerve 16:1049-1055, 1993.
validation of visual analogue scales as ratio scale 22. Stanton-Hicks M,Janig W, Hassenbusch S, Haddox
measures for chronic and experimental pain. Pain JD, Boas R, Wilson P. Reflex sympathetic dystrophy:
17:45-56, 1983. changing concepts and taxonomy. Pain 63:127-
6. Price DD. Psychological and Neural Mechanisms of 133, 1995.
Pain. Raven Press, New York, 1988, pp 18-49. 23. Roberts WJ. A hypothesis on the physiological ba-
sis for causalgia and related pains. Pain 24:297-311, 41. Schmidt R, Schmelz M, Forster C, Ringkamp M,
1986. Torebjork E, Handwerker H. Novel classes of re-
24. Campbell JN, Meyer RA, Raja SN. Is nociceptor ac- sponsive and unresponsive C nociceptors in human
tivation by alpha-1 adrenoreceptors the culprit in skin. J Neurosci 15:333-341, 1995.
sympathetically maintained pain? Am Pain Soc J 42. Schmidt R, Schmelz M, Torebjork HE, Handwerker
1:3-11, 1992. HO. Mechano-insensitive nociceptors encode pain
25. Levine JD, Taiwo YO, Collins SD, Tarn JK. Nora- evoked by tonic pressure to human skin. Neuro-
drenaline hyperalgesia is mediated through inter- science 98:793-800, 2000.
action with sympathetic postganglionic neurone 43. Schmelz M, Schmid R, Handwerker HO, Torebjork
terminals rather than activation of primary afferent HE. Encoding of burning pain from capsaicin-
nociceptors. Nature 323:158-160, 1986. treated human skin in two categories of unmyeli-
26. SatoJ, Perl ER. Adrenergic excitation of cutaneous nated nerve fibres. Brain 123:560-571, 2000.
pain receptors induced by peripheral nerve injury. 44. Dotson R, Ochoa J, Cline M, Roberts W, Yarnitsky
Science 251:1608-1610, 1991. D, Simone D. Sympathetic effects on human low
27. SatoJ, Suzuki S, Iseki T, Kumazawa T. Adrenergic threshold mechanoreceptors (abstract). Abstracts
excitation of cutaneous nociceptors in chronically Soc Neurosci 16:1280, 1990.
inflamed rats. Neurosci Lett 164:225-228, 1993. 45. Dotson R, Ochoa J, Cline M, Marchettini P, Yarnit-
28. Gold MS, White DM, Ahlgren SC, Guo M, Levine sky D. Intraneural microstimulation of low thresh-
JD. Catecholamine-induced mechanical sensitiza- old mechanoreceptors in patients with causal-
tion of cutaneous nociceptors in the rat. Neurosci gia/RSD/SMP (abstract). Abstracts Soc Neurosci
Lett 175:166-170, 1994. 18:290, 1992.
29. Low PA. Autonomic nervous system function. J Clin 46. Casale R, Elam M. Normal sympathetic nerve ac-
Neurophysiol 10:14-27, 1993. tivity in a reflex sympathetic dystrophy with marked
30. Chelimsky TC, Low PA, Naessens JM, Wilson PR, skin vasoconstriction.J Auton Nerv Syst 41:215-219,
Amadio PC, O'Brien PC. Value of autonomic test- 1992.
ing in reflex sympathetic dystrophy. Mayo Clin Proc 47. Rossi P, Serrao M, Amabile G, Parisi L, Pierelli F,
70:1029-1040, 1995. Pozzessere G. A simple method for estimating con-
31. Sandroni P, Low PA, Ferrer T, Opfer-Gehrking TL, duction velocity of the spinothalamic tract in
Willner CL, Wilson PR. Complex regional pain syn- healthy humans. Clin Neurophysiol 111:1907-1915,
drome I (CRPS I): prospective study and laboratory 2000.
evaluation. ClinJ Pain 14:282-289, 1998. 48. Carmon A, Mor J, Goldberg J. Evoked cerebral re-
32. Shahani BT, Halperin JJ, Boulu P, Cohen J. Sym- sponses to noxious thermal stimuli in humans. Exp
pathetic skin response—a method of assessing Brain Res 25:103-107, 1976.
unmyelinated axon dysfunction in peripheral neu- 49. Beydoun A, Morrow TJ, Shen JF, Casey KL. Vari-
ropathies. J Neurol Neurosurg Psychiatry 47:536- ability of laser-evoked potentials: attention, arousal
542, 1984. and lateralized differences. Electroencepha'ogr
33. Cronin KD, Kirsner RL, Fitzroy VP. Diagnosis of re- Clin Neurophysiol 88:173-181, 1993.
flex sympathetic dysfunction. Use of the skin po- 50. Wu Q, Garcia-Larrea L, Mertens P, Beschet A, Sin-
tential response. Anaesthesia 37:848-852, 1982. dou M, Mauguiere F. Hyperalgesia with reduced
34. Rommel O, Tegenthoff M, Pern U, Strumpf M, laser evoked potentials in neuropathic pain. Pain
Zenz M, Malin JP. Sympathetic skin response in pa- 80:209-214, 1999.
tients with reflex sympathetic dystrophy. Clin Au- 51. Bromm B. Consciousness, pain, and cortical activ-
ton Res 5:205-210, 1995. ity. Adv Pain Res Ther 22:35-59, 1995.
35. Ochoa J. The newly recognized painful ABC syn- 52. Kazarians H, Scharein E, Bromm B. Laser evoked
drome: thermographic aspects. Thermology 2:65- brain potentials in response to painful trigeminal
107, 1986. nerve activation. IntJ Neurosci 81:111-122, 1995.
36. Ochoa J, Cline M, Comstock W, et al. Painful syn- 53. Treede R-D, LorenzJ, Kunze K, Bromm B. Assess-
drome newly recognized: polymodal hyperalgesia ment of nociceptive pathways with laser-evoked po-
with cross modality threshold modulation and ru- tentials in normal subjects and patients. Adv Pain
bor (abstract). Abstracts Soc Neurosci 13:189, 1987. Res Ther 22:377-392, 1995.
37. Ochoa JL, Yarnitsky D, Marchettini P, Dotson R, 54. Spiegel J, Hansen C, Treede RD. Clinical evalua-
Cline M. Interactions between sympathetic vaso- tion criteria for the assessment of impaired pain
constrictor outflow and C nociceptor-induced an- sensitivity by thulium-laser evoked potentials. Clin
tidromic vasodilatation. Pain 54:191-196, 1993. Neurophysiol 111:725-735, 2000.
38. Vallbo AB, Hagbarth KE, Torebjork HE, Wallin BG. 55. Gibson SJ, LeVasseur SA, Helme RD. Cerebral
Somatosensory, proprioceptive, and sympathetic event-related responses induced by CO2 laser stim-
activity in human peripheral nerves. Physiol Rev ulation in subjects suffering from cervico-brachial
59:919-957, 1979. syndrome. Pain 47:173-182, 1991.
39. Torebjork E. Human microneurography and intra- 56. Tarkka IM, Treede RD. Equivalent electrical source
neural microstimulation in the study of neuro- analysis of pain-related somatosensory evoked po-
pathic pain. Muscle Nerve 16:1063-1065, 1993. tentials elicited by a COg laser. J Clin Neurophysiol
40. Torebjork HE, Lundberg LE, LaMotte RH. Central 10:513-519, 1993.
changes in processing of mechanoreceptive input 57. Treede RD, Bromm B. Reliability and validity of ul-
in capsaicin-induced secondary hyperalgesia in hu- tra-late cerebral potentials in response to C-fibre ac-
mans. J Physiol 448:765-780, 1992. tivation in man. In Dubner R, Gebhart GF, Bond
MR (eds). Proceedings of the Vth World Congress myelinated afferents in diabetic polyneuropathy, as
on Pain. Elsevier, Amsterdam, 1988, pp 567-573. assessed by laser evoked potentials. Clin Neuro-
58. Antonini G, Gragnani F, Romaniello A, et al. Sen- physiol 111:270-276, 2000.
sory involvement in spinal-bulbar muscular atrophy 60. Agostino R, Cruccu G, lannetti GD, et al. Trigemi-
(Kennedy's disease). Muscle Nerve 23:252-258, nal small-fibre dysfunction in patients with diabetes
2000. mellitus: a study with laser evoked potentials and
59. Agostino R, Cruccu G, Romaniello A, Innocent! P, corneal reflex. Clin Neurophysiol 111:2264-2267,
Inghilleri M, Manfredi M. Dysfunction of small 2000.
SECTION 2
Electrophysiologic
Assessment of
Neural Function
Part G
Sleep and
Consciousness
Only recently have the prevalence and clinical importance of sleep disor-
ders been recognized adequately. The development of electrophysiologic
tools for the assessment of sleep has helped define the normal and altered
physiology of sleep (Chapter 39). Sleep disorders include inadequate, ex-
cessive, and disordered sleep. The latter includes excessive or abnormal
movements during sleep. The recording of the surface muscle electromyo-
gram during these movements while monitoring blood pressure, pulse, and
respiration, as described in this section, can help physicians identify, char-
acterize, and define the type and severity of sleep disorder (Chapter 40).
Electrophysiologic assessment of sleep disorders is a superb example of
the importance of combining the methods of clinical neurophysiology for
assessing the condition of a patient. The combination of newly defined pat-
terns of surface electromyographic recordings, electroencephalographic
recordings, and measurements of autonomic function is a critical part of
the assessment of sleep disorders. New advances in sleep studies include un-
attended polysomnography, split-night recordings, excess daytime sleepiness
evaluation, and the interpretation of electro-oculography and multiple sleep
latency testing in the parasomnias.
Chapter 39
PHYSIOLOGIC ASSESSMENT
OF SLEEP
Peter J. Hauri
Cameron D. Harris
Michael H. Silber
DEFINITIONS Blood Gases

Polysomnography Cardiac Rhythm
Multiple Sleep Latency Test Summary Statistics for Respiratory
Maintenance of Wakerulness Test Variables
EVALUATION OF SLEEP ASSESSING PERIODIC LIMB
Traditional Sleep Stages MOVEMENTS
Nontraditional Measures of Sleep Summary Statistics for Periodic Limb
Summary Statistics for Sleep Variables Movements
ASSESSING RESPIRATION DURING ASSESSING OTHER PHYSIOLOGIC
SLEEP VARIABLES
Definitions Core Temperature
Airflow Esophageal pH Measurements
Effort SUMMARY
Snoring Sounds
The clinical evaluation of sleep and its dis- before either awakening spontaneously or
orders began in the early 1970s. The goal of being awakened by a technician. The goal of
this chapter is to review the recording and polysomnography is to quantify the amount
interpretation of sleep studies and to discuss of time spent in various stages of sleep dur-
how specific clinical issues determine what ing the night and to document clinically rel-
data need to be obtained. evant events that disrupt sleep, such as car-
diopulmonary abnormalities or abnormal
motor activity.
DEFINITIONS The basic format of a comprehensive
polysomnogram is standardized. The re-
Polysomnography quired minimum includes (1) continuous
monitoring of at least one channel of an
Polysomnograms are recorded during the electroencephalogram (EEC) (C3-A2orC4-
normal sleeping hours of a patient. Patients Al), (2) one or two channels of an electro-
come to the sleep laboratory 1-2 hours be- oculogram (EOG), (3) three channels of res-
fore their usual bedtime. After the elec- piratory data (airflow, effort, and oxyhemo-
trodes and sensors have been applied, pa- globin saturation), (4) electrocardiographic
tients may watch television until they are (EGG) recording, and (5) electromyography
ready to go to sleep. They sleep 6-8 hours (EMG) of the submental and anterior tibial
493
494 Sleep and Consciousness
muscles. Other variables may be monitored Polysomnograms usually are interpreted

as clinically indicated, including snoring, visually. If sleep stages and other polysomno-
position of the sleeper, additional EEG, sur- graphic variables are scored by computer,1
face EMG recordings from all four limbs, visual inspection of the raw data is still nec-
esophageal pressure, and esophageal pH. essary. The interpretation of many noctur-
Currendy, polysomnograms typically are nal events requires clinical skill (for exam-
recorded with digital systems that allow re- ple, other sleep disorders may masquerade
view at variable screen widths. A 30-second as sleep-disordered breathing), and some
screen window is most frequently used, but patients may have an abnormal sleep struc-
longer windows (60-180 seconds) may aid ture tiiat is not easily scored by machine.
in evaluating respiratory or other periodic The analysis of polysomnograms of neo-
events and a shorter window (10 seconds) in nates and infants requires special experi-
assessing EEG abnormalities. If an analog ence and skill because, for diese groups,
paper system is used, the data usually are EEG stages are defined differendy, and res-
recorded at a paper speed of 10 mm/sec- piratory and other behaviors during sleep
ond. The patient's nighttime behavior are unique.
should also be observed and recorded dur- An excerpt from a typical polysomnogram
ing polysomnography. Digital systems allow is shown in Figure 39-1, and a typical sum-
16 to 32 channels of data to be recorded. mary of a polysomnographic recording (in
Figure 39-1. Typical polysomnogram montage, including channels for eye movement, electroencephalography, chin
and leg EMG, airflow, snoring sounds, oxyhemoglobin saturation, and respiratory effort. Abdomen, abdominal
breathing effort by inductive plethysmograph; Airflow, combined signal from nose and mouth; EGG, electrocar-
diogram; EMG, electromyogram; LOG, left outer canthus; Rib cage, thoracic breathing effort by inductive plethys-
mograph; ROC, right outer canthus; sonogram, sound recording; SpOg, oxyhemoglobin saturation by pulse oxime-
try; Sum, electronic summation of rib cage and abdominal movements. (These abbreviations are used in the following
figures.)
Physiologic Assessment of Sleep 495
Figure 39-2. A typical polysomnogram report. Later, narcolepsy was diagnosed in this patient (partly on the basis
of results of multiple sleep latency test). The summary statistics are prepared by a polysomnographic technologist,
and the impression is provided by the supervising physician after reviewing the raw data. NREM, non-REM sleep;
REM, rapid eye movement sleep.
this case, from a patient who complained of Multiple Sleep Latency Test
excessive sleepiness) is shown in Figure
39-2. The usual placement of electrodes and The multiple sleep latency test (MSLT)3 consists
sensors and the amplification variables typi- of four or five 20-minute rest periods in bed
cally used for a polysomnogram are sum- in a dark room spaced 2 hours apart and al-
marized in Table 39-1. most always recorded on the day following a
The EOG derivations used allow vertical night with polysomnography. The goal of
and horizontal eye movements to be distin- the test is to quantify physiologic sleepiness
guished (vertical movements produce in- during waking hours and to determine the
phase and horizontal movements out-of- occurrence of REM sleep near sleep onset.
phase deflections). This is helpful in the For this test, patients are asked to remain in
differentiation of rapid eye movements of the laboratory for the entire day. They may
rapid eye movement (REM) sleep (most read, watch television, or engage in other
commonly horizontal) from blinks of wake- quiet activities, but they must not sleep be-
fulness (resulting in vertical movements), es- tween scheduled naps. Recordings per-
pecially in REM sleep without atonia. The formed during the MSLT are simpler than
EEG derivations allow sampling of frontal, those performed during polysomnography,
central, and occipital activity and provide and typically, only EEG, EOG, submental
bipolar as well as referential representation EMG, and ECG data are recorded. For each
of vertex activity. The 10 Hz low-frequency nap, sleep latency and the occurrence of
filter setting used in the EMG derivations re- REM sleep are noted. For the MSLT, sleep
duces movement artifact. latency is defined as the time between "lights
Table 39-1. Typical Settings for a Polysomnogram

Signal Sensor Type and Placement Sensitivity LFF, Hz* HFF, Hz*
EOG Electrodes placed 1 cm lateral and 1 cm 5-7.5 /iV/mm 0.5 70
inferior to outer canthi referenced to Fpz
EEC Electrodes placed at Fz, Cz, Oz, Cg, C4, AI, 5-7.5 /iV/mm 0.5 70
and Aaf
Chin EMG Electrodes placed over mentalis, mylohyoid, 2 ju,V/mm 10 70
and digastric muscles
Leg EMG Electrodes placed over belly of anterior 2 juV/mm 10 70
tibialis muscle of each leg
EGG Electrodes placed on each shoulder 50 /iV/mm 0.5 70
Oxygen Light absorbance sensor placed on earlobe 100 mV/cmJ DC 5
saturation or finger
Airflow Thermocouple placed in front of both nares Transducer 0.5 5
and the mouth, connected in series to dependent!
provide one signal
Breathing Inductors (elastic bands with embedded wires) Patient DC 5
effort placed around the chest, just under the dependent!
axillae, and around the abdomen at the
level of the navel
Snoring Microphone placed in contact with the skin Transducer 10 70
slightly superior and lateral to thyroid cartilage dependent!
*Low (LFF) or high (HFF) frequency cut-off may vary depending on manufacturer of polygraph.
fFz, Cz, Oz are not required by Rechtschaffen and Kales.2
Ôptimum sensitivity may depend on manufacturer or equipment used or recording conditions.
DC, direct current; EGG, electrocardiogram; EEG, electroencephalogram; EMG, electromyogram; EOG, electro-
oculogram.
out" and the beginning of the first epoch of cies by disrupting sleep, no differences in
any stage of sleep. If no sleep occurs during latencies have been found after nights of
the first 20 minutes in bed, the test is dis- home vs. laboratory polysomnography.4 Sec-
continued and the sleep latency is recorded ond, all treatment with psychotropic drugs
as 20 minutes. After sleep onset has oc- that can safely be discontinued should be
curred, patients are allowed to sleep for 15 stopped a minimum of 2 weeks before the
minutes to determine whether they will en- study. Normal values for the MSLT and its
ter REM sleep. Accurate interpretation of interpretation are discussed in Chapter 40.
the results depends on fulfillment of certain
preconditions. First, adequate amounts of
sleep must be obtained for 1-2 weeks before Maintenance of Wakefulness Test
the study to ensure that the patient is not
voluntarily sleep deprived. This is assessed by A less frequently used variant of the MSLT
having the patient complete a sleep log and is the maintenance of wakefulness test. This is
often wear a wrist actigraph, a device that not a test for sleepiness but measures the
monitors motor activity of an arm and allows ability of the patient to remain awake. It
periods of rest and activity to be differenti- should not be used to diagnose sleep disor-
ated. At least 6 and preferably 7 hours of ders, but sometimes it is helpful in assessing
sleep should be obtained during the night the response to stimulants or a patient's fit-
before the MSLT, as measured by the ness to drive or to fly an airplane. Conditions
polysomnogram. Although it might be imag- are the same except that the patient sits in
ined that a laboratory sleep study the pre- a comfortable recliner or bed with the back
ceding night might affect the MSLT laten- and head supported by a bedrest in a dimly
lit room and is asked to try to stay awake sleeper is suddenly awakened by a sensory
rather than to try to sleep. The patient may stimulus. Nevertheless, to make analysis eas-
not use extraordinary methods to remain ier, a polysomnogram is analyzed in arbitrary
awake, such as slapping the face or singing. epochs, each usually 30 seconds long. Each
Two variants of the test have been described: epoch is classified into one of six possible
one lasting 20 minutes and one lasting 40 stages: wakefulness, REM sleep, and NREM
minutes.5 Because many normal subjects re- sleep stages 1-4. The epoch is classified ac-
main awake for more than 20 minutes, we cording to the stage that comprises a greater
use the 40-minute protocol and measure percentage of the epoch than any other
sleep onset to the first epoch or first 10 sec- stage. The scoring for these six stages is done
onds of any stage of sleep. Published normal according to rules that have been modified
mean sleep latency for this protocol5 is 32.6 only slightly since they were first published
minutes, and 1 standard deviation below the by Rechtschaffen and Kales2 in 1968.
mean is 22.7 minutes and 2 standard devia-
tions below the mean is 12.8 minutes. For
the purposes of practical measurement of Traditional Sleep Stages
adequate alertness, we aim at mean latencies
longer than 22 minutes, while recognizing Wakefulness is scored if there is activity in
that about 15% of normal subjects have the 8-12 Hz (alpha) range over the major-
mean latencies less than this. ity of a 30-second epoch or low-voltage
mixed frequency activity (Fig. 39-3). Rapid
eye movements, eye blinks, and high tone in
EVALUATION OF SLEEP the chin EMG are also typical of wakefulness.
Stage 1 is defined by a relatively low-
During a night's sleep, periods of wakeful- amplitude mixed-frequency EEG predomi-
ness, non-rapid eye movement (NREM) nantly in the 3-7 Hz range (Fig. 39-4).
sleep and REM sleep gradually wax and There are typically slow, usually horizontal,
wane. As wakefulness changes to sleep, eye movements and often decreased tone on
slower EEG activity gradually replaces the al- the chin EMG. Vertex sharp waves and rudi-
pha rhythm, and delta waves very gradually mentary spindles less than 0.5 second in du-
increase as sleep progressively deepens. ration may occur toward the end of stage 1.
There is little abrupt switching from one Stage 2 is characterized by the appear-
sleep stage to the next, except when the ance of sleep spindles (bursts of 12-14 Hz
Figure 39-3. Wakefulness. Note the prominent alpha rhythm, rapid eye movements, and high EMG activity in the
chin and legs.
Figure 39-4. Stage 1 sleep. Note the vertex sharp waves, the slow rolling eye movements, and the absence of alpha
rhythm.
activity lasting 0.5-1.5 seconds) and K com- rapid eye movements and attenuation of
plexes (biphasic EEG waves, initially nega- EMG activity).
tive followed by a positive component, last- Stages 3 and 4 are frequently combined
ing a minimum of 0.5 second) (Fig. 39-5). and called slow wave sleep (Figs. 39-6 and
Because sleep spindles and K complexes are 39-7). The defining criteria are large-am-
discrete and intermittent, intervals as long plitude slow waves (minimally 75 ^V peak-
as 3 minutes of stage 1-appearing EEG be- to-peak, 0.5-2 Hz). Stage 3 is scored if an
tween K complexes or spindles are still epoch contains 20%-50% of slow waves,
scored as stage 2 unless there is evidence to whereas stage 4 is scored if more than 50%
the contrary (for example, large body of the epoch consists of slow waves. With in-
movements followed by alpha waves or creasing age, the overall amplitude of the
Figure 39-5. Stage 2 sleep. Note sleep spindles and K complexes that characterize stage 2 sleep.
Figure 39-6. Stage 3 sleep. Between 20 and 50% of the epoch contains 0.5-2 Hz activity of amplitude greater than
75 uV.
EEG diminishes. Considerable activity may tivity and episodic bursts of rapid eye move-
still occur in the 0.5-2 Hz range, but it no ments (Fig. 39-8). Sawtooth waves (sharp
longer produces amplitudes exceeding 75 waves in the theta range with a small saw-
/*V. Most sleep clinicians still score stages 3 tooth about halfway up or down the main
and 4 in the elderly if the amplitudes of these wave) are seen only in REM sleep, but not
waves reach at least 50 /x,V. in all patients. Transient phasic twitches are
Stage REM is defined by a relatively low- often seen in REM sleep, but the baseline
amplitude mixed-frequency EEG, similar to EMG amplitude should be no higher than
that seen in stage 1, in combination with that seen in any other sleep stage. REM sleep
markedly decreased tone in chin EMG ac- may be divided into phasic or tonic periods.
Figure 39-7. Stage 4 sleep. More than 50% of the epoch contains 0.5-2 Hz activity of amplitude greater than 75 /*V.
Figure 39-8. Stage REM sleep. REM sleep is characterized by rapid eye movements, atonic chin electromyography
(EMG), and an electroencephalography (EEC) pattern similar to that seen in stage 1 sleep. Sawtooth waves may
also be seen.
During phasic REM, eye movements are must last for a minimum of 3 seconds, and
dense, many chin EMG twitches occur, and (3) arousals scored in REM sleep must have
there is increased variability in EGG and res- both increased EEG frequency and in-
piration. Tonic REM epochs are relatively creased chin EMG activity. Scoring these
devoid of these phasic features. arousals is important for an assessment of
Movement time is not a sleep stage but is sleep quality: the more arousals there are,
scored whenever movement artifacts ob- the worse one's perception is of how well
scure EEG channels for at least 50% of the one has slept and the sleepier one is during
epoch and if preceding and subsequent the subsequent day.7 Of course, if an arousal
epochs are scored as sleep. In such a case, lasts longer than 15 seconds, wakefulness
one does not know whether the patient ac- predominates during the epoch in question
tually awakened during the movement or and a full awakening is scored.
whether the movement occurred exclusively Alpha intrusions are scored if alpha activity
during sleep. (8-12 Hz) that is easily visible to the eye is
superimposed on the normal activity of
NREM sleep (Fig. 39-9). This phenomenon,
Nontraditional Measures of Sleep also called alpha-delta sleep,8 or nonrestorative
sleep,9 is often associated with chronic pain
The four sleep-stage phenomena that are (for example, in fibromyalgia or rheumatoid
clinically significant but not included in the arthritis), but it also may occur in patients
original Rechtschaffen and Kales scoring who have no known medical disorder.
system are arousals, alpha intrusions, REM REM sleep without atonia is scored when
sleep without atonia, and body position. both the EEG and EOG suggest REM sleep
Arousals are defined as abrupt shifts in EEG but the chin EMG does not show the ex-
frequency to the alpha, fast beta, or occa- pected muscle atonia. This usually takes the
sionally theta frequency bands. A subcom- form of a marked increase in phasic twitches,
mittee of the American Sleep Disorders As- but sometimes sustained tonic muscle activ-
sociation6 has defined EEG arousal scoring ity is present. The technician often observes
rules, which require that ( 1 ) the subject must vigorous twitching and apparently purpose-
be asleep for at least 10 seconds before and ful movements during this type of sleep,
after the arousal, (2) the EEG frequency shift such as punching in the air10 (Fig. 39—10).
Figure 39-9. Alpha intrusions into nonrapid eye movement sleep. This polysomnogram of a 50-year-old woman with
a complaint of chronic fatigue illustrates intrusion of diffuse alpha activity into slow-wave sleep.
REM sleep without atonia is the neurophys- ical polysomnograms. This is done eidier
iologic marker of REM sleep behavior dis- with a special position indicator worn by the
order (see Chapter 40). patient or by observing the patient through
Body position may affect the severity of dis- a closed circuit video monitor. If the patient
ordered breathing. It is always scored in clin- does not spontaneously sleep part of the
Figure 39-10. REM sleep without muscle atonia. The montage is modified to record electromyography (EMG) ac-
tivity from all four limbs from an older man with Parkinson disease. Note frequent bursts of activity in the limb and
submental EMG leads. Typically, the movements associated with this activity are related to dream content. (From
Daube JR., Cascino GD, Dotson RM, Silber MH, Westmoreland BF. Continuum: Lifelong Learning in Neurology
[Clinical Neurophysiology]. Vol. 4, Part A. Lippincott Williams & Wilkins, Baltimore, 1998, p 169. By permission of
the American Academy of Neurology.)
time on the back and part of the time on the morning wakefulness the time between the
side during the recording, he or she is usu- patient's last sleep episode and finally get-
ally awakened and asked to sleep in the ting up. Although the latter scoring method
other position to assess respiration in both is preferred because it gives more pertinent
positions. information, it is used less frequently by
sleep disorders centers.
Initial REM latency is scored by adding all
Summary Statistics for the time spent sleeping (stages 1, 2, 3, and
Sleep Variables 4) before the first epoch scored as REM.
Initial REM latency decreases with age
After all epochs have been scored, summary (Table 39-3). Initial REM latency consider-
statistics are computed (Fig. 39-2). They in- ably shorter than expected for a patient's age
clude the following: is nonspecific, but it may suggest considera-
1. Time in bed is from "lights out" to get- tion of diagnoses of major depression, sleep
ting out of bed in the morning for the deprivation, narcolepsy, or withdrawal of
last time minus time spent out of bed REM-suppressing medications. If narcolepsy
during the night (for example, for trips is suspected, an MSLT is essential.
to the bathroom). To have a valid Arousal index counts the number of awak-
polysomnogram, typically a minimum enings and arousals per sleep hour. The
of 6 hours in bed is required. arousal indices considered normal depend
2. Total sleep time includes all epochs on the scoring criteria used and may also
scored as stages 1, 2, 3, and 4 or as REM vary from laboratory to laboratory. A study
sleep. of arousals in normal subjects12 revealed
3. Sleep efficiency is computed as the per- a mean of 4/hour using the original
centage of the time in bed that is spent Rechtschaffen and Kales definition2 and a
asleep. mean of 20/hour using the newer Ameri-
The norms for sleep stages and percent- can Sleep Disorders Association criteria de-
ages are reported in Table 39-2. It is ironic scribed above.6 In our laboratory, we con-
that in most sleep laboratories sleep is scored sider 15 or fewer arousals per hour to be
according to the Rechtschaffen and Kales2 normal. To interpret arousal indices, it is im-
rules, but only weak norms exist for that sys- portant to know how many of the arousals
tem. The norms given in Table 39-2 are and awakenings are caused by disordered
from the Williams, Karacan, and Hursch11 breathing or by periodic limb movements.
system, which scores sleep from a frontal-oc- Arousals may be associated with subtle in-
cipital derivation and, therefore, reports creases in upper airway resistance that are
more stages 3 and 4 sleep and slightly more not easily recognized as hypopnea.13 Mark-
wakefulness than the Rechtschaffen and edly increased arousal indices unassociated
Kales system. with periodic limb movements or disordered
Initial sleep latency is the time from "lights breathing are often seen in patients with
out" to the first epoch of scored sleep. Some pain or other medical disorders or in those
compute latency to stage 1 and others to with psychologic distress, especially anxiety.
stage 2 sleep. Because insomniacs often fall Interpretation of the various sleep scores
asleep quite early for a minute or two and and indices requires clinical judgment and
then remain awake for a considerable time should be made conservatively. Studies have
after that, some laboratories score sleep la- shown that the patient's sleep in the labora-
tency to the first sleep epoch that is followed tory on the first night may be atypically poor
by a minimum of 10 minutes of sleep. (first-night effect). >l5 Alternatively, some in-
Among the three measures of sleep latency, somniacs sleep especially well on the first
latency to stage 1 is preferred by clinicians. night in the laboratory (reverse first-night ef-
Wake after sleep onset is the total amount of fect).16 Because the first night spent in the
wakefulness scored between the time a pa- laboratory is frequently atypical, research
tient falls asleep and the time he or she gets studies often use a minimum of two labora-
up in the morning. Others score it only be- tory nights to assess sleep and discard the
tween first and last sleep, scoring as early first as adaptation. Economically, however,
Table 39-2. Normative Data for Sleep Stages Related to Age
AGE (YEARS)
4 7.5 11 14 17.5 25 35 45 55 65 75
Time in bed (minutes) 618.7 599.7 587.3 506.4 477.4 443.9 439.1 435.3 444.7 458.6 500.1
Total sleep time (minutes) 593.5 580.8 559.9 484.6 451.7 424.6 423.6 407.1 410.3 406.1 393.2
Stage 1
Percentage 2.12 2.30 2.96 3.63 3.88 4.31 4.94 6.60 6.20 8.71 8.03
Minutes 12.7 13.4 16.8 17.8 17.8 18.5 21.3 28.0 26.7 38.6 36.9
Stage 2
Percentage 45.05 47.92 47.76 46.33 49.24 48.96 55.3 54.38 59.76 55.78 53.86
Minutes 270.8 279.7 271.3 226.5 225.9 209.8 238.1 230.4 257.5 246.9 247.6
Delta
Percentage 20.55 20.98 20.93 22.82 23.24 19.22 13.2 10.30 7.78 4.92 5.70
Minutes 123.5 122.4 118.9 111.8 106.6 82.4 56.8 43.6 33.5 21.8 26.2
REM
Percentage 30.01 28.32 27.21 26.16 22.07 26.62 24.84 24.76 21.62 22.26 18.57
Minutes 180.4 165.3 154.6 128.2 101.3 114.1 107.0 104.9 93.2 98.5 85.4
Data from Williams et al.11
Table 39-3. Mayo Sleep Disorders ing sleep. The basic information obtained
Center Guidelines for Evaluation from polysomnography about disordered
of Normal Lower Bounds for breathing includes the frequency and type
REM Latency of breathing disturbances, how severely oxy-
hemoglobin saturation is affected, and if it
Age (years) Initial REM Latency (minutes) is associated with cardiac arrhythmias. It is
also of clinical relevance to determine if
15-24 70 there is a difference in the degree of disor-
25-34 60 dered breathing related to body position or
35-44 45 sleep stage.
45-60 35
61 + 30
Definitions
Disordered breathing events associated with
sleep traditionally have been classified as ei-
this is not feasible for clinical studies. There- ther apneas or hypopneas. Recent study of
fore, clinical studies are rarely performed to the upper airway resistance syndrome has
assess sleep architecture alone but usually led to the additional definition of respiratory
to assess factors that disrupt sleep, such as effort-related arousals (RERAs).13 Apnea is de-
breathing or movement disorders. fined as complete cessation of airflow, and
A comparison between the objectively hypopnea is a partial decrease in airflow. In
recorded sleep variables as discussed in this theory, apnea and hypopnea are distinct, but
chapter and the patients' self-reports of their the difference in their clinical relevance is
sleep (obtained by their answers to a ques- slight, and reliable discrimination of the two
tionnaire the following morning) may yield types of events is technically difficult. Recent
clinical insight, especially if there is a large recommendations from the American Acad-
discrepancy. Small discrepancies of up to ap- emy of Sleep Medicine17 combine apnea and
proximately 90 minutes in total sleep time hypopnea as a single event type. For a tran-
are within normal limits. If the discrepancy sient reduction in breathing to be scored as
is much larger, it is important to consider apnea/hypopnea, it must (1) last 10 seconds
whether it can be explained by (1) a high or longer, with a clear decrease (> 50%)
arousal index or alpha intrusions into sleep from baseline in the amplitude of a valid
or (2) a sleep-state misperception, that is, to- measure of airflow during sleep or (2} be as-
tally normal sleep lasting a minimum of 6.5 sociated with either an oxyhemoglobin de-
hours in a patient who reports sleeping lit- saturation greater than 3% or an arousal.
tle or not at all. Development of sleep over Respiratory effort-related arousals are re-
an entire night is best represented by results lated to transient increases in airway resis-
of a sleep histogram (Fig. 39-11). This can tance that lead to arousal. There must be
indicate whether the sleep disturbances (for evidence of progressively increasing respira-
example, disordered breathing and periodic tory effort terminated by an arousal and re-
limb movements) are evenly distributed over sumption of normal respiratory efforts.
the entire sleep period or are associated with It is also necessary to determine if an ap-
specific sleep stages or times of the night. nea or hypopnea is obstructive or central.
They can also show whether excessive wake- Obstructive apnea is defined as a cessation of
fulness occurs early or late (possibly sug- airflow in the presence of measurable, often
gesting influences of the circadian rhythm gradually increasing, respiratory effort. Dur-
or psychiatric factors). ing obstructive apnea, paradoxical breathing
is often observed, that is, the chest expands
as the abdomen contracts, but this is not
ASSESSING RESPIRATION essential for diagnosis (Fig. 39-12A). Obstruc-
DURING SLEEP tive hypopnea has features similar to obstruc-
tive apnea and is often marked by crescendo
Most polysomnographic studies are per- inspiratory snoring. Central apnea shows a
formed to assess disordered breathing dur- cessation of airflow coupled with a lack of
Figure 39-11. Sleep histogram of, A, a 19-year-old woman with no history of a sleep disorder and, B, a 58-year-old man with severe obstructive sleep apnea. The man has severely
fragmented sleep during the diagnostic study, with marked rebound of rapid eye movement sleep during the CPAP trial. MT, movement time; PLM, periodic limb movements
of sleep; CPAP, air pressure (cm HaO) of continuous positive airway pressure system, (continued)
Stokes breathing, during sleep indicates ei-

ther poor cardiac output (longer circulation
time between lung and blood-gas sensors in
the carotid body) or problems with neuronal
control of respiration.
Apnea events with an initial central com-
ponent followed by an obstructive compo-
nent are called mixed apneas (Fig. 39-12 Q.
To evaluate disordered breathing events,
a polysomnographic montage includes air-
flow through the nose and mouth, breath-
ing effort, snoring intensity, a continuous
measure of oxyhemoglobin saturation, and
cardiac rhythm. As mentioned above, the
body position of the sleeper also has to be
assessed. If continuous positive airway pres-
sure (CPAP) is considered as a treatment
for disordered breathing during sleep,
polysomnography is used to determine the
minimal CPAP pressure that eliminates all
disordered breathing (including snoring) in
all sleep stages and in all positions.18 Ther-
apeutic trials in the laboratory help the pa-
tient to adapt to wearing the mask while
sleeping under constant professional super-
vision. During the CPAP trial, pressure must
be monitored either by continuous record-
ing on the polygraph or by careful written
notation of pressure changes.
Airflow
The reference standard for measurement of
airflow is a pneumotachometer, a device that
requires a snug-fitting mask over the face of
Figure 39-12. Types of sleep apnea. A, Obstructive ap- the subject. Pneumotachometers provide an
nea is characterized by loud intermittent snoring, com-
plete cessation of airflow, paradoxical movement of the accurate quantitative measurement of air-
rib cage and abdomen, and moderate-to-severe oxygen flow, but the constrictive mask is uncom-
desaturation. B, Central apnea has absence of snoring, fortable and tends to disturb sleep. Its use is
simultaneous cessation of airflow and respiratory effort, limited to research applications. The most
and usually only mild-to-moderate oxygen desaturation.
C, Mixed apnea—an initial central apnea followed by
common method for detecting airflow dur-
obstructive apnea that usually produces moderate-to- ing clinical polysomnography is the use of
severe oxygen desaturation. (From Kaplan J. Diagnosis thermal sensors (thermocouples or thermis-
and therapy of sleep-disordered breathing. In Burton tors) placed in the airstream at each nostril
GG, Hodgkin JE, Ward JJ [eds]. Respiratory Care: A and the mouth. Thermal sensors detect the
Guide to Clinical Practice, 3rd ed. JB Lippincott Com-
pany, Philadelphia, 1991, pp 279-287. By permission of temperature change between inspired and
Lippincott Williams & Wilkins.) expired air. Although this does not quanti-
tatively measure how much air is flowing,
changes in signal amplitude indicate relative
respiratory effort (Fig. 39-125). Sleep-onset increases or decreases in airflow. Less com-
central apnea may be relatively benign; it of- monly used techniques for detecting airflow
ten simply indicates anxious overbreathing are capnometry and airway pressure. A cap-
during wakefulness, with normalization dur- nometer senses the variations in COg con-
ing sleep. Periodic central apnea, or Cheyne- centration of inspired and expired air. Air-
way pressure variations are detected with a Some laboratories record breathing effort
cannula placed at the airway openings and by monitoring EMG activity in intercostal
connected to a pressure transducer. muscles. Because intercostal activity is in-
When CPAP is used during polysomnog- hibited during REM sleep, electrodes are
raphy, thermal sensors and capnometers do placed over the sixth or seventh intercostal
not function well. Most CPAP systems de- space in an attempt to also record di-
signed for laboratory use provide analog sig- aphragmatic EMG.
nals for pressure and airflow. These signals
can be interfaced with the polygraph to al-
low continuous monitoring of airflow during Snoring Sounds
the CPAP trial.
In many sleep laboratories, a small micro-
phone is attached to the throat of the pa-
Effort tient or mounted on the headboard or sus-
pended above the bed. The output of the
The purpose of the respiratory pump mus- device may be filtered and recorded directly
cles is to create a pressure gradient between on the polygraph or processed through an
the thoracic cavity and the atmosphere. Re- integrator or sound level meter before
sulting pressure variations can be measured recording. The signal is often recorded on
accurately by placing a transnasopharyngeal audio or video tape for review after the study.
balloon-tipped, transducer-tipped, or water- The recorded signal may be supplemented
filled catheter into the esophagus. Although with the technician's personal judgment, us-
this technique is the reference standard for ing a grading scale for loudness of snoring
measuring respiratory effort, it is invasive in four steps: 1, barely audible; 2, audible at
and can be an unpleasant experience and, the bedside; 3, audible from the open door
thus, is not used routinely. Esophageal pres- to the bedroom; and 4, audible through the
sure is used to definitively identify central ap- closed door.
nea and RERAs.
Inductive plethysmography and various
strain-gauge methods are noninvasive tech- Blood Gases
niques for detecting respiratory effort based
on measuring changes in the dimensions of Continuous monitoring of a blood gas (Og
the thoracic cavity. The most elegant tech- or COg) is mandatory because it provides in-
nique for monitoring respiratory effort dur- formation about the consequences of respi-
ing sleep is inductive plethysmography.19 ratory dysfunction. Monitoring of oxyhemo-
Motion of the diaphragm changes the vol- globin saturation is performed easily with a
ume of the abdominal cavity, and motion of pulse oximeter, a device that measures the
the rib cage changes the volume of the tho- light absorption of two wavelengths of red
racic cavity. The inductive plethysmograph light passed through a capillary bed, such as
has wires embedded in elastic bands that are that of the earlobe or the nail bed of a fin-
placed around the chest and abdomen to ger. The wavelengths used match the peak
sense changes in the cross-sectional area of absorption factors of oxyhemoglobin and
each of these two cavities. An electrical sum- deoxyhemoglobin. The oximeter then cal-
mation of the signals from rib cage and ab- culates the ratio of oxyhemoglobin to total
domen provides a rough estimate of overall hemoglobin and translates it into a digital
tidal volume. display and an analog voltage that is written
Rather than measuring cross-sectional out on the polygraph. Because of lung-ear
area, strain gauges are sensitive to changes or lung-digit circulation time, the nadir of
in thoracic or abdominal circumference. the oxyhemoglobin saturation graph usually
The most commonly used type of strain follows the termination of the respiratory
gauge consists of a piezoelectric transducer event by 7-9 seconds.
connected to a set of straps encircling the In addition to using one channel of the
body. Typically, two piezo belts are used: one polysomnogram for ear oximetry, it is often
around die chest and the other around the useful to record oximetry on an indepen-
abdomen. dent strip-chart recorder, running this strip
Figure 39-13. Oximetry strip chart recording. The repetitive desaturations occurring in the early portion of the
tracing are caused by obstructive apneas. Initiation of continuous positive airway pressure (CPAP) eliminates the
apnea during the later portion of the recording.
at a slow speed of about 2-5 inches/hour. 1. Number of central, obstructive, or

This provides a quick overview of oxygena- mixed apneas and hypopneas. Usually,
tion for the entire night and allows easy iden- this is given as the average number of
tification of periods of maximal respiratory these events per hour of sleep in NREM
dysfunction (Fig. 39-13). If CO2 retention is and REM sleep. The mean duration of
of concern in a patient, end-tidal (capnog- episodes of apnea and hypopnea dur-
raphy) or transcutaneous PCOg may be ing NREM, REM, and total sleep time
recorded. The PCC>2 of air sampled from the is often reported as well as the longest
airway at end-expiration (end-tidal) approx- duration of an apneic episode in each
imates alveolar PCOg. Transcutaneous PCOg stage of sleep.
uses a skin surface electrode to measure the 2. The apnea-hypopnea index (AHI) (also
PCOg in tissue underlying the skin. Both called respiratory disturbance index or dis-
techniques provide data about trends in ar- ordered breathing index) indicates the
terial PCOg but should be supplemented combined number of apneas and hy-
with blood gas analysis of arterial blood popneas per sleep hour. This is usually
samples for accurate diagnosis of alveolar reported for total sleep time, NREM
hypoventilation. and REM sleep time, and sleep time in
the supine vs. other positions. Apnea-
hypopnea index is the standard indi-
Cardiac Rhythm cator of severity for obstructive apnea.
Fewer than five events per hour is con-
The EGG is recorded to detect changes in sidered normal. Five to 15 events per
the cardiac rhythm related to disordered hour is mild, 15—30 is moderate, and
breathing during sleep. A single lead I is de- more than 30 is severe. The recent
rived from electrodes placed on the right trend of including RERA in the sever-
and left shoulders. Additional leads can be ity index for sleep-related disordered
derived during the recording from other breathing has yet to gain general ac-
electrodes in use on the patient. For instance ceptance.17
a lead II can be derived from electrodes on 3. Mean oxygen saturation values during
the right ear and left leg. This provides flex- wakefulness, REM sleep, and NREM
ibility to maintain a signal if electrode in- sleep are usually given or a range is in-
tegrity is lost. dicated. Also important is the percent-
age of time during sleep that the pa-
tient spends with oxygen saturation
above 90%, 80%, or 70%, and so forth,
Summary Statistics for 4. Snoring is often reported on a scale
Respiratory Variables from 1 to 4 (see above under Snoring
Sounds). Clinically, one looks for
The following respiratory variables are typi- crescendo snoring that leads to arousal,
cally reported on the polysomnographic as the sleeper struggles harder and
summary sheet: harder for air.
5. The number of arousals caused by dis- one-channel recording from both legs. Peri-
ordered breathing is especially impor- odic limb movements of sleep are scored
tant because occasionally breathing is only if at least four occur in sequence, with
not disordered enough to be scored as a duration of 0.5-5.0 seconds each and in-
apnea or hypopnea but nevertheless termovement intervals of 4-90 seconds20
awakens the sleeper (upper airway resis- (Fig. 39-14). Periodic limb movements show
tance syndrome). flexions of hip, knee, and ankle joints to-
gether with extension of the toes, similar to
a withdrawal response. Periodic limb move-
ASSESSING PERIODIC ments need to be distinguished from gross
LIMB MOVEMENTS body movements, which also show EMG ar-
tifact in the EGG and EEG leads, and from
Periodic limb movements of sleep (PLMS), usu- body movements associated with arousal
ally of the legs but rarely the arms, may oc- from sleep apnea.
cur at regular intervals during sleep. They Periodic limb movements of sleep occur
are recorded with surface electrodes applied in many normal, especially older, sleepers
over the anterior tibialis muscle of both legs. but may also occur in several disorders, in-
Either one channel of EMG is recorded from cluding restless legs syndrome (RLS). Gener-
both legs or two channels are recorded, one ally, polysomnography is not indicated for
per leg. The latter arrangement eliminates the diagnosis of RLS, which is made on the
much of the EGG artifact that is typical in a basis of the history because the presence of
Figure 39-14. Periodic limb movements (PLM) of sleep. Bursts of anterior tibial surface EMG (Leg EMG) occur-
ring at approximately 20-second intervals are accompanied by subtle arousals lasting 3 seconds or longer.
PLMS is neither sufficiently specific nor sen- ASSESSING OTHER

sitive. However, in situations in which the PHYSIOLOGIC VARIABLES
clinical probability of a patient having RLS
is estimated at about 50%, a polysomnogram Core Temperature
showing PLMS may occasionally be helpful
in establishing the diagnosis.21 When limb Monitoring core temperature may provide
movements are not associated with EEG information about the phase and amplitude
arousal, their significance in causing day- of the circadian cycle. Theoretically, this may
time sleepiness is uncertain. In many cases, be helpful in the diagnosis of circadian
PLMS may be an epiphenomenon accom- rhythm disorders, but in practice it is used
panying other disorders such as narcolepsy only as a research technique.
or idiopathic hypersomnia and may not in
themselves cause symptoms.22 However, if a
high percentage of the movements cause the
sleeper to awaken, PLMS can potentially Esophageal pH Measurements
contribute to excessive daytime somnolence,
although, first, care should be taken to rule Many patients experience reflux of stomach
out other causes of sleepiness.21 In a few lab- contents into the esophagus during sleep.
oratories, accelerometers, which better indi- Clearing of this material from the esopha-
cate the force of the movement, are used. gus is delayed markedly during sleep. Occa-
They may also be useful in patients with var- sionally, the reflux may be related to apnea
ious tremors if there is a question about the or may be associated with insomnia. Occa-
degree to which these tremors interfere with sionally, esophageal pH is measured in sleep
sleep. laboratories.
Periodic limb movements of sleep also
can be recorded during wakefulness, espe-
cially in patients with RLS. This has given SUMMARY
rise to the suggested immobilization test for the
syndrome. In this test, the patient is asked The tests and recording techniques used for
to lie down but to remain awake and not to assessing sleep and physiologic variables dur-
move the legs voluntarily. Anterior tibial ing sleep are discussed, including a review
EMG is recorded, and the periodic limb of polysomnography (all-night sleep stud-
movement index over 30-60 minutes is cal- ies), the MSLT (daytime nap studies to as-
culated. More than 40 PLMS per hour sess excessive daytime fatigue and sleep-
(counting movements with a duration of onset REM periods), and the maintenance
0.5—10 seconds) have both a sensitivity and of wakefulness test. The recording and scor-
specificity of 81% for the diagnosis of RLS.23 ing of sleep stages are explained, as is the as-
The test is often uncomfortable because pa- sessment of respiration and PLMS.
tients with RLS need to move their legs to
relieve the discomfort.
REFERENCES
1. Hirshkowitz M, Moore CA. Issues in computerized
polysomnography. Sleep 17:105-112, 1994.
Summary Statistics for Periodic 2. Rechtschaffen A, Kales A. A manual of standardized
Limb Movements terminology, techniques and scoring system for
sleep stages of human subjects. Brain Information
Service/Brain Research Institute, Los Angeles, Cal-
A periodic limb movement index (that is, ifornia, 1968.
the number of periodic limb movements 3. Carskadon MA, Dement WC, Mitler MM, Roth T,
per sleep hour) is reported, followed by a Westbrook PR, Keenan S. Guidelines for the mul-
periodic limb movement-arousal index. tiple sleep latency test (MSLT): a standard measure
of sleepiness. Sleep 9:519-524, 1986.
This is the number of periodic limb move- 4. Kingshott RN, Douglas NJ. The effect of in-labora-
ments per sleep hour that led to arousal or tory polysomnography on sleep and objective day-
awakening. time sleepiness. Sleep 23:1109-1113, 2000.
5. Doghramji K, Mitler MM, Sangal RB, et al. A nor- ("first night effect"). Electroencephalogr Clin Neu-
mative study of the maintenance of wakefulness test rophysiol 22:556-558, 1967.
(MWT). Electroencephalogr Clin Neurophysiol 103: 16. Hauri PJ, Olmstead EM. Reverse first night effect
554-562, 1997. in insomnia. Sleep 12:97-105, 1989.
6. Sleep Disorders Adas Task Force of the American 17. The Report of an American Academy of Sleep Med-
Sleep Disorders Association. EEG arousals: scoring icine Task Force. Sleep-related breathing disorders
rules and examples. Sleep 15:173-184, 1992. in adults: recommendations for syndrome defini-
7. Bonnet MH. Performance and sleepiness as a function and measurement techniques in clinical re-
tion of frequency and placement of sleep disrup- search. Sleep 22:667-689, 1999.
tion. Psychophysiology 23:263-271, 1986. 18. Sullivan CE, Crunstein RR. Continuous positive air-
8. Hauri P, Hawkins DR. Alpha-delta sleep. Electro- way pressure in sleep disordered breathing. In
encephalogr Clin Neurophysiol 34:233-237, 1973. Kryger MH, Roth T, Dement WC (eds). Principles
9. Moldofsky H, Lue FA. The relationship of alpha and and Practice of Sleep Medicine. WB Saunders, Phil-
delta EEG frequencies to pain and mood in 'fi- adelphia, 1994, pp 694-705.
brositis' patients treated with chlorpromazine and 19. Staats BA, Bonekat HW, Harris CD, Offord KP.
L-tryptophan. Electroencephalogr Clin Neurophys- Chest wall motion in sleep apnea. Am Rev Respir
iol 50:71-80, 1980. Dis 130:59-63, 1984.
10. Schenck CH, Mahowald MW. REM sleep parasom- 20. The Adas Task Force. Recording and scoring leg
nias. Neurol Clin 14:697-720, 1996. movements. Sleep 16:748-759, 1993.
11. Williams RL, Karacan I, Hursch CJ. Electroen- 21. Silber MH. Commentary on controversies in sleep
cephalography (EEG) of Human Sleep: Clinical Ap- medicine (Montplaisir et al. Periodic leg move-
plications. John Wiley & Sons, New York, 1974. ments are not more prevalent in insomnia or hy-
12. Mathur R, Douglas NJ. Frequency of EEG arousals persomnia but are specifically associated witii sleep
from nocturnal sleep in normal subjects. Sleep disorders involving a dopaminergic mechanism).
18:330-333, 1995. Sleep Med 2:367-369, 2001.
13. Guilleminault C, Stoohs R, Clerk A, Cetel M, 22. Montplaisir J, Michaud M, Denesle R, Gosselin A.
Maistros P. A cause of excessive daytime sleepiness. Periodic leg movements are not more prevalent in
The upper airway resistance syndrome. Chest 104: insomnia or hypersomnia but are specifically asso-
781-787, 1993. ciated with sleep disorders involving a dopaminer-
14. Agnew HW Jr, Webb WB, Williams RL. The first gic impairment. Sleep Med 1:163-167, 2000.
night effect: an EEG study of sleep. Psychophysiol- 23. Montplaisir J, Boucher S, Nicolas A, et al. Immobi-
ogy 2:263-266, 1966. lization tests and periodic leg movements in sleep
15. Mendels J, Hawkins DR. Sleep laboratory adapta- for die diagnosis of restiess leg syndrome. Mov Dis-
tion in normal subjects and depressed patients ord 13:324-329, 1998.
Chapter 40
ASSESSING SLEEP DISORDERS
Peter J. Hauri
Cameron D. Harris
Michael H. Silber
INDICATIONS FOR POLYSOMNOGRAPHY

PARTIAL EVALUATIONS
PERFORMANCE OF A SLEEP STUDY
DISORDERS OF EXCESSIVE DAYTIME
SOMNOLENCE
PARASOMNIAS
SUMMARY
INDICATIONS FOR 3. Assessment of parasomnias. An extended

POLYSOMNOGRAPHY polysomnographic study using addi-
tional electroencephalographic (EEG)
Polysomnographic studies are indicated in and electromyographic (EMG) chan-
the diagnosis of the following disorders:1 nels is indicated for sleep-related vio-
1. Assessment of sleep-disordered breath- lent or potentially injurious behavior or
ing, including suspected obstructive behavior that is highly disturbing to the
sleep apnea or hypopnea syndrome or sleep of relatives. Such a study is also
central sleep apnea syndrome. Poly- indicated for sleep behaviors that are
somnography is also indicated for titra- unusual or atypical because of age at
tion of continuous positive airway pres- onset, frequency or duration of occur-
sure (CPAP) in patients with diagnosed rence, or die specifics of the particular
sleep-disordered breathing. Polysom- motor pattern in question. Paroxysmal
nography is not routinely indicated for arousals thought to be possible noctur-
nocturnal hypoxemia in the presence nal seizures are a further indication,
of chronic lung disease unless sleep- when the results of standard EEG are
related upper airway obstruction is also inconclusive. Polysomnography is not
suspected. The mere presence of snor- routinely indicated for typical, uncom-
ing, obesity, systemic hypertension, or plicated, and noninjurious parasom-
nocturnal cardiac arrhythmia without nias that can easily be diagnosed on the
other symptoms is not an indication for basis of the medical history or for pa-
polysomnography. tients with known nocturnal seizures
2. Assessment of excessive daytime som- but no other sleep complaints.
nolence (narcolepsy, idippathic hy- 4. Assessment of periodic limb movements
persomnia). This assessment requires of sleep. Polysomnography is indicated
polysomnography and a subsequent when periodic limb movements of
multiple sleep latency test (MSLT). sleep are suspected to be the cause of
513
insomnia or excessive daytime sleepi- sleep without causing significant oxy-

ness, especially in the absence of rest- hemoglobin desaturation. The patient
less legs syndrome. The frequency of may not have slept much during the
periodic limb movements of sleep and night when oximetry was performed or
their effect on sleep continuity and ar- may have positional sleep apnea and
chitecture need to be determined. happened to sleep only on the side dur-
Polysomnography is not indicated for ing the study.
the diagnosis of restless legs syndrome. 2. Hotter monitoring—Frequent bradycar-
Polysomnography is not indicated for the dia alternating with tachycardia, dem-
routine evaluation of transient or chronic in- onstrated with Holter monitoring, dur-
somnia.2 However, it may be indicated if ing the night suggests sleep apnea
sleep-disordered breathing or periodic limb syndrome. However, most patients with
movements of sleep are thought to comprise a serious obstructive sleep apnea syn-
a significant component of the insomnia, if drome show entirely normal cardiac
the initial diagnosis is uncertain, or if be- rhythms throughout the period of sleep.
havioral or pharmacologic treatment has 3. Wrist actigraphy—A small, watch-like de-
been unsuccessful. In rare cases, poly- vice is worn by the patient on the wrist
somnography may be performed if a sleep- of the nondominant hand, usually for
state misperception syndrome is suspected. an entire week (Fig. 40-1). This device
Such a syndrome is defined as a complaint counts and stores in its memory the
of insomnia or excessive sleepiness that oc- number of wrist movements that occur
curs without any objective evidence of sleep for each 1-minute epoch. Periods of rel-
disturbance. In this case, the polysomno- ative absence of such movements are
gram would demonstrate normal sleep la- interpreted as sleep and periods of
tency (fewer than 15-20 minutes), a normal high activity as wakefulness. Although
number of arousals and awakenings (fewer the validity of this method is adequate
than 10/hour), and normal sleep duration for estimating sleep in normal sub-
(longer than 6.5 hours). The MSLT may or jects,4 its accuracy is significantly less in
may not show abnormal sleep latencies. people with sleep disorders such as ap-
nea or insomnia.5 Patients with peri-
odic limb movements or obstructive
sleep apnea show an excessive number
PARTIAL EVALUATIONS of movements and on the actigraph
look as though they are awake when
Because complete polysomnographic stud- they are actually asleep. Because pa-
ies are expensive, screening with partial tients with depression or psychophysi-
recordings is sometimes performed. The ologic insomnia may move very little
American Academy of Sleep Medicine has while lying in bed awake, their total
reviewed this practice in the case of assess- sleep may be overestimated by wrist
ing obstructive sleep apnea syndrome and actigraphy.
has found that under certain, carefully de- 4. Continuous ambulatory EEG monitoring—
fined conditions, unattended (portable) re- This may be adequate to assess sleep ar-
cording for the assessment of obstructive chitecture, but it cannot determine the
sleep apnea is acceptable.3 These include cause of frequent arousals, such as dis-
the following: ordered breathing or periodic limb
1. Nocturnal oximetry—Although the pres- movements.
ence of repetitive desaturations in an 5. Unattended portable polysomnography—
oximetric tracing strongly suggests This may be indicated for patients with
sleep apnea syndrome, normal findings suspected severe obstructive sleep ap-
on oximetry do not rule out disordered nea when initiation of treatment is ur-
breathing. Patients, especially younger gent and standard polysomnography is
ones, may have sleep apneas serious not readily available, for medically un-
enough to cause repeated arousals from stable patients unable to be studied in
Assessing Sleep Disorders 515
figure 48-1. Wrist actigraphy data recorded from 18:00 (6:00 P.M.) Friday through 18:00 the following Friday. The
height of each vertical black line is proportional to the level of activity for a 1-minute epoch. Areas where the lines
are very short or absent represent periods of immobility and probable sleep.
the sleep laboratory, and for follow-up ensure the possibility of normal sleep
studies to evaluate the effects of treat- (no telephone calls, children crying,
ment. In our opinion, such a system and so forth).
should contain a minimum of four
channels to evaluate sleep (two electro-
oculographic channels, one EEC chan- PERFORMANCE OF
nel, and one chin EMG channel) and A SLEEP STUDY
other channels to evaluate airflow,
oxyhemoglobin saturation, respiratory Before evaluation, patients are usually sent
effort, anterior tibialis EMG, and electro- questionnaires about sleep and a sleep log
cardiography (EGG). However, Ameri- to be filled out for a minimum of 1 week. A
can Academy of Sleep Medicine stan- sleep disorder specialist, ideally board cer-
dards of practice allow monitoring with tified in Sleep Medicine then interviews the
a minimum of ventilation (at least two patient. If indicated, polysomnography with
channels), EGG or heart rate, and oxy- or without a subsequent MSLT is then
hemoglobin saturation in these limited scheduled.
circumstances. The home environment On the appointed night, patients come to
also has to be adequately monitored to the sleep center at about 8 P.M. First, they fill
out a questionnaire about their activity dur- up conference with the patient is scheduled
ing the past day, about their intake of med- to discuss the results, to make a final diag-
icine, coffee, and alcohol, and about emo- nosis, and to initiate treatment if indicated.
tional issues that might disrupt sleep. A
technician applies the electrodes and sen-
sors necessary for the polysomnographic DISORDERS OF EXCESSIVE
study. Patients then watch television or read DAYTIME SOMNOLENCE
until they are ready to go to bed. After the
patients are connected to the polysomno- Excessive daytime somnolence is defined as the
graph, biocalibrations are made (patients tendency to fall asleep very easily. The abil-
are asked to move and blink their eyes, move ity to fall asleep quickly differentiates exces-
their legs, grit their teeth, breathe exclusive daytime somnolence from chronic fa-
sively through either the nose or mouth, and tigue, malaise, or low-grade depression, in
perform an isovolume maneuver to maxi- which patients may be exhausted and lie in
mize the recording of paradoxical breath- bed for most of the day but are usually un-
ing) . Patients stay in bed for a minimum of able to fall asleep quickly when asked to do
6 hours after "lights out" (often longer) ex- so. Excessive daytime somnolence may be
cept for short trips to the bathroom. A tech- caused by an extrinsic or intrinsic cause. The
nician monitors the recording throughout most common extrinsic cause is voluntary
the night, observes the patients through sleep deprivation (insufficient sleep syndrome),
video monitoring, and notes on die record often found in people who work two jobs or
a patient's position and any potentially sig- have other reasons for not allowing them-
nificant events (for example, patient vocal- selves enough time in bed. Normal sleep
ization, environmental noise that arouses needs vary from person to person, ranging
the patient, and loudness of snoring). from 3 to 4 hours to 9 to 10 hours per night.
If a diagnosis of obstructive sleep apnea If insufficient sleep syndrome is suspected,
syndrome is made early in the night, the sec- it is often useful to ask patients with exces-
ond half of the night may be used for a ther- sive daytime somnolence to stay in bed 1
apeutic trial of nasal CPAP. The American hour longer for an entire week, regardless
Academy of Sleep Medicine recommends of how long they have stayed in bed until
that such "split-night" studies be restricted that time, and observe whether excessive
to patients with apnea-hypopnea indices daytime somnolence wanes. Other extrinsic
greater than 40 (or greater than 20 if apneas causes include the use of sedating medica-
are long with major desaturations) in the tions, alcohol or recreational drugs, shift
first half of the night. However, economic work, or environmental noise (including the
pressures have forced an increasing number snoring of a spouse8).
of sleep laboratories to perform "split-night" The most common intrinsic cause of ex-
studies routinely if disordered breathing of cessive sleepiness is obstructive sleep apnea
any severity is detected in the first half of the or hypopnea syndrome. Other causes in-
study. No differences in medium6 and long- clude central sleep apnea syndrome, nar-
term7 CPAP use have been found in patients colepsy, idiopathic hypersomnia, disorders
titrated with a split-night protocol compared of circadian rhythm such as delayed sleep
with those undergoing a full-night CPAP phase syndrome, and periodic limb move-
study. Occasionally, medications (such as ments of sleep (if these are associated with
levodopa/carbidopa for periodic move- a high percentage of arousals). Polysomnog-
ments) are administered during the second raphy (with or without a subsequent MSLT)
part of the night. is needed to diagnose most intrinsic causes
In the morning, patients fill out another (circadian rhythm disorders are excep-
questionnaire about their perception of the tions). Initially, polysomnography is per-
quality of their sleep in the laboratory. A formed. If a cause, such as sleep apnea, is
technician then scores the recordings and found, the MSLT scheduled for the follow-
computes various indices. After the sleep ing day is canceled and the problem is
specialist has reviewed the record, a follow- treated. If the polysomnogram is normal
and nighttime sleep is adequate (6 or 7 PARASOMNIAS

hours), the MSLT is performed (see Chap-
ter 38). A mean initial sleep latency over Parasomnias are undesirable, mainly motor,
four or five nap opportunities is calculated, phenomena that occur during sleep. They
as described in Chapter 38. A value of fewer include disorders of arousal (for example,
than 5 minutes indicates moderate to severe sleepwalking and sleep terrors), wake-sleep
excessive daytime sleepiness, and a value of transition disorders (for example, sleep
5-8 minutes usually is regarded as mild day- starts), REM-related parasomnias (for ex-
time sleepiness. Values greater than 10 min- ample, REM sleep behavior disorder, night-
utes are normal, and those between 8 and mares) , and miscellaneous parasomnias (for
10 minutes are considered to be in a "gray example, bruxism, nocturnal enuresis). The
area," requiring clinical judgment for inter- indications for studying parasomnias are dis-
pretation. If one of the initial sleep laten- cussed earlier under Indications for Poly-
cies is an oudier (for example, 3-, 4-, 5-, 4-, somnography. The laboratory assessment of
and 18-minute initial latencies), it is more parasomnias requires more sophisticated
accurate to use the median than the mean recording systems and additional interpre-
sleep latency.9 Rapid eye movement (REM) tive abilities than routine polysomnogra-
sleep occurring within 15 minutes after phy.11 Additional polygraphic channels are
sleep onset (sleep onset REM [SOREM]) in needed, including one or more channels for
any of the nap opportunities is also recording arm EMG (usually over the ex-
recorded. The presence of SOREMs in two tensor digitorum muscles) and additional
or more naps, or in one nap and the pre- EEG channels (preferably 16), to rule out
ceding overnight polysomnographic study, seizures that can mimic parasomnias (Fig.
is considered abnormal. Although this find- 40-2). Digital recording systems are ideal for
ing is regarded as the neurophysiologic this technique, but analog recorders are still
marker of narcolepsy, it is not specific and used occasionally, with paper speed in-
other causes need to be considered. These creased to 15 mm/second to allow for bet-
include abrupt withdrawal from REM-sup- ter EEG resolution. High-resolution, time-
pressant medication (for example, most synchronized videotape recording under
stimulants and antidepressants), moderate infrared light is essential. Digital video is the
or severe obstructive sleep apnea syn- current state of the art because it allows ex-
drome,10 or sleep deprivation. In addition, tremely rapid correlation with polysomno-
a history suggestive of narcolepsy that is con- graphic data but it uses large amounts of
firmed by a short mean initial sleep latency computer memory. Physicians interpreting
on the MSLT is necessary to make the di- parasomnia studies should be familiar with
agnosis. If the sleepy patient has no history both routine polysomnographic interpreta-
of cataplexy, the results of the polysomnog- tion and ictal EEG patterns.
raphy are normal, and the MSLT shows Disorders of arousal (sleepwalking, sleep
pathologically short sleep latency with fewer terrors, or confusional arousals) are charac-
than two SOREMs, the likely diagnosis is id- terized by sudden arousals from slow-wave
iopathic hypersomnia. sleep, followed by abnormal motor behavior.
Occasionally, a problem in assessing ex- They are most common in children but may
cessive daytime somnolence is caused by de- commence or persist in adulthood. Al-
layed sleep phase syndrome, that is, the in- though the conditions are usually benign,
ability to fall asleep until the early morning occasional patients may suffer or inflict se-
hours coupled with routine sleep until noon vere injuries. The polysomnogram shows a
or later. If another disorder, such as nar- rapid, usually unprovoked arousal from
colepsy, is suspected in these patients, stage 3 or 4 non-REM sleep. This is some-
polysomnography is performed during rou- times preceded by a series of hypersynchro-
tine sleeping hours (for example, 4 A.M. un- nous delta waves without epileptiform activ-
til 1 P.M.) and the MSLT is conducted dur- ity, seen in the scoring channel in 47% of
ing routine wakefulness (for example, naps events.12 The actual event usually is ob-
at 3 P.M., 5 P.M., 7 P.M., 9 P.M., and 11 P.M.). scured by movement artifact, but it may show
Figure 40-2. Parasomnia recording using an expanded electroencephalographic montage, illustrating arousal para-
somnia. There is a sudden partial arousal from stage 4 nonrapid eye movement sleep without epileptiform activity.
Note that the arousal is preceded by a series of hypersynchronous delta waves and that some slow activity contin-
ues throughout the arousal. In this 9-year-old boy, the episodes of sleep terror were precipitated by stridor (visible
as deflections on the sonogram channel) as a result of vocal cord paresis following surgery for a posterior fossa
medulloblastoma. LOG, left outer canthus; ROC, right outer canthus; sonograph, recording of upper airway sound.
(From Daube JR, Cascino GD, Dotson RM, Silber MH, Westmoreland BF. Continuum: Lifelong Learning in Neu-
rology [Clinical Neurophysiology]. Vol. 4, Part A. Lippincott Williams & Wilkins, Baltimore, 1998, p 166. By per-
mission of the American Academy of Neurology.)
tachycardia and varied EEC patterns, in- nature of the event. Even if a typical episode
cluding rhythmic delta activity, irregular is not recorded the night of the study, care-
mixed frequency activities (predominantly ful review of the tracing often reveals the
delta and theta), or alpha rhythm.12 Com- presence of minor confusional arousals.
pared with normal age-matched controls, REM sleep behavior disorder is characterized
the polysomnogram of patients with disor- by an abnormal persistence of muscle tone
ders of arousal shows a higher percentage of during REM sleep with dream enactment be-
slow wave sleep, more frequent arousals havior.14 This usually takes the form of arm
from slow wave sleep, and a more even dis- flailing and kicking with vocalizations. If th
tribution of slow wave sleep through the patient is wakened, a violent dream is often
night.13 Although these findings shed inter- recalled. Injuries to the patient and bed part-
esting light on the pathogenesis of the dis- ner are common. The condition occurs most
order, they are not specific enough to be of frequently in older men and is often associ-
diagnostic help. The polysomnographic ap- ated with neurodegenerative disease such
pearances of sleep terrors, sleepwalking, and as Parkinson's disease, dementia with Lewy
confusional arousals are identical, and video bodies, or multiple system atrophy. The poly-
recording is essential to delineate fully the somnogram shows abnormally increased
phasic muscle activity in REM sleep and oc- for the use of polysomnography in the evaluation
casionally a persistent tonic EMG. Even if no of insomnia. Sleep 18:55-57, 1995.
3. Standards of Practice Committee of the American
gross movements are recorded the night of Sleep Disorders Association. Practice parameters
the study, muscle tone during REM sleep for the use of portable recording in the assessment
usually remains high and can be recognized. of obstructive sleep apnea. Sleep 17:372-377, 1994.
It is essential to record an additional upper 4. Sadeh A, Sharkey KM, Carskadon MA. Activity-
based sleep-wake identification: an empirical test of
extremity EMG channel, because not all methodological issues. Sleep 17:201-207, 1994.
skeletal muscles are involved in any one 5. Hauri PJ, WisbeyJ. Wrist actigraphy in insomnia.
patient. Sleep 15:293-301, 1992.
6. Anonymous. The impact of split-night polysomnog-
raphy for diagnosis and positive pressure therapy
titration on treatment acceptance and adherence
in sleep apnea/hypopnea. Sleep 23:17-24, 2000.
SUMMARY 7. McArdle N, Grove A, Devereux G, Mackay-Brown
L, Mackay T, Douglas NJ. Split-night versus full-
This chapter describes the indications and night studies for sleep apnoea/hypopnoea syn-
drome. Eur Respir J 15:670-675, 2000.
practice variables for polysomnographic 8. Beninati W, Harris CD, Herold DL, Shepard JWJr.
studies. The roles of partial evaluations and The effect of snoring and obstructive sleep apnea
screening tests are discussed. The practical on the sleep quality of bed partners. Mayo Clin Proc
performance of a sleep study is described. 74:955-958, 1999.
The disorders of excessive daytime somno- 9. Benbadis SR, Perry M, Wolgamuth BR, Turnbull J,
Mendelson WB. Mean versus median for the mul-
lence and their evaluation, including the tiple sleep latency test. Sleep 18:342-345, 1995.
clinical use and interpretation of the MSLT, 10. Aldrich MS, Chervin RD, Malow BA. Value of the
are reviewed. Extended polysomnography multiple sleep latency test (MSLT) for the diagno-
for evaluation of parasomnias is outlined, sis of narcolepsy. Sleep 20:620-629, 1997.
with special discussion of disorders of arousal 11. Aldrich MS, Jahnke B. Diagnostic value of video-
EEG polysomnography. Neurology 41:1060-1066,
and REM sleep behavior disorder. 1991.
12. Schenck CH, Pareja JA, Patterson AL, Mahowald
MW. Analysis of polysomnographic events sur-
rounding 252 slow-wave sleep arousals in thirty-
eight adults with injurious sleepwalking and sleep
REFERENCES terrors. J Clin Neurophysiol 15:159-166, 1998.
13. Espa F, Ondze B, Deglise P, Billiard M, Besset A.
1. Polysomnography Task Force, American Sleep Dis- Sleep architecture, slow wave activity, and sleep
orders Association Standards of Practice Commit- spindles in adult patients with sleepwalking and
tee. Practice parameters for the indications for sleep terrors. Clin Neurophysiol 111:929-939,2000.
polysomnography and related procedures. Sleep 14. Olson EJ, Boeve BF, Silber MH. Rapid eye move-
20:406-422, 1997. ment sleep behaviour disorder: demographic, clin-
2. Standards of Practice Committee of the American ical and laboratory findings in 93 cases. Brain
Sleep Disorders Association. Practice parameters 123:331-339, 2000.
SECTION 2
Electrophysiologic
Assessment of
Neural Function
Part H
Intraoperative
Monitoring
The central and peripheral nervous systems are at risk for damage during
surgical procedures, particularly vascular, orthopedic, and neurosurgical
procedures. Although some types of damage may be expected because of
the nature of the procedure, other types can occur unexpectedly. In either
case, damage may be reversible if the surgeon is made aware of the change
and takes appropriate action. Standard clinical tools cannot assess neural
function during surgical procedures. Therefore, surgeons have relatively lit
tle information on which to base decisions about modifying the procedure
in response to impending damage to neural tissue.
Most electrophysiologic measurements described in this book can be made
intraoperatively to monitor neural function. Electroencephalography can be
used to monitor the status of cortical function; somatosensory evoked po-
tentials, to monitor sensory pathways in the periphery, spinal cord, and brain;
auditory evoked potentials, to monitor peripheral and central auditory path-
ways; nerve conduction studies and electromyography, to monitor periph-
eral nerve damage; and motor evoked potentials, to monitor descending
motor pathways in the brain stem and spinal cord. Each of these techniques
has been modified so it can be used in the operating room for a wide vari-
ety of surgical procedures. These monitoring techniques provide helpful
guidance to the surgeon during the procedure and have reduced morbid-
ity associated with certain procedures. The optimal monitoring methods vary
with the level and type of surgical procedure: supratentorial (Chapter 41),
posterior fossa (Chapter 42), spinal column (Chapter 43), and peripheral
nerve (Chapter 44).
As surgeons have become more familiar with the benefits of intraopera-
tive monitoring of neural function, the demand for it has increased. This
522 Intraoperative Monitoring
change is reflected in the expanded discussion in this edition of the com-

pound muscle action potential, motor evoked potential, electromyographic
recordings, carotid stump, and compressed spectral array and revision of
the discussion on intraoperative monitoring during nerve entrapment pro-
cedures. With increased experience with intraoperative monitoring, more
specific criteria have been developed to prevent damage.
Continuous electrophysiologic monitoring can be helpful in the intensive
care unit. Patients in an intensive care unit either have a major neurologic
disease or are at risk of having one develop as a complication of another
disorder. Electrophysiologic techniques can be used to monitor neural func-
tion in this setting just as they can be in the operating room to identify early
or otherwise unrecognizable neural damage. The chapters in this section il-
lustrate the applications of electrophysiologic techniques both in the oper-
ating room and in the intensive care unit.
Chapter 41
CEREBRAL FUNCTION MONITORING
Elson L. So
Frank W. Sharbrough
TECHNICAL FACTORS IN ELECTROENCEPHALOGRAPHY AND

INTRAOPERATIVE OTHER MONITORING TECHNIQUES IN
ELECTROENCEPHALOGRAPHIC RELATION TO SURGICAL STRATEGY
MONITORING DURING CAROTID ENDARTERECTOMY
SYMMETRICAL ELECTROENCEPHALOGRAPHIC
ELECTROENCEPHALOGRAPHIC MONITORING DURING CARDIAC
PATTERNS DURING ANESTHESIA SURGERY
PREOPERATIVE FOCAL ABNORMALITIES SUMMARY
FOCAL ELECTROENCEPHALOGRAPHIC
CHANGES DURING CAROTID
ENDARTERECTOMY
The advent of digital electroencephalo- tions.3 However, EEG monitoring during the
graphic (EEG) recording has made intraop- latter procedure is limited by the effect of
erative EEG monitoring easier to perform. hypothermia, which often suppresses EEG
It eliminates the inconvenience and the inactivity,4 making it ineffective as a monitor-
efficiency associated with continuous paper ing tool for ischemia. Electroencephalo-
recording. Any portion of the recording can graphic monitoring for cardiac surgery is
be readily accessed and displayed in differ- further limited because easily correctable
ent formats, and one segment of the record- causes of ischemia occur less commonly than
ing can be displayed side-by-side with an- during operations on the carotid artery.
other segment for visual comparison. Digital
recording also allows monitoring by the elec-
trophysiologist from a location other than TECHNICAL FACTORS IN
the operating room. Also, digital EEG sig-
nals can be subjected to computer analysis INTRAOPERATIVE
for detecting subde changes in EEG activity ELECTROENCEPHALOGRAPHIC
during surgery. MONITORING
Among the first applications of intraoper-
ative recording of cerebral electrical activity To ensure reliable EEG recordings during
were corticography and, later, acute depth intraoperative monitoring, it is important to
studies during epilepsy surgery.1 More re- pay special attention to technical factors:5
cently, scalp EEG recordings (with or with- application of electrodes with collodion for
out special computer processing) have been stability, an adequate number of electrodes
used routinely in many medical centers to (at least 8 and preferably 16 channels), fil-
monitor cerebral electrical activity during tering of background noise with a 60 Hz
endarterectomy2 and cardiac bypass opera- notch filter, use of adequate sensitivity, and
523
proper grounding for patient safety. The (Figs. 41-1 to 41-4). Even a longer time
time scale used to display EEG data is a par- scale display, corresponding to a paper speed
ticularly important technical factor, because of 5 mm/second (one-sixth of that used in
a large amount of data is generated during standard recordings), is adequate for de-
intraoperative EEG monitoring and it must tecting important EEG changes of ischemia
be compressed without essential data being during intraoperative clamping (Fig. 41-5).
lost. Computer processing with spectral anal- If unusual EEG changes are suspected while
ysis presented as a compressed spectral a slower time scale is being used, returning
array has been used for this purpose." How- to the regular time scale or paper speed of
ever, data can be compressed visually with- 30 mm/second will permit prompt identifi-
out using spectral analysis simply by altering cation of abnormal patterns (Fig. 41-1). Re-
the time scale in digital EEG recordings, for cent reports suggest that computerized
example, slowing the paper speed initially to quantitative analysis of EEG may comple-
a rate of 15 mm/second, which is half the ment visual analysis in detecting the EEG
speed used in standard EEG recordings7 changes.8'9
Figure 41-1. Electroencephalogram (EEG) with isoflurane anesthesia. A, EEG recorded with routine paper speed
(30 mm/second). Common patterns of anesthetic—anterior maximum, rhythmic fast (ARF) alpha pattern, ante-
rior maximum, triangular, slow (ATS) pattern, and posterior, arrhythmic, slow (PAS) pattern—are seen in the left
hemisphere. In the right hemisphere, there is an increase in the amount of irregular slowing, especially in tem-
poral distribution, and reduction of the normal ARF alpha pattern. B, The same changes can be appreciated qual-
itatively even at a reduced paper speed of 5 mm/second (one-sixth the usual paper speed). Note that this patient
has a focal abnormality while under anesthesia, even though the patient had had only a right cerebral transient
ischemic attack (TIA) without residual deficit. (From Daube JR, Harper CM, Litchy WJ, Sharbrough FW. Intraop-
erative monitoring. In Daly DD, Pedley TA [eds]. Current Practice of Clinical Electroencephalography, 2nd ed.
Raven Press, New York, 1990, p 743. By permission of Mayo Foundation.)
Figure 41-2. The effect of painful stimulation (skin incision) on the electroencephalogram during levels of anes-
thesia below minimal alveolar concentration; such stimulation tends to reduce the amount of slow activity and to
accentuate the amount of fast activity seen with a given concentration of anesthetic agent. Paper speed is 15 mm/
second (half the usual speed). (From DaubeJR, Harper CM, Litchy WJ, Sharbrough FW. Intraoperative monitor-
ing. In Daly DD, Pedley TA [eds]. Current Practice of Clinical Electroencephalography, 2nd ed. Raven Press, New
York, 1990, p 742. By permission of Mayo Foundation.)
Figure 41-3. A-C, Electroencephalogram (EEG) of a 54-year-old man undergoing left carotid endarterectomy. The
EEG result demonstrates the type of reduction in the faster anesthetic components and retention of rhythmic slow-
ing that occurs with less severe decrease in cerebral blood flow (CBF) below the critical level. This change is easy
to identify despite significant artifact affecting the posterior electrodes. Paper speed is 15 mm/second (half the
usual paper speed). (From Daube JR, Harper CM, Litchy WJ, Sharbrough FW. Intraoperative monitoring. In Daly
DD, Pedley TA [eds]. Current Practice of Clinical Electroencephalography, 2nd ed. Raven Press, New York, 1990,
p 742. By permission of Mayo Foundation.)
525
Figure 41-4. A-C, Electroencephalogram (EEG) of a 74-year-old man undergoing right carotid endarterectomy.
The EEG in B demonstrates the more severe type of EEG "suppression" that occurs in association with a more se-
vere decrease in cerebral blood flow (CBF) below the critical level. Paper speed is 15 mm/second (half the usual
paper speed). (From Daube JR, Harper CM, Litchy WJ, Sharbrough FW. Intraoperative monitoring. In Daly DD,
Pedley TA [eds]. Current Practice of Clinical Electroencephalography, 2nd ed. Raven Press, New York, 1990, p 743.
By permission of Mayo Foundation.)
Figure 41-5. Electroencephalogram (EEG) of a 68-year-old man undergoing left carotid endarterectomy. The EEG
shows the type of dramatic, rapid attenuation of all components often associated with cerebral blood flow (CBF)
of less than 6 or 7 mL/100 g per minute. Paper speed is 5 mm/second (one-sixth the usual paper speed). (From
Daube JR, Harper CM, Litchy WJ, Sharbrough FW. Intraoperative monitoring. In Daly DD, Pedley TA [eds]. Cur-
rent Practice of Clinical Electroencephalography, 2nd ed. Raven Press, New York, 1990, p 744. By permission of
Mayo Foundation.)
526
Cerebral Function Monitoring 527
SYMMETRICAL higher in amplitude, and slower in fre-

ELECTROENCEPHALOGRAPHIC quency—from the upper to the lower beta
range and finally into the alpha or even theta
PATTERNS DURING range. In humans, the anesthetic ARF pa
ANESTHESIA tern tends to be more continuous than nor-
mal sleep spindles.
Although much has been written about the In addition to the alpha-frequency ARF
effects of different anesthetic agents on the anesthetic pattern, there are often anterior
EEG during anesthesia, most of these agents maximum, triangular, slow (ATS) waves,
produce similar EEG patterns when used at which commonly are diphasic and have a
concentrations below their minimal alveolar sharply contoured, initially negative phase
concentration, which is the level necessary that is followed by a more rounded positive
to prevent movement in response to a phase (Fig. 41-1). These ATS waves charac-
painful stimulus in approximately 50% of teristically have a duration of less than 1 sec-
patients. The common anesthetic agents, in- ond and may occur either as a single tran-
cluding thiopental, halothane, enflurane, sient or in brief trains. When preceded by a
isoflurane, and nitrous oxide, produce cer- waxing ARF alpha pattern, they may pro-
tain EEG changes. At subanesthetic concen- duce a "mitten-like" pattern.10 In addition to
trations, thiopental characteristically pro- the ATS waves, more posterior, arrhythmic,
duces maximal beta activity in the anterior slow (PAS) waves, which are often of lower
midline, and halothane, enflurane, isoflu- amplitude, may become prominent. Dura-
rane, and 50% nitrous oxide produce a sim- tion of these individual polymorphic slow
ilar but less prominent pattern of beta waves is usually longer than 1 second. The
activity.10 Most surgical inductions are per- maximum of this activity is often difficult to
formed relatively rapidly with thiopental and identify clearly, but it is always more poste-
are accompanied by the following charac- rior and may be more prominent over the
teristic sequence of changes: the subanes- temporal regions (Fig. 41-1). This pattern is
thetic pattern of beta activity rapidly be- least obvious with halothane, more obvious
comes widespread, increases in amplitude, with enflurane, and most prominent with
and slows in frequency toward the alpha isoflurane. Strictly, nitrous oxide alone is not
range.11 During this phase, the faster rhythm potent enough to be an anesthetic agent at
may be intermixed with bursts of high-am- atmospheric pressure; nonetheless, it po-
plitude, frontal-intermittent, rhythmic delta tentiates the effects of other inhalation
activity, the FIRDA pattern. With a slower rate agents. During "balanced anesthesia" with
of induction with nonbarbiturate inhalation 50% nitrous oxide in combination with an-
agents, there is less tendency toward inter- other agent, the PAS pattern tends to be
mittent rhythmic bursting. more prominent than when only a single
When a light level of steady-state anesthe- agent is used.
sia is reached, a characteristic anterior max-
imum, rhythmic fast (ARF) pattern, usually
in the lower beta- or alpha-frequency range, PREOPERATIVE FOCAL
is seen with all agents, including halothane, ABNORMALITIES
enflurane, isoflurane, and even thiopental
(see Fig. 41-1). The frequency of this pat- Most patients undergoing endarterectomy
tern tends to slow with increasing concen- show a symmetrical baseline pattern of the
trations of the agent. This anesthetic ARF type described above. However, depending
pattern is strikingly similar to postanoxic, on patient selection, 30%-40% of patients
widespread, anterior maximum alpha-coma may have a focal abnormality of varying
pattern. The origin of the postanoxic ARF severity.12 This focal abnormality consists of
pattern is uncertain. However, evidence a unilateral decrease in the amplitude of
suggests that the anesthetic ARF pattern, the ARF anesthetic pattern by more than
usually in the alpha-frequency range, is a 30%-40%. This commonly is associated with
drug-induced beta-variant pattern11 that increased wave length and amplitude of per-
with anesthesia becomes more widespread, sistent polymorphic slowing on the side of
the decreased amplitude. The latter gener- focal slowing. In such a case, preanesthetic
ally can be distinguished from the usual PAS beta activity is usually also symmetrical.
pattern during anesthesia because it has a The nonlocalizing FIRDA pattern or more
longer wave length and is more irregular and persistent generalized slowing is easily rec-
higher in amplitude on the pathologic side. ognized as abnormal during the waking
In most patients, focal baseline EEG abnor- state, but it cannot be identified as abnor-
mality correlates with preoperative deficits. mal during anesthesia because these pat-
However, a small percentage of patients with terns commonly occur in most patients at
such baseline abnormalities have experi- some stage of anesthesia, whether or not
enced only transient ischemic attacks, pre- they have had symptoms of a central nervous
sumably caused by a hemodynamic mecha- system lesion.
nism. These patients usually have normal
findings on computed tomography, with an
ipsilateral decrease in retinal artery pressure FOCAL
and a low baseline cerebral blood flow
thought sufficient to cause EEG abnormali- ELECTROENCEPHALOGRAPHIC
ties in the absence of residual neurologic CHANGES DURING CAROTID
signs or computed tomographic abnormali- ENDARTERECTOMY
ties7 (Fig. 41-1).
In general, a preoperative focal anesthetic Electroencephalographic changes com-
EEG abnormality correlates with a preoper- monly occur with carotid artery clamping—
ative waking EEG abnormality. Occasionally, some change occurs in approximately 25%
in exceptional patients, anesthesia activates of patients. These changes almost always oc-
an abnormality that was either not apparent cur within 20-30 seconds after clamping and
or less apparent during the waking trace.13 are associated with decreased cerebral blood
However, anesthesia may obscure an abnor- flow below a critical level, which varies with
mality that was present during the waking the anesthetic agent. With halothane, the
trace and may also activate an abnormality critical level is between 15 and 18 mL/100
that is minimal or not apparent during the g per minute; it is slightly lower with enflu-
waking state, as in the case of an anterior rane and isofiurane. The reasons for these
hemispheric insult that does not affect the differences are uncertain. Roughly the sever-
normal symmetrical alpha pattern. Despite ity and rapidity of the onset of EEG change
such a normal symmetrical alpha pattern, vary in direct proportion to the degree of
the anesthetic record may show a major de- lowering of blood flow below the critical
crease in the ARE pattern and an increase level. Minor changes consist of a 25%-50%
in the irregular slowing in the anterior dis- decrease in the faster components and an
tribution. In these patients, the drug-in- increase in amplitude and wave length of
duced beta activity seen during the pre- slower components (Fig. 41-3). Changes
anesthetic state is nearly always decreased on that occur with more severe decreases in
the side of a reduced anesthetic ARF pattern, blood flow (in the range of 6 or 7 mL/100
which is further evidence that the anesthetic g per minute or less) are associated with an
alpha-frequency ARF pattern is related more even greater reduction of anesthetic faster
directly to drug-induced beta activity than to components, along with decreased ampli-
the normal waking alpha rhythm. tude of the slower components, producing
The anesthetic EEG also tends to activate a lower amplitude and featureless EEG on
an abnormality when intermittent rhythmic the side of clamping (Figs. 41-4 and 41-5).
slow waves are present in the temporal re- Although up to 25% of EEGs may show some
gion on the side of ischemia. This intermit- change, only l%-3% show the more severe
tent abnormality is often converted into a degree of change.
more obvious and persistent focal slowing Focal transient changes that occur other
along with reduction in the ARF pattern dur- than during clamping can be seen in up to
ing anesthesia. A more posterior lesion, 10% of patients. In most, this is caused by
which produces significant abnormality in transient asymmetrical effects of changing
the alpha pattern, may leave the anesthetic levels of anesthesia on a preexisting focal ab-
ARF pattern symmetrical, without obvious normality and is of no consequence.14 Em-
bolization rarely occurs in association with of stroke is higher than when there is no
shunting (in less than 1% of our shunted pa- change (4.5% vs. 1.4%).16 Selective shunt
tients) . Some EEG changes likely are caused placement in patients with major EEG
by the reversible effects of embolization changes significantly decreases the rate of
from the operative site. However, in ap- stroke.17 The rate of intraoperative stroke
proximately 1% of patients undergoing end- when no monitoring is used is six times
arterectomy, a focal EEG change develops higher than that when EEG monitoring is
intraoperatively that is not associated with used to select patients for shunt place-
carotid artery clamping and persists ment.18
throughout the procedure; ultimately, it is Intraoperative changes in a somatosensory
associated with a new neurologic deficit in evoked potential recording are measured
the immediate postoperative period. These and quantified more easily than EEG
changes almost always prove to be caused by changes. Moreover, somatosensory evoked
embolization. An effective way of identifying potentials are less likely to be influenced by
embolization is to measure cerebral blood factors such as anesthetic effects, preopera-
flow when a new persistent focal EEG change tive cerebral abnormalities, and recording
develops. artifacts. Nonetheless, a major drawback is
that, after carotid cross-clamping, somato-
sensory evoked potential changes are not de-
ELECTROENCEPHALOGRAPHY tected as promptly as EEG changes, because
AND OTHER MONITORING it often takes a minute or longer to complete
the averaging and analysis of somatosensory
TECHNIQUES IN RELATION TO evoked potential signals,19 but nearly all
SURGICAL STRATEGY DURING EEG changes occur within 20-30 seconds af-
CAROTID ENDARTERECTOMY ter cross-clamping. A study that compared
the two techniques concluded that they are
Intraoperative monitoring is useful only if it complementary.20
can accurately and promptly forewarn sur- Intraoperative neurologic examination
geons of the occurrence of cerebral compli- has been advocated as a method of deter-
cations in time for corrective measures to be mining intolerance to carotid cross-clamp-
instituted. The greatest susceptibility to cere- ing. However, the method cannot be used in
bral ischemic injury occurs when the carotid nearly 25% of the patients who prefer or re-
artery is cross-clamped just before it is inquire general anesthesia.21 Although intra-
cised. To avoid this potential complication, operative neurologic examination has not
some surgeons have routinely placed a shunt been shown to be superior to EEG moni-
from the common carotid artery to the in- toring in decreasing stroke rate, the combi-
ternal carotid artery to bypass the site of the nation of the two techniques reportedly can
clamp. However, the rate of embolic stroke reduce the need for shunt placement.22
with routine placement of a shunt is nearly A method of monitoring the integrity of
10 times greater than that with selective use vascular flow during carotid surgery is in-
of a shunt.15 Thus, reliable and accurate tracarotid injection of xenon.5 The deter-
techniques of intraoperative monitoring are mination of cerebral blood flow with this
needed to select patients who require a technique is clearly influenced by the Paco2
shunt to decrease the risk of cerebral isch- level and requires that xenon be delivered
emic injury during carotid cross-clamping. directly to the internal carotid artery while
Intraoperative monitoring techniques can the external carotid artery is clamped. Blood
be divided into those that detect cerebral flow usually is determined only three or four
dysfunction (EEG, somatosensory evoked times intraoperatively: before clamping, im-
potentials, and neurologic examination) mediately after clamping, immediately after
and those that measure the integrity of cere- placing a shunt (if one is used), and at the
bral perfusion (carotid stump pressure, trans- end of the procedure. Blood flow measure-
cranial Doppler, and cerebral blood flow ment is complementary to EEG findings. Fo-
measurements). Recent studies have con- cal EEG changes as a result of ischemia
firmed that when a major EEG change oc- caused by decreased perfusion pressure dis-
curs during carotid cross-clamping, the risk tal to a clamped carotid artery are always as-
sociated with low blood flow, as measured because the profound hypothermia that is
with the xenon technique. An unexpected induced intraoperatively suppresses most of
EEG change that persists and is associated the EEG activities needed for monitoring.
with "normal blood flow" is essentially Diffuse EEG changes are difficult to distin-
pathognomonic of embolization. The pres- guish from global ischemia even when the
ence of normal blood flow after emboliza- hypothermia is mild or when PaGOg is high.
tion is explained on the basis of the so-called To address the limitations of EEG monitor-
look-through phenomenon.23 A simplified ing in cardiac surgery, a recent review pro-
interpretation of this phenomenon is as fol- posed that EEG monitoring be used con-
lows: if the ischemia is a result of embolic comitantly with transcranial Doppler and
occlusion of one-half the blood vessels to a near-infrared spectroscopy. Near-infrared
region (with the other one-half being spectroscopy is a noninvasive optical method
patent), injection of xenon produces a nor- for continuously measuring cerebral venous
mal or, at times, an increased flow and oxygenation. Its measurement is relatively
washout of xenon through the patent blood unaffected by anesthesia effect, altered cere-
vessels. Totally occluded vessels receive no bral metabolism, or markedly decreased per-
xenon and thus do not contribute to the fusion pressure.3
overall measurement of flow.
Carotid stump pressure determination is
a measurement of the back pressure of flow SUMMARY
at the distal carotid stump after cross-clamp-
ing. Although low carotid stump pressure Intraoperative electrophysiologic monitor-
values are more likely to be associated with ing of cerebral function during cardiovas-
clinically important EEG changes24 and cular surgery requires a thorough knowl-
higher stroke rates,16 the specificity of ca- edge of the effect of anesthetic agents on
rotid stump pressure measurement is only electrophysiologic signals. Although there
approximately 60% to 80%.16'17'24 Also, are variations among anesthetic agents and
carotid stump pressure measurement is not their effects on the EEG, most of the agents
as reliable as xenon blood flow measure- produce similar changes that can be recog-
ment.25 Recent studies that compared EEG nized and distinguished from the effects of
with carotid stump pressure monitoring sug- ischemia. Success of the monitoring also de-
gested that EEG monitoring is more accu- pends heavily on the technical aspects of
rate17 and the use of carotid stump pressure recording, such as the time scale of the EEG
measurement alone may result in a high display and adjustment of the anesthetic
rate of unnecessary shunts.24 Transcranial agent used.
Doppler insonates the temporal bone win- Currently, intraoperative EEG monitoring
dow to measure blood flow velocity at the is used mostly during carotid surgery be-
middle cerebral artery ipsilateral to the cause of its favorable sensitivity and speci-
carotid cross-clamping. A recent review of ficity in promptly detecting cerebral intoler
the method concluded that it complements ance to carotid cross-clamping. Recent
EEG monitoring, probably because the two studies have continued to demonstrate con-
techniques measure different potential ef- vincingly the usefulness of intraoperative
fects of carotid cross-clamping.26 An advan- monitoring in decreasing the risk of stroke
tage of transcranial Doppler is that it can de- in carotid endarterectomy. For the most
tect abnormalities of intracranial emboli part, the advent of non-EEG monitoring
from carotid debris and hyperperfusion pat- techniques has not replaced EEG monitor-
terns after carotid repair.2' ing. Many recent studies have shown that
these non-EEG monitoring techniques com-
plement EEG monitoring, largely because
ELECTROENCEPHALOGRAPHIC the aspects of intraoperative cerebral hy-
MONITORING DURING poperfusion or ischemia that these tests
CARDIAC SURGERY measure are different from those that EEG
measures. To date, studies that have com-
Electroencephalographic monitoring has pared the different modalities used in in-
not been used widely during cardiac surgery traoperative monitoring have lacked the
scientific rigor of randomized controlled sults of surgery and hemodynamics of cerebral isch-
studies. emia. Mayo Clin Proc 56:533-543, 1981.
15. Salvian AJ, Taylor DC, Hsiang YN, et al. Selective
shunting with EEG monitoring is safer than routine
shunting for carotid endarterectomy. Cardiovasc
Surg 5:481-485, 1997.
REFERENCES 16. McCarthy WJ, Park AE, Koushanpour E, Pearce
WH, YaoJS. Carotid endarterectomy. Lessons from
1. Meyers R, Knott JR, Hayne RA, Sweeney DB. The intraoperative monitoring—a decade of experi-
surgery of epilepsy: limitations of the concept of the ence. Ann Surg 224:297-305, 1996.
cortico-electrographic "spike" as an index of the 17. Lacroix H, Van Gertruyden G, Van Hemelrijck J,
epileptogenic focus. J Neurosurg 7:337-346, 1950. Nevelsteen A, Suy R. The value of carotid stump
2. Sharbrough FW, MessickJM Jr, Sundt TM Jr. Cor- pressure and EEG monitoring in predicting carotid
relation of continuous electroencephalograms with cross-clamping intolerance. Acta Chir Belg 96:269-
cerebral blood flow measurements during carotid 272, 1996.
endarterectomy. Stroke 4:674-683, 1973. 18. Plestis KA, Loubser P, Mizrahi EM, Kantis G, Jiang
3. Edmonds HL Jr, Rodriguez RA, Audenaert SM, ZD, Howell JF. Continuous electroencephalo-
Austin EH III, Pollock SB Jr, Ganzel BL. The role graphic monitoring and selective shunting reduces
of neuromonitoring in cardiovascular surgery. J neurologic morbidity rates in carotid endarterec-
Cardiothorac Vase Anesth 10:15-23, 1996. tomy. J Vase Surg 25:620-628, 1997.
4. Quasha AL, Sharbrough FW, Schweller TA, Tinker 19. Guerit JM, Witdoeckt C, de Tourtchaninoff M, et
JH. Hypothermia plus thiopental: synergistic EEC al. Somatosensory evoked potential monitoring in
suppression (abstract). Anesthesiology 51 (Suppl): carotid surgery. I. Relationships between qualitative
S20, 1979. SEP alterations and intraoperative events. Elec-
5. Sundt TM Jr, Sharbrough FW, Trautmann JC, troencephalogr Clin Neurophysiol 104:459-469,
Gronert GA. Monitoring techniques for carotid 1997.
endarterectomy. Clin Neurosurg 22:199-213,1975. 20. Fiori L, Parenti G. Electrophysiological monitoring
6. Chiappa KH, Burke SR, Young RR. Results of elec- for selective shunting during carotid endarterec-
troencephalographic monitoring during 367 carotid tomy. J Neurosurg Anesthesiol 7:168-173, 1995.
endarterectomies. Use of a dedicated minicom- 21. Stoughton J, Nath RL, Abbott WM. Comparison of
puter. Stroke 10:381-388, 1979. simultaneous electroencephalographic and mental
7. Yanagihara T, Klass DW. Discrepancy between CT status monitoring during carotid endarterectomy
scan and EEC in hemodynamic stroke of the carotid with regional anesthesia. J Vase Surg 28:1014—1021,
system. Trans Am Neurol Assoc 104:141-144,1979. 1998.
8. Visser GH, Wieneke GH, van Huffelen AC. Carotid 22. Fiiorani P, Sbarigia E, Speziale F, et al. General
endarterectomy monitoring: patterns of spectral anaesthesia versus cervical block and perioperative
EEG changes due to carotid artery clamping. Clin complications in carotid artery surgery. Eur J Vase
Neurophysiol 110:286-294, 1999. Endovasc Surg 13:37-42, 1997.
9. Minicucci F, Cursi M, Fornara C, et al. Computer- 23. Donley RF, Sundt TM, Anderson RE, Sharbrough
assisted EEG monitoring during carotid endarter- FW. Blood flow measurements and the "look
ectomy. J Clin Neurophysiol 17:101-107, 2000. through" artifact in focal cerebral ischemia. Stroke
10. Stockard J, Bickford R. The neurophysiology of 6:121-131, 1975.
anaesthesia. Monogr Anesthesiol 2:3-46, 1975. 24. Harada RN, Comerota AJ, Good GM, Hashemi HA,
11. Sharbrough FW. Nonspecific abnormal EEG pat- Hulihan JF. Stump pressure, electroencephalo-
terns. In Niedermeyer E, Lopes da Silva F (eds). graphic changes, and the contralateral carotid
Electroencephalography: Basic Principles, Clinical artery: another look at selective shunting. Am J Surg
Applications and Related Fields. Urban &: 170:148-153, 1995.
• Schwarzenberg, Baltimore-Munich, 1982, pp 135- 25. McKay RD, Sundt TM, Michenfelder JD, et al. In-
154. ternal carotid artery stump pressure and cerebral
12. Sundt TM Jr, Sharbrough FW, Anderson RE, blood flow during carotid endarterectomy: modifi-
Michenfelder JD. Cerebral blood flow measure- cation by halothane, enflurane, and innovar. Anes-
ments and electroencephalograms during carotid thesiology 45:390-399, 1976.
endarterectomy. J Neurosurg 41:310-320, 1974. 26. Ackerstaff R, Moll F. Use of EEG and TCD for as-
13. Hansotia PL, Sharbrough FW, Berendes J. Activa- sessment of brain function during operations on
tion of focal delta abnormality with methohexital carotid artery. In Horsch S, Ktenidis K (eds).
and other anesthetic agents (abstract). Electroen- Peri-Operative Monitoring in Carotid Surgery.
cephalogr Clin Neurophysiol 38:554, 1975. Steinkopf Verlag, Darmstadt, 1998, pp 110-120.
14. Sundt TM Jr, Sharbrough FW, Piepgras DG, Kearns 27. Fiori L, Parenti G, Marconi F. Combined transcra-
TP, Messick JM Jr, O'Fallon WM. Correlation of nial Doppler and electrophysiologic monitoring for
cerebral blood flow and electroencephalographic carotid endarterectomy. J Neurosurg Anesthesiol
changes during carotid endarterectomy: with re- 9:11-16, 1997.
Chapter 42
BRAIN STEM AND CRANIAL
NERVE MONITORING
METHODS Posterior Cranial Fossa

Electromyography Head and Neck Surgery
Nerve Conduction Studies SUMMARY
Evoked Potentials
APPLICATIONS
Middle Cranial Fossa
Cranial nerves can be injured during surgi- die electrodes. Electrodes placed in muscle
cal procedures performed in the middle and record various spontaneous and stimulus
posterior cranial fossae as well as in the head evoked activity arising from individual mus-
and neck region. Damage results from com- cle fibers or motor units. Movement and
pression, stretch, abrasion, or ischemia of electric artifacts, fibrillation and fascicula-
the nerve. If axonal disruption occurs, re- tion potentials, and random motor unit po-
covery is limited. Cranial nerve function can tential (MUP) activity related to inadequate
be monitored with the patient under anes- anesthesia are regularly recorded from mus-
thesia by recording spontaneous or stimulus- cle intraoperatively. Electric stimulation of
evoked electrical activity directly from the the innervating nerve produces a response
nerve or the cranial muscles. These methods by activation of MUPs in the area immedi-
can detect damage to either the intra-axial ately surrounding the electrode. Mechanical
or extra-axial portion of cranial nerves. Ac- irritation (abrasion, stretch, or compres-
tivity in other pathways in the brain stem can sion), saline irrigation, and ischemia pro-
be monitored by following changes in blood duce high-frequency bursts of MUPs, called
pressure, pulse, respiration, temperature, neurotonic discharges that have a characteris-
and evoked potentials in various sensory and tic sound and appearance.1 Neurotonic dis-
motor pathways. charges provide surgeons immediate feed-
back about nerve location and potential
injury to nerves in the surgical field. Neuro-
METHODS tonic discharges can be recorded in situa-
tions that require neuromuscular blockade
Electromyography by titrating the dose of the neuromuscular
blocking agent with the motor response ob-
Special small-diameter flexible wire electrodes tained with peripheral nerve stimulation
are used to record motor unit activity from intraoperatively.2
within muscles. These electrodes are less Intraoperative electromyography (EMG)
traumatic to local tissue and are more easily is performed with the same sweep speed and
secured than monopolar or concentric nee- filter settings as standard diagnostic needle
532
Brain Stem and Cranial Nerve Monitoring 533
Figure 42-1. Neurotonic dis-

charges recorded from facial
nerve-innervated muscles during
posterior fossa surgery.
EMG. Recordings are possible from almost and muscle fibers. Whenever possible,
any cranial muscle, including extraocular CMAPs recorded from the skin surface over-
and facial muscles, muscles of mastication lying the motor point are used because they
and tongue, and pharyngeal and laryngeal give more quantifiable information about
muscles. The activity from multiple muscles the total number of functioning motor ax-
is often monitored simultaneously with a ons in the nerve than CMAPs recorded from
multichannel recording instrument. As with intramuscular electrodes. The optimal sur-
standard EMG, auditory signals are very im- face electrodes are 5-mm disks that are ap-
portant in the analysis of the origin and re- plied firmly to the skin with collodion. Com-
lationship of the potentials to intraoperative pound muscle action potential recordings
events. are made with filter settings of 2 Hz-20 kHz,
Neurotonic discharges have various forms, sweep speeds of 1-10 ms/cm, and sensitivi-
but all of them consist of MUPs that fire in ties of 100 juV-5 mV/cm.
a rapid and irregular manner (Fig. 42-1). At Nerve action potentials (NAPs) are
times, multiple discharges firing asynchro- recorded directly from mixed or sensory
nously and independently are recorded nerves in the surgical field or subcuta-
from a single muscle. Neurotonic discharges neously. Although NAPs are lower in ampli-
are often precipitated by mechanical stimu- tude and more difficult to record than
lation of the axonal membrane of periph- CMAPs, they may provide useful information
eral nerves. They are sensitive indicators of when sensory nerves are involved or when
nerve irritation and occur in virtually all CMAPs cannot be recorded. The amplitude
monitored patients.3'4 Sharp transection of of the CMAP or NAP is proportional to the
a nerve may not produce neurotonic dis- number of axons conducting the response.
charges,5 and damaged nerves are less likely Therefore, when a goal of monitoring is to
to produce neurotonic discharges than determine the number of intact axons, the
healthy nerves. In addition, irrigation of a amplitude or area of the response can be
nerve with saline frequently produces long measured at any time and compared with
trains of neurotonic discharges lasting 2-60 values recorded earlier intraoperatively or
seconds. Therefore, the density and fre- with preoperative baseline measurements.
quency of neurotonic discharges recorded Several different stimulators are used to
during surgery correlates only roughly with activate peripheral nerve axons intraopera-
the severity of postoperative neurologic tively. Handheld stimulators of various sizes
deficit.6 and configurations that can be gas-sterilized
are commercially available. Stimulators that
are insulated to the very tip of the electrode
Nerve Conduction Studies have fewer problems with current shunting,
but they may also produce subthreshold
Two types of nerve conduction studies can stimuli if they are not applied properly to
be performed on cranial nerves intraopera- the surface of the nerve. Other stimulators
tively. Compound muscle action potentials have a hooked configuration that allows a
(CMAPs) represent activity in motor axons nerve or fascicles within a nerve to be sepa-
rated from surrounding tissue. This reduces

artifact from the stimulus or surrounding
muscles and allows the nerve elements of in-
terest to be stimulated selectively. Bipolar
stimulators have the cathode and anode at-
tached to the same handle and within sev-
eral centimeters of each other. This provides
a localized stimulus that reduces the risk of
current spread to adjacent nerves. The dis-
advantage of the bipolar stimulator is that
activation may be inadequate if the nerve is
distant or there is too much fluid in the sur-
gical field. Monopolar stimulators use a sin-
gle handheld cathode placed on the nerve,
with a separate anode placed some distance
away, usually a needle in the edge of the sur-
gical field or a distant surface electrode. Figure 42-2. Flash-evoked visual evoked potential
Monopolar stimulation reduces the chance recorded intraoperatively. The signal is often unreliable
when recorded with patient under general anesthesia.
of inadequate stimulation, but it increases
the likelihood of current spread to other
nerves and shock artifact.
The size of the stimulating electrodes AUDITORY EVOKED POTENTIALS
varies depending on the nerve stimulated.
Small cranial nerves require stimulator tips Auditory evoked potentials reflect electric
as small as 1 mm, and larger peripheral activity in the auditory nerve and brain stem
nerves may require 2-3 mm electrodes to in response to cochlear stimulation. Brain
provide an adequate stimulus. If a surgical stem auditory evoked potentials (BAEPs)
forceps is modified for use as a bipolar stim- are recorded from surface electrodes placed
ulator, the surgeon can dissect tissue with either in the external ear canal or on the
the stimulator. ear pinna and referenced to the vertex. Five
distinct waveforms can be recorded: waves I
Evoked Potentials and II originate from the auditory nerve and
waves III to V reflect activity in brain stem
VISUAL EVOKED POTENTIALS structures. The electrocochleogram can be
recorded from a needle electrode placed
Visual evoked potentials (VEPs) reflect elec- through the tympanic membrane into the
tric activity in optic pathways in response to wall of the middle ear cavity. The first ma-
photic stimulation of the retina. The most jor negative waveform (Nl) of the electro-
reproducible response is recorded as a broad cochleogram represents activity in the lat-
positivity over the occipital head region ap- eral portion of the auditory nerve and is
proximately 100 milliseconds after the stim- analogous to wave I of the BAEP. Auditory
ulus. This PI 00 waveform represents electri- nerve action potentials (NAPs) can also be
cal activity in the occipital cortex. It is well recorded from a small cotton-wick electrode
defined in all awake patients when a pattern- placed directly on the nerve in the cerebel-
reversal illuminated stimulus is given to the lopontine angle at surgery. Auditory evoked
retina. Visual evoked potentials can also be potentials are not affected significantly by
recorded in anesthetized patients with a general anesthetics.
strobe flash stimulus applied through the Brain stem auditory evoked potentials ap-
closed eyelid or by way of specially designed pear to be the best technique for most sur-
contact lenses. However, flash evoked VEPs gical procedures, because it has fewer tech-
are greatly attenuated by general anesthesia nical problems, can be used during the
and, thus, have limited use intraoperatively7 entire surgical procedure, monitors the en-
(Fig. 42-2). tire auditory system, and correlates well with
postoperative hearing status.8"10 The elec- tumors in the region of the pituitary and hy-
trocochleogram is better defined in patients pothalamus and during operations on vas-
with acoustic neuromas, but technical diffi- cular lesions that may affect the optic nerves,
culties are common and only the lateral por- chiasm, or tracts. Because of the variability
tion of the auditory nerve located in the in latency and amplitude of the response
auditory canal of the temporal bone is mon- and the poor correlation with postoperative
itored. Recording of auditory NAPs provides visual function, visual evoked responses are
immediate feedback, but technical problems not a reliable monitor of the function of the
are common and the technique can be used visual pathway during surgery.7
only when the auditory nerve is exposed in
the surgical field. If they can be reliably
recorded, they show better preservation of CAVERNOUS SINUS REGION
hearing.10 Sudden changes in the latency Tumors and vascular lesions of the orbital,
and amplitude of BAEPs occur when the sphenoidal, or cavernous sinus regions can
nerve is avulsed or the internal auditory damage cranial nerves directly or distort
artery is damaged.11 Recovery after sudden normal anatomical relationships, making it
loss of BAEPs is unusual, but correlation of difficult to distinguish between normal and
the change with intraoperative events may abnormal nervous system structures. Types
help prevent hearing loss in future cases. of cases that may benefit from monitoring
Gradual changes in BAEPs may reflect ex- include tumors such as meningiomas, lym-
cessive traction on or manipulation of the phomas, carcinomas, pituitary adenomas,
nerve and often resolve with adjustment of and vascular lesions such as carotid or
retractors or modification of the surgical ophthalmic aneurysms. The oculomotor,
approach. trochlear, and abducens nerves can be
monitored with EMG or CMAPs.13 Wire
SOMATOSENSORY EVOKED electrodes are placed in the extraocular
POTENTIALS AND MOTOR muscles after the patient is anesthetized.
EVOKED POTENTIALS Neurotonic discharges are recorded in the
appropriate muscle when its nerve is me-
Median, ulnar, and tibial somatosensory chanically stimulated in the surgical field.
evoked potentials (SEPs) can be used to The surgeon may also use a handheld stim-
monitor activity in the medial lemniscus of ulator to identify selected cranial nerves by
the brain stem. Changes may occur only recording a CMAP in the appropriate tar-
when sensory pathways are involved or get muscle14 (Fig. 42-3). A NAP can be
when damage to the brain stem is extensive. recorded directly from the ophthalmic di-
Monitoring of motor evoked potentials vision of the trigeminal nerve with a small
(MEPs) may enhance the sensitivity of brain cotton-wick electrode.15
stem monitoring by detecting early com-
promise of pyramidal tract neurons. Use of
MEPs to monitor brain stem function is still
investigational and is limited by the ex-
quisite sensitivity of these potentials to in-
halation anesthetics.12
APPLICATIONS
Middle Cranial Fossa
PITUITARY REGION Figure 42-3. Compound muscle action potentials from
extraocular muscles obtained by direct electric stimu-
Visual evoked potentials have been used to lation of the oculomotor nerve in the surgical field in
monitor the visual system during removal of the region of the cavernous sinus.
Posterior Cranial Fossa talis muscles correlates with the number of

functioning axons in the nerve16 (Fig.
CEREBELLOPONTINE ANGLE 42-4). Preservation of the facial CMAP at
the end of the operation predicts good re-
The trigeminal, facial, auditory, and vestibu- covery of facial nerve function within 1 year
lar nerves can be injured during the removal postoperatively.6'17 Brain stem auditory
of posterior fossa tumors and during mi- evoked potentials can be monitored simul-
crovascular decompression or neurectomy taneously with EMG and CMAPs.9 Changes
for trigeminal neuralgia, hemifacial spasm, in BAEPs correlate well with the postopera-
or vertigo. Also, the brain stem may be in- tive level of hearing.9'11'17'18 Gradual
volved by mass lesions larger than 3 cm in changes are often reversible by altering the
diameter. The facial nerve and the motor surgical approach or by moving retrac-
division of the trigeminal nerve are monitors.19'20 Sudden loss of BAEPs is usually ir-
tored with intramuscular electrodes placed reversible and represents either ischemia or
in muscles of facial expression and mastica- avulsion of the auditory nerve11 (Fig. 42-5).
tion, respectively. Mechanical stimulation Changes that correlate with postoperative
produces neurotonic discharges in the re- function help determine the mechanism of
spective muscles. Electric stimulation can be nerve injury, thereby improving future sur-
used to identify the various nerves in the gical results. When brain stem compression
cerebellopontine angle. The amplitude of is present, monitoring SEPs or MEPs may
CMAPs recorded over the nasalis or men- also be useful.
Figure 42-4. Monitoring of electromyographic (EMG) potentials and facial compound muscle action potentials
(CMAP) intraoperatively for acoustic neuroma in a 55-year-old woman. Examples of neurotonic discharges observed
at various times intraoperatively and gradual loss of facial CMAPs indicate iatrogenic injury of the facial nerve.
(From Daube JR, Harper CM. Surgical monitoring of cranial and peripheral nerves. In Desmedt JE [ed]. Neu-
romonitoring in Surgery. Elsevier Science Publishers, Amsterdam, 1989, p 118. By permission of the publisher.)
Figure 42-5. Monitoring of brain stem auditory evoked potentials during operation for acoustic neuroma. The sud-
den loss of the response correlated with inadvertent coagulation of the internal auditory artery. (From Harper CM,
Daube JR. Surgical monitoring with evoked potentials: the Mayo Clinic experience. In Desmedt JE [ed]. Neu-
romonitoring in Surgery. Elsevier Science Publishers, Amsterdam, 1989, p 281. By permission of the publisher.)
JUGULAR FORAMEN REGION damage to branches of the facial nerve. Elec-

tromyographic monitoring has been shown
Glomus tumors, meningiomas, metastatic to decrease the risk of facial nerve injury dur-
cancer, and other lesions in the region of ing parotidectomy.23'24
the jugular foramen may involve the facial, The recurrent branch of the laryngeal
auditory, glossopharyngeal, vagal, and spinal nerve can be injured during carotid endar-
accessory nerves. Electromyographic activity terectomy,25 anterior cervical diskectomy,26
is monitored by placing intramuscular wire or resection of thyroid tumors.27 Wire elec-
electrodes in facial, laryngeal, and trapezius trodes can be placed by direct laryngoscopy
muscles. Electric stimulation can be used to into the vocalis muscle for monitoring the
distinguish between rootlets of the glos- recurrent laryngeal nerve. Monitoring has
sopharyngeal and vagus nerves in patients been shown to decrease the frequency of
undergoing neurectomy for glossopharyn- vocal cord paralysis associated with these
geal neuralgia.21'22 Electromyographic mon- procedures.24
itoring of the hypoglossal nerve (with elec-
trodes placed in the tongue), in addition to
monitoring of the vagus and spinal accessory SUMMARY
nerves, is useful during operations to remove
chordomas, meningiomas, and other lesions Various modalities are available for moni-
in the region of the clivus and foramen toring the function of the cranial nerves and
magnum. brain stem during intracranial or extracra-
nial head and neck operations. After con-
sideration of the surgical risks, a multi-
Head and Neck Surgery modality approach can be tailored to the
needs of each patient. Intraoperative moni-
Operations for neoplasms of the parotid toring has been shown to decrease the inci-
gland may injure one or more branches of dence of cranial nerve injury during poste-
the facial nerve that course through the rior fossa surgery.
gland. Each branch can be monitored se-
lectively by placing wire electrodes in the
frontalis, orbicularis oculi, orbicularis oris, REFERENCES
and mentalis muscles. Mechanical irritation
of the nerve produces neurotonic discharges 1. Harner SG, Daube JR, Ebersold MJ. Electrophysio-
in the target muscle. Electric stimulation in logic monitoring of facial nerve during temporal
the surgical field can locate and prevent bone surgery. Laryngoscope 96:65-69, 1986.
2. Blair EA, Teeple E Jr, Sutherland RM, Shih T, Chen and application in the surgical management of fa-
D. Effect of neuromuscular blockade on facial cial pain. Neurosurgery 38:76-81, 1996.
nerve monitoring. AmJ Otol 15:161-167, 1994. 16. Goldbrunner RH, Schlake HP, Milewski C, Tonn
3. Romstock J, Strauss C, Fahlbusch R. Continuous JC, Helms J, Roosen K. Quantitative parameters
electromyography monitoring of motor cranial of intraoperative electromyography predict facial
nerves during cerebellopontine angle surgery. J nerve outcomes for vestibular schwannoma surgery.
Neurosurg 93:586-593, 2000. Neurosurgery 46:1140-1146, 2000.
4. Kombos T, Suess O, Kern BC, Funk T, Pietila T, 17. Moller AR, Jannetta PJ. Preservation of facial func-
Brock M. Can continuous intraoperative facial elec- tion during removal of acoustic neuromas. Use of
tromyography predict facial nerve function follow- monopolar constant-voltage stimulation and EMG.
ing cerebellopontine angle surgery? Neurol Med J Neurosurg 61:757-760, 1984.
Chir (Tokyo) 40:501-505, 2000. 18. Wazen JJ. Intraoperative monitoring of auditory
5. Nelson KR, Vasconez HC. Nerve transection with- function: experimental observations and new ap-
out neurotonic discharges during intraoperative plications. Laryngoscope 104:446-455, 1994.
electromyographic monitoring. Muscle Nerve 18: 19. Hatayama T, Moller AR. Correlation between la-
236-238, 1995. tency and amplitude of peak V in the brain stem
6. Harner SG, Daube JR, Ebersold MJ, Beatty CW. Im- auditory evoked potentials: intraoperative record-
proved preservation of facial nerve function with ings in microvascular decompression operations.
use of electrical monitoring during removal of Acta Neurochir 140:681-687, 1998.
acoustic neuromas. Mayo Clin Proc 62:92-102, 20. Matthies C, Samii M. Management of vestibular
1987. schwannomas (acoustic neuromas): the value of
7. Raudzens PA. Intraoperative monitoring of evoked neurophysiology for intraoperative monitoring of
potentials. Ann N YAcad Sci 388:308-326, 1982. auditory function in 200 cases. Neurosurgery
8. Radtke RA, Erwin CW, Wilkins R, Rozear M. Intra- 40:459-466, 1997.
operative brain stem auditory evoked potentials 21. TahaJM, TewJM Jr, Keith RW, Payner TD. Intra-
(BAEPs): significant decrease in postoperative au- operative monitoring of the vagus nerve during in-
ditory deficit (abstract). Neurology 37 (Suppl 1): tracranial glossopharyngeal and upper vagal rhizot-
219, 1987. omy: technical note. Neurosurgery 35:775-777,
9. Harper CM, Harner SG, Slavit DH, et al. Effect of 1994.
BAEP monitoring on hearing preservation during 22. TahaJM, TewJM Jr. Long-term results of surgical
acoustic neuroma resection. Neurology 42:1551- treatment of idiopathic neuralgias of the glos-
1553, 1992. sopharyngeal and vagal nerves. Neurosurgery
10. Jackson LE, Roberson JBJr. Acoustic neuroma sur- 36:926-930, 1995.
gery: use of cochlear nerve action potential moni- 23. Terrell JE, Kileny PR, Yian C, et al. Clinical outcome
toring for hearing preservation. AmJ Otol 21:249- of continuous facial nerve monitoring during pri-
259, 2000. mary parotidectomy. Arch Otolaryngol Head Neck
11. Neu M, Strauss C, Romstock J, Bischoff B, Surg 123:1081-1087, 1997.
Fahlbusch R. The prognostic value of intraopera- 24. Brennan J, Moore EJ, Shuler KJ. Prospective analy-
tive BAEP patterns in acoustic neurinoma surgery. sis of the efficacy of continuous intraoperative nerve
Clin Neurophysiol 110:1935-1941, 1999. monitoring during thyroidectomy, parathroidec-
12. Zentner J. Noninvasive motor evoked potential tomy, and parotidectomy. Otolaryngol Head Neck
monitoring during neurosurgical operations on the Surg 124:537-543, 2001.
spinal cord. Neurosurgery 24:709-712, 1989. 25. Markand ON, Dilley RS, Moorthy SS, Warren C Jr.
13. Schlake HP, Goldbrunner R, Milewski C, et al. Monitoring of somatosensory evoked responses
Technical developments in intra-operative moni- during carotid endarterectomy. Arch Neurol 41:
toring for the preservation of cranial motor nerves 375-378, 1984.
and hearing in skull base surgery. Neurol Res 26. Jellish WS, Jensen RL, Anderson DE, SheaJF. In-
21:11-24, 1999. traoperative electromyographic assessment of re-
14. Sekiya T, Hatayama T, Shimamura N, Suzuki S. In- current laryngeal nerve stress and pharyngeal in-
traoperative electrophysiological monitoring of jury during anterior cervical spine surgery with
oculomotor nuclei and their intramedullary tracts Caspar instrumentation. J Neurosurg 91 (Suppl 2):
during midbrain tumor surgery. Neurosurgery 170-174, 1999.
47:1170-1176, 2000. 27. Timon CI, Rafferty M. Nerve monitoring in thyroid
15. Stechison MT, Moller A, Lovely TJ. Intraoperative surgery: Is it worthwhile? Clin Otolaryngol 24:487-
mapping of the trigeminal nerve root: technique 490, 1999.
Chapter 43
SPINAL CORD MONITORING
Jasper R. Daube
APPLICATIONS TYPES OF SURGERY

MONITORING METHODS Primary Spine Disease
Somatosensory Evoked Potentials Primary Neural Disease
Motor Evoked Potentials PRIMARY VASCULAR DISEASE
ELECTROMYOGRAPHIC AND NERVE SUMMARY
CONDUCTION STUDIES
APPLICATIONS might not be because of the risk of an ad-

verse outcome.
The neural structures and their associated The primary goal of intraoperative moni-
risks during spine surgery include the fol- toring is to prevent new neurologic deficits
lowing: by identifying impairment sufficiently early
• Spinal cord—ischemia, slow compres- to allow prompt correction of the cause. This
sion, stretching, and direct trauma is accomplished best by demonstrating nor-
• Nerve roots and spinal nerve—stretch- mal function early in a procedure and re-
ing, blunt trauma, pinching, and isch- peatedly testing function to search for
emia changes that signal impending damage. In-
• Cauda equina—stretching, blunt trauma, traoperative testing can assist the surgeon in
pinching, and ischemia identifying neural structures in the region of
Damage may occur at one or more levels of a tumor. Continuous monitoring of neural
the spine; thus optimally, each level at risk function demonstrates physiologic changes
should be monitored. The modalities and caused by alterations in temperature, anes-
variables of monitoring change with the thetic, or hypotension that typically return
level. The level most commonly monitored to baseline after the alteration has been re-
is the thoracic level. versed. For example, irritation of neural
Monitoring is of benefit in many surgical tissue or mild local compression produces
procedures on the spine.1"3 Immediately changes that can be reversed by stopping the
postoperatively, persistent neurologic deficit offending activity. Their early recognition al-
develops in fewer than 0.5% of patients who lows the surgeon to modify the procedures
undergo corrective operations for scoliosis to reduce the likelihood of a persistent
or other surgical procedures on the spine, deficit. Neural destruction is readily recog-
but this deficit can be devastating. Of the nized but not reversible. Thus, monitoring
complications that occur, one-half are com- provides immediate evidence not only of the
plete paraplegia and one-half are incom- damage but of the severity and location of
plete paraplegia, with one-third of patients the damage.
having no recovery of function.4 With surgi- Two publications provide specific recom-
cal monitoring, some patients can be con- mendations for ensuring quality and safety
sidered surgical candidates who otherwise during surgical monitoring.2'5 With some
539
modifications, the equipment used for sur- with the anesthesiologist about the anes-
gical monitoring is the same as that used in thetic options.
outpatient testing. Electromyographic (EMG) In each case, the methods of surgical mon-
equipment should allow both audio and vi- itoring are selected on the basis of the struc-
sual presentations and should have auto- tures at risk, the optimal methods for mon-
matic artifact rejection to minimize opera- itoring their function, and the anticipated
tive interference, particularly that caused by anesthetic regimen. Because the spinal cord,
cautery. Electric safety is critically important. nerve roots, spinal nerves, and cauda equina
Monitoring equipment must not be able to may all need to be monitored, somatosen-
cause a current greater than 100 îA to pass sory evoked potentials (SEPs), motor evoked
through the patient if equipment grounding potentials (MEPs), nerve conduction stud-
fails. The most common danger is from im- ies, EMG, and reflex testing may each need
proper and malfunctioning grounding. Cur- to be applied. Some surgical procedures put
rent neurophysiology equipment should more than one anatomical structure at risk,
have optical isolation of each patient contact and all the structures warrant monitoring.
to prevent the inadvertent conduction of Such monitoring requires simultaneous use
electric currents between the patient and the of multiple modalities.
equipment. The neurophysiologist needs to
take responsibility for ensuring that the leak-
age current of the equipment has been MONITORING METHODS
tested and is safe.
The wide variety of surgical procedures on Several monitoring modalities are available.
the spine (bony spine disease, spinal cord They are reviewed briefly in Table 43-1, but
disease, spinal nerve and root disorders), the each one is discussed more fully in textbooks
many underlying diseases that require sur- and reviews of surgical monitoring.7"11
gical treatment, the many levels of deficit
that the patient may have, and the varying
anesthetic needs all require flexibility on the Somatosensory Evoked Potentials
part of the monitoring team.6 It is not pos-
sible to have a single, standard protocol for As clearly shown in animal models, isch-
monitoring. The unique problems of each emia or compression of the spinal cord pro-
patient require that the monitoring be de- longs, reduces, and then obliterates SEPs in
signed individually after review of the clini- proportion to the amount of damage.
cal problem, discussion with the surgeon Somatosensory evoked potentials have be-
about the structures at risk, and discussion come an accepted method of monitoring
Table 43-1. Modalities for Spinal Cord Monitoring*

Modality Stimulation Recording
SEPs Peripheral nerve, dermatome, Peripheral nerve, plexus, cauda
cauda equina, or spinal cord equina, directly from spinal
cord (lumbar, thoracic, or
cervical) or cerebral cortex
MEPs Cerebral cortex or spinal cord Spinal cord, peripheral nerve,
(magnetic or electric) or muscle
Nerve conduction studies Root Peripheral nerve or muscle
EMG Mechanical or electric Muscle
Reflex testing Peripheral nerve, root, or Root, peripheral nerve, or muscle
pudendal nerve
*Electrophysiologic modifications include (1) rates and pattern of stimulation, (2) type or placement of record-
ing electrodes, and (3) facilitation of MEPs and SEPs.
EMG, electromyography; MEPs, motor evoked potentials; SEPs, somatosensory evoked potentials.
Spinal Cord Monitoring 541
spinal cord function during a wide variety • Esophageal or nasopharyngeal elec-

of surgical procedures on the spine or cord. trodes at the cervical levels; either of
A large multicenter study of scoliosis sur- these can provide a good stable record-
gery showed that the incidence of postop- ing anterior to the spinal cord.
erative neurologic deficits was 0.46% with A readily performed, technically reliable
SEP monitoring and 1.04% without SEP approach to SEP monitoring of spinal cord
monitoring.12 function during spine surgery using tibial
SEPs recorded from the sciatic nerve, cervi-
STIMULATING AND RECORDING cal spinal cord, and scalp is shown in Figure
ELECTRODES 43-1. Several methods of recording in the
surgical field have been developed so that
The most commonly used monitoring mo- recordings can be made closer to the neural
dality is SEPs, because of the ease of appli- tissue. Needle electrodes can be placed in
cation, limited susceptibility to anesthesia, the intervertebral ligaments or spinal lam-
and reasonable sensitivity to detecting dam- ina. Several insulated wires are available that
age. Somatosensory evoked potentials may can be placed in the epidural space to record
be used to monitor the spinal cord, spinal particularly high-amplitude well-defined re-
nerves, or nerve roots, but they are much less sponses from lumbar and thoracic spinal lev-
effective in monitoring the latter two. Stim- els above or below the level of stimulation.
ulation typically is simultaneous for the right The combination of epidural stimulation
and left sides of the body, using serial aver- and epidural recordings often produces su-
aging techniques from a single location on perior SEP amplitudes with well-defined
one trace with a 100-ms interval between the peaks. Direct stimulation of the spinal cord
stimuli on the two sides. The nerves most rein the surgical field activates ascending and
liable for monitoring are the median or ul- descending potentials that may be in motor
nar nerve at the wrist and the posterior tib- or sensory tracts.
ial nerve at the ankle. Recording electrodes
should be placed proximal and distal to the TECHNICAL FACTORS
site of possible damage to determine
whether a loss of response is a peripheral The anesthetic sensitivity of spinal cord-
technical problem or is related to the surgi- evoked potentials is different from that of
cal procedure. The electrodes may be ap- scalp recordings. For example, with nitrous
plied in various ways: oxide, the amplitude of scalp-recorded po-
• Surface electrodes on the limb, lumbar tentials is decreased approximately 50%;
spine, cervical spine, and scalp. Larger, spinal cord-evoked potentials, in compari-
more reliable tibial SEPs may be ob- son, are relatively unaffected by most anes-
tained with C3-C4 recordings than with thetic agents and drugs. A combination of
the standard Cz-Fz recordings, and spinal and scalp recordings provides the ad-
they should always be tested to be sure vantages of both types of monitoring. The
that the optimal potential is selected. percutaneous electrodes on the lamina allow
• Needle electrodes adjacent to the pe- continued recording of spinal potentials
ripheral nerve or lamina; 30-75 mm when reproducible scalp potentials cannot
percutaneous needle electrodes placed be obtained. Occasionally, because of a pa-
directly on the lamina outside the sur- tient's preoperative neurologic deficit, scalp
gical field can record well-defined, re- potentials cannot be recorded. In many of
producible potentials (or in the in- these patients, SEPs may be recorded at the
traspinous ligament at any spinal level neck.
in the surgical field). Both the level of anesthesia and blood
• Epidural recordings can be made at any pressure change the latency and amplitude
level in the operating field with multi- of SEPs, especially if the mean blood pres-
lead electrodes, cable electrodes, strip sure is less than 70 mm Hg. Rarely, the scalp
electrodes, fine wires, or needles in the response is enhanced after the induction of
intraspinous ligament at any spinal anesthesia; generally, it is decreased. In a
level. small proportion of cases, the response is lost
Figure 43-1. Standard placement of electrodes for monitoring somatosensory evoked potentials during thora-
columbar spine surgery. The sciatic response is recorded with a needle electrode near the nerve. Cervical responses
are recorded from an electrode on the lamina of the spine of C7 and from an esophageal electrode. Scalp responses
are recorded from standard vertex electrodes. (From Daube JR., Harper CM, Litchy WJ, Sharbrough FW. Intraop-
erative monitoring. In Daly DD, Pedley TA [eds]. Current Practice of Clinical Electroencephalography, 2nd ed.
Raven Press, New York, 1990, p 747. By permission of Lippincott, Williams & Wilkins.)
immediately after the induction of anesthe- Several technical factors must be consid-
sia. This occurs more frequently in children ered with intraoperative recordings. The
and adolescents than in adults. Similarly, the rate of stimulation cannot be as fast as when
amplitude gradually decreases and the la- testing an awake patient. Somatosensory
tency increases the longer the period of evoked potentials can be recorded with
anesthesia. Somatosensory evoked potential stimulation rates of 5 Hz or even 10 Hz in
changes caused by anesthesia are much less most awake patients, but when a patient is
prominent at the neck than on the scalp. anesthetized, the scalp SEPs fatigue at rates
The response varies depending on the anes- greater than 3 Hz. Stimulation rates as low
thetic agent. Of the common induction as 0.5 or 1.0 Hz may be needed, especially
agents, propofol has the least effect. Nitrous with deeper levels of anesthesia. The num-
oxide has some effect, but enflurane and ber of stimuli that are averaged varies with
isoflurane reduce SEPs in more than half of the magnitude of the response and the level
the patients in whom monitoring is per- of background noise. Because many of the
formed. patients are paralyzed, muscle activity is min-
imal. With ideal conditions, clear responses can be distinguished from background vari-
can be obtained with only 64 or 128 stimuli. ability.14 Ideally, the ordinary background
However, in most cases, 500 stimuli are of- variability of the potentials should be no
ten needed to obtain reproducible traces, more than 30% in amplitude and 1.0 ms in
because there are sources of artifact other latency. Infrequent and unusual nonsurgical
than muscle. causes of rapid changes in SEPs must always
A frequent problem during surgical mon- be considered before concluding that the
itoring is SEP variability on sequential operation is the cause. These causes may be
recordings. Somatosensory evoked poten- any of the physiologic changes in tempera-
tials may change for many reasons other ture, blood pressure, or subdural air.
than surgical damage to the sensory path- Accuracy of monitoring can also be im-
way. During cervical surgery with the patient proved with appropriate use of peripheral
in the sitting position, a marked reduction recordings, including monitoring arm SEPs
in SEPs can occur because of the accumula- over Erb's point with median or ulnar nerve
tion of subdural air. This change is recog- stimulation and monitoring leg SEPs from
nized readily by comparing the standard the sciatic nerve at the gluteal fold or the
vertex electrode recording with that of elec- N22 lumbar potential from T12-L1. Moni-
trodes placed just above the ear, where there toring median nerve responses during tho-
is less subdural air. To differentiate changes racic spinal procedures can help to distin-
caused by technical factors from those guish changes caused by anesthesia from
caused by pathway damage requires that the those caused by the surgical procedure.
alteration in the amplitude or latency be Preoperative testing with SEPs, EMG, and
consistent at both the neck and scalp record- nerve conduction studies often assists with
ing sites and the peripheral response be in- determining the optimal combination of
tact. The change must be greater than the recordings for each patient. When SEPs have
baseline variation documented during the a very low amplitude or are absent, bilateral
initial period of the operation. Although a peripheral nerve stimulation, stimulation of
50% decrease in amplitude from baseline the sciatic nerves, or stimulation of the
and a 5% change in latency generally indi- cauda equina may be needed to elicit the po-
cate likely damage,13 no absolute change in tentials. Recordings may have to be made
amplitude can be considered evidence of from the epidural space rather than from an
spinal cord damage. In some patients with extraspinal location. If EMG and nerve con-
no damage, the scalp response appears to be duction studies show abnormalities in the
lost transiently whereas other responses are distribution of particular nerve roots, these
intact. In patients with damage, smaller, con- roots will be more susceptible to further
sistent alterations at two sites of stimulation damage and warrant intramuscular record-
provide evidence of compression before ma- ings to monitor neurotonic discharges.
jor changes appear. Somatosensory evoked potential monitor-
Unnecessary alarms, caused by random ing during spine surgery has proved valuable
variability of the signals, during somatosen- in warning surgeons of potential damage to
sory monitoring have been a concern, but the spinal cord. Because such damage is un-
they generally are infrequent.13 Such false common, few large studies have defined the
alarms, or false positives, can be reduced change in SEPs in relation to spinal cord
substantially by paying attention to tech- damage. In surgical monitoring of spinal
nique, for example, with the use of a record sensory pathways, approximately 20%
stricted filter band pass, with the low filter of patients have significant changes in SEPs
increased to 30 Hz and high filter to 2000 intraoperatively. Up to half of these changes
Hz. Recording variables should be chosen may be reversed surgically (removal of he-
for each patient to maximize the stability of matoma, rods, spine wires, etc.). Patients
the recording, but after these variables have without any change in SEPs may have post-
been selected, they need to be kept constant operative deficits consistent with anterior
throughout the operation. The variability of spinal artery syndrome, but this is rare. If
the recorded potential should be assessed SEP changes occur, the amplitude reduction
early in an operation so that adverse events usually begins at the neck and scalp 10-30
minutes after localized spinal cord damage, trains of stimuli are to be given), and mag-
with an increase in latency of up to 3 ms. So- netic MEPs are more sensitive than electric
matosensory evoked potentials can return to MEPs to anesthetic agents.15 Motor evoked
baseline value within 5—10 minutes after cor- potentials are obtained easily and reliably
rection. The abrupt loss of SEPs is less likely with direct stimulation of the spinal cord or
to be reversible. In a few patients, SEP am- cerebral hemispheres of patients given a
plitude shows improvement intraoperatively. combination of nitrous oxide and narcotic
In summary, several patterns of change in anesthesia. Halogenated anesthetics mark-
SEPs have been observed and correlated edly decrease MEP amplitude through ef-
with postoperative neurologic function. The fects on both cortical neurons and anterior
change may be late or gradual, emphasizing horn cells. Some medical centers have be-
the importance of continuing the monitor- gun to use transcranial MEP stimulation for
ing until the patient is awake. Gradual surgical monitoring, but this is still consid-
changes in SEPs may be caused also by isch- ered experimental in the United States.
emia of the spinal cord or peripheral nerves, Transcranial stimuli are best delivered at a
as during operations on the thoracic or ab- low rate, fewer than 1 per 3 seconds. The op-
dominal aorta. Less frequently, SEPs change timal stimulus is a capacitively coupled stim-
abruptly, usually in relation to contusion of ulus of modest intensity, 250 V-450 V, with
the spinal cord or to compression by an a time constant of 100 microseconds. Usu-
epidural or subdural hematoma. When the ally, 8-10 sweeps are averaged to improve
loss of SEPs is abrupt, the site of injury reproducibility.
should be localized by recording with epi- The spinal cord can be stimulated with an
dural or direct spinal electrodes, followed by epidural electrode, a needle in or between
careful inspection of the involved level for spinous processes, or with a nasopharyngeal
hematoma or other potentially reversible active electrode and a laminar reference
causes of spinal cord injury. However, if the electrode. Interspinous or epidural elec-
loss of SEPs is abrupt and not reversible, trodes in the surgical field use a distant an-
paraplegia is likely. Improvement in the am- ode in the subcutaneous tissue and are
plitude of SEPs intraoperatively usually is as- technically more difficult to use than the
sociated with improved neurologic status combination of laminar and esophageal
postoperatively. Postoperative motor deficits electrodes. Typically, MEPs are recorded
without associated changes in SEPs intra- with partial neuromuscular block to reduce
operatively are infrequent but well docu- body movement. Monitoring with spinal
mented. The motor deficits may occur as cord-evoked MEPs has been applied suc-
part of the anterior spinal artery syndrome cessfully in thoracic and lumbar surgery but
or they may be caused by direct injury to the not in cervical spine surgery. Cervical spine
ventral spinal cord, ventral horn cells, or surgery requires cranial stimulation to evoke
nerve roots. MEPs.
RECORDING ELECTRODES
Motor Evoked Potentials Motor evoked potentials can be recorded di-
STIMULATING ELECTRODES
rectly from the spinal cord as a cord-evoked
potential, from a peripheral nerve as a nerve
Stimulation of the motor cortex or spinal action potential, or from muscle as a com-
cord can evoke MEPs for monitoring the in- pound muscle action potential (CMAP).
tegrity of corticospinal motor pathways. Stimulation of the spinal cord activates both
Studies with both electric and magnetic stim- motor and sensory fibers, even at threshold.
ulation have been reported. In an anesthe- Therefore, the MEPs recorded from the
tized patient, magnetic stimulation has no spinal cord in response to direct spinal cord
advantages over electric stimulation and has stimulation (cord-to-cord) are a mixture of po-
major disadvantages: the coil is cumbersome tentials in ascending and descending path-
and hard to immobilize relative to the skull, ways, and this mixture may be less sensitive
the apparatus is expensive (particularly if to spinal insult than monitoring of selective
pathways. Recordings performed in the sur- ulation may reflect activity primarily of pos-
gical field are most useful for operations on terior column axons, because spinal motor
the spinal cord (for example, tumors or ar- neuron activity is reduced. With the neuro-
teriovenous malformations). The directly muscular block that typically is used, a sig-
recorded potentials can localize the area of nificant component of the neurogenic po-
damage or record responses that are too tential may be end plate potential from
small to be obtained with other methods. surrounding muscle as well as potentials
from the motor and sensory fibers in the pe-
SPINAL CORD RECORDINGS ripheral nerve.
Epidural bipolar cardiac pacing electrodes MUSCLE RECORDINGS

with an interelectrode separation of 2-3 cm
are suitable for recording descending MEPs Compound muscle action potentials are re-
(the same electrode can be used to record corded with surface electrodes in response
SEPs directly from the spinal cord or to to stimulation of either the cerebral cortex
stimulate the spinal cord). The major ad- or the spinal cord, but the potentials are dif-
vantages and disadvantages of transcranial ferent for the two modes of stimulation. Nar-
MEP monitoring from the spinal cord in- cotic anesthesia is necessary for CMAPs to be
clude the following: elicited, and even then the CMAPs produced
• It is relatively immune to the effects of by single stimuli to the motor cortex are too
anaesthetic agents. variable for monitoring motor function reli-
• Full muscle relaxation is possible and, ably. Pairs or trains of stimuli, especially to
indeed, desirable. the spinal cord, produce enough temporal
• Somatosensory evoked potentials can summation of excitatory input to activate
be recorded reliably in the same sweeps most motor neurons. Large, stable CMAPs
if the cerebral cortex and peripheral can be recorded, but this often requires par-
nerve are stimulated simultaneously. tial neuromuscular block to control move-
• Often, MEPs can be recorded in pa- ment (Fig. 43-2). The level of neuromuscu-
tients with a preexisting neural deficit. lar block is monitored best with recording
• Abnormality is identified promptly. of CMAPs in response to peripheral nerve
• It is feasible only when epidural leads stimulation along with the CMAPs from cen-
can be inserted, but this usually requires tral stimulation.
a posterior approach to the spinal cord. With spinal cord stimulation, optimal
• It does not identify the side responsible CMAPs are obtained with paired stimuli at
for any deterioration in the recorded intervals of 3-5 ms. With cortical stimu-
volleys. lation, trains of 5 or 6 stimuli of high inten-
• Motor evoked potentials are not as reli- sity are needed at intervals of 2-5 ms.
ably recorded from the lumbar cord Compound muscle action potentials are
When a preexisting abnormality or the sur- recorded best from multiple muscles in both
geon's preference prevents MEPs and SEPs legs. A major advantage of the technique is
from being recorded with epidural elec- that monitoring can be adapted by choosing
trodes, monitoring can be performed by us- muscles to suit the specific clinical need, for
ing other potentials. example, muscles innervated by specific
nerve roots when the operation is low spinal
PERIPHERAL NERVE RECORDINGS or segmental or when a nerve root is known
to be at risk. The signal-to-noise ratio for
Neurogenic potentials can be recorded in the CMAPs is sufficient for single trials to be
region of the sciatic or tibial nerve in re- recorded without averaging. Although the
sponse to stimulation of the cerebral cortex reproducibility of evoked CMAPs is good, it
or the spinal cord, but they are less well de- may not be as high as that of MEPs in
fined with cortical stimulation. One hun- epidural recordings.
dred or more responses need to be averaged. It is critical for the neurophysiology team
If inhalation anesthesia is used, the neuro- to be in constant contact with the anesthe-
genic MEPs obtained with spinal cord stim- sia team (even though they may be at op-
to change is alteration of the level of neu-

romuscular block.
The major advantages and disadvantages
of MEP monitoring with CMAPs include the
following:
* Unilateral dysfunction can be identi-
fied.
* Evoked potentials with spinal cord stim-
ulation are resistant to anesthesia.
* Compound muscle action potentials
evoked with spinal stimulation vary with
the level of neuromuscular block.
" Compound muscle action potentials
evoked with spinal stimulation can be
recorded simultaneously with SEPs.
* Cortically evoked potentials can be
adapted for virtually all spinal and cere-
bral operations.
* Compound muscle action potential
recording is equally useful for opera-
tions on the low spinal cord, cauda
equina, or nerve root.
* Spinal stimulation cannot be used with
cervical spine surgery.
* There is no intrusion into the operative
field.
* Cortically evoked CMAPs intrinsically
are more variable and more sensitive to
anesthesia than spinal evoked CMAPs.
Many neurophysiologists have found that di-
rect spinal cord stimulation and recording
give highly reproducible motor responses
when recorded in leg muscles (Fig. 43-2).
ELECTROMYOGRAPHIC AND
NERVE CONDUCTION STUDIES
Damage to cervical or lumbosacral nerve
Figure 43-2. Motor evoked potentials (MEPs) elicited
by stimulation of the spinal cord with paired stimuli ap-
roots or motor neurons during a surgical
plied between a laminar needle and esophageal elec- procedure on the spine can be minimized
trodes. A, Stimulating electrode location. B, MEPs with a combination of EMG recordings and
recorded as surface compound muscle action potentials nerve conduction studies.16 Anterior horn
from leg muscles bilaterally, with partial neuromuscu- cells can be damaged during dissection of
lar block. Ant. rib., anterior tibialis; reel, fem., rectus
femoris. intraspinal tumors or by ischemia caused by
compression or traction. Radiculopathies
are an occasional complication of scoliosis
surgery; Aey likely are caused by local com-
posite ends of the patient) to prevent loss of pression or traction of a root.17 Neurotonic
the evoked response because of anesthesia. discharges in limb muscles innervated by the
The most common reason for deterioration affected motor neurons or axons can warn
of cortically evoked CMAPs is the adminis- of potential damage caused by manipula-
tration of a supplemental narcotic agent. tion, traction, or ischemia of nerve roots
The most common reason for spinal CMAPs (Fig. 43-3B).
Figure 43-3. Three examples of monitoring with electromyography (EMG) and nerve conduction studies during
lumbosacral surgery. A, Compound muscle action potentials evoked in the anal sphincter by direct stimulation of
tissue in the surgical field identified the tissue as axons in the L2-L4 nerve roots (lipomeningocele resection). B,
Neurotonic discharges in the anal sphincter (SPH) during lipomeningocele dissection warned the surgeon of irri-
tation of the L2-L4 axons. C, Motor unit potential firing during lumbar fusion warned the surgeon of irritation of
dorsal root axons. D, Compound muscle action potentials evoked in L5-innervated muscles by a stimulating elec-
trode in a pedicle screw hole with less than 20 mA current warned the surgeon that the pedicle screw was close
enough to the dorsal root to irritate it or damage it.
Neurotonic discharges are irregular bursts of muscle response to stimulation can help dis-
motor unit potentials recorded in muscles in tinguish among nerve roots and differenti-
response to axonal irritation.18 They can be ate the roots from non-neural structures.
recorded with surface, subcutaneous, or in- The studies record CMAPs produced by lo-
tramuscular electrodes. Fine-wire electrodes cal stimulation of individual nerves (Fig.
inserted in each of the muscles at risk are 43-3A). Stimulation may be applied with
the most convenient and specific recording monopolar, bipolar, or forceps electrodes.
electrodes. Recordings can be made from Recordings can be made with surface, sub-
any somatic muscle; for example, monitor- cutaneous, or intramuscular electrodes.
ing of L3-S3 muscles, including the anal
sphincter, is helpful in operations for my-
elomeningocele. Also, intramuscular record- TYPES OF SURGERY
ings of motor unit potential firing caused
by a reflex response can detect irritation of Electrophysiologic monitoring can be bene-
the sensory axons in the dorsal root (Fig. ficial in many surgical procedures in infants,
43-3C). children, and adults by reducing the extent
Direct stimulation of nerves in the surgi- and duration of damage. The surgeon should
cal field can provide information about the decide whether monitoring is needed be-
location and integrity of the nerves.19 If the cause he or she can best judge the risk of
normal anatomy is distorted, recording the neural damage and the structures at risk.
The anesthesiologist selects the optimal anesthesia, presumably an idiosyncratic re-

anesthesia, and, after discussion with the sur- action to the anesthetic agent. In these cases,
geon and anesthesiologist, the clinical neu- a cervical cord recording from a nasopha-
rophysiologist selects the optimal monitor- ryngeal, esophageal, or laminar needle elec-
ing methods. The spine surgery procedures trode is necessary to monitor spinal cord
commonly monitored include scoliosis, ky- function.
phoscoliosis, cervical spondylosis and steno- Surgical risk varies according to the amount
sis, lumbar spondylosis and stenosis, spine of spinal cord deformity, the severity of
trauma, rheumatoid arthritis, spine tumor, preoperative deficit, the size and type of le-
and herniated disk. sion to be excised, bony stability, previous
The largest group of patients who have operations, and other medical disorders.
monitoring is teenagers undergoing correc- The surgical risk may be low enough that in-
tive surgery for scoliosis. Another large traoperative monitoring of neural function
group is elderly persons who are operated is not necessary. Also, monitoring may not
on for cervical or lumbar spondylosis, often provide any benefit because the deficit is
with associated spinal stenosis or foramina complete and cannot be made worse or the
stenosis. Monitoring is also performed for structure has to be sacrificed to complete the
procedures for bony spine tumors, thoracic operation.
aneurysms, traumatic spinal damage, and For each level of spine surgery, the risk to
spondylitis. each neural structure should be assessed sep-
arately. If the risk is primarily to nerve roots
rather than the spinal cord, spinal cord mon-
Primary Spine Disease itoring may not be needed. For example, the
small risk to upper extremity nerves and
The optimal methods of monitoring differ plexus during thoracic and lumbar spine
from patient to patient depending on the surgery may be identified with monitoring.
age of the patient, preoperative deficit, type The risk to nerve roots during spine surgery
of surgery, spinal level of surgery, anesthetic increases when hardware fixation is used in
agents used, and other individual patient procedures for spondylolisthesis, stenosis,
factors regardless of the spinal level of the scoliosis, or instability. Fixation for spine fu-
operation. The most important factors to sion can be achieved with several different
consider are the age of the patient, the sur- surgical approaches. At the cervical level, ad-
gical risk, and the spinal level of surgery. jacent vertebrae commonly are wired to-
Three major factors must be considered gether to obtain the stabilization needed for
when monitoring is performed in patients bone healing and fusion. The risk to neural
younger than 21 years; each of these factors structures can occur at any time during the
presents a unique challenge to the clinical procedure, but the risk is particularly high
neurophysiologist. First, infants and small during fixation (Fig. 43-4).
children usually require different stimulat- Also, the hardware is often fixed with a
ing and recording electrodes and great care metal screw. There is a risk of root damage,
in electrode placement. Second, scalp SEP with pain or radiculopathy, if the screw im-
averaging is more difficult in children, es- pinges on the root. Thresholds to stimula-
pecially younger ones, because the ampli- tion can determine the likelihood of dam-
tude of slow-wave activity is much higher age.20 Thresholds less than 20 Vindicate that
when the child is anesthetized. Because of the nerve root is close enough to the pedi-
this, slower rates of stimulation, a larger cle screw to be damaged (Fig. 43-3Z)). Mo-
number of averaged stimuli, or a lower level tor unit potential firing may occur if a screw
of anesthesia (or a combination of these) is is too close to the dorsal root or if the root
often required. In all children, recordings is otherwise irritated.
from the cervical cord are ideal to demon- Although the major purpose of SEP mon-
strate that the spinal cord is intact, even if itoring is to help recognize subclinical
the scalp response is not clearly recogniz- changes that could herald new postoperative
able. Third, in some children and adoles- neural deficit, SEPs occasionally show im-
cents, the scalp response is lost early during provement when the procedure reduces
Figure 43-4. Gradual loss of somatosensory evoked potentials (SEPs) during stabilization procedure for cervical
spine fracture in a 60-year-old man. Responses were lost within a few minutes after wiring C5-C7, but diey returned
quickly after the wires were removed. The patient awoke with no deficit. B, bilateral; L, left; R, right. (From Daube
JR, Harper CM, Litchy WJ, Sharbrough FW. Intraoperative monitoring. In Daly DD, Pedley TA [eds]. Current Prac-
tice of Clinical Electroencephalography, 2nd ed. Raven Press, New York, 1990, p 765. By permission of Lippincott,
spinal cord compression or ischemia (or C5-T1 level if there is significant preopera-
both). This is often seen with cervical stabi- tive radiculopathy or if the roots or spinal
lization in patients with cervical rheumatoid nerves may be compromised during the op-
arthritis (Fig. 43-5). eration (Fig. 43-6). A prospective study has
shown that monitoring during cervical spine
CERVICAL SPINE DISEASE surgery is beneficial when specific criteria of
change are used.21
Monitoring during cervical spine surgery re-
quires initially determining which neural THORACIC SPINE DISEASE
structures are at risk and the level of risk.
The major concern for most patients having Somatosensory evoked potential monitoring
an operation at the C1-C4 level is myelopa- is sufficient for most patients undergoing
thy. Preoperative neurologic deficit, a mul- thoracic spine surgery.2 However, because
tilevel operation, upper cervical surgery, and motor function may be lost despite intact
instrumentation increase the risk. Monitor- SEPs, a combination of MEP and SEP mon-
ing of tibial and median SEPs is sufficient itoring has been recommended as the "stan-
unless the spinal cord is already compro- dard of care" for scoliosis surgery.23 In some
mised, in which case MEP monitoring may academic medical centers, MEP monitoring
also be needed. If the cervical spine is un- can easily be applied and is used routinely.24
stable, monitoring should begin before anes- However, anesthetic and technical consider-
thesia, because positioning the head after ations make MEP monitoring more difficult
the patient has been anesthetized may com- to apply in many settings. In these cases,
promise the spinal cord. Monitoring of MEP monitoring should be considered if
nerve roots with EMG should be considered there is marked deformity, preoperative
for patients having spine surgery at the deficit, anterior vertebrectomy, or other
Figure 43-5. Gradual improvement of somatosensory evoked potential amplitude and latency during spinal cord
decompression for rheumatoid arthritis. The patient's neurologic deficit improved postoperatively. (From Daube
JR, Harper CM, Litchy WJ, Sharbrough FW. Intraoperative monitoring. In Daly DD, Pedley TA [eds]. Current Prac-
tice of Clinical Electroencephalography, 2nd ed. Raven Press, New York, 1990, p 766. By permission of Lippincott,
Figure 43-6. Simultaneous recording of somatosensory evoked potentials (SEPs) from the scalp and arm muscle
electromyography (EMG) during an operation on a cervical cord tumor with syrinx. A, Magnetic resonance image
of the syrinx (oval). B, Tibial SEPs were absent from the onset of recording. The left median SEP was lost at 1:00.
The only postoperative deficit was loss of proprioceptive sensation in the left arm. C, Neurotonic discharges warned
the surgeon of irritation of the local root or anterior horn cells. Bil., bilateral.
550
evidence of considerable risk to the spinal little opportunity to modify the procedure,
cord.25 The possible loss of sensory function other than to stop it because of electro-
alone is sufficient reason to monitor SEPs physiologic changes. Consequently, moni-
even if MEP monitoring is available.26 toring is not often performed during these
procedures, but if it is, simultaneous MEP
LUMBOSACRAL SPINE DISEASE and SEP monitoring maximizes the possi-
bility of recognizing significant change.28
Often with surgery at the lumbosacral level, Some groups have found that monitoring
spinal nerves are at greater risk than the
during intramedullary cord surgery de-
spinal cord itself. Because SEPs are less ef-
creases the frequency of major complica-
fective than EMG in identifying nerve root
tions.29"31 Multimodality monitoring can in-
damage, EMG monitoring is often an im-
clude EMG recordings (Fig. 43-7).
portant part of monitoring during proce-
dures at the lumbosacral level.27
DORSAL RHIZOTOMY
Primary Neural Disease In patients with spasticity, especially children
with cerebral palsy, improvement in the
Operations on tumors of the spinal cord, spasticity and function reportedly has been
particularly intramedullary tumors, are high- obtained by cutting a proportion of the
risk procedures, and the surgeon usually has fibers in the L2-S1 dorsal roots, but the re-
Figure 43-7. Loss of left and right anterior tibial motor evoked potentials (MEPs), with electric stimulation of the
cerebral cortex (transcranial [Tc]-MEP), immediately after the aorta was clamped at level Thl2 during thoracoab-
dominal aneurysm surgery. Responses recovered after reperfusion through reattached lumbar arteries. LSA, lum-
bar segmental artery. (From Jacobs et al.37 By permission of Thieme Medical Publishers.)
ports are anecdotal.32 Although this proce- other tissue that will be dissected or sec-
dure had been known for many years, only tioned. Continuous EMG monitoring from
recently has Staudt et al.33 reported electro- multiple limb and sphincter muscles is the
physiologic monitoring in these patients. most effective method for doing this by iden-
Two to five fascicles are dissected apart in tifying mechanical irritation of neural tissue
each dorsal root and stimulated with single during dissection and immediately warning
stimuli and trains of stimuli to elicit a reflex. the surgeon of possible damage. Direct
The character of the responses bilaterally in stimulation of unidentified tissue elements
L2- to S2-innervated limb muscles is assessed will quickly identify it as neural tissue if a
to determine the contribution of each fasci- CMAP is evoked in a limb or sphincter mus-
cle to the spasticity. It is critical to identify cle. Although no controlled study has tested
the motor root to avoid increased weakness the value of monitoring, surgeons who have
from sectioning motor axons. The motor had the opportunity to use it agree that it
and sensory roots are distinguished readily provides valuable feedback that reduces the
by measuring the threshold to single stimuli. amount of neural damage and allows the op-
Despite many reports on the benefits of eration to proceed more expeditiously.
dorsal rhizotomy, no controlled study has
demonstrated that electrophysiologic testing
provides a better outcome than blindly sec- PRIMARY VASCULAR DISEASE
tioning 30%-50% of the roots that innervate
spastic muscles. In addition to its well known use during cere-
bral aneurysm, carotid artery, and other
cerebral vascular operations, electrophysio-
CAUDAEQUINAAND
logic monitoring is used during two spinal
TETHERED CORD
cord procedures: thoracoabdominal aortic
Surgical procedures below the spine of LI aneurysm and vascular malformation opera-
pose a risk to the cauda equina rather than tions. Direct surgical ablation of arteriove-
to the spinal cord. The overlap of multiple nous malformations can be monitored with
roots innervating individual dermatomes SEPs or MEPs if the surgeon deems the risk
makes SEP monitoring insensitive to nerve sufficient. Monitoring with SEPs or MEPs
root damage in the cauda equina. There- during temporary occlusion of the major
fore, monitoring cauda equina function re- feeder vessels of an arteriovenous malfor-
lies heavily on a combination of EMG and mation can indicate the risk of ablating the
nerve conduction studies. The presence and vessel by embolization. The difficulty of per-
distribution of evoked responses in limb and forming MEP monitoring has precluded its
anal sphincter muscles with direct stimula- use during embolization of an arteriovenous
tion of tissue in the area of the cauda equina malformation.
allow neural tissue and specific roots to be Monitoring is important in thoracoab-
identified. Continuous EMG monitoring of dominal aortic aneurysm surgery because
the same muscles can warn the surgeon of the risk of paraplegia is as high as 15%. To
impending damage. For example, neuro- decrease this rate, the surgical procedure
tonic discharges indicate that a ventral root has been modified, including spinal cord
is mechanically irritated and motor unit po- cooling, cerebrospinal fluid drainage, pre-
tential firing indicates that a dorsal root is medication, cross-clamping at short dis-
irritated.19 The most difficult operations to tances to minimize the segment of spinal
perform in the region of the cauda equina cord exposed to ischemia, femoral bypass,
are those for congenital abnormalities. Sev- and measurement of spinal cord blood flow.
eral congenital abnormalities of the lum- For each of these, the functional measures
bosacral cord and cauda equina can result of SEPs and MEPs are invaluable in identi-
in progressive neurologic deficit referred to fying significant ischemia. Analysis of the
as tethered cord syndrome?* changes in both SEPs and MEPs requires un-
The primary purpose of electrophysio- derstanding and distinguishing the effects of
logic monitoring is to preserve neural tissue the ischemia that can occur in these patients
by identifying it and distinguishing it from at the cortical (blood pressure or carotid oc-
elusion), peripheral nerve (femoral artery 5. Burke D, Nuwer MR, Daube J, et al. Intraoperative
clamping), entire spinal cord (aorta clamp- monitoring. The International Federation of Clin-
ical Neurophysiology. Electroencephalogr Clin Neu-
ing) , and segmental spinal cord (segmental rophysiol Suppl 52:133-148, 1999.
spinal artery occlusion) levels.35 Initial stud- 6. Nuwer JM, Nuwer MR. Neurophysiologic surgical
ies of the benefit of revascularization using monitoring staffing patterns in the USA. Elec-
SEPs as a guide36 were less successful than troencephalogr Clin Neurophysiol 103:616-620,
1997.
more recent studies using MEPs.37 The rapid 7. Andrews RJ (ed). Intraoperative Neuroprotection.
alterations that can occur with occlusion and Williams & Wilkins, Baltimore, 1996.
revascularization are shown in Figure 43-7. 8. Kalkman CJ, Been HD, Ongerboer de Visser BW.
Motor evoked potentials recorded from pe- Intraoperative monitoring of spinal cord function.
ripheral muscles are a better indicator of A review. Acta Orthop Scand 64:114-123, 1993.
9. Loftus CM, Traynelis VC (eds). Intraoperative Mon-
dangerous ischemia, because the motor neu- itoring Techniques in Neurosurgery. McGraw-Hill
rons in the anterior horn are more sensitive Health Professions Division, New York, 1994.
to ischemia than the motor pathways in the 10. Russell GB, Rodichok LD (eds). Primer of Intra-
spinal cord.1 operative Neurophysiologic Monitoring. Butter-
worth-Heinemann, Boston, 1995.
11. Stalberg E, Sharma HS, Olsson Y (eds). Spinal Cord
Monitoring: Basic Principles, Regeneration, Patho-
physiology, and Clinical Aspects. Springer-Verlag,
SUMMARY Wien/NewYork, 1998.
12. Nuwer MR, Dawson EG, Carlson LG, Kanim LE,
Continuous electrophysiologic monitoring Sherman JE. Somatosensory evoked potential
of spinal cord or spinal nerve (or both) func- spinal cord monitoring reduces neurologic deficits
after scoliosis surgery: results of a large multicenter
tion intraoperatively can minimize potential survey. Electroencephalogr Clin Neurophysiol 96:
damage that may occur during spine sur- 6-11, 1995.
gery. Somatosensory evoked potentials are 13. Nuwer MR. Spinal cord monitoring. Muscle Nerve
easiest to use for monitoring function and 22:1620-1630, 1999.
14. Papastefanou SL, Henderson LM, Smith NJ, Hamil-
have had the widest application. Unless spinal ton A, Webb JK. Surface electrode somatosensory-
cord injury is caused by a vascular insult, evoked potentials in spinal surgery: implications for
with purely motor damage, SEP monitoring indications and practice. Spine 25:2467-2472,
can identify the damage early enough to 2000.
alert the surgeon. Motor evoked potential 15. Burke D, Hicks RG. Surgical monitoring of motor
pathways. J Clin Neurophysiol 15:194-205, 1998.
monitoring is useful as well, but technical 16. Holland NR. Intraoperative electromyography dur-
problems in some settings can limit its ap- ing thoracolumbar spinal surgery. Spine 23:1915-
plication. Neurotonic discharges recorded 1922, 1998.
from peripheral muscle are sensitive to 17. Harper CM Jr, Daube JR, Litchy WJ, Klassen RA.
Lumbar radiculopathy after spinal fusion for sco-
nerve root irritation and, thus, can help sur- liosis. Muscle Nerve 11:386-391, 1988.
geons recognize when and where damage 18. Obi T, Mochizuki M, Isobe K, Mizoguchi K, Takatsu
may be occurring. M, Nishimura Y. Mechanically elicited nerve root
discharge: mechanical irritation and waveform.
Acta Neurol Scand 100:185-188, 1999.
19. Kothbauer K, Schmid UD, Seller RW, Eisner W. In-
traoperative motor and sensory monitoring of the
REFERENCES cauda equina. Neurosurgery 34:702-707, 1994.
20. Moed BR, Ahmad BK, Craig JG, Jacobson GP, An-
1. Daube JR. Intraoperative monitoring reduces com- ders MJ. Intraoperative monitoring with stimulus-
plications and is therefore useful. Muscle Nerve evoked electromyography during placement of il-
22:1151-1153, 1999. iosacral screws. An initial clinical study. J Bone Joint
2. Guerit JM. Neuromonitoring in the operating Surg Am 80:537-546, 1998.
room: Why, when, and how to monitor? Electro- 21. Dennis GC, Gehkordi O, Millis RM, Cole AN,
encephalogr Clin Neurophysiol 106:1-21, 1998. Brown DS, Paul OA. Monitoring of median nerve
3. Herdmann J, Deletis V, Edmonds HL Jr, Morota N. Somatosensory evoked potentials during cervical
Spinal cord and nerve root monitoring in spine spinal cord decompression. J Clin Neurophysiol
surgery and related procedures. Spine 21:879-885, 13:51-59, 1996.
1996. 22. Nuwer MR. Spinal cord monitoring with somato-
4. MacEwen GD, Bunnell WP, Sriram K. Acute neu- sensory techniques. J Clin Neurophysiol 15:183-
rological complications in the treatment of scolio- 193, 1998.
sis. A report of the Scoliosis Research Society. J 23. Padberg AM, Wilson-Holden TJ, Lenke LG, Brid-
Bone Joint Surg Am 57:404-408, 1975. well KH. Somatosensory- and motor-evoked poten-
tial monitoring without a wake-up test during idio- clinical overview. In Stalberg E, Sharma HS, Olsson
pathic scoliosis surgery. An accepted standard of Y. Spinal Cord Monitoring: Basic Principles, Re-
care. Spine 23:1392-1400, 1998. generation, Pathophysiology, and Clinical Aspects.
24. Burke D, Hicks R, Stephen J, Woodforth I, Craw- Springer-Verlag, Wien/New York, 1998, pp 509-
ford M. Assessment of corticospinal and so- 520.
matosensory conduction simultaneously during 32. Gul SM, Steinbok P, McLeod K. Long-term out-
scoliosis surgery. Electroencephalogr Clin Neuro- come after selective posterior rhizotomy in children
physiol 85:388-396, 1992. with spastic cerebral palsy. Pediatr Neurosurg
25. Deutsch H, Arginteanu M, Manhart K, et al. So- 31:84-95, 1999.
matosensory evoked potential monitoring in ante- 33. Staudt LA, Nuwer MR, Peacock WJ. Intraoperative
rior thoracic vertebrectomy. J Neurosurg 92 (Suppl): monitoring during selective posterior rhizotomy:
155-161, 2000. technique and patient outcome. Electroen-
26. Lorenzini NA, Schneider JH. Temporary loss of in- cephalogr Clin Neurophysiol 97:296-309, 1995.
traoperative motor-evoked potential and perma- 34. McQuillan PM, Newberg N. Intraoperative elec-
nent loss of somatosensory-evoked potentials asso- tromyography. In Russell GB, Rodichok LD (eds).
ciated with a postoperative sensory deficit. J Primer of Intraoperative Neurophysiologic Moni-
Neurosurg Anesthesiol 8:142-147, 1996. toring. Butterworth-Heinemann, Boston, 1995, pp
27. Weiss DS. Spinal cord and nerve root monitoring 171-187.
during surgical treatment of lumbar stenosis. Clin 35. Sueda T, Okada K, Watari M, Orihashi K, Shikata
Orthop 384:82-100, 2001. H, Matsuura Y. Evaluation of motor- and sensory-
28. Nagle KJ, Emerson RG, Adams DC, et al. Intraop- evoked potentials for spinal cord monitoring dur-
erative monitoring of motor evoked potentials: a re- ing thoracoabdominal aortic aneurysm surgery. Jpn
view of 116 cases. Neurology 47:999-1004, 1996. J Thorac Cardiovasc Surg 48:60-65, 2000.
29. Jones SJ, Harrison R, Koh KF, Mendoza N, 36. Galla JD, Ergin MA, Lansman SL, et al. Use of so-
Crockard HA. Motor evoked potential monitoring matosensory evoked potentials for thoracic and tho-
during spinal surgery: responses of distal limb mus- racoabdominal aortic resections. Ann Thorac Surg
cles to transcranial cortical stimulation with pulse 67:1947-1952, 1999.
trains. Electroencephalogr Clin Neurophysiol 37. Jacobs MJHM, de Haan P, Meylaerts SA, de Mol BA,
100:375-383, 1996. Kalkman CJ. Benefits of monitoring motor-evoked
30. Kothbauer K, Deletis V, Epstein FJ. Intraoperative potentials during thoracoabdominal aortic aneu-
spinal cord monitoring for intramedullary surgery: rysm repair: technique of choice to assess spinal
an essential adjunct. Pediatr Neurosurg 26:247- cord ischemia? In Gloviczki P, Goldstone J (eds).
254, 1997. Perspectives in Vascular Surgery, Vol. 12. Thieme
31. Tamaki T. Intraoperative spinal cord monitoring— Medical Publishers, New York, 2000, pp 1-16.
Chapter 44
PERIPHERAL NERVOUS SYSTEM
MONITORING
METHODS Repair of Traumatic Peripheral Nerve Injury

Nerve Conduction Studies Prevention of Injury During Peripheral
Electromyography Nerve Surgery
Somatosensory Evoked Potentials SUMMARY
APPLICATIONS
Entrapment Neuropathies
Electrophysiologic studies are performed bipolar electric stimulator placed directly on

during surgery for nerve trauma, entrap- the nerve within the surgical field. The size
ment neuropathies, and primary or metasta- of the stimulator is matched to the size of
tic neoplasms that involve peripheral nerves. nerve. Larger stimulators similar to those
These studies provide information about the used in routine nerve conduction studies are
number, location, type, and severity of nerve used to stimulate large nerve trunks that are
lesions.1 This information can be used to an- isolated from other nerves. Hooked stimu-
swer important questions not resolved with lating electrodes can be used to elevate
preoperative electrodiagnostic studies and to nerves from surrounding tissue when better
help surgeons make important therapeutic stimulus isolation is required. Small elec-
decisions about decompression, neurolysis, trodes are used when individual or small
or grafting of nerves. With minor modifica- groups of fascicles are stimulated. Intraop-
tions, standard techniques of electrodiagno- erative stimulation requires careful consid-
sis such as nerve conduction studies, elec- eration of the location and strength of the
tromyography (EMG), and somatosensory stimulus. Normal nerve is activated with as
evoked potentials (SEPs) are used to moni- little current as 1-5 mA for 0.5 ms.2 Over-
tor the peripheral nervous system during stimulation can produce nonspecific stimu-
surgery. Appropriate monitoring protocols lation of surrounding nerves or stimulus
can be designed for each patient after the spread along the course of the nerve. The
findings of the preoperative neurologic ex- latter leads to inaccurate calculation of la-
amination, nerve conduction studies, EMG, tency and conduction velocity. Diseased
and surgical goals are reviewed. nerve typically has a higher threshold for
stimulation.3 Thus, stimulus threshold
should be monitored closely during intra-
METHODS operative nerve conduction studies. Orien-
tation of the stimulator is also important.
Nerve Conduction Studies Bipolar stimulation produces a more focal
distribution of current than monopolar stim-
Mixed motor and sensory or sensory cuta- ulation and thus is preferred for intraoper-
neous nerves are stimulated with a handheld ative stimulation of most peripheral nerves.
555
The cathode and anode should be aligned Somatosensory Evoked Potentials

with the long axis of the nerve, and the cath-
ode should be proximal to the desired di- Somatosensory evoked potentials are re-
rection of the current. corded from surface electrodes over the cer-
Compound muscle action potentials vical spine and contralateral parietal scalp
(CMAPs) are recorded with surface or sub- following direct electric stimulation of pe-
cutaneous needle electrodes placed over dis- ripheral nerve elements in the surgical field.
tal muscles. Compound nerve action poten- The presence of a response indicates conti-
tials (NAPs) are recorded from large mixed nuity of sensory axons between the spinal
nerves with a handheld bipolar recorder cord and the site of peripheral nerve stimu-
placed directly on the nerve or from small lation. This technique has several advantages
cutaneous nerves with needle electrodes over standard SEP recordings made in the
placed next to the nerve. Compound mus- laboratory, including increased selectivity of
cle action potential recordings are larger stimulation and enhanced sensitivity caused
and less contaminated with artifact, and they by amplification of the response by the cen-
monitor exclusively the function of motor tral nervous system.
axons. Nerve action potential recordings are
technically more difficult to perform but
more sensitive than CMAPs in localizing ab-
normalities along nerves.4 Nerve action po- APPLICATIONS
tentials also are able to detect early axonal
regeneration through an area of nerve in- Nerve conduction studies, EMG, and SEPs
jury long before regeneration is reflected in are used alone or in combination during
CMAP recordings.5 The presence of CMAPs operations for various entrapment neu-
or NAPs indicates that some axons are in ropathies,7'8 repair of traumatic nerve in-
continuity; the amplitude and area of the re- juries,5'10"12 and resection of tumors that
sponse are proportional to the number of affect peripheral nerves.13 The most proxi-
functioning axons. Focal slowing of conduc- mal segments of the peripheral nervous sys
tion velocity, conduction block, or increased tem (roots and spinal nerves) are monitored
threshold of stimulation localizes a lesion during spinal surgery. Distal elements (that
along the nerve with 1-2 cm of accuracy.6"8 is, brachial and lumbar plexuses and indi-
vidual nerves of the extremities) are studied
to improve localization of peripheral nerve
Electromyography lesions, to determine the status of axonal re-
generation, and to protect nerve fascicles
Intraoperatively, EMG is recorded best with from iatrogenic injury.
small intramuscular wire electrodes. The Preoperative nerve conduction studies
wires are introduced percutaneously with a and needle EMG are used to localize and to
hollow needle, which is then withdrawn, characterize most lesions affecting the pe-
leaving the wire in place. When intramus- ripheral nervous system.7 When chronic le-
cular electrodes are used, electric stimula- sions produce segmental demyelination, me-
tion of the nerve produces a polyphasic EMG chanical distortion of paranodal myelin, or
response that, although difficult to quantify, impaired function of ion channels at the
is less likely than surface electrodes to record node of Ranvier, conduction block and fo-
nonspecific activity from adjacent muscles. cal slowing of conduction velocity are ob-
Mechanical irritation of the nerve produces served on routine nerve conduction studies.
a high-frequency discharge of motor unit po- Two important criteria are required to dem-
tentials (MUPs) (neurotonic discharge) that onstrate conduction block or focal slowing
can be distinguished easily from artifact and of conduction velocity. First, stimulation
other MUP activity in EMG recordings.6 The must be performed proximal and distal to
occurrence of neurotonic discharges is used the lesion (preferably in short 1-2 cm seg-
to locate nerves within the surgical field and ments through the area of the lesion). Sec-
to warn of the potential for nerve injury ond, the fascicles that contain the lesion
should the irritation continue.9 should be stimulated or recorded in isola-
Peripheral Nervous System Monitoring 557
tion from fascicles belonging to other nerves NAP recording. Placement proximal to the
in proximity. These criteria cannot always be lesion eliminates movement artifact and
fulfilled when a lesion affects proximal or CMAPs caused by contraction of adjacent
deep nerves. In this setting, performing muscles. The stimulator is moved from dis-
CMAPs or NAPs (or both) over short seg- tal to proximal in successive 1-2 cm seg-
ments of exposed nerve at surgery is help- ments. The lesion is identified by a change
ful.8'14 in amplitude and slowed conduction of the
When partial or complete axon loss is the NAP over the short nerve segment. Nerve ac
major pathologic substrate, localization with tion potentials may provide useful informa-
preoperative nerve conduction studies and tion before regeneration has reached the
EMG is much less precise. In this setting, in- muscle, a time when CMAPs are absent.
traoperative nerve conduction studies can
often localize the main site of a peripheral
nerve lesion and identify functioning nerve Entrapment Neuropathies
fascicles that are preserved or regenerating
across the lesion. In the absence of conduc- Localization by clinical signs and preopera-
tion block, there is no change in CMAP am- tive nerve conduction studies and EMG is ad-
plitude between proximal and distal stimu- equate in most cases of entrapment neu-
lation sites. In contrast, a segmental change ropathy.1 Carpal tunnel release and most
in NAP amplitude can frequendy be dem- cases of ulnar transposition are not moni-
onstrated even in purely axonal lesions (Fig. tored. However, complicated cases of ulnar
44-1). To show this, a bipolar hooked nerve or median neuropathy or radial, femoral, sci-
electrode is placed proximal to the lesion for atic, tibial, or peroneal neuropathy are often
Figure 44-1. Intraoperative recording of compound muscle action potentials and nerve action potentials (NAP)
during ulnar nerve exploration and stimulation at 1-cm intervals. The "0" point indicates the location of the me-
dial epicondyle. The greatest chance in latency and amplitude occurred over a 3-cm segment spanning the origin
of the cubital tunnel. FCU, flexor carpi ulnaris; Hypo, hypothenar. (From Daube JR, Harper CM. Surgical moni-
toring of cranial and peripheral nerves. In Desmedt JE [ed]. Neuromonitoring in Surgery. Elsevier Science Pub-
lishers, Amsterdam, 1989, p 133. By permission of the publisher.)
monitored because of inherent difficulties in of the two heads of the flexor carpi ulnaris
defining the number and location of lesions may be treated with a cubital tunnel release
with preoperative studies.8'11'12 Ulnar neu- or medial epicondylectomy, whereas a trans-
ropathy at the elbow is one of the most position may be performed if the lesion is
common mononeuropathies. The ulnar more diffuse or localized at or proximal to
nerve can be injured by repeated trauma to the medial epicondyle. Lesions at unusual
the nerve in the region of the medial epi- sites, such as the distal aspect of the cubital
condyle, compression by bony or soft tissue tunnel, can be localized and explored,
deformities around the elbow joint, recur- thereby avoiding unnecessary decompres-
rent subluxation over the medial epicon- sion or transposition at more proximal sites.
dyle, or entrapment between the heads of The results of nerve conduction studies per-
the flexor carpi ulnaris (cubital tunnel syn- formed during ulnar nerve exploration are
drome).15 The ulnar nerve also may be com- illustrated in Figure 44-1. Preoperative
pressed in the mid-forearm or at the wrist. nerve conduction studies demonstrated lo-
Traditionally, ulnar transposition has been calized slowing with increased CMAP dis-
the most popular surgical procedure for ul- persion approximately 3 cm distal to the
nar neuropathy at the elbow. However, be- medial epicondyle. Compound muscle ac-
cause transposition is not always successful tion potentials were recorded intraopera-
and may increase morbidity,16 more conser- tively over the abductor digiti minimi and
vative procedures such as simple cubital tun- flexor carpi ulnaris muscles, and NAPs were
nel release5 or medial epicondylectomy17 recorded from the proximal ulnar nerve.
have been advocated. Changes were found in the amplitude and
The exact site of entrapment can often be latency of both CMAPs and NAPs over a
determined with preoperative electrodiag- 3-cm segment at the origin of the cubital tun-
nostic studies. Useful findings include a lo- nel. Because there was no area of slowing
calized area of conduction block or slowing proximal or distal to this point, a cubital tun-
on short segmental stimulation during mo- nel release was performed.
tor nerve conduction studies. The distribu-
tion of abnormalities on needle EMG may
also help localize the lesion, especially if the Repair of Traumatic Peripheral
flexor carpi ulnaris or flexor digitorum pro- Nerve Injury
fundus muscle is affected. Sometimes pre-
operative studies provide inadequate or Intraoperative monitoring is particularly
inaccurate information. Factors that con- .useful when multiple, deep, proximal nerves
tribute to this include variability in the loca- are injured and several potential mecha-
tion of the cubital tunnel in relation to the nisms of injury are involved (for example,
medial epicondyle, selective damage to cer- traction, contusion, and ischemia). The pri-
tain fascicles within the ulnar nerve, techni- mary purpose of monitoring in this setting
cal difficulties with the recording (for ex- is to localize the injured segment(s) and to
ample, overstimulation causing current assess the status of axonal continuity across
spread), and the occurrence of lesions in the injured area.
unusual locations.14 Many of these difficul- Monitoring techniques can be helpful
ties can be avoided with intraoperative nerve during any peripheral nerve repair. This is
conduction studies. Conduction block and best illustrated by examining the role of
focal slowing are easier to detect and to lo- these techniques in the surgical repair of
calize accurately when the nerve is exposed. traumatic injuries of the brachial plexus.10'18
In addition, overstimulation is easier to de- The complexity of brachial plexus anatomy,
tect and to correct intraoperatively, and ar- the multiplicity and severity of injury to its
eas of increased threshold help identify dam- elements, and the frequent occurrence of
aged nerve segments. nerve root avulsion make lesions of this
The sensitivity and accuracy of intraoper- structure particularly difficult to evaluate
ative nerve conduction studies may help the and treat.10'19 The presence or absence of
surgeon choose the most appropriate treat- nerve root avulsion is one of the most im-
ment. A well-localized lesion in the region portant factors in determining prognosis
and the need for surgical intervention in bra- mary ramus of a given nerve root innervates
chial plexus injuries. If root avulsion is pres- paraspinal muscles at multiple levels, the dis-
ent, then repair of postganglionic elements tribution of fibrillation potentials may over-
innervated by the avulsed root will be of no estimate the number of roots involved. Fur-
benefit. The clinical examination, nerve thermore, the presence of a postganglionic
conduction studies, needle EMG, and myel- lesion with diminished sensory NAPs may
ography are used preoperatively to assess the mask an associated lesion involving pregan-
integrity of cervical nerve roots.10'19 The glionic segments.
combination of Horner's syndrome, dener- The predictive value of preoperative SEPs
vation of paraspinal and other proximal in detecting root continuity as well as the
muscles, preserved sensory NAPs, and the presence of mixed preganglionic and post-
presence of a meningocele on myelography ganglionic lesions has been disappointing.20
in association with a paralyzed anesthetic These uncertainties usually can be resolved
limb strongly suggest multiple root avul- by performing SEP recordings intraopera-
sions. However, any one of these findings in tively.21"23 In this setting, the exposed spinal
isolation is less predictive. Examples of false- nerve is stimulated directly by the surgeon
positive and false-negative myelograms have while SEP recordings are made from the cer-
been reported.10 Because the posterior pri- vical spine or scalp (or both) (Fig. 44-2). If
Figure 44-2. Electrophysiologic techniques for monitoring brachial plexopathy. In the upper right, somatosensory
evoked potentials (SEP) recorded over the scalp during root stimulation. In the middle, nerve action potentials
recorded directly (DNAP) from short segments of the plexus. In the lower left, compound muscle action poten-
tials (CMAP) recorded from distal muscles during selective stimulation of plexus elements. S, stimulating electrodes;
R, recording electrodes. (From Daube JR, Harper CM. Surgical monitoring of cranial and peripheral nerves. In
Desmedt JE [ed]. Neuromonitoring in Surgery. Elsevier Science Publishers, Amsterdam, 1989, p 135. By permis-
sion of the publisher.)
conduction velocity over short segments in

either CMAP or NAP recordings (Fig. 44-4)
sometimes localizes a lesion. When no
CMAP is recorded, there may be complete
disruption of axons at the site of injury or
there may be regeneration across the lesion,
but insufficient time for the regenerating ax-
ons to reach a distal muscle. In this setting,
the recording of a NAP across the injured
segment suggests regeneration is occurring,
whereas the absence of a NAP suggests se-
vere and complete axon loss. The latter find-
ing would suggest the need for lesion resec-
tion and subsequent grafting.4'5 Recording
the NAP from individual nerve fascicles has
been reported to help guide partial fas-
Figure 44-3. Intraoperative recording of somatosensory
evoked potentials over the scalp with stimulation of cer- cicular repair in cases of incomplete le-
vical nerve roots directly in the surgical field. Well- sions.11'12-25 Technical difficulties related to
defined responses were seen with stimulation of the C7 the size of the electrodes, occurrence of
and C8 roots. No response was obtained with stimula- shock artifact, and avoidance of current
tion of the C5 or C6 root, indicating avulsion of the spread to adjacent fascicles have limited the
root at these levels. (From Daube JR, Harper CM. Sur-
gical monitoring of cranial and peripheral nerves. In widespread use of fascicular recordings.
Desmedt JE [ed]. Neuromonitoring in Surgery. Elsevier
Science Publishers, Amsterdam, 1989, p 136. By per-
Prevention of Injury During
Peripheral Nerve Surgery
sensory fibers within the root are intact, a
well-defined SEP is recorded. The absence As are cranial nerves, the peripheral nerves
of a response confirms the presence of root that innervate the trunk and extremity
avulsion at that level (Fig. 44-3). Because muscles are susceptible to mechanical or
this technique tests sensory function, it is ischemic injury during various surgical pro-
theoretically possible that the ventral motor cedures. Electromyographic and SEP moni-
root could be avulsed with sparing of the toring has been used in an attempt to pre-
dorsal root. Motor evoked potentials with vent injury to the phrenic nerve26 and
transcranial stimulation and recording of brachial plexus during cardiac surgery.27
NAPs at the root level have been reported During these procedures, EMG monitoring
recently.24 If this technique proves to be refor neurotonic discharges helps localize and
liable, the combination of SEPs and MEPs warn of potential injury to the nerve, and di-
would be the most reliable way to assess the rect electric stimulation with a handheld
integrity of both sensory and motor compo- stimulator is used to identify viable nerves or
nents of the nerve root. fascicles.
When root continuity is present, attention Orthopedic procedures that involve disar-
is turned to assessment and possible repair ticulation or extensive manipulation of the
of postganglionic elements. Injured ele- limbs are often monitored. Electromyo-
ments of the plexus are stimulated proxi- graphic or SEP monitoring (or both) has
mally, with attempts to record a CMAP from been reported as a means to help detect and
a distal muscle or a NAP from the nerve dis- prevent injury to the axillary, radial, and
tal to the site of injury (Fig. 44-2). When a musculocutaneous nerves during shoulder
CMAP is present, the surgeon can be confi- surgery and to the femoral, obturator, and
dent that some motor axons are intact. In sciatic nerves during high-risk hip sur-
this setting, the lesion is left alone or, at gery.28"30 Electromyographic monitoring
most, simple neurolysis is performed. De- also may be useful during the resection of
tecting conduction block or focal slowing of primary or metastatic peripheral nerve neo-
Figure 44-4. Intraoperative nerve conduction studies showing conduction block along the medial cord of the bra-
chial plexus. (From Daube JR, Harper CM. Surgical monitoring of cranial and peripheral nerves. In Desmedt JE
[ed]. Neuromonitoring in Surgery. Elsevier Science Publishers, Amsterdam, 1989, p 136. By permission of the
publisher.)
plasms and in recognizing arterial occlu- multaneously preventing iatrogenic injury to

sions of the limb.31 In this case, the goal is peripheral nerve elements.
to resect the tumor with as little damage as
possible to normal nerve fascicles. During
dissection, individual or groups of fascicles REFERENCES
are stimulated mechanically or electrically
while EMG activity is monitored in distal 1. Brown WF, Veitch J. AAEM minimonograph #42:
muscles. An attempt is made to preserve fas- Intraoperative monitoring of peripheral and cra-
nial nerves. Muscle Nerve 17:371-377, 1994.
cicles that produce a distal EMG response, 2. Selesnick SH. Optimal stimulus duration for intra-
whereas those that do not are sacrificed. operative facial nerve monitoring. Laryngoscope
109:1376-1385, 1999.
3. Mandpe AH, Mikulec A, Jackler RK, Pitts LH, Ym-
gling CD. Comparison of response amplitude ver-
SUMMARY sus stimulation threshold in predicting early post-
operative facial nerve function after acoustic
Intraoperative monitoring of the peripheral neuroma resection. Am J Otol 19:112-117, 1998.
nervous system can be performed with EMG, 4. Tiel RL, Happel LT Jr, Kline DG. Nerve action po-
CMAPs, NAPs, and SEPs. Application of tential recording method and equipment. Neuro-
surgery 39:103-108, 1996.
these techniques requires adequate preop- 5. Oberle JW, Antoniadis G, Rath SA, Richter HP.
erative clinical and electrophysiologic as- Value of nerve action potentials in the surgical man-
sessment and consultation with the surgeon agement of traumatic nerve lesions. Neurosurgery
about the goals of the procedure. These 41:1337-1342, 1997.
6. Daube JR, Harper CM. Surgical monitoring of cra-
techniques can assist the surgeon in deter- nial and peripheral nerves. In Desmedt JE (ed).
mining the number, location, type, and Neuromonitoring in Surgery. Elsevier Science Pub-
severity of peripheral nerve lesions while si- lishers, Amsterdam, 1989, pp 115-138.
7. Brown WF, Ferguson GG, Jones MW, Yates SK. The of somatosensory evoked potentials and nerve con-
location of conduction abnormalities in human en- duction studies in the surgical management of bra-
trapment neuropathies. Can J Neurol Sci 3:111- chial plexus injuries. J Bone Joint Surg Br 62:492-
122, 1976. 496, 1980.
8. Campbell WW, Sahni SK, Pridgeon RM, Riaz G, 22. Sugioka H, Tsuyama N, Hara T, Nagano A,
Leshner RT. Intraoperative electroneurography: Tachibana S, Ochiai N. Investigation of brachial
management of ulnar neuropathy at the elbow. plexus injuries by intraoperative cortical so-
Muscle Nerve 11:75-81, 1988. matosensory evoked potentials. Arch Orthop
9. Harper CM, Daube JR. Facial nerve electromyog- Trauma Surg 99:143-151, 1982.
raphy and other cranial nerve monitoring. J Clin 23. Mahla ME, Long DM, McKennett J, Green C,
Neurophysiol 15:206-216, 1998. McPherson RW. Detection of brachial plexus dys-
10. Kline DG, Hackett ER, Happel LH. Surgery for le- function by somatosensory evoked potential moni-
sions of the brachial plexus. Arch Neurol 43:170- toring: a report of two cases. Anesthesiology 60:
181, 1986. 248-252, 1984.
11. Kline DG, Kim D, Midha R, Harsh C, Tiel R. Man- 24. Turkof E, Millesi H, Turkof R, Pfundner P, Mayr
agement and results of sciatic nerve injuries: a 24- N. Intraoperative electroneurodiagnostics (tran-
year experience. J Neurosurg 89:13-23, 1998. scranial electrical motor evoked potentials) to eval-
12. Kim DH, Kline DG. Management and results of per- uate the functional status of anterior spinal roots
oneal nerve lesions. Neurosurgery 39:312-319, and spinal nerves during brachial plexus surgery.
1996. Plast Reconstr Surg 99:1632-1641, 1997.
13. Gruen JP, Mitchell W, Kline DG. Resection and 25. Kline DG, Dejonge BR. Evoked potentials to eval-
graft repair for localized hypertrophic neuropathy. uate peripheral nerve injuries. Surg Gynecol Obstet
Neurosurgery 43:78-83, 1998. 127:1239-1248, 1968.
14. Campbell WW, Pridgeon RM, Sahni KS. Entrap- 26. Mazzoni M, Solinas C, Sisillo E, Bortone F, Susini
ment neuropathy of the ulnar nerve at its point of G. Intraoperative phrenic nerve monitoring in car-
exit from the flexor carpi ulnaris muscle (abstract). diac surgery. Chest 109:1455-1460, 1996.
Muscle Nerve 9:662, 1986. 27. Seal D, Balaton J, Coupland SG, et al. Somatosen-
15. Dawson DM, Hallet M, Millender LH. Entrapment sory evoked potential monitoring during cardiac
Neuropathies. Little, Brown, Boston, 1983, pp surgery: an examination of brachial plexus dys-
87-122. function. J Cardiothorac Vase Anesth 11:187-191,
16. Asami A, Morisawa K, Tsuruta T. Functional out- 1997.
come of anterior transposition of the vascularized 28. Helfet DL, Anand N, Malkani AL, et al. Intraoper-
ulnar nerve for cubital tunnel syndrome. J Hand ative monitoring of motor pathways during opera-
Surg [Br] 23:613-616, 1998. tivefixationof acute acetabular fractures. J Orthop
17. Geutjens GG, Langstaff RJ, Smith NJ, Jefferson D, Trauma 11:2-6, 1997.
Howell CJ, Barton NJ. Medial epicondylectomy or 29. Mills WJ, Chapman JR, Robinson LR, Slimp JC. So-
ulnar-nerve transposition for ulnar neuropathy at matosensory evoked potential monitoring during
the elbow? J Bone Joint Surg Br 78:777-779, 1996. closed humeral nailing: a preliminary report. J Or-
18. Slimp JC. Intraoperative monitoring of nerve re- thop Trauma 14:167-170, 2000.
pairs. Hand Clin 16:25-36, 2000. 30. Arlington ED, Hochschild DP, Steinagle TJ, Mon-
19. Davis DH, Onofrio BM, MacCarty CS. Brachial gan PD, Martin SL. Monitoring of somatosensory
plexus injuries. Mayo Clin Proc 53:799-807, 1978. and motor evoked potentials during open reduc-
20. Yiannikas C, Shahani BT, Young RR. The investi- tion and internal fixation of pelvis and acetabular
gation of traumatic lesions of the brachial plexus fractures. Orthopedics 23:1081-1083, 2000.
by electromyography and short latency somatosen- 31. Vossler DG, Stonecipher T, Millen MD. Femoral
sory potentials evoked by stimulation of multiple artery ischemia during spinal scoliosis surgery de-
peripheral nerves. J Neurol Neurosurg Psychiatry tected by posterior tibial nerve somatosensory-
46:1014-1022, 1983. evoked potential monitoring. Spine 25:1457-1459,
21. Landi A, Copeland SA, Parry CB, Jones SJ. The role 2000.
SECTION 3
Applications of
C inical
Neurophysiology:
Assessing Symptom
Complexes and
Disease Entities
Electrophysiologic assessments described in preceding chapters outline the
wide range of measurements that can be made in patients with suspected
disease of the central or peripheral nervous system. Each approach has ad-
vantages and shortcomings. The clinical neurophysiologic testing technique
that is most appropriate for a patient depends on the clinical problem. Of-
ten, some combination of techniques best provides the necessary data.
Selection of a technique requires, first, taking a medical history and ex-
amining the patient and, second, formulating a differential diagnosis. This
differential diagnosis should include many possible disorders; for practical
purposes, disorders that most likely are considered first in selecting a diag-
nostic technique. Findings of the clinical history and examination must also
be considered in selecting individual components of the techniques, for ex-
ample, which nerves to test with nerve conduction studies in a patient with
suspected peripheral nerve disease.
Decisions need to be made during clinical neurophysiologic testing to op-
timize data collection and to minimize the time, discomfort, and cost to the
patient. The approach to particular groups of clinical problems, with sug-
gestions on the approach to patients, is reviewed in Chapter 43, focusing on
the use of electroencephalography, electromyography, and nerve conduc-
tion studies. These suggestions will not be entirely correct for any individ-
ual patient because assessment depends on the unique history and physical
examination findings of each patient.
Chapter 45
APPLICATION OF CLINICAL
NEUROPHYSIOLOGY: ASSESSING
SYMPTOM COMPLEXES
Jasper R. Daube
Elson L. So
CLINICAL NEUROPHYSIOLOGY IN THE Electroencephalographic Evaluation of

ASSESSMENT OF DISEASE Cognitive Dysfunction
WHEN CLINICAL NEUROPHYSIOLOGY Electroencephalographic Evaluation of
CAN HELP Seizures and Other Paroxysmal Disorders
SYMPTOM COMPLEXES AND NEURAL Electroencephalography in the Intensive
SYSTEMS Care Unit
Assessment of Motor Symptoms Intraoperative Electroencephalography
Assessment of Sensory Symptoms Electroencephalography in the Newborn
Assessing Impairment of Consciousness and Electroencephalography in the Epilepsy
Cognition Monitoring Unit
Assessing Impairment of Visceral Function ASSESSING CLINICAL DISORDERS:
and Sleep ASSESSMENT WITH
LOCALIZATION OF DISEASE ELECTROMYOGRAPHY AND NERVE
Localization at the Supratentorial Level CONDUCTION STUDIES
Localization at the Posterior Fossa Level Cervical and Lumbar Radiculopathies
Localization at the Spinal Cord Level Electromyography in the Evaluation of
Localization at the Peripheral Level Peripheral Neuropathy
IDENTIFYING DISEASE TYPES Electromyography in the Evaluation of
PROGNOSIS Weakness
Electrophysiologic Classification of Nerve Unexpected Findings on Nerve Conduction
Injury Studies: Cause and Action
Focal Neuropathies and Radiculopathies Interpretation of Nerve Conduction Studies
Motor Neuron Disease SUMMARY
Myositis
ASSESSING CLINICAL DISORDERS:
ASSESSMENT WITH
Electroencephalographic Evaluation of
Impaired Consciousness or Delirium
565
566 Assessing Symptom Complexes and Disease Entities
The many techniques of clinical neurophys- potentials (BAEPs), and visual evoked po-
iology described in earlier chapters are ap- tentials (VEPs).
plied either to assist clinicians in assessing Anatomical localization is approached
disease of the central or peripheral nervous first by major level: supratentorial, posterior
system or, less commonly, in monitoring fossa, spinal cord, or peripheral nervous sys-
changes in neural function. These tech- tem. More specific localization is possible
niques can be used to monitor neural func- within each of these major levels. At the
tion in observing progression of disease or supratentorial level, the disease may involve
improvement in a patient's condition with the cerebral cortex, subcortical white mat-
specific treatment. They also are used in the ter, thalamus, hypothalamus, or basal gan-
intensive care unit and operating room to glia. At the posterior fossa level, the mid-
identify progressive neural damage. The fo- brain, pons, medulla, or cerebellum may be
cus of this chapter is the application of clin- involved. Spinal cord levels from C2 to S2
ical neurophysiologic techniques in assess- may be distinguished clinically. In the pe-
ing clinical problems. riphery, disease can be localized to the spinal
nerve, plexus, peripheral nerve, neuromus-
cular junction, or muscle. Specific clinical
neurophysiologic techniques can help local-
CLINICAL NEUROPHYSIOLOGY ize disease to each of these levels.
IN THE ASSESSMENT Selecting a treatment for the cause of
OF DISEASE symptoms requires identifying the underly-
ing disease and its stage. It is not sufficient
A patient comes to a physician with specific to know that a patient's symptoms are caused
complaints that require explanation and by a localized lesion in the midthoracic
treatment. The assessment of neurologic dis- spinal cord without knowing the nature of
ease begins with a hypothesis about the lo- the lesion. Thoracic cord disease caused by
cation and type of disease based on the pa- a herniated disk, intraspinal tumor, arterio-
tient's symptoms and history and the results venous malformation, multiple sclerosis, or
of neurologic examination. Clinical neuro- vitamin B^ deficiency are treated differ-
physiology can assist in localizing disease ently. Clinical neurophysiology can assist in
and defining the likelihood of different dis- distinguishing these diseases as well as in lo-
eases. After the disease has been localized calizing them. Also, it can help define prog-
and identified, selection of the appropriate nosis by classifying changes as acute, suba-
treatment requires determining the stage of cute, chronic, or residual from an old
the disease, including the severity of in- process. Whether the disease is rapid, inter-
volvement, the rate of evolution of the dis- mediate, slow, stable, or improving produces
order, and the prognosis. Clinical neuro- different clinical neurophysiologic findings.
physiology can assist in defining the stage of An acute process develops within seconds to
a disease. The patient's symptoms and the a few days. A subacute disorder evolves over a
signs localize the disease to a neural system few days to weeks and a chronic disorder, over
and to an anatomical level. Neurologic dis- months to years. In progressive diseases, there
eases may involve any combination of the is increasing damage and impairment of
motor system, sensory system, internal regu- function. Improvement occurs when the dis-
lation (autonomic nervous) system, systems ease process subsides and neural mecha-
regulating consciousness and cognition, and nisms of repair can begin to reduce the
vascular system. One or more of the tech- severity of damage. In a stable process, dam-
niques of clinical neurophysiology can test age has occurred but remains unchanged,
each of these systems, and the technique that because either the rate of neural reparative
is selected is the one that tests the portion processes is able to keep pace with the rate
of the nervous system usually responsible for of neural damage in a chronic continuous
the symptoms. For example, sensory symp- disorder or the disease process has subsided
toms are assessed with sensory nerve con- entirely but the damage cannot be repaired.
duction studies, somatosensory evoked po- This usually is referred to as a residual of the
tentials (SEPs), brain stem auditory evoked disease. Clinical neurophysiology can help
Application of Clinical Neurophysiology 567
identify or classify disorders according to situations in which the physician cannot ob-
these categories. Physicians and clinical neu- tain an adequate clinical history and perform
rophysiologists must be aware of the poten- an adequate neurologic examination, clini-
tial applications of clinical neurophysiology cal neurophysiology may provide the infor-
and make full use of them. mation needed to make a diagnosis. These
situations include patients who are in coma,
have dementia or psychiatric disease, or may
not be able to cooperate. A language barrier
WHEN CLINICAL may interfere with taking a medical history
NEUROPHYSIOLOGY and performing a neurologic examination.
CAN HELP When traumatic injuries such as fractures or
postoperative immobilization preclude thor-
The clinical neurophysiologic assessment of ough neurologic examination, clinical neu-
a suspected disorder of the nervous system rophysiology may be able to assess function
can help define each of the disease features and provide essential information.
described in the section. The focus of this Several electrophysiologic techniques can
section is on considerations important in de- be used to identify subclinical disease by de-
ciding whether one or more clinical neuro- tecting an abnormality that is either below
physiologic techniques are warranted for a threshold for clinical identification or has no
particular clinical problem. After taking the clinical accompaniments. Examples include
medical history and performing a neuro- epileptiform discharges between seizures,
logic examination, the physician often has slowing of conduction in a hereditary neu-
formulated a hypothesis about the neural ropathy with no deficit, and fibrillation po-
system involved, the level of involvement, the tentials in a radiculopathy with no clinical
type of disease, and the prognosis. If the deficit.
physician is sufficiently certain that the hy- In clinical situations in which the physi-
pothesis is correct, electrophysiologic testing cian is able to identify a category of disease,
is not needed. However, the information a clinical neurophysiology may be needed to
physician has is not always definitive enough characterize the disease. A patient who has
to determine the best approach to treat- brief spells that are clearly seizures may re-
ment. Clinical neurophysiology can provide quire EEG to determine whether the spells
the additional information that often is are absence seizures or partial complex
needed to determine the best treatment. seizures. A patient with a peripheral neu-
The most common application of clinical ropathy may require nerve conduction stud-
neurophysiology is to confirm a tentative ies to determine whether the neuropathy is
clinical diagnosis. Uncertainty about the di- axonal or demyelinating. Clinical neuro-
agnosis usually reflects an atypical or in- physiology may help localize the disease with
complete symptom complex, incomplete or a precision not otherwise possible clinically.
mixed findings that do not all fit with the For example, EEG may define the origin of
suspected disorder, a relatively mild stage of frontal lobe seizures or nerve conduction
the disease with a minimum of symptoms studies may identify compression at the
and signs, or unexpected findings that are aponeurosis of the flexor carpi ulnaris mus-
not consistent with the diagnosis. Elec- cle as the cause of an ulnar neuropathy.
troencephalography (EEG) can readily con- Clinical assessment usually can define the
firm that a patient with atypical symptoms severity of a disease as mild, moderate, severe,
has complex partial seizures. or with total loss of function. Clinical neuro-
Also, clinical neurophysiology can be of physiology can quantify the severity with
value when a specific disease is strongly sus- reproducible measures of the extent of
pected, but other diseases with similar find- abnormality. Quantification can assist in
ings have to be excluded. For example, a making decisions about the best treatment
patient with a C6 radiculopathy may have and prognosis, especially the likelihood of
features suggestive of carpal tunnel syn- improvement. As an example, nerve con-
drome. Clinical neurophysiology can be duction studies can define a 30% block,
used to help exclude the latter possibility. In rather than axonal loss, as the cause of weak-
ness in a peroneal neuropathy. Such quan- tem, clinical neurophysiologic testing with
tification provides evidence that the process one of the following modalities that assesses
is relatively mild and a good recovery can the motor system should be considered:
be expected. In summary, potential applica- • Electroencephalography should be con-
tions of clinical neurophysiology include the sidered if the motor symptom or deficit
following: may be caused by disease at the cortical
• Confirm diagnosis level, either a seizure or a destructive
• Exclude other disease process (for example, tumor or infarct).
• Identify unrecognized or subclinical Electroencephalography is of limited
disease value in assessing movement disorders,
• Localize the abnormality except when an epileptic mechanism is
• Define severity under consideration.
• Define pathophysiology (pathology/ • Transcranial motor evoked potentials
disease) (MEPs) (approved for clinical use in
• Define evolution, stage, and prognosis Canada and Europe, but not in the
United States) test the entire motor
pathway from the cerebral cortex
SYMPTOM COMPLEXES AND (where the motor system is activated by
NEURAL SYSTEMS a magnetic or electric pulse) to muscle
(where the response is recorded). Mo-
Neurologic symptoms will suggest whether tor evoked potentials are most useful
one or more neural systems are involved. in distinguishing primary motor system
Confirmatory signs on examination provide disease from functional or psychiatric
more certainty about the involvement of a disease. For example, a patient with hys-
particular neural system. Clinical neuro- terical paralysis of the legs can be eval-
physiologic testing can often provide further uated with this technique.
confirmation if needed. The symptoms that • Disorders of movement such as tremors,
suggest disorders of specific systems are as jerks, and twitches cannot always be clas-
follows: sified clearly by clinical observation.
• Motor system—paralysis, weakness, Quantitative measurements with multi-
tremor, other extraneous movements, channel surface electromyographic
and posture abnormalities (EMG) recordings from widespread
• Sensory system—sensory loss, paresthe- muscle groups can often characterize
sia, pain, and impairment of vision, the presence or absence of a motor sys-
hearing, or balance tem disorder and can sometimes define
• Disorders of consciousness—confusion, it precisely.
coma, syncope, and seizures • Motor nerve conduction studies can be
• Cognitive disorders—dementia, confu- critically important in identifying dis-
sion, and disorders of language ease of the peripheral motor pathways
• Internal control system and autonomic in the plexus, peripheral nerve, neuro-
disorders—perspiration abnormalities, muscular junction, or muscle.
fatigue, vascular changes, pain, and • Repetitive stimulation can separate
emotional disorders functional fatigue from defects of neu-
romuscular transmission caused by dis-
ease at the neuromuscular junction.
Assessment of Motor Symptoms Amplitudes of evoked responses can
help define the severity of the disease
Symptoms or signs involving movement are process and the presence of conduction
strong evidence that disease affects the mo- block. Conduction slowing can define
tor system at some level of the nervous sys- the precise location of the damage.
tem. If there is atrophy, loss of power, jerk- • Needle EMG can identify and charac-
ing, shaking, stiffness, or any of the many terize peripheral nerve and muscle
manifestations of disease of the motor sys- diseases.
Assessment of Sensory Symptoms their specific nature and, thereby, ap-

propriate treatment modalities.
Complaints of numbness, tingling, localized • Somatosensory evoked potentials and
pain, loss of vision, hearing impairment, and BAEPs can determine whether central
poor balance suggest involvement of sensory sensory pathways from the level of the
systems. Clinical neurophysiologic testing brain stem to the cerebral cortex are in-
can assist in identifying and characterizing tact and functioning normally. This
the disease involving the sensory system at helps define the status of patients in
any level of the nervous system. The follow- coma. Severity of impairment of SEPs
ing tests should be considered for patients after head trauma is helpful in judging
with sensory symptoms: prognosis.
• Electroencephalography can help iden- Electroencephalography is the most pre-
tify sensory symptoms arising at the cise laboratory technique available for as-
cortical level. For example, unilateral sessing cortical function. Disorders of cog-
paresthesia caused by a seizure dis- nition with impairment of mentation,
charge or a local destructive lesion may language, and memory are caused primarily
be associated with focal spikes or slow- by disorders at the level of the cerebral cor-
ing. tex. Only clinical testing of thinking and
• Somatosensory evoked potentials can memory can provide more information than
help determine the presence of impair- EEG. Characteristic EEC abnormalities can
ment along the sensory pathways, par- help define the nature of the disorders that
ticularly if there is an identifiable sen- produce impaired mentation, including
sory loss. Alzheimer's disease, chronic infection, isch-
• Sensory nerve conduction studies can emic vascular disease, subdural hematoma,
localize disease in peripheral nerves. and frontal lobe mass lesions.
Slowing of sensory nerve action poten-
tials (NAPs) may distinguish between
primarily axonal and demyelinating Assessing Impairment of Visceral
disorders. Function and Sleep
• Visual evoked potentials and auditory
evoked potentials (AEPs) can help lo- Impairment of visceral function, including
calize the site of abnormality and char- autonomic disorders, syncope, and disor-
acterize the type of abnormality, even ders of the reproductive organs, can be as-
without visual or hearing impairment. sessed with two groups of electrophysio-
logic tests: autonomic function testing and
polysomnography.
Assessing Impairment of • Autonomic function testing can assess
Consciousness and Cognition peripheral sympathetic function in vas-
cular disease or peripheral nerve dis-
Episodic or continuous confusion, uncon- ease. It also can assess central vascular
sciousness, and disorders of sleep can be control mechanisms that may be altered
evaluated and characterized with the follow- in autonomic diseases such as multisys-
ing electrophysiologic tests: tem atrophy.
• Electroencephalography, often critical • Polysomnography assesses sleep disor-
in defining the nature of an impairment ders by measuring both EEG activity
of consciousness, can distinguish sei- and associated autonomic function. Dis-
zure disorders from metabolic distur- orders of the autonomic nervous system
bances, focal lesions with increased may be manifested on polysomnogra-
pressure, hysterical disorders, and sleep phy. Polysomnography is particularly
disorders. valuable in central disorders of the in-
• Polysomnography provides a precise as- ternal regulatory system, but it can also
sessment and characterization of sleep be helpful in assessing impotence and
disorders, often assisting in defining sleep apnea.
LOCALIZATION OF DISEASE posterior fossa level and occasionally in lo-

calizing it within that level.
Clinical neurophysiologic testing often can • Auditory evoked potentials and audi-
be more precise than clinical evaluation in tory testing can specifically identify in-
defining the location of a disease process. volvement of the peripheral auditory
Even in cases in which clinical assessment in- pathways and distinguish it from in-
dicates the possibility of localized disease, volvement of auditory pathways at the
clinical neurophysiologic testing is often level of the pons or midbrain. Auditory
necessary to confirm localization. Clinical evoked potentials also can distinguish
neurophysiology can localize disease to and the nature of these disorders.
within the supratentorial, posterior fossa, • Posturography can identify the pres-
spinal, and peripheral levels. ence of a peripheral vestibular disorder
and distinguish it from central disease
of the vestibular pathways.
Localization at the • Comparison of tibial and median SEPs
recorded from the scalp with recordings
Supratentorial Level at the neck can identify involvement at
the posterior fossa level.
Each of the following clinical neurophysio- • Blink reflexes and facial nerve conduc-
logic measures may provide evidence of spe- tion studies can identify involvement of
cific involvement at the supratentorial level: cranial nerves V and VII and distinguish
• Electroencephalography is one of the their involvement from diseases of the
most helpful techniques for distin- midbrain or pons.
guishing among supratentorial diseases • Needle EMG of cranial muscles can pro-
because of its ability to identify localized vide evidence of damage to brain stem
areas of cerebral involvement caused by motor neurons or peripheral motor
either an epileptogenic or a destructive pathways in primary neurogenic pro-
process. Electroencephalography can cesses, myasthenia gravis, or primary
also help distinguish between subcorti- myopathies. Patients with bulbar symp-
cal and cortical diseases. toms can often be separated into those
• Evoked potentials can be used to dis- with an upper motor neuron-pseudo-
tinguish subcortical from cortical in- bulbar disorder and those with lower mo-
volvement. Somatosensory evoked po- tor neuron involvement in amyotrophic
tentials may increase or decrease in size lateral sclerosis or myasthenia gravis.
in cortical disease and be reduced or de-
layed in subcortical processes such as
multiple sclerosis. Visual evoked poten- Localization at the Spinal
tials are able to distinguish disease in Cord Level
the optic nerves and tracts from that in
the cerebral cortex. Clinical findings can usually identify the
• The origin of disorders with tremor, level of spinal cord disease unless multiple
jerking, and twitching may be identified disorders or multilevel involvement is pres-
as cortical or basal ganglionic by the char- ent, in which case clinical neurophysiology
acter of the pattern of firing and distri- can be particularly helpful, as indicated by
bution on multichannel movement re- the following:
cordings. • Somatosensory evoked potentials re-
corded with median, ulnar, or tibial
nerve stimulation can distinguish in-
Localization at the Posterior volvement of the peripheral and spinal
Fossa Level nerves from direct spinal cord damage
and separate lumbar and thoracic cord
Clinical neurophysiology can be helpful in lesions from cervical cord lesions. Oc-
identifying the presence of a lesion at the casionally, SEPs are able to determine
the nature of the disorder, especially de- ders from spinal cord disease and in-
myelinating disorders. volvement of central autonomic path-
• On nerve conduction studies, F waves, ways. Patterns of distribution of tem-
patterns of amplitude changes with mo- perature change, alteration in sweating,
tor and sensory conduction studies, and and vascular reflexes are combined to
H reflexes can sometimes identify in- provide this information.
volvement of specific levels of the spinal
cord.
• Paraspinal fibrillation potentials on IDENTIFYING DISEASE TYPES
needle EMG provide evidence of lower
motor neuron involvement at the spinal Clinical neurophysiology can facilitate iden-
cord level; they can also help define dis- tification of specific diseases. Electrophysio-
tribution of lower motor neuron loss logic testing can sometimes supplement the
along the spinal cord. For example, in initial classification of a disease as vascular,
amyotrophic lateral sclerosis, EMG may inflammatory, degenerative, or neoplastic,
demonstrate evidence of subclinical in- but it does so in different ways for different
volvement at the thoracic level. tests. In many instances, only a broad cate-
• Autonomic function testing in primary gory of disease can be suggested, but in oth-
spinal cord disease, particularly local- ers, specific diseases can be identified. These
ized disease with trauma, inflammatory are described in detail in the chapters on
disease, or ischemia, will show localized each of the techniques. The following are
changes at a specific segmental level. some examples. Electroencephalography
can distinguish epileptogenic from destruc-
tive lesions, but rarely can it categorize the
Localization at the changes as neoplastic, inflammatory, infec-
Peripheral Level tious, or degenerative. At times, specific
clinical entities can be suggested, such as
Peripheral lesions can be localized to the the Lennox-Gastaut syndrome, hepatic en-
level of a spinal nerve, plexus, peripheral cephalopathy, and hypsarrhythmia of in-
nerve, neuromuscular junction, or muscle fancy. Evoked potentials rarely provide evi-
and specifically to individual nerves and dence of a specific clinical entity. Marked
muscles. Four electrophysiologic tests assist increase in latency is usually evidence of a
in localizing peripheral lesions: demyelinating process in patients with mul-
• Motor and sensory nerve conduction tiple sclerosis. Evoked potentials are useful
studies identify localized areas of dam- particularly in identifying areas of subclini-
age to individual nerves. cal involvement to confirm the presence of
• Repetitive stimulation identifies and multiple lesions in multiple sclerosis.
characterizes disorders of the neuro- Nerve conduction studies infrequently
muscular junction. identify specific disease categories. Marked
• Needle EMG localizes lesions in cases in slowing and dispersion are evidence of a
which nerve conduction studies are not demyelinating neuropathy, which may be
successful, either because the damage is caused by either an inherited or an acquired
primarily axonal or the nerves are not disorder. Nerve conduction studies also help
accessible for stimulation and record- distinguish demyelinating disease from ax-
ing. Electromyography also can assist in onal disease. Repetitive stimulation can
distinguishing primary neurogenic dis- demonstrate specific patterns of abnormali-
ease from neuromuscular junction and ties seen in myasthenia gravis and distin-
muscle diseases. guish them from Lambert-Eaton myasthenic
• Autonomic function tests (quantitative syndrome. Occasionally, EMG can assist in
sudomotor autonomic response testing, identifying specific disorders by characteris-
thermoregulatory sweat test, and skin tic findings such as changes with polymyosi-
vascular reflexes) separate peripheral tis, periodic paralysis, and radiation damage
sympathetic and parasympathetic disor- with myokymia. Autonomic function testing
can provide evidence of specific disorders nerve conduction studies in many periph-
such as multisystem atrophy or reflex sym- eral nerve disorders. Identification of dis-
pathetic dystrophy. Peripheral audiologic ease type with EMG is well known, but
testing and posturography are able to dis- changes in EMG findings with time are less
tinguish central from peripheral structural familiar. They are reviewed below. Recogni-
lesions. Polysomnography testing can iden- tion of different stages in the evolution of
tify sleep apnea and some forms of sleep im- a disease depends on understanding the
pairment such as periodic movements of pathophysiologic changes that occur in
sleep. nerve and muscle. The three types of nerve
damage—conduction block, slowing of con-
duction, and axonal destruction—evolve
PROGNOSIS over very different time courses. Secondary
changes in muscle with each of these evolve
Although several clinical neurophysiologic over time courses that vary with severity of
procedures can define the severity of a dis- disease.
ease process, the procedures vary in their
ability to characterize the stage of evolution
of the disease or to provide prognostic in- Electrophysiologic Classification
formation. Electroencephalography is the of Nerve Injury
most established and widely used laboratory
procedure that provides nonstructural—but Conduction block is a localized area of abnor-
objective—evidence of the severity and pro- mality that is unable to conduct an action
gression of cerebral disorders. However, ex- potential in an axon or group of axons. The
cept for a few disorders, the association be- proportion of fibers that are blocked in a
tween an abnormal EEG finding and disease nerve is a direct measurement of the amount
severity varies among patients, partly be- of clinical deficit. Nerve function proximal
cause of the variability in the manner and and distal to the conduction block can be
the severity with which the cerebral symp- entirely normal. Conduction block may be
toms present for similar stages of the disor- caused by either metabolic or structural
der. A key principle in using EEG is that se- changes. Conduction block caused by local
rial EEG recordings must be obtained to anesthetics, anoxia, and some toxins may im-
optimize the value of this test in assessing prove over minutes to hours and may not be
disease progression and prognosis in each associated with histologic alteration. Con-
person. duction block caused by distortion or loss of
The physiologic and pathologic mecha- myelin persists for days to weeks, because it
nisms that underlie normal and abnormal requires remodeling of the histologic ab-
EEG are not as well understood as those in normality. Conduction block caused by ei-
peripheral nerve and muscle disorders. One ther of these mechanisms may not improve
reason is that brain specimens of living pa- if the offending mechanism is not elimi-
tients are not as readily available for study nated. Slowing of conduction is usually caused
and correlation with EEG findings. None- by myelin changes and, thus, requires weeks
theless, EEG is used to evaluate nearly all dis- to months for improvement to occur if the
eases affecting the brain. It remains the best underlying cause can be eliminated. In con-
tool for assessing severity and prognosticat- trast to conduction block, slowing of con-
ing the outcome of many categories of cere- duction alone may be associated with little
bral dysfunction, including toxic, metabolic, or no clinical deficit.
infectious, vascular, degenerative, traumatic, Axonal disruption or degeneration is associ-
and seizure disorders. Compared with other ated with a loss of axons through wallerian
neurophysiologic techniques, EMG and nerve degeneration. Therefore, recovery of func-
conduction studies are more consistent for tion depends on reinnervation. Reinnerva-
detecting abnormalities that typify the stage tion can occur rapidly, within days to weeks,
of development of an underlying neurologic if the number of axons lost is not great and
disorder. Clinically valuable prognostic in- the remaining axons can provide reinnerva-
formation can be gained from EMG and tion by local collateral sprouting. Reinner-
vation is much slower, over months to years, curs) . The corollary is that muscles closer to
if it requires sprouting and growth of the the lesion show fibrillation potentials sooner
damaged axons. than muscles more distant to the damage.
Electrophysiologic changes in muscle as- The reinnervation of muscle is associated
sociated with nerve damage depend primar- with a defined sequence of changes in the
ily on whether the degeneration is wallerian. estimate of the number of motor units and
Slowing of conduction is not associated with in MUPs. If reinnervation is by collateral
measurable changes on needle EMG or in sprouting, the estimate of the number of mo-
estimates of the number of motor units. tor units remains low and increases only in
Conduction block is associated with reduced proportion to the number of axons that re-
recruitment of motor unit potentials (MUPs) generate and reach the muscle. With initial
on needle EMG and reduced estimates of reinnervation, MUPs consist of activity in
the number of motor units proximal to the only a small number of muscle fibers, with
site of damage. Wallerian degeneration pro- poor synchrony of firing and unstable neu-
duces reduced recruitment of MUPs, re- romuscular junctions. Therefore, potentials
duced estimates of the number of motor are low amplitude, polyphasic, and unstable.
units, and muscle changes associated with These have been called nascent motor unit po-
denervation and reinnervation. tentials. As reinnervation proceeds to include
Denervation of muscle results in a loss of more muscle fibers with better synchrony of
the trophic factors that maintain normal firing, MUPs become higher in amplitude,
membrane function. With the loss of inner- longer in duration, less polyphasic, and
vation, a muscle fiber discharges sponta- more stable. Therefore, late after reinnerva-
neously and contracts in a rhythmical fash- tion, MUPs are of long duration and high
ion. The contractions are called fibrillation amplitude. Reinnervation usually is com-
and the associated discharges are fibrillation pleted by less than the normal number of
potentials. Fibrillation potentials develop 1-3 axons; thus, recruitment is decreased be-
weeks after acute denervation. Delay in their cause there are fewer motor units. These
appearance varies with species and muscle changes are summarized in Tables 45-1 and
characteristics. In humans, the delay de- 45—2. Both conduction block and axonal dis-
pends most on the length of axon attached ruption can have different time courses, de-
to the muscle fiber. If axonal destruction oc- pending on the underlying mechanism and
curs close to a muscle fiber, fibrillation de- the number of axons involved. The changes
velops more quickly than if the damage is over time in compound muscle action po-
more proximal (that is, the shorter the seg- tentials (CMAPs) and the results of needle
ment of axon attached to the muscle, the EMG examination after a localized nerve in-
more quickly wallerian degeneration oc- jury are shown in Tables 45-1 and 45-2.
Table 45-1. Compound Action Potential Amplitude after

Peripheral Nerve Injury*
AMPLITUDE
0-5 Days After 5 Days During Recovery
Conduction block
Distal stimulation Normal Normal Normal
Axonal disruption
Distal stimulation Normal Low Increases
*Supramaximal stimulation.
Table 45-2. Results of Needle Examination after Peripheral

Nerve Injury
0-15 Days After 15 Days During Recovery
Conduction block
Fibrillation potentials None None None
Motor unit potentials 4 number I number 1 number
Axonal disruption
Fibrillation potentials None Present Reduced
Motor unit potentials I number | number Nascent
I , decrease; f , increase.
Focal Neuropathies and ropathies and radiculopathies can be made

Radiculopathies on the basis of an analysis of the combina-
tion of these changes. Examples of these in-
Electrophysiologic changes on nerve con- terpretations are given in Tables 45-3 and
duction studies in mononeuropathies vary 45-4.
with the rapidity of development, the dura-
tion of damage, the severity of damage, and
the underlying pathologic condition. Local- Motor Neuron Disease
ized narrowing of axons or paranodal or in-
ternodal demyelination caused by a chronic Gradual loss of anterior horn cells in motor
compressive lesion produces localized slow- neuron disease produces changes in the
ing of conduction. Narrowing of axons dis- EMG findings during the course of the dis-
tal to chronic compression results in slowing ease. These changes allow electromyogra-
of conduction along the entire length of the phers to assess the evolution of the disease
nerve. Telescoping of axons with intussus- as well as its severity. In the initial stages of
ception of one internode into another pro- the disease, before clinical weakness is evi-
duces distortion and obliteration of the dent, collateral sprouting of viable motor
nodes of Ranvier and, thus, conduction neuron axons maintains innervation of all
block. Moderate segmental demyelination muscle fibers; thus, few if any fibrillation po-
and local metabolic alterations are often as- tentials are evident. However, the loss of
sociated with conduction block. With stimu- MUPs can be recognized. Later, MUP size
lation proximal to the site of damage, the increases with innervation of greater num-
conduction block is manifested as lower am- bers of muscle fibers. If a significant amount
plitude evoked responses. In an acute lesion of collateral sprouting has occurred, some
with disruption of the axons, the segment of MUPs vary in configuration. As these
nerve distal to the lesion may continue to changes progress to the stage where rein-
function normally for up to 5 days; then, as nervation cannot keep pace with denerva-
the axons undergo wallerian degeneration, tion, fibrillation potentials become promi-
they cease to conduct and the amplitude of nent. During this time, larger numbers of
the evoked response diminishes and finally regenerating fibers are present and inter-
disappears. One week after an acute injury, mittent blocking of the components of a
the amplitude of the evoked response is a MUP, motor unit potential variation, becomes
rough gauge of the number of intact viable more evident. The potentials become in-
axons. creasingly polyphasic, with satellite poten-
Interpretations about the duration and tials. This combination of polyphasic and
severity of nerve injury after focal neu- varying MUPs is evidence of a severe, pro-
Table 45-3. Interpretation of Electromyographic Findings

after Peripheral Nerve Injury
Finding Interpretation
0-5 Days
Motor unit potentials present Nerve intact, functional axons
Fibrillation potentials present Old lesion
Low-amplitude compound action Old lesion
potential
5-15 Days
Compound action potential, Conduction block
distal only
Low-amplitude compound action Axonal disruption
potential
After 15 days
Compound action potential, Conduction block
distal only
Fibrillation potentials Axonal disruption
Recovery
Increasing compound action Block clearing
potential
Decreasing number of fibrillation Reinnervation
potentials
Nascent motor unit potentials Reinnervation
gressing disorder. At times, it is accompa- tor units increases, resulting in larger than
nied by a decrement on slow repetitive stim- normal MUPs late in the disorder.
ulation (Table 45-5).
ASSESSING CLINICAL
Myositis DISORDERS:
ASSESSMENT WITH
Inflammatory myopathies also evolve over ELECTROENCEPHALOGRAPHY
time, beginning with small MUPs and
quickly developing fibrillation potentials As indicated by the discussion above, clini-
and polyphasic MUPs. The regenerating cal neurophysiologic testing cannot be ap-
muscle fibers and fibers that have lost their plied in a routine fashion. For each patient,
innervation because of nerve terminal dam- testing should be designed to answer the
age, segmental necrosis, or fiber splitting clinical question posed by the patient's prob-
produce a number of fibrillation potentials lem. This requires giving careful thought to
that roughly parallel the degree of disease the selection of the testing procedures. The
severity. As the disease subsides, fibrillation ensuing sections discuss the practical ap-
potentials become less prominent and mo- proach in using EEG, EMG, and nerve con-
tor unit potentials have a more normal size. duction studies to assess neurologic disor-
The number of muscle fibers in some mo- ders.
Table 45-4. Evolution of Electromyographic Changes in Radiculopathy
Acute (< 7 days) Subacute (weeks) Progressive Residual
Nerve conduction study
Compound muscle action Normal Low if severe Low if severe Low if severe
potential amplitude
Motor conduction velocity Normal < 30% slow if severe < 30% slow if severe (< 30% slow if severe)*
Motor distal latency Normal < 30% long if severe < 30% long if severe (< 30% long if severe)
F wave or H reflex Prolonged or absent Absent or prolonged Absent or prolonged (Prolonged or absent)
Needle electromyography
Fibrillation potentials None Proximal, brief Many proximal and distal Distal, small, few if any
Fasciculation potentials Rare Rare Contraction fasciculation Contraction fasciculation
Motor unit potentials Reduced recruitment Reduced recruitment Long duration, high Long duration, high
in weak muscles in weak muscles, amplitude, reduced amplitude, reduced
polyphasic, may recruitment in weak recruitment in weak
vary muscles, polyphasic, muscles, changes distal
may vary > proximal
" ( ), May recover to normal level.
Table 45-5. Evolution of Electromyographic (EMG) Changes in Muscle and Motor Neuron Disease
Subacute Active Chronic Active Inactive or Residual
Myositis
Nerve conduction studies Normal Low Normal
Fibrillation potentials Proximal, paraspinal muscles Widespread None
Complex repetitive discharges None Present None
Motor unit potentials Short duration, few polyphasic Short duration, polyphasic, Short duration, polyphasic,
rarely long duration rarely long duration
Motor neuron disease
Nerve conduction studies
Compound muscle action Normal unless severe Low if severe Low if severe
potential
Motor conduction velocity Normal < 30% slow if severe < 30% slow if severe
and F wave
Motor distal latency Normal < 30% long if severe < 30% long if severe
Repetitive stimulation Decrement in some Decrement in some Normal
Needle EMG
Fibrillation potentials Many Many Few, small, distal
Fasciculation potentials Frequent in mildly affected Frequent in mildly affected Rare or absent
muscles, absent in severely muscles, absent in severely
affected muscles affected muscles
Complex repetitive None Rare Occasional
discharges
Motor unit potentials Reduced recruitment, increased Reduced recruitment, long Long duration, high amplitude
duration, may be unstable duration, high amplitude,
polyphasic, unstable
Unlike other clinical neurophysiologic achieving, especially when the EEG is per-
tests, EEG is used frequently and regularly formed to evaluate impaired consciousness.
for patients of all age groups and with vari- The best approach to discussing the use
ous conditions. Therefore, the following in- of EEG in evaluating neurologic disorders is
formation about the patient and recording a symptom-oriented approach. The discus-
conditions must be obtained and docu- sion also should include the recording of
mented with each EEG procedure: (/) age, EEGs in special environments such as the op-
(2) clinical history, (3) reason for the pro- erating room and intensive care units. Pre-
cedure, (4) medications, (5} time of last ceding chapters have already presented the
meal, (6) number of hours of previous abnormal EEG features or patterns of spe-
night's sleep, (7) time of last occurrence of cific cerebral disorders. In the following
symptom, (8) current level of consciousness discussion, comments relevant to analog
and alertness, and (9) previous EEG records. recording (paper recording) are made be-
The following are the general technical cause there are several technical constraints
considerations when recording or reviewing in many clinical situations with such record-
an EEG: ing. Although digital EEG is being used in-
• Avoid setting low linear frequency filter creasingly in clinical practice, analog record-
at greater than 1.0 Hz. (Analog recording is still common in the United States and
ing should include pages with record- elsewhere in the world.
ing at 0.5 Hz.)
• Minimize distortion of both benign
sharp transients and abnormal sharp
waves by avoiding use of high linear
frequency filters below 70 Hz unless Electroencephalographic
necessary. Evaluation of Impaired
• Adjust sensitivity or gain settings if nec- Consciousness or Delirium
essary to properly display waveforms of
different amplitude range. (With ana- Impairment of consciousness is one of the
log recording, pages of different set- most frequent clinical manifestations of
tings may be necessary.) acute neurologic illnesses. Electrophysiologic
• Display the electrocardiographic moni- recording is the only laboratory method that
tor at proper amplitude so that com- objectively assesses the severity of distur-
plexes will be easily visualized yet not in- bance in cerebral function. This property of
terfere with EEG tracings. electrophysiologic recording makes it valu-
• Monitor respiratory movements and able as a method for objectively following
oximeter readings if necessary. the progress of the patient's condition to
• Stimulate the patient to elicit the most supplement clinical observation. In general,
alert state and its corresponding EEG the severity of EEG abnormality parallels the
appearance. observed depth of mental obtundation.
• Document the patient's behavior, espe- Several abnormal EEG patterns indicate
cially when there is an EEG discharge. the severity and the prognosis of the pa-
A segment of the recording should show tient's condition (for example, burst-sup-
EEG activity when the patient is most alert. pression, spindle coma pattern, alpha coma
This is important because many patients be- pattern). Moreover, a number of abnormal
come drowsy or sleepy during the EEG pro- EEG patterns suggest strongly the probable
cedure. The EEG activity associated with cause or mechanism underlying the mental
drowsiness are difficult to differentiate from obtundation (for example, triphasic waves,
abnormal background slowing of a mild de- seizure discharges, periodic lateralized epilep-
gree. Drowsy and sleep EEG activities can tiform discharges [PLEDs]). Not uncom-
also mask background abnormality of gen- monly, an EEG shows objective signs of cere-
eralized or focal slowing. The patient should bral dysfunction before other test results are
be stimulated verbally or physically during positive or available (for example, slowing or
the recording to determine the highest level PLEDs in infectious, ischemic, traumatic,
of arousal that the patient is capable of metabolic, or toxic disorders). The follow-
ing should be considered when performing dent of the patient's cognitive effort and
or reviewing EEGs for evaluation of mental performance at the time of the recording.
obtundation: Thus, EEG findings can complement the
• Use compressed time scale (slow paper clinical assessment of cognitive disorders by
speed) if subtle focal slowing or asym- detecting objective evidence of cerebral
metries are suspected. dysfunction. Furthermore, memory impair-
• Verbally stimulate the patient, and ob- ment or forgetfulness is frequently a symp-
serve and document any behavioral or tom of anxiety and depression. Recently, the
motor response. Use physical stimula- public's enhanced awareness of Alzheimer's
tion if necessary. disease has resulted in more patients be-
• Observe the patient continually for ab- coming concerned about the significance of
normal movements. their own symptoms of forgetfulness. Ab-
• When paroxysmal or periodic EEG pat- normal EEG findings help determine the or-
terns occur or the background of the ganic nature of the complaint. However,
EEG changes spontaneously, observe and normal EEG findings in combination with
document any corresponding change in normal clinical and laboratory results can be
motor or behavioral activity. Also, stim- reassuring. Although normal EEG findings
ulate the patient to determine whether do not completely exclude an organic cause
the EEG changes react. of cognitive disorder, an abnormal wake
• Use extraocular eye leads to help re- EEG background eliminates psychologic
solve the nature of frontal waveforms disorder as the only explanation of cognitive
that are difficult to distinguish from eye dysfunction.
movement artifact. The correlation between the severities of
• Reposition the head if ambiguous wave- EEG abnormality and cognitive dysfunction
forms occur in the cranial region that is not as good as that between EEG and men-
rests against the examination table. tal obtundation. This is particularly true for
• Consider recording sleep activity to ac- subcortical dementias. Despite this, a few
tivate spikes and sharp waves in patients cognitive disorders have characteristic EEG
with mild to moderate mental obtun- patterns. Conditions characterized by rapid
dation, unless the recording already cognitive decline in adults often require con-
shows abnormalities highly suggestive of sideration of Creutzfeldt-Jakob disease. Be-
the underlying cause or mechanism of cause brain biopsy is often avoided in these
delirium. patients, the characteristic pattern of peri-
• Consider serial EEGs to help monitor odic sharp waves may be the only supportive
the course of the patient's disorder. laboratory evidence available.
• In making the diagnosis of electrocere- Some specific causes of childhood de-
bral inactivity, adhere to the recom- mentia or progressive encephalopathy can
mendations of the American Clinical be suggested by characteristic abnormal
Neurophysiology Society (formerly the EEG patterns. Examples are the patterns of
American EEG Society). periodic waveforms in patients with meta-
bolic encephalopathy, subacute sclerosing
panencephalitis (SSPE), or abnormal stor-
age diseases. Epileptic encephalopathy can
Electroencephalographic also be suggested by the detection of fre-
Evaluation of Cognitive quent clinical or subclinical seizures or
Dysfunction widespread epileptiform discharges. Because
formal neuropsychologic testing may be dif-
Cognitive dysfunction is another frequent ficult to perform during early childhood,
neurologic complaint and presenting symp- EEG may be especially useful as an objective
tom. Whereas neuropsychologic testing di- tool in the serial assessment of cognitive
rectly measures the symptom of cognitive function.
deficit, EEG indirectly evaluates the severity The following should be considered when
of the cognitive dysfunction. However, the recording or reviewing EEGs for evaluation
information provided by EEG is indepen- of cognitive function:
• Verbally stimulate the patient, and ob- tural or functional derangement associated
serve and document any behavioral or with the seizure disorder.
motor response. Electroencephalography helps in deter-
• Use physical stimulation if necessary. mining the specific seizure-type or epilepsy
• Use compressed time scale (slow paper syndrome. The classifications of epileptic
speed) if subtle focal slowing or asym- seizures and epilepsy syndromes are based
metries are suspected. on both clinical information and the type of
• Observe the patient continually for ab- EEG abnormality. The location and distri-
normal movements. bution of lEDs or ictal discharges help de-
• When paroxysmal or periodic EEG pat- termine the seizure type and the epilepsy
terns occur or the background of the syndrome. At the minimum, the distinction
EEG changes spontaneously, observe between focal and generalized discharges
and document any corresponding ab- contributes importantly to the initial step
normal motor or behavioral activity. of seizure diagnosis, that is, establishing
Also, stimulate the patient to determine whether the disorder is focal or primary
whether the EEG changes react. generalized. This is an essential step in sei-
• Use extraocular eye leads to help re- zure management, because selecting the
solve the nature of frontal waveforms antiepileptic drug appropriate for treatment
that are difficult to distinguish from eye depends largely on distinguishing between
movement artifact. these two main seizure types. Although some
• Consider recording sleep activity to ac- antiepileptic drugs are effective for both sei-
tivate spikes and sharp waves in patients, zure types, many are effective in controlling
unless the recording already shows ab- one type and not the other. Knowledge of
normalities highly suggestive of the un- the location and distribution of lEDs and ic-
derlying cause or mechanism of delir- tal discharges is also essential in localizing
ium. the surgical focus and selecting candidates
• Consider serial EEGs to help monitor for epilepsy surgery.
the course of the patient's disorder. Certain interictal EEG discharge patterns
are characteristic of specific epilepsy syn-
dromes. In the appropriate clinical setting,
the hypsarrhythmia pattern is specific for in-
Electroencephalographic fantile spasms, whereas centrotemporal dis-
Evaluation of Seizures and charges induced by sleep are highly sugges-
Other Paroxysmal Disorders tive of the syndrome of benign rolandic
epilepsy (benign epilepsy with centrotem-'
Epileptic seizure episodes result from the ab- poral spikes). The diagnosis of the epilepsy
normal and excessive discharge of neurons. syndrome determines the clinical manage-
Thus, EEG is commonly used to detect ab- ment and prognosis of many seizure disor-
normal interictal and ictal electric dis- ders. Many epilepsy syndromes are age-de-
charges that are highly associated with pendent in onset and remission, and the
epileptic seizure disorders. Interictal abnor- likelihood of spontaneous remission is good.
malities that have the best association with In comparison, some syndromes typically are
epileptic seizure disorders are spikes, sharp intractable to drug treatment. Also, certain
wave discharges, and temporal intermittent epilepsy syndromes are highly associated
rhythmic delta activity (TIRDA). These in- with an underlying structural abnormality.
terictal epileptiform discharges (lEDs) oc- The presence or absence of lEDs can serve
cur in only 40% of the initial EEGs of pa- as a prognostic factor in assessing the risk of
tients with seizures and in approximately seizure recurrence. Some studies have sug-
1% of persons without epilepsy. Nonethe- gested that with first unprovoked seizure dis-
less, their presence strongly supports a di- orders the presence of lEDs is associated
agnosis of seizure disorder when other ap- with a higher risk of seizure recurrence.
propriate clinical or laboratory data are Many studies also support the finding that if
present. Furthermore, focal or generalized the current EEG shows lEDs in seizure-free
EEG slowing may reveal an underlying struc- patients who discontinue taking antiepilep-
tic medication, seizures are more likely to positive leg pressure devices. Placing elec-
recur. trodes on the patient and making the EEG
The following should be considered when recording may interfere with nursing care
evaluating patients for epileptic seizure dis- and vice versa. Generally, the patients in in-
orders and other paroxysmal events: tensive care who have an EEG study have al-
• Obtain a sleep recording as well as a tered mentation or are experiencing seizures
wake recording, unless the wake record- and other paroxysmal events. Thus, recom-
ing has already disclosed IED activity mendations made above about EEG record-
that is sufficient for clinical manage- ing for specific clinical situations should be
ment. followed when applicable (that is, recording
• Consider partial sleep deprivation be- the EEG of a patient in coma or with
fore the EEG procedure, especially if a seizures). Additional recommendations for
previous EEG did not show epileptiform making EEG recordings in the intensive care
abnormalities. unit are the following:
• Schedule sleep-deprived patients for • Ensure electrical safety (Chapter 2).
EEG to be performed the following morn- Avoid introducing the patient into
ing and not the afternoon. the path of a ground loop or double
• If the patient still is unable to fall asleep ground, especially a patient with an in-
during the procedure despite sleep dep- dwelling cardiovascular catheter.
rivation, consider administering chloral • Observe closely for artifacts. Determine
hydrate to promote sleep. (Precautions and document their origin.
of conscious sedation should be exer- • If artifacts from other equipment inter-
cised, particularly for children. Instruct fere excessively with the EEG recording,
the patient not to drive for the rest of inquire whether the equipment respon-
the day if a sedative is given.) sible for the artifact can be turned off
• Use anterior temporal electrodes to en- or removed temporarily.
hance the probability of recording tem- • Modify and document electrode place-
poral lEDs. ments if head dressings or wounds in-
• Perform photic stimulation and hyper- terfere with standard electrode place-
ventilation unless contraindicated med- ments.
ically. • Consider prolonged recording or in-
• Use precipitating measures for patients termittent recordings to monitor the
whose spells have known precipitants. clinical course of the patient.
• Consider supplementing the "routine"
EEG recording with simultaneous video
recording if the patient is experiencing
daily spells. Intraoperative
• Use one channel for the electrocardio- Electroencephalography
graphic monitor and another channel
for oximeter monitoring. Electroencephalographic monitoring of the
cerebral cortex is performed most often dur-
ing carotid endarterectomy and epilepsy sur-
gery. Recording in the operating room set-
Electroencephalography in the ting presents challenges similar to those of
Intensive Care Unit recording in the intensive care unit. Intra-
operative EEG has additional constraints,
Recording EEGs in the intensive care unit such as anesthetic agents and limited ability
presents special challenges. Several devices to physically adjust the patient or the equip-
and pieces of equipment in the intensive care ment. The recording must be interpreted
unit can introduce artifacts into the EEG immediately to provide the information nec-
recording and make EEG recording difficult, essary to guide the surgical procedure. Be-
such as electrocardiographic and blood pres- cause intraoperative recordings are essen-
sure monitors, indwelling catheters, respira- tially prolonged monitoring that extends
tors, intravenous pumps, surgical drains, and over hours, digital EEG should be used. Dig-
ital recording allows prompt retrieval of seg- sons, EEG is frequently used to detect ob-
ments of the recording for side-by-side com- jective abnormalities that help in determin-
parison to assess the course of the patient ing the mechanism or the nature of the
and the effect of surgical intervention. The clinical manifestations in the newborn.
following should also be considered when Recording EEG in the newborn presents
performing intraoperative EEG and evoked unique challenges and requires special skills.
potential studies: Considerable skill is needed in applying elec-
• Obtain a baseline recording before trodes on a small head, especially in prema-
anesthetic agents are administered. To ture neonates. The scalp of the newborn is
ensure the adequacy of the recording, more delicate than that of an older child or
the baseline record may have to be adult. Extracerebral monitors such as those
made outside the operating room. for eye movements, respiration, and muscle
• Plan the type of recording needed for activity are needed to help define the wake
the operation with the surgical and and sleep states in the newborn. Many EEGs
anesthesia staff. of premature newborns are performed in
• Ensure that the electrodes and cable the neonatal intensive care unit. Thus, the
connections are stable after the patient requirements and constraints discussed in
has been positioned but before being the preceding section about recording in the
draped. Label each electrode clearly. intensive care unit setting also apply. In ad-
• Establish that access to the recording dition, the following should be considered:
surface and the equipment is possible • Assure the parents or caregiver about
even when the patient is draped and the the nature of the study.
operation has commenced. • If possible, perform the recording dur-
• Document the anesthetic agents used, ing or right after feeding.
and promptly communicate with the • Use miniature cup electrodes for
anesthesiology staff any need to modify recording the EEG, surface EMG, eye
the level or type of anesthesia or seda- movements, and electrocardiogram; use
tion. piezoelectric transducers or imped-
• Verify and graphically document the ance pneumographs for recording res-
electrode placements, especially intra- piration.
cranial electrodes. • In the newborn, oximeter recordings
• Promptly inform the surgical staff about can identify seizures that result in oxy-
the status and the quality of the record- gen desaturation.
ing. • Use a nasal thermistor for recording air-
• Document on the recording the stages flow and simultaneous repiratory piezo-
of the operation and any developments. electric transducers or impedance pneu-
mographs if apnea is suspected.
• Make certain that the setting is well ven-
Electroencephalography in tilated if using collodion or acetone. Do
the Newborn not use either of these substances inside
an isolette.
Frequently, EEG is performed in a full-term • Use a heat lamp, and monitor body tem-
or premature newborn for evaluation of sus- perature if the newborn is removed
pected abnormal movements and apneic from the isolette for applying elec-
episodes. Clinical manifestations of seizures trodes.
in the newborn differ from those in older • Note if there is scalp edema, which may
children and adults. Many of the seizure be- affect the EEG recording.
haviors in the newborn are subtle, and many • Use the newborn montage, with fewer
also mimic normal physiologic events. In the electrodes.
newborn, apnea is much more frequently an • Use the appropriate frequency filter set-
epileptic manifestation than it is in older pa- tings and gain to optimize recordings.
tients. However, apnea in the newborn is • Perform part of the recording with a
also commonly a manifestation of cerebral montage that includes the midline or
injury or severe prematurity. For these rea- central regions.
Adjust or prop the patient's head to • Make sure the electrode connections
minimize electrocardiographic and bal- are stable and the patient is in view
listocardiographic or other movement throughout the monitoring. Use a high-
artifacts. resolution color camera during the day-
time and an infrared camera at night to
obtain the best video image.
• Facilitate the occurrence of seizures or
Electroencephalography in the symptoms by activating procedures if
Epilepsy Monitoring Unit necessary (for example, withdrawal of
antiepileptic drug therapy, sleep depri-
Video-EEG monitoring has become a major vation, photic stimulation, hyperventi-
procedure in clinical neurophysiology (see lation, or psychologic suggestion).
Chapter 11). Currently, epilepsy monitoring • Provide continuous visual monitoring
is conducted in many hospitals and clinics. unless the type of spell or seizure is not
The monitoring can be done in a dedicated likely to cause injury. Be aware that pa-
facility with fixed equipment or in other lo- tients with epilepsy are at risk for the
cations with mobile recording equipment. development of prolonged convulsive
The advent of digital EEC and video has seizures after antiepileptic drug treat-
made it easier to store and to access data for ment has been withdrawn.
review. The correlation between EEG activ- • For each patient, provide a plan for
ity and clinical behavior of the patient is en- preventing or interrupting prolonged
hanced by the simultaneous display of video seizures or spells. This may include hav-
and EEG data. The recording can be re- ing venous access with a heparin lock.
trieved and reviewed at remote locations as Certified equipment and qualified per-
needed if the equipment used for recording, sonnel must be immediately available
storing, and reviewing data are linked in a for cardiorespiratory resuscitation.
network. • Except for brief seizures such as ab-
Successful use of video-EEG monitoring sence or myoclonic seizures, evaluate
depends on both technical and nontechni- the patient when each seizure occurs,
cal factors. Epilepsy monitoring is per- noting the time of occurrence.
formed best as part of a comprehensive • Ictal and postictal neurologic evalua-
program of patient evaluation and manage- tions should include tests for alertness,
ment. Essential participants in the program orientation, comprehension, language,
include nurses, occupational therapists, psy- and motor function.
chologists or psychiatrists, and social work • Encourage assisted physical activity to
personnel. The need for their services minimize complications of long-term
should be individualized according to the bed rest.
medical, psychologic, and social conditions • Adjust settings of spike or seizure de-
of each patient. The following should also tection programs according to patient's
be considered when conducting epilepsy sleep—wake state and physical activity to
monitoring: optimize detection of events.
• Counsel patients and guardians about
the nature and the requirements of the
procedure, including the need to push ASSESSING CLINICAL
the event button when symptoms occur. DISORDERS: ASSESSMENT
• Ensure a safe monitoring environment.
Minimize or strategically locate equip- WITH ELECTROMYOGRAPHY
ment and furniture. Protect the patient AND NERVE CONDUCTION
from hard surfaces or protruding fix- STUDIES
tures. Use padding if necessary.
• The patient should be accompanied Because the types and locations of neuro-
when ambulating if spontaneous or ac- muscular disorders are relatively well defined,
tivated seizures are likely to occur. Have algorithms for testing can be developed. The
selected patients wear helmets. algorithms must take into account findings
obtained during the test in order to deter- rillation potentials (evidence of axonal de-
mine the amount and types of testing that struction) and also the earliest to show im-
should be performed. They are suggestions provement. Changes in paraspinal muscles
for EMG and nerve conduction studies in can localize the damage proximal to the
neuromuscular disorders and nearly always plexus. Persistent abnormalities in para-
need to be modified according to the par- spinal muscles after neck or back surgery
ticular problem and findings in individual preclude postoperative testing of these mus-
cases. The following sections oudine a set of cles. If diagnosis is uncertain, preoperative
possible algorithms for several neuromuscu- EMG studies may be helpful. The peripheral
lar problems. distribution of an abnormality defines the
root involved. Severity of damage can be es-
timated by the amount of motor fiber de-
generation (fibrillation potentials and MUP
Cervical and Lumbar changes). Electromyography is particularly
Radiculopathies valuable in differentiating relatively recent
nerve damage with abundant fibrillations
The diagnostic value of EMG in assessing pa- (especially in proximal muscles) from the re-
tients who may have radiculopathy includes sidual of old disease with scanty fibrillation
answering the questions: (1) Is there evi- potentials (mainly in distal muscles).
dence of radiculopathy? (2) Which nerve Electromyography does not define the
root is involved in the radiculopathy? (3) cause of the radiculopathy. Electromyo-
How severe is the neural damage caused by graphic signs of a localized radiculopathy
the radiculopathy? (4) Is the radiculopathy could be similar whether caused by a disk,
of recent onset, is it ongoing, or is it a re- tumor, or diabetes mellitus. Because disor-
sidual of an old lesion? (5) Is there evidence ders of the nerve roots produce changes only
of other peripheral nerve disease? Nerve if the nerve fibers are damaged, EMG can
conduction studies primarily answer ques- never exclude the presence of a radiculopa-
tion 5. However, F-wave latencies and H- thy and EMG findings may be normal even
reflex latencies can measure conduction when the radiculopathy causes severe pain.
through the nerve roots. In a small propor- The following algorithms suggest specific ap-
tion of patients with lesions of the C7, C8, proaches to suspected radiculopathy.
L5, or SI nerve root, particularly those with
recent damage, the F waves or H reflexes CERVICAL RADICULOPATHY
may be abnormal when other measurements (ARM PAIN)
are normal. Damage to the C8, L5, or SI root
may also cause low-amplitude motor re- Nerve Conduction Studies to Identify and
sponses and mild slowing of conduction ve- Localize Peripheral Nerve Damage
locity in the median, ulnar, peroneal, or tib- 1. Median motor conduction study with F
ial nerves. Determining the amount of waves.
amplitude reduction that occurs with radicu- 2. Ulnar motor conduction study with F
lopathy helps define the amount of axonal waves. Record first dorsal interosseous
destruction. Sensory NAPs generally are nor- for C8-T1 symptoms.
mal in nerve root disease and are helpful 3. If either of the above show unexpected
in differentiating it from more peripheral abnormal findings, check technical fac-
disease. tors, temperature, anomalies, and con-
Needle electrode examination of muscles sider another algorithm, as appropri-
is still the most useful method for identify- ate.
ing radiculopathy. Because EMG changes 4. If symptoms are nonspecific or suggest
evolve over time, the age of the lesion can need for C4-C8 sensory conduction
be judged from both the distribution and studies:
the type of abnormality. Well-defined fibril- a. Median sensory conduction study
lation potentials are not seen until 3 weeks b. Ulnar sensory conduction study
after nerve damage. Proximal and para- 5. If symptoms suggest C5-C6 or upper
spinal muscles are the earliest to show fib- trunk, consider musculocutaneous/
biceps and lateral antebrachial sensory 7. If peroneal amplitude is low or pe-

conduction study. roneal conduction velocity is 35-42
6. If the only abnormality is a prolonged m/second, consider superficial per-
F wave or low CMAP, compare the op- oneal sensory testing.
posite side. 8. Sural sensory conduction study. Con-
Needle Examination sider medial plantar sensory conduc-
1. First dorsal interosseous, pronator tion study if patient is younger than
teres, biceps, triceps, infraspinatus, cer- 55 years.
vical paraspinal muscles (unless there 9. If peroneal or tibial amplitude is low, F
has been posterior neck surgery), and waves are abnormal or borderline, H
one weak muscle. reflex is abnormal or borderline, or ve-
2. If one root is suspected clinically or if locity is slow, compare with the oppo-
any abnormality is seen, examine two site side.
more muscles in the distribution of the Needle Examination
suspected nerve root (proximal and 1. With no specific root suspected: ante-
distal) and demonstrate normal muscle rior tibial, peroneus longus, posterior
above and below the level of the in- tibial, medial gastrocnemius, lateral
volved root. gastrocnemius, vastus medialis, rectus
3. Check the paraspinal muscles if limb femoris, gluteus medius, gluteus max-
muscles are abnormal, if symptoms are imus, sacral paraspinal, low lumbar
of recent onset, if radiculopathy is paraspinal, mid-lumbar paraspinal, up-
likely, or if symptoms are only in the per lumbar paraspinal muscles (if no
neck. previous spine surgery).
4. If the results of the needle examination 2. If a specific root is suspected: add mus-
are abnormal, check at least one con- cles in that distribution. If any muscle
tralateral muscle. is abnormal, compare the most abnor-
mal muscle with that on the other side.
LUMBOSACRAL RADICULOPATHY
Nerve Conduction Studies to Identify and
Localize a Peripheral Nerve Disorder Electromyography in the
1. Peroneal motor conduction study with Evaluation of Peripheral
F waves Neuropathy
2. If knee amplitude is greater than ankle
amplitude, check for supramaximal Testing for a neuropathy includes both
stimulation at the ankle, excess stimu- nerve conduction studies and needle EMG.
lation at the knee, or accessory per- Electromyography can confirm the presence
oneal nerve. of peripheral nerve dysfunction and distin-
3. If ankle amplitude is 20% greater than guish patients with radiculopathy from those
knee amplitude or velocity is less than with complaints caused by spinal cord or
35 m/second, consider peroneal neu- nonorganic disease. In the presence of neu-
ropathy. ropathy, EMG may provide localizing or eti-
4. Tibial motor conduction study with F ologic information. Different patterns of ab-
waves normality are found in demyelinating and
5. If tibial ankle amplitude is 50% greater axonal neuropathies, in large-fiber and
than knee amplitude or velocity is less small-fiber neuropathies, in polyradiculopa-
than 35 m/second, consider another thy, and in mononeuritis multiplex. Normal
disease. findings do not exclude peripheral neu-
6. If SI radiculopathy is strongly sus- ropathy. The results of nerve conduction
pected, no clinical signs are present studies and EMG are often normal with
and motor nerve conduction studies small-fiber involvement in diabetic, amyloid,
are normal, check tibial H reflex from or hereditary sensory neuropathy. In some
the knee (must be compared with op- patients with complaints of vague or non-
posite side). specific pain (especially those who have dia-
betes), EMG may provide evidence of nerve 7. If arm is tested first, do at least one
damage before it is evident clinically. Pain- motor and sensory conduction study
ful diabetic radiculopathies are often con- in a leg.
firmed with EMG. 8. Median sensory conduction study.
After a peripheral neuropathy has been di- a. Palmar stimulation if antidromic
agnosed, EMG can help classify the disease response is absent or carpal tunnel
and suggest a possible cause. Findings of seg- syndrome is suspected
mental demyelination suggest an inherited, b. If no responses are seen in distal
autoimmune, or inflammatory neuropathy; sensory nerves, consider proximal
axonal degeneration suggests a toxic, meta- testing, for example, antebrachial
bolic, or nutritional neuropathy. Some cutaneous.
neuropathies have characteristic electrodi- c. Consider ulnar sensory conduction
agnostic patterns, such as prominent study if symptoms are appropriate
paraspinal fibrillation potentials with a or median sensory conduction is
mixed demyelinating and axonal neuropa- abnormal.
thy in diabetes or bilateral carpal tunnel d. Consider radial sensory conduc-
syndrome superimposed on a largely axonal tion study if compression syndromes
neuropathy in amyloidosis. F-wave latencies are present.
provide a measure of proximal conduction 9. Ulnar motor conduction study with F
and, in some disorders, may show early ab- waves.
normality. With many nerves available to Four-point ulnar study only if there is
test, the most appropriate ones must be se- a significant decrease in amplitude
lected. For example, plantar nerves show across the elbow
earlier abnormalities in neuropathy than 10. Test additional nerves if findings are
more proximal sensory nerves. equivocal.
11. Test additional proximal conduction
if there are no responses or poly-
PERIPHERAL NEUROPATHY radiculopathy is possible:
a. Blink reflex: unilateral Rl and R2
1. Test the most involved extremity (arm responses
or leg) first if the deficit is mild or b. Musculocutaneous motor conduc-
moderate; test the least involved ex- tion study
tremity if the deficit is severe. or
2. Peroneal motor conduction study c. Somatosensory evoked potential
with F waves study in arm or leg or both
3. Tibial motor conduction study with 12. Needle examination—The following
F waves muscles are appropriate in each case:
4. If no responses, test peroneal nerve anterior tibial, medial gastrocnemius,
conduction to the anterior tibial lumbar paraspinal, and first dorsal in-
muscle. terosseous muscles.
5. Sural sensory conduction study a. If these are normal, test foot
a. Consider superficial peroneal sen- muscles.
sory conduction study if no sural b. If any of these are abnormal, test
response. the opposite extremity and proxi-
b. If patient is younger than 55 years mal muscles of leg (and arm), in-
or sural sensory conduction study cluding paraspinal muscles.
is normal (or both): plantar sen-
sory conduction CARPAL TUNNEL SYNDROME
6. If any of the above are abnormal and
a mononeuritis multiplex or other 1. Median motor conduction study with
asymmetrical process is suspected, test F waves
the opposite leg. If not, or if upper ex- Consider median motor distal latency
tremities are symptomatic, test the to 2nd lumbrical, with or without inch-
arm. ing, compared with ulnar distal latency
recorded to the 2nd dorsal interossei tude/area difference is greater than

over the same distance (normal, < 0.4- 20%, add below-elbow and upper-arm
ms difference) if no thenar response or stimulation (four points must be stim-
other studies are normal. ulated) .
2. Ulnar motor conduction study with F 4. Check for median-to-ulnar anomalous
waves anastomosis if there is only an ampli-
3. Check for anomalous innervation if tude change.
amplitude or configuration changes 5. If abnormal, use 2 cm inching from
from elbow to wrist on ulnar or median below to above elbow.
motor conduction studies. 6. If clinically indicated, do ulnar study
4. Median palmar sensory conduction with wrist, elbow, upper arm, and
study with wrist and elbow recording: supraclavicular and/or root stimula-
Consider antidromic median sensory tion (arm down) for possible plex-
conduction study if motor study is ab- opathy.
normal or technical problems are an- 7. If ulnar and median study distal la-
ticipated. tencies are prolonged, check for tech-
5. Ulnar sensory conduction using the nical problems (for example, low tem-
same method as median sensory con- perature, anode distal).
duction, including conduction velocity 8. If ulnar nerve conduction is slow only
Consider recording antidromic rein forearm, consider needle stimula-
sponses from thumb or middle finger tion to localize the site of damage.
or using radial sensory conduction if 9. If ulnar conduction velocity is normal
both ulnar and median sensory con- with atrophy or only a long distal
duction values are abnormal. latency, consider ulnar stimulation
6. Check the temperature of the hand if while recording first dorsal interosse-
ulnar and median sensory distal laten- ous muscle for lesion in Guyon's canal
cies are long. or flexor carpi ulnaris recording for
7. Contralateral median and ulnar sen- selective lesion of branch to that
sory conduction studies if the ipsilat- muscle.
eral median conduction study is ab- a. Dorsal ulnar sensory recordings or
normal or symptoms are bilateral comparison of ulnar conduction
8. Contralateral median motor conduc- with the 2nd dorsal interosseous
tion if the Contralateral median sensory with median to the 2nd lumbrical
conduction is abnormal may help identify ulnar damage in
9. Needle examination of ipsilateral first the hand.
dorsal and thenar muscles: b. Consider using flexor carpi ulnaris
a. Examine flexor pollicis longus and muscle recording if there is no
pronator teres if thenar muscle hand response.
EMG is abnormal. 10. Ulnar antidromic sensory conduction
b. Examine additional muscles, espe- study.
cially those innervated by C6 if nerve a. Add below elbow and upper arm
conduction values are normal. stimulation if abnormal.
b. Use palmar stimulation if wrist le-
ULNAR NEUROPATHY sion is suspected. Near-nerve nee-
dle recording may show selective
1. Ipsilateral median motor conduction slowing in some fascicles.
study with F waves. 11. Median antidromic sensory conduc-
2. Ipsilateral ulnar motor conduction tion study (palmar stimulation if
study to both the hypothenar and first carpal tunnel syndrome is also sus-
dorsal interosseous muscle, with F pected)
waves. 12. Perform the same study contralater-
3. If ulnar conduction velocity is less ally if motor amplitude decreases
than 52 m/second (compare with more than 20% across the elbow,
median) or the elbow-wrist ampli- motor conduction slows more than
8 m/second across the elbow (long- 6. Consider serratus anterior and di-
segment calculation), ulnar sensory aphragm.
conduction is absent, ulnar sensory B. Middle or Lower Trunk Damage
conduction is slow, other specific ab- 1. Median motor conduction study with
normality is identified, or symptoms F waves
are bilateral. 2. Ulnar motor conduction study with
13. Needle examination: first dorsal in- stimulation at the wrist, elbow, upper
terosseous, abductor digiti quinti, and arm, and supraclavicular with F waves
flexor carpi ulnaris 3. If normal with a lower trunk deficit,
14. If ulnar muscles are normal: flexor consider nerve root stimulation.
pollicis longus, extensor indicis pro- 4. Median sensory conduction study with
prius, pronator teres, triceps, and wrist, elbow, upper arm, and supra-
biceps clavicular stimulation. Consider nerve
15. If ulnar muscles are abnormal: add ab- root stimulation.
ductor pollicis brevis, extensor carpi 5. Ulnar sensory conduction study with
ulnaris, extensor indicis proprius, and wrist, elbow, upper arm, and supra-
contralateral first dorsal interosseous clavicular stimulation. Consider nerve
to the above. root stimulation.
16. If other C8 muscles are abnormal, ex- 6. Radial sensory conduction study
amine paraspinal muscles. 7. Conduction to specific muscle group,
17. If only distal muscles are abnormal, for example, infraspinatus, if clinically
examine leg muscles. indicated (compared with the opposite
side)
BRACHIAL PLEXOPATHY 8. Compare abnormal results with the op-
posite side.
Because brachial plexopathy may involve any 9. Needle examination: first dorsal inter-
of the many nerves of the upper extremity, osseous, abductor digiti minimi, exten-
evaluation of a brachial plexopathy is best sor indicis proprius, pronator teres,
modified on the basis of the deficit and pos- biceps, triceps, extensor digitorum
sible causes. communis, cervical paraspinals. Con-
A. Upper Trunk Damage sider deltoid, brachioradialis, infra-
1. Musculocutaneous/biceps nerve con- spinatus, rhomboid, serratus anterior.
duction studies, stimulating upper arm Define the upper border of the seg-
and supraclavicular or axillary/deltoid, mental abnormality.
stimulating supraclavicular
2. Consider root stimulation recording
over biceps or deltoid if these muscles PERONEAL NEUROPATHY
are clinically weak, with reduced re-
cruitment of MUPs on the needle ex- 1. Ipsilateral peroneal motor conduc-
amination, and the CMAPs recorded tion study with F waves
with supraclavicular stimulation are 2. Use slow sweep and high gain if there
normal. is a twitch with no CMAP.
3. Radial, lateral antebrachial sensory 3. Measure to initial negativity if there is
conduction. Compare with the other a positive dip at the knee and not at
side if borderline. the ankle.
4. If the above are normal, compare with 4. If the knee CMAP is larger:
the other side and perform distal mo- a. Recheck the ankle for supramaxi-
tor and sensory studies. mal stimulation (slide stimulator);
5. Needle examination: biceps, deltoid, use long-duration high-voltage stim-
brachioradialis, pronator teres, triceps, ulus, especially with low amplitude.
infraspinatus, rhomboid, cervical para- b. Recheck the knee for current
spinal. Define the lower border of the spread and tibial stimulation (watch
abnormality. twitch and move lateral).
c. If knee CMAP remains larger, c. If any muscle EMG is abnormal,

check for accessory peroneal test the most involved muscles on
anomaly. the opposite side.
5. If the only abnormality is prolonged
distal latency, check temperature.
6. If the ankle CMAP is 20% larger than Electromyography in the
that of the knee or velocity is not in Evaluation of Weakness
the normal range:
a. Stimulate below the head of the Generalized weakness is a common com-
fibula with at least a 10-cm distance plaint; most often, it is not caused by neuro-
between the knee site and the site muscular disease. The following algorithms
below the head of the fibula. can help identify the specific causes of weak-
b. If the fibular CMAP is 20% larger ness, such as myopathy, neuromuscular junc-
than that of the knee, "inch" prox- tion disease, or motor neuron disease.
imally along the nerve starting at 1. If weakness is generalized or the arms
the head of the fibula. are weaker: ulnar motor conduction
7. If no response from extensor digito- study with F waves and 2 Hz repetitive
rum brevis or if there is anterior tib- stimulation with the hand immobilized
ial muscle weakness with normal pe- 2. If the ulnar study is equivocal: median
roneal studies: motor conduction study with F waves
Peroneal conduction to anterior and 2 Hz repetitive stimulation
tibial muscle, stimulating the fibula 3. If proximal muscles are clearly weaker:
head and 10 cm proximal a. Musculocutaneous motor conduc-
8. Tibial motor conduction study with F tion study with 2 Hz repetitive stim-
waves ulation and the arm immobilized
9. Superficial peroneal sensory conduc- b. Trapezius conduction study with 2
tion study Hz repetitive stimulation and the
a. If amplitude is greater than 5 /iV, arm immobilized
record above the head of the 4. If the legs are clearly weaker: peroneal
fibula. motor conduction study with F waves
b. If no response or the velocity and peroneal study to anterior tibial
clearly is not normal above the with repetitive stimulation and the leg
fibula, record below the head of immobilized
the fibula. 5. Median sensory conduction study
c. Use needle stimulation if no re- 6. If the legs clearly are weaker, sural sen-
sponse. sory conduction study
10. Sural conduction study 7. If an abnormality is found at any point,
11. If sural study is abnormal, assess for use the appropriate algorithm, for ex-
peripheral neuropathy. ample, polyradiculopathy, myopathy,
12. If any part of the peroneal recording myasthenia gravis.
is abnormal or borderline, repeat pe- 8. Needle examination
roneal studies on the opposite leg. a. If the arms are involved, examine
13. Needle examination the first dorsal, interosseous, biceps,
a. Always test anterior tibial, per- triceps, and infraspinatus muscles.
oneus longus, medial gastrocne- b. If the legs are involved, examine the
mius, and flexor digitorum longus anterior tibial, gluteus medius, lum-
muscles. bar paraspinal, and other muscles as
b. If indicated, study the short head clinically indicated.
of the biceps femoris, extensor hal-
lucis, extensor digitorum brevis, MYOPATHY/MYOSITIS
abductor hallucis, quadriceps, glu-
teus medius, and lumbar para- Consider needle examination first, particu-
spinal muscles. larly when predominately proximal muscles
are affected and there are no associated sen- ple, brachioradialis, forearm flexor,
sory symptoms. sternocleidomastoid, cervical para-
1. For nerve conduction studies, if the spinal, or facial muscles.
lower extremities are weaker: e. Look for fibrillation potentials, my-
a. Peroneal motor conduction study otonic discharges, small MUPs, vari-
with F waves and 2 Hz repetitive ation of MUPs, or a mixture of small
stimulation; brief (10 seconds) ex- and large MUPs.
ercise if CMAP is low amplitude. Use f. If the findings are uncertain or un-
myasthenia gravis algorithm if there usual, quantify MUPs.
is a decrement.
b. Tibial motor conduction study with
F waves if peroneal conduction is
not satisfactory MYASTHENIA GRAVIS
c. Sural sensory conduction study
d. Plantar sensory conduction study if 1. The patient should not take pyri-
the patient is younger than 55 years dostigmine (Mestinon) for at least 4
e. Needle examination if all are nor- hours (preferably more) before the
mal test, if possible.
2. If there is generalized weakness or the 2. If there is generalized myasthenia or
arms are weaker: if arm or bulbar muscles are primar-
a. Ulnar motor conduction study with ily involved, perform an ulnar motor
F waves and 2 Hz repetitive stimula- conduction study on the symptomatic
tion; brief exercise if CMAP is low side:
amplitude; repetitive stimulation a. Keep the hand warm (above 32°C)
studies if there is a decrement and immobilized on a board.
b. Median motor conduction study b. Measure the amplitude, latency,
with F waves if the ulnar study is conduction velocity, and F wave.
equivocal or not satisfactory Look for repetitive CMAPs follow-
c. If a proximal myopathy or neuro- ing the main M wave with single
muscular junction abnormality is supramaximal stimuli, seen in the
possible: presence of pyridostigmine or the
(1) Musculocutaneous motor con- slow channel and acetylcholine es-
duction study with 2 Hz repeti- terase deficiency congenital myas-
tive stimulation thenic syndromes.
or c. If motor conduction is abnormal,
(2) Spinal accessory motor conduc- consider another algorithm, for ex-
tion study with 2 Hz repetitive ample, peripheral neuropathy or
stimulation ulnar neuropathy. Do repetitive stim-
d. Median sensory conduction study ulation on an uninvolved nerve.
3. Needle examination—one side only d. Repetitive stimulation at the wrist:
(same as previous EMG if done be- (1) 2 Hz, 4 shocks 3 times for a re-
fore): producible response
a. Check to ensure that the creatine ki- (2) If normal, 1 minute of exercise
nase level has been measured. (if low amplitude or decre-
b. Test moderately weak muscles, but ment, 10 seconds of exercise)
not severely weak or atrophic mus- (3) 2 Hz, 4 shocks at 5, 30, 60,120,
cles. and 180 seconds (longer if re-
c. Examine first dorsal interosseous, bi- sults are equivocal)
ceps, triceps, infraspinatus, deltoid, 3. Repetitive stimulation as above until
anterior tibial, vastus lateralis, glu- two nerves are clearly abnormal with
teus medius, and lumbar paraspinal a reproducible decrement, selecting
muscles. nerve by clinical findings and symp-
d. If weakness is focal or selective, sam- toms. Routine motor nerve conduc-
ple the involved muscles, for exam- tion studies should be performed first
on each nerve; findings should be nor- b. Superficial peroneal sensory con-

mal before proceeding with repetitive duction study if only peroneal mo-
stimulation. tor conduction study is abnormal
a. Median—thumb immobilized by 4. If the clinical features are asymmetri-
hand; repetitive stimulation at wrist cal, compare with the other leg. If not,
as above test ipsilateral arm.
b. Musculocutaneous—immobilize 5. Median sensory conduction study with
arm on arm board. proximal stimulation
c. Spinal accessory—immobilize arm 6. Ulnar motor conduction study with F
with strap. waves:
d. Axillary—immobilize arm with a. If normal, consider wrist, elbow, ax-
strap. illa, supraclavicular, and root stim-
e. Facial ulation.
f. Peroneal with anterior tibial b. Watch for dispersion.
recording—immobilize leg on leg 7. If findings in step 4 or 5 are normal
board. or borderline:
g. Femoral—technically difficult a. Median motor conduction study
4. If the leg is more involved clinically, with F waves
start with 3f above. b. Ulnar sensory conduction study
5. If a decrement is found, consider the with proximal stimulation
testing effect of edrophonium (Ten- 8. If the above studies are normal or bor-
silon) or 3,4-diaminopyridine. derline or no responses are obtained,
6. Median sensory conduction study consider:
(sural conduction study if findings are a. Musculocutaneous motor conduc-
primarily in the leg) tion study with proximal stimula-
7. Needle examination: first dorsal in- tion
terosseous, biceps, triceps, deltoid, b. Unilateral blink reflex testing
infraspinatus, sternocleidomastoid, c. Facial motor conduction studies
masseter, facial, cervical paraspinals, d. Tibial H reflex
anterior tibial, rectus femoris, ilio- 9. Consider tibial and median SEPs if
psoas, and gluteus medius muscles nerve conduction studies are all
(with isolation of single MUPs to normal.
check for variation) 10. Needle examination: First dorsal in-
8. Examine leg muscles if clinically indi- terosseous, biceps, triceps, anterior
cated. tibial, abductor hallucis, vastus later-
9. If normal, perform needle examina- alis, gluteus medius, and lumbar and
tion on other symptomatic muscles. thoracic paraspinal muscles. Others as
10. If all test findings are normal and it is needed for focal or questionable ab-
clinically indicated, perform single normality or as dictated by neurologic
fiber EMG. deficit.
MOTOR NEURON DISEASE

POLYRADICULOPATHY
The diagnosis of definite amyotrophic lat-
1. Peroneal motor conduction study eral sclerosis requires upper motor neuron
with F waves. Stimulate below the and lower motor neuron signs at three lev-
fibula if the amplitude is lower at the els of the nervous system. Other lower mo-
knee without dispersion. tor neuron syndromes may have similar
2. Tibial motor conduction study with F EMG findings, including spinal muscular
waves atrophy, residuals of poliomyelitis, hex-
3. Sural sensory conduction study: osaminidase A deficiency, multifocal motor
a. Consider plantar sensory conduc- neuropathy, pure motor inflammatory neu-
tion study if patient is younger than ropathy, demyelinating neuropathy, lead
55 years. neuropathy, porphyria, Fazio-Londe disease
(cranial), focal motor neuron disease b. Fasciculation potentials are usually

(Sobue's), arteriovenous malformation of present; if not, consider another
the cord, syrinx, and paraneoplastic syn- disease.
dromes such as lymphoma or radiation. c. First, select muscles that are most
1. Ulnar motor conduction study with likely to be abnormal, such as weak,
F waves: atrophic, or distal muscles.
a. Consider repetitive stimulation at d. Do not attempt to localize damage
2 Hz for evaluation of progression. to one nerve before looking for
b. If there is a decrement, stimulate widespread changes.
after exercise for 10 seconds. e. If clinically involved muscles show
2. If there is a decrement in the ulnar no definite abnormality, do not
nerve, repetitive stimulation at 2 Hz spend much time or effort exam-
3. Consider motor unit number esti- ining many normal muscles in the
mate. same distribution.
4. If median nerve involvement is sus- f. Examine some muscles in each
pected, ulnar nerve conduction is ab- limb even if there is no clinical ab-
normal, or legs are uninvolved: normality. Paraspinal muscles, es-
a. Median motor conduction study pecially thoracic ones, should also
with F waves be examined.
b. Consider motor unit number esti- g. Look for MUP variation.
mate. h. Consider needle examination of
5. If the legs are involved more than the diaphragm.
arms or if a motor polyradiculopathy 12. If needle EMG shows clear abnormal-
is being considered: ity in only one or two extremities, con-
a. Peroneal motor conduction study sider single fiber EMG for jitter, using
with F waves standard concentric electrode (500
b. Consider motor unit number esti- Hz low-frequency filter; 0.5 ms/cm
mate. sweep) to look for MUP variation.
6. If peroneal motor conduction is ab-
normal or if a leg F wave is needed:
tibial motor conduction study with F POSTPOLIO SYNDROME
waves
7. If legs are tested, consider sural sen- Postpolio syndrome is a clinical diagnosis
sory conduction study, medial plan- that cannot be confirmed with EMG or nerve
tar/ankle, or superficial peroneal if conduction studies. The purpose of these
the motor response is lost. tests in this syndrome is to look for other su-
8. Median sensory conduction study perimposed, new diseases in the presence of
9. If median sensory conduction is ab- known residuals of the old poliomyelitis.
normal or ulnar motor conduction is 1. Nerve conduction studies:
low amplitude: ulnar sensory conduc- a. These studies are used to search for
tion study, same method as median carpal tunnel syndrome, ulnar neu-
sensory study ropathy, peroneal neuropathy, pe-
10. Consider phrenic nerve conduction ripheral neuropathy, or any other
study if there are respiratory symp- neuropathy that may be suggested
toms. clinically. If the symptoms are mild
11. Needle examination must be based on and diffuse, test the most clinically
clinical findings in conjunction with involved limb. If the symptoms are
the following considerations: severe and diffuse, test the least in-
a. Two to three muscles innervated by volved limb.
different nerves and roots must b. Examine the extremity with great-
have fibrillation potentials and est number of new symptoms.
MUP changes at two distinct levels c. Motor conduction study with F
to confirm diagnosis (the four lev- waves (note size of F waves)
els are the brain stem, cervical, tho- d. Sensory conduction study with ve-
racic, and lumbar spinal cord). locities
e. If findings are abnormal or bor- 4. Bilateral blink reflex testing

derline, compare with the opposite 5. Blink reflex testing with additional
extremity. orbicularis oris recording if synkinesis
2. Needle EMG: is suspected (hemifacial spasm, old fa-
a. This test is used to search for cial neuropathy)
radiculopathy or myopathy. 6. Limb nerve conduction (polyradicu-
b. Test the most symptomatic ex- lopathy protocol) if two cranial nerves
tremity. are slowed, if diere is a decrement, or
(1) Least atrophic muscles: if a generalized disorder is suspected
Distribution of major roots 7. Jaw jerk if abnormalities suggest mul-
Proximal and distal tiple cranial nerve disease
(2) Markedly atrophic muscles are 8. Lateral spread responses if hemifacial
always abnormal and will not spasm is suspected
give useful information. 9. Needle examination: examine ipsilat-
c. Always compare opposite, less symp- eral orbicularis oris, orbicularis oculi,
tomatic extremity: mentalis, and frontalis muscles.
(1) Muscles with comparable at- 10. If findings on ipsilateral needle ex-
rophy amination are abnormal, check mas-
(2) Muscles in same distribution seter, sternocleidomastoid muscles,
d. Large MUPs are of no significance, and a contralateral facial muscle.
but MUP variation may be. 11. If MUP variation or a decrement is
e. Fibrillation potentials are signifi- found, use the protocol for myasthe-
cant only if they are found in dis- nia gravis.
tribution of—
(1) Mononeuropathy or radicu- MYOTONIC SYNDROMES
lopathy:
Limited to distribution of one Many different central and peripheral dis-
nerve or root orders may present with muscle stiffness.
Present proximally and distally Central processes such as rigidity and spas-
Present in muscles without ticity are best assessed clinically, because
atrophy EMG test results are normal. The evaluation
(2) Peripheral neuropathy or poly- of peripheral disorders is based on the char-
radiculopathy: acter of the symptoms. If the major com-
Present bilaterally plaint is episodic weakness, with or without
Present in muscles without at- muscle stiffness, the patient should be tested
rophy for periodic paralysis. If the major complaint
f. Decreased insertional activity with is episodic myalgia without true muscle stiff-
increased resistance to needle ness, the patient should be tested for a my-
movement occurs with muscle fi- opathy. A muscle biopsy or ischemic fore-
brosis in atrophic muscles arm exercise/lactate test (or both) is helpful
if myalgia or contractures develop with ex-
FACIAL WEAKNESS ercise. If the major complaint is muscle stiff-
ness, consider doing die needle examination
1. Ipsilateral facial motor conduction first to confirm the presence and nature of
study: spontaneous activity. If myotonic discharges
a. 2-Hz repetitive stimulation if a de- are found, nerve conduction studies may
fect of neuromuscular junction is help define their source.
suspected 1. Ulnar motor conduction study with F
b. Reorient stimulating electrode if waves and the hand immobilized:
the initial positive dip occurs be- a. 2 Hz repetitive stimulation before
cause of masseter activation. and at 3,15, and 30 seconds after 10
2. Opposite facial conduction study if seconds of exercise
comparison is needed b. Interpretation—after 10 seconds of
3. Record from another facial muscle if exercise, an immediate decrease in
weakness is focal. CMAP amplitude and repair of a
decrement (if present) are specific Also, myotonic discharges, on aver-

for myotonic syndromes except age, fire at slower rates in para-
proximal myotonic myopathy, which myotonia.
does not show these changes. De- 7. Conduct the following special needle
creased CMAP amplitude after 10 examination on a clinically involved
seconds of exercise will return to muscle with myotonic discharges:
normal in less than 2 minutes (usu- a. After 20 repeated forceful contrac-
ally 30 to 40 seconds) in all forms of tions, examine for changes in myo-
myotonia, except for paramyotonia, tonia, loss of MUP, or postexercise
in which recovery may take up to 90 fibrillation.
minutes. b. After exercise in paramyotonia,
2. Consider additional testing after cool- myotonia increases and the MUPs
ing in ice water bath for 5—10 minutes, drop out. In all other forms of pri-
exercise after cooling, and after re- mary myotonia, myotonia subsides
warming to test for paramyotonia. and motor unit potentials do not
3. Consider median or peroneal motor change. Postexercise fibrillation may
conduction studies with F waves with occur in paramyotonia.
immobilization. c. After cooling a muscle to 20°C (in-
4. Repetitive stimulation and exercise as tramuscular temperature measured
for ulnar nerve with needle thermistor), examine
5. Median or sural sensory conduction for spontaneous discharges, change
study in myotonia, loss of motor unit po-
6. Needle examination—symptomatic mus- tential, and the effect of 20 strong
cles are more likely to show abnormal- contractions. Observe EMG activity
ity. The following abnormalities should as the muscle warms up:
be looked for: (1) After cooling, myotonia increases
a. Fibrillation potentials—nonspecific, except in paramyotonia, in
found in muscle membrane disor- which the muscle becomes elec-
ders (channelopathies) and muscle trically silent. Spontaneous dis-
fiber necrosis, splitting, or vacuolar charges indistinguishable from
change fibrillation potentials occur only
b. Complex repetitive discharges— in paramyotonia, especially as
nonspecific, found in myopathies, the temperature decreases from
neuropathic disease, and in some 32°C to 28°C (intramuscular
normal, older persons temperature).
c. Fasciculation potentials and cramp (2) If poor recruitment develops in
discharges—nonspecific, found in cooled muscle, exercise will
nerve hyperexcitability, cramp-fascic- rapidly produce normal pat-
ulation syndrome, neuropathy, or terns except in paramyotonia,
motor neuron disease in which all MUPs are lost.
d. Myokymic discharges—found in (3) Any changes seen in cold mus-
nerve hyperexcitability with radia- cle are rapidly reversed by
tion damage, multiple sclerosis, warming except in paramyoto-
some brain stem tumors, and com- nia, in which abnormalities re-
pression neuropathy cover slowly over several hours.
e. Motor unit potentials are of short 8. If the diagnosis is still unclear, addi-
duration in rested muscle in myo- tional nerve conduction studies should
tonic dystrophy and in some cases of be considered:
recessive myotonia congenita. On a. Rapid repetitive stimulation (greater
average, myotonic discharges occur than 25 Hz) of ulnar nerve with
in shorter bursts, are more variable hand restrained
in rate and amplitude, and fire at (1) Rapid repetitive stimulation pro-
higher rates in myotonia congenita duces a waxing-and-waning pat-
than in other forms of myotonia. tern that is most prominent in
myotonia congenita, particu- g. Gain setting may be wrong. Check

larly the autosomal recessive amplification.
form. h. Sweep speed may be wrong, with
(2) More than 30 seconds of stimu- the response off the oscilloscope
lation may be needed before screen. Check sweep speed.
this pattern appears. i. Filter setting may be incorrect. Check
(3) The waxing-and-waning effect filters.
disappears in exercised muscle. j. There may be stimulator malfunc-
b. Cool the contralateral hypothenar tion. Test on self.
muscle to 28°C and apply 2 Hz stim- k. The amplifier may be turned off.
ulation to the ulnar nerve with the 1. Innervation may be anomalous.
hand restrained on a board to check Stimulate other nerves as appropri-
for occurrence or enhancement of ate.
a decrement. Occurrence or en- m. It may be a manifestation of disease.
hancement of a decrement at 2 Hz 2. If the CMAP amplitude differs between
stimulation during cooling is char- proximal and distal stimulations
acteristic of paramyotonia. a. At the site of the lower amplitude
c. If periodic paralysis is suspected, response—
consider prolonged exercise, potas- (1) The stimulating electrode may
sium challenge, and intra-arterial be off the nerve. Slide the elec-
epinephrine tests. trode medially and laterally
without voltage change.
(2) The nerve may be deeper (see
Unexpected Findings on Nerve step Ic above).
Conduction Studies: (3) Stimulation may not be supra-
Cause and Action maximal for other reasons listed
in step 1 above.
COMMON PROBLEMS b. At the site of higher amplitude re-
sponse—There may be excessive
1. If CMAP or sensory NAP is low ampli- stimulation, with current spread to
tude or absent: activate other nerves. Slide the stim-
a. Preamplifier input may not be ulating electrode toward the other
turned on. nerve while watching the configura-
b. Stimulator may not be over the tion of the evoked response and the
nerve: slide the stimulating elec- muscle twitch.
trode without changing the stimu- c. Dispersion of the action potential:
lating voltage. (1) It may be a normal variant; for
c. The nerve may be deep or the pa- instance, tibial nerve to abduc-
tient obese: firmly push in stimulat- tor hallucis brevis muscle re-
ing electrode, separate cathode and sponse.
anode, increase voltage and dura- (2) It may be disease with late com-
tion to maximum, try monopolar ponents or increasing duration.
stimulating electrode, or consider d. Anomalous innervation: stimulate
using needle stimulation. other nerves as above.
d. Anode and cathode may be re- e. Diseases with conduction block
versed, causing anodal block, or 3. If an initial positivity precedes a nega-
they may be too close with current tive M wave of the CMAP:
shunt. Check cathode location. a. The active recording electrode may
e. Recording electrode may be placed not be over the end plate region.
incorrectly. Check the positions of Check the position of the recording
the active and reference electrodes. electrode and slide it while stimu-
f. Wrong electrodes may be plugged lating.
into the recording system. Check the b. The wrong nerve may be stimulated.
input plugs. Check twitch and configuration for
current spread. Check the location 5. High threshold (excessive voltage re-
of the stimulating electrode. quired to achieve supramaximal stimu-
c. Active and reference input may be lation):
reversed. Check the input of the a. The stimulating electrode may not
electrodes. be over the nerve. Slide the stimu-
d. There may be a volume-conducted lator without changing the voltage.
response from a distant muscle, es- b. The nerve may be deep; push in the
pecially at high gain (for example, stimulating electrode firmly, spread
peroneal with knee stimulation). the cathode and anode, or consider
Check twitch and change the stim- monopolar stimulation.
ulation site. c. Intervening tissue may be excessive
e. Innervation may be anomalous. (for example, scar or fat).
Check change in the response with d. Contact between the skin and elec-
stimulation at different sites on dif- trode may be poor. Abrade the skin
ferent nerves. or add electrode paste.
f. The reference electrode may not be e. Amplification may be incorrect or
in the correct location or may not be the amplifier may not be turned on.
in contact with the skin. Check the f. The anode and cathode may be too
electrode. close, with current shunt. Separate
g. There may be disease with atrophy. them.
4. If there is excessive stimulus artifact or g. There may be disease with small, re-
baseline shift (that is, active and refer- generating, or hypertrophic axons.
ence electrodes are not isopotential): 6. If distal latency is long:
a. The contact of the recording elec- a. The extremity may be cold. Make
trode with the skin may be poor. sure the thermistor is working cor-
Clean and abrade the skin and reap- rectly.
ply electrodes. b. Distance may be wrong. Check
b. There may be a current bridge be- against normal distance.
tween the stimulating and recording c. The cathode and anode may be re-
or ground electrodes. Check for versed. Check cathode location.
smeared paste and clean the skin be- d. Stimulus delay set may be wrong.
tween the electrodes. Check delay.
c. The active, reference, or ground e. Stimulation may be submaximal.
electrode may be broken, not in con- f. Gain may be too low.
tact, or not plugged in. Check elec- g. There may be local or diffuse dis-
trodes, especially if 60 cycle artifact ease.
is present. 7. If there are no recognizable F waves:
d. The stimulating electrode may be a. The gain may be too low. Increase
too close to the recording electrode. amplification.
Check their relative position and b. Sweep speed may be too fast. Try a
distance. slower sweep speed.
e. Stimulating electrodes may be ori- c. There may be poor relaxation. Ma-
ented incorrectly. Rotate the stim- nipulate the extremity and stimulate
ulating electrodes if the artifact is only when the audio EMG recording
not bidirectional. The skin must be is quiet.
cleaned if the artifact is bidirec- d. Too few stimuli may have been
tional. given. Increase the number.
f. There may be equipment malfunc- e. Voltage may not be supramaximal.
tion. Check the effect of voltage Check for supramaximal voltage
change. with the cathode proximal.
g. The stimulus may be excessive, es- f. There may be anodal block, with the
pecially duration. cathode distal. Place the cathode
h. There may be amplifier/preampli- proximal.
fier overload. Reduce stimulus in- g. F waves may be confused with an
tensity. axon reflex or dispersed M wave.
Check persistence and latency ous muscle with stimulation of the

change with the site of stimulation. median and ulnar nerves at the el-
h. F waves may be lost in the M wave bow and wrist.
with proximal stimulation. Stimu- 3. If the extensor digitorum brevis re-
late distally. sponse changes configuration or has a
i. It may be a normal variant, especially higher amplitude with knee than with
with peroneal stimulation. ankle peroneal nerve stimulation:
j. It may be caused by disease. a. Check for supramaximal stimula-
tion at the ankle.
ANOMALOUS INNERVATION b. Check for current spread to the tib-
ial nerve at the knee. Observe for
1. If the hypothenar CMAP with ulnar plantar flexion caused by tibial
nerve stimulation differs in configura- nerve stimulation and move die
tion between the wrist and the elbow knee stimulating electrode laterally
or if the elbow amplitude/area is less and distally if present.
than 80% of the wrist amplitude: c. Stimulate behind the lateral malleo-
a. Check for overstimulation at the wrist lus while recording extensor digi-
or understimulation at the elbow. torum brevis muscle. A response
b. Record hypothenar CMAP with me- confirms the presence of a deep ac-
dian stimulation at the wrist and el- cessory branch of the superficial pe-
bow. A higher proximal amplitude roneal nerve. Make sure that you are
or configuration difference between not stimulating the tibial nerve at
the elbow and wrist confirms me- the ankle.
dian/ulnar nerve anastomosis.
c. If these steps give no difference,
stimulate the ulnar nerve at the Interpretation of Nerve
wrist, below the elbow, at the elbow, Conduction Studies
and in the upper arm to check for
ulnar conduction block (ulnar neu- DEMYELINATING NEUROPATHY
ropathy protocol).
d. If further characterization is de- 1. Conduction velocity less than 70% of
sired, record first dorsal interosse- normal with normal amplitude
ous muscle with stimulation of the 2. Conduction velocity less than 50% of
median and ulnar nerves at the el- normal with amplitude larger than
bow and wrist or record with a nee- 50% of normal
dle electrode to localize the site of 3. Distal latencies greater than 150% of
anomaly. normal
2. If the thenar CMAP with median nerve 4. Compound muscle action potential dis-
stimulation differs in configuration be- persion or conduction block
tween the elbow and wrist or if the el- 5. F waves are prolonged out of propor-
bow amplitude/area exceeds the wrist tion to peripheral slowing.
amplitude: 6. Blink reflexes are prolonged (Rl > 16
a. Check for overstimulation at the el- ms).
bow or understimulation at the wrist. 7. Note: similar changes can occur after
b. Record the thenar CMAP with ulnar regeneration of a severe nerve injury.
stimulation at the wrist and elbow.
Higher amplitude at the wrist or AXONAL NEUROPATHY
configuration difference between
the wrist and elbow responses con- 1. Amplitude less than 70% of normal
firms median/ulnar nerve anasto- with conduction velocity greater than
mosis. 70% of lower limit of normal
c. An initial thenar positivity is the 2. Amplitude less than 50% with any de-
usual finding with ulnar stimulation. gree of conduction velocity slowing
d. If further characterization is de- 3. Low amplitude with distal latencies less
sired, record first dorsal interosse- than 130% of normal
4. Fibrillation potentials and large MUPs SUMMARY

5. No dispersion or block
6. Normal F waves and blink reflexes The major value and primary application of
7. Absence of sensory potentials with nor- clinical neurophysiology is in the assessment
mal motor conduction and fibrillation and characterization of neurologic disease.
potentials Selection of appropriate studies for the
problem of an individual patient requires a
FOCAL CONDUCTION BLOCK careful clinical evaluation to determine pos-
Definite Block sible causes of the patient's symptoms. The
1. Compound muscle action potential nature of the symptoms and the conclusions
area difference proximal to distal of the clinical evaluation are the best guides
greater than 50%, regardless of dis- to appropriate use of clinical neurophysio-
tance or CMAP duration (except for logic testing.
tibial nerve study) The approach to testing can be assisted by
2. Compound muscle action potential deciding which structures are likely to be in-
area difference proximal to distal volved. For example, motor and sensory
greater than 30%, less than 15% dif- symptoms are best assessed using the differ-
ference in CMAP duration, regardless ent methods of motor and sensory nerve
of distance conduction studies. Electroencephalography,
3. Compound muscle action potential autonomic function testing, and polysomnog-
area difference proximal to distal raphy provide distinct assessment of distur-
greater than 20% over distances of 10 bances of consciousness, cognition, visceral
cm or less, regardless of CMAP dura- function, and sleep. The level of the nervous
tion system that is likely to be involved by the dis-
4. Compound muscle action potential ease process can also guide selection of the
area difference proximal to distal neurophysiologic methods that will be most
greater than 10% over a distance of helpful in sorting out the clinical problem.
2 cm Disorders of the cerebral hemisphere are
Possible Block best characterized electrically by electroen-
Compound muscle action potential area dif- cephalography, somatosensory evoked po-
ference proximal to distal greater than 30%, tentials, polysomnography, and movement
regardless of distance or CMAP duration recordings. Lesions in the posterior fossa
may benefit from the addition of cranial con-
FOCAL SLOWING OF CONDUCTION duction studies and brain stem auditory
VELOCITY evoked potentials. Spinal cord disease pro-
duces alterations in EMG, nerve conduction
1. Conduction velocity slowing greater studies, and somatosensory evoked poten-
than 10 m/second over a 10 cm seg- tials. Peripheral diseases show changes on
ment (faster conduction in the seg- nerve conduction studies, EMG, and auto-
ments of nerve both proximal and dis- nomic function testing.
tal to the slowed segment) The multiplicity of different neurophysio-
2. Conduction delay longer than 0.4 ms logic measures that can be applied in pe-
over a 1 cm segment of nerve ripheral disorders is sometimes assisted by
3. Distal latency prolonged by more than applying guideline protocols based on the
130% of normal, with other distal la- patient's clinical findings and what is found
tencies normal during testing. Although a clinical neuro-
4. Distal latency prolonged by more than physiologic assessment rarely provides evi-
0.4 ms when comparing nerves in the dence for a specific diagnosis, it can provide
same limb over the same distance valuable information about the severity, pro-
5. Temperature higher than 32°C gression, and prognosis of the disease.
GLOSSARY OF
ELECTROPHYSIOLOGIC TERMS
A Wave: A compound muscle action potential size and configuration. See also compound
that follows the M wave, evoked consis- action potential, motor unit action potential
tently from a muscle by submarimal elec- Activation: (1) In physiology, a general term
tric stimuli and frequently abolished by for the initiation of a process. (2) The pro-
supramaximal stimuli. Its amplitude is simi- cess of motor unit action potentialfiring. The
lar to that of an F wave, but the latency force of muscle contraction is determined
is more constant. Usually occurs before by the number of motor units and their
the F wave, but may occur afterwards. firing rate.
Thought to be due to extra discharges in Activation Procedure: A technique used to
the nerve, ephapses, or axonal branching. detect defects of neuromuscular transmis-
This term is preferred over axon reflex, sion during repetitive nerve stimulation test-
axon wave, or axon response. Compare with ing. Most commonly a sustained voluntary
the F wave. contraction is performed to elicit facilitation
Absolute Refractory Period: See refractory or postactivation depression. See also tetanic
period. contraction.
Accommodation: In neuronal physiology, a Active Electrode: Synonymous with exploring
rise in the threshold transmembrane depo- electrode. See recording electrode.
larization required to initiate a spike, when Acute Inflammatory Neuropathy: An acute,
depolarization is slow or a subthreshold monophasic polynewopathy. Characterized
depolarization is maintained. In the older by a time course of progression to maxi-
literature, the observation that the final mum deficit within 4 weeks of onset of
intensity of current applied in a slowly symptoms. Most common clinical presen-
rising fashion to stimulate a nerve was tation is an ascending sensory-motor neu-
greater than the intensity of a pulse of cur- ropathy. Electrodiagnostic studies most
rent required to stimulate the same nerve. commonly reveal evidence for demyelina-
The latter may largely be an artifact of the tvm, but axonal degeneration also occurs.
nerve sheath and bears little relation to Distinguish from chronic inflammatory de-
true accommodation as measured intra- myelinating polyradicukmeuropathy (GDP).
cellularly. See also Guttlain-Bam syndrome.
Adaptation: A decline in thefrequencyof the
Accommodation Curve: See strength-duration spike discharge as typically recorded from
curve. sensory axons in response to a maintained
Acoustic Myography: The recording and stimulus,
analysis of sounds produced by contract- ADEMG: Abbreviation for automatic decom-
ing muscle. The muscle contraction may be position electromyography.
produced by stimulation of the nerve sup- AEP: Abbreviation for auditory-evoked poten-
ply to the muscle or by volitional activa- tial
tion of the muscle. Afterdischarge: (1) The continuation of ac-
Action Potential (AP): The brief regeneration potentials in a neuron, axon or muscle
tive electric potential that propagates along fiber following the termination of an ap-
a single axon or muscle fiber membrane. plied stimulus. (2) The continuation of fir-
An all-or-none phenomenon; whenever ing of muscle action potentials after cessation
the stimulus is at or above threshold, the ac- of voluntary activation, for example in
tion potential generated has a constant myotonia.
*Reproduced by permission of the American Association of Electrodiagnostic Medicine,
599
600 Glossary of Electrophysiologic Terms
Afterpotential: The membrane potential be- other than the ones of interest. The stim-
tween the end of the spike and the time ulus artifact (or shock artifact) represents
when the membrane potential is restored cutaneous spread of stimulating current
to its resting value. The membrane during to the recording electrode and the delay in re-
this period may be depolarized or hyper- turn to baseline which is dependent on the
polarized at different times. ability of filters to respond to high voltage.
Akinesia: Lack or marked delay of intended Stimulus artifacts may precede or overlap
movement, often observed in patients with the activity of interest. Movement artifact
Parkinson's disease. Often used synony- refers to a change in the recorded activity
mously with bradykinesia. caused by movement of the recording
Amplitude: With reference to an action po- electrodes.
tential, the maximum voltage difference be- Asterixis: A quick involuntary movement
tween two points, usually baseline-to-peak caused by a brief lapse in tonic muscle ac-
or peak-to-peak. By convention, the am- tivation. It can be appreciated only during
plitude of potentials which have an initial voluntary movement. Is usually irregular,
negative deflection from the baseline, but can be rhythmic and confused with ac-
such as the compound muscle action potential tion tremor.
and the antidromic sensory nerve action po- Ataxia: Clumsiness of movement. Specific
tential are measured from baseline to the features include dysmetria (incorrect dis-
most negative peak. In contrast, the am- tance moved) and dysdiadochokinesis (ir-
plitude of a compound sensory nerve action regularity of attempted rhythmic move-
potential, motor unit potential, fibrillation po- ments) . Most commonly due to a disorder
tential, positive sharp wave, fasciculation po- of the cerebellum or proprioceptive sen-
tential, and most other action potentials is sory system. Referred to, respectively, as
measured from the most positive peak to cerebellar ataxia or sensory ataxia.
the most negative peak. Auditory Evoked Potential (AEP): Electric
Amplitude Decay: The percent change in waveforms of biologic origin elicited in re-
the amplitude of the M wave or the com- sponse to sound stimuli. Classified by their
pound sensory nerve action potential between latency as short-latency brain stem auditory
two different stimulation points along evoked potential (BAEP) with a latency of up
the nerve. Decay = 100 • (amplitudeaistai ~ to 10 ms, middle-latency with a latency of
amplitudeproximai) / amplitudedistai • Useful 10 to 50 ms, and long-latency with a la-
in the evaluation of conduction block. Ab- tency of over 50 ms. See brain stem audi-
normal decay without increased temporal tory evoked potential.
dispersion may indicate a conduction Automatic Decomposition EMG (ADEMG):
block. Computerized method for extracting in-
Anodal Block: A local block of nerve con- dividual motor unit action potentials from an
duction caused by membrane hyperpolar- interference pattern.
ization under a stimulating anode. Does not Averager: See signal averager.
occur in routine clinical studies, since it is Averaging: A method for extracting time-
possible for the anode to routinely result locked potentials from random back-
in nerve depolarization if sufficient current ground noise by sequentially adding traces
intensities are used. and dividing by the total number of traces.
Anode: The positive terminal of an electric Axon Reflex: Use of term discouraged as it
current source. See stimulating electrode. is incorrect. No re/lex is thought to be in-
Antidromic: Propagation of a nerve impulse volved. See preferred term, A wave.
in the direction opposite to physiologic Axon Response: See preferred term, A wave.
conduction; for example, conduction Axon Wave: See A wave.
along motor nerve fibers away from the Axonal Degeneration: Degeneration of the
muscle and conduction along sensory segment of a nerve distal to the cell body
fibers away from the spinal cord. Contrast with preferential distal pathology.
with orthodromic. Axonotmesis: Nerve injury characterized by
AP: Abbreviation for action potential. axon and myelin sheath disruption with
Artifact (also Artefact): A voltage change gen- supporting connective tissue preservation,
erated by a biologic or nonbiologic source resulting in axonal degeneration distal to
Glossary of Electrophysiologic Terms 601
the injury site. Compare neurapraxia, neu- Bizarre Repetitive Potential: See preferred
rotmesis. term, complex repetitive discharge.
Backaveraging: Averaging a signal which oc- Blink Reflex: See blink responses.
curs in a time epoch preceding a trigger- Blink Responses: Compound muscle action po-
ing event. Often used to extract a time- tentials evoked.from orbicularis oculi mus-
locked EEG signal preceding voluntary or cles as a result of brief electric or me-
involuntary movement, usually triggered chanical stimuli applied to the cutaneous
by the onset of the EMG activity of area innervated by the supraorbital (or
the movement. An example is the less commonly, the infraorbital) branch of
Bereitschaftspotential. the trigeminal nerve. Typically, there is an
Backfiring: Discharge of an antidromically ac- early compound muscle action potential
tivated motor neuron. (Rl wave) ipsilateral to the stimulation site
BAEP: Abbreviation for brain stem auditory with a latency of about 10 ms and a bilat-
evoked potential. eral late compound muscle action poten-
BAER: Abbreviation for brain stem auditory tial (R2 wave) with a latency of approxi-
evoked response. See preferred term, brain mately 30 ms. Generally, only the R2 wave
stem auditory evoked potential. is associated with a visible contraction of the
Baseline: (1) The potential recorded from a muscle. The configuration, amplitude, du-
biologic system while the system is at rest. ration, and latency of the two components,
(2) A flat trace on the recording instru- along with the sites of recording and stim-
ment; an equivalent term, isoekctric line, ulation, should be specified. The Rl and
may be used. R2 waves are oligosynaptic and polysnap-
Benign Fasciculation Potential: A firing pat- tic brain stem reflexes, respectively. To-
tern of fasciculation potentials occurring in gether they are called the blink reflex. The
association with a clinical syndrome offas- afferent arc is provided by the sensory
ciculations in an individual with a nonpro- branches of the trigeminal nerve and the
gressive neuromuscular disorder. Use of efferent arc is provided by facial nerve mo-
term discouraged. tor fibers.
BER: Abbreviation for brain stem auditory Blocking: Term used in single fiber electromyo-
evoked responses. See preferred term, brain graphy to describe dropout of one or more
stem auditory evoked potentials. components of the potential during se-
Bereitschaftspotential (BP): A component quential firings. If more than one com-
of the movement-related cortical potential. The ponent drops out simultaneously it is de-
slowly rising negativity in the EEG prescribed as concomitant blocking. Usually
ceding voluntary movement. The German seen when jitter values exceed 80 to 100
term means "readiness potential." Has two /AS. A sign of abnormal neuromuscular
phases called BP1 and BP2 or BP and NS' transmission, which may be due to pri-
(negative slope). See backaveraging. mary neuromuscular transmission disorders,
Biphasic Action Potential: An action potential such as myasthenia gravis and other myas-
with one baseline crossing, producing two thenic syndromes. Also seen as a result of
phases. degeneration and reinnervation in neu-
Biphasic End-Plate Activity: See end-^late ac- ropathies or myopathies. Concomitant block-
tivity (biphasic). ing may be generated by a split muscle
Bipolar Needle Electrode: Recording electrode fiber or failure of conduction at an axon
that measures voltage between two insu- branch serving several muscle fibers.
lated wires cemented side-by-side in a steel BP: Abbreviation for Bereitschaftspotential.
cannula. The bare tips of the electrodes Brachial Plexus: An anatomical structure
are flush with the level of the cannula which is formed by the spinal roots from
which may serve as a ground. C5 to Tl, traverses the shoulder region,
Bipolar Stimulating Electrode: See stimulat- and culminates in the named peripheral
ing electrode. nerves in the arm. It is composed of roots,
Bizarre High-Frequency Discharge: See pre- trunks, divisions, cords, and terminal
ferred term, complex repetitive discharge. nerves.
Bizarre Repetitive Discharge: See preferred Bradykinesia: Slowness of movement, often
term, complex repetitive discharge. observed in patients with Parkinson's
disease. Often used synonymously with flexes and the control of movement by the
akinesia. spinal cord and brain. See electrodiagnosis.
Brain Stem Auditory Evoked Potential Central Motor Conduction: The time taken
(BAEP): Electric waveforms of biologic ori- for conduction of action potentials in the
gin elicited in response to sound stimuli. central nervous system from motor cortex
Normally consists of a sequence of up to to alpha motoneurons in the spinal cord
seven waves, designated I to VII, which oc- or brain stem. Calculated from the laten-
cur during the first 10 ms after the onset cies of the motor evoked potentials produced
of the stimulus and have positive polarity by transcranial magnetic stimulation or trans-
at the vertex of the head. cranial electrical stimulation, subtracting the
Brain Stem Auditory Evoked Response time for peripheral conduction.
(BAER, BER): See preferred term, brain Chorea: Clinical term used to describe ir-
stem auditory evoked potentials. regular, random, brief, abrupt, involun-
BSAP: Abbreviation for brief, small, abun- tary movements of the head or limbs due
dant potentials (see BSAPP). Use of term to a disorder of the basal ganglia. Most
is discouraged. commonly observed in patients with Hunt-
BSAPP: Abbreviation for brief, small, abun- ington's disease and Sydenham's chorea.
dant, polyphasic potentials. Used to de- Chronaxie (also Choronaxy): See strength-
scribe a recruitment pattern of brief duration, duration curve.
small amplitude, overly abundant, polypha- Chronic Inflammatory Demyelinating Poly-
sic motor unit action potentials, with respect radiculoneuropathy (CIDP): A polyneuro-
to the amount of force generated; usually pathy or polyradiculoneuropathy character-
a minimal contraction. Use of term dis- ized by generalized demyelination of the
couraged. Quantitative measurements of peripheral nervous system. In most cases
motor unit action potential duration, am- there is also a component of axonal de-
plitude, numbers of phases, and recruitment generation. Some cases are associated with
frequency are preferred. See motor unit ac- a monoclonal gammopathy of undeter-
tion potential. mined significance (MGUS). Distinguish
C Reflex: An abnormal reflex response repre- from acute inflammatory neuropathy.
senting the electrophysiologic correlate of Clinical Electromyography: Term used com-
sensory evoked myoclonus. The term "C" monly to describe the scientific methods
was chosen to indicate that the reflex of recording and analysis of biologic elec-
might be mediated in the cerebral cortex. trical potentials from human peripheral
This is sometimes, but not always, true. nerve and muscle. See preferred term, elec-
c/s (also cps): Abbreviation for cycles per sec- trodiagnostic medicine.
ond. See preferred term, Hertz (Hz). CMAP: Abbreviation for compound muscle ac-
Carpal Tunnel Syndrome: A mononeuropathy tion potential.
affecting the median nerve at the wrist. As Coaxial Needle Electrode: See synonym, con-
the nerve passes through the carpal tun- centric needle electrode.
nel, a space bounded dorsally by the bones Collision: When used with reference to nerve
of the wrist, laterally by the forearm flexor conduction studies, the interaction of two
tendons, and volarly by the transverse action potentials propagated toward each
carpal ligament, it is subject to compres- other from opposite directions on the
sion by any of these structures. Repetitive same nerve fiber so that the refractory peri-
hand and wrist movement is thought to ods of the two potentials prevent propa-
contribute to the compression. gation past each other.
Cathode: The negative terminal of an elec- Complex Motor Unit Action Potential: A mo-
tric current source. See stimulating elec- tor unit action potential that is polyphasic or
trode. serrated. See preferred terms, polyphasic
Center Frequency: The mean or median fre- action potential or serrated action potential.
quency of a waveform decomposed by fre- Complex Repetitive Discharge: A type of
quency analysis. Employed in the study of spontaneous activity. Consists of a regularly
muscle fatigue. repeating series of complex polyphasic or
Central Electromyography: Use of electrodi- serrated potentials that begin abruptly af-
agnostic recording techniques to study re- ter needle electrode movement or sponta-
neously. The potentials have a uniform produced by stimulation of the nerve di-
shape, amplitude, and discharge frequency recdy or indirecdy. Details of the method
ranging from 5 to 100 Hz. The discharge of stimulation and recording should be
typically terminates abruptiy. May be seen specified, togedier with the fiber type (sen-
in both myopathic and neurogenic disor- sory, motor, or mixed nerve).
ders, usually chronic. Thought to be due Compound Sensory Nerve Action Potential
to ephaptic excitation of adjacent muscle (Compound SNAP): A compound nerve ac-
fibers in a cyclic fashion. This term is pre- tion potential recorded from the afferent
ferred to bizarre high frequency discharge, fibers of a sensory nerve, a sensory branch
bizarre repetitive discharge, bizarre repetitive po- of a mixed nerve or in response to stimula-
tential, pseudomyotonic discharge, and syn- tion of a sensory nerve or a dorsal nerve
chronized fibrillation. See also ephapse and root. May also be elicited when an ade-
ephaptic transmission. quate stimulus is applied synchronously to
Compound Action Potential: A potential or sensory receptors. The amplitude, latency,
waveform resulting from the summation of duration, and configuration should be
multiple individual axon or musclefiberac- noted. Generally, the amplitude is mea-
tion potentials. See compound mixed nerve ac- sured as the maximum peak-to-peak volt-
tion potential, compound motor nerve action age when there is an initial positive de-
potential, compound nerve action potential, flection or from baseline-to-peak when
compound sensory nerve action potential, and there is an initial negative deflection. The
compound muscle action potential latency is measured as either the time to
Compound Mixed Nerve Action Potential: A the initial deflection or the negative peak,
compound nerve action potential recorded and the duration as the interval from the
from a mixed nerve when an electric stimu- first deflection of the waveform from the
lus is applied to a segment of the nerve baseline to its final return to the baseline.
that contains both afferent and efferent Also referred to by the less preferred
fibers. The amplitude, latency, duration, and terms sensory response, sensory potential, or
phases should be noted. SNAP.
Compound Motor Nerve Action Potential Concentric Needle Electrode: Recording elec-
(Compound Motor NAP): A compound trode that measures an electric potential dif-
nerve action potential recorded from effer- ference between a centrally insulated wire
ent fibers of a motor nerve or a motor and the cannula of the needle through
branch of a mixed nerve. Elicited by stimu- which it runs.
lation of a motor nerve, a motor branch Conditioning Stimulus: See paired stimuli.
of a mixed nerve, or a ventral nerve root. Conduction Block: Failure of an action po-
The amplitude, latency, duration, and num- tential to propagate past a particular point
ber of phases should be noted. Distinguish in the nervous system whereas conduction
from compound muscle action potential is possible below the point of the block.
Compound Muscle Action Potential (CMAP): Documented by demonstration of a re-
The summation of nearly synchronous duction in the area of a compound muscle
muscle fiber action potentials recorded from action potential greater than that normally
a muscle, commonly produced by stimu- seen with stimulation at two different
lation of the nerve supplying the muscle points on a nerve trunk; anatomic varia-
either directly or indirectly. Baseline-to- tions of nerve pathways and technical fac-
peak amplitude, duration, and latency of the tors related to nerve stimulation must be
negative phase should be noted, along with excluded as the cause of the reduction in
details of the method of stimulation and area.
recording. Use of specific named potentials Conduction Distance: The length of nerve
is recommended, for example, M wave, F or muscle over which conduction is de-
wave, H wave, T wave, A wave, and Rl or termined, customarily measured in cen-
R2 wave (blink responses). timeters or millimeters.
Compound Nerve Action Potential (Com- Conduction Time: See conduction velocity.
pound NAP): The summation of nearly Conduction Velocity (CV): Speed of propa-
synchronous nerve fiber action potentials gation of an action potential along a nerve
recorded from a nerve trunk, commonly or muscle fiber. The nerve fibers studied
(motor, sensory, autonomic, or mixed Cramp Discharge: Involuntary repetitive fir-

nerve) should be specified. For a nerve ing of motor unit action potentials at a high
trunk, the maximum conduction velocity frequency (up to 150 Hz) in a large area of
is calculated from the latency of the evoked a muscle usually associated with painful
potential (muscle or nerve) at maximal or muscle contraction. Both discharge frequency
supramaximal intensity of stimulation at and number of motor unit action poten-
two different points. The distance be- tials activated increase gradually during
tween the two points (conduction distance) development, and both subside gradually
is divided by the difference between the with cessation. See muscle cramp.
corresponding latencies (conduction time). Cross Talk: ( 1 ) A general term for abnormal
The calculated result is the conduction ve- communication between excitable mem-
locity of the fastest fibers and is usually ex- branes. See ephapse and ephaptic transmis-
pressed as meters per second (m/s). As sion. (2) Term used in kinesiologicEMGfor
commonly used, refers to the maximum signals picked up from adjacent muscles.
conduction velocity. By specialized tech- Crossed Leg Palsy: Synonym for peroneal
niques, the conduction velocity of other neuropathy at the knee.
fibers can also be determined and should Cubital Tunnel Syndrome: A mononeuropathy
be specified, for example, minimum con- involving the ulnar nerve in the region of
duction velocity. the elbow. An entrapment neuropathy caused
Congenital Myasthenia: A heterogeneous by compression of the nerve as it passes
group of genetic disorders of the neuro- through the aponeurosis (the cubital tun-
muscular junction manifest by muscle nel) of the two heads of the flexor carpi
weakness and fatigue. ulnaris approximately 1.5-3.5 cm distal to
Contraction: A voluntary or involuntary re- the medial epicondyle of the elbow. The
versible muscle shortening that may or mechanism of entrapment is presumably
may not be accompanied by action poten- narrowing of the cubital tunnel during el-
tials from muscle. Contrast the term con- bow flexion. See also tardy ulnar palsy and
tracture. ulnar neuropathy at the elbow.
Contraction Fasciculation: Clinical term for Cutaneous Reflex: A reflex produced by cu-
visible twitching of a muscle with weak vol- taneous stimulation. There are several
untary or postural contraction which has phases to cutaneous reflexes, and, if the
the appearance of a fasciculation. More muscle has a background contraction, the
likely to occur in neuromuscular disorders phases can be seen to be inhibitory as well
in which the motor unit territory is enlarged as excitatory.
and the tissue covering the muscle is thin, CV: Abbreviation for conduction velocity.
but may also be observed in normal indi- Cycles Per Second (c/s, cps): Unit of fre-
viduals. quency. See preferred term hertz (Hz).
Contracture: (1) Fixed resistance to stretch Decomposition EMG: Synonym for auto-
of a shortened muscle due to fibrous con- matic decomposition EMG.
nective tissue changes and loss of sarco- Decrements! Response: See preferred term,
meres in the muscle. Limited movement of decrementing response.
a joint may be due to muscle contracture Decrementing Response: A reproducible de-
or to fibrous connective tissue changes in cline in the amplitude and/or area of the
the joint. Contrast with contraction, which M wave of successive responses to repetitive
is a rapidly reversible painless shortening nerve stimulation. The rate of stimulation
of the muscle. (2) The prolonged, painful, and the total number of stimuli should be
electrically silent, and involuntary state specified. Decrementing responses with
of temporary muscle shortening seen in disorders of neuromuscular transmission
some myopathies (for example, muscle are most reliably seen with slow rates (2-5
phosphorylase deficiency). Hz) of nerve stimulation. A decrementing
Coupled Discharge: See preferred term, response with repetitive nerve stimulation
satellite potential. commonly occurs in disorders of neuro-
cps (also c/s): Abbreviation for cycles per sec- muscular transmission, but can also be
ond. See preferred term, Hertz (Hz). seen in some neuropathies, myopathies, and
motor neuron disease. An artifact resembling tentials occur in groups, the rate of re-
a decrementing response can result from currence of the group and rate of repeti-
movement of the stimulating or recording tion of the individual components in the
electrodes during repetitive nerve stimulation groups should be specified. See also firing
(see pseudodecrement). Contrast with incre- rate.
menting response. Discrete Activity: See interference pattern.
Delay: (1) The time between the beginning Distal Latency: The interval between the de-
of the horizontal sweep of the oscilloscope livery of a stimulus to the most distal point
and the onset of an applied stimulus. (2) A of stimulation on a nerve and the onset of
synonym for an information storage de- a response. A measure of the conduction
vice (delay line) used to display events oc- properties of the distal-most portion of
curring before a trigger signal. motor or sensory nerves. See motor latency
Delay Line: An information storage device and sensory latency.
used to display events which occur before Double Discharge: Two sequential firings of
a trigger signal. A method for displaying a motor unit action potential of the same
a waveform at the same point on a sweep form and nearly the same amplitude, oc-
from a free-running ekctromyogram. curring consistently in the same relation-
Demyelination: Disease process affecting the ship to one another at intervals of 2-20
myelin sheath of central or peripheral ms. See also multiple discharge, triple dis-
nerve fibers, manifested by conduction ve- charge.
locity slowing, conduction block, or both. Doublet: Synonym for the preferred term,
Denervation Potential: Sometimes used as a
double discharge.
synonym for fibrillation potential. Use of this
term is discouraged, since fibrillation po- DSEP: Abbreviation for dermatomal so-
tentials can occur in the absence of de- matosensory evoked potential.
nervation. See preferred term, fibrillation Duration: The time during which something
potential. exists or acts. (1) The interval from the
Depolarization: A change in the existing beginning of the first deflection from the
membrane potential to a less negative baseline to its final return to the baseline
value. Depolarizing an excitable cell from of an action potential or waveform, unless
its resting level to threshold typically gen- otherwise specified. If only part of the
erated an action potential. waveform is measured, the points of the
Depolarization Block: Failure of an excitable measurement should be specified. For ex-
cell to respond to a stimulus due to pre- ample, the duration of the M wave may be
existing depolarization of the cell mem- measured as the negative phase duration
brane. and refers to the interval from the deflec-
Depth Electrodes: Electrodes which are in- tion of the first negative phase from the
serted into the substance of the brain for baseline to its return to the baseline. (2)
electrophysiological recording. Most of- The interval of the applied current or volt-
ten inserted using stereotactic techniques. age of a single electric stimulus. (3) The in-
Dermatomal Somatosensory Evoked Poten- terval from the beginning to the end of
tial (DESP): Scalp recorded waveforms gen- a series of recurring stimuli or action
erated from repeated stimulation of a potentials.
specific dermatome. Different from typical Dynamic EMG: See kinesiologic EMG.
somatosensory evoked potentials which are Dyskinesia: An abnormal involuntary move-
recorded in response to stimulation of a ment of a choreicor dystonictype. The term
named peripheral nerve. is nonspecific and is often used in associ-
Discharge: The firing of one or more ex- ation with a modifier that describes its eti-
citable elements (neurons, axons, or mus- ology, for example, tardive dyskinesia or
cle fibers); as conventionally used, refers L-DOPA dyskinesia.
to all-or-none potentials only. Synonymous Dystonia: A disorder characterized by invol-
with action potential. untary movements caused by sustained
Discharge Frequency: The rate at which a muscle contraction, producing prolonged
potential discharges repetitively. When po- movements or abnormal postures.
E-l: Synonymous with input terminal 1. See ripheral, and autonomic nervous systems
recording electrode. and muscles. See also clinical electromyog-
E-2: Synonymous with input terminal 2. See raphy, electromyography, electroneurography, elec-
recording electrode. troneuromyography, evoked potentials, electrodi-
E:I Ratio: In autonomic testing, the ratio of agnostic medicine, electrodiagnostic medicine
the longest electrocardiographic R-R in- consultation, and electrodiagnostic medicine
terval during expiration to the shortest consultant.
during inspiration. Primarily a measure of Electrodiagnostic Medicine: A specific area
parasympathetic control of heart rate. of medical practice in which a physician
Early Recruitment: A recruitment pattern integrates information obtained from the
which occurs in association with a reduc- clinical history, observations from physical
tion in the number of muscle fibers per examination, and scientific data acquired
motor unit or when the force generated by by recording electrical potentials from the
the fibers is reduced. At low levels of mus- nervous system and muscle to diagnose or
cle contraction more motor unit action po- diagnose and treat diseases of the central,
tentials are recorded than expected, and a peripheral, and autonomic nervous sys-
full interference pattern may be recorded at tems, neuromuscular junctions, and mus-
relatively low levels of muscle contraction. cle. See also electrodiagnosis, electrodiagnos-
Most often encountered in myopathy. tic medicine consultation, and electrodiagnostic
Earth Electrode: Synonymous with ground medicine consultant.
electrode. Electrodiagnostic Medicine Consultant: A
EDX: Abbreviation for electrodiagnosis. Can physician specially trained to obtain a
also be used for electrodiagnostic and ekc- medical history, perform a physical exam-
trodiagnostic medicine. ination, and to record and analyze data ac-
Electric Inactivity: See preferred term, elec- quired by recording electrical potentials
tric silence. from the nervous system and muscle to di-
Electric Silence: The absence of measurable agnose and/or treat diseases of the cen-
electric activity due to biologic or nonbi- tral, peripheral, and autonomic nervous
ologic sources. The sensitivity and signal- systems, neuromuscular junction, and
to-noise level of the recording system muscle. See also electrodiagnosis, electrodi-
should be specified. agnostic medicine, and electrodiagnostic medi-
Electrocorticography: Electrophysiologic re- cine consultation.
cording directly from the surface of the Electrodiagnostic Medicine Consultation:
brain. In the intraoperative setting, The medical evaluation in which a spe-
recordings are made of ongoing sponta- cially trained physician (electrodiagnostic
neous electroencephalogram activity, or medicine consultant) obtains a medical his-
potentials evoked by stimulation of periph- tory, performs a physical examination,
eral sensory pathways. and integrates scientific data acquired by
Electrode: A conducting device used to recording electrical potentials from the
record an electric potential (recording elec- nervous system and muscle to diagnose
trode) or to deliver an electric current and/or treat diseases of the central, pe-
(stimulating electrode). In addition to the ripheral, and autonomic nervous systems,
ground electrode used in clinical recordings, neuromuscular junction, and muscle. See
two electrodes are always required either also electrodiagnosis, electrodiagnostic medi-
to record an electric potential or to de- cine, and electrodiagnostic medicine consul-
liver a stimulus. See ground electrode, record- tant.
ing electrode, and stimulating electrode. Also Electromyogram: The record obtained by
see specific needle electrode configurations: electromyography.
monopolar, unipolar, concentric, bifilar record- Electromyograph: Equipment used to acti-
ing, bipolar stimulating, multilead, single fiber, vate, record, process, and display electri-
and macro-EMG needle electrodes. cal potentials for the purpose of evaluating
Electrodiagnosis (EDX): The scientific meth- the function of the central, peripheral,
ods of recording and analyzing biologic and autonomic nervous systems, neuro-
electrical potentials from the central, pe- muscular junction, and muscles.
Electromyographer: See preferred term, ekc- in the postsynaptic membrane of a mus-

trodiagnostic medicine consultant. cle fiber by release of acetylcholine from
Electromyography (EMG): Strictly defined, the presynaptic axon terminal in response
the recording and study of insertion, spon- to an action potential.
taneous, and voluntary activity of muscle End Plate Spike: See end plate activity (bi-
with a recording electrode (either a needle elec- phasic) .
trode for invasive EMG or a surface electrode End Plate Zone: The region in a muscle
for kinesiologic studies). The term is also where neuromuscular junctions are con-
commonly used to refer to an ekctrodiag- centrated.
nostic medicine consultation, but its use in ENG: Abbreviation for electroneurography.
this context is discouraged. ENMG: Abbreviation for electroneuromyogra-
Electroneurography (ENG): The recording phy.
and study of the action potentials of pe- Entrapment Neuropathy: A mononeuropathy
ripheral nerve. Synonymous with nerve caused by compression of nerve as it passes
conduction studies. through an area of anatomical narrowing.
Electroneuromyography (ENMG): The com- Ephapse: A point of abnormal communica-
bined studies of electromyography and elec- tion where an action potential in one mus-
troneurography. Synonymous with clinical cle fiber or axon can cause depolarization
electromyography. See preferred term elec- of an adjacent muscle fiber or axon to
trodiagnostic medicine consultation. generate an action potential.
EMG: Abbreviation for electromyography. Ephaptic Transmission: The generation of a
End Plate Activity: Spontaneous electric ac- nerve fiber action potential from one muscle
tivity recorded with a needle electrode close fiber or axon to another through an
to muscle end plates. These potentials may ephapse. Postulated to be the basis for com-
have several different morphologies. plex repetitive discharges, myokymic discharges,
1. Monophasic: Low-amplitude (10-20 /*V), and hemifacial spasm.
short-duration (0.5-1.0 ms), negative po- EPSP: Abbreviation for excitatory postsynaptic
tentials occurring in a dense, steady pat- potential.
tern, the exact frequency of which cannot Erb's Point: The site at the anterolateral base
be defined. These nonpropagated poten- of the neck where percutaneous nerve
tials are probably miniature end-plate poten- stimulation activates the axons comprising
tials recorded extracellularly. Referred to the upper trunk of the brachial plexus.
as end plate noise or sea-shell sound (sea shell Erb's Point Stimulation: Percutaneous supra-
roar or noise). clavicular nerve stimulation during which
2. Biphasic: Moderate-amplitude (100-300 the upper trunk of the brachial plexus is
/iV), short-duration (2-4 ms), initially activated. See the more general and pre-
negative spike potentials occurring irregu- ferred term, supraclavicular nerve stimula-
larly in short bursts with a high frequency tion.
(50-100 Hz). These propagated potentials Evoked Potential: Electric waveform elicited
are generated by muscle fibers excited by - by and temporally related to a stimulus,
activity in nerve terminals. These poten- most commonly an electric stimulus de-
tials have been referred to as biphasic livered to a sensory receptor or nerve, or
spike potentials, end-plate spikes, and, in- applied directly to a discrete area of the
correctly, nerve potentials. May also have a brain, spinal cord, or muscle. See auditory
biphasic initially positive morphology. evoked potential, brain stem auditory evoked po-
3. Triphasic: Similar to biphasic potentials, tential, spinal evoked potential, somatosensory
but the waveforms have three phases with an evoked potential, visual evoked potential, com-
initial positive deflection. Fire in an ir- pound muscle action potential, and compound
regular fashion; contrast with fibrillation sensory nerve action potential.
potential. Evoked Potential Studies: Recording and
End Plate Noise: See end plate activity analysis of electric waveforms of biologic
(monophasic). origin elicited in response to electrical,
End Plate Potential (EPP): The graded non- magnetic, or physiological stimuli. Stimuli
propagated membrane potential induced are applied to specific motor or sensory
receptors, and the resulting waveforms are producible increase in the amplitude and
recorded along their anatomic pathways area of successive M waves during repetitive
in the peripheral and central nervous sys- nerve stimulation. Postactivation or postte-
tem. A single motor or sensory modality is tanic facilitation—Nerve stimulation stud-
typically tested in a study, and the modal- ies performed within a few seconds after
ity studied is used to define the type of a brief period (2-60 seconds) of nerve
study performed. See auditory evoked po- stimulation producing tetanus or after a
tentials, brain stem auditory evoked potentials, strong voluntary contraction may show
visual evoked potentials, and somatosensory changes in the configuration of the M
evoked potentials. wave (s) compared to the results of identi-
Evoked Response: Tautology. Use of term cal studies of the rested muscle as follows:
discouraged. See preferred team, evoked (a) repair of the decrement—A diminution of
potential. the decrementing response with slow rates
Excitability: Capacity to be activated by or (2-5 Hz) of repetitive nerve stimulation;
react to a stimulus. (b) increment after exercise—an increase in
Excitatory Postsynaptic Potential (EPSP): A the amplitude and area of the M wave
local, graded depolarization of a neuron in elicited by a single supramaximal stimu-
response to activation by a nerve terminal. lus. Distinguish from pseudofacilitation,
Contrast with inhibitory postsynaptic poten- which occurs in normal individuals in re-
tial. sponse to repetitive nerve stimulation at
Exploring Electrode: Synonymous with ac- high rates (20-50 Hz) or after strong vo-
tive electrode. See recording electrode. litional contraction. It probably reflects a
F Reflex: An incorrect term for F wave. reduction in the temporal dispersion of the
F Response: Synonymous with F wave. See summation of a constant number of mus-
preferred term, F wave. clefiber action potentials and is characterized
F Wave: An action potential evoked intermit- by an increase in the amplitude of the suc-
tently from a muscle by a supramaximal cessive M waves with a corresponding de-
electric stimulus to the nerve due to an- crease in their duration. There is no net
tidromic activation of motor neurons. When change in the area of the negative phase
compared with the maximal amplitude of of successive M waves. (2) An increase in
the Mwave, it is smaller (l%-5% of the M the amplitude of the motor evoked potential
wave) and has a variable configuration. Its as a result of background muscle activa-
latency is longer than the M wave and is tion.
variable. It can be evoked in many mus- Far-Field: A region of electrical potential
cles of the upper and lower extremities, where the isopotential voltage lines associ-
and the latency is longer with more distal ated with a current source change slowly
sites of stimulation. Named "F" wave by over a short distance. Some use the term
Magladery and McDougal in 1950, be- far-field potential to designate a potential
cause it was first recorded from foot mus- that does not change in latency, amplitude,
cles. Compare with the H wave and the or polarity over infinite distances; alter-
A wave. One of the late responses. native designations include "boundary po-
Facial Neuropathy: Clinical diagnosis of fa- tential" and "junctional potential." The
cial weakness or paralysis due to pathol- terms near-field and far-field are arbitrary
ogy affecting the seventh cranial nerve (fa- designations as there are no agreed-upon
cial nerve). Bell's palsy refers to a facial criteria defining where the near-field ends
neuropathy due to inflammation of the fa- and the far-field begins. Compare with
cial nerve. near-field.
Facilitation: An increase in an electrically Fasciculation: The random, spontaneous
measured response following identical stim- twitching of a group of muscle fibers be-
uli. Occurs in a variety of circumstances: longing to a single motor unit. The twitch
( 1 ) Improvement of neuromuscular trans- may produce movement of the overlying
mission resulting in activation of previously skin (if in limb or trunk muscles) or mu-
inactive muscle fibers. May be identified cous membrane (if in the tongue). If the
in several ways: Incrementing response—a re- motor unit is sufficiently large, an associ-
ated joint movement may be observed. amplitude of less than 1 mV. May also have
The electric activity associated with the a biphasic, initially negative phase when
twitch is termed a fasciculation potential. recorded at the site of initiation. It has an
See also myokymia. Historically, the term associated high-pitched regular sound de-
fibrillation was used incorrectly to describe scribed as "rain on a tin roof." In addition
fine twitching of muscle fibers visible to this classic form, positive sharp waves may
through the skin or mucous membranes. also be recorded from fibrillating muscle
This usage is no longer accepted. fibers when the potential arises from an
Fasciculation Potential: The electric activity area immediately adjacent to the needle
associated with a fasciculation which has electrode.
the configuration of a motor unit activation Firing Pattern: Qualitative and quantitative
potential but which occurs spontaneously. descriptions of the sequence of discharge
Most commonly occur sporadically and of electric waveforms recorded from mus-
are termed "single fasciculation poten- cle or nerve.
tials." Occasionally the potentials occur as Firing Rate: Frequency of repetition of a po-
a grouped discharge and are termed a "brief tential The relationship of the frequency
repetitive discharge." The repetitive firing of to the occurrence of other potentials and
adjacent fasciculation potentials, when the force of muscle contraction may be de-
numerous, may produce an undulating scribed. See also discharge frequency.
movement of muscle (see myokymia}. Use Flexor Reflex: A reflex produced by a nox-
of the terms benign fasciculation and ma- ious cutaneous stimulus, or a train of elec-
lignant fasciculation is discouraged. In- trical stimuli, that activates the flexor
stead, the configuration of the potentials, muscles of a limb and thus acts to with-
peak-to-peak amplitude, duration, number draw it from the stimulus. In humans, it is
of phases, stability of configuration, and fre- well-characterized only in the lower ex-
quency of occurrence, should be specified. tremity.
Fatigue: A state of depressed responsiveness Frequency: Number of compelte cycles of a
resulting from activity. Muscle fatigue is a repetitive waveform in 1 second. Measured
reduction in contraction force following in hertz (Hz) or cycles per second (cps or c/s).
repeated voluntary contraction or electric Frequency Analysis: Determination of the
stimulation. range of frequencies composing a waveform,
Fiber Density: (1) Anatomically, a measure with a measurement of the absolute or
of the number of muscle or nerve fibers relative amplitude of each component
per unit area. (2) In single fiber electromyog- frequency.
raphy, the mean number of muscle fiber ac- Full Interference Pattern: See interference
tion potentials fulfilling amplitude and rise pattern.
time criteria belonging to one motor unit Full Wave Rectified EMG: The absolute
within the recording area of a single fiber value of a raw EMG signal. Involves in-
needle electrode encountered during a sys- verting all the waveforms below the isopo-
tematic search in a weakly, voluntarily con- tential line and displaying them with op
tracting muscle. See also single fiber elec- posite polarity above the line. A technique
tromyography, single fiber needle electrode. used to analyze kinesiologic EMG signals.
Fibrillation: The spontaneous contractions of Functional Refractory Period: See refractory
individual muscle fibers which are not vis- period.
ible through the skin. This term has been Gl, G2: Abbreviation for grid 1 and grid 2.
used loosely in electromyography for the pre- Generator: In volume conduction theory, the
ferred term, fibrillation potential. source of electrical activity, such as an ac-
Fibrillation Potential: The action potential of tion potential See far-field and near-field.
a single muscle fiber occurring sponta- "Giant" Motor Unit Action Potential: Use of
neously or after movement of a needle elec- term discouraged. Refers to a motor unit
trode. Usually fires at a constant rate. Con- action potential with a peak-to-peak ampli-
sists of biphasic or triphasic spikes of short tude and duration much greater than the
duration (usually less than 5 ms) with an range found in corresponding muscles in
initial positive phase and a peak-to-peak normal subjects of similar age. Quantita-
tive measurements of amplitude and du- tion. The reflex and wave are named in
ration are preferable. honor of Hoffmann's description (1918).
Giant Somatosensory Evoked Potential: En- Compare the F wave and A wave.
larged somatosensory evoked potentials seen as Habituation: Decrease in size of a re/lex mo-
a characteristic of cortical reflex myoclonus tor response to an afferent stimulus when
and reflecting cortical hyperexcitability. the latter is repeated, especially at regular
Grid 1: Synonymous with Gl, input terminal and recurring short intervals.
1 (E-l), or active or exploring electrode. Use Hemifacial Spasm: Clinical condition char-
of the term Grid 1 is discouraged. See acterized by frequent, repetitive, unilat-
recording electrode. eral, involuntary contractions of the facial
Grid 2: Synonymous with G2, input terminal muscles. Electrodiagnostic studies dem-
2 (E-2), or reference electrode. Use of the term onstrate brief discharges of groups of motor
Grid 2 is discouraged. See recording elec- unit action potentials occurring simultane-
trode. ously in several facial muscles. Occasion-
Ground Electrode: A connection from the ally high frequency discharges occur.
patient to earth. Used as a common return Hertz (Hz): Unit of frequency. Synonymous
for an electric circuit and as an arbitrary with cycles per second.
zero potential reference point. Hoffmann Reflex: See H wave.
Grouped Discharge: Term used historically Hyperekplexia: Clinical condition charac-
to describe three phenomena: (1) irregu- terized by exaggerated startle reflexes. Star-
lar, voluntary grouping of motor unit action tie reflexes can be exaggerated by being
potentials as seen in a tremulous muscular more extreme than expected (larger am-
contraction, (2) involuntary grouping of plitude or more widespread) or by lack of
motor unit action potentials as seen in normal habituation to repeated similar
myokymia, (3) general term to describe re- stimuli. Can be either genetic or acquired.
peated firing of motor unit action po- Hyperpolarization: A change in the existing
tentials. See preferred term, repetitive dis- membrane potential to a more negative
charge. value.
Guillain-Barre Syndrome: Eponym for acute Hypertonia: See tone.
inflammatory neuropathy. Also referred to as Hypotonia: See tone.
Landry-Guillain-Barre syndrome or Landry- Hz: Abbreviation for hertz.
Guillain-Barre-Strohl syndrome. Impulse Blocking: See blocking.
H Reflex: Abbreviation for Hoffmann re- Inching: A nerve conduction study technique
flex. See H wave. consisting of applying stimuli at multiple
H Response: See preferred term H wave. short distance increments along the
H Wave: A compound muscle action potential course of a nerve. This technique is used
with a consistent latency recorded from to localize an area of focal slowing or con-
muscles after stimulation of the nerve. duction block.
Regularly found in adults only in a limited Incomplete Activation: Motor unit action po-
group of physiologic extensors, particu- tentials firing, on requested maximal ef-
larly the calf muscles. Compared to the M fort, in decreased numbers at their nor-
wave of the same muscle, has a longer la- mal physiological rates, within the basal
tency and thus is one of the late responses firing range of 5-10 Hz. Causes include
(see A and F wave). Most reliably elicited upper motor neuron syndrome, pain on mus-
with a stimulus of long duration (500-1000 cle contraction, hysteria/conversion reac-
/AS) . A stimulus intensity sufficient to elicit tion, and malingering. Contrast with re-
a maximal amplitude M wave reduces or duced recruitment.
abolishes the H wave. Thought to be due Increased Insertion Activity: See insertion ac-
to a spinal reflex, with electric stimulation tivity.
of afferent fibers in the mixed nerve and Increment After Exercise: See facilitation.
activation of motor neurons to the muscle Incremental Response: See preferred term,
mainly through a monosynaptic connec- incrementing response.
tion in the spinal cord. The latency is Incrementing Response: A reproducible in-
longer with more distal sites of stimula- crease in amplitude and/or area of succes-
sive M waves to repetitive nerve stimulation. Input Terminal 2: The input of a differen-
The rate of stimulation and the number tial amplifier at which negativity, relative
of stimuli should be specified. Commonly to the other input terminal, produces a
seen in two situations. First, in normal sub- downward deflection. Synonymous with
jects the configuration of the M wave may reference electrode, E-2 or less preferred term,
change in response to repetitive nerve grid 2. See recording electrode.
stimulation so that the amplitude pro- Insertion Activity: Electric activity caused by
gressively increases as the duration de- insertion or movement of a needle electrode
creases, leaving the area of the M wave within a muscle. The amount of the activ-
unchanged. This phenomenon is termed ity may be described as normal, reduced,
pseudofacilitation. Second, in neuromuscular or increased (prolonged), with a descrip-
transmission disorders, the configuration of tion of the waveform and repetition rate.
the M wave may change with repetitive See also fibrillation potential and positive
nerve stimulation so that the amplitude sharp wave.
and the area of the M wave progressively Integrated EMG: Mathematical integration
increase. This phenomenon is termed of the full wave rectified EMG signal. Re-
facilitation. Contrast with decrementing flects tile cumulative EMG activity of a
response. muscle over time. See also linear envelope
Indifferent Electrode: Synonymous with ref- EMG.
erence electrode. Use of term discouraged. Interdischarge Interval: Time between con-
See recording electrode. secutive discharges of the same potential.
Infraclavicular Plexus: Segments of the bra- Measurements should be made between
chial plexus inferior to the divisions; in- the corresponding points on each wave-
cludes the three cords and the terminal form.
peripheral nerves. This clinically descrip- Interference: Unwanted electric activity
tive term is based on the fact that the clav- recorded from the surrounding environ-
icle overlies the divisions of the brachial ment.
plexus when the arm is in the anatomic Interference Pattern: Electric activity
position next to the body. recorded from a muscle with a needle elec-
Inhibitory Postsynaptic Potential (IPSP): A trode during maximal voluntary effort. A
local graded hyperpolarization of a neuron full interference pattern implies that no
in response to activation at a synapse by individual motor unit action potentials can be
a nerve terminal. Contrast with excitatory clearly identified. A reduced interference
postsynaptic potential pattern (intermediate pattern) is one in
Injury Potential: (1) The potential difference which some of the individual motor unit
between a normal region of the surface of action potentials may be identified while
a nerve or muscle and a membrane region others cannot due to superimposition of
that has been injured; also called a "de- waveforms. The term discrete activity is used
marcation," or "killed end" potential. Ap- to describe the electric activity recorded
proximates the potential across the mem- when each of several different motor unit
brane because the injured surface has action potentials can be identified in an
nearly the same potential as the interior ongoing recording due to limited super-
of the cell. (2) In electrodiagnostic medicine, imposition of waveforms. The term single
the term is also used to refer to the elec- unit pattern is used to describe a single mo-
trical activity associated with needle electrode tor unit action potential, firing at a rapid
insertion into muscle. See preferred terms rate (should be specified) during maxi-
fibrillation potential, insertion activity, and mum voluntary effort. The force of con-
positive sharp wave. traction associated with the interference
Input Terminal 1: The input terminal of a dif- pattern should be specified. See also early
ferential amplifier at which negativity, rel- recruitment, recruitment pattern, reduced re-
ative to the other input terminal, produces cruitment pattern.
an upward deflection. Synonymous with Interference Pattern Analysis: Quantitative
active or exploring electrode, E-l or less pre- analysis of the interference pattern. This can
ferred term, grid 1. See recording electrode. be done either in the frequency domain us-
ing fast Fourier transformation (FFT) or early reinnervation and myasthenic disor-
in the time domain. Can be done using a ders, the variability can be sufficiently
fixed load (for example, 2 kg), at a given large to be easily detectable by eye. Quan-
proportional strength (for example, 30% titative methods for estimating this vari-
of maximum) or at random strengths. The ability are not yet widely available.
following are measured in the time do- Jitter: The variability of consecutive dis-
main: a) the number of turns per second charges of the interpotential interval between
and b) the amplitude, defined as the mean two muscle fiber action potentials belonging
amplitude between peaks. to the same motor unit. Usually expressed
Intermediate Interference Pattern: See in- quantitatively as the mean value of the dif-
terference pattern. ference between the interpotential inter-
International 10-20 System: A system of ekc- vals of successive discharges (the mean con-
trode placement on the scalp in which elec- secutive difference, MCD). Under certain
trodes are placed either 10% or 20% of conditions, it is expressed as the mean
the total distance on a line on the skull value of the difference between interpo-
between the nasion and inion in the sagit- tential intervals arranged in the order of
tal plane and between the right and left decreasing interdischarge intervals (the
preauricular points in the coronal plane. mean sorted difference, MSD). See single fiber
Interpeak Interval: Difference between the electromyography.
peak latencies of two components of a wave- Jolly Test: A technique named for Friedrich
form. Jolly, who applied an electric current to
Interpotential Interval: Time between two excite a motor nerve repetitively while re-
different potentials. Measurement should cording the force of muscle contraction.
be made between the corresponding parts Use of the term is discouraged. Inappro-
of each waveform. priately used to describe the technique of
Intraoperative Monitoring: The use of elec- repetitive nerve stimulation.
trophysiological stimulating and record- Kinematics: Technique for description of
ing techniques in an operating room set- body movement without regard to the un-
ting. The term is usually applied to derlying forces. See kinesiologic EMG.
techniques which are used to detect injury Kinesiologic EMG: The muscle electrical ac-
to nervous tissue during surgery or to tivity recorded during movement. Gives
guide the surgical procedure. information about the timing of muscle
Involuntary Activity: Motor unit action poten- activity and its relative intensity. Either sur-
tials that are not under volitional control. face electrodes or intramuscular fine wire elec-
The condition under which they occur trodes are used. Synonymous with dynamic
should be described, for example, spon- EMG.
taneous or reflex potentials. If elicited by a Kinesiology: The study of movement. See
stimulus, its nature should be described. kinesiologic EMG.
Contrast with spontaneous activity. Kinetics: The internal and external forces af-
IPSP: Abbreviation for inhibitory postsynaptic fecting the moving body. See kinesiologic
potential. EMG.
Irregular Potential: See preferred term, ser- Late Component (of a Motor Unit Action
rated action potential. Potential): See preferred term, satellite po-
Isoelectric Line: In electrophysiologic tential.
recording, the display of zero potential dif- Late Response: A general term used to de-
ference between the two input terminals scribe an evoked potential in motor nerve
of the recording apparatus. See baseline. conduction studies having a longer latency
Iterative Discharge: See preferred term, than the M wave. Examples include A
repetitive discharge. wave, F wave, and H wave.
Jiggle: Shape variability of motor unit action Latency: Interval between a stimulus and a re-
potentials recorded with a conventional sponse. The onset latency is the interval be-
EMG needle electrode. A small amount oc- tween the onset of a stimulus and the on-
curs normally. In conditions of disturbed set of the evoked potential. The peak latency
neuromuscular transmission, including is the interval between the onset of a stim-
ulus and a specified peak of the evoked recording range of a macro-EMG electrode.
potential. Characterized by consistent appearance
Latency of Activation: The time required for when the small recording surface of the
an electric stimulus to depolarize a nerve macro-EMG electrode is positioned to
fiber (or bundle of fibers as in a nerve record action potentials from one muscle
trunk) beyond threshold and to initiate an fiber. The following characteristics can be
action potential in the fiber(s). This time is specified quantitatively: (1) maximal peak-
usually of the order of 0.1 ms or less. An to-peak amplitude, (2) area contained un-
equivalent term, now rarely used, is the der the waveform, (3} number of phases.
"utilization time." Macro MUAP: Abbreviation for macro motor
Latent Period: See preferred term, latency. unit action potential.
Linear Envelope EMG: Moving average of Macroelectromyography (Macro-EMG): Gen-
the full wave rectified EMG. Obtained by low eral term referring to the technique and
pass filtering the full wave rectified EMG. conditions that approximate recording of
See also integrated EMG. all muscle fiber action potentials arising from
Linked Potential: See preferred term, satel- the same motor unit. See macro motor unit
lite potential action potential.
Lipoatrophy: Pathologic loss of subcuta- Macro-EMG: Abbreviation for macroelec-
neous fat and connective tissues overlying tromyography.
muscle which mimics the clinical appear- Macro-EMG Needle Electrode: A modified
ance of atrophy of the underlying muscle. single fiber electromyography electrode insu-
Long-Latency Reflex: A reflex with many lated to within 15 mm from the tip and
synapses (polysynaptic) or a long pathway with a small recording surface (25 //,m in
(long-loop) so that the time to its oc- diameter) 7.5 mm from the tip.
curence is greater than the time of oc- Malignant Fasciculation: Used to describe
currence of short-latency reflexes. See also large, polyphasic fasciculation potentials fir-
long-loop reflex. ing at a slow rate. This pattern has been
Long-Loop Reflex: A reflex thought to have seen in progressive motor neuron disease,
a circuit that extends above the spinal seg- but the relationship is not exclusive. Use
ment of the sensory input and motor of this term is discouraged. See fascicula-
output. May involve the cerebral cortex. tion potential.
Should be differentiated from reflexes Maximal Stimulus: See stimulus.
arising from stimulation and recording Maximum Conduction Velocity: See conduc-
within a single segment or adjacent spinal tion velocity.
segments (that is, a segmental reflex). See MCD: Abbreviation for mean consecutive dif-
also long-latency reflex. ference. See jitter.
M Response: See preferred term, M wave. Mean Consecutive Difference (MCD): See
M Wave: A compound muscle action potential jitter.
evoked from a muscle by an electric stim- Mean Sorted Difference (MSD): See jitter.
ulus to its motor nerve. By convention, the Membrane Instability: Tendency of a cell
M wave elicited by a supramaximal stimu- membrane to depolarize spontaneously in
lus is used for motor nerve conduction stud- response to mechanical irritation or fol-
ies. Ideally, the recording electrodes should be lowing voluntary activation. May be used to
placed so that the initial deflection of the describe the occurrence of spontaneous
evoked potential from the baseline is nega- single muscle fiber action potentials such as
tive. Common measurements include la- fibrillation potentials during needle electrode
tency, amplitude, and duration. Also referred examination.
to as the motor response. Normally, the con- MEP: Abbreviation for motor evoked potential
figuration is biphasic and stable with re- MEPP: Abbreviation for miniature end-plate
peated stimuli at slow rates (1-5 Hz). See potential.
repetitive nerve stimulation. Microneurography: The technique of re-
Macro Motor Unit Action Potential: The av- cording peripheral nerve action poten-
erage electric activity of that part of an tials in humans by means of intraneural
anatomic motor unit that is within the electrodes.
Miniature End Plate Potential (MEPP): The Motor Nerve: A nerve containing axons
postsynaptic muscle fiber potentials pro- which innervate extrafusal and intrafusal
duced through the spontaneous release of muscle fibers. These nerves also contain
individual acetylcholine quanta from the sensory afferent fibers from muscle and
presynaptic axon terminal. As recorded other deep structures.
with monopolar or concentric needle electrodes Motor Nerve Conduction Velocity (MNCV):
inserted in the end-plate region, MEPPs The speed of propagation of action poten-
are monophasic, negative, short duration tials along a motor nerve. See conduction
(less than 5 ms), and generally less than velocity.
20 /aV in amplitude. Motor Neuron Disease: A clinical condition
Minimum Conduction Velocity: The nerve characterized by degeneration of motor
conduction velocity measured from slowly nerve cells in the brain, brain stem, and
conducting nerve fibers. Special tech- spinal cord. The location of degeneration
niques are needed to produce this mea- determines the clinical presentation. Pri-
surement in motor or sensory nerves. mary lateral sclerosis occurs when degen-
Mixed Nerve: A nerve composed of both eration affects mainly corticospinal tract
motor and sensory axons. motor fibers. Spinal muscular atrophy oc-
MNCV: Abbreviation for motor nerve conduc- curs when degeneration affects lower mo-
tion velocity. See conduction velocity. tor neurons. Amyotrophic lateral sclerosis
Mononeuritis Multiplex: A disorder charac- occurs when degeneration affects both
terized by axonal injury and/or demyeli- corticospinal tracts and lower motor
nation affecting nerve fibers in multiple neurons.
nerves (multiple mononeuropathies). Usu- Motor Point: The site over a muscle where
ally occurs in an asymmetric anatomic its contraction may be elicited by a minimal
distribution and in a temporal sequence intensity short duration electric stimulus.
which is not patterned or symmetric. Motor Response: (1) The compound muscle ac-
Mononeuropathy Multiplex: A disorder tion potential (M wave) recorded over a
characterized by axonal injury and/or de- muscle in response to stimulation of the
myelination affecting nerve fibers exclu- nerve to the muscle. (2) The muscle twitch
sively along the course of one named or contraction elicited by stimulation of the
nerve. nerve to a muscle. (3) The muscle twitch
Monophasic Action Potential: An action po- elicited by the muscle stretch reflex.
tential with the waveform entirely on one Motor Unit: The anatomic element consist-
side of the baseline. ing of an anterior horn cell, its axon, the
Monophasic End Plate Activity: See end plate neuromuscular junctions, and all of the
activity (monophasic). muscle fibers innervated by the axon.
Monopolar Needle Electrode: A solid wire Motor Unit Action Potential (MUAP): The
electrode coated with Teflon™, except at compound action potential of a single motor
the tip. Despite the term monopolar, a unit whose muscle fibers lie within the
separate surface or subcutaneous refer- recording range of an electrode. With vol-
ence electrode is required for recording untary muscle contraction, it is character-
electric signals. May also be used as a cath- ized by its consistent appearance and re-
ode in nerve conduction studies with another lationship to the force of the contraction.
electrode serving as an anode. The following measures may be specified,
Motor Evoked Potential (MEP): A compound quantitatively if possible, after the record-
muscle action potential produced by either ing electrode is placed randomly within the
transcranial magnetic stimulation or trans- muscle:
cranial electrical stimulation. 1. Configuration
Motor Latency: Interval between the onset a. Amplitude, peak-to-peak (juV or mV).
of a stimulus and the onset of the resultant b. Duration, total (ms).
compound muscle action potential (M wave). c. Number of phases (monophasic, bi-
The term may be qualified, as proximal phasic, triphasic, tetraphasic, poly-
motor latency or distal motor latency, de- phasic) .
pending on the relative position of the d. Polarity of each phase (negative, pos-
stimulus. itive).
e. Number of turns. Multi MUP Analysis: A template matching, de-

f. Variation of shape (jiggle), if any, composition EMG method used for MUAP
with consecutive discharges. analysis.
g. Presence of satellite (linked) potentials, Multielectrode: See multilead electrode.
if any. Multifocal Motor Neuropathy: A disease
h. Spike duration, including satellites. characterized by selective focal block of
2. Recruitment characteristics motor nerve conduction in multiple nerves.
a. Threshold of activation (first recruited, Motor nerve conduction studies may permit
low threshold, high threshold). identification and localization of the seg-
b. Onset frequency. ments of nerve affected by the underlying
c. Recruitment frequency (Hz,) or recruit- pathology.
ment interval (ms) of individual po- Multilead Electrode: Three or more insu-
tentials. lated wires inserted through apertures in
Descriptive terms implying diagnostic sig- a common metal cannula with their bared
nificance are not recommended, for ex- tips flush with the cannula's outer cir-
ample, myopathic, neuropathic, regeneration, cumference. The arrangement of the bare
nascent, giant, BSAP, and BSAPP. See poly- tips relative to the axis of the cannula and
phasic action potential, serrated action potential the distance between each tip should be
Motor Unit Fraction: See scanning EMG. specified. See electrode.
Motor Unit Number Counting: See the pre- Multiple Discharge: Four or more motor unit
ferred term motor unit number estimate action potentials of the same form and
(MUNE). nearly the same amplitude occurring con-
Motor Unit Number Estimate (MUNE): A sistently in the same relationship to one
quantitative technique for determining another and generated by the same axon.
the number of functioning motor units in See double and triple discharge.
a muscle. A variety of methods, including Multiple!: See multiple discharge.
spike-triggered averaging, incremental mo- MUNE: Abbreviation for motor unit number
tor nerve stimulation, F-wave measurement, estimate, motor unit number estimation, and
or a Poisson statistical technique can be motor unit number estimating.
used. Synonyms can include motor unit MUP: Abbreviation for motor unit potential.
number estimation and motor unit number See preferred term, motor unit action po-
estimating. tential.
Motor Unit Number Estimating (MUNE): Muscle Action Potential: Term commonly
See motor unit number estimate (MUNE). used to refer to a compound muscle action
Motor Unit Number Estimation (MUNE): potential.
See motor unit number estimate (MUNE). Muscle Atrophy: Decrease in size of a mus-
Motor Unit Potential (MUP): See synonym, cle that may be due to disease of nerve or
motor unit action potential. muscle, or to disuse.
Motor Unit Territory: The area of a muscle Muscle Cramp: An involuntary, painful mus-
cross-section within which the muscle cle contraction associated with electrical ac-
fibers belonging to an individual motor tivity. Cramp discharges are most common,
unit are distributed. but other types of repetitive discharges can
Movement Artifact: See artifact. also be seen.
Movement-Related Cortical Potential: Elec- Muscle Fiber Action Potential: Action poten-
troencephalogram activity associated with tial recorded from a single muscle fiber.
(before and after) a voluntary movement. Muscle Fiber Conduction Velocity: The
There are several components including speed of propagation of a single muscle
the Bereitschaftspotential before the move- fiber action potential, usually expressed as
ment and the motor potential at about meters per second. Usually less than most
the time of the movement. See also nerve conduction velocities, varies with the
Bereitschaftspotential. rate of discharge of the muscle fiber, and
MSD: Abbreviation for mean sorted difference. requires special techniques for measure-
See jitter. ment.
MUAP: Abbreviation for motor unit action Muscle Hypertrophy: Increase in the size of
potential. a muscle due to an increase in the size of
the muscle fibers or replacement or dis- at a fairly uniform firing rate (1-5 Hz).
placement of muscle fibers by other tis- Myokymic discharges are a subclass of
sues. The latter is also referred to by the grouped discharges and repetitive discharges.
term pseudohypertrophy, because the muscle See also ephapse and ephaptic transmission.
is enlarged but weak. Muscle fibers in- Myopathic Motor Unit Potential: Low am-
crease in size as a physiologic response to plitude, short duration, polyphasic motor
repetitive and forceful voluntary contrac- unit action potentials. Use of term discour-
tion or as a pathologic response to invol- aged. It incorrectly implies specific diag-
untary electric activity in a muscle, for nostic significance of a motor unit action
example, myotonic discharges or complex potential configuration. See motor unit ac-
repetitive discharges. tion potential.
Muscle Stretch Reflex: Activation of a mus- Myopathic Recruitment: Used to describe an
cle which follows stretch of the muscle, for increase in the number and firing rate of
example, by percussion of a muscle ten- motor unit action potentials compared with
don. See stretch reflex, T wave. normal for the strength of muscle con-
Muscle Tone: See tone. traction. Use of term discouraged.
Myasthenia Gravis: A disease characterized Myopathy: Disorder affecting the structure
by muscle weakness which increases with and/or function of muscle fibers. Etiolo-
repetitive muscle activation. Most com- gies include hereditary, congenital, mi-
monly, an autoimmune disease caused by tochondrial, inflammatory, metabolic,
the presence of antibodies to the acetyl- infectious, neoplastic, vascular, and trau-
choline receptors at the neuromuscular matic diseases. Most, but not all of these
junction. disorders, show abnormalities on needle
Myoclonus: A quick jerk of a body part electromyography.
produced by a brief muscle contraction typ- Myotonia: Delayed relaxation of a muscle af-
ically originating from activity in the center voluntary contraction or percussion. As-
tral nervous system. Based on the ana- sociated with propagated electric activity,
tomic location of the pathology, may be such as myotonic discharges, complex repetitive
classified as spinal, segmental, brain stem, discharges, or neuromyotonic discharges.
or cortical. Myotonic Discharge: Repetitive discharge
Myoedema: Focal muscle contraction pro- which occurs at rates of 20-80 Hz. There
duced by muscle percussion. Not associ- are two types: (1) biphasic (positive-nega-
ated with propagated electric activity. May tive) spike potentials less than 5 ms in du-
be seen in hypothyroidism (myxedema) ration resembling fibrillation potentials. (2)
and chronic malnutrition. positive waves of 5-20 ms duration resem-
Myokymia: Continuous quivering or undu- bling positive sharp waves. Both potential
lating movement of surface and overlying forms are recorded after needle electrode in-
skin and mucous membrane associated sertion, after voluntary muscle contraction,
with spontaneous, repetitive discharge of or after muscle percussion and are due to
motor unit action potentials. See myokymic dis- independent, repetitive discharges of sin-
charge, fasciculation, and fasciculation poten- gle muscle fibers. The amplitude and fre-
tial. quency of the potentials must both wax and
Myokymic Discharge: A form of involuntary wane. This change produces a character-
activity in which motor unit action potentials istic musical sound in the audio output of
fire repetitively and may be associated with the electromyograph due to the correspond-
clinical myokymia. Two firing patterns have ing change in pitch, which has been
been described: (7) Commonly, the dis- likened to the sound of a "dive bomber."
charge is a brief, repetitive firing of single Contrast with waning discharge.
motor unit action potentials for a short pe- Myotonic Potential: See preferred term, my-
riod (up to a few seconds) at a uniform otonic discharge.
rate (2-60 Hz) followed by a short period NAP: Abbreviation for nerve action potential.
(up to a few seconds) of silence, with rep- See compound nerve action potential.
etition of the same sequence for a partic- Nascent Motor Unit Potential: From the
ular potential at regular intervals. (2) Latin nascens, "to be born." Refers to very
Rarely, the potential recurs continuously low amplitude, short duration, highly
polyphasic motor unit action potentials ob- Nerve Conduction Velocity (NCV): The
served during early states of reinnervation. speed of action potential propagation along
Use of term is discouraged, as it incor- a nerve fiber or nerve trunk. Generally as-
recdy implies diagnostic significance of a sumed to refer to the maximum speed of
motor unit action potential configuration. propagation unless otherwise specified.
See motor unit action potential. See conduction velocity.
NCS: Abbreviation for nerve conduction study. Nerve Fiber Action Potential: Action potential
NCV: Abbreviation for nerve conduction ve- recorded from a single axon.
locity. See conduction velocity. Nerve Potential: Equivalent to nerve action po-
tential. Also commonly, but inaccurately,
Near-Field: A region of electrical activity
used to refer to the biphasic form of end
where the isopotential voltage lines associ-
plate activity observed during needle electrode
ated with a current source change rapidly
examination of muscle. The latter use is in-
over a short distance. The terms near-field
correct, because muscle fibers, not nerve
and far-field are arbitrary designations, as
fibers, are the source of these potentials.
there are no agreed-upon criteria defin-
Nerve Trunk Action Potential: See preferred
ing where the near-field ends and the far-
term, compound nerve action potential.
field begins. Compare widi far-field.
Neurapraxia: Clinical term used to describe
Needle Electrode: An electrical device used the reversible motor and sensory deficits
for recording or stimulating that is posi- produced by focal compressive or traction
tioned near the tissue of interest by pen- lesions of large myelinated nerve fibers.
etration of the skin. See specific elec- It is due to conduction block, most often
trodes: bifilar (bipolar) needle recording caused by focal demyelination, but, when
electrode, concentric needle electrode, macro- very short lived, presumably caused by fo-
EMG needle electrode, monopolar needle elec- cal ischemia. The axon is not injured at
trode, multilead electrode, single fiber needle the lesion site. Compare with axonotmesis
electrode, and stimulating electrode. and neurotmesis.
Nerve Action Potential (NAP): Strictly de- Neuromuscular Transmission Disorder:
fined, refers to an action potential recorded Clinical disorder associated with pathol-
from a single nerve fiber. The term is com- ogy affecting the structure and function of
monly used to refer to the compound nerve the neuromuscular junction and interfer-
action potential. See compound nerve action ing with synaptic transmission at that site.
potential. Specific diseases include myasthenia gravis,
Nerve Conduction Study (NCS): Recording Lambert-Eaton myasthenic syndrome, and
and analysis of electric waveforms of bio- botulism.
logic origin elicited in response to electric Neuromyopathy: Clinical disorder associ-
or physiologic stimuli. The waveforms are ated with pathology affecting both nerve
compound sensory nerve action potentials, com- and muscle fibers.
pound muscle action potentials, or mixed nerve Neuromyotonia: Clinical syndrome of con-
action potentials. The compound muscle tinuous muscle fiber activity manifested as
action potentials are generally referred to continuous muscle rippling and stiffness.
by letters which have historical origin: It may be associated with delayed relax-
M wave, F wave, H wave, T wave, A wave, ation following voluntary muscle contrac-
and Rl, R2 waves. It is possible under stan- tion. The accompanying electric activity
dardized conditions to establish normal may be intermittent or continuous. Terms
ranges for amplitude, duration, and latency used to describe related clinical syn-
of the waveforms and to calculate the max- dromes are continuous muscle fiber ac-
imum conduction velocity of sensory and tivity syndrome, Isaac syndrome, Isaac-
motor nerves. The term generally refers to Merton syndrome, quantal squander
studies of waveforms generated in the pe- syndrome, generalized myokymia, pseudo-
ripheral nervous system, whereas evoked po- myotonia, normocalcemic tetany, and neu-
tential studies refers to studies of waveforms rotonia. Distinguish from myotonia.
generated in both the peripheral and cen- Neuromyotonic Discharge: Bursts of motor
tral nervous systems. Synonymous with unit action potentials that fire at high rates
electroneurography. (150-300 Hz) for a few seconds, often
starting or stopping abruptly. The ampli- that can be voluntarily maintained by a

tude of the waveforms typically wanes. Dis- subject.
charges may occur spontaneously or be Order of Activation: The sequence of ap-
initiated by needle electrode movement, vol- pearance of different motor unit action po-
untary effort, ischemia, or percussion of a tentials with increasing strength of volun-
nerve. The activity originates in motor ax- tary contraction. See recruitment.
ons. Distinguish from myotonic discharges Orthodromic: Propagation of a nerve im-
and compkx repetitive discharges. One type pulse in the same direction as physiologic
of electrical activity recorded in patients conduction; for example, conduction
who have clinical neuromyotonia. along motor nerve fibers toward the muscle
Neuropathic Motor Unit Potential: Ab- and conduction along sensory nerve fibers
normally high-amplitude, long-duration, toward the spinal cord. Contrast with
polyphasic motor unit action potential. Use antidromic.
of term discouraged. Incorrectly implies a Paired Stimuli: Two consecutive stimuli de-
specific diagnostic significance of a motor livered in a time-locked fashion. The time
unit action potential configuration. See interval between the two stimuli and the
motor unit action potential. intensity of each stimulus can be varied but
Neuropathic Recruitment: A recruitment pat- should be specified. The first stimulus is
tern characterized by a decreased number called the conditioning stimulus and the sec-
of motor unit action potentials firing at a ond stimulus is the test stimulus. The con-
rapid rate. Use of term discouraged. See ditioning stimulus may modify tissue ex-
preferred terms, reduced interference pattern, citability, which is then evaluated by the
discrete activity, single unit pattern. response to the test stimulus.
Neuropathy: Disorder of the peripheral Parasite Potential: See preferred term, satel-
nerves. May be classified by the anatomi- lite potential.
cal structure of the nerve most affected by
Peak Latency: Interval between the onset of
the disease: the cell body (neuronopathy),
a stimulus and a specified peak of an
the axon (axonopathy), or the myelin
evoked waveform.
sheath (demyelinating neuropathy). May
selectively affect motor or sensory nerves or Peroneal Neuropathy at the Knee: A
both simultaneously. The etiology may be mononeuropathy involving the common
hereditary, metabolic, inflammatory, toxic, peroneal nerve as it passes around the
or unknown. head of the fibula. The presumed mecha-
Neurotmesis: Partial or complete nerve sev- nism is compression of the nerve against
erance including the axons, associated the fibula. See also crossed leg palsy.
myelin sheaths, and supporting connec- Phase: That portion of a waveform between
tive tissues, resulting in axonal degeneration the departure from and the return to the
distal to the injury site. Compare with ax- baseline.
onotmesis, neurapraxia. Plexopathy: Axonal and/or demyelinating
Neurotonic Discharges: Repetitive motor unit disorder affecting the nerve fibers exclu-
action potentials recorded from intramus- sive to the cervical, brachial, lumbar, or
cular electrodes during intraoperative moni- sacral rearrangement of spinal nerve roots
toring. Thought to arise from irritation or into peripheral nerves.
injury of nerves supplying the muscle from Polarization: The presence of an electric po-
which the recording is made. tential difference usually across an ex-
Noise: Electric activity not related to the sig- citable cell membrane.
nal of interest. In electrodiagnostic medicine, Polyneuropathy: Axonal and/or demyeli-
waveforms generated by electrodes, cables, nating disorder affecting nerve fibers, usu-
amplifier, or storage media and unrelated ally in a symmetrical fashion. The distal
to potentials of biologic origin. The term segments of the longer nerves in the lower
has also been used loosely to refer to one extremities are usually the most severely
form of end-plate activity. affected. May be classified as sensory, mo-
Onset Frequency: The lowest stable firing tor, or sensorimotor depending on the
rate for a single motor unit action potential function of nerve fibers affected.
Polyphasic Action Potential: An action poten- cular transmission after sustained activa-
tial with four or more baseline crossings, tion at the neuromuscular junction. The
producing five or more phases. See phase. changes in the configuration of the M
Contrast with serrated action potential wave due to postactivation exhaustion are
Polyradiculoneuropathy: See radiculopathy. referred to as postactivation depression.
Positive Sharp Wave: A biphasic, positive Postactivation Facilitation: See facilitation.
then negative action potential of a single Postactivation Potentiation: An increase in
muscle fiber. It is initiated by needle elec- the force of contraction (mechanical re-
trode movement (insertional or unsus- sponse) after a strong voluntary contrac-
tained positive sharp wave) or occurs tion. Contrast postactivation facilitation.
spontaneously. Typically discharge in a uni- Posttetanic: The period following tetanus.
form, regular pattern at a rate of 1-50 Hz; Contrast with postactivation.
the discharge frequency may decrease Posttetanic Depression: See postactivation
slightly just before cessation of discharge. depression.
The initial positive deflection is rapid (<1 Posttetanic Facilitation: See facilitation,
ms), its duration is usually less than 5 ms, posttetanic.
and the amplitude is up to 1 mV. The neg- Posttetanic Potentiation: (1) The increment-
ative phase is of low amplitude, and its du- ing mechanical response of muscle during
ration is 10-100 ms. A sequence of posi- and after repetitive nerve stimulation. (2) In
tive sharp waves is commonly referred to central nervous system physiology, en-
as a train of positive sharp waves. Assumed hancement of excitability or reflex outflow
to be recorded from a damaged area of a of neuronal systems following a long pe-
muscle fiber. This configuration may re- riod of high-frequency stimulation. See fa-
sult from the position of the needle elec- cilitation, potentiation.
trode which is believed to be adjacent to Potential: (1) A difference in charges, mea-
the depolarized segment of a muscle fiber surable in volts, that exists between two
injured by the electrode. Note that the points. Most biologically produced poten-
positive sharp waveform is not specific for tials arise from the difference in charge
muscle fiber damage. May occur in asso- beween two sides of a cell membrane. (2)
ciation with fibrillation potentials and are A term for a physiologically recorded
thought by some to be equivalent dis- waveform.
charges. Motor unit action potentials and po- Potentiation: Physiologically, the enhance-
tentials in myotonic discharges may have the ment of a response. The convention used
configuration of positive sharp waves. in this glossary is to use the term potentia-
Positive Wave: Loosely defined, the term tion to describe the incrementing me-
refers to a positive sharp wave. See preferred chanical response of muscle elicited by
term positive sharp wave. repetitive nerve stimulation, for example,
Postactivation: The period following volun- posttetanic potentiation, whereas the term
tary activation of a nerve or muscle. Con- facilitation is used to describe the incre-
trast with posttetanic. menting electrical response elicited by
Postactivation Depression: A reduction in repetitive nerve stimulation, for example,
the amplitude and area of the M wave(s) in postactivation facilitation.
response to a single stimulus or train of Prolonged Insertion Activity: See insertion
stimuli which occurs within a few minutes activity.
following a 10-60 second strong voluntary Propagation Velocity of a Muscle Fiber: The
contraction. Postactivation exhaustion refers speed of transmission of a muscle fiber ac-
to the cellular mechanisms responsible for tion potential.
the observed phenomenon of postactiva- Pseudodecrement: An artifact produced by
tion depression. Also used to describe re- movement of the stimulating or recording
duction of the M wave following a tetanus, electrodes during repetitive nerve stimulation.
which should more logically be termed The amplitude and area of the M wave can
posttetanic depression. vary in a way that resembles a decrementing
Postactivation Exhaustion: A reduction in response; however, the responses are gener-
the safety factor (margin) of neuromus- ally irregular and not reproducible.
Pseudofacilitation: See facilitation. Raster: A method for display of a free-

Pseudohypertrophy: See muscle hypertrophy. running sweep in electromyography. Sweeps
Pseudomyotonic Discharge: Formerly used are off-set vertically so that each successive
to describe complex repetitive discharges. Use sweep is displayed below the one preced-
of term discouraged. ing it.
Pseudopolyphasic Action Potential: Use of RawEMG: Unprocessed EMG signal recorded
term discouraged. See preferred term, ser- with surface or intramuscular electrodes.
rated action potential. Reciprocal Inhibition: Inhibition of a motor
QEMG: Abbreviation for quantitative elec- neuron pool secondary to the activation of
tromyography. the motor neuron pool of its antagonist.
QSART: Abbreviaton for quantitative sudo- It is one of several important spinal mech-
motor axon reflex test. anisms of motor control that help to make
QST: Abbreviaton for quantitative sensory movements smoother and utilize less en-
testing. ergy. There are multiple mechanisms for
Quantitative Electromyography (QEMG): A reciprocal inhibition, including one me-
systematic method for measuring the diated by the la inhibitory interneuron
recordings made by an intramuscular nee- that actives la afferents and disynaptically
dle electrode. Measurements include motor inhibits the muscle that is the antagonist
unit action potential characteristics such as to the source of the la afferents.
amplitude, duration, and phases, or interfer- Recording Electrode: Device used to record
ence pattern characteristics. See turns and electric potential difference. All electric
amplitude analysis. recordings require two electrodes. The elec-
Quantitative Sensory Testing (QST): An in- trode close to the source of the activity to
strumented method for measuring cuta- be recorded is called the active or explor-
neous sensation. ing electrode, and the other recording elec-
Quantitative Sudomotor Axon Reflex Text trode is called the reference electrode. Active
(QSART): Test of post-ganglionic sym- electrode is synonymous with input termi-
pathetic sudomotor axons function by nal 1, or E-l (or older terms whose use is
measuring sweat output following acti- discouraged, i.e., grid 1 and Gl). Refer-
vation of axon terminals by local appli- ence electrode is synonymous with input
cation of acetylcholine. Antidromic trans- terminal 2, or E-2 (or older terms whose
mission of the impulse from the nerve use is discouraged, i.e., grid 2 and G2). In
terminals reaches a branch point, then some recordings it is not certain which
travels orthodromically to release acetyl- electrode is closer to the source of the bi-
choline from the nerve terminals, in- ologic activity, for example, recording
ducing a sweating response. In small fiber with a bifilar needle recording electrode, or
polyneuropathy, the response may be re- when attempting to define far-field poten-
duced or absent. In painful neuropathies, tials. In this situation, it is convenient to
and in reflex sympathetic dystrophy, the refer to one electrode as input electrode
response may be excessive and persistent 1, or E-l, and the other as input electrode
or reduced. 2, or E-2. By present convention, a poten-
Rl, R2 Waves: See blink responses. tial difference that is negative at the active
Radiculopathy: Axonal and/or demyelinat- electrode (input terminal 1, E-l) relative
ing disorder affecting the nerve fibers ex- to the reference electrode (input terminal
clusive to one spinal nerve root or spinal 2, E-2) causes an upward deflection on the
nerve. May affect the anterior (motor) or display screen. The term "monopolar
posterior (sensory) spinal nerve roots, or recording" is not recommended, because
both, at one spinal cord segment level. all recordings require two electrodes;
The resulting clinical syndrome may in- however, it is commonly used to describe
clude pain, sensory loss, paresthesia, weak- the use of one type of intramuscular nee-
ness, fasciculations, and muscle atrophy. If dle electrode. A similar combination of nee-
more than one spinal root is involved, the dle electrodes has been used to record
term polyradiculopathy may be used as a nerve activity and also has been referred
descriptor. to as "monopolar recording."
Recruitment: The successive activation of the flexes involving autonomic effector or-
same and additional motor units with in- gans also occur. Examples include the H
creasing strength of voluntary muscle con- reflex and the sudomotor reflex. See H
traction. See motor unit action potential wave, quantitative sudomotor axon reflex test.
Recruitment Frequency: Firing rate of a mo- Refractory Period: General term for the
tor unit action potential (MUAP) when a time following an action potential when an
different MUAP first appears during grad- excitable membrane cannot be stimulated
ually increasing voluntary muscle contrac- to produce another action potential. The
tion. This parameter is essential to assess- absolute refractory period is the time follow-
ment of recruitment pattern. ing an action potential during which no
Recruitment Interval: The interdischarge in- stimulus, however strong, evokes a further
terval between two consecutive discharges of response. The relative refractory period is the
a motor unit action potential (MUAP) when time following an action potential during
a different MUAP first appears during which a stimulus must be abnormally large
gradually increasing voluntary muscle con- to evoke a second response. 'The functional
traction. The reciprocal of the recruitment refractory period is the time following an ac-
interval is the recruitment frequency. See also tion potential during which a second ac-
interdischarge interval. tion potential cannot yet excite the given
Recruitment Pattern: A qualitative and/or region.
quantitative description of the sequence Refractory Period of Transmission: Interval
of appearance of motor unit action potentials following an action potential during which
during increasing voluntary muscle con- a nerve cannot conduct a second one.
traction. The recruitment frequency and re- Distinguish from refractory period, as com-
cruitment interval are two quantitative monly used, which deals with the ability of
measures commonly used. See interference a stimulus to produce an action potential.
pattern, early recruitment, reduced recruitment Regeneration Motor Unit Potential: Use of
for qualitative terms commonly used. term discouraged. See motor unit action
Recurrent Inhibition: Decreased probability potential.
of firing of a motor neuron pool mediated Relative Refractory Period: See refractory
by Renshaw cells. Renshaw cells are acti- period.
vated by recurrent collaterals from the Repair of the Decrement: See facilitation.
axons of alpha-motoneurons. Such inhi- Repetitive Discharge: General term for the
bition influences the same cells that orig- recurrence of an action potential with the
inate the excitatory impulses and their same or nearly identical form. May refer
neighbors. to recurring potentials recorded in mus-
cle at rest, during voluntary contraction, or
Reduced Insertion Activity: See insertion ac- in response to a single nerve stimulus. See
tivity.
double discharge, triple discharge, multiple dis-
Reduced Interference Pattern: See interfer- charge, myokymic discharge, complex repeti-
ence pattern. tive discharge, neuromyotonic discharge, and
Reduced Recruitment Pattern: A descriptive cramp discharge.
term for the interference pattern when the Repetitive Nerve Stimulation: The tech-
number of motor units available to gener- nique of repeated supramaximal stimula-
ate a muscle contraction are reduced. One tion of a nerve while recording successive
cause for a reduced interference pattern. See M waves from a muscle innervated by the
interference pattern, recruitment pattern. nerve. Commonly used to assess the in-
Reference Electrode: See recording electrode. tegrity of neuromuscular transmission.
Reflex: A stereotyped motor response elicited The number of stimuli and the frequency of
by a sensory stimulus and a response. Its an- stimulation should be specified. Activation
atomic pathway consists of an afferent, sen- procedures performed as a part of the test
sory input to the central nervous system, at should be specified, for example, sus-
least one synaptic connection, and an ef- tained voluntary contraction or contraction
ferent output to an effector organ. The induced by nerve stimulation. If the test
response is most commonly motor, but re- includes an activation procedure, the time
elapsed after its completion should also be Scanning EMG: A technique by which a nee-
specified. For a description of specific pat- dle electrode is advanced in defined steps
terns of responses, see incrementing response, through muscle while a separate SFEMG
decrementing response, facilitation, and postac- electrode is used to trigger both the dis-
tivation depression. play sweep and the advancement device.
Repolarization: A return in membrane po- Provides temporal and spatial information
tential from a depolarized state toward die about the motor unit. Distinct maxima in
normal resting level. the recorded activity are considered to be
Residual Latency: The calculated time dif- generated by muscle fibers innervated
ference between the measured distal la- by a common branch of an axon. These
tency of a motor nerve and the expected la- groups of fibers form a motor unit fraction.
tency, calculated by dividing the distance Sea Shell Sound (Sea Shell Roar or Noise):
between the stimulating cathode and the Use of term discouraged. See end plate ac-
active recording electrode by the maximum tivity, monophasic.
conduction velocity measured in a more Sensory Latency: Interval between the onset
proximal segment of the nerve. It is due of a stimulus and the onset of the negative
in part to neuromuscular transmission deflection of the compound sensory nerve
time and to slowing of conduction veloc- action potential. This term has been used
ity in terminal axons due to decreasing di- loosely to refer to the sensory peak latency.
ameter and the presence of unmyelinated May be qualified as proximal sensory la-
segments. tency or distal sensory latency, depending
Response: An activity elicited by a stimulus. on the relative position of the stimulus.
Resting Membrane Potential: Voltage across Sensory Nerve: A nerve containing only sen-
the membrane of an excitable cell in the sory fibers, composed mainly of axons in-
absence of a stimulus. See polarization. nervating cutaneous receptors.
Rheobase: See strength-duration curve. Sensory Nerve Action Potential (SNAP): See
Rigidity: A velocity-independent increase in compound sensory nerve action potential.
muscle tone and stiffness with full range of Sensory Nerve Conduction Velocity: The
joint motion as interpreted by the clinical speed of propagation of action potentials
examiner from the physical examination. along a sensory nerve.
Often associated with simultaneous low-
Sensory Peak Latency: Interval between the
grade contraction of agonist and antagonist
onset of a stimulus and the peak of the neg-
muscles. Like muscle spasticity, the invol-
ative phase of the compound sensory nerve ac-
untary motor unit action potential activity in-
tion potential. Contrast with sensory latency.
creases with activity or passive stretch.
Sensory Potential: Synonym for the more
Does not seem to change with the veloc-
precise term, compound sensory nerve action
ity of stretch, and, on passive stretch, the
potential.
increased tone has a "lead pipe" or con-
stant quality. It is a cardinal feature of cen- Sensory Response: Synonym for the more
tral nervous system disorders affecting the precise term, compound sensory nerve action
basal ganglia. Contrast with spasticity. potential.
Rise Time: The interval from the onset of a SEP: Abbreviation for somatosensory evoked
polarity change of a potential to its peak. potential.
The method of measurement should be Serrated Action Potential: A waveform with
specified. several changes in direction (turns) which
Satellite Potential: A small action potential do not cross the baseline. Most often used
separated from the main motor unit action to describe a motor unit action potential. The
potential by an isoelectric interval which term is preferred to complex motor unit ac-
fires in a time-locked relationship to the tion potential and pseudopolyphasic action po-
main action potential. It usually follows, tential. See also turn and polyphasic action
but may precede, the main action poten- potential.
tial. Less preferred terms include late com- SFEMG: Abbreviation for single fiber elec-
ponent, parasite potential, linked potential, tromyography.
and coupled discharge. Shock Artifact: See artifact.
Short-Latency Reflex: A reflex with one train consists of a series of up to 10 or

(monosynaptic) or few (oligosynaptic) more potentials in which the individual
synapses. Used in contrast to long-latency components fire at irregular intervals. The
reflex. potentials consistently vary in amplitude,
Short-Latency Somatosensory Evoked Po- duration, and configuration. Individual po-
tential (SSEP): That portion of the wave- tentials may be mono-, bi-, tri-, or multi-
forms of a somatosensory evoked potential phasic in appearance; they often have a
normally occurring within 25 ms after positive waveform. The variation on se-
stimulation of the median nerve in the up- quential firings produces a distinctive
per extremity at the wrist, 40 ms after stim- sound, hence the name. See most often in
ulation of the common peroneal nerve in those with mesomorphic builds, especially
the lower extremity at the knee, and 50 young adult males. Found most often in
ms after stimulation of the posterior tibial lower extremity muscles, especially the
nerve at the ankle. medial gastrocnemius.
Signal Averager: A digital device that im- Somatosensory Evoked Potential (SEP):
proves the signal-to-noise ratio of an elec- Electric waveforms of biologic origin
trophysiological recording by adding elicited by electric stimulation or physio-
successive time-locked recordings to pre- logic activation of peripheral sensory nerves
ceding traces and computing the average and recorded from peripheral and central
value of each data point. A signal acquired nervous system structures. Normally is a
by this method is described as an "aver- complex waveform with several compo-
aged" waveform. nents which are specified by polarity and
Silent Period: A pause in the electric activ- average peak latency. The polarity and la-
ity of a muscle that may be produced by tency of individual components depend
many different stimuli. Stimuli used com- upon (1) subject variables, such as age,
monly in clinical neurophysiology include gender, and body habitus, (2) stimulus
rapid unloading of a muscle, electrical characteristics, such as intensity and rate
stimulation of a peripheral nerve or trans- of stimulation, and (3) recording param-
cranial magnetic stimulation. eters, such as amplifier time constants,
Single Fiber Electromyography (SFEMG): electrode placement, and electrode combi-
The technique and conditions that permit nations. See short-latency somatosensory evoked
recording of single muscle fiber action po- potentials.
tentials. See single fiber needle electrode, block- Spasticity: A velocity-dependent increase in
ing, and jitter. muscle tone due to a disease process that in-
Single Fiber EMG: See single fiber ekctromy- terrupts the suprasegmental tracts to the
ography. alpha motor neurons, gamma motor neu-
Single Fiber Needle Electrode: A needle elec- rons, or segmental spinal neurons. May be
trode with a small recording surface (usu- elicited and interpreted by the clinical ex-
ally 25 /mi in diameter) which permits the aminer during the physical examination
recording of single muscle fiber action po- by brisk passive movement of a limb at the
tentials between the recording surface and joint. Almost uniformly accompanied by
the cannula. See single fiber electromyogra- hyperreflexia, a Babinski sign, and other
phy. signs of upper motor neuron pathology,
Single Unit Pattern: See interference pattern. including clonus and the clasp-knife phe-
SNAP: Abbreviation for sensory nerve action nomenon. The clasp-knife phenomenon
potential. See compound sensory nerve action is a rapid decrease of tone following a pe-
potential. riod of increased tone during passive ro-
Snap, Crackle, and Pop: A benign type of in- tation of the joint. The pathophysiology is
creased insertion activity that follows, after a not certain and may include more than
very brief period of electrical silence, the dysfunction of the corticospinal tracts.
normal insertion activity generated by nee- Spike: (1) A short-lived (1-3 ms), all-or-none
dle electrode movement. It consists of trains waveform that arises when an excitable
of potentials that vary in length; however, membrane reaches threshold. (2) The elec-
they can persist for a few seconds. Each tric record of a nerve or muscle impulse.
Spinal Evoked Potential: Electric waveforms ulating electrodes are called bipolar if they
of biologic origin recorded over the spine are encased or attached together and are
in response to electric stimulation or phys- called monopolar if they are not. Electric
iologic activation of peripheral sensory stimulation for nerve conduction studies
fibers. See preferred term, somatosensory generally requires application of the cath-
evoked potential. ode in the vicinity of the neural tissue to
Spontaneous Activity: Electric activity reproduce depolarization.
corded from msucle at rest after insertion Stimulation Single Fiber Electromyography
activity has subsided and when there is not (Stimulation SFEMG): Use of electrical
voluntary contraction or an external stimu- stimulation instead of voluntary activation
lus. Compare with involuntary activity. of motor units for the analysis of single fiber
SSEP: Abbreviation for short-latency somatosen- electromyography. The method is used in pa-
sory evoked potential. tients who are unable to produce a steady
Staircase Phenomenon: The progressive involuntary muscle contraction. The stimula-
crease in muscle contraction force observed tion can be delivered to intramuscular ax-
in response to continued low rates of mus- ons, nerve trunks, or muscle fibers.
cle activation. Stimulus: Any external agent, state, or
Startle (Reflex): A response produced by an change that is capable of influencing the
unanticipated stimulus that leads to alert- activity of a cell, tissue, or organism. In
ing and protective movements such as eye clinical nerve conduction studies, an electric
lid closure and flexion of the limbs. Au- stimulus is applied to a nerve. It may be
ditory stimuli are typically most effica- described in absolute terms or with re-
cious. spect to the evoked potential of the nerve or
Stiff-man Syndrome: A disorder character- muscle. In absolute terms, it is defined by
ized by continuous muscle contraction giv- a duration (ms), a waveform (square, expo-
ing rise to severe stiffness. Axial muscles nential, linear, etc.), and a strength or in-
are typically affected most severely. Pa- tensity measured in voltage (V) or current
tients have difficulty moving. Walking and (mA). With respect to the evoked poten-
voluntary movements are slow. Sensory tial, the stimulus may be graded as sub-
stimulation often induces severe spasms. threshold, threshold, submaximal, maximal, or
Electromyography demonstrates continuous supramaximal. A threshold stimulus is one
activity of motor unit action potentials in a just sufficient to produce a detectable re-
normal pattern that cannot be silenced by sponse. Stimuli less than the threshold
contraction of the antagonist muscle. It is stimulus are termed subthreshold. The
often associated with circulating antibod- maximal stimulus is the stimulus intensity
ies to glutamic acid decarboxylase (GAD), after which a further increase in intensity
and the resulting deficiency of GABA may causes no increase in the amplitude of the
play a role in its pathophysiology. Since evoked potential. Stimuli of intensity be-
women are affected in equal or greater low this level but above threshold are sub-
numbers than men, the term stiff-person maximal. Stimuli of intensity greater than
syndrome may be preferable. the maximal stimulus are termed supra-
Stiff-person Syndrome: Synonym for stiff- maximal. Ordinarily, supramaximal stim-
man syndrome. uli are used for nerve conduction studies.
Stigmatic Electrode: A term of historic in- By convention, an electric stimulus of ap-
terest. Used by Sherrington for active or proximately 20% greater voltage/current
exploring electrode. than required for the maximal stimulus is
Stimulated SFEMG: See preferred term stim- used for supramaximal stimulation. The
ulation SFEMG. frequency, number, and duration of a series
Stimulating Electrode: Device used to deliver of stimuli should be specified.
electric current. All electric stimulation re- Stimulus Artifact: See artifact.
quires two electrodes; the negative terminal Strength-Duration Curve: Graphic presenta-
is termed the cathode, and the positive ter- tion of the relationship between the in-
minal is the anode. By convention, the stim- tensity (Y axis) and various durations (X
axis) of the threshold electric stimulus of a Synkinesis: Involuntary movement made by

nerve or muscle. The rheobase is the in- muscles distant from those activated vol-
tensity of an electric current of infinite du- untarily. It is commonly seen during re-
ration necessary to produce a minimal covery after facial neuropathy. It is due to
action potential The chronaxie is the time aberrant reinnervation and/or ephaptic
required for an electric current twice the transmission.
rheobase to elicit the first visible action po- T Wave: A compound muscle action potential
tential. Measurement of the strength- evoked from a muscle by rapid stretch of
duration curve is not a common practice its tendon, as part of the muscle stretch reflex.
in modern ekctrodiagnostic medicine. Tardy Ulnar Palsy: A type of mononeuropathy
Stretch Reflex: A reflex produced by passive involving the ulnar nerve at the elbow.
lengthening of a muscle. The principal The nerve becomes compressed or en-
sensory stimuli come from group la and trapped due to deformity of the elbow
group II muscle spindle afferents. It con- from a previous injury. See also cubital tun-
sists of several phases. The earliest compo- nel syndrome and ulnar neuropathy at the
nent is monosynaptic and is also called the elbow.
myotatic reflex, or tendon reflex. There Template Matching: An automated method
are also long-latency stretch reflexes. See used in quantitative ekctromyography for
also muscle stretch reflex, T wave. selecting motor unit action potentials for
Submaximal Stimulus: See stimulus. measurement by extracting only potentials
Subnormal Period: A time interval that im- which resemble an initially identified po-
mediately follows the supernormal period of tential.
nerve which is characterized by reduced Temporal Dispersion: Relative desynchro-
excitability compared to the resting state. nization of components of a compound
Its duration is variable and is related to the muscle action potential due to different rates
refractory period. of conduction of each synchronously
Subthreshold Stimulus: See stimulus. evoked component from the stimulation
Supernormal Period: A time interval that im- point to the recording electrode. It may be
mediately follows the refractory period which due to normal variability in individual
corresponds to a very brief period of par- axon conduction velocities, especially when
tial depolarization. It is characterized by in- assessed over a long nerve segment, or to
creased nerve excitability and is followed by disorders that affect myelination of nerve
the subnormal period. fibers.
Supraclavicular Plexus: That portion of the Terminal Latency: Synonymous with pre-
brachial plexus which is located superior to ferred term, distal latency. See motor latency
the clavicle. and sensory latency.
Supraclavicular Stimulation: Percutaneous TES: Abbreviation for transcranial electrical
nerve stimulation at the base of the neck stimulation.
which activates the upper, middle, and/or Test Stimulus: See paired stimuli.
lower trunks of the brachial pkxus. This Tetanic Contraction: The contraction pro-
term is preferred to Erb'spoint stimulation. duced in a muscle through repetitive max-
Supramaximal Stimulus: See stimulus. imal direct or indirect stimulation at a suf-
Surface Electrode: Conducting device for ficiently high frequency to produce a
stimulating or recording placed on the smooth summation of successive maxi-
skin surface. The material (metal, fabric, mum twitches. The term may also be ap-
etc.), configuration (disk, ring, etc.), size, plied to maximum voluntary contractions
and separation should be specified. See in which the firing frequencies of most or
electrode (ground, recording, stimulating). all of the component motor units are suffi-
Sympathetic Skin Response: Electrical po- ciently high that successive twitches of in-
tential resulting from electrodermal activ- dividual motor units fuse smoothly. Their
ity in sweat glands in response to both di- combined tensions produce a steady,
rect and reflex peripheral or sympathetic smooth, maximum contraction of the
trunk stimulation of autonomic activity. whole muscle.
Tetanus: (1) The continuous contraction of potential is initiated in a single axon or

muscle caused by repetitive stimulation or muscle fiber or a group of axons or mus-
discharge of nerve or muscle. Contrast with cle fibers.
tetany. (2) A clinical disorder caused by Threshold Stimulus: See stimulus.
circulating tetanus toxin. Signs and symp- Tic: Clinical term used to describe a sudden,
toms are caused by loss of inhibition in the brief, stereotyped, repetitive movement.
central nervous system and are character- When associated with vocalizations, may
ized by muscle spasms, hyperreflexia, be the primary manifestation of Tourette
seizures, respiratory spasms, and paralysis. syndrome.
Tetany: A clinical syndrome manifested by Tilt Table Test: A test of autonomic func-
muscle twitching, cramps, and carpal and tion that is performed by measuring blood
pedal spasm. These clinical signs are man- pressure and heart rate before and a spec-
ifestations of peripheral and central ner- ified period of time after head up tilt. The
vous system nerve irritability from several duration of recording and amount of tilt
causes. In these conditions, repetitive dis- should be specified.
charges (double discharge, triple discharge, TMS: Abbreviation for transcranial magnetic
multiple discharge) occur frequently with stimulation.
voluntary activation of motor unit action po- Tone: The resistance to passive stretch of a
tentials or may appear as spontaneous activ- joint. When the resistance is high, this is
ity. This activity is enhanced by systemic al- called hypertonia, and when the resistance
kalosis or local ischemia. is low, this is called hypotonia. Two types of
Tetraphasic Action Potential: Action potential hypertonia are rigidity and spasticity.
with three baseline crossings, producing Train of Positive Sharp Waves: See positive
four phases. sharp wave.
Thermography: A technique for measuring Train of Stimuli: A group of stimuli. The du-
infrared emission from portions of the ration of the group or the number of
body surface. The degree of emission de- stimuli as well as the stimulation frequency
pends upon the amount of heat produced should be specified.
by the region that is studied. Its use in Transcranial Electrical Stimulation (TES):
the diagnosis of radiculopathy, peripheral Stimulation of the cortex of the brain
nerve injury, and disorders of the auto- through the intact skull and scalp by
nomic nervous system is controversial. means of a brief, very high voltage, elec-
Thermoregulatory Sweat Test: A technique trical stimulus. Activation is more likely un-
for assessing the integrity of the central der the anode rather than the cathode. Be-
and peripheral efferent sympathetic path- cause it is painful, this technique has
ways. It consists of measuring the sweat dis- largely been replaced by transcranial mag-
tribution using an indicator powder while netic stimulation.
applying a controlled heat stimulus to raise Transcranial Magnetic Stimulation (TMS):
body temperature sufficient to induce Stimulation of the cortex of the brain
sweating. through the intact skull and scalp by
Thoracic Outlet Syndrome: An entrapment means of a brief magnetic stimulus. In
neuropathy caused by compression of the practice, a brief pulse of strong current is
neurovascular bundle as it traverses the passed through a coil of wire in order to
shoulder region. Compression arises from produce a time-varying magnetic field in
acquired or congenital anatomic varia- the order of 1-2 Tesla. Contrast with trans-
tions in the shoulder region. Symptoms cranial electrical stimulation.
can be related to compression of vascular Tremor: Rhythmical, involuntary oscillatory
structures, portions of the brachial plexus, movement of a body part.
or both. Triphasic Action Potential: Action potential
Threshold: The level at which a clear and with two baseline crossings, producing
abrupt transition occurs from one state to three phases.
another. The term is generally used to re- Triple Discharge: Three motor unit action po-
fer to the voltage level at which an action tentials of the same form and nearly the
same amplitude, occurring consistently in results in a characteristic pattern of heart

the same relationship to one another and rate and blood pressure changes that can
generated by the same axon. The interval be used to quantify autonomic function.
between the second and third action po- See Valsalva ratio.
tentials often exceeds that between the first Valsalva Ratio: The ratio of the fastest heart
two, and both are usually in the range of rate occurring at the end of a forced ex-
2-20 ms. See also double discharge, multiple halation against a closed glottis (phase II
discharge. of the Valsalva maneuver), and the slowest
Triplet: Synonym for the preferred term, heart rate within 30 seconds after the
triple discharge. forced exhalation (phase IV). In patients
Turn: Point of change in polarity of a wave- with disorders of the autonomic nervous
form and the magnitude of the voltage system, the ratio may be reduced.
change following the turning point. It is VEP: Abbreviation for visual evoked potential
not necessary that the voltage change pass VER: Abbreviation for visual evoked response.
through the baseline. The minimal excur- See visual evoked potential.
sion required to constitute a change Visual Evoked Potential (VEP): Electric
should be specified. waveforms of biologic origin recorded over
Turns and Amplitude Analysis: See pre- the cerebrum and elicited in response to
ferred term interference pattern analysis. visual stimuli. They are classified by stim-
Refers to the interference pattern analysis ulus rate as transient or steady state, and
developed by Robin Willison in the 1960s. they can be further divided by stimulus
Ulnar Neuropathy at the Elbow: A mononeu- presentation mode. The normal transient
ropathy involving the ulnar nerve in the re- VEP to checkerboard pattern reversal or
gion of the elbow. At least two sites of en- shift has a major positive occipital peak at
trapment neuropathy have been recognized. about 100 ms (PI00), often preceded by
The nerve may be entrapped or com- a negative peak (N75). The precise range
pressed as it passes through the retro- of normal values for the latency and am-
condylar groove at the elbow. Alterna- plitude ofPI 00 depends on several factors:
tively, it may be entrapped just distal to (1) subject variables, such as age, gender,
the elbow as it passes through the cubital and visual acuity, (2) stimulus characteris-
tunnel. Anatomic variations or deformi- tics, such as type of stimulator, full-field or
ties of the elbow may contribute to nerve half-field stimulation, check size, contrast
injury. See also cubital tunnel syndrome and and luminescence, and (3) recording pa-
tardy ulnar palsy. rameters, such as placement and combi-
Unipolar Needle Electrode: See synonym, nation of recording electrodes.
monopolar needle recording electrode. Visual Evoked Response (VER): Synonym
Upper Motor Neuron Syndrome: A clinical for preferred term, visual evoked potential
condition resulting from a pathological Volitional Activity: Synonymous with volun-
process affecting descending motor path- tary activity.
ways including the corticospinal tract or Voltage: Potential difference between two
its cells of origin. Signs and symptoms in- recording sites usually expressed in volts
clude weakness, spasticity, and slow and (V) or millivolts (mV).
clumsy motor performance. On electro- Volume Conduction: Spread of current from
myographic examination of weak muscles, a potential source through a conducting
there is slow motor unit action potential fir- medium, such as body tissues.
ing at maximal effort. Voluntary Activity: In electromyography, the
Utilization Time: See preferred term, latency electric activity recorded from a muscle
of activation. with consciously controlled contraction.
Valsalva Maneuver: A forcible exhalation The effort made to contract the muscle
against the closed glottis which creates an may be specified relative to that of a cor-
abrupt, transient elevation of intratho- responding normal muscle, for example,
racic and intra-abdominal pressure. This minimal, moderate, or maximal. If the
recording remains isoelectric during the the segment of an axon distal to nerve in-
attempted contraction and equipment jury that destroys its continuity.
malfunction has been excluded, it can be Waning Discharge: A repetitive discharge that
concluded that there is no voluntary gradually decreases in frequency or ampli-
activity. tude before cessation. Contrast with myo-
Wake-up Test: A procedure used most com- tonic discharge.
monly in spinal surgery. During critical Wave: A transient change in voltage repre-
portions of an operation in which the sented as a line of differing directions over
spinal cord is at risk for injury, the level time.
of general anesthesia is allowed to de- Waveform: The shape of a wave. The term
crease to the point where the patient can is often used synonymously with wave.
respond to commands. The patient is then Wire Electrodes: Thin wires that are insu-
asked to move hands and feet, and a move- lated except for the tips, which are bared.
ment in response to commands indicates The wire is inserted into muscle with a
the spinal cord is intact. This procedure needle. After the needle is withdrawn, the
is used routinely in some centers. Soma- wire remains in place. Wire electrodes are
tosensory-evoked potential monitoring has superior to surface electrodes for kinesiologic
supplanted its use in most centers, except EMG, because they are less affected by cross
sometimes in the situation where they in- talk from adjacent muscles. They also
dicate the possibility of spinal cord injury. record selectively from the muscle into
Wallerian Degeneration: Degeneration of which they are inserted.
INDEX
Acetylcholine, in synaptic transmission, 269, 269/ Anesthesia
270/ preoperative focal EEC abnormalities, 527-528
Acoustic neuroma symmetrical EEG patterns during, 527
brain stem auditory evoked potentials, 208, 209/ Anhidrosis
210-211 distribution patterns, 462, 462/
masseter reflex, 390, 390/ in elderly, 464
Acoustic reflex Annotation, of digital electroencephalograms, 43-44
decay, 217 Anomalous innervation, nerve conduction studies,
definition of, 216 597
measurement, with immittance unit, 216-217, 217/ Anoxia, outcome prediction, somatosensory evoked
testing, 214 potentials and, 198
Action potentials Anterior horn cell disorders, single fiber
mathematical model, 56, 56/ electromyography, 356
measurement, voltage clamp technique for, 55-56 Anterior temporal spike (sharp wave), 91, 91/
motor unit. See Motor unit potentials Antiepileptic drugs, 146
muscle, 59 Aortic arch baroreceptors, 451
propagation velocity, 55 Apnea
Action tremors, 402 central, 504, 507, 507/
Active zones, 58 definition of, 504
Adolescent developmental changes, mixed, 507
electroencephalography, 113-115, 114/ 115/ obstructive, 504, 507/
Adrenergic function assessment, by skin vasomotor Apnea-hypopnea index (AHI), 509
reflexes, 451-456, 453/ 453*, 454*, 455/ Arousals
AEEG. See Ambulatory electroencephalography number of, 510
Afterdepolarizations, 58 respiratory effort-related, 504
Afterhyperpolarizations, 58 Arousals, EEG, 500
Age, compound muscle action potentials and, 249 Arterial baroreflexes, 441-442
Ageotropic positional nystagmus, 420 Artifacts
AHI (apnea-hypopnea index), 509 electric, SEP amplitude and, 182
Airflow measurement, during sleep, 507-508 in electroencephalography, 82
Aliasing, 47, 47/ muscle, SEP amplitude and, 182
Alpha activity, 82 nonphysiologic, 73-74, 74/
Alpha-delta sleep, 500, 501/ physiologic, 72-73, 72*, 73/
Alpha-frequency coma pattern, 103 Asymmetry
Alpha intrusions, 500, 50 If neonatal, 111
Alpha rhythm, 82-83, 83/ quantitiative measurement, 138
Alpha variants, 87 Athetosis, surface electromyography, 411
Alternation, 364, 364/ Atypical spike-and-wave pattern, 93
Alzheimer's disease, P300 abnormalities, 164 Audiogram, 215-216, 215/
Ambulatory electroencephalography (AEEG) Auditory evoked potentials. See Brain stem auditory
clinical applications, 126-127 evoked potentials
duration of study, 126-127 Auditory pathway, 61
indications, 124-125 Auditory pattern recognition, in needle
limitations, 125-126 electromyography, 298
technical aspects, 125-126 Auditory system
Aminoacidurias, EEC abnormalities, 112-113 anatomy, 205
Amplification, 15 evaluation
Amplifiers audiogram for, 214, 215-216, 215/
differential, 15-16, 15/ evoked potentials. See Brain stem auditory evoked
for single fiber electromyography, 345 potentials
Amyotrophic lateral sclerosis physiology, 205
compound muscle action potentials, 257, 257* symptoms, with vertigo, 414—415
diagnosis, 591 Autocorrelation analysis, 50
motor evoked potentials, 288 Autoimmune myasthenia gravis, single fiber
motor unit number estimate, 368, 369/ electromyography, 353-354
Analog signals, 44-47, 45/ 46/ Automatic decomposition electromyography, 337-339
629
630 Index
Autonomic axons, 59 in infinite homogenous media, 30-34, 31/-33/

Autonomic nervous system, 443 in nonhomogenous media, 34—36, 34/j 35/
anatomy, 437 sources
visceral afferents, 438 active, 28
visceral efferents, 438-439 dipole, 30-31, 31/
cardiovascular reflexes, 441-442 monopole, 30, 31/
diseases of, 437 passive, 28, 29
evaluation, 474 quadrupole, 31-32, 32/
pupil cycle time, 474 spatial distributions from, 32-34, 32/ 33/
Valsalva maneuver for. See Valsalva maneuver spatial distributions, 32-34, 32/ 33/
function testing, 569, 571 volume conductor and, 28
heart rate control, 441 Bizarre repetitive potentials (complex repetitive
laboratory evaluation of. See also Quantitative discharges), 310, 310/ 311*
sudomotor axon reflex test Blink reflex
indications for, 445, 446* for disease localization in posterior fossa, 570
pain tests, 481-483 neuroanatomy, 384, 384/
postural normotension, maintenance of, 442-443 testing
power spectrum analysis, 470 caveats, 388-389
sympathetic nervous system clinical applications, 384-389, 385/-3S9/
assessment in humans, 440 pitfalls, 388-389
function, 439-440 responses, 383-384
functional anatomy of outflow, 439 stimulation method, 382-383, 383/
testing, 437-438 Blocking measurement, by single fiber
Averaging electromyography, 344, 349
digital, 48-49 Blood gas analysis, during sleep, 508-509, 509/
methods, for somatosensory evoked potentials, 182 Blood pressure, beat-to-beat response
A waves, 246, 248/ to tilt-up, 452-454
Axial spasms, of propriospinal origin, 408-409 to Valsalva maneuver, 451-452, 453/ 453*, 454*
Axonal disruption/degeneration, 572-573, 573*, 574* Body position, disordered breathing in sleep and,
Axonal neuropathies. See also Amyotrophic lateral 501-502
sclerosis Bony spine disease, motor evoked potentials, 287
compound muscle action potentials, 256-257, 257* Booster zones, 58
nerve action potentials, 171 Botulism
nerve conduction studies, 597-598 repetitive stimulation, 278-279
Axons single fiber electromyography, 354-355
autonomic, 59 BPPV (benign paroxysmal positioning vertigo),
motor, 59 417-419, 418/
reflexes, 246-247 Brachial plexopathy
sensory, 59 electromyography, 588
repair, intraoperative somatosensory evoked
potentials, 559-560, 559/ 560/
Brachial plexus lesions, compound muscle action
/3-Adrenergic blockade, heart rate response to deep potentials, 264
breathing and, 468 Brain abscess, electroencephalography in children,
BAER. See Brain stem auditory evoked potentials 120
Balance mechanisms, 413 Brain death
Ballistic movement overflow, 408 brain stem auditory evoked potentials, 212
Baroreceptors, 451 electroencephalographic criteria, 104-105, 104/
Barorefiex indices, 456 somatosensory evoked potentials, 199
Baroreflexes, arterial, 441-442 Brain stem
Battery, 4 intraoperative monitoring
Benign epileptiform transients of sleep (BETS), 89, clinical applications, 535-537, 535/-£37/
89/ 89* electromyography for, 532-533, 533/
Benign paroxysmal positioning vertigo (BPPV), evoked potentials for, 534-535, 534/
417-419, 418/ lesions
Benign rolandic epilepsy of childhood (BREC), 91-92 acoustic reflex, 217-218, 218*, 219/
Benign sporadic sleep spikes (BSSS), 89, 89/ 89* audiogram, 217-218, 218*
Benign temporal slow transients, 85 brain stem auditory evoked potentials, 211-212,
Bereitschaftspotential, 161 21 If, 212/
Beta activity, 83, 83/ somatosensory evoked potentials, 196-197, 196/
Beta-frequency coma pattern, 103 as structural generator, 60
BETS (benign epileptiform transients of sleep), 89, Brain stem auditory evoked potentials (BAEPs)
89/ 89* advantages, 208
Bioelectric potentials. See also specific bioelectric age-related effects, 206, 206/
potentials clinical applications, 207-212, 208/-212/
calculating acoustic neuroma, 208, 209/ 210-211
Index 631
brain death, 212 Cauda equina surgery, intraoperative monitoring, 552

coma, 212 Cavernous sinus tumor removal, intraoperative
demyelinating disease, 211 monitoring, 535, 535/
intrinsic brain stem lesions, 211-212, 21 If, 212/ CDP (computerized dynamic posturography),
multiple sclerosis, 208/-210/ 209-210, 211 430-431, 431/
peripheral acoustic disorders, 224-226, 224/-22G/ Central apnea, 504, 507, 507/
for disease localization in posterior fossa, 570 Central control disorders, needle electromyography,
electrodes, 224, 224/ 319-320, 321/
generators, 206 Central nervous system
indications, 204, 208 disorders. See also specific central nervous system
interpeak intervals, 223-224 disorders
interpretation, 223-224 somatosensory evoked potentials, 194-200,
for intraoperative monitoring, 534-535, 536, 537/ 194/-197/
latencies excitabilty measurement, H reflex for, 379-380
absolute, 223 lesions, causing vertigo, 414
interaural differences, 223, 224 Central-temporal spikes, in children, 91-92, 117-118,
methods, 207 117/
in normal subjects, 205-207, 205/ Cerebellar disease, movement-related cortical
stimulus, 204-206, 222-223 potentials, 163
wave II, 61 Cerebellar tremor, 403
waveforms, 206 Cerebellopontine angle tumor removal, intraoperative
Brain stem auditory evoked response (BAER). See monitoring, 536, 536/ 537/
Brain stem auditory evoked potentials Cerebral blood flow monitoring, during carotid
Brain stem myoclonus, 405 endarterectomy, 529-530
Breach rhythm, 87-88, 88/ Cerebral cortex
Breathing, sleep-disordered, 513 anatomy, 61
BREC (benign rolandic epilepsy of childhood), 91-92 function monitoring, 530-531
BSSS (benign sporadic sleep spikes), 89, 89/ 89* during anesthesia, 527
Burst-suppression pattern during cardiac surgery, 530
after cardiac arrest, 103-104, 104/ during carotid endarterectomy, 528-530
in neonate, 111 preoperative focal abnormalities, 527-528
technical factors, 523-524, 524f 526/
postsynaptic potentials, 61
stimulation methods, 282-283, 284/
Calcium-dependent potassium channels, 58 as structural generator, 61-62. See also
Caloric irrigation, 423-425, 423f426/ 427* Electroencephalography
Canalith repositioning procedure, 418-419 synaptic activation, current flow from, 29-30, 29/
Capacitance, 5 Cerebral hemorrhage, somatosensory evoked
Capacitors potentials, 199
properties of, 4-5 Ceroid lipofuscinosis, electroencephalography, 82
reactance, 11 Cerumen plug, diagnosis, 218<
rules for, 5-6, 6/ Cervical radiculopathies, electromyography, 584-585
Cardiac arrest, burst-suppression pattern and, Cervical spine surgery, intraoperative monitoring of,
103-104, 104/ 548, 549, 549/ 550/
Cardiac surgery, intraoperative Cervical spondylotic myelopathy, somatosensory
electroencephalographic monitoring, 530 evoked potentials, 194-195, 194/
Cardiopulmonary reflexes, 442 Cheater adapter, 24
Cardiovagal function testing Chemical driving force, for resting membrane
cold face test, 469-470 potential, 53-54
diving reflex, 469-470 Chiasmatic lesions, visual evoked potentials, 231-233,
heart rate response 231/
to deep breathing, 467-469, 469/ Children. See also Infants
to standing, 469 central-temporal spikes in, 91-92, 117-118, 117/
Cardiovascular reflexes, 441-442 comatose, somatosensory evoked potentials, 198-199
Carotid baroreceptors, 451 electroencephalography
Carotid endarterectomy abnormalities, 116-122, 116f 122/
focal electroencephalographic changes during, benign variants, 115-116
528-529 developmental changes, 113-115, 114/ 115/
intraoperative monitoring techniques, 529-530 in disease. See specific diseases
shunt placement, 529 H reflexes in, 378
Carotid sinus baroreceptors, 443 spine disease surgery, intraoperative monitoring of,
Carotid stump pressure measurement, during carotid 548
endarterectomy, 530 Chorea, surface electromyography, 411
Carpal tunnel syndrome Chronic inflammatory demyelinating
compound muscle action potentials, 260-261, 261/ polyradiculoneuropathy, masseter inhibitory
electromyography, 586-587 reflex, 391-392, 391/
632 Index
Chronic intracranial electroencephalographic focal neuropathies, 258-265, 259*, 260*, 261/

monitoring 263/ 264/
depth electrode studies, 152-155, 152/-154/ peripheral nerve injuries, 251-252, 252*, 253*
indications, 151-152, 152* peripheral neuropathies, 255-258, 257*, 258*
prognostic importance, 156 F-wave latency, 249*
subdural electrode studies, 155-156, 155/ F waves, 245-248, 24S/-248/
Chrono-dispersion, 246 initial positivity preceding negative M wave,
CIDP. See Chronic inflammatory demyelinating 595-596
polyradiculoneuropathy measurements, 242-245
Circuits amplitude, 242
filter, 12-13, 12/ conduction velocity, 243, 243/
inductive-capacitive, 8-10, 9/ distal latency, 243
inductive-resistive-capacitive, 10, 10/ errors in, 243-245, 244/
resistive-capacitive, time constant and, 7, If latency, 242-243
resistive-inductive, time constant and, 7, 8/ normal values, 245
root-mean-square potentials of, 10-11 size and configuration, 243, 244/
Clamping, intraoperative, EEG changes during, 524, normal values, 249
526/ number/size of muscle fibers and, 360-361
Click sensation level, 205 clinical applications, 237, 238
Click stimuli, for brain stem auditory evoked in disease states, 250-251, 265
potentials, 222-223 physiologic variables, 249-250
Clinical neurophysiology quantitative MUNE and, 362-365, 363/ 364/
in disease assessment, 566-567 recording methods, 556
for evaluation of severity, 567-568 electrodes for, 238-239
for localization, 570-571 location of recording electrodes, 239, 239/ 240/
of symptom complexes, 568-569 stimulation, 240-242
for identifying disease types, 571-572 electrodes for, 240-241
for prognostic prediction, 572-575, 573*-^577* location of electrodes for, 241-242
techniques. See specific clinical neurophysiologic Compound nerve action potential, 29
techniques Computerized dynamic posturography (CDP),
Closed-field potentials, 29, 29/ 430-431, 431/
CMAPs. See Compound muscle action potentials Computerized rotatory chair test, 428, 429/ 430, 430/
C nociceptors, sensitized, 481, 48 If Concomitant blocking, in single fiber
Cochlear disorders electromyography, 352, 352/
brain stem auditory evoked potentials, absolute Conduction block, 572, 573, 573*, 574*
latencies, 223 Conduction velocity
diagnosis, 217, 218* calculation, 243, 243/
Cognitive dysfunction, electroencephalography, nerve action potentials, 178, 179/
579-580 slowing, 572, 598
Cogwheel pursuit, 421 Conductive hearing loss, 216, 222
Coherence function, 140 Conductor, 4
Coils (inductors), 5 Cone of stability, 431, 431/
Cold face test, 469-470 Cone-shaped waves, 115
Cold presser test, 439 Congenital abnormalities. See also specific congenital
Collector, 14 abnormalities
Coma EEG abnormalities, 113
brain stem auditory evoked potentials, 212 Consciousness impairments
electroencephalography, 101, 103, 103/ assessment of, 569
prognosis, somatosensory evoked potentials, electroencephalography, 578-579
198-199 Contingent negative variation testing, 164
Combined sensory index (CSI), 260 Continuous ambulatory EEG monitoring, for sleep
Common mode, 16 disorder assessment, 514
Complex regional pain syndrome (CRPS), 481-482, Core temperature, assessment during sleep, 511
484 Cortical blindness, visual evoked potentials, 231/J
Complex repetitive discharges, 298, 310, 310/ 311* 233-234
Compound muscle action potentials (CMAPs) Cortical malformations, EEG abnormalities, 113, 113/
after CNS motor structure stimulation. See Motor Cortical myoclonus, 406-407
evoked potentials Cortical projection techniques, 143
amplitude, 344 Cortical reflex myoclonus, 406
after nerve injury, 573, 573* Cortical tremor, 404-407
differing between proximal and distal Corticectomy, 146
stimulation, 595 Coulomb, 3
low or absent, 595 Coupling discharges. See Satellite potentials
in disease states. See also specific disease states Cramp potentials, 312-313/ 315/ 315*
conduction block, 251 Cranial fossa tumor removal, intraoperative
conduction slowing, 250-251, 252 monitoring, 535-537, 535/-537/
Index 633
Cranial nerves segmental, compound muscle action potentials,

intraoperative monitoring, 532-537 257-258, 258*
clinical applications, 535-537, 535/-^537/ somatosensory evoked potentials, 195, 195/
evoked potentials, 534-535, 534/ visual evoked potentials, 231-232, 231/
intraoperative nerve conduction studies, 533-534 Dendritic spikes, 58
lesions Denervation supersensitivity, of skin vasomotor
acoustic reflex, 217, 218* reflexes, 455
acoustic reflex decay and, 217, 218< Dermatomal nerve stimulation, SEP amplitude and,
audiogram, 2181 183
brain stem auditory evoked potentials, 224-225, Dermatomal somatosensory evoked potentials, 189
224/-22G/ Diabetic peripheral neuropathy, thermoregulatory
reflexes sweat test, 458-459, 459/
blink. See Blink reflex Differential amplifiers, 15-16, 15/
masseter inhibitory, 391-392, 391/ Diffuse demyelination, nerve action potentials, 171
C response, 406 Digital computers, in clinical neurophysiology, 41-44,
Creutzfeldt-Jakob disease, electroencephalography, 43*
82, 101, 102/ Digital electroencephalography. See
Cross-correlation analysis, 139-140, 140/ Electroencephalography, digital
Cross-correlation function, 139-140, 140/ Digital filters, 49-50, 50/
Cross-spectral analysis, 140, 141/ Digital recording technology
CRPS (complex regional pain syndrome), 481-482, advantages, 42
484 disadvantages, 42
CSI (combined sensory index), 260 for electroencephalography. See
Cumulative amplitude, 334 Electroencephalography, digital
Current systems, construction of, 44
definition of, 4, 10-11 Digital signal processing
leakage. See Leakage currents averaging, 48-49
physiologic effects, 19-20, 19/ usage, 47-48
requirements to flow through body, 18-19 Digital signals, 44
Current density, 4 Digitization
Current-limiting devices, 23, 23/ principles, 44
Cutaneous nerve stimulation process, 44-45
SEP amplitude and, 183 quantization and, 45-46, 45/ 46/
somatosensory evoked potentials, 189 sampling and, 46-47
Cutaneous reflex, 396, 396/ Diodes, 14, 14/
Cut-off frequency, of low-pass filter, 13 Dipole
definition of, 30-31, 31/
source, localization of, 39-40, 39/
Direction-changing nystagmus, 422
Daytime somnolence, excessive, 513, 516-517 Direction-fixed nystagmus, 422
DCV (dorsal column volley), 187 Discharges. See also specific types of discharges
Deafness, 216 event recording, 66-67, 67/
Deep breathing, heart rate response to, 467-469, 469/ Discrete data, 51
Degenerative disorders, in children, Disordered breathing index, 509
electroencephalography of, 118-119, 119/ Distal anhidrosis, 462
Dejerine-Sottas disease, 258 Distortion product otoacoustic emissions (DPOAEs),
Delirium, electroencephalography, 578-579 218-220, 220/
Delta activity Diving reflex, 469-470
definition of, 83-84 Dix-Hallpike test, 417-418, 418/ 427*
intermittent rhythmic, 99-100, 100/ Dizziness, causes of, 413
polymorphic focal Doping, semiconductors and, 13-14
of focal intracranial lesions, 96-98, 97/ 100, 100/ Dorsal column volley (DCV), 187
sequential alterations of, 100, 100/ Dorsal rhizotomy, intraoperative monitoring,
slow 551-552
in diffuse encephalopathies, 101, 101/ Doublets, 318-319, 320/ 320*
neonatal, 112 DPOAEs (distortion product otoacoustic emissions),
temporal intermittent rhythmic, of focal 218-220, 220/
intracranial lesions, 98-99, 99/ Drowsiness
Demyelination diseases. See aho specific demyelination electroencephalography, 85
diseases benign variants, 88, 88/
brain stem auditory evoked potentials, 211 children, 114, 115/
mechanisms of slow conduction in, 250-251 rhythmic temporal theta bursts of, 88, 88/
motor evoked potentials, 286-287, 287/ Drugs. See also specific drugs
multifocal motor conduction block, 258 causing vertigo, 415
neuropathy EEG abnormalities, neonatal, 112
nerve conduction studies, 597 withdrawal, neonatal EEG abnormalities and, 112
634 Index
Duration of straining, effect on Valsalva maneuver, in elderly, 84-85

472 in sleep state, 85-86, 85/ 86/
D waves, 283 ambulatory, 124-127
Dying-back neuropathies, 251 amplification and, 15
Dynamic visual acuity, 410 artifacts, 82
Dynamic walking tests, 419 assessment, 575, 578
Dysgenetic disorders, EEG abnormalities, 113 of cognitive dysfunction, 579-580
Dystonia, surface electromyography, 409-411, 410/ of cognitive impairment, 569
of consciousness impairment, 569
of delirium, 578-579
of impaired consciousness, 578-579
ECoG. See Electrocorticography of motor symptoms, 568
EEG. See Electroencephalography of paroxysmal disorders, 580-581
E:I ratio, 468 of seizures, 580-581
Elderly, anhidrosis in, 464 of sensory symptoms, 569
Electric artifact, SEP amplitude and, 182 technical considerations, 578
Electric charges, 3. See also Current asymmetry, neonatal, 111
Electric field, 3 benign variants, 87-90, 87/-90/ 87t, 89*, 90*
Electric force, 3 in children, 115-116
Electric potentials, 4. See also Bioelectric potentials during drowsiness, 88, 88/
Electric power distribution systems, 17, 18/ during sleep, 88-90, 88/-90/ 89*
Electric shock, 17-20 during wakefulness, 87-88, 87/ 87*, 88/
Electrical driving force, for resting membrane brain death criteria, 104-105, 104/
potential, 53-54 children
Electrical safety abnormalities, 116-122, 116/^122/
equipment grounding and, 24—25, 26/ benign variants, 115-116
procedures for technicians, 27 developmental changes, 113-115, 114/ 115/
rules for, 25-27 chronic intracranial monitoring, 151-156,
Electrical stimulation, for motor evoked potentials, 152/-156/ 152*
283-284 clinical utility, 77-78
Electricity of cortical processes. See Movement-related cortical
circuit analysis, 5-6, 6/ potentials
principles of, 3-5 definition of, 78
Electrocardiogram artifacts, physiologic, 72-73 digital
Electrocochleogram, 61 capabilities, 42-44, 43*
Electrocorticography (ECoG) reliability of interpretation, 43, 43*
in epilepsy surgery, 157-159, 157/j 158/ for disease localization, at supratentorial level, 570
limitations, 158 for disease type identification, 571
prognostic importance of, 159 display of activity, 78, 78/-80/ 80
Electrodecremental pattern, 93 in epilepsy monitoring unit, 583
Electrodes extracranial
bipolar concentric, 295/ 296 limitations of, 149
macroelectrodes, 296 for partial epilepsy, 148-151, 148/-151/
monopolar, 295-296, 295/ 305, 327 focal changes
needle during carotid endarterectomy, 528-529
for compound muscle action potentials, 238-241 preoperative, 527-528
for electromyography, 295-296, 295/ ictal, depth electrode studies, 153-155, 153/ 154/
for single fiber electromyography, 344, 345 in intensive care unit, 581
recording interictal discharge patterns, 580-581
for compound muscle action potentials, 238-239, intraoperative monitoring, 581-582
239/ 240/ during anesthesia, 527
for quantitative electromyography, 326-328, technical factors, 523-524, 524y^526/
326/-S28/ isoelectric, 111
for spinal cord intraoperative monitoring, 541, magnetoencephalography, 144
542/ neonatal, 582-583
single fiber, 295/ 296, 344, 345 abnormal patterns, 109-113, Illf-I13f
standard concentric, 295, 295/ normal patterns, 109, 110/
stimulating referrals for, 108-109
for compound muscle action potentials, 240-242 pathologic activity. See also under specific pathologic
for spinal cord intraoperative monitoring, 541, disorders
542/ in children, 116-122, 116/^122/
Electroencephalography (EEG) distinctive epileptiform patterns, 91-94, 91/ 94/
activation procedures, 80-82, 81/ 95/
adult activity, normal, 61-62 in neonate, 109-113, llljMIS/
in awake state, 82-85, 83/ 84/ nonspecific, 90
in drowsiness state, 85 preoperative focal abnormalities, 527-528
Index 635
for prognostic prediction, 572 EOAEs (evoked otoacoustic emissions), 214, 218-220
prolonged video. See Prolonged video Epilepsy
electroencephalography ambulatory electroencephalography, 126-127
quantitative methods differential diagnosis, prolonged video EEC for,
cortical projection, 143 131
cross-correlation analysis, 139-140, 140/ electroencephalography, 120, 580-581
cross-spectral analysis, 140, 141/ lesional, epilepsy surgery for, 147
Fourier analysis, 137-138 medial temporal lobe, epilepsy surgery for, 147
interpolation techniques, 140-141 monitoring unit, electroencephalography in, 583
montage reformatting, 138-139, 139/ partial, extracranial EEC for, 148-151, 148>151/
multivariate statistical methods, 142-143 partial or focal or localization-related, 146
for seizure detection, 138 treatment
source dipole localization, 143-144, 143/ antiepileptic drug, 146
for spike and sharp-wave detection, 138 surgical. See Epilepsy surgery
topographic mapping, 141-142, 142/ Epilepsy surgery
volume conduction principles, 36-39, 38/ candidate evaluation, 147-148
waveform alterations, 69 ambulatory EEC for, 127
Electromagnet, 5 prolonged video EEC for, 133-134
Electromotive force (EMF), 4 contraindications, 147-148
Electromyography (EMG) corticectomy, 146
amplification and, 15 efficacy, 146
artifacts, physiologic, 73 electrocorticography in, 157-159, 157/ 158/
brain stem intraoperative monitoring, 532-533, outcome, 147
533/ preoperative evaluation
cervical radiculopathies, 584-585 chronic intracranial EEC monitoring, 151-156,
cranial nerve intraoperative monitoring, 532-533, 152/-156/ 152*
533/ extracranial electroencephalography for,
for disease assessment, 583-584 148-151, 148f 151/
focal neuropathies, 574, 575* presurgical evaluation, 147-148
intraoperative monitoring, 536, 536/ prognosis
of entrapment neuropathies, 557-558 chronic intracranial monitoring and, 156
spinal cord, 540*, 546-547, 547/ electrocorticography and, 159
lumbosacral radiculopathies, 584, 585 Epileptic myoclonus, 405, 406
motor neuron disease, 574, 577* Epileptiform discharges
motor unit number estimate, 359-360 abnormal, in children, 116-118, H6J-118f
for peripheral nervous system monitoring, 556 ictal, 93-94, 94/ 95/
peripheral neuropathies, 585-589 interictal
potentials, origins of, 299, 299/ diffuse, 92-93
radiculopathies, 574, 575*, 576* focal, 91-92, 91/
semiquantitative, 298 neonatal, 111, 111/
single potential changes, 69 periodic lateralized, 99, 99/
for weakness evaluation, 589-595 Epileptogenic zone, localization, 147-148
Electronic devices. See specific electronic devices EPP (end plate potential), 269-270
Electronystagmography test battery EPSPs (excitatory postsynaptic potentials), 29, 57
caloric irrigation, 423-425, 42S/426/ Equipment grounding, 21, 24-25, 26/
gaze testing, 420-421, 427* Equipotential grounding system, 21
oculomotor testing, 421 Ergoreflexes, 442
positional testing without fixation, 421-423, 422/ Esophageal pH, measurements during sleep, 511
preparation for, 420 Essential myoclonus, 405, 408
Electro-oculographic recordings, of nystagmus, 415, Essential tremor, 403
416/ Event recording, 66-68,' 66/ 67/
Electroretinogram, 60-61 Event-related potentials, 164
Electrotonic membrane potential, 54-55, 55/ Evoked otoacoustic emissions (EOAEs), 214, 218-220
EMF (electromotive force), 4 Evoked potentials. See also specific types of evoked
EMG. See Electromyography potentials
Emitter, 14 averaging, 48-49, 48/
Empyema, electroencephalography in children, 120 brain stem intraoperative monitoring, 534-535,
Encephalitis, in children, electroencephalography of, 534/
119-120 cranial nerve intraoperative monitoring, 534-535,
Encephalopathy, diffuse, 100-101, 101j^l04/ 103-104 534/
End plate noise, 300, 300/ for disease localization at supratentorial level, 570
End plate potential (EPP), 269-270 Exaggerated startle, 408
End plate spikes, 59, 300, 300/ Excessive daytime somnolence, 516-517
End plate zone, 326 Excitatory postsynaptic potentials (EPSPs), 29, 57
Entrapment neuropathies, intraoperative monitoring, Expiratory pressure, effect on Valsalva maneuver,
557-558 472
636 Index
Extrapyramidal disease size, electroencephalography and, 95

blink reflex, 388 stage, electroencephalography and, 95-96
thermoregulatory sweat test, 459 Focal neuropathies. See Mononeuropathies
Extrapyramidal myoclonus, 405 Focal sweat loss, 462
Eye movement Forward biasing, 14
blinks, posterior slow waves of, 115 Forward problem, 39
in nystagmus, 415, 417/ Fourier analysis (spectral), 50-52, 137-138
opening/closing, as EEG activation procedure, 80 Frequency, of continuous waveforms, 63-64, 64/ 65/
Frequent stimulus, 164
Friedreich's ataxia, somatosensory evoked potentials,
196, 196/
Facial nerve conduction studies, for disease Frontal seizures, 133
localization, 570 Frontal spike discharge, 91
Facial nerve lesions, blink reflex, 385-386, 385/ Fukuda step test, 419
Facial paralysis, diagnosis, 217, 218* Functional myoclonus, 409
Facial synkinesis assessment, blink reflex for, 386, Functional tremor, 404-405
386/ 387/ Fundamental, 51
Facial weakness, electromyography, 593 F waves
Familial infantile myasthenia, repetitive stimulation, definition of, 245
279 latency, 245-247, 24S/-247/
Familial tremor, 403 limb length and, 248
Far-field potentials (FFPs), 184-185, 207 normal values, 2491
Fasciculation potentials, in needle electromyography, measurements, for quantitative MUNE, 365
310-311, 31 If, 31 It recording method, 246-248, 247/ 248/
Fast alpha variant, 87 repeater, 257
Fast spike-and-wave, 89, 90/ 90* stimulation method, 246
Feedback data flow, 49 vs. H reflex, 376-377, 377*
Feedforward data flow, 49
Femoral neuropathy, nerve action potential, 174—175
Fencing, 133
FFPs (far-field potentials), 184-185, 207 Gastrocnemius muscle, H reflex testing technique,
Fibrillation, 300, 573 377-378
Fibrillation potentials Gaze testing, 416, 420-421, 427*
definition of, 299, 573 G2 electrode (inactive), 239
in denervated muscle, 309, 309/ Generators
density, 309, 309/ for brain stem auditory evoked potentials, 206
diseases associated with, 3081 definition of, 4
positive waveform, 308, 308/ physiologic, 53-58, 54*, 55/-5S/ See also Membrane
spike form, 308, 308/ potentials
Filters action potentials, 55-56, 56/
digital, 49-50, 50/ electro tonic membrane potential, 54—55, 55/
high-pass, 12-13, 12/ postsynaptic potentials, 56-57
low-pass, 13 resting membrane potential, 53-54, 54*
settings, SEP amplitude and, 182-183 propagating, production of stationary potentials,
Final common pathway, 291 36, 37/
Finite impulse response filters, 49 structural, 58-62
FIRDA pattern, 527 Geotropic positional nystagmus, 420
First-night effect, 502 Global anhidrosis, 462
Flexor carpi radialis, H reflex testing technique, 378 G-protein-activated inward rectifying potassium
Flexor reflex, 396 channels (GIRKs), 467
Flexor reflex afferents, 396 Gray matter disease, in children,
Focal conduction block, nerve conduction studies, electroencephalography of, 119
598 Grid, 15
Focal cortical myoclonus, 406 Ground, 16, 18
Focal demyelinating lesions, nerve action potentials, Ground loops, 21, 22/
171 Grounding, of electrical equipment. See Equipment
Focal intracranial lesions grounding
density, electroencephalography and, 95 Guillain-Barre syndrome
electroencephalography compound muscle action potentials, 255, 256
accuracy, 96 somatosensory evoked potentials, 191-192, 192/
local effects, 96-98, 97/
regional effects, 96, 98-99, 99/
remote effects, 99-100
spatial factors, 95 Hair cells, 61
temporal factors, 95-96 Hard of hearing, 216
location, electroencephalography and, 95 Harmonics, 51
Index 637
Head injury Ictal discharges, 91-94, 94/ 95/

in children, electroencephalography of, 121, 121/ ICU. See Intensive care unit
outcome prediction, somatosensory evoked lEDs (interictal epileptiform discharges), 580-581
potentials and, 198 Imbalance, 433-434
Head-shaking nystagmus, 419-420 Immittance unit, 216, 217/
Head surgery, intraoperative monitoring, 537 Impedance, 10-12, ll/
Hearing impairment, 216 Inborn errors of metabolism, EEG abnormalities,
Hearing loss 112-113
categories, 214, 215-216 Inching method, 255
conductive, 216 Indirect discharges (backfiring activation), 246, 248/
sensorineural, 216 Inductance, 5
Hearing tests, 214 Inductive plethysmography, 508
Heart period range, 468 Inductors (coils)
Heart rate reactance, 11
autonomic control, 441 rules for, 5-6, 6/
response Infantile spasms, ambulatory electroencephalography,
to deep breathing, 467-469, 469/ 126
to standing, 469 Infants
Heart rate range, 468, 469/ comatose, somatosensory evoked potentials,
Heart rhythm, during sleep, 509 198-199
Hemiconvulsions, electroencephalography, in electroencephalography, of developmental changes,
children, 120 113-114
Hemifacial spasm, blink reflex testing, 386 H reflexes in, 378
Hemiplegia, in children, electroencephalography of, neonates
120 electroencephalography, 108-109, 582-583
Hepatic coma, electroencephalography, 101, 103, evoked otoacoustic emissions, 220
103/ spine disease surgery, intraoperative monitoring of,
Hereditary hyperekplexia, 408 548
Hereditary motor and sensory neuropathy type I, Infectious diseases, EEG abnormalities, 112
laser evoked potentials, 486, 487/ Infinite impulse response filters, 49
Hering-Breuer respiratory reflex, 467 Inflammatory disorders, electroencephalography in
Herpes simplex encephalitis children, 119-120, 120/
electroencephalography, in children, 119 Inflammatory myopathy, single fiber
periodic lateralized epileptiform discharges, 101 electromyography, 355
Hertz (Hz), 63 Inflammatory polygangliopathies, masseter reflex, 390
Hiccups, 407 Inhibitory postsynaptic potentials (IPSPs), 29, 57
High-frequency repetitive potentials (complex Innervation ratio, 304, 325
repetitive discharges), 310, 310/ 31 If Insertional activity, in needle electromyography, 306,
Histogram, 64, 65/ 306/
Hoffman reflex. See H reflex Instrument ground, 20
Holmes tremor, 403 Insufficient sleep syndrome, 516
Holler monitoring, for sleep disorder assessment, Intensive care unit (ICU)
514 electroencephalography, 581
H reflex somatosensory evoked potentials in, 199-200
amplitude, 376, 376f Intention tremors, 402
clinical applications, 375, 379-380, 380/ Interaural latency differences, 223, 224
historical aspects, 375 Interference, 73-74, 74/
latency, 376, 379 Interference pattern
pediatric, 378 analysis, 333-334, 333/
physiologic basis, 375-377, 376£ 376<, 377f decomposition into constitutent MUPs, 337-339
testing technique, 377-378, 379/ 380f motor unit number estimate and, 359-360
vs. F wave, 376-377, 377< in needle electromyography, 307
Huntmgton's disease turns and amplitude analysis, 339-340, 340/
long latency reflexes, 395 Interictal activity
somatosensory evoked potentials, 196-197 definition of, 91
Hydrocephalus, electroencephalography in children, diffuse discharges, 92-93
121, 121/ focal discharges, 91-92, 91/
Hyperventilation Interictal epileptiform discharges (lEDs), 580-581
as EEC activation procedure, 81, 81/ Intermittent rhythmic delta activity, 99-100, 100/
electroencephalography, in children, 114, 114/ Interpolation techniques, 140-141
Hypnic jerks, 407-408 Interval analysis, 49/j 50, 5 If
Hypopnea, 504 Intraoperative monitoring
Hypothalamus tumor removal, intraoperative of brain stem. See Brain stem, intraoperative
monitoring, 535 monitoring
Hypoxia-ischemic insult, EEC abnormalities, 112 of cerebral cortex. See Cerebral cortex, function
Hypsarrhythmia, 93, 116, 116/ monitoring
638 Index
Intraoperative monitoring (Continued) Liberatory maneuver, 419

of cranial nerves. See Cranial nerves, intraoperative Linear display, of digital electroencephalography, 42
monitoring Linked potentials. See Satellite potentials
electroencephalographic, 530-531 Localization, of disease, 570-571
1
electroencephalography, 581-582 Long latency reflexes
of peripheral nervous system. See Peripheral to cutaneous nerve stimulation, 396, 396/
nervous system, intraoperative monitoring definition of, 394
of spinal cord. See Spinal cord, intraoperative flexor reflex and, 396
monitoring historical aspects, 394
in vascular disease surgery, 552-553 to mixed nerve stimulation, 395-396
Intraventricular hemorrhage, EEG abnormalities, 112 to stretch stimulus, 395
Inverse problem, 39 supporting evidence, 394-395
Inward current flow, 29 transcortical loop and, 394
IPSPs (inhibitory postsynaptic potentials), 29, 57 Lower motor neuron diseases, fibrillation potentials
Isoelectric electroencephalography, 111 and, 308<
Isolation transformers, 21-22, 22/ Lumbosacral radiculopathies, electromyography, 584,
Isometric tremors, 402 585
I waves, 283
Macroelectrodes, 296
Jaw jerk. See Masseter inhibitory reflex Macro-electromyography (macro-EMG), 332-333
Jerk-locked averaging, 163-164, 163/ Magnetic resonance imaging (MRI), for pre-epilepsy
Jerk nystagmus, 415 surgery evaluation, 147
Jitter Magnetic stimulation
measurement, by single fiber electromyography, for compound muscle action potentials, 241
344, 347-349 for motor evoked potentials, 284-285
single fiber electromyography, incorrect Magnetoencephalography (MEG), 39/ 40, 144
measurement position for, 350-351, 351/ Maintenance of wakefulness test, 496-497
Jugular foramen tumors, intraoperative monitoring, Masseter inhibitory reflex (jaw jerk)
537 clinical indications, 391-392, 391/
Junctional folds, 269, 269/ neuroanatomy, 389
testing
indications for, 390-391, 390/
methods for, 389-390, 390/
K complex, 86 testing method, 391, 39If
Kirchhoffs first law, 5-6, 6f, 7, 7/ MCD (mean consecutive difference), 347
Kirchhoff s second law, 6, 6/ Mean amplitude, 334
Mean consecutive difference (MCD), 347
Medial plantar nerve, nerve action potentials, I79f
Median nerve
La Crosse encephalitis, 101 nerve action potentials, 179<
Lambda waves, 84, 84/ somatosensory evoked potentials, 183, 184f,
Lambert-Eaton myasthenic syndrome 186-188, 186/
repetitive stimulation, 273/ 275/ 276, 278 Median neuropathies, compound muscle action
single fiber electromyography, 354—355 potentials, 260-262, 261/
Laser evoked potentials (LEPs), for pain evaluation, MEG (magnetoencephalography), 39/ 40, 144
484-486, 48S/487/ Membrane, electrical properties, 54
Lateral spread response, 264 Membrane potentials
blink reflex testing, 386, 386/ 387/ electrotonic, 54—55, 55/
Laterocollis, 410 resting, 53-54, 541
Leakage currents types of, 57-58, 58/
electric shock and, 20 Membrane time constant, 54, 55/
origin, 20-21 Meningitis, electroencephalography in children, 119
reaching patients Mental obtundation, electroencephalography of,
methods for, 21 578-579
reducing, 21-24, 21/-24/ MEPPs (miniature end plate potentials), 59, 269
tests for, 25, 26/ MEPs. See Motor evoked potentials
Learning set, 52 Mesencephalic lesions, masseter reflex, 390-391
Lennox-Gastaut syndrome Metabolic disorders, EEG abnormalities, 112
generalized slow spike-and-wave, in children, 117, Microneurography, 483-484
117/ Midbrain tremor, 403
prolonged video EEG, 133 Migraine headache, electroencephalography in
slow spike-and-wave pattern in, 92 children, 122, 122/
LEPs (laser evoked potentials), for pain evaluation, Miniature end plate potentials (MEPPs), 59, 269
484-486, 485>487/ Mitten patterns, 90
Index 639
Moments of the probability distribution, 52 changes, neuropathic vs. myopathic, 320-321, 3211
Morioneuropathies (focal neuropathies) characteristics, 325-326
compound muscle action potentials, 258-265, 259*, complexity, 330-332, 331/
260*, 261/ 263/ 264/ computer-assisted quantitative analysis, 335-337,
electromyography, 574, 5751 335/ 336/
nerve action potentials, 173-175, 174/ conduction block and, 573
Monopole, 30, 31/ distribution of electrical activity, 332
Montage reformatting, 43, 138-139, 139/ duration, 304/ 305, 328-S29/
Motor axons, 59 duration measurement, by single fiber
Motor conduction studies, compound muscle .action electromyography, 344, 349
potentials. See Compound muscle action interference pattern, 333-334
potentials long-duration, 315, 316/ 316*
Motor control test, 431 measurement, 305-306
Motor evoked potentials (MEPs) measurement characteristics, 303, 304/
amplitude, 285 mixed long-duration and short-duration, 317, 318/
central conduction time, 286 neurotonic discharges, 532-533, 533/
clinical applications, 282, 286-288, 287/ phases, 329/ 330-331
complications, 288 polyphasic, 315-316, 316/ 317/ 330-331
definition of, 282 power-spectrum analysis, 340-341
electrodes, 544-545 properties
intraoperative monitoring, 535 measurable only with special electrodes, 332-333
in primary neural disease, 551, 55 If with standard electrodes, 328-330, 329/ 330/
of vascular disease surgery, 552-553 rise time, 330
latency, 285-286 satellite potentials, 329/ 331
of motor symptoms, 568 short-duration, 314, 315*
recordings, 285 single, manual analysis of, 334-335
muscle, 545-546, 546/ size, 361-362
peripheral nerve, 545 spike duration, 330, 330/
spinal cord, 545 terminal component, 329/330
for spinal cord intraoperative monitoring, 540*, terminal negative afterpotential, 330
544-546, 546/ thickness, 330
stimulation, 282-283, 284/ turn, 329/ 331
electric, technical aspects, 283-284 variation, 317-318, 318/-320/ 319*, 331-332, 331/
magnetic, technical aspects, 284—285 574-575
Motor neuron disease Motor unit territory, 325
electromyography, 574, 5771, 591-592 Motor units
somatosensory evoked potentials, 197, 197/ definition of, 359
Motor system symptoms, assessment of, 568 number of, 325, 359
Motor unit Movement-associated potentials, 49
assessment, 291-292 Movement disorders
with needle electromyography. See Needle diagnosis, 399
electromyography surface electromyography, 399-401, 400/
with quantitative electromyography. See Movement-related cortical potentials
Quantitative electromyography contingent negative variation testing, 164
with single fiber electromyography. See Single definition of, 161
fiber electromyography jerk-locked averaging and, 163-164, 163/
number in muscle. See Motor unit number neural generators, 162
estimate technique, 161-162
Motor unit fractions, 332 waveforms
Motor unit number estimate (MUNE) abnormalities in disease, 163
clinical applications, 368, 369/ normal, 162-163, 162/
definition of, 359 MRI (magnetic resonance imaging), for pre-epilepsy
physiologic methods, 359 surgery evaluation, 147
quantitative, 301, 361-368 MSLT (multiple sleep latency test), 495-496, 516-517
all-or-none increments in CMAP, 362-365, 363/ Mu rhythm, 84, 84/
364/ Multifocal motor conduction block, 258
F-wave measurements, 365 Multi-MUAP analysis, 338-339
spike-triggered MUP averaging, 362, 363/ Multiple sclerosis
statistical measurements, 365-366, 366/-36S/ 368 blink reflex, 387-388, 389/
underlying assumptions, 361-362 brain stem auditory evoked potentials, 208/L210/
by standard electromyography, 359-360 209-210, 211
by standard nerve conduction studies, 360-361 motor evoked potentials, 286-287, 287/
Motor unit potentials (MUPs) visual evoked potentials, 231-232, 231/
amplitude, 304/ 305, 329-330 Multiple sleep latency test (MSLT), 495-496, 516-517
area, 329 Multiplets, 318-319, 320/ 320*
automated analysis, 337 Multispike-and-slow-wave pattern, 93
640 Index
Multivariate statistical methods, of complex repetitive discharges, 310, 310/ 31 It

electroencephalography, 142-143 cramp potentials, 312-313/ 315/ 315*
MUNE. See Motor unit number estimate doublets, 318-319, 320/ 320*
MUPs. See Motor unit potentials fasciculation potentials, 310-311, 31 If, 311*
Muscle fibrillation potentials, 308-309, 308/ 3081, 309/
contracting, needle electromyography, 297 insertional, 307
reinnervation, 573 long-duration motor unit potentials, 315, 316/j
resting, needle electromyography, 297 316*
stiffness, 593 mixed long-duration and short-duration MUPs,
as structural generators, 59 317, 318/
sympathetic activity, 440-441 motor unit potential variation, 317-318,
weakness 318/-320/ 319*
electromyography, 589-595 multiplets, 318-319, 320/ 320*
episodic, 593 myokymic discharges, 311-312, 312/ 312*
evaluation algorithm, 589 myotonic discharges, 309-310, 309/ 310*
Muscle artifact, SEP amplitude and, 182 neurotonic discharges, 312, 313/ 314/ 314*
Muscle disorders patterns of, 320-321, 321*
with fibrillation potentials, 3081 polyphasic motor unit potentials, 315-316, 316/
primary, single fiber electromyography, 355 317/
Muscle end plate potentials, 59 short-duration motor unit potentials, 314, 315*
Muscle fibers synkinesis, 320
density triplets, 318-319, 320*
definition of, 304, 325 cervical radiculopathies, 584, 585
single fiber electromyography, 344, 346-347 data collection, 297
innervation of, 268-269, 269/ for disease localization
myelinated, conduction in, 250 in peripheral nervous system, 571
synchrony of firing, motor unit potentials and, in posterior fossa, 570
325-326 electrodes, 295-296, 295/
Muscular dystrophy, single fiber electromyography, knowledge base, 294
355 limitations, 297-298
M wave, 242 lumbosacral radiculopathies, 584, 585
Myasthenia, congenital muscle selection for, 294-295
repetitive stimulation, 279 needle electrodes
single fiber electromyography, 354 characteristics, 296
Myasthenia gravis for deep muscles, 297
autoimmune, single fiber electromyography, insertion, 296-297
353-354 movement of, 297
electromyography, 590-591 for obese patients, 297
repetitive stimulation study, 272, 272/ 273/ 278 for small muscles, 297
Myelinated fibers, conduction in, 250 normal activity, 298-301, 300/
Myoclonic dystonia syndrome, 408 pattern of firing, 298
Myoclonus spontaneous, 298-299, 300/
etiologic classification, 405-406 voluntary, 300-301, 300/
muscle jerks, 405 pain control, 297
somatosensory evoked potentials, 197, 197/ patient preparation, 294
Myokymic discharges, in needle electromyography, pattern recognition, 298
311-312,312/ 312* postpolio syndrome, 593
Myopathies recording technique, 306-307, 306/
classification, 358 recruitment, 301, 303-305
electromyography, 589-590 abnormal, 319
Myorhythmia, 403 frequency, 301, 302/
Myositis, electromyography, 589-590 pattern, 319
Myotonia rapid, 319
electromyography, 593-595 reduced, 303, 319
repetitive stimulation, 279, 279/ semiquantitative, 298
Myotonic discharges, in needle electromyography, signal analysis, 305-307, 306/
309-310, 309/ 310* Negative afterpotential, 327-328, 327/
Negative myoclonus, 405
Neonates. See Infants, neonates
Nernst potential, 53, 54*
NAPs. See Nerve action potentials Nerve action potentials (NAPs)
Near-field potentials (NFPs), 184-185, 207 advantages of, 170
Neck surgery, intraoperative monitoring, 537 amplitude, 175, 176
Needle electromyography, 321-322 low or absent, 595
abnormal activity, 307-321 measurements, 177-178, 178/
in central control disorders, 319-320, 321/ antidromic technique, 170, 173
Index 641
clinical importance, 169-172, 170/ Neurologic symptoms, neural system involvement

conduction velocity, 178 and, 568-571
human factors, 176-177, 176/ Neuromuscular junction
measurements, 177-178, 178/ anatomy, 268-269, 269/
medial plantar nerve, 179* disorders, 268. See also Lambert-Eaton myasthenic
median nerve, 179* syndrome
methods botulism, 278-279, 354-355
for averaging, 177 congenital myasthenia, 279, 354
for nerve stimulation, 177 fibrillation potentials and, 308*
of recording, 177 physiology, 269-270, 270/
orthodromic technique, 170, 174 repetitive stimulation technique. See Repetitive
in pathologic conditions, 171-172 stimulation
mononeuropathies, 173-175, 174/ Neuromuscular transmission disorders, primary,
plexopathy, 173 single fiber electromyography of, 353-355
radiculopathy, 172-173 Neuromyotonia, 312, 313/ 314/ 314*
peak latency, 178, 178/ Neurons, preganglionic sympathetic, 439
phase cancellation and, 170-171 Neuropathic pain. See Pain
planning of study, 172 Neurotonic discharges
superficial radial nerve, 1791 electromyography of, 532-533, 533/
sural nerve, 179* needle electromyography of, 312, 313/ 314/ 314*
technical factors, 175-176 NFPs (near-field potentials), 184-185, 207
temperature effects, 175 Nocturnal myoclonus, 409
diphasic appearance, 170, 170/ Nocturnal oximetry, for sleep disorder assessment,
ulnar nerve, 179* 514
Nerve conduction studies Node, 6
anomalous innervation, 597 Noise, 74
artifacts, 596 Nonepileptic myoclonus, 405-406
assessment Nonrestprative sleep, 500, 501/
of motor symptoms, 568 Norepinephrine, skin vasomotor reflexes and,
of sensory symptoms, 569 455-456
averaging, 48-49, 48/ Nylen maneuver (Dix-Hallpike test), 417-418, 418/
baseline shift, 596 427*
for disease localization Nyquist theorem, 47
in peripheral nervous system, 571 Nystagmus
in spinal cord, 571 congenital, 415
distal latency, long, 596 horizontal and vertical, 415, 416/
focal slowing of conduction velocity, 598 2nd-degree, 416-417
F wave, absence, 596-597 positioning-induced, 417-419, 418/
for identifying disease types, 571-572 3rd-degree, 416
interpretation, 597-598
intraoperative monitoring
of cranial nerve, 533-534
of entrapment neuropathies, 558 Obstructive apnea, 504, 508-509, 509/
of peripheral nerve injury repair, 560, 561/ Obstructive hypopnea, 504
of peripheral nervous system, 555-556 Obstructive sleep apnea
of spinal cord, 540*, 546-547, 547/ causing excessive daytime somnolence, 516-517
for motor unit number estimate, 360-361 diagnosis, 516
myotonic syndromes, 593-595 Occipital cortex, 228
postpolio syndrome, 592-593 Occipital spikes, 92, 118, 118/
stimulation Occupational cramps, 410-411
for nerve action potentials, 177 Oculomotor testing, 421
supramaximal, 596 Oddball stimulus, 164
unexpected findings, 595-597 Oddball technique of auditory stimulation, 164
Net driving force, 54 Ohms, 4
Net ionic flux, 54 Ohm's law, 4
Neuralgic amyotrophy (Parsonage-Turner syndrome), Open-field potentials, 29, 29/
264 Optic nerve tumors, visual evoked potentials, 232
Neurocutaneous disorders, EEG abnormalities, Optic neuritis, visual evoked potentials, 232, 232/
113 Optic pathway, 61
Neurogenic blocking, in single fiber Optical coupling devices, 23-24, 24/
electromyography, 352, 352/ Optokinetic after nystagmus test, 421
Neurogenic disorders, single fiber electromyography, Optokinetic nystagmus, 421
355-356 Oromandibular dystonia, 410
Neurogenic potentials, 545 Orthostatic tremor, 404, 404/
Neurologic examination, during carotid Oscillopsia, 414
endarterectomy, 529 Ossicular discontinuity, diagnosis, 218*
642 Index
Otitis media, diagnosis, 218* Periodic short-interval diffuse discharges (PSIDDs),

Otoconia, 413 101, 102/
Otolith organ testing, 427 Peripheral lesions, causing vertigo, 414, 416
Otolith organs, 413 Peripheral nerves
Otosclerosis stapes fixation, diagnosis, 2182 evaluation, by nerve action potentials. See Nerve
Outward current flow, 29 action potentials
O waves, 115 injury
Oxygen saturation, during sleep, 509 compound muscle action potentials, 255, 256/
Oxyhemoglobin saturation, monitoring during sleep, electrophysiologic classification, 572-573, 573*
508-509, 509/ repair, intraoperative monitoring of, 558-560,
559/-561/
as structural generators, 59
Peripheral nervous system
P300, 164 disorders. See also specific peripheral nervous system
Pain disorders
autonomic tests, 481-483 localization of, 571
electrophysiology, 477-488 motor evoked potentials, 287
evaluation somatosensory evoked potentials, 191-194, 192/
by laser evoked potentials, 484-486, 485y^487/ 193/
by microneurography, 483-484 intraoperative monitoring, 561
hyperalgesia, quantitative sensory test, 480-481, electromyography for, 556
480/ of entrapment neuropathies, 557-558
magnitude estimation, 478 nerve conduction studies for, 555-556
physiologic mechanisms, microneurography of, to prevent injury, 560-561
483-484 somatosensory evoked potentials for, 556
thresholds of traumatic nerve injury repair, 558-560,
appearance, 478 559/-561/
determining, method of limits for, 478 preoperative evaluation, 556-557, 557/
disappearance, 478 Peripheral neuromuscular diseases
4,2,1 stepping algorithm, 478 myopathic, 358
Palatal tremor, 404 neurogenic, 358
Paradoxical localization, 189, 228 Peripheral neuropathy. See also specific peripheral
Parasites. See Satellite potentials neuropathies
Parasomnias, 513, 517-519, 518/ blink reflex, 387, 388/
Paraspinal fibrillation potentials, for disease compound muscle action potentials, 251-252, 252*,
localization, 571 253*, 255-258, 257?, 258*
Parasympathetic nervous system, 437 focal lesions, 252, 254, 254/ 254*
Parkinson's disease pressure palsies, 254-255
long latency reflexes, 395 diabetic, thermoregulatory sweat test, 458-^59, 459/
movement-related cortical potentials, 163 electromyography, 585-589
P300 abnormalities, 164 focal, compound muscle action potentials, 255,
surface electromyography, 411 258-265, 259*, 260*, 261/ 263/ 264/
tremor, 403 nerve action potentials, 171-172
Paroxysmal disorders, electroencephalography, 580-581 single fiber electromyography, 355-356
Paroxysmal rhythmic fast activity, 93 somatosensory evoked potentials, 191-192,192/ 193/
Paroxysmal tachyrhythmia, 93 Peroneal neuropathy
Parsonage-Turner syndrome (neuralgic amyotrophy), compound muscle action potentials, 262, 264/
264 electromyography, 588-589
Partial epilepsy, extracranial electroencephalography nerve action potential, 174-175
ictal, 149-151, 149/-151/ Persistent low voltage, in neonate, 111
interictal, 148-149, 148/ Phantom spike-and-wave, 89, 90/ 90*
Passive source or sink, 28, 29 Phase cancellation, 251-252
Past-pointing test, 419 nerve action potentials and, 170-171
Patient ground, 21 Phase spectrum, 140
Pattern recognition, 52 Photic stimulation
Peak latency, nerve action potential, 178, 178/ as EEG activation procedure, 81-82
Pendular nystagmus, 415 electroencephalography, children, 114
Periodic lateralized epileptiform discharges (PLEDs), for electroencephalography, benign variants, 87
99, 99/112 Photomyogenic or photomyologic response, as EEG
Periodic limb movements (PLMs) activation procedure, 81-82
polysomnography, 513-514 Photoparoxysmal response (photoconvulsive), as EEG
in sleep, 510-511, 510/ activation procedure, 81
surface EMG patterns, 409 Physiologic myoclonus, 405, 407-408
Periodic long-interval diffuse discharges (PLIDDs), Pituitary tumor removal, intraoperative monitoring,
101, 102/ 535
Periodic paralysis, repetitive stimulation, 279, 280/ Plastic imprints, for sweat measurement, 449
Index 643
PLEDs (periodic lateralized epileptiform discharges), in delirium, EEC patterns and, 578-579
99, 99/ 112 Projected rhythm, 99-100
Plexopathy, nerve action potentials, 173 Prolonged video electroencephalography (PVEEG)
PLIDDS (periodic long-interval diffuse discharges), advantages, 129
101, 102/ clinical applications, 131-134
PLMs. See Periodic limb movements epileptic vs. nonepileptic events, 131
Pneumotachometers, 507-508 nonepileptic behavioral events, 131-132
Polyminimyoclonus, 406 psychogenic seizure diagnosis, 131-132
Polyneuropathy, laser evoked potentials, 486, 487/ seizure classification, 132-133
Polyphasic motor unit potentials, 315-316, 316/ 317/ surgical evaluation, 133-134
Polyphasic potentials, 304 clinical utility, 129
Polyradiculopathy, electromyography, 591 development of, 129-130
Polysomnography equipment, 130-131
assessment, of consciousness impairment, 569 Proximal conduction, measurement, H reflex for,
definition of, 493 379, 380/
format, 493-494 Pseudohypacusis, 220
indications, 513-514 Pseudomyotonic discharges (complex repetitive
interpretation, 494-495, 494/ 495/ discharges), 310, 310/ 31 If
recording method, 494 PSIDDs (periodic short-interval diffuse discharges),
settings, typical, 495, 496< 101, 102/
for sleep assessment, 569 PSPs (postsynaptic potentials), 56-57, 61
unattended portable, for sleep disorder assessment, Psychogenic seizures, differential diagnosis,
514-515 131-132
Positional nystagmus, 422 Psychomotor-variant pattern, 88, 88/
Positional testing without fixation, 421-423, 422/ 427< Pupil cycle time, 474
Positioning-induced nystagmus, 417-419, 418/ Pure-tone and speech audiometric testing, 214,
Positioning nystagmus, 422 215-216, 215/
Positive bursts (spikes) PVEEG. See Prolonged video electroencephalography
14&4-Hz, 88-89, 88/ Pyramidal myoclonus, 405
14&6-Hz, 88-89, 88/ 116
Positive myoclonus, 405
Positive rolandic sharp waves, neonatal, 111
Positive sharp transients of sleep (POSTS), 86, 86/ QSART. See Quantitative sudomotor axon reflex test
Postactivation facilitation, 272-273, 272/ 273/ QST. See Quantitative sensory test (QST)
Posterior fossa, disease localization in, 570 Quadrupole, 31-32, 32/
Posterior fossa lesions, blink reflex, 387 Quantitative electromyography
Posterior slow waves advantages, 325
of eye blinks, 115 amplifier, 328, 328/
of youth, 115-116 computer-assisted, 335-337, 335/ 336/
Posterior tibial mixed somatosensory evoked disadvantages, 325
potentials, 188-189, 188/ interference pattern
Postictal slowing, in children, 122 decomposition into constitutent MUPs, 337-339
Postpolio syndrome, electromyography, 592-593 turns and amplitude analysis of, 339-340, 340/
POSTS (positive sharp transients of sleep), 86, 86/ interference pattern, analysis, 333-334, 333/
Postsynaptic potentials (PSPs), 56-57, 61 motor unit potential properties, 325-326, 328-330,
Postural normotension, maintenance of, 442-443 329/ 330/
Postural tremors, 402 power-spectrum analysis, 340-341
Posture, effect on Valsalva maneuver, 472 rationale for, 324
Posturography, for disease localization, 570 recording equipment, 326-328, S26/-328/
Potassium channels Quantitative sensory test (QST)
calcium-dependent, 58 clinical applications, 477-478
early, 58 interpretation, 479-481, 480/ 481/
Potentials, electric, 4. See also Bioelectric potentials pain thresholds, 478
Power, 4 procedure, 478-479, 479/
Power spectrum, 51 temperature stimulus, 478
Power spectrum analysis, 340-341, 470 Quantitative sudomotor axon reflex test (QSART)
Prechiasmatic lesions, visual evoked potentials, abnormal patterns, 448, 448/
231-233, 231/ clinical applications, 448-449
Precision decomposition, 338 equipment, 446-447, 447t
Preganglionic process, 171-172 indications, 445, 446tf
Preganglionic sympathetic fibers, 451 normal response, 446-448, 447/
Pressure palsies, compound muscle action potentials, patient preparation, 446, 446f
254-255 side-to-side differences, 482-483
Prognosis significance, 448
in consciousness impairment, EEC patterns and, Quantization, 45-46, 45/ 46/
578-579 Quantum, 269
644 Index
Radiculopathy Restless legs syndrome (RLS), 510-511

compound muscle action potentials, 265 Reticular myoclonus, 405
electromyography, 574, 575f, 576* Reticular reflex myoclonus, 407
nerve action potentials, 172-173 Retrochiasmatic lesions, visual evoked potentials,
somatosensory evoked potentials, 193-194 233-234, 233/
Rare stimulus, 164 Retrocollis, 410
Rasmussen encephalitis, electroencephalography in Reverse biasing, 14
children, 119-120 Reverse first-night effect, 502
Reactance, calculation of, 11 Rhythmic temporal theta bursts, of drowsiness, 81
Reafferant potential, 162 88/
Record storage/retrieval, digital Rise time (rate of rise), 67, 67/ 304/ 305
electroencephalography and, 42-43 Rolandic spikes, 91-92
Recording methods, for nerve action potentials, 177 Romberg test, 419
Recruitment Root-mean-square potentials of circuits, 10-11
analysis, for motor unit number estimate, 360 Rotational torticollis, 409
definition of, 360 Rotatory nystagmus, 415
Rectification, 14, 14/ RSD (reflex sympathetic dystrophy), 481-482
Reflex sympathetic dystrophy (RSD), 481-482 Rubral tremor, 403
Reflexes
arterial baroreflexes, 441-442
blink. See Blink reflex
cardiopulmonary, 442 Saccadic eye movement testing, 421, 4271
cardiovascular, 441-442 Safety factor, 270
diving, 469-470 Sampling
ergoreflexes, 442 digitization and, 46-47
skin vasomotor. See Skin vasomotor reflexes frequency, 46
for spinal cord monitoring, 5401 interval, 46
sympathetic, 439-440 aliasing and, 47, 47/
venoarteriolar, 442, 454-455, 455/ Sampling theorem, 47
Regional anhidrosis, 462 Satellite potentials, 304-305, 304£ 315-316, 316/
REM sleep behavior disorder, 518-519 317/ 329/ 331
Repetitive stimulation Scalp potentials, 187
abnormalities, 274-276 Sciatic neuropathy, nerve action potential, 174-175
criteria for, 276 Sclerosing myeloma neuropathy, somatosensory
technical factors, 275, 275/ evoked potentials, 191
acetylcholine release and, 269-270, 270/ Seats of electromotive force, 4-6, 6/ See also Battery;
amplitude, 274, 275f Generators
clinical correlations, 278-279, 279/ Segmental anhidrosis, 462
motor symptoms assessment, 568 Segmental demyelinating neuropathies, compound
nerve-muscle combinations, selection of, 276, 2771, muscle action potentials, 257-258, 258<
278 Segmental myoclonus, 407
rapid rates of stimulation, 276 Seizures
technical artifacts, 275, 275/ classification, prolonged video EEG for, 132-133
technique, 270-276 detection, quantitative, 138
electrode placement, 271 differential diagnosis, prolonged video EEG for,
immobilization, 271, 271/ 131
postactivation facilitation, 272-273, 272/ 273/ electroencephalography, 580-581
stimulus for, 271-272, 272/ 273/ occipital, in children, 118, 118/
train of four stimulus, 272, 272/ partial
x-shifted display, 273-274, 274/ impact on lifestyle, 147
Repetitive transient waveforms, 49 localization of, 146
RERAs (respiratory effort-related arousals), 504 surgery for. See Epilepsy surgery
Residual of the disease, 566-567 Selectivity, of single fiber electromyography, 343-344
Resistance, 4 Semicircular canals, 413
Resistors, 4-6, 6/ Semiconductors
Resonance, 11-12, ll/ doping and, 13-14
Respiration during sleep, assessment of, 504, n-type, 14
505/^507/ 507-510, 509/ p-type, 14
Respiration phase, effect on Valsalva maneuver, Semiquantitation, 307
472-473 Semiquanutative electromyography, 298
Respiratory disturbance index, 509 Sensorineural hearing loss, 216
Respiratory effort-related arousals (RERAs), 504 brain stem auditory evoked potentials, 222
Respiratory sinus arrhythmia, 441 Sensory analysis, for vertigo, 433<
Resting membrane potential, 53-54, 54< Sensory axons, 59
definition of, 53 Sensory organization test, 431-432, 432/
driving forces for, 53-54 Sensory receptors, as structural generator, 60-61
Nernst potential and, 53, 54t Sensory symptoms assessment, 569
Index 645
SEPs. See Somatosensory evoked potentials electroencephalogram, 82

SFEMG. See Single fiber electromyography benign variants, 88-90, 88f-90/ 89*
Shaky legs syndrome, 404, 404/ children, 114-115
Sharp-and-slow-wave pattern, 92 K complex, 86
Sharp-wave detection, quantitative, 138 neonatal, 109, HO/
Shut-eye waves, 115 normal adult, 85-86, 85/ 86/
Signal-to-noise ratio, 48-49, 74, 175 positive sharp transients, 86, 86/
Silent period, 396-397, 397/ vertex sharp transients (V waves), 85-86, 86/
Silicone imprints, for sweat measurement, 449 latency, initial, 502
Simple kinetic tremors, 402 neonatal, active, 109, HO/
Single fiber electromyography (SFEMG) non-rapid eye movement, 85, 497
amplifier, 345 periodic limb movements in, 510-511, 510/
amplitude trigger, 345 quiet, 109
clinical applications, 353-356. See also specific rapid-eye movement, 497
disorders electroencephalogram, 86
delay, 345 electroencephalography, 499-500, 500/
display, 345 latency, initial, 502, 504*
electrode, 344, 345 without atonia, 500-501, 501/
hardware, 345 respiration during, assessment of, 504, 505/-507/
interpretation, 343, 356-357 507-510, 509/
measurements respiratory assessment
blocking, 344-345, 349 of airflow, 507-508
duration, 345, 349 of blood gases, 508-509, 509/
fiber density, 344, 346-347 of cardiac rhythm, 509
jitter, 344, 347-349 of effort, 508
pitfalls, 349-352, 350/-352/ of snoring sounds, 508
damaged fiber, 351, 352/ statistical summary of variables, 509-510
ephaptic activation, 351-352, 352/ stages of, 497-500, 497/-500/
false trigger, 350, 351/ variables, summary statistics for, 502, 503*, 504*
incorrect measurement position, 350-351, Sleep deprivation, 516
351/ Sleep disorders
neurogenic blocking, 352, 352/ differential diagnosis, prolonged video EEG for,
selectivity of, 343-344 131
software, 345 excessive daytime somnolence, 516-517
split fiber, 351-352, 352/ parasomnias, 517-519, 518/
stimulated, 346, 352 partial evaluations, 514-515, 515/
technique, 344-349 polysomnography, 513-514
unstable trigger, 350, 350/ sleep study performance for, 515-516
voluntary, 345-346 Sleep histogram, 505j^506/
Single motor unit potentials (SMUPs), 361 Sleep onset REM (SOREM), 517
Single potential variables, measurable, 69-70, 70* Sleep spindles, 497-498, 498/
Skin Sleep-state misperception syndrome, 514
blood flow Slow alpha variant, 87
control of, 440 Slow calcium spikes, 57-58, 58/
measurement of, 451-452 Slow fused transients, 115
sympathetic innervation of, 440 Slow lambdas of childhood, 115
Skin vasomotor reflexes Slow spike-and-wave pattern, 92, 117, 117/
barorefiex indices, 456 Slow wave sleep, 498-499, 499/
blood pressure beat-to-beat response Small fiber neuropathy, thermoregulatory sweat test,
to tilt-up, 452-454 458
to Valsalva maneuver, 451-452, 453/ 453*, 454* Small sharp spikes, 89, 89/ 89*
denervation supersensitivity, 455 Smearing effect, 143
norepinephrine plasma levels and, 455-456 Smooth ocular pursuit test, 421, 427*
sustained handgrip, 456 SMUPs (single motor unit potentials), 361
venoarteriolar reflex, 454-455, 455/ Snoring sounds, assessment of, 508, 509
Sleep Soleus muscle, H reflex testing technique, 377-378,
active, 109 378/
ambulatory electroencephalography, 127 Somatosensory evoked potentials (SEPs)
assessment, 569 amplitude, 191
of core temperature, 511 influencing factors, 182-183
of esophageal pH, 511 variables in, 190
by maintenance of wakefulness test, 496-497 assessment
by multiple sleep latency test, 495-496 of consciousness impairment, 569
nontraditional measures, 500-502, 501/ of sensory symptoms, 569
by polysomnography, 485/J 493-495, 494/ 496* brain death criteria, 199
disordered breathing in, 504, 507, 507/ during carotid endarterectomy, 529
efficiency, 502 clinical applications
646 Index
Somatosensory evoked potentials (SEPs) (Continued) event recording, 66-67, 67/

central nervous system disorders, 194-200, frontal, 91
194fl97/ occipital, 92, 118, 118/
peripheral nervous system disorders, 191-194, small sharp, 89, 89/ 89*
192/ 193/ Spike-triggered averaging, for motor unit number
cutaneous nerve stimulation, 189 estimate, 362, 363/
dermatomal, 189 Spinal cord
for disease localization intraoperative monitoring
in posterior fossa, 570 advantages of, 539
in spinal cord, 570-571 areas of risk and, 539
electrodes, 541, 542/ clinical applications, 539-540
indications, 181-182 electromyography, 546-547, 547/
in intensive care unit, 199-200 goal of, 539-540
interpretation, 190-191 motor evoked potentials for, 540*, 544—546,
for intraoperative monitoring, 535 546/
of lumbosacral spine surgery, 551 nerve conduction, 540*, 546-547, 547/
of peripheral nerve injury repair, 559-560, 559/ in primary spine disease surgery, 548-549,
560/ 549/-551/ 551-552
of spinal surgery, 540-544, 5401, 542/J 548-549, reflex testing for, 540*
549/ 550/ somatosensory evoked potentials for, 540-544,
of thoracic spine surgery, 549, 551 540*, 542/
of vascular disease surgery, 552-553 surgical procedures and, 547-548
latencies, variables in, 190 lesions, somatosensory evoked potentials, 196
limitations, 182 localization of disease in, 570-571
localization, 190-191 as structural generator, 60
median nerve, 186-188, 186/ tethered, intraoperative monitoring of, 552
methods, 182-190, 183/ 184*, 186/ 188/ tumors, somatosensory evoked potentials, 196
for averaging, 182 Spindle coma pattern, 103
far-field potentials, 184-185 Spine disease, primary, intraoperative monitoring,
montage, 184-185 548-549, 549/-551/ 551-552
near-field potentials, 184-185 of infants/children, 548
peak nomenclature, 183-184, 184< surgical risk and, 548, 549/
neuroanatomy and, 185-186 Spontaneous activity, in needle electromyography,
N20/P25 complex, 187-188 307
for peripheral nervous system monitoring, 556 Spontaneous nystagmus, 415-416
posterior tibial mixed, 188-189, 188/ SQUID (superconducting quantum interference
recording montages, 189-190 device), 39/40
Fz, 187 SREDA (subclinical rhythmic electrographic
Fz-CSS, 189 discharge of adults), 87, 87/ 87*
scalp-to-noncephalic, 187 Standard stimulus, 164
standard methods, 189-190 Standing, heart rate response to, 469
technical factors, 541-544 Startle reflex, 408
trigeminal nerve, 189 Static posture testing, 419
ulnar nerve, 186-188, 186/ Stationary potentials, produced by propagating
waveforms, origin of, 185-186 generators, 36, 37/
SOREMs, 517 Statistical analysis, 52
Source dipole localization, 143r-144, 143/ Status epilepticus, 94, 94/ 95/
Spasmodic torticollis, 409 Step test, 428, 430
Spastic dysphonia, 410 Stimulation
Spatial filter, 143 for brain stem auditory evoked potentials, 204,
Spatial-temporal biodynamics, 95 205-206, 222-223
Spectral analysis (Fourier), 50-52, 137-138 unilateral vs. bilateral, SEP amplitude and, 183,
Spectral edge frequency, 137 183/
Speech recognition thresholds, 215/ 216 variables, SEP amplitude and, 183
Spike-and-slow-wave complex, 91 Stroboscopic intermittent photic stimulation, as EEG
Spike-and-wave pattern activation procedure, 81
atypical, 93 Stroke
definition of, 91 motor evoked potentials, 287
3-Hz, 92, 116f, 116£ 117 somatosensory evoked potentials, 199
6-Hz, 89, 90/ 90* Subacute sclerosing panencephalitis
slow, 92 electroencephalography in children, 120
Spikes periodic long-interval diffuse discharges, 101, 102/
anterior temporal, 91, 91/ Subclinical electrographic seizure discharge, 94, 94/
central-temporal, 91-92 95/
definition of, 91 Subclinical rhythmic electrographic discharge of
detection, quantitative, 138 adults (SREDA), 87, 87/ 87*
Index 647
Subdural effusion, electroencephalography in TEOAEs (transient evoked otoacoustic emissions),

children, 120 218-220, 219/
Sudomotor axon reflex test, 60, 60/ Terminal-two electrode, 239
Suggested immobilization test, 511 Test for adaptation, toes up/down, 431
Superconducting quantum interference device Tetanic stimulation, 270
(SQUID), 39/ 40 Tethered cord, intraoperative monitoring, 552
Superficial radial nerve, nerve action potentials, I79t Thalamic tremor, 403
Superior semicircular canal dehiscence, 427-428 Thermal hyperalgesia, quantitative sensory test,
Supratentorial lesions, somatosensory evoked 480-481, 480/
potentials, 196-197, 196/ Thermoregulatory sweat test
Supratentorial level, disease localization, 570 clinical applications, 458-459, 459/
Sural nerve, nerve action potentials, 179* description, 458
Surface electromyography interpretation pitfalls/difficulties, 464-465
abnormal patterns method, 459-462
in dystonia, 409-411, 410/ equipment for, 459-461, 460/
in myoclonus, 406-409 procedure for, 461-462, 4611
in tremor, 402-405 normal distribution patterns, 462, 462/
limitations, 400-401 reporting of results, 463, 463/ 464/
of movement disorders, 399-400, 400* Theta activity, 83, 112
dystonia, 409-411, 410/ Thoracic outlet syndrome, 264
myoclonus, 405-409 somatosensory evoked potentials, 193
techniques for, 400-401 Thoracic spine disease, intraoperative monitoring,
tremor, 402-405 549, 551
normal patterns, 401-402, 401/ Thoracoabdominal aneurysm surgery, intraoperative
ballistic, 401-402, 401/ monitoring, 551, 551/ 552-553
reflex, 401, 40 If Tibial nerve, somatosensory evoked potentials, 184<
tonic, 401, 401/ Tics, surface electromyography, 411
recording techniques Time and frequency domain analysis, 49/-51/J 50-52
for dystonia, 409 Time constant
for myoclonus, 406 of inductive-capacitive circuits, 9-10
for tremor, 402 resistive-capacitive circuits and, 7, 7/
signal quality, 401 resistive-inductive circuits and, 7, 8/
of voluntary movement disorders, 411 Time scale, for intraoperative electroencephalography
Sustained handgrip, skin vasomotor reflexes and, monitoring, 523-524, 524/-S26/
456 TIRDA (temporal intermittent rhythmic delta
Sway-referenced, 432 activity), 98-99, 99/
Sweat Topographic mapping, electroencephalographic,
measurement, imprint methods for, 449 141-142, 142/
normal distribution, 462, 462f Torticollis, 409-410, 41 Of
Sweat cell, multicompartment, 446, 447* Total cell capacitance, 54
Sweat glands, as structural generators, 59-60, 60/ Trace alternant pattern, 109
Sylvian seizures, 91-92 Transient evoked otoacoustic emissions (TEOAEs),
Sylvian spikes, 91-92 218-220, 219/
Sympathetic nervous system, 437 Transistors
assessment in humans, 440 amplification, 15
function, 439-440 npn, 14-15
innervation of skin, 440 pnp, 14-15
muscle activity, 440-441 semiconductors for, 13-14
reflexes, 439-440 Trauma, motor evoked potentials, 287
Sympathetic skin response (SSR), 483 Traumatic brachial plexopathy, somatosensory evoked
Symptomatic myoclonus, 405 potentials, 192-193, 193/
Synkinesis, 320 Traveling waves, 185
Syringomyelia, somatosensory evoked potentials, 195 Tremor
cerebellar or intention, 403
cortical, 406-407
definition of, 402
Tachyrhythmia, paroxysmal, 93 essential, 403
Tandem walking test, 419 exaggerated physiologic, 402-403
Tangential + somatosensory radial theory, 188 familial, 403
Tardive dyskinesia, movement-related cortical functional, 404-405
potentials, 163 Holmes, 403
Task- and position-specific tremor, 403-404 orthostatic, 404, 404/
Temperature, compound muscle action potentials palatal, 404
and, 249 in Parkinson's disease, 403
Temporal intermittent rhythmic delta activity surface electromyography of, 402
(TIRDA), 98-99, 99/ task- and position-specific, 403-404
Trigeminal nerve superior semicircular canal dehiscence, 427-428
lesions, blink reflex, 384-385, 385/ symptoms, 414-415
somatosensory evoked potentials, 189 work-up, 433-434
Triphasic wave pattern, in hepatic coma, 101, 103, electronystagmography test battery, 420-425,
103/ 422/-426/ 427*
Triplets, 318-319, 320* history taking, 414-415
Tumors. See also specific tumors vestibular system physical examination, 415-420,
electroencephalography in children, 120 416/418/
Turns/amplitude ratio, 334 vestibulospinal reflexes, 419-420
Turns and amplitude analysis, of interference pattern, Vestibular ocular reflex (VOR), 414
339-340, 340/ Vestibular rehabilitation, 433
Tympanic membrane Vestibular system
perforated, 218* physical examination, 415-420, 416/-418/
thickened, 218* sensory organs in, 413
Vestibular visual ocular reflex test (WOR), 428
Vestibulospinal reflex testing, for vertigo work-up,
419-420
Ulnar nerve Visceral function impairment, assessment of, 569
nerve action potentials, 179* Visual evoked potentials (VEPs)
repetitive stimulation, in myotonia, 279, 279/ clinical applications
somatosensory evoked potentials, 184t, 186-188, prechiasmatic and chiasmatic lesions, 231-233,
186/ 231/
stimulation, SEP amplitude and, 183 retrochiasmatic lesions, 233-234, 233/
Ulnar neuropathy interpretation, 230, 230*
compound muscle action potentials, 262, 263/ for intraoperative monitoring, 534, 534/ 535
electromyography, 587-588 method, 227
nerve action potential, 174 methods, 229-230
Upper airway resistance syndrome, 510 in normal subjects, 228-229, 228/ 229/
PI latency, normal, 228-229, 229/ 230, 230*
Pi-wave or P100, 227
retinal. See Electroretinogram
Valsalva maneuver sensory symptom assessment, 569
beat-to-beat blood pressure response to, 452, 453/ stimulation, 227, 229-230
453* Visual system, anatomy, 227-228
clinical usage in autonomic function testing Voice tremor, isolated, 403-404
technical variables, 472-473 Voltage clamp technique, for measuring action
technique, 472 potentials, 55-56
normal response, phases of, 470-471, 471/ Volume conductors
physiologic basis for, 470-471, 471/ applications, 36-40, 37/-S9/
pitfalls of, 473 definition of, 28
Valsalva ratio electric properties, 30
definition of, 473 VOR (vestibular ocular reflex), 414
pitfalls of, 473 WOR (vestibular visual ocular reflex test), 428
Vascular lesions, electroencephalography in children, V waves (vertex sharp transients), 85-86, 86/
120-121
Velocity recovery function, 348
Venoarteriolar reflex, 443, 454-455, 455/
Venoarteriolar reflexes, 442 Wake after sleep onset, 502
Ventral thalamotomy, postoperative movement-related Wakefulness, electromyography, 497, 497/
cortical potentials, 163 Wallenberg syndrome, somatosensory evoked
VEPs. See Visual evoked potentials potentials, 196
Vertex sharp transients (V waves), 85-86, 86/ Wallerian degeneration, 254
Vertigo Waveforms
causes, 413 artifactual, 72-74, 72*, 73/ 74/
central lesions, 414 nonphysiologic, 73-74, 74/
drug-induced, 415 physiologic, 72-73, 72*, 73/
peripheral lesions, 414, 416 continuous, 63-64, 64/ 65/ 66
computerized dynamic posturography, 430-431, frequency analysis of, 64, 65/ 66
431/ frequency of, 63-64, 64/ 65/
computerized rotatory chair test, 428, 429/ 430, signal displays, 71-72
430/ summation of, 63, 64/
motor control test, 431 variables, measurable, 70-71, 70*, 71*
otolith organ testing, 427 intermittent or event
sensory analysis, 433* recording, 66-68, 66/| 67/
sensory organization test, 431-432, 432/ signal displays, 71-72
variables, measurable, 69-70, 70t Wind-up, 486
movement-related cortical potentials, normal, Word recognition, audiogram and, 215/ 216
162-163, 162/ Wrist actigraphy, for sleep disorder assessment, 514,
physiologic alterations, 69-72, 70*, 711 515/
Weakness. See Muscle, weakness Writers' cramp, 410-411
Weber syndrome, somatosensory evoked potentials, Writing tremor, primary, 403
196
West's syndrome, 93
White matter disease, electroencephalography in
children, 119, 119/ Xenon injection measurement, of cerebral blood flow
Wicket spikes (waves), 85, 89-90 during carotid endarterectomy, 529-530

(Contemporary Neurology Series, 66) Jasper R. Daube - Clinical Neurophysiology (Contemporary Neurology Series, 66) - Oxford University Press, USA (2002)

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

(Contemporary Neurology Series, 66) Jasper R. Daube - Clinical Neurophysiology (Contemporary Neurology Series, 66) - Oxford University Press, USA (2002)

Uploaded by

Copyright:

Available Formats

CLINICAL NEUROPHYSIOLOGY

Contemporary Neurology Series

Oxford New York

Copyright ©2002 by Mayo Foundation for Medical Education and Research

Oxford is a registered trademark of Oxford University Press

All rights reserved. This book is protected by copyright.

Library of Congress Cataoging-in-Publication Data

Mark Hallett, M.D.

tom complexes with electroencephalography (EEG). The discussion of pediatric

Rochester, Minnesota J.R.D.

SECTION 1. ANALYSIS OF ELECTROPHYSIOLOGIC WAVEFORMS

1. ELECTRICITY AND ELECTRONICS FOR CLINICAL

2. ELECTRIC SAFETY IN THE LABORATORY AND HOSPITAL 17

4. DIGITAL SIGNAL PROCESSING 41

5. ELECTROPHYSIOLOGIC GENERATORS IN CLINICAL

6. CLASSIFICATION OF WAVEFORM CHARACTERISTICS 63

7. ALTERATION OF WAVEFORMS AND ARTIFACTS 69

SECTION 2. ELECTROPHYSIOLOGIC ASSESSMENT OF

Part A. Cortical Function

8. ELECTROENCEPHALOGRAPHY: GENERAL PRINCIPLES AND

ELECTROENCEPHALOGRAPHIC MANIFESTATIONS OF FOCAL

10. AMBULATORY ELECTROENCEPHALOGRAPHY 124

11. PROLONGED VIDEO ELECTROENCEPHALOGRAPHY 129

12. ELECTROENCEPHALOGRAPHIC SPECIAL STUDIES 137

13. ELECTROENCEPHALOGRAPHIC RECORDINGS FOR

14. MOVEMENT-RELATED POTENTIALS AND EVENT-RELATED

Part B. Sensory Pathways

15. NERVE ACTION POTENTIALS 169

19. BRAIN STEM AUDITORY EVOKED POTENTIALS IN

Part C. Motor Pathways

21. COMPOUND MUSCLE ACTION POTENTIALS 237

RAPID RATES OF STIMULATION 276

Part D. Assessing the Motor Unit

24. ASSESSING THE MOTOR UNIT WITH NEEDLE

COMPUTER=-ASSISTED QUANTITATIVE ANALYSIS OF MOTOR

Part £. Reflexes and Central Motor Control

28. H REFLEXES 375

30. LONG LATENCY REFLEXES AND THE SILENT PERIOD 394

Part F. Autonomic Function

33. CLINICAL PHYSIOLOGY OF THE AUTONOMIC NERVOUS

MUSCLE SYMPATHETIC ACTIVITY 440

34. QUANTITATIVE SUDOMOTOR AXON REFLEX TEST AND

35. ADRENERGIC FUNCTION 451

36. THERMOREGULATORY SWEAT TEST 458

37. CARDIOVAGAL AND OTHER REFLEXES 467

38. ELECTROPHYSIOLOGY OF PAIN 477

Part G. Sleep and Consciousness

39. PHYSIOLOGIC ASSESSMENT OF SLEEP 493

40. ASSESSING SLEEP DISORDERS 513

Part H. Intraoperative Monitoring

41. CEREBRAL FUNCTION MONITORING 523

42. BRAIN STEM AND CRANIAL NERVE MONITORING 532

43. SPINAL CORD MONITORING 539

SECTION 3. APPLICATIONS OF CLINICAL NEUROPHYSIOLOGY:

45. APPLICATION OF CLINICAL NEUROPHYSIOLOGY: ASSESSING

GLOSSARY OF ELECTROPHYSIOLOGIC TERMS 599

CHRISTOPHER D. BAUCH, PH.D. GREGORY D. CASCINO, M.D.

JOHN N. CAVINESS, M.D.

JAMES H. BOWER, M.D. JASPER R. DAUBE, M.D.

ROBERT H. BREY, PH.D. ROSE M. DOTSON, M.D.

ROBERT D. FEALEY, M.D. TERRENCE D. LAGERLUND, M.D., PH.D.

C. MICHEL HARPER, JR., M.D. PHILLIP A. Low, M.D.

CAMERON D. HARRIS, B.S., R.P.S.G.T. JOSEPH Y. MATSUMOTO, M.D.