Journal of Electromyography and Kinesiology 16 (2006) 586–602

www.elsevier.com/locate/jelekin

2006 ISEK Congress Keynote Lecture

Clinical applications of high-density surface EMG: A systematic review

a,b,* a,b
Gea Drost , Dick F. Stegeman , Baziel G.M. van Engelen b, Machiel J. Zwarts a,b

a
Department of Clinical Neurophysiology, Institute of Neurology, Radboud University Nijmegen Medical Centre,
P.O. Box 9101, 6500 HB, Nijmegen, The Netherlands
b
Neuromuscular Centre Nijmegen, Institute of Neurology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands

Abstract

High density-surface EMG (HD-sEMG) is a non-invasive technique to measure electrical muscle activity with multiple (more than
two) closely spaced electrodes overlying a restricted area of the skin. Besides temporal activity HD-sEMG also allows spatial EMG activ-
ity to be recorded, thus expanding the possibilities to detect new muscle characteristics. Especially muscle fiber conduction velocity
(MFCV) measurements and the evaluation of single motor unit (MU) characteristics come into view. This systematic review of the lit-
erature evaluates the clinical applications of HD-sEMG. Although beyond the scope of the present review, the search yielded a large
number of ‘‘non-clinical’’ papers demonstrating that a considarable amount of work has been done and that significant technical pro-
gress has been made concerning the feasibility and optimization of HD-sEMG techniques. Twenty-nine clinical studies and four reviews
of clinical applications of HD-sEMG were considered. The clinical studies concerned muscle fatigue, motor neuron diseases (MND),
neuropathies, myopathies (mainly in patients with channelopathies), spontaneous muscle activity and MU firing rates. In principle,
HD-sEMG allows pathological changes at the MU level to be detected, especially changes in neurogenic disorders and channelopathies.
We additionally discuss several bioengineering aspects and future clinical applications of the technique and provide recommendations for
further development and implementation of HD-sEMG as a clinical diagnostic tool.
 2006 Elsevier Ltd. All rights reserved.

Keywords: Review; High-density surface EMG; Multi-channel EMG; High-density surface EMG; High-spatial-resolution surface EMG

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
1.1. Clinical application: EMG as a diagnostic tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
1.2. Introduction of a new sEMG technique: high-density surface EMG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
2.1. Literature searches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
2.2. Classification of the relevant literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
3.1. Literature reviewed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
3.2. Clinical applications of HD-sEMG. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
3.2.1. Fatigue in neuromuscular disorders and chronic fatigue syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590
3.2.2. MND and neuropathies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590

Abbreviations: HD-sEMG, high-density surface EMG; MU, motorunit; MUAP, motor unit action potential; sEMG, surface EMG; MFCV, muscle
fiber conduction velocity.
*
Corresponding author. Address: Department of Clinical Neurophysiology, Institute of Neurology, Radboud University Nijmegen Medical Centre,
P.O. Box 9101, 6500 HB, Nijmegen, The Netherlands. Tel.: +31 24 3613491; fax: +31 24 3615097.
E-mail address: g.drost@neuro.umcn.nl (G. Drost).

1050-6411/$ - see front matter  2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jelekin.2006.09.005
 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602 587

3.2.3. Combination studies of MND, neuropathies and myopathies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592

3.2.4. Myopathies (including channelopathies) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
3.2.5. Positive muscle phenomena in patients and HD-sEMG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
3.2.6. Motor unit firing rate and HD-sEMG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
4.1. HD-sEMG and bioengineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
4.2. Clinical applications to date . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
4.2.1. HD-sEMG and fatigue studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
4.2.2. HD-sEMG and neurogenic changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
4.2.3. HD-sEMG and myopathies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
4.2.4. HD-sEMG and positive muscle phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
4.2.5. HD-sEMG and the central nervous system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
4.3. How to review a diagnostic tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
4.4. HD-sEMG and future perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597

1. Introduction for assessing neurophysiological characteristics of neuro-

Surface EMG (sEMG) is a non-invasive technique to
measure muscle activity where surface electrodes are placed
on the skin overlying a muscle or group of muscles (Her-
mens et al., 2000). In the context of routine motor–nerve
conduction studies in clinical neurophysiology, the use of
surface electrodes is widely used to record compound mus-
cle action potentials. Outside electrodiagnostic medicine, in
rehabilitation research, sport sciences, kinesiology and
ergonomics (Hogrel, 2005), sEMG is used to record on–
off switching of muscles and to estimate muscle force during
voluntary dynamic contractions. Clinical neurophysiology
uses EMG recordings during voluntary activity as a diag-
nostic tool especially in patients with neuromuscular disor-
ders. Intramuscular needle EMG electrodes are
predominantly used to evaluate motor unit (MU) function
but since it is a painful procedure, sEMG constitutes a good
alternative, particularly in the examination of children.
1.1. Clinical application: EMG as a diagnostic tool
In clinical practice, needle EMG evaluation, in combina-
tion with nerve conduction studies, is the standard method
muscular diseases. Although considerable effort has been
made to quantify needle EMG, most clinical neurophysiol-
ogists continue to base their diagnoses on subjective assess-
ments of motor unit action potentials (MUAPs) by visual
inspection of the on-screen signals and auditory evaluation
of the discharging units. In contrast, sEMG can hardly be
judged without proper data quantification, especially when
simultaneous recordings from multiple electrodes need to
be interpreted, which is the subject of this review. To facil-
itate the assessment of such large volumes of data, informa-
tion needs to be compressed.
Another difference between the two techniques is their
scope (see Fig. 1). Insertional activity, for instance, cannot
be evaluated with sEMG because of the absence of needle
insertions. Also spontaneous activity occurring at the sin-
gle fiber level, such as fibrillations, positive spikes and
myotonia, is too small to be recorded via the skin’s sur-
face. Where in principle the two techniques meet is in
the information at MU level. Single MU information
can be recorded rather easily with needle EMG. Because
the distance between the active muscle fibers and the nee-
dle electrode is small, and force levels are moderate,
MUAPs can be recorded fairly selectively with needle

movement muscle motor unit muscle fiber muscle

membrane

Conventional sEMG

HD-sEMG

Needle EMG

Fig. 1. The scope of the various EMG techniques. Conventional bipolar sEMG, with one bipolar electrode pair over each muscle, is mainly used in
movement studies. It yields concurrent information of muscle activity in different muscles. With the development of HD-sEMG, a technique that used
multiple electrodes on one muscle, it became possible to also subtract information at the single motor unit (MU) level. With HD-sEMG information on
muscle–fiber conduction velocity (MFCV) can be used to supplement the information at the muscle–fiber level obtained with needle EMG.
588 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602
electrodes. Conversely, conventional sEMG techniques
comprising a single bipolar signal from two electrodes
placed on one muscle do not allow a non-ambiguous
extraction of single MU information, which is why the
technique is of little or no use for clinical neurophysio-
logical purposes. Two reports that assessed the clinical
utility of conventional sEMG as a tool in the diagnosis
and treatment of neuromuscular diseases (Haig et al.,
1996; Pullman et al., 2000) both concluded that sEMG
was unacceptable as a tool in clinical diagnosis. However,
since their publication important technical advances have
been made in sEMG recording techniques.
1.2. Introduction of a new sEMG technique: high-density
surface EMG
The development of sEMG equipment that records the
input of multiple electrodes placed on one muscle has
increased the possibility of detecting single MU character-
istics (Stegeman et al., 2000b; Blok et al., 2002b; Merletti
et al., 2003).
The complex nature of this new multi-channel sEMG
signal, to be henceforth referred to as ‘high-density surface
EMG’ or HD-sEMG, places high demands on the instru-
mentation and signal analyses (Blok et al., 2002b, 2005;
Lapatki et al., 2006; Farina et al., 2002a). New spatiotem-
poral features at MU level have been introduced (Rau and
DisselhorstKlug, 1997; Stegeman et al., 2000b; Zwarts and
Stegeman, 2003; Drost et al., 2004a; Kleine et al., 2006).
Despite the technical possibility to obtain single MU infor-
mation from sEMG, to date HD-sEMG has not yet been
widely used as diagnostic tools in the clinical neurophysio-
logical practice.
We have conducted the current systematic review to
gain a more comprehensive insight into the current status
of the clinical applications of HD-sEMG. With a proper
inventory of the studies performed so far we aimed to
derive suggestions and recommendations for future stud-
ies fostering the development of HD-sEMG as a clinical
tool.
Studies using HD-sEMG with either electrode arrays or
matrix electrodes are discussed. Those that applied HD-
sEMG in patients were included as clinical applications.
We excluded HD-sEMG studies of the lumbar area.
Although we by no means consider such studies less impor-
tant and they frequently involve multiple electrodes at the
back muscles, we deemed them unsuitable for the present
review in that they were too far removed from routine clin-
ical diagnostics.
2. Methods
2.1. Literature searches
The literature search was performed on May 24, 2006 and
comprised two databases: PubMed and Web of Science. Table 1
lists the search terms used. Only full papers were included.
Abstracts published as conference or workshop proceedings were
not considered as the information provided was insufficient to
allow in-depth analyses of the presented work. Reviews were
analyzed separately. We also included two studies of clinical
applications our group has recently conducted, one of which is in
press, the other submitted.
2.2. Classification of the relevant literature
In the Pubmed search we confined ourselves to the system’s
human studies category. Reports from both sources that were
based on the data from one bipolar electrode pair over each
examined muscle were excluded as were animal studies, non-
English language studies, studies whose main topic was other than
sEMG, sEMG studies in which activity was not evoked volun-
tarily (i.e. electrically stimulation), and studies of non-skeletal
muscles. We classified the relevant literature as follows:
I. Bioengineering in HD-sEMG, subdivided into: (i) improve-
ments or developments at an instrumentation level, (ii)
physiological modeling of sEMG signals and (iii) signal
acquisition and analysis procedures.
II. Physiological applications of HD-sEMG.
III. Clinical applications of HD-sEMG, subdivided into: (i)
fatigue studies in neuromuscular disorders and chronic fati-
gue syndrome (CFS), (ii) motor neuron diseases (MND)
and neuropathies, (iii) combination studies of MND, neur-
opathies and myopathies (iv) myopathies (including chan-
nelopathies), (v) positive muscle phenomena and (vi) MU
firing rate.
As explained above, we focused on the clinical applications of
HD-sEMG. All reports on studies on patients with neurological
or neuromuscular diseases and the use of sEMG with more than
one bipolar electrode pair on a single muscle were considered as
such. Hence, all studies using a three-electrode linear electrode
array up to large two-dimensional matrixes of electrodes were
included. The term neuromuscular disorder refers to isolated or
associated primary or secondary disorders of the MU. We
adopted the subdivision of the clinical application studies on the
basis of the literature our searches yielded.
3. Results
3.1. Literature reviewed
The Pubmed search initially generated a total of 205
articles and the Web of Science search 160 references.
Eighty-three articles proved to have been mentioned twice.
We successively excluded meeting abstracts (5), articles that
were not in English (10), animal studies (6), articles not pri-
marily dealing with surface EMG (21), reports on bipolar
sEMG (67), studies on evoked MUAPs (4), paraspinal
muscles (7) and non-skeletal muscles (1). After assessing
the full text of 160 articles we classified the articles (Table
2a). We took the liberty to include two recent articles of
our own group one of which is in press and the other has
recently been submitted, concerning clinical applications
of HD-sEMG. Table 2b shows the references concerning
bioengineering and physiological applications of HD-
systematic review of the literature describing the applica-
HD-sEMG offers for clinical studies. Hogrel published a
Zwarts and Stegeman focused on the new possibilities
2000b; Zwarts et al., 2000; Zwarts and Stegeman, 2003).
cations of HD-sEMG (Hogrel, 2005; Stegeman et al.,

3.2. Clinical applications of HD-sEMG

clinical applications are discussed.

sEMG. Only the 29 articles regarding HD-sEMG and

III Clinical applications of HD-sEMG

II Physiological applications of HD-

Signal acquisition and analysis

Physiological modeling of sEMG

Instrumentation level
I Bioengineering in HD-sEMG

HD-sEMG
number of reports per category
Classification of the HD-sEMG studies included for review and the
Table 2a

multichannel surface-electromyography OR (multi-channel AND

Web of Science

(surface emg AND (mfcv OR muscle fiber conduction velocity)

PubMed
Overview of the search terms and strategy
Table 1
conduction velocity) AND patients)
clinical applications) OR (surface emg AND (mfcv OR muscle fiber
(surface emg AND neuromuscular disorders) OR (surface emg AND
multielectrode emg OR hsr-emg OR multi-electrode array emg OR
MCSEMG OR hsr-emg OR multi-electrode array emg OR
electrode-array EMG OR Surface EMG multi-electrode OR
arrays OR surface electrode-array electromyogram OR surface
density surface emg OR surface electrode array OR surface electrode
semg OR high-density surface emg OR high density semg OR high
OR (surface electromyography AND high-density) OR high-density
channel semg OR multichannel semg OR multichannel surface emg
surface electromyography) OR multi-channel surface emg OR multi-

arrays’’ OR (surface emg AND muscle cramp)

surface emg OR ’’surface electrode array’’ OR ‘‘surface electrode
OR high density semg OR high-density semg OR high density
high-density) OR high-density semg OR high-density surface emg
multichannel surface emg OR (’’surface electromyography’’ AND
surface emg OR multi-channel semg OR multichannel semg OR
channel AND ‘‘surface electromyography’’) OR multi-channel
electromyogram OR surface EMG multi-electrode) OR (multi-
OR surface electrode-array EMG OR surface electrode-array
(channel surface emg AND neuromuscular disorders) OR hsr-emg
AND patients) OR (surface emg AND clinical applications) OR
Anal sphincter muscle
Facial muscles

Anal sphincter muscle

Facial muscles
sEMG

procedures

signals
Our search yielded four review articles of clinical appli-

G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602

29

31

48

18

(N = 162)
No. of articles
1
8

2
1
8
4
2

Reviews
Table 2b
All references concerning bioengineering in HD-sEMG and physiological applications of HD-sEMG
I Bioengineering in HD-sEMG
 Instrumentation (Blok et al., 2002b; Farina and Cescon, 2001; Lapatki et al., 2004; Masuda et al., 1985; Merletti et al., 2004; Pozzo et al., 2004) (Prutchi, 1995; Sadoyama et al., 1985; Schneider et al., 1989; Thus-
neyapan and Zahalak, 1989; van Vugt and van Dijk, 2001) Review: (Merletti et al., 2003)
 Modeling:(Blok et al., 2002a; Dimitrov and Dimitrova, 2001; Dimitrova et al., 2001, 2003; DisselhorstKlug et al., 1998; Farina and Merletti, 2001, 2002a, 2004d,e,f, 2005; Fattorini et al., 2005; Merletti et al., 1999;
Miyano and Sadoyama, 1979, 2004, 2005, 2006; Schneider et al., 1991) Review: (McGill, 2004; Stegeman et al., 2000a)
 Signal acquisition and signal analysis:(Beck et al., 2005; Blok et al., 2005; Cescon et al., 2004; Chauvet et al., 2003; DisselhorstKlug et al., 1997, 1999; Falla et al., 2002; Farina et al., 2000, 2001, 2002c,d,b, 2003b,a,
2004a,g,h; Farina and Merletti, 2004; Garcia et al., 2005; Gazzoni et al., 2004; Grassme et al., 2003; Gronlund et al., 2005b,a; Hagg, 1993; Harba and Lynn, 1981; Harba and Teng, 1999; Holtermann et al., 2005a;
Kaneko et al., 1996; Kiryu et al., 1997; Kleine et al., 2000a; Knaflitz and Bonato, 1999; Koh and Grabiner, 1993; Masuda and Sadoyama, 1986, 1989; Merlo et al., 2003; Nakamura et al., 1997, 2004a,b; Okajima
et al., 1995; Ollivier et al., 2005; Ostlund et al., 2004; Reucher et al., 1987a,b; Roeleveld et al., 1997; Schulte et al., 2004b, 2005; Staudenmann et al., 2005, 2006)
Review (Arendt-Nielsen and Zwarts, 1989; DisselhorstKlug et al., 2000; Farina et al., 2004c) (Rau et al., 1997)
II Physiological applications of HD-sEMG
(Arendt-Nielsen et al., 1992; Cescon et al., 2006; Christova et al., 1999; Falla and Farina, 2005; Farina et al., 2004b, 2005a,b,c,d; Gerilovsky et al., 1991; Hanayama, 1994; Holtermann et al., 2005b; Houtman et al., 2003;
Kleine et al., 2000b; Maisetti et al., 2002; Masuda et al., 2001, 1983a,b; Masuda and Sadoyama, 1987; Masuda and DeLuca, 1991; Nishizono et al., 1989, 1990; Okajima et al., 1998; Sadoyama et al., 1988; Saitou et al.,
2000; Sbriccoli et al., 2003; Schulte et al., 2004a, 2006a,b; Roeleveld et al., 2000; Zwarts, 1989)
Review: (Roeleveld and Stegeman, 2002; Rau et al., 2004)
HD-sEMG applied to anal sphincter muscle: (Enck et al., 2005) Review: (Azpiroz et al., 2005; Enck et al., 2004)
HD-sEMG applied to facial muscles: (Castroflorio et al., 2005a; Castroflorio et al., 2005b; Geister et al., 1980; Iwasaki et al., 1990; Konno et al., 2005; Lapatki et al., 2006; Tokunaga et al., 1998; Tokunaga et al., 1991)

589
590 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602

tion of sEMG within the framework of neuromuscular dis-
orders. Next, we will discuss the various clinical HD-
sEMG studies categorized as:
1. Fatigue in neuromuscular disorders and CFS: five stud-
ies (Table 3).
2. MND and neuropathies: seven studies (Table 4).
3. Combination studies of MND, neuropathies and myo-
pathies: four studies (Table 5).
4. Myopathies (including channelopathies): eight studies of
which six concerned channelopathies (Tables 6a and 6b).
5. Positive muscle phenomena: two studies (Drost et al.,
2006, submitted for publication).
6. MU firing rate: three studies (Kleine et al., 2001; Sun
et al., 2000; Chen et al., 1997).
3.2.1. Fatigue in neuromuscular disorders and chronic fatigue
syndrome
Five studies examined fatigue (see Table 3 for an over-
view). The four publications by Linssen and his team
(Linssen et al., 1990, 1991b,a, 1996) all concern fatigue
studies into different myopathies. Two studies (Linssen
et al., 1991a,b) describe fatigue at high ischemic levels
in congenital myopathy characterized by type-I predomi-
nance. The patients showed an almost complete lack of
decline in muscle–fiber conduction velocity (MFCV) dur-
ing the fatiguing exercise. In their 1990 and 1996 studies,
Linssen et al. found an unimpaired muscle membrane
excitability during high-level fatiguing exercises, indicating
a dominant role for the intramuscular lactic acid forma-
tion in the normal decline of muscle fiber conduction in
healthy subjects. Schillings and colleagues (2004) com-
pared patients with CFS to healthy subjects during max-
imal voluntary contraction and reported that MFCV
declined less in the patients and that their peripheral-mus-
cle-fatigue at the end of the contraction was less pro-
nounced than it was in the healthy controls. A
diminished central activation of the muscle was thought
to be responsible for this phenomenon.
3.2.2. MND and neuropathies
Table 4 lists the six HD-sEMG studies of MND and
neuropathies (van der Hoeven et al., 1993; Roeleveld
et al., 1998; Wood et al., 2001; El Dassouki and Lefauc-
heur, 2002; Bahm et al., 2003; Drost et al., 2004a). The case
report of Lanzetta et al. (2005), who evaluated recovery of
a transplanted hand, is not listed in Table 4. They recorded
HD-sEMG signals with a linear electrode array of 16 elec-
trodes and an interelectrode distance of 2.5 mm, from
intrinsic muscles of the transplanted hand to assess their
degree of reinnervation. First signs of reinnervation of sin-
gle MUs were reported as reflected by the voluntary mod-
ulation of discharge rates of the MUs measured.
van der Hoeven et al. (1993) studied MFCV in patients
with amyotrophic lateral sclerosis (ALS) and in patients
with traumatic lesions of the plexus brachialis using both
needle and surface EMG. With the needle EMG potentials
were evoked with a stimulation needle electrode and mean
MFCV as well as fastest and slowest MFCV ratios were
calculated. For the sEMG measurements they used a rigid
array with three silver electrodes (diameter 2 mm) with a
common center electrode and interelectrode distances of
10 mm. The mean MFCV and the mean median frequency
(Fmed) were calculated from the bipolarly derived data at
different force levels. The sEMG signals in the ALS

Table 3
Overview of the HD-sEMG studies on muscle fatigue in patients with neuromuscular disorders and chronic fatigue syndrome (CFS)
Authors EMG variables (No. of electrodes) Controls Patients Muscles
Linssen et al. (1990) MFCV 26 5 McA BB
Frequency spectra
Amplitude
(5)
Linssen et al. (1991) Amplitude 26 3 CCD BB
MFCV 1 NRM
(5)
Linssen et al. (1991a) Amplitude 6 CCD VM
MFCV 1 NRM
Frequency spectra 1 EM
(5)
Linssen et al. (1996) Amplitude 7 3 McA BB
MFCV
Frequency spectra
(5)
Schillings et al. (2004) MFCV 14 14 CFS BB
CAF
Abbreviations: MFCV, muscle–fiber conduction velocity; McA, McArdle’s disease; BB, biceps brachii muscle; CCD, central core disease; NRM, nemaline
rod myopathy; EM, extertiaonal myalgia patient with 80% type-I dominance in muscle biopsy; VM, vastus medalis muscle; CAF, central activation
failure; CFS, chronic fatigue syndrome.
 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602 591

Table 4
Overview of HD-sEMG studies in patients with neurogenic disorders
Authors EMG variables Controls Patients Muscles
(No. of electrodes) (Age in years) (Age in years)
van der Hoeven et al. (1993) MFCV 28 16 PBL BB
Frequency spectra (34–74 y) (16–32 y)
(3) 22 ALS
(33-78 y)
Roeleveld et al. (1998) MU size 7 10 PPP BB
(58) (20–36 y) (51–72 y)
Wood et al. (2001) MU size 10 12 MND FDS
(16) (30–59 y) (28–75 y)
El Dassouki and Lefaucheur (2002) MFCV 20 20 CTS APB
(16) (24–78 y) (30–76 y) APB
Bahm et al. (2003) visual inspection 51 PBL BB
(1.5–16 y) ISM
Drost et al. (2004a,b,c) Amplitude 9 9 PPS VL
MFCV (38–68 y) (37–68 y)
Frequency spectra
LOI
MU size
(126)
Abbreviations: MFCV, muscle–fiber conduction velocity; BB, biceps brachii muscle; PBL, plexus brachialis lesion; ALS, amyotrophic lateral sclerosis; MU
size, motor unit size; PPP, prior polio patients; MND, motor neuron disease; FDS, flexor digitorum superficialis muscle; PPS, postpoliomyelitis syndrome;
VL, vastus lateralis muscle; LOI, level of interference (representing the density of the interference pattern); APB, abductor pollicis brevis muscle; ISM,
infraspinatus muscle.

patients revealed increased MFCV values in combination
with a decrease of the Fmed. The authors attributed the ele-
vated MFCV to muscle–fiber hypertrophy in the surviving,
voluntarily recruited MUs and the Fmed decrease to the
broadening of MUAP wave shapes.
Roeleveld et al. (1998) also compared the results of the
two EMG techniques. They obtained MUAPs during vol-
untary contractions of the biceps muscle of patients with
enlarged MUs resulting from prior poliomyelitis using
macroEMG, a needle technique in which a substantial part
of the electrode shaft is used as a recording electrode, and a
HD-sEMG technique with 58 electrodes with a diameter of
1.2 mm. Averaged surface MUAPs were obtained by nee-
dle EMG and sEMG triggered averaging. The MUAPs
area and peak amplitudes in the HD-sEMG recordings
showed similar correlations as found in the needle EMG
data. HD-sEMG MUAPs obtained both by needle trigger-
ing and by HD-sEMG triggering were equally sensitive to
pathology as the macro MUAPs. The most sensitive differ-
ence between the polio patients and the healthy controls
was the final positive peak amplitude. The authors hence
suggested to use the late wave as a measure of MU size
and concluded that sEMG is largely equivalent to macro
EMG in detecting enlarged MUs.
In their attempt to detect changes in MU architecture
through an increased understanding of the volume conduc-
tor effect Wood et al. (2001) utilized simple surface
waveform variables. A 16 · 2 electrode array (overall
dimension 5 · 45 mm) with an electrode diameter of
1 mm was placed transversely to the muscle fibers of the
flexor digitorum superficialis muscle. To identify repeated
firings of single, large amplitude MUAPs the authors
devised a pattern-recognition algorithm, the resultant spa-
tial information facilitating the decomposition process. To
estimate the size of a MU, they used the transversal width
of the potential distribution from that MU to correct for
MU depth. Using this model the differences in MU size
estimate between the patients with MND and the healthy
subjects became statistically significant.
El Dassouki and Lefaucheur (2002) studied the correla-
tion between the MFCV in the thenar muscle in patients
with moderate carpal tunnel syndrome (CTS). In healthy
subjects they established a correlation between MFCV
and distal motor latency or mean nerve-conduction veloc-
ity (MNCV) but found no such correlation in patients,
due to the nerve pathology and normal MFCV values.
Bahm et al. (2003) examined 51 children with obstetric
brachial plexus lesions with an electrode grid with an elec-
trode diameter of 0.5 mm and a small interelectrode dis-
tance of 2.5 mm. HD-sEMG was recorded from the
abductor pollicis brevis muscle, the biceps muscle and the
infraspinatus muscle of the affected side. Recordings for
the abductor pollicis brevis muscle were possible for all
children but the biceps muscle recordings demanded more
cooperation of the patients. Although follow-up studies
after reconstructive surgery were performed for a limited
number of patients the authors failed to present quantifica-
tion of variables or group results.
Drost et al. (2004a) studied MU characteristics in nine
healthy subjects and nine patients with postpoliomyelitis
syndrome to look for variables that would most clearly dis-
criminate between the two groups. A grid with 126 elec-
592 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602

Table 5
Overview of HD-sEMG studies combining patients with neurogenic and patients with myopathic disorders
Authors EMG variables Controls Patients Muscles
(No. of electrodes) (Age in years) (Age in years)
Yamada et al. (1991) MFCV 28 2 DMD BB
(6 · 10) (5–40 y) 1 MM TA
1 MD
1 SMA (KW)
2 GBS
(5–28 y)
Kumagai and Yamada (1991) MFCV 28 4 DMD BB
Amplitude (5–40 y) 1carrier DMD TA
(6 · 10) 1 MD
2 MM
1 SMA (K-W)
(5–28 y)
Ramaekers et al. (1993) Seven variables + 63 35 DMD APB
(17 electrodes) (0–25 y) 21 SMA
2 GBS
(0–24 y)
Huppertz et al. (1997) Nine variables++ 61 35 DMD APB
(32 electrodes) (0–25y) 6 BMD
21 SMAa
10 HMSN
Abbreviations: MFCV, muscle–fiber conduction velocity; DMD, Duchenne muscular dystrophy; MM, myotubular myopathy; MD, myotonic dystrophy;
MD, myotonic dystrophy; BMD, Becker muscular dystrophy; SMA, spinal muscular atrophy.
(KW, Kugelberg–Welander type); GBS, Guillain-Barré syndrome.
+ Ramaekers parameters (1–7): 1. MFCV in single MUAPs, 2. Signal entropy, 3. First zero crossing of ACF, 4. Dwell-time over RMS, 5. Chi-value of
amplitude distribution, 6. Gradient of peak edge, 7. Peak frequency distribution.
++ Huppertz: 1. MFCV of single MUAPs, 2. Signal entropy, 3. Gradient of peak edge, 4. Dwell time over RMS, 5. Chi-value of amplitude distribution,
6. Peak/gap energy, 7. Maximum peak amplitude, 8. RMS, 9. Autocorrelation function.
a
1 Werdnig-Hofmann disease; 6 intermediate; 3 SMA type-III; 11 not classified.

trodes with a diameter of 1.5 mm and an interelectrode dis-
tance of 5 mm was used and both the single MUAPs
extracted from the HD-sEMG signals and the raw signal
itself were analyzed. Force and HD-sEMG were recorded
simultaneously at different force levels. Extraction of single
averaged MUAPs was performed in accordance with the
method described by Kleine and coworkers (2000a) and
proved possible at force levels of 5% and 20% of the
MVC only. The estimated MUAP size, determined as the
highest area under the curve of these extracted MUAPs,
perfectly separated the patients from the controls. Raw sig-
nal analysis showed significant between-group differences
for the monopolarly recorded amplitude at force levels
up to 60% MVC. At high force levels the level-of-interfer-
ence (LOI), a variable closely related to the so-called turns-
and-amplitude analysis in needle EMG, differed signifi-
cantly between patients and healthy controls.
3.2.3. Combination studies of MND, neuropathies and
myopathies
Four studies examined both patients with myopathic
and patients with neurogenic disorders (Yamada et al.,
1991; Kumagai and Yamada, 1991; Ramaekers et al.,
1993; Huppertz et al., 1997; for an overview, see Table
5). Kumagai and Yamada studied various patient samples
with a ‘‘matrix’’ type of surface electrode with a diameter
of 5 mm and an interelectrode distance of 7 mm. In the
normal subjects they found MFCV values in both the tibial
anterior and the biceps brachii muscles to be higher at
moderate contractions than those during weak contrac-

Table 6a
Overview of HD-sEMG studies in myopathies (excluding channelopathies)
Authors EMG variables Controls Patients Muscles
(No. of electrodes) (Age in years) (Age in years)
(Hilfiker and Meyer, 1984) Mean MFCV 17 5 PMD (15–46 y) BB
(30) (6–68 y) 5 DMD (6–8 y)
Yamada et al. (1987) Mean MFCV 14 2 DMD (7–8 y) BB
Amplitude distribution (4–50 y) TA
(60)
Abbreviations: MFCV, muscle–fiber conduction velocity; BB, biceps brachii muscle; TA, tibialis anterior muscle; PMD, progressive muscular dystrophy.
 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602 593

Table 6b
Overview of HD-sEMG studies in channelopathies
Authors EMG variables Controls Patients Muscles
(No. of electrodes) (Age in years) (Age in years)
Zwarts et al. (1988) MFCV 15 15 asymptomatic relatives BB
Frequency spectra (32–52 y) (15–47 y)
(3) 10 HOPP
(19–71 y)
van der Hoeven et al. (1994) MFCV 46 33 HOPP carriers BB
Frequency spectra (19–74 y) (16–57 y)
EMG amplitude 56++
(3) (16–76 y)
Zwarts and van Weerden (1989) EMG amplitude 3 3 RMC BB
MFCV (29–35 y) (23–40 y) BR
Frequency spectra 8 MD ADM
(3) (23–64 y)
Drost et al. (2001) MFCV 5 7 RMC BB
PPA (22–64 y) (22–57 y)
(126)
Drost et al. (2004) Amplitude 7 RMC BB
Visual analysis (22–57 y)
(126)
Van Beekvelt et al. (2006) Amplitude 9 3 RMC BR
MFCV (32–48 y)
(32)
Abbreviations: HOPP, hypokalemic periodic paralysis; RMC, recessive myotonia congenita; BR, brachioradialis muscle; ADM, abductor digiti minimi
muscle; ++ asymptomatic first degree carriers; PPA, peak-peak amplitude.

tions. No significant changes were found for age or sex.
Furthermore, MFCV was significantly related to the ampli-
tude of the MUAPs. Both groups of researchers concluded
that the MFCV cutoff point between healthy subjects and
patients with myopathy was about 3.0 m/s. The mean
MFCV was significantly reduced in patients with Duch-
enne muscular dystrophy (DMD) and patients with myotu-
bular myopathy. In the myopathic patients the MFCV in
both muscles did not increase with muscle force as it did
in the healthy subjects. In an asymptomatic DMD carrier,
the MFCV proved reduced during weak and moderate iso-
metric contractions.
In neurogenic patients the mean MFCV values were
within the normal range. In two patients with Guillain-
Barré syndrome (GBS) high-amplitude potentials are
described that were similar to those in a patient with spinal
muscular atrophy (SMA; Kumagai and Yamada, 1991).
Ramaekers et al. (1993) studied the clinical application
of HD-sEMG for the recording of single MU activity in
children with DMD and SMA whose data were compared
with those from 63 healthy children ranging from the neo-
natal period to the age of 24 years. They concluded that in
healthy neonates the MFCV was lowest at 1–2 m/s and
gradually reached a plateau of 2.9–4 m/s from the age of
4 years onwards. MUAPs in DMD patients had a more
blunted waveform and an abnormal pattern with signifi-
cantly lower amplitude MUAPs in 30 of the 31 patients.
In 10 SMA patients the sEMG showed abnormally high
MUAP amplitudes and a reduced recruitment of firing
MUs. The MFCV remained within the normal range.
Huppertz et al. (1997) retrospectively determined the
diagnostic accuracy of the HD-sEMG method. The best
result was achieved by calculating a set of three more
advanced signal characteristics and MFCV. Using a special
evaluation procedure they were able to assign the subject to
the correct group in about 81% of the cases. The sensitivity
for the detection of abnormal vs. normal was reported to
be 82%, specificity 97%, and PPV 97%. Myopathic and
neuropathic disorders were recognized with a sensitivity
of 85% and 68%, specificity 97% and 98%, and PPV 92%
and 91%, respectively. The authors accordingly claimed
to have obtained a specificity and sensitivity to distinguish
patients from controls that was within the same range as
the results obtained with needle EMG.
3.2.4. Myopathies (including channelopathies)
Eight studies (Zwarts et al., 1988; van der Hoeven et al.,
1994; Zwarts, 1989; Drost et al., 2001, 2004b; Van Beekvelt
et al., 2006; Hilfiker and Meyer, 1984; Yamada et al., 1987)
were found using HD-sEMG in patients with myopathies,
from which six concerned channelopathies (see Tables 6a
and 6b, respectively). Hilfiker and Meyer (1984) reported
that MFCV could not be determined in some patients in
a late stage of progressive muscular dystrophy, as there
was no discernible time shift of the bipolar potentials. To
explain the lack of propagation, they postulated the
increased longitudinal spread of motor endplates, as can
be observed in myopathic muscles. In DMD patients, the
MFCV could be measured and was significantly reduced,
with a mean value of 2.81 m/s. Based on the MFCV value
594 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602
each DMD patient could be clearly separated from the
controls. Yamada et al. (1987) also found abnormal prop-
agation patterns of MUAPs in de TA muscles of 2 DMD
patients. In their biceps muscles a decreased mean MFCV
to 2.8 and 3.2 m/s was observed. Most of the HD-sEMG
studies were performed in patients with channelopathies
(as summarized in Table 6b), which are myopathies due
to a genetic defect in the gene encoding for a voltage-gated
channel in the muscle membrane.
Hypokaliemic periodic paralysis (HOPP): The two stud-
ies performed on HOPP patients were based on the same
extended family (van der Hoeven et al., 1994; Zwarts
et al., 1988). Eight of the nine HOPP patients had a
reduced MFCV (2 SDs lower than controls) compared to
their controls (Zwarts et al., 1988). The mean MFCV of
the controls was 4.55 ± 0.33 m/s. In six of the asymptom-
atic relatives MFCV was also reduced. The Fmed of the
patients and the relatives with a low MFCV proved to be
significantly lower than the values of the healthy controls.
The MFCV was more often abnormal than the Fmed in
HOPP patients. van der Hoeven et al. (1994) observed a
significantly lower mean MFCV in HOPP patients and in
asymptomatic carriers, both with invasive MFCV measure-
ments and with three surface electrodes. However, only the
invasive method showed a lower MFCV in all proven car-
riers. Detecting HOPP carriers with the surface MFCV
method had a sensitivity of 77%.
The amplitude of the sEMG signal at maximum volun-
tary contraction was lower in the carrier group. There was
a positive correlation between the MFCV and the neuro-
muscular efficiency (the quotient of force and EMG ampli-
tude) in the controls but not in the HOPP carriers. Since
type-II fibers have a higher neuromuscular efficiency, this
suggested a preferential involvement of type-II fibers in
HOPP.
Recessive myotonia congenita (RMC) – Zwarts and van
Weerden (1989) studied transient paresis in myotonic syn-
dromes. EMG was recorded during five maximal 10-s vol-
untary contractions followed by 13-s recovery. In the
recessive myotonia congenita patients a decline of EMG
amplitude was found during an initial loss of force. This
transient paresis was accompanied by a dramatic fall in
the MFCV concomitant with a shift of the power spectrum
to lower frequencies. During the decline in force, amplitude
and MFCV lessened with each successive contraction
(warming-up phenomenon).
Using a 126 HD-sEMG grid Drost et al. (2001) showed
that transient paresis was explained by a deteriorating mem-
brane function, ending in conduction block and paresis.
During the period of force decline in transient paresis, a
decrease in amplitude of the MUAPs from endplate towards
the tendon was observed. The disturbance increased with
time and place, indicating a deteriorating membrane func-
tion, and ended in a complete blocking of propagation within
seconds. Furthermore, it was shown that transient paresis
was a force-dependent phenomenon in that at low force lev-
els it remained absent in recessive myotonia congenita
patients. In a later study Drost et al. (2004b) tested the
hypothesis that although muscle–membrane function is
already disturbed at low force levels, abnormal MU recruit-
ment acts as a compensatory mechanism to obtain normal
force stability. The sEMG abnormalities included disturbed
propagation of MUAPs over the muscle–fiber membrane
and abnormal MU recruitment pointing to central adapta-
tion mechanisms. Prolonged periods of low sEMG ampli-
tudes were sometimes present despite normal force,
indicating that mechanisms other than MU recruitment also
contribute to force preservation. During the periods of low
sEMG, the density of myotonic discharges, recorded simul-
taneously with needle EMG, was not increased. It was con-
cluded that abnormal MU recruitment as a compensatory
mechanism occurs in recessive myotonia congenita, but can-
not fully explain constant force.
Besides transient paresis and constant force, Van Beek-
velt et al. (2006) also examined the warming-up phenome-
non (the diminishing of the initial temporary weakness that
occurs in RMC patients with repetitive contractions) with
HD-sEMG. The underlying hypothesis of their study was
that the warming-up phenomenon is associated with the
exercise-related activation of Na+–K+-ATPase. The con-
traction was preceded by an infusion of the Na+–K+-ATP-
ase inhibitor ouabain into the brachial artery of the
exercising arm. Transient paresis occurred after initiation
of exercise, accompanied by electrophysiological changes
indicating sarcolemmal conduction block. Ouabain infu-
sion did not affect the recovery from transient paresis or
the electromyographic changes, indicating that the warm-
ing-up phenomenon in RMC is not mediated by Na+–
K+-ATPase.
3.2.5. Positive muscle phenomena in patients and HD-sEMG
In two studies of our group we examined positive invol-
untary muscle phenomena, fasciculation potentials and
involuntary contractions in Satoyoshi syndrome (Drost
et al., submitted for publication, 2006). Fasciculations can
easily be observed with sEMG as Kumagai and Yamada
(1991) demonstrated with their measurements of the fore-
arm flexor muscles of a patient with SMA. We showed that
HD-sEMG is suitable to examine fasciculations because of
its non-invasive character allowing long, stable recording
times and because of the spatiotemporal information (Drost
et al., submitted for publication). The unique characteristics
of fasciculation potentials across all channels facilitate
extensive fasciculation potential characterization.
In the second study, we examined involuntary painful
muscle contractions in the legs of a patient with Satoyoshi
syndrome using two non-invasive EMG techniques: HD-
sEMG (with 126-channels) of the vastus lateralis muscle
and polymyographic sEMG recordings on various muscles
(Drost et al., submitted for publication). During the invol-
untary contractions, HD-sEMG showed a fourfold
increase in amplitude relative to the MVCs. These high
potentials were distributed across the entire muscle and
showed a pronounced oscillatory behavior with a fre-
 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602
quency around 45 Hz. Together with polymyographic
sEMG that had revealed that the involuntary contractions
often occur simultaneously in various muscles, these find-
ings point to hyperactivity or a disinhibition at the alpha
motor neuron level, probably originating at that level,
although a central origin cannot be excluded.
3.2.6. Motor unit firing rate and HD-sEMG
Three articles primarily used HD-sEMG to measure fir-
ing patterns of single MUs. In fact, our earlier study (Drost
et al., 2004b) also focused on central mechanisms besides
membrane disturbances to explain constant force.
Kleine et al. (2001) compared 126-channel HD-sEMG
recordings from the biceps muscles of seven patients with
Parkinson’s disease with those of five healthy controls to
explore whether reduced inhibition in the motor cortex of
the patients was accompanied by similar changes in their
MU firing modulation by transcranial magnetic stimula-
tion as was found in ALS. They applied a HD-sEMG
decomposition technique to extract the firing pattern of
up to five simultaneously active MUs. The peristimulus
time histograms indicated a longer duration of the evoked
excitatory postsynaptic potential. The authors attributed
the results to prolonged corticospinal volleys resulting from
impaired intracortical inhibition.
Chen et al. (1997) also exploited HD-sEMG signals to
analyze the temporal changes and spatial distribution of
the dominant MU firing rate. They applied seven elec-
trodes on the biceps muscle and seven electrodes on the tri-
ceps muscle from lateral to medial, with an interelectrode
distance about 20 mm, to compare five healthy subjects,
five Parkinson patients and five patients with cerebrovascu-
lar accidents (CVA). To estimate the dominant firing rate
they used autoregressive spectrum analysis of the sEMG
because the EMG spectrogram was expected to allow
examination of the time-varying characteristics of the firing
rates and the recruitment of MUs from the sEMG signals.
The authors found abnormalities of the mean MU firing
rates and spatial patterns in both patient groups and sug-
gest that via this new representation HD-sEMG spatiotem-
poral firing patterns were of clinical use in the
electrophysiological detection and characterization of
motor control disorders. Sun et al. (2000) also conducted
a study of single MU firings involving a computer simula-
tion and a clinical study with HD-sEMG recordings of the
interosseous muscle in patients with stroke, myopathies
and neuropathies and in healthy controls (n = 16 for each
group). The HD-sEMG consisted of a 2 · 2 matrix with
a 0.2 mm diameter for each electrode and an interelectrode
distance of 2.5 mm. The authors proposed a singular value
decomposition method to study the MU firing patterns.
The firing variability was found to be increased in the
stroke patients and in the myopathy and neuropathy
groups the mean firing rates were elevated. Relative to
the healthy controls, the firing variation index, i.e. the fir-
ing standard deviation (SD) normalized by the firing mean,
was elevated in every patient group.
595
4. Discussion
The current overview of HD-sEMG application
research clearly demonstrates that as a tool in clinical prac-
tice the technique is still in its infancy although the large
number of ‘‘non-clinical’’ papers our search generated
shows that the technique is being probed extensively and
progress has been made in its technical and practicable
aspects. It proved difficult to find studies that used HD-
sEMG in clinical applications, which was largely due to
the lack of uniformity in the designations for this surface
EMG technique involving multiple electrodes on one mus-
cle. As the overview of our search strategy in Table 1 illus-
trated, a wide range of names was used, among which
multi-channel sEMG, high-resolution sEMG and multi-
array sEMG. Because the term multi-channel sEMG is also
used in kinesiology studies to denote an EMG comprising
bipolar electrodes on different muscles, we advise against
the use of this term. There is a clear need for consensus
regarding the technique’s name.
Another salient detail was the great variability in elec-
trodes, electrode-positions, interelectrode distances and
configuration of electrodes. This caused a similar variabil-
ity in the results in the various pathologies, impeding their
reproducibility and a sound interpretation. Especially for
clinical applications, uniformity in electrode configuration
and electrode grids will become increasingly essential. This
is also the case for the recording systems and possibly also
for the key features of the analysis software. Diversity in
equipment complicates the clinical application of the tech-
nique unnecessarily.
In sum, further improvements of HD-sEMG as well as
future applications of the technique in the clinical practice
warrant uniformity in various aspects. As Hogrel (2005)
already stated, a lack of standards for electrodes, configura-
tions, electrode placement and recording protocols have
adversely affected the possibility of the technique’s integra-
tion into routine clinical use. Accordingly, it is recommended
to establish a consortium of leading investigators in this field
to devise technical guidelines for clinical applications as was
done by Hermens et al. (2000) for kinesiology studies.
4.1. HD-sEMG and bioengineering
As mentioned above, needle EMG is the standard to
compare HD-sEMG signals with as it is virtually the only
EMG technique that is used as diagnostic tool in clinical
neurophysiology. However, diagnoses are mostly based
on a subjective assessment of the signals. In contrast,
for HD-sEMG interpretations quantitative variables are
essential. Because of the magnitude of the data, automatic
quantification and analysis of the signals is indispensable,
which poses a challenge for the introduction of HD-
sEMG in the clinical practice. Since we chose to restrict
our review to the clinical applications of HD-sEMG,
the discussion of studies concerning the biophysics and
bioengineering of HD-sEMG was not exhaustive. For a
596 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602
more comprehensive overview of the state of the art on
these topics in sEMG applications we refer to the review
by Merletti and Parker (2004).
4.2. Clinical applications to date
4.2.1. HD-sEMG and fatigue studies
The scope of HD-sEMG applications is different from
that of needle EMG. Because of spatial information allow-
ing MFCV measurements and because of the longer
recording times the technique affords, the greater majority
of fatigue studies exploit the domain of sEMG. In most
clinical studies using HD-sEMG MFCV, which is inacces-
sible to conventional needle EMG, proved an important
variable providing useful information about the symptom
in neuromuscular disorders. Because fatigue is such a pro-
nounced symptom in nearly all neuromuscular disorders,
this new EMG variable may prove a valuable addition
for use in the clinical practice.
4.2.2. HD-sEMG and neurogenic changes
The studies investigating neurogenic changes (see Table
3) showed that HD-sEMG can identify these changes rela-
tively easily. Variables from ‘‘raw’’ HD-sEMG data such
as MFCV and Fmed were studied (van der Hoeven et al.,
1993), and also amplitude measurements and variables that
provide information about the MUAPs’ interference pat-
terns were used in an attempt to differentiate between sub-
jects with neurogenic diseases and healthy controls (Drost
et al., 2004a). No uniform pattern was found: in ALS
patients MFCV was increased and Fmed decreased, and
no differences were found in the MFCVs of patients with
postpolio syndrome and healthy subjects.
Some studies additionally performed MU size estima-
tions (Roeleveld et al., 1997; Drost et al., 2004a; Wood
et al., 2001) with good results, demonstrating that the
extraction of single MUAPs from HD-sEMG data to esti-
mate MU size facilitates the differentiation between neuro-
genic MUAPs and MUAPs of healthy subjects. It needs to
be noted that the technique is still very time-consuming.
The study by El Dassouki and Lefaucheur (2002) did
not provide useful information about the diagnostic role
of HD-sEMG and neither did (Bahm et al.’s, 2003) meth-
odological observation.
4.2.3. HD-sEMG and myopathies
In muscular dystrophy several studies report a slowed
MFCV, which is probably based on muscle fiber atrophy
(Yamada et al., 1991; Kumagai and Yamada, 1991). Dis-
turbed MFCVs seem to be a hallmark in channelopathies.
In the transient paresis phase of recessive myotonia our
group found a unique conduction block along the muscle
fiber (Drost et al., 2001).
It was found that the disorder in the HOPP family
described earlier was caused by a mutation (Arg528His)
in the dihydropyridine receptor CACNL1A3. These find-
ings underscore that HD-sEMG seems especially valuable
and yields novel pathophysiological insights into disorders
characterized by sarcolemmal disturbances.
4.2.4. HD-sEMG and positive muscle phenomena
Spontaneous muscle–fiber activity cannot be recorded
via sEMG as fibrillation potentials, positive sharp waves
and myotonic discharges, being single muscle fiber dis-
charges, are beyond its scope. Indeed, researchers who
applied HD-sEMG to examine patients with myotonic syn-
dromes were unable to detect myotonic discharges (Kuma-
gai and Yamada, 1991; Zwarts and van Weerden, 1989;
Drost et al., 2001; Drost et al., 2004b; Van Beekvelt
et al., 2006).
Rather, and better than needle EMG, HD-sEMG is
intrinsically suited to detect spontaneous MU activity like
fasciculation potentials (Drost et al., submitted for publica-
tion). Future studies in clinical settings will exploit this
property further. To illustrate the technique’s potential,
the HD-sEMG study of spontaneous MU activity in a
child with Satoyoshi syndrome revealed new pathophysio-
logical information about the involuntary contractions
(Drost et al., 2004c, 2006).
4.2.5. HD-sEMG and the central nervous system
Decomposition of HD-sEMG signals not only yields
information about the MUs themselves, but also about
how MUs are controlled by the central nervous system
(CNS) in generating force (De Luca et al., 2006). In patients
with Parkinson’s disease, ALS, CVA and other disorders,
studies of MU firings provide a non-invasive access to cen-
tral processes and their pathophysiology. Decomposing the
HD-sEMG signals and identifying single MUs will not only
further our understanding of the pathophysiology of these
diseases, but also our understanding of the raw EMG sig-
nal. Because of the recording stability of HD-sEMG, MU
information that can only be derived from long recording
session with many thousands of discharges of the same
MU come into view (Kleine et al., 2006).
4.3. How to review a diagnostic tool
Frequently, due to their potential clinical values, expec-
tations with respect to new, innovative diagnostic tests are
high and sometimes even overheated. Because of the rapid
advances in technology there is an increasing need for ade-
quate evaluation techniques to test new diagnostic tools
and several systematic methods have been developed
(Tatsioni et al., 2005; Bossuyt et al., 2003). To improve
the accuracy and comprehensiveness of reports on diagnos-
tic accuracy a group of scientists and editors has drawn up
the Standard for Reporting of Diagnostic Accuracy or
STARD (Bossuyt et al., 2003). It comprises a 25-item
checklist and a flow diagram to help authors secure well
performed studies. Unfortunately, as HD-sEMG is not
yet utilized as a clinical application, its diagnostic accuracy
cannot yet be evaluated in keeping with these new
conventions.
 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602
Tatsioni et al. (2005) state that, as opposed to the
quality of studies of therapeutic interventions, little is
known about the methodological quality of studies of
diagnostic tests and that outcomes often still have an
‘‘anecdotal’’ character. They distinguish six levels in their
proposed review of diagnostic technologies: (1) technical
feasibility and optimization, (2) diagnostic accuracy, (3)
diagnostic thinking impact, (4) therapeutic thinking
impact, (5) patient outcome impact and (6) social impact.
Evidently, these levels are not only applicable for evalu-
ation purposes but can also serve as guidelines in the
refinement of novel diagnostic tools like HD-sEMG.
The present review has shown that the research into
the clinical applications of HD-sEMG is almost entirely
restricted to testing the first level. The only level-two
study that examined the diagnostic accuracy of the
HD-sEMG method is the study by Huppertz et al.
(1997). Although the results they report look promising,
prospective studies using blinded observers are needed to
corroborate the outcomes.
4.4. HD-sEMG and future perspectives
For proper diagnostics in the clinical neurophysiological
practice it is important to examine at MU level. However, a
decomposition problem HD-sEMG shares with that of
needle EMG is that the analysis of single MU firing pat-
terns and MU characteristics is still too complicated and
time-consuming, rendering it less viable for use in the clin-
ical practice. To overcome this problem, more and better-
validated automated analyses are crucial. A first, promising
step in that direction has recently been taken by De Luca
et al. (2006). For HD-sEMG to become a sophisticated,
clinically accepted diagnostic tool it requires a more sys-
tematic evaluation process that should also be exploited
to guide its refinement.
A large body of research has been dedicated to engineer-
ing and physiology, which has yielded valuable insights and
new methods. In effect, there is a ‘‘mismatch’’ between the
number of practical application studies and the number of
supporting bioengineering reports. A similar tendency can
be found for studies on special needle EMG techniques and
advanced data quantification techniques, e.g. in MU
decomposition studies: the development in clinical applica-
tions lags behind. This may in part be explained by the
intrinsic attraction of EMG-related signal analysis prob-
lems to the (bio)engineering world. Physiologically rela-
tively well-understood signal sources, the firing motor
units, occur in an intermingled and blurred signal, a typical
problem for the signal processing sciences and a challeng-
ing aspect for modeling. It also explains why ‘‘signal’’ pro-
cessing is the largest discipline within bioengineering, larger
than, for instance, instrumentation.
Such rather negative conclusion can be turned into a
positive prospect, as many building blocks for a successful
journey of HD-sEMG possibilities into the clinical practice
are available already.
597
Acknowledgements
We like to thank Alice Tillema for her help in the liter-
ature search and Hanneke Meulenbroek-van der Meulen
for her help with the English language.
References
Arendt-Nielsen L, Zwarts M. Measurement of muscle fiber conduction
velocity in humans: techniques and applications. J Clin Neurophysiol
1989;6:173–90.
Arendt-Nielsen L, Gantchev N, Sinkjaer T. The influence of muscle length
on muscle–fiber conduction–velocity and development of muscle
fatigue. Electroencephalogr Clin Neurophysiol 1992;85: 166–72.
Azpiroz F, Fernandez-Fraga X, Merletti R, Enck P. The puborectalis
muscle. Neurogastroenterol Motil 2005;17(Suppl. 1):68–72.
Bahm J, Meinecke L, Brandenbusch V, Rau G, Disselhorst-Klug C. High
spatial resolution electromyography and video-assisted movement
analysis in children with obstetric brachial plexus palsy. Hand Clin
2003;19:393.
Beck RBJ, Houtman CJ, O’Malley MJ, Lowery MM, Stegeman DF. A
technique to track individual motor unit action potentials in surface
EMG by monitoring their conduction velocities and amplitudes. IEEE
Trans Biomed Eng 2005;52:622–9.
Blok JH, Stegeman DF, van OA. Three-layer volume conductor model
and software package for applications in surface electromyography.
Ann Biomed Eng 2002a;30:566–77.
Blok JH, Van Dijk JP, Drost G, Zwarts MJ, Stegeman DF. A high-density
multichannel surface electromyography system for the characterization
of single motor units. Rev Scientific Instrum 2002b;73:1887–97.
Blok JH, Van Dijk JP, Zwarts MJ, Stegeman DF. Motor unit action
potential topography and its use in motor unit number estimation.
Muscle Nerve 2005;32:280–91.
Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig
LM, et al. Towards complete and accurate reporting of studies of
diagnostic accuracy: the STARD initiative43. Ann Clin Biochem
2003;40:357–63.
Castroflorio T, Farina D, Bottin A, Debernardi C, Bracco P, Merletti R,
et al. Non-invasive assessment of motor unit anatomy in jaw-elevator
muscles. J Oral Rehabil 2005a;32:708–13.
Castroflorio T, Farina D, Bottin A, Piancino MG, Bracco P, Merletti R.
Surface EMG of jaw elevator muscles: effect of electrode location and
inter-electrode distance. J Oral Rehabil 2005b;32:411–7.
Cescon C, Gazzoni M, Gobbo M, Orizio C, Farina D. Non-invasive
assessment of single motor unit mechanomyographic response and
twitch force by spike-triggered averaging. Med Biol Eng Comput
2004;42:496–501.
Cescon C, Sguazzi E, Merletti R, Farina D. Non-invasive characterization
of single motor unit electromyographic and mechanomyographic
activities in the biceps brachii muscle. J Electromyogr Kinesiol
2006;16:17–24.
Chauvet E, Fokapu O, Hogrel JY, Gamet D, Duchene J. Automatic
identification of motor unit action potential trains from electromyo-
graphic signals using fuzzy techniques. Med Biol Eng Comput
2003;41:646–53.
Chen JJ, Sun TY, Lin TH, Lin TS. Spatio-temporal representation of
multichannel EMG firing patterns and its clinical applications. Med
Eng Phys 1997;19:420–30.
Christova P, Kossev A, Kristev I, Chichov V. Surface EMG recorded by
branched electrodes during sustained muscle activity. J Electromyogr
Kinesiol 1999;9:263–76.
De Luca CJ, Adam A, Wotiz R, Gilmore LD, Nawab SH. Decomposition
of surface EMG signals. J Neurophysiol 2006;96:1646–57.
Dimitrov GV, Dimitrova NA. Effect of parameters altering with muscle
fibre functional state on power spectra of spatially filtered extracellular
potentials. J Med Eng Techn 2001;25:74–8.
598 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602
Dimitrova NA, Dimitrov GV, Nikitin A. Longitudinal variations of
characteristic frequencies of skeletal muscle fibre potentials detected by
a bipolar electrode or multi-electrode. J Med Eng Techn 2001;25:
34–40.
Dimitrov GV, Disselhorst-Klug C, Dimitrova NA, Schulte E, Rau G.
Simulation analysis of the ability of different types of multi-electrodes
to increase selectivity of detection and to reduce cross-talk. J
Electromyogr Kinesiol 2003;13:125–38.
DisselhorstKlug C, Silny J, Rau G. Improvement of spatial resolution in
surface-EMG: A theoretical and experimental comparison of different
spatial filters. IEEE Trans Biomed Eng 1997;44:567–74.
DisselhorstKlug C, Silny J, Rau G. Estimation of the relationship between
the noninvasively detected activity of single motor units and their
characteristic pathological changes by modelling. J Electromyogr
Kinesiol 1998;8:323–35.
DisselhorstKlug C, Rau G, Schmeer A, Silny J. Non-invasive detection of
the single motor unit action potential by averaging the spatial potential
distribution triggered on a spatially filtered motor unit action
potential. J Electromyogr Kinesiol 1999;9:67–72.
DisselhorstKlug C, Bahm J, Ramaekers V, Trachterna A, Rau G. Non-
invasive approach of motor unit recording during muscle contractions
in humans. Eur J Appl Physiol 2000;83:144–50.
Drost G, Blok JH, Stegeman DF, Van Dijk JP, Van Engelen BG, Zwarts
MJ. Propagation disturbance of motor unit action potentials during
transient paresis in generalized myotonia: a high-density surface EMG
study. Brain 2001;124:352–60.
Drost G, Stegeman DF, Schillings ML, Horemans HL, Janssen HM,
Massa M, et al.. Motor unit characteristics in healthy subjects and
those with postpoliomyelitis syndrome: a high-density surface EMG
study. Muscle Nerve 2004a;30:269–76.
Drost G, Van Dijk JP, Stegeman DF, Van Engelen BG, Zwarts MJ.
Maintaining constant voluntary force in generalized myotonia despite
muscle membrane disturbances: insights from a high-density surface
EMG study. J Clin Neurophysiol 2004b;21:114–23.
Drost G, Verrips A, Hooijkaas H, Zwarts M. Glutamic acid
decarboxylase antibodies in Satoyoshi syndrome. Ann Neurol
2004c;55:450–1.
Drost G, Verrips A, Van Engelen BGM, Stegeman DF, Zwarts M.
Involuntary painful muscle contractions in Satoyoshi syndrome. A
surface EMG study. Movement Disorders 2006 [in press].
Drost G, Kleine BU, Stegeman DF, Van Engelen BGM, Zwarts M.
Fasciculation potentials in high-density suface EMG [submitted for
publication].
El Dassouki M, Lefaucheur JP. The correlation between muscle and nerve
fiber conduction velocities in thenar muscle is lost in case of carpal
tunnel syndrome. Clin Neurophysiol 2002;113:1121–4.
Enck P, Hinninghofen H, Wietek B, Becker HD. Functional asymmetry of
pelvic floor innervation and its role in the pathogenesis of fecal
incontinence. Digestion 2004;69:102–11.
Enck P, Hinninghofen H, Merletti R, Azpiroz F. The external anal
sphincter and the role of surface electromyography. Neurogastroen-
terol Motil 2005;17(Suppl. 1):60–7.
Falla D, Farina D. Muscle fiber conduction velocity of the upper trapezius
muscle during dynamic contraction of the upper limb in patients with
chronic neck pain. Pain 2005;116:138–45.
Falla D, Dall’Alba P, Rainoldi A, Merletti R, Jull G. Location of
innervation zones of sternocleidomastoid and scalene muscles–a basis
for clinical and research electromyography applications. Clin Neuro-
physiol 2002;113:57–63.
Farina D, Cescon C. Concentric-ring electrode systems for noninvasive
detection of single motor unit activity. IEEE Trans Biomed Eng
2001;48:1326–34.
Farina D, Merletti R. A novel approach for precise simulation of the
EMG signal detected by surface electrodes. IEEE Trans Biomed Eng
2001;48:637–46.
Farina D, Merletti R. Estimation of average muscle fiber conduction
velocity from two-dimensional surface EMG recordings. J Neurosci-
ence Methods 2004;134:199–208.
Farina D, Mesin L. Sensitivity of surface EMG-based conduction velocity
estimates to local tissue in –homogeneities – influence of the number of
channels and inter-channel distance. J Neurosci Methods 2005;142:
83–9.
Farina D, Fortunato E, Merletti R. Noninvasive estimation of motor unit
conduction velocity distribution using linear electrode arrays. IEEE
Trans Biomed Eng 2000;47:380–8.
Farina D, Muhammad W, Fortunato E, Meste O, Merletti R, Rix H.
Estimation of single motor unit conduction velocity from surface
electromyogram signals detected with linear electrode arrays. Med Biol
Eng Comput 2001;39:225–36.
Farina D, Cescon C, Merletti R. Influence of anatomical, physical, and
detection-system parameters on surface EMG. Biol Cybernet
2002a;86:445–56.
Farina D, Fosci M, Merletti R. Motor unit recruitment strategies
investigated by surface EMG variables. J Appl Physiol 2002b;92: 235–47.
Farina D, Merletti R, Indino B, Nazzaro M, Pozzo M. Surface EMG
crosstalk between knee extensor muscles: experimental and model
results. Muscle Nerve 2002c;26:681–95.
Farina D, Arendt-Nielsen L, Merletti R, Graven-Nielsen T. Assessment of
single motor unit conduction velocity during sustained contractions of
the tibialis anterior muscle with advanced spike triggered averaging. J
Neurosci Methods 2002d;115:1–12.
Farina D, Arendt-Nielsen L, Merletti R, Indino B, Graven-Nielsen T.
Selectivity of spatial filters for surface EMG detection from the tibialis
anterior muscle. IEEE Trans Biomed Eng 2003a;50:354–64.
Farina D, Schulte E, Merletti R, Rau G, Disselhorst-Klug C. Single motor
unit analysis from spatially filtered surface electromyogram signals.
Part I: spatial selectivity. Med Biol Eng Comput 2003b;41:330–7.
Farina D, Fevotte C, Doncarli C, Merletti R. Blind separation of linear
instantaneous mixtures of nonstationary surface myoelectric signals.
IEEE Trans Biomed Eng 2004a;51:1555–67.
Farina D, Gazzoni M, Camelia F. Low-threshold motor unit membrane
properties vary with contraction intensity during sustained activation
with surface EMG visual feedback. J Appl Physiol 2004b;96:
1505–15.
Farina D, Merletti R, Enoka RM. The extraction of neural strategies from
the surface EMG. J Appl Physiol 2004c;96:1486–95.
Farina D, Mesin L, Martina S. Advances in surface electromyographic
signal simulation with analytical and numerical descriptions of the
volume conductor. Med Biol Eng Comput 2004d;42:467–76.
Farina D, Mesin L, Martina S, Merletti R. A surface EMG generation
model with multilayer cylindrical description of the volume conductor.
IEEE Trans Biomed Eng 2004e;51:415–26.
Farina D, Mesin L, Martina S, Merletti R. Comparison of spatial filter
selectivity in surface myoelectric signal detection: influence of the
volume conductor model. Med Biol Eng Comput 2004f;42:114–20.
Farina D, Pozzo M, Merlo E, Bottin A, Merletti R. Assessment of average
muscle fiber conduction velocity from surface EMG signals during
fatiguing dynamic contractions. IEEE Trans Biomed Eng 2004g;51:
1383–93.
Farina D, Zagari D, Gazzoni M, Merletti R. Reproducibility of muscle–
fiber conduction velocity estimates using multichannel surface EMG
techniques. Muscle Nerve 2004h;29:282–91.
Farina D, Gazzoni M, Camelia F. Conduction velocity of low-threshold
motor units during ischemic contractions performed with surface
EMG feedback. J Appl Physiol 2005a;98:1487–94.
Farina D, Arendt-Nielsen L, Graven-Nielsen T. Experimental muscle pain
decreases voluntary EMG activity but does not affect the muscle
potential evoked by transcutaneous electrical stimulation. Clin Neu-
rophysiol 2005b;116:1558–65.
Farina D, Arendt-Nielsen L, Graven-Nielsen T. Experimental muscle pain
reduces initial motor unit discharge rates during sustained submaximal
contractions. J Appl Physiol 2005c;98:999–1005.
Farina D, Arendt-Nielsen L, Graven-Nielsen T. Spike-triggered average
torque and muscle fiber conduction velocity of low-threshold motor
units following submaximal endurance contractions. J Appl Physiol
2005d;98:1495–502.
 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602
Fattorini L, Felici F, Filligoi GC, Traballesi M, Farina D. Influence
of high motor unit synchronization levels on non-linear and
spectral variables of the surface EMG. J Neurosci Methods
2005;143:133–9.
Garcia GA, Okuno R, Akazawa K. A decomposition algorithm for
surface electrode-array electromyogram. A noninvasive, three-step
approach to analyze surface EMG signals. IEEE Eng Med Biol Mag
2005;24:63–72.
Gazzoni M, Farina D, Merletti R. A new method for the extraction and
classification of single motor unit action potentials from surface EMG
signals. J Neurosci Methods 2004;136:165–77.
Geister DE, Steusloff PM, Ash MM. Application of a surface electrode
array in the measurement of Emg activity during masticatory function.
J Dental Res 1980;59:487.
Gerilovsky L, Karadimov D, Ianakiev B. Hypoxia reduces the conduction
velocity of the excitation along the striated muscles in man. Electro-
myogr Clin Neurophysiol 1991;31:203–8.
Grassme R, Stegeman DF, Drost G, Schumann NP, Scholle HC. Selective
spatial information from surface EMG after temporal filtering: the
application to interference EMG using cross-covariance analysis. Clin
Neurophysiol 2003;114:2338–46.
Gronlund C, Ostlund N, Roeleveld K, Karlsson JS. Simultaneous
estimation of muscle fibre conduction velocity and muscle fibre
orientation using 2D multichannel surface electromyogram. Med Biol
Eng Comput 2005a;43:63–70.
Gronlund C, Roeleveld K, Holtermann A, Karlsson JS. On-line signal
quality estimation of multichannel surface electromyograms. Med Biol
Eng Comput 2005b;43:357–64.
Hagg GM. Action-potential velocity-measurements in the upper trapezius
muscle. J Electromyogr Kinesiol 1993;3:231–5.
Haig AJ, Gelblum JB, Rechtien JJ, Gitter AJ. Technology assessment: the
use of surface EMG in the diagnosis and treatment of nerve and
muscle disorders46. Muscle Nerve 1996;19:392–5.
Hanayama K. Recovery of conduction velocity of muscle fiber action
potential after strenuous isometric contraction. Jpn J Physiol
1994;44:75–88.
Harba MI, Lynn PA. Optimizing the acquisition and processing of surface
electromyographic signals. J Biomed Eng 1981;3:100–6.
Harba MI, Teng LY. Reliability of measurement of muscle fiber
conduction velocity using surface EMG. Front Med Biol Eng
1999;9:31–47.
Hermens HJ, Freriks B, sselhorst-Klug C, Rau G. Development of
recommendations for SEMG sensors and sensor placement proce-
dures. J Electromyogr Kinesiol 2000;10:361–74.
Hilfiker P, Meyer M. Normal and myopathic propagation of surface
motor unit action potentials. Electroencephalogr Clin Neurophysiol
1984;57:21–31.
Hogrel JY. Clinical applications of surface electromyography in neuro-
muscular disorders. Neurophysiol Clin 2005;35:59–71.
Holtermann A, Roeleveld K, Karlsson JS. Inhomogeneities in muscle
activation reveal motor unit recruitment. J Electromyogr Kinesiol
2005a;15:131–7.
Holtermann A, Roeleveld K, Vereijken B, Ettema G. Changes in agonist
EMG activation level during MVC cannot explain early strength
improvement. Eur J Appl Physiol 2005b;94:593–601.
Houtman CJ, Stegeman DF, Van Dijk JP, Zwarts MJ. Changes in muscle
fiber conduction velocity indicate recruitment of distinct motor unit
populations. J Appl Physiol 2003;95:1045–54.
Huppertz HJ, sselhorst-Klug C, Silny J, Rau G, Heimann G. Diagnostic
yield of noninvasive high spatial resolution electromyography in
neuromuscular diseases. Muscle Nerve 1997;20:1360–70.
Iwasaki S, Tokunaga T, Baba S, Tanaka M, Kawazoe T. Noninvasive
estimation of the location of the end plate in the human masseter
muscle using surface electromyograms with an electrode array. J Osaka
Dent Univ 1990;24:135–40.
Kaneko H, Kiryu T, Saitoh Y. Compensation for the distortion of bipolar
surface EMG signals caused by innervation zone movement. IEICE
Trans Informat Systems 1996;E79D:373–81.
599
Kiryu T, Takahashi K, Ogawa K. Multivariate analysis of muscular
fatigue during bicycle ergometer exercise. IEEE Trans Biomed Eng
1997;44:665–72.
Kleine BU, Blok JH, Oostenveld R, Praamstra P, Stegeman DF.
Magnetic stimulation-induced modulations of motor unit firings
extracted from multi-channel surface EMG. Muscle Nerve 2000a;23:
1005–15.
Kleine BU, Schumann NP, Stegeman DF, Scholle HC. Surface EMG
mapping of the human trapezius muscle: the topography of
monopolar and bipolar surface EMG amplitude and spectrum
parameters at varied forces and in fatigue. Clin Neurophysiol
2000b;111:686–93.
Kleine BU, Praamstra P, Stegeman DF, Zwarts MJ. Impaired motor
cortical inhibition in Parkinson’s disease: motor unit responses to
transcranial magnetic stimulation. Exp Brain Res 2001;138:477–83.
Kleine BU, Van Dijk JP, Lapatki BG, Zwarts MJ, Stegeman DF (2006)
Using two-dimensional spatial information in decomposition of
surface EMG signals. J Electromyogr Kinesiol [in press].
doi:10.1016/j.jelekin.2006.05.003.
Knaflitz M, Bonato P. Time-frequency methods applied to muscle fatigue
assessment during dynamic contractions. J Electromyogr Kinesiol
1999;9:337–50.
Koh TJ, Grabiner MD. Evaluation of methods to minimize cross-talk in
surface electromyography. J Biomech 1993;26:151–7.
Konno M, Sato K, Mito T, Mitani H. Relationship between the direction
of mandibular growth and masseter muscle conduction velocity. Am J
Orthod Dentofacial Orthop 2005;128:35–43.
Kumagai K, Yamada M. The clinical use of multichannel surface
electromyography. Acta Paediatr Jpn 1991;33:228–37.
Lanzetta M, Pozzo M, Bottin A, Merletti R, Farina D. Reinnervation of
motor units in intrinsic muscles of a transplanted hand. Neurosci Lett
2005;373:138–43.
Lapatki BG, Van Dijk JP, Jonas IE, Zwarts MJ, Stegeman DF. A thin,
flexible multielectrode grid for high-density surface EMG. J Appl
Physiol 2004;96:327–36.
Lapatki BG, Oostenveld R, Van Dijk JP, Jonas IE, Zwarts MJ, Stegeman
DF. Topographical characteristics of motor units of the lower facial
musculature revealed by means of high-density surface EMG. J
Neurophysiol 2006;95:342–54.
Linssen WHJP, Jacobs M, Stegeman DF, Joosten EMG, Moleman J.
Muscle fatigue in mcardles disease – muscle–fiber conduction-velocity
and surface EMG frequency-spectrum during ischemic exercise. Brain
1990;113:1779–93.
Linssen WH, Stegeman DF, Joosten EM, Binkhorst RA, Merks MJ, ter
Laak HJ, et al.. Fatigue in type I fiber predominance: a muscle force
and surface EMG study on the relative role of type I and type II
muscle fibers. Muscle Nerve 1991a;14:829–37.
Linssen WH, Stegeman DF, Joosten EM, Merks HJ, ter Laak HJ,
Binkhorst RA, et al.. Force and fatigue in human type I muscle fibres.
A surface EMG study in patients with congenital myopathy and type I
fibre predominance. Brain 1991b;114(Pt 5):2123–32.
Linssen WHJP, Stegeman DF, Merks MJH, Binkhorst RA, Notermans
SLH. Electromyographic evidence of delayed fatigue-induced sarco-
lemmal excitability impairment in McArdle’s disease. J Electromyogr
Kinesiol 1996;6:147–57.
Maisetti O, Guevel A, Legros P, Hogrel JY. Prediction of endurance
capacity of quadriceps muscles in humans using surface electromyo-
gram spectrum analysis during submaximal voluntary isometric
contractions. Eur J Appl Physiol 2002;87:509–19.
Masuda T, DeLuca CJ. Recruitment threshold and muscle–fiber conduc-
tion-velocity of single motor units. J Electromyogr Kinesiol
1991;1:116–23.
Masuda T, Sadoyama T. The propagation of single motor unit action
potentials detected by a surface electrode array. Electroencephalogr
Clin Neurophysiol 1986;63:590–8.
Masuda T, Sadoyama T. Skeletal muscles from which the propagation of
motor unit action potentials is detectable with a surface electrode
array. Electroencephalogr Clin Neurophysiol 1987;67:421–7.
600 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602
Masuda T, Sadoyama T. Processing of myoelectric signals for estimating
the location of innervation zones in the skeletal muscles. Front Med
Biol Eng 1989;1:299–314.
Masuda T, Miyano H, Sadoyama T. The distribution of myoneural
junctions in the biceps brachii investigated by surface electromyogra-
phy. Electroencephalogr Clin Neurophysiol 1983a;56:597–603.
Masuda T, Miyano H, Sadoyama T. The propagation of motor unit
action potential and the location of neuromuscular junction investi-
gated by surface electrode arrays. Electroencephalogr Clin Neuro-
physiol 1983b;55:594–600.
Masuda T, Miyano H, Sadoyama T. A surface electrode array for
detecting action potential trains of single motor units. Electroencep-
halogr Clin Neurophysiol 1985;60:435–43.
Masuda T, Kizuka T, Zhe JY, Yamada H, Saitou K, Sadoyama T, et al..
Influence of contraction force and speed on muscle fiber conduction
velocity during dynamic voluntary exercise. J Electromyogr Kinesiol
2001;11:85–94.
McGill KC. Surface electromyogram signal modelling. Med Biol Eng
Comput 2004;42:446–54.
Merletti R, Parker PA. Electromyography. Physiology, engineering and
noninvasive applications. 1st ed. Hoboken, New Jersey: IEEE Press,
Wiley; 2004.
Merletti R, Roy SH, Kupa E, Roatta S, Granata A. Modeling of surface
myoelectric signals – Part II: Model-based signal interpretation. IEEE
Trans Biomed Eng 1999;46:821–9.
Merletti R, Farina D, Gazzoni M. The linear electrode array: a useful tool
with many applications. J Electromyogr Kinesiol 2003;13:37–47.
Merletti R, Bottin A, Cescon C, Farina D, Gazzoni M, Martina S, et al..
Multichannel surface EMG for the non-invasive assessment of the anal
sphincter muscle. Digestion 2004;69:112–22.
Merlo A, Farina D, Merletti R. A fast and reliable technique for muscle
activity detection from surface EMG signals. IEEE Trans Biomed Eng
2003;50:316–23.
Mesin L, Farina D. Simulation of surface EMG signals generated by
muscle tissues with inhomogeneity due to fiber pinnation. IEEE Trans
Biomed Eng 2004;51:1521–9.
Mesin L, Farina D. A model for surface EMG generation in volume
conductors with spherical inhomogeneities. IEEE Trans Biomed Eng
2005;52:1984–93.
Mesin L, Farina D. An analytical model for surface EMG generation in
volume conductors with smooth conductivity variations. IEEE Trans
Biomed Eng 2006;53:773–9.
Miyano H, Sadoyama T. Theoretical analysis of surface EMG in
voluntary isometric contraction. Eur J Appl Physiol Occup Physiol
1979;40:155–64.
Nakamura Y, Kiryu T, Saitoh Y. Multivariate analysis of multichannel
surface myoelectric signals to determine muscular fatigue. Methods Inf
Med 1997;36:302–5.
Nakamura H, Yoshida M, Kotani M, Akazawa K, Moritani T. The
application of independent component analysis to the multi-channel
surface electromyographic signals for separation of motor unit action
potential trains: part I-measuring techniques. J Electromyogr Kinesiol
2004a;14:423–32.
Nakamura H, Yoshida M, Kotani M, Akazawa K, Moritani T. The
application of independent component analysis to the multi-channel
surface electromyographic signals for separation of motor unit action
potential trains: part II-modelling interpretation. J Electromyogr
Kinesiol 2004b;14:433–41.
Nishizono H, Kurata H, Miyashita M. Muscle fiber conduction velocity
related to stimulation rate. Electroencephalogr Clin Neurophysiol
1989;72:529–34.
Nishizono H, Fujimoto T, Ohtake H, Miyashita M. Muscle fiber
conduction velocity and contractile properties estimated from surface
electrode arrays. Electroencephalogr Clin Neurophysiol 1990;75:
75–81.
Okajima Y, Tsubahara A, Kondo K, Chino N, Noda Y, Tomita Y. A new
method of estimating the distribution of muscle fiber conduction
velocities. Electroencephalogr Clin Neurophysiol 1995;97:310–7.
Okajima Y, Toikawa H, Hanayama K, Ohtsuka T, Kimura A,
Chino N. Relationship between nerve and muscle fiber conduc-
tion velocities of the same motor unit in man. Neurosci Lett
1998;253:65–7.
Ollivier K, Portero P, Maisetti O, Hogrel JY. Repeatability of surface
EMG parameters at various isometric contraction levels and during
fatigue using bipolar and Laplacian electrode configurations. J
Electromyogr Kinesiol 2005;15:466–73.
Ostlund N, Yu J, Roeleveld K, Karlsson JS. Adaptive spatial filtering of
multichannel surface electromyogram signals. Med Biol Eng Comput
2004;42:825–31.
Pozzo M, Bottin A, Ferrabone R, Merletti R. Sixty-four channel wearable
acquisition system for long-term surface electromyogram recording
with electrode arrays. Med Biol Eng Comput 2004;42:455–66.
Prutchi D. A high-resolution large array (HRLA) surface EMG system.
Med Eng Phys 1995;17:442–54.
Pullman SL, Goodin DS, Marquinez AI, Tabbal S, Rubin M. Clinical
utility of surface EMG: report of the therapeutics and technology
assessment subcommittee of the American Academy of Neurology.
Neurology 2000;55:171–7.
Ramaekers VT, DisselhorstKlug C, Schneider J, Silny J, Forst J, Forst R,
et al.. Clinical-application of a noninvasive multielectrode array EMG
for the recording of single motor unit-activity. Neuropediatrics
1993;24:134–8.
Rau G, DisselhorstKlug C. Principles of high-spatial-resolution surface
EMG (HSR-EMG): Single motor unit detection and application in the
diagnosis of neuromuscular disorders. J Electromyogr Kinesiol
1997;7:233–9.
Rau G, Disselhorst-Klug C, Silny J. Noninvasive approach to motor unit
characterization: muscle structure, membrane dynamics and neuronal
control. J Biomech 1997;30:441–6.
Rau G, Schulte E, Disselhorst-Klug C. From cell to movement: to what
answers does EMG really contribute?. J Electromyogr Kinesiol
2004;14:611–7.
Reucher H, Rau G, Silny J. Spatial-filtering of noninvasive multielectrode
Emg .1. Introduction to measuring technique and applications. IEEE
Trans Biomed Eng 1987a;34:98–105.
Reucher H, Silny J, Rau G. Spatial-filtering of noninvasive multielectrode
Emg .2. Filter performance in theory and modeling. IEEE Trans
Biomed Eng 1987b;34:106–13.
Roeleveld K, Stegeman DF. What do we learn from motor unit action
potentials in surface electromyography?. Muscle Nerve 2002(Suppl. 11):
S92–7.
Roeleveld K, Stegeman DF, Falck B, Stalberg EV. Motor unit size
estimation: confrontation of surface EMG with macro EMG. Elec-
tromyogr Motor Control-Electroencephalogr Clin Neurophysiol
1997;105:181–8.
Roeleveld K, Sandberg A, Stalberg EV, Stegeman DF. Motor unit size
estimation of enlarged motor units with surface electromyography.
Muscle Nerve 1998;21:878–86.
Roeleveld K, Van Engelen BG, Stegeman DF. Possible mechanisms of
muscle cramp from temporal and spatial surface EMG characteristics.
J Appl Physiol 2000;88:1698–706.
Sadoyama T, Masuda T, Miyano H. Optimal conditions for the
measurement of muscle fibre conduction velocity using surface
electrode arrays. Med Biol Eng Comput 1985;23:339–42.
Sadoyama T, Masuda T, Miyata H, Katsuta S. Fibre conduction velocity
and fibre composition in human vastus lateralis. Eur J Appl Physiol
Occup Physiol 1988;57:767–71.
Saitou K, Masuda T, Michikami D, Kojima R, Okada M. Innervation
zones of the upper and lower limb muscles estimated by using
multichannel surface EMG. J Hum Ergol (Tokyo) 2000;29:35–52.
Sbriccoli P, Bazzucchi I, Rosponi A, Bernardi M, De VG, Felici F.
Amplitude and spectral characteristics of biceps Brachii sEMG depend
upon speed of isometric force generation. J Electromyogr Kinesiol
2003;13:139–47.
Schillings ML, Kalkman JS, van der Werf SP, Van Engelen BG,
Bleijenberg G, Zwarts MJ. Diminished central activation during
 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602
maximal voluntary contraction in chronic fatigue syndrome. Clin
Neurophysiol 2004;115:2518–24.
Schneider J, Rau G, Silny J. A noninvasive EMG technique for
investigating the excitation propagation in single motor units.
Electromyogr Clin Neurophysiol 1989;29:273–80.
Schneider J, Silny J, Rau G. Influence of tissue inhomogeneities on
noninvasive muscle–fiber conduction-velocity measurements – investi-
gated by physical and numerical modeling. IEEE Trans Biomed Eng
1991;38:851–60.
Schulte E, Ciubotariu A, Arendt-Nielsen L, Disselhorst-Klug C, Rau G,
Graven-Nielsen T. Experimental muscle pain increases trapezius
muscle activity during sustained isometric contractions of arm muscles.
Clin Neurophysiol 2004a;115:1767–78.
Schulte E, Farina D, Merletti R, Rau G, Disselhorst-Klug C. Influence of
muscle fibre shortening on estimates of conduction velocity and
spectral frequencies from surface electromyographic signals. Med Biol
Eng Comput 2004b;42:477–86.
Schulte E, Dimitrova NA, Dimitrov GV, Rau G, Disselhorst-Klug C.
Estimation of the muscle fibre semi-length under varying joint
positions on the basis of non-invasively extracted motor unit action
potentials. J Electromyogr Kinesiol 2005;15:290–9.
Schulte E, Kallenberg LAC, Christensen H, Disselhorst-Klug C, Hermens
HJ, Rau G, et al.. Comparison of the electromyographic activity in
the upper trapezius and biceps brachii muscle in subjects with muscular
disorders: a pilot study. Eur J Appl Physiol 2006a;96:185–93.
Schulte E, Miltner O, Junker E, Rau G, Disselhorst-Klug C. Upper
trapezius muscle conduction velocity during fatigue in subjects
with and without work-related muscular disorders: a non-invasive
high spatial resolution approach. Eur J Appl Physiol 2006b;96:
194–202.
Staudenmann D, Kingma I, Stegeman DF, van Dieen JH. Towards
optimal multi-channel EMG electrode configurations in muscle force
estimation: a high density EMG study. J Electromyogr Kinesiol
2005;15:1–11.
Staudenmann D, Kingma I, Daffertshofer A, Stegeman DF, van Dieen
JH. Improving EMG-based muscle force estimation by using a high-
density EMG grid and principal component analysis. IEEE Trans
Biomed Eng 2006;53:712–9.
Stegeman DF, Blok JH, Hermens HJ, Roeleveld K. Surface EMG models:
properties and applications. J Electromyogr Kinesiol 2000a;10:313–26.
Stegeman DF, Zwarts MJ, Anders C, Hashimoto T. Multi-channel surface
EMG in clinical neurophysiology. Suppl Clin Neurophysiol
2000b;53:155–62.
Sun TY, Chen JJ, Lin TS. Analysis of motor unit firing patterns in
patients with central or peripheral lesions using singular-value
decomposition. Muscle Nerve 2000;23:1057–68.
Tatsioni A, Zarin DA, Aronson N, Samson DJ, Flamm CR, Schmid C,
et al.. Challenges in systematic reviews of diagnostic technologies.
Ann Intern Med 2005;142:1048–55.
Thusneyapan S, Zahalak GI. A practical electrode-array myoprocessor for
surface electromyography. IEEE Trans Biomed Eng 1989;36:295–9.
Tokunaga T, Kimura K, Tanaka M, Kawazoe T. Measurement of muscle
fiber conduction velocity in the human masseter muscle using the
surface EMG cross-correlation technique. Front Med Biol Eng
1991;3:215–9.
Tokunaga T, Baba S, Tanaka M, Kashiwagi K, Kimura K, Kawazoe T.
Two-dimensional configuration of the myoneural junctions of human
masticatory muscle detected with matrix electrode. J Oral Rehabil
1998;25:329–34.
Van Beekvelt MC, Drost G, Rongen G, Stegeman DF, Van Engelen BG,
Zwarts MJ. Na+–K+-ATPase is not involved in the warming-up
phenomenon in generalized myotonia. Muscle Nerve 2006;33:514–23.
van der Hoeven JH, Zwarts MJ, van Weerden TW. Muscle fiber
conduction velocity in amyotrophic lateral sclerosis and traumatic
lesions of the plexus brachialis. Electroencephalogr Clin Neurophysiol
1993;89:304–10.
van der Hoeven JH, Links TP, Zwarts MJ, van Weerden TW.
Muscle fiber conduction velocity in the diagnosis of familial
601
hypokalemic periodic paralysis–invasive versus surface determina-
tion. Muscle Nerve 1994;17:898–905.
van Vugt JP, van Dijk JG. A convenient method to reduce crosstalk in
surface EMG. Cobb Award-winning article, 2001. Clin Neurophysiol
2001;112:583–92.
Wood SM, Jarratt JA, Barker AT, Brown BH. Surface electromyography
using electrode arrays: a study of motor neuron disease. Muscle Nerve
2001;24:223–30.
Yamada M, Kumagai K, Uchiyama A. The distribution and propagation
pattern of motor unit action potentials studied by multi-channel
surface EMG. Electroencephalogr Clin Neurophysiol 1987;67:
395–401.
Yamada M, Kumagai K, Uchiyama A. Muscle fiber conduction velocity
studied by the multi-channel surface EMG. Electromyogr Clin
Neurophysiol 1991;31:251–6.
Zwarts MJ. Evaluation of the estimation of muscle fiber conduction
velocity. Surface versus needle method. Electroencephalogr. Clin
Neurophysiol 1989;73:544–8.
Zwarts MJ, Stegeman DF. Multichannel surface EMG: basic aspects and
clinical utility. Muscle Nerve 2003;28:1–17.
Zwarts MJ, van Weerden TW. Transient paresis in myotonic syndromes.
A surface EMG study. Brain 1989;112(Pt 3):665–80.
Zwarts MJ, van Weerden TW, Links TP, Haenen HT, Oosterhuis HJ. The
muscle fiber conduction velocity and power spectra in familial
hypokalemic periodic paralysis. Muscle Nerve 1988;11:166–73.
Zwarts MJ, Drost G, Stegeman DF. Recent progress in the diagnostic use
of surface EMG for neurological diseases. J Electromyogr Kinesiol
2000;10:287–91.
Baziel van Engelen started his medical science
education at the University of Nijmegen
(KUN) in 1975. Between 1975 and 1984 he
received his MD in medical science, and also
studied philosophy at the University of
Amsterdam which he finished cum laude in
1992. Between 1985 and 1992 he did his Neu-
rology residency at the Universities of Berlin
and Nijmegen, followed by a research fellow-
ship at the department of neurology and
immunology of the Mayo Clinic and Mayo
Foundation in Rochester MN, USA in 1993. In
1994, Baziel became a staff member of the Neurology department of the
University of Nijmegen, with a focus on neuromuscular diseases. Baziel
became associate professor for neuromuscular diseases at Nijmegen Uni-
versity in 2000, and full professor in 2003.
Baziel has published over 160 articles in his research field, and has been
entitled to a large number of research grants. His focus is on clinical and
translational research of neuromuscular disorders, especially myopathies.
He is secretary of the research school for fundamental and clinical
movement sciences (IFKB), and is member of various neurological
organisations.
Dick F. Stegeman received the M.Sc. degree in
electrical engineering from the University of
Twente, Enschede, The Netherlands in 1976.
He worked towards his PhD degree (obtained
in 1981) on model studies of human electric
nerve activity at the Medical Physics and Bio-
physics Department at the Radboud Univer-
sity, Nijmegen, The Netherlands. After a post
doc position at the otolaryngology department,
he was appointed in 1984 at the Department of
Clinical Neurophysiology as head of the
Physics Group of that department. Since 2003,
he has a part-time professorship in applied electrophysiology at the Fac-
ulty of Human Movement Sciences at the Vrije Universiteit Amsterdam.
He is director of the Interuniversity institute of Fundamental and Clinical
Human Movement Sciences (IFKB, Nijmegen, Amsterdam).
602 G. Drost et al. / Journal of Electromyography and Kinesiology 16 (2006) 586–602
His main professional interests concern electrophysiological modelling,
the theory of volume conduction and the measurement and quantitative
analysis of electroneurographic, EMG and EEG data. His work concerns
the study of neuromuscular and degenerative neurological disorders. A
key development under his supervision is the system for high-density
surface EMG. He is the co(author) of about 125 peer reviewed interna-
tional publications. He is on the editorial board of Clinical Neurophysi-
ology and in the advisory board of the Journal of electromyography and
Kinesiology. The present research is concerned with the use of spatio-
temporal information in EMG and EEG data.
Machiel J. Zwarts received his MD from the
Faculty of Medicine, University of Groningen
in 1978 and completed his residency in Neu-
rology and Clinical Neurophysiology in 1984
at the University hospital, Groningen. In
1989 he received his PhD degree on the thesis
‘‘Applications of muscle fiber conduction
velocity estimation’’ – A surface EMG study.
Since 1997 he is professor of Clinical Neu-
rophysiology at the Radboud University
Medical Centre, Nijmegen, the Netherlands.
In 1992, he was awarded the first ‘‘Storm van
Leeuwen-Magnus prijs’’ of the Dutch society of Clinical Neurophysi-
ology and in 2001 he presented the 26th Annual Edward H. Lambert
Lecture on invitation of the American Association of Electrodiagnostic
Medicine. He is chairman of the Dutch Society for Clinical Neuro-
physiology and coauthor of the second edition of ‘‘Electrodiagnostic
Medicine’’ (2002) by Dumitru D., Amato A.A. and Zwarts M.J. His
research concerns the electrophysiological measurements and diagnosis
of neuromuscular disorders and muscle fatigue. A major goal is the
development and application of multi-channel surface electromyography
as a new tool in the noninvasive diagnosis of neuromuscular disorders.
Other areas of interest are ultrasound diagnosis of neuromuscular dis-
ease, magnetic stimulation of the nervous system, electrophysiological
predictors of recovery, EMG guided botulinum toxin treatment for
movement disorders and central aspects of local muscle fatigue.
Gea Drost received her MD in 1994 from the
Radboud University of Nijmegen, The Neth-
erlands. She completed her residency in
Neurology with an emphasis on Clinical
Neurophysiology in 2000. Before, she got her
BSc in physical therapy from the Academy of
Physical Therapy in Arnhem, the Nether-
lands. At the moment, works as a Neurolo-
gist and Clinical Neurophysiologist at the
Department of Neurology (staff member) at
the Radboud University Nijmegen Medical
Centre, the Netherlands. She combines her
clinical work with PhD research concerning pathophysiological insights
and clinical applications of HD-sEMG (expected 2007). Her special
interests are EMG and neuromuscular disorders, in particular muscle
channelopathies.