Journal of Electromyography and Kinesiology 24 (2014) 11–17
Contents lists available at ScienceDirect
Journal of Electromyography and Kinesiology
journal homepage: www.elsevier.com/locate/jelekin
Electromyographic analysis of upper limb muscles during standardized
isotonic and isokinetic robotic exercise of spastic elbow in patients with
stroke
Minki Sin a,1, Won-Seok Kim b,1, Daegeun Park a, Yu-Sun Min b, Woo Jin Kim c, Kyujin Cho a,⇑
Nam-Jong Paik b
a
School of Mechanical and Aerospace Engineering, Seoul National University/IAMD, Seoul, Republic of Korea
Department of Rehabilitation Medicine, Seoul National University College of Medicine and Seoul National University Bundang Hospital, Seongnam, Republic of Korea
c
Department of Physical Medicine and Rehabilitation, Haeundae Paik Hospital, Inje University of Medicine, Busan, Republic of Korea
b
a r t i c l e
i n f o
Article history:
Received 12 April 2013
Received in revised form 3 September 2013
Accepted 4 October 2013
Keywords:
Stroke
Muscle spasticity
Isotonic contraction
Isokinetic contraction
Co-contraction
Surface electromyography
Rehabilitation robotics
a b s t r a c t
Although it has been reported that strengthening exercise in stroke patients is beneficial for their motor
recovery, there is little evidence about which exercise method is the better option. The purpose of this
study was to compare isotonic and isokinetic exercise by surface electromyography (EMG) analysis using
standardized methods.
Nine stroke patients performed three sets of isotonic elbow extensions at 30% of their maximal voluntary isometric torque followed by three sets of maximal isokinetic elbow extensions with standardization
of mean angular velocity and the total amount of work for each matched set in two strengthening modes.
All exercises were done by using 1-DoF planner robot to regulate exact resistive torque and speed. Surface electromyographic activity of eight muscles in the hemiplegic shoulder and elbow was recorded.
Normalized root mean square (RMS) values and co-contraction index (CCI) were used for the analysis.
The isokinetic mode was shown to activate the agonists of elbow extension more efficiently than the
isotonic mode (normalized RMS for pooled triceps: 96.0 ± 17.0 (2nd), 87.8 ± 14.4 (3rd) in isokinetic,
80.9 ± 11.0 (2nd), 81.6 ± 12.4 (3rd) in isotonic contraction, F[1, 8] = 11.168; P = 0.010) without increasing
the co-contraction of muscle pairs, implicating spasticity or synergy.
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
Upper limb muscle weakness is a common impairment
(Go et al., 2013) after stroke and has been known to be associated
with functional ability (Harris and Eng, 2007). Strength training
has been suggested as one of the therapies to restore strength in
the paretic arm of stroke patients in several clinical trials (Hammami et al., 2012; Harris and Eng, 2010). For this purpose, many
strength training strategies have been suggested to reduce the disability after stroke (Pinter and Brainin, 2012), and some of them
use robotic device for their rehabilitation study (Jezernik et al.,
2003; MinKi et al., 2011).
For strength training, variable contraction type of exercises can
be applied; isotonic and isokinetic modes are the most commonly
used in clinical settings. Some studies investigated the effect of isotonic strength training (Gelber et al., 1995; Logigian et al., 1983)
and other studies used the isokinetic mode as an intervention
⇑ Corresponding author. Tel.: +82 2 880 1703.
1
E-mail address: [email protected] (K. Cho).
These authors contributed equally to this work.
1050-6411/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jelekin.2013.10.002
(Chang et al., 2007), which proved beneficial effects on upper limb
strength and function in stroke patients. However, there have been
no studies to compare the differential effect between the two
modes in stroke patients.
Although studies using surface electromyography (EMG) reported some differences in EMG parameters between the isotonic
and isokinetic modes in healthy individuals (Purkayastha et al.,
2006; Remaud et al., 2009), these results cannot be extrapolated
to patients with stroke directly because this group of individuals
may display spasticity and abnormal synergy during the strength
training. It has been considered that strength training of the hemiplegic arm can aggravate the tone, abnormal synergies and pain in
the view of classical rehabilitation (Bobath, 2000). Although some
previous studies reported that strength training did not increase
the tone in stroke patients, a definite conclusion for patients with
spasticity is difficult because most studies include stroke patients
with only a low level of tone (Harris and Eng, 2010). EMG complementary analysis in spasticity and synergisms of stroke patients is
suitable for the isotonic and isokinetic contraction exercises.
In addition, to compare the isotonic and the isokinetic exercises,
it has been suggested that standardization method is needed. How-
12
M. Sin et al. / Journal of Electromyography and Kinesiology 24 (2014) 11–17
ever, standardized comparison of the two modes was only performed in the lower extremities of healthy people (Remaud et al.,
2009). Furthermore, the isotonic exercise in the previous study cannot simulate the real isotonic exercise in usual clinical settings, because it was designed to increase the velocity directly proportional
to the force measured by the dynamometer, while the real isotonic
mode should give constant load throughout overall range of motion
(ROM) (Kovaleski et al., 1995; Remaud et al., 2009). Therefore, in this
study, 1 degree of freedom (DoF) planner robot was used to provide
actual isotonic and isokinetic exercises for our experiments.
Accurate measurement and accurate motion generation of the
robotic device allows regulation of exact desired force or motion,
designation of certain amount of exercise load and intention driven
exercise, which facilitates advanced rehabilitation protocols.
The purpose of this study was to compare the efficiency of isotonic and isokinetic exercises in stroke patients. Surface EMG was
analyzed while the patients performed standardized exercise. In
addition to agonist and antagonist activities, co-contraction index
(CCI) for selected muscle pairs in the shoulder and elbow was included for EMG analysis to compare the spasticity and synergy between the two modes.
2. Methods
2.1. Subjects
Inclusion criteria were as follows: (1) unilateral hemiparesis in
the upper extremity caused by unilateral first-ever stroke, (2)
20 years or older, (3) elbow joint spasticity in the range of one plus
to three in the modified Ashworth scale in the hemiparetic arm, (4)
voluntary elbow extension strength of three or above in the
hemiparetic arm as measured by the manual muscle test proposed
by Medical Research Council, (5) no previous disease affecting the
function of the hemiparetic arm, except for stroke itself (6) free of
cognitive, language, visuospatial or attention deficits that would
prevent subjects from following the experimental procedures,
and (7) free of medical conditions which would cause hemodynamic instability. Patients were recruited from June 2012 to September 2012. The subjects were inpatients or outpatients with
stroke in one Department of Rehabilitation Medicine of a tertiary
hospital. One physiatrist is affiliated with the Department of Rehabilitation Medicine screened the patients, and total of nine patients
who met the above-mentioned criteria and provided their written
consent were enrolled in this study. This research was approved by
the local institutional review board and was conducted in accordance with the regulatory standards of Good Clinical Practice and
the Declaration of Helsinki (World Medical Association Declaration
of Helsinki: Ethical Principles for Medical Research Involving Human Subjects, 2008).
2.2. System and devices
For the experiment, the customized experimental platform has
been built. The platform consists of 5 parts: exoskeleton part, control unit, measurement unit, base frame and chair. The detailed
composition of the platform is in a Supplementary file.
2.3. Experimental design
The experiment was designed to compare isotonic exercise and
isokinetic exercise during elbow extension. The target motion, planar elbow extension at shoulder height, was decided to remove the
effect of gravity and minimize the effect of action of shoulder
external rotator and abductor (Hu et al., 2007). The experimental
protocol suggested by Remaud was used with modifications to
compare the isotonic and the isokinetic exercise with standardiza-
tion of mean angular velocity and total amount of work (Remaud
et al., 2009). The experiment was divided into three sessions: setup, familiarization and actual test.
In the setup session, Brunnstrom and Fugl-Meyer Assessment
(FMA) scale in the affected arm were checked. After that, surface
electrodes (1.8 1.2 mm Ag–AgCl, Bioprotech Inc., Wonju, Korea)
were attached on the following muscles in the patient’s hemiplegic
upper extremity based on the ‘‘Surface EMG for Non-Invasive
Assessment of Muscles (SENIAM)’’ (Hermens et al., 2000): anterior,
middle and posterior deltoid, biceps short and long head, triceps
long and lateral head and brachioradialis. One ground electrode
was attached at the backside of the neck (C7). The EMG data were
corrected for DC bias removal, band-pass filtered from 10 to 450 Hz
and sampled at 1024 Hz.
After EMG electrodes were attached, the maximum voluntary
isometric contraction (MVIC) of each muscle was performed to
measure maximum EMG value. First, MVIC of shoulder and forearm muscles (anterior, middle and posterior deltoid and brachioradialis) was performed (Hurst et al., 2012). In the anatomical
position of the hemiparetic upper extremity, three 5-s MVIC were
performed while an experienced physiatrist blocked the shoulder
from moving. In detail, the physiatrist stabilized the shoulder with
one hand and the upper arm with the other hand while the patient
was asked to exert maximum voluntary force in three directions;
flexion, extension and abduction. 30-s rest period between trial
and 2-min rest period between each action were given. EMG activities were measured during MVIC of each muscle.
Before MVIC of upper arm muscles (biceps short and long head,
triceps long and lateral head), subjects were set to experimental
posture. Trunk of subject was restricted by shoulder and abdominal straps that attached to the chair and the height of the chair
was adjusted to let the patient’s shoulder and the robot lie in the
same horizontal plane. The shoulder of the subject was abducted
90°, forearm was in the neutral position and elbow was positioned
at 90° in the horizontal plane. The rotation axis of the robot was
aligned to the anatomical axis of the elbow as shown in Fig. 1. Forearm was fastened to the manipulandum using straps. Elbow angle
of 90° (with 180° for full extension) was selected because it is
known that the maximum torques of the elbow can be obtained
around this angle (Koo et al., 2003).
Following the posture setting, MVIC of the upper arm muscles
was assessed (Hu et al., 2007). Three sets of 5-s elbow extension
and flexion were performed with the same rest periods as the
shoulder and forearm MVIC assessment. In addition to EMG activities, torques were measured during MVIC of the upper arm muscle
(see Fig. 2).
In the familiarization session, subjects were asked to try both
isotonic and isokinetic exercise. Resistive torque during isotonic elbow extension was set at 30% of the maximal elbow extension torque of each subjects from the setup session. Several elbow isotonic
extensions from 60° to 110° were tried. In cases the subjects cannot extend the elbow from 60° to 110° during the isotonic mode,
the range was adjusted (Table 1). After deciding the angle, several
isotonic elbow extensions were performed. Next, several isokinetic
elbow extensions with the same range and mean velocity of the
previous isotonic session were performed.
Before actual test session, the subjects took a 10-min rest period. Resistive torque and range of exercise were set using the same
parameters determined in the familiarization session. Firstly, in the
actual test session, the subjects performed three sets of six isotonic
elbow extensions with a 2-min rest period between the sets. During the isotonic elbow extension, angle, angular velocity and external torque were measured by the robot and mean angular velocity
and total work done by the subjects were calculated for each exercise sets using external torque and angular velocity. After 10 min of
rest period, the subjects performed standardized three sets of
M. Sin et al. / Journal of Electromyography and Kinesiology 24 (2014) 11–17
13
influence of spasticity and synergism during both the elbow effort
exercises.
Immediately after completing the three sets of each isotonic
and isokinetic sessions, the 15-point Borg scale was used to evaluate the degree of subjective muscle fatigue in the affected upper
extremity (Oberg et al., 1994). A numeric rating scale (NRS) for
pain ranging from 0 to 10 was used in the affected upper extremity
before and after each isotonic and isokinetic session (Breivik et al.,
2008).
(a)
2.4. Data analysis (EMG data processing and statistical analysis)
To analyze the difference between two exercise modes regarding standardization, efficiency, activation level, synergy pattern
and muscle fatigue, EMG data processing and statistical analysis
using ANOVA were performed. The detailed EMG data processing
and statistical analysis are in a Supplementary file.
(b)
3. Results
3.1. Controlled parameters
Fig. 1. The experimental setup.
A two-way repeated measures ANOVA for the total external
amount of work revealed no significant effects of the modes
(F[1, 8] = 1.613; P = 0.240), set orders (F[1,8] = 0.401; P = 0.544) and
interactions between the modes and set orders (F[1,8] = 0.572;
P = 0.471).
A two-way repeated measures ANOVA for the mean angular
velocity revealed no significant effects of the modes (F[1,8] = 4.662;
P = 0.063), set orders (F[1, 8] = 0.021; P = 0.888) and interactions
between the modes and set orders (F[1, 8] = 0.146; P = 0.712).
3.2. Number of repetitions
ANOVA results showed that there is a significant differences on
the modes of contraction for repetition number variable. The number of repetitions was smaller for isotonic (6 ± 0, control variable)
as compared to isokinetic (4.60 ± 1.60 and 4.67 ± 1.32 in the second
and the third isokinetic sets, respectively) mode (F[1, 8] = 9.776;
P = 0.014). However, the effects of series (F[1, 8] = 0.024;
P = 0.880) and interaction between the modes and set orders
(F[1, 8] = 0.024; P = 0.880) were not significant.
3.3. Surface EMG activity levels according to the mode of exercise
Fig. 2. Diagram of the control system.
isokinetic elbow extension with the same rest period with isotonic
elbow extensions. During the familiarization session, mean velocity of the isotonic extension session of each set was used as moving
velocity of each isokinetic extension set. Total work done by the
subjects was also calculated using the same method with the isotonic extension and each set of the isokinetic extension continued
until the work reached the value of the matched isotonic extension
set. No resistive torque was applied during reciprocal elbow flexion
for both exercise modes. The detailed methods of standardization
of isokinetic sets to the corresponding isotonic sets were described
elsewhere (Remaud et al., 2009). EMG of the 8 muscles was measured during the only elbow extension contraction to verify the
The two-way repeated measures ANOVA revealed significant
differences between the modes for the RMS values of the long head
of the triceps (F[1, 8] = 7.950; P = 0.023), lateral head of the triceps
(F[1, 8] = 5.798; P = 0.043) and the pooled values from both heads
of the triceps (F[1, 8] = 11.168; P = 0.010); RMS values for the isokinetic mode were higher (Fig. 3A). There were no significant effects
of the modes and set orders on the RMS values of the short and
long head of the biceps, brachioradialis and pooled values from
both heads of the biceps and brachioradialis (Fig. 3B).
3.4. Analysis of CCI for muscle pairs
Two way repeated measures ANOVA for the CCI from all possible muscle pairs of 8 muscles (total 28 muscle pairs) revealed significant effects of the modes of exercise in only two muscle pairs,
the lateral head of the triceps/long head of the biceps
(F[1, 8] = 5.988; P = 0.040) and the posterior deltoid/lateral head
of the triceps (F[1, 8] = 15.152; P = 0.005) (Fig. 4). There were no
significant effects of the set orders and interaction between the
modes and set orders on the CCI from these two muscle pairs.
14
M. Sin et al. / Journal of Electromyography and Kinesiology 24 (2014) 11–17
Table 1
Baseline characteristics of the subjects.
Patient
no.
1
2
3
4
5
6
7
8
9
Age
(years)
29
80
89
47
46
60
63
44
47
Gender
F
M
M
F
M
M
F
M
M
Stroke onset to
test (days)
1768
60
92
1093
1205
1985
80
17
15
Stroke type
Ischemic
Ischemic
Ischemic
Hemorrhagic
Ischemic
Ischemic
Hemorrhagic
Ischemic
Hemorrhagic
Hemiplegic
side
Left
Left
Right
Right
Right
Left
Right
Right
Right
MAS in hemiplegic
arm
Elbow
extensor
Elbow
flexor
3
1+
2
3
3
3
1+
1+
2
3
1+
2
3
3
3
1+
1+
2
Brunnstromstage in arm
3
3
4
3
4
3
5
4
5
FMA
in U/E
34
34
47
23
38
30
57
49
60
Muscle strength in
elbow
Elbow
extensor
Elbow
flexor
4
3
3
3
3
4
4
3
3
4
3
3
3
3
4
4
3
3
Tested elbow
range of motion
(°)
60–110
60–110
50–80
60–100
60–100
60–110
60–110
60–90
60–110
MAS: Modified Ashworth Scale, FMA: Fugl-Meyer assessment, U/E: Upper extremity.
3.5. Muscle fatigues and upper extremity pain according to the mode
of exercise
The Borg scale after completing each mode of exercise was
11.89 ± 2.67 in the isotonic mode and 11.56 ± 2.30 in the isokinetic
mode, and no significant differences between the modes were
found (P = 0.863).
All subjects reported a zero score of NRS for pain in the affected
upper extremity just before each mode of exercise. No subject
complained of pain after the isokinetic mode of exercise. Only
one subject (subject number 8) reported pain which scored two
in NRS after completing the isotonic mode of exercise, but it disappeared during the 10-min resting period.
4. Discussion
This is the first study to investigate the differences between the
isotonic and isokinetic modes during concentric elbow extension
using surface EMG and robotic device in stroke patients with
spasticity.
The standardized methods for angular velocity and total work
realized 30 revealed that the isokinetic mode showed a significantly less number of total repetitions and greater agonist activity
during elbow extension than those in the isotonic mode. In addition, CCI in two muscle pairs were significantly lesser in the isokinetic mode compared to those in the isotonic mode (Fig. 4).
The results (greater agonist activity and lesser number of repetitions in isokinetic mode compared to those in isotonic mode)
suggest the isokinetic mode makes efficient dynamic muscle action
and contributes more to increase motor unit recruitment without
applying unnecessary heavy loads (Purkayastha et al., 2006). Also
these results are compatible with conventional theoretical concept.
Theoretically, it has been thought that isokinetic contraction works
better than isotonic contraction because the isokinetic mode can
induce maximal loading to muscles through the overall ROM in
contrast to the isotonic mode which can load only at the weakest
mechanical points during motion (Kovaleski et al., 1995; Schmitz
and Westwood, 2001; Smith and Melton, 1981).
However, the previous two studies revealed that agonist
activity was greater in isotonic mode than that in isokinetic mode
(Purkayastha et al., 2006; Schmitz and Westwood, 2001). Remaud
et al. (2009)) argued that these two studies did not standardize the
exercise modes. However, after using the standardization method
which is similar with that of our study (Remaud et al., 2005), they
also reported that agonist muscles activity was greater in the isotonic mode than that in isokinetic mode and the number of repetitions was same in the two modes.
These different results may be caused by different methods
used to realize the isotonic mode between the study of Remaud
and ours. According to Remaud et al. (2009), the isotonic exercise
mode described in the study of Remaud is explained as it has
threshold resistance and its moving velocity is directly proportional to surplus force. This can be expressed by following
equation:
T Human T Threshold
8
>
< K x ðDT > 0Þ
ðDT 6 0Þ
¼ DT ¼ 0
>
:
ð1Þ
where THuman is a force that a subject generates, TThreshold is a
threshold resistance force, K is a proportional constant and x is
angular velocity.
Although this isotonic mode simulates actual isotonic exercise
by producing various velocities along the force, it is hard to say
the isotonic mode is an actual isotonic exercise. The ideal isotonic
exercise mode should produce constant resistive force on the muscle regardless of motion. However, the resistive force of the isotonic mode used in the study of Remaud is a function of the
angular velocity (Eq. (2)) which means that it is hard to impose
constant resistive force to subjects and the experiment conditions
cannot be regulated well:
T Resistive ¼ T Human ¼ T Threshold þ K x ðDT > 0Þ
ð2Þ
To solve this problem, isotonic exercise mode of our study was
implemented with torque control, inertia compensation and a robot with high backdrivability to make the resistive force constant.
It is a usual phenomenon that the co-contraction of antagonist
muscles is exaggerated during recovery after stroke, and this may
be associated with increased velocity-dependent stretch reflexes,
called spasticity (Frisoli et al., 2012; Marciniak, 2011). Because
stretch reflexes have been known to be proportionally increased
to the level of muscular preactivation (Lin and Sabbahi, 1999;
Smeets and Erkelens, 1991), strengthening exercises would increase the stretch reflex. From the concept of classical neurodevelopmental treatment, an increase in stretch reflexes or synergy has
been thought to be something to be avoided (Bobath, 2000).
The experiment results show that co-contraction during elbow
extensions is less severe in the isokinetic mode than that in the isotonic mode (Fig. 4). The degree of stretch reflex is known to be
associated with stretch velocities and joint angle (McPherson
et al., 2011). However, the velocities and joint angle between the
isotonic and isokinetic modes are controlled, these two parameters
cannot explain the differences in the CCI. One possible explanation
is that the presence of acceleration during the isotonic mode,
which is minimal during isokinetic motion, due to faster and more
M. Sin et al. / Journal of Electromyography and Kinesiology 24 (2014) 11–17
15
Fig. 3. RMS values for elbow extensors (A) and elbow flexors (B) according to the modes of exercise (isotonic vs. isokinetic) and set orders (second and third). Values are
mean ± SEM. P < 0.05 for the differences between the mode of exercise with two-way repeated measures ANOVA.
irregular acceleration, resulting in more increased stretch reflexes
(Berardelli et al., 1983; Burke et al., 1978; Perot et al., 1992).
Cortical damage after stroke induces cortical overlap of shoulder and elbow joint representations which results in flexion or
extension synergies during voluntary motion (Yao et al., 2009). Because preloading increases the synergies, strengthening exercises
are expected to increase the synergies during training (Miller and
Dewald, 2012). However, the CCI results show that co-contraction
of posterior deltoid and lateral head of the triceps, the component
of extension synergy in stroke patients (Brunnstrom, 1970), is less
severe in isokinetic elbow extension. Because flexion synergy is related with stretch reflexes during elbow extension (McPherson
et al., 2011), the larger stretch reflexes during isotonic elbow
extension were expected due to faster and more irregular acceleration can lead to more increased flexion synergy. However, there
were no differences in CCI of the muscle pairs associated with flexion synergy between the two modes. The reason why isotonic elbow extension induced more extension synergies than the
isokinetic mode is difficult to explain at this point in time.
The initial target of elbow extension angle was from 60° to 110°,
but the range had to be adjusted in several patients because they
could not perform isotonic elbow extension fully within this range
although they could do in the isokinetic. Patients cannot extend
the elbow further if they cannot generate extension torque more
than the preset load in the isotonic mode, which is different from
the isokinetic mode. The terminal range had to be reduced in patients who needed adjustment of the ROM in isotonic elbow extension (Table 1). ROM limitation may be due to reduced torque
generation and stability in the terminal range of extension
(Remaud et al., 2009; Trumbower et al., 2009).
This study has several limitations to be considered.
First, healthy controls were not included. Therefore, it is hard to
conclude that the results of this study are examples of the unique
phenomena that occur in stroke patients. However, because there
has been concerned about increased abnormal spasticity and synergy during strengthening exercises in the field of stroke rehabilitation, and no studies comparing the different modes of
strengthening exercises in stroke patients have been published
up to date, this study can give implications for selecting appropriate modes of strengthening and for designing a longitudinal study
to investigate the effects of different modes of strengthening in
stroke rehabilitation.
Second, the order of two strengthening exercise was not randomized. Because the isotonic exercise always preceded the
isokinetic exercise to standardize two exercise mode, this can
affect the results of our study due to fatigue. However, resting
intervals between the two exercise modes are sufficient in
reference to the previous study (Remaud et al., 2009), the Borg
scale which indicates muscle fatigue (Oberg et al., 1994) was not
different between the two modes after completing each exercise,
16
M. Sin et al. / Journal of Electromyography and Kinesiology 24 (2014) 11–17
Fig. 4. Normalized co-contraction index from the pairs of the lateral head of the triceps/long head of biceps and the posterior deltoid/lateral head of the triceps according to
the modes of exercise (isotonic vs. isokinetic) and set orders (second and third).
and the only one patient complained of mild pain (score of 2 in
NRS) after the isotonic exercise, but it was transient.
Finally, because the sample size is small, it is possible that the
power to detect the small differences of CCI between the two
modes is not sufficient. In addition, it is inconvenienced by not performing subgroup analysis according to the different degrees of
spasticity.
5. Conclusion
In this first study comparing isotonic and isokinetic elbow
extensions in stroke patients with spasticity, the isokinetic mode
was shown to be able to activate the agonists of elbow extension
more efficiently than the isotonic mode without increasing the
co-contraction of muscle pairs which can implicate spasticity or
synergy. A further study with a longitudinal design is required to
compare the effects of isotonic and isokinetic strengthening training on functional recovery including strength gain, spasticity and
synergy in stroke patients with spasticity.
Acknowledgments
This study was supported by the Seoul National University
Bundang Hospital Research Fund (02-2012-052), the Technology
Innovation Program (10036492) funded by the MKE/KEIT, Korea
and National Research Foundation of Korea (NRF) Grant funded
by the Korean Government (A100249).
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the
online
version,
at
http://dx.doi.org/10.1016/
j.jelekin.2013.10.002.
References
Berardelli A, Sabra AF, Hallett M, Berenberg W, Simon SR. Stretch reflexes of triceps
surae in patients with upper motor neuron syndromes. J Neurol, Neurosurg,
Psychiatr 1983;46:54–60.
Bobath B. Adult hemiplegia: evaluation and treatment. Oxford, UK: ButterworthHeinemann; 2000.
Breivik H, Borchgrevink PC, Allen SM, Rosseland LA, Romundstad L, Hals EK, et al.
Assessment of pain. Br J Anaesth 2008;101:17–24.
Brunnstrom S. Movement therapy in hemiplegia. A neurophysiological
approach. New York: Harper & Row; 1970.
Burke D, Hagbarth KE, Lofstedt L. Muscle spindle activity in man during shortening
and lengthening contractions. J Physiol 1978;277:131–42.
Chang JJ, Tung WL, Wu WL, Huang MH, Su FC. Effects of robot-aided bilateral forceinduced isokinetic arm training combined with conventional rehabilitation on
arm motor function in patients with chronic stroke. Arch Phys Med Rehabil
2007;88:1332–8.
Frisoli A, Procopio C, Chisari C, Creatini I, Bonfiglio L, Bergamasco M, et al. Positive
effects of robotic exoskeleton training of upper limb reaching movements after
stroke. J Neuroeng Rehabil 2012;9:36.
Gelber DA, Josefczyk B, Herrman D, Good DC, Verhulst SJ. Comparison of two
therapy approaches in the rehabilitation of the pure motor hemiparetic stroke
patient. Neurorehabil Neural Repair 1995;9:191–6.
Go AS, Mozaffarian D, Roger VL, Benjamin EJ, Berry JD, Borden WB, et al. Heart
disease and stroke statistics–2013 update: a report from the American Heart
Association. Circulation 2013;127:e6–e245.
Hammami N, Coroian FO, Julia M, Amri M, Mottet D, Herisson C, et al. Isokinetic
muscle strengthening after acquired cerebral damage: a literature review. Ann
Phys Rehabil Med 2012;55:279–91.
Harris JE, Eng JJ. Paretic upper-limb strength best explains arm activity in people
with stroke. Phys Ther 2007;87:88–97.
Harris JE, Eng JJ. Strength training improves upper-limb function in individuals with
stroke. Stroke; J Cereb Circulation 2010;41:136–40.
Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations
for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol
2000;10:361–74.
Hu X, Tong KY, Song R, Tsang VS, Leung PO, Li L. Variation of muscle coactivation
patterns in chronic stroke during robot-assisted elbow training. Arch Phys Med
Rehabil 2007;88:1022–9.
Hurst HT, Swarén M, Hébert-Losier K, Ericsson F, Sinclair J, Atkins S, et al. Influence
of course type on upper body muscle activity in elite cross-country and
downhill mountain bikers during off road downhill cycling. J Sci Cycl
2012;1:2–9.
Jezernik S, Colombo G, Keller T, Frueh H, Morari M. Robotic orthosis lokomat: a
rehabilitation and research tool. Neuromodulation: J Int Neuromodulation Soc
2003;6:108–15.
Koo TK, Mak AF, Hung L, Dewald JP. Joint position dependence of weakness during
maximum isometric voluntary contractions in subjects with hemiparesis. Arch
Phys Med Rehabil 2003;84:1380–6.
M. Sin et al. / Journal of Electromyography and Kinesiology 24 (2014) 11–17
Kovaleski JE, Heitman RH, Trundle TL, Gilley WF. Isotonic preload versus isokinetic
knee extension resistance training. Med Sci Sports Exercise 1995;27:895–9.
Lin FM, Sabbahi M. Correlation of spasticity with hyperactive stretch reflexes and
motor dysfunction in hemiplegia. Arch Phys Med Rehabil 1999;80:526–30.
Logigian MK, Samuels MA, Falconer J, Zagar R. Clinical exercise trial for stroke
patients. Arch Phys Med Rehabil 1983;64:364–7.
Marciniak C. Poststroke hypertonicity: upper limb assessment and treatment. Top
Stroke Rehabil 2011;18:179–94.
McPherson JG, Stienen AH, Drogos JM, Dewald JP. The relationship between the
flexion synergy and stretch reflexes in individuals with chronic hemiparetic
stroke. IEEE Int Conf Rehabil Robotics: [Proceedings] 2011;2011:5975516.
Miller LC, Dewald JP. Involuntary paretic wrist/finger flexion forces and EMG
increase with shoulder abduction load in individuals with chronic stroke. Clin
Neurophysiol: Off J Int Fed Clin Neurophysiol 2012;123:1216–25.
MinKi S, DaeGeun P, Kyu-Jin C. Comparison and evaluation of robotic strength
rehabilitation algorithms: isokinetic, isotonic and shared control method. In:
Advanced mechatronic systems (ICAMechS), 2011 international conference on
2011. p. 438–41.
Oberg T, Sandsjo L, Kadefors R. Subjective and objective evaluation of shoulder
muscle fatigue. Ergonomics 1994;37:1323–33.
Perot C, Mora I, Goubel F. Muscle stretch used as conditioning stimulus for assessing
a reciprocal inhibition. Electromyogr Clin Neurophysiol 1992;32:331–9.
Pinter MM, Brainin M. Rehabilitation after stroke in older people. Maturitas
2012;71:104–8.
Purkayastha S, Cramer JT, Trowbridge CA, Fincher AL, Marek SM. Surface
electromyographic amplitude-to-work ratios during isokinetic and isotonic
muscle actions. J Athl Train 2006;41:314–20.
Remaud A, Cornu C, Guevel A. A methodologic approach for the comparison
between dynamic contractions: influences on the neuromuscular system. J Athl
Train 2005;40:281–7.
Remaud A, Cornu C, Guével A. Agonist muscle activity and antagonist muscle coactivity levels during standardized isotonic and isokinetic knee extensions. J
Electromyogr Kinesiol 2009;19:449–58.
Schmitz RJ, Westwood KC. Knee extensor electromyographic activity-to-work ratio
is greater with isotonic than isokinetic contractions. J Athl Train
2001;36:384–7.
Smeets JB, Erkelens CJ. Dependence of autogenic and heterogenic stretch reflexes on
pre-load activity in the human arm. J Physiol 1991;440:455–65.
Smith MJ, Melton P. Isokinetic versus isotonic variable-resistance training. Am J
Sports Med 1981;9:275–9.
Trumbower RD, Krutky MA, Yang BS, Perreault EJ. Use of self-selected postures to
regulate multi-joint stiffness during unconstrained tasks. PloS One
2009;4:e5411.
Yao J, Chen A, Carmona C, Dewald JP. Cortical overlap of joint representations
contributes to the loss of independent joint control following stroke.
NeuroImage 2009;45:490–9.
Minki Sin received a B.S degree in mechanical and
aerospace engineering from Seoul National University,
Seoul, Korea, in 2010. He is currently a Ph.D. candidate
in mechanical engineering at Biorobotics Laboratory in
Seoul National University, Seoul, Korea. His current
research interests include rehabilitation robotics and
robot control.
Won-Seok Kim received B.S degree from Seoul National
University College of Medicine, Seoul, Korea in 2004,
and M.S degree from Graduate School of Public Health,
Seoul National University, Seoul, Korea in 2011. He is
now working as a physiatrist at the Department of
Rehabilitation Medicine in Seoul National University
Bundang Hospital, Seongnam-si, Gyeonggi-do Korea. He
is also a Ph.D candidate for Rehabilitation Medicine in
Seoul National University College of Medicine, Seoul,
Korea. He is interested in the researches for the neurorehabilitaion, geriatric rehabilitation and epidemiology
related with the disability issues.
17
Daegeun Park received a B.S. degree in mechanical
engineering from Pohang University of Science and
Technology (Postech), Pohang, Republic of Korea, in
2010. He is currently working toward a Ph.D. degree at
the Biorobotics Laboratory in Seoul National University,
Seoul, Korea. His current research interests include
rehabilitation and assistive robot design with soft
robotics.
Yu-Sun Min was graduated from Seoul National University College of Medicine. She achieved a M.D degree
in 2006. She was certified as a member of rehabilitation
specialist in 2011. She had worked for Seoul National
University Bundang Hospital as a research fellow with
interest in robotic rehabilitation. Currently she is a
clinical assistant professor in the Department of Rehabilitation Medicine in Kyoungpook National University
Medical Center, Daegu, Korea.
Woo Jin Kim graduated Inje University School of Medicine and received license of medical doctor in 2006. She
finished her Internship and Residency at Samsung
Changwon Hospital, Sungkyunkwan University School
of Medicine in 2011. In 2012, she acquired her Masters
degree in physical medicine and rehabilitation from
Kyungsang University School of Medicine. After finishing her fellowship at Bundang Seoul National Hospital
from 2011 to 2012, she is now an assistant professor in
Department of Physical Medicine and Rehabilitation at
Haeundae Paik Hospital, Inje University School of
Medicine. Currently she is working toward a Ph.D.
degree at physical medicine and rehabilitation department in Busan National
University of Medicine. Her current research interest include stroke rehabilitation,
brain plasticity and mapping associated with rehabilitation and rTMS.
Kyujin Cho received B.S and M.S. degrees from Seoul
National University, Seoul, Korea in 1998 and 2000,
respectively, and a Ph.D. degree in mechanical engineering from Massachusetts Institute of Technology in
2007. He was a post-doctoral fellow at Harvard Microrobotics Laboratory until 2008. At present, he is an
associate professor of Mechanical and Aerospace Engineering and the director of Biorobotics Laboratory at
Seoul National University. His research interests include
biologically inspired robotics, robotic systems using
smart actuators, novel mechanisms using smart structures, and rehabilitation and assistive robotics.
Nam-Jong Paik received B.S, M.S and Ph.D degrees from
Seoul National University College of Medicine, Seoul,
Korea in 1990, 1995 and 2000, respectively. At present,
he is a professor of Department of Rehabilitation Medicine, Seoul National University of College of Medicine
and is working at Seoul National University Bundang
Hospital, Seongnam-si, Gyeonggi-do Korea. His research
interests are neurorehabilitation and geriatric rehabilitation including stroke rehabilitation, new therapeutic
interventions such as neuromodulation, robotic rehabilitation, tele-rehabilitation and virtual rehabilitation.