Page 1 Page 2 changes in frequency pr。perties 。f eーectr。encepha

Shinshu University Institutional Repository SOAR-IR
Changes in frequency properties of electroencephalograph and
electromyography in motor learning process
木村, 貞治; 藤原, 孝之; 西村, 尚志; 大平, 雅美; GOH, Ah
Cheng; 神子嶋, 誠; 三好, 圭; 楊箸, 隆哉; 杉田, 勇; 許, 明峻
Issue Date
紀要 23: 25-36(1998)
Vo123, 1997
Changes in frequency properties of electroencephalograph
and electromyography in motor learning process
木村貞治1),藤原孝之1),西村尚志1),大平雅美1), GOHAhChengl)
神子嶋誠1),三好 圭1),楊箸隆哉2),杉田 勇3),許 明唆4)
Changes infrequency properties of electroencephalograph
and eleetromyography in motor learmimg process
The purpose of this study was to investigate the modifications of motor performance and
changes in electrophysiological properties such as electroencephalogram (EEC) and electr0-
myography (EMG). The subjects consisted of six healthy males aged 18 to 24 years (mean
- 21.8 years). The task of motor learning was maximum reciprocal tapping of theNandVfinger in the non-dominant hand. Practices were executed 5 days per week for 4 weeks. Parame-
ters of analysis were rate of tapping, median power frequency properties in EEC and EMG
during 10 seconds.
The results showed significant increases in the mean rate of tapping and median power
frequencies in EEC. Correlation between rate of tapping and median power frequency of EEC
on the motor cortex showed a gradual increase with each week. There was no significant cor-
relation between rate of tapping and median power frequency properties of EMG.
It is suggested that an increase of neural activities in the cortex may correlate with the
evolution of motor learnlng.
Key Words :
Motorlearning (運動学習), EEC (脳波), EMG (筋電図)
1 )信州大学医療技術短期大学部理学療法学科; KIMURA Teiji. FUJIWARA Takayuki. NISHIMURA Naoshi,
OHIRA Masayoshi, GOH Ah Cheng. KAMIKOJIMA Makoto. MIYOSHI Kei. Dept. of PhysicalTherapy. School
of Allied Medical Sciences, Shinshu Univ.
2 )信州大学医療技術短期大学部看護学科; YANAGIHASHI Ryuya, DepL of Nursing. School of Allied Medical
Sciences, Shinshu Univ.
3 )諏訪中央病院; SUGITA Isamu, DepL of Rehabilitation, Suwa Central Hospital
4 )順天郷大学校工科大学; FUR Myung Joon. Dept. of Electricalengineering, College of Electrical engineering.
University of Sooncllunhyamg
learnlng in physical therapy.
1. lntroduction
With regards to electrophysiological stud-
ln physical therapy, some therapeutic in-
ies related to motor control or motor learn-
puts such as verbal orders, demonstrations.
ing. Murthy and Fets8) indicated that the
manual assistance and mechanical assistance
band ratio of the β wave of electroencephalo一
are applied to the patients in order to im-
gram was observed during tasks requlrlng
prove the performance of motor control.
fine finger movements and focused attention
The series of processes concerned with
in monkey motor cortical cells. Karni etal.9)
relative permanent changes of motor per-
reported that expansion of activated regions
formance caused by these repetitive thera-
in primary motor cortex during motor skill
peutic Inputs is defined as motor learn-
learning observed by functional magnetic
resonance image (f-MRI) may be caused by
The conceptual models of motor learning
theory, such as closed loop theory3), schema
the organization of synapse network related
to the evolution of motor skill learnlng.
theory4), impulse variability theory5)and
Concerning the changes of cortical activity
knowledge of results (KR) 6)7) have been
when executing learned motor sequences.
previously discussed.
Lang et al.10) reported that in a learned se-
Experimental approaches based on the
quence of four movements (flex index finger,
above motor learning theories have been de-
extend hand. extend index finger,flex
veloped from the fields of psychology, physi-
hand) , large negative DC potentials were
cal education, neuroscience, bio-information
recorded in positions located above the me-
engineering and othersl).
sial fronto-central cortex (Cz) and sensori-
In previous experimental studies, analyses
motor hand areas of either hemisphere (C
of motor learning have been executed based
3 andC 4 ) in the beginningofeachperiod.
on the change of performance criteria such
Furthermore, DC potentials were absent in
as speed, accuracy ln movement, aS Well as
Cz and the end of the period of execution.
analyses based on the changes of electro-
On the other hand, an electromyographical
physiological properties such as EEC and
study conducted by Sadoyama et al. ll) ex-
amined the modification of auricle muscles
In physical therapy, both analyses of per-
using EMG to evaluate the effect of motor
formance and electrophysiologiCalproperties
learnlng quantitatively. Since the results in-
are important in order to prove the scientific
dicated that the acquisition process of auricle
evidence related to acquisition of motor skill
movement was effectively shown, as demon-
by therapeutic exercise.
strated by the gradual increase in mean am-
However there are only a few studies
plitude of the EMGs of auricle muscles, they
which attempt to explain the neuromuscular
suggested that EMG signal processing
changes which accompany skilled motor
method has wider application in the field of
Changes in frequency properties of electroencephalograph and electromyographyinmotor leaming process 27
neuromuscular rehabilitation and sports.
24 years (mean 21.8 years) who had no pre-
Vallbo et al.12) investigated the human
vious experiences about continuous skill mo-
muscle spindle response in a motor learning
tor learnlng Such as playlng the plan°. In-
task by using microneurographic technique,
formed consent to the experiment was taken
and they concluded that focusing attention
from all subjects. Subjects were all right
on the kinaesthetic input during imposed
movement was not associated with a consis-
3. Methods
tent increase of fusimotor drive.
Although the above previous studies re-
1 ) Task of Motor Learning
ported on the elecrophysiological evidences
The maximal reclprOCal finger tapplng Of
based on the analysis of EEC or EMG in mo-
Ⅳ and V finger in the non-dominant hand
tor control or motor learnlng process,
with maximum veloclty Was Selected as a
changes in the frequency properties of EEC
motor learnlng task. The subjects were in-
or EMG in motor learning processes have
structed to execute maximum reclprOCal fin-
not been discussed. Furthermore, correlation
ger tapplng in N and V finger in the non-
between evolution of motor performance and
dominant hand, placing their left palms on
changes in electrophysiologlCal properties
the desk and observing their own move-
have not been discussed.
Fets et al 13). reported that a cortical motor
The task was selected because it is not
neuron projects alpha motoneuron and it
used in daily activity, and improvements of
regulates the generation of muscle force.
motor skill by practice can be expected.
Consequently, lt appears that changes of
2 ) procedure
frequency propel・tiesinEEG or EMG may
During practice, subjects were seated on a
reflect modification in cortical neuron activ-
chair, with their back supported, and only
ity and myoelectrical activity by motor
their left hand was placed on the desk.
learnlng. These quantitative analyses of mo-
The task consisted of a block of50 trials
tor learning may be necessary in order to
per day, and one trial includes reclprOCal fin-
clarify the effects of therapeutic exercises in
ger tapping of leftⅣandVfinger for 10 sec-
onds with maximal velocity and a rest period
The purpose of this study was to investi-
of 10 seconds. The practice was executed for
gate the effect of motor learning based on
5 days (from Monday to Friday) per week
the modification of motor performance and
for 4 weeks.
changes in frequency properties in EEC and
The whole sequence of practice for a subject Was aS follows :
2. Subjects
1 block/day x5 daysx4 weeks- 1.000 trials.
Subjects were six healthy males aged 18 to
The practice of motor learning was limited
to only the practice time in a day to prevent
and touch sensors were adjusted to position
the bias induced by over training performed
the ⅣandV丘nger pulpwithin the box. The
by the subjects during their private time.
touch sensor signals were amplified by a DC
3 ) Measurement
The contents of measurement consisted of
the rate of reciprocal tapping in Ⅳ and V fin-
amplifier (Unipulse, Digital Indicator F
430), and were stored to a data recorder
(Shinko, RCD-728).
ger. electroencephalogram. and electr0-
Recordings of EEC signals during tapping
myographyinleft finger extensor and nexor
were made by an EEC machine (NEC Sanei,
during lOseconds. The trigger sound to
MK - 930705) with 8 silver-silver chloride
start and stop the tapping was generated by
cup electrodes. The electrodes were placed
a phono-stimulator (Sanei, 3 G 22) with lO-
overthe scalp atFz', Cl',Cz,C2', RHM,P
second intervals. Measurements were car-
l', Pz', P2'based on the movement related
ried out once per week. The data were cor-
cortical potentials (MRCP) using a conduc-
rected from 5 th to 7 th trials during a block
tive electrode gel (Fig. 1). The electr0-ocu-
in a day per week.
logram (BOG) was recorded from an elec-
Measurements were executed in a
shielded room to prevent noise.
trode placed 1.5cm below the right outer
canthus in order to check the blinking. The
The contact pressure of tapping by Ⅳ and
subjects were required to keep their eyes
V fingers was measured by using touch sen-
opened during each trial in order to prevent
sor equipment (ME, specially made) on the
artifacts caused by blinking.
desk. The touch sensor equipment was made
by an aluminum box and two touch sensors,
Two common reference electrodes were
placed on the earlobe.
EMG signals of left Extensor Digitorum
Communis and Flexor Digitorum Sublimis
during reciprocal finger tapping were de-
tected by a telemetric EMG machine (NEC
Sam-Ei. Multitelemeter 511) with surface
electrodes. Before placing the silver-silver
chloride disposal type electrode, the skin was
first prepared with an abrasive alcohol prep-
ping solution. and electrode impedance was
always below 2 kEL A pair of electrodes were
placed in the center of the muscle belly, parallel to the direction of muscle fibers.
Detected EMG signals were stored on a
Figt 1 Placementof Electrodes in EEG
RHM was focused in this investigation.
data recorder together with tapping signals
and EEC signals.
Changes in frequency properties of electroencephalograph and electromyography in motor learning process 29
by band-pass filter ranged from 0.5 Hz to 30
-EEG 占竿空-画亘]
Hz. EMG signals were filtered by band-pass
filter ranged from 10 Hz to 250 Hz14).
Frequency properties of EEC signals and
EMG signals were analyzed through Fast
欝≡ND (signalProcessor)CPU
Fourier Transform (FFT) with Hanning
window function, and median power freFig. 2 Block Diagram of Experiment
quency (MdPF) was calculated using BIMU-
4 ) Data Analysis
T A S-E.
The analog data of tapping signals, EEC
Data analyses of maximum tapplng, MdPF
slgnals and EMG signals were converted to
of EEC and EMG were performed using av-
digital signals by an A/D converter with 1
eraglng Value of data in5th to7th trial of
kHz sampling frequency from the data re-
subjects in each week.
Among the placement of scalp electrodes,
Signal processing was executed on a per-
RHM was focused in this study since RHM
sonal computer (NEC, PC-9821 Ⅹp) With a
lies close to right primary motor cortex for
software for processlng Of bi0-information
finger movement15).
signals (Kissei Comtec, BIMUTAS-E, ver.
5 ) Statistical Analysis
2.ll) (Fig.2,Fig.3).
To test 4 weeks practice effects on maxi-
Digital filters were used to reduce the
mum tapping, MdPF of EEC and EMG,
noise in each signals using BIMUTAS-E. The
Friedman test was performed by using SPSS
tapplng Slgnals were filtered by low pass fil-
6.1 for Windows (SPSS Inc. ) , because nor一
ter below 10 Hz. EEC signals were丘ltered
mal distribution was not observed in these
a/.J4r._■(.I:■ LT'
イ貭粭、ツ粐貽、ニDヲトツ荀箸薐粭「 、「躔ニ
粨 、ト「粫偵
2粨 粐粨
tp: 畔
-JJ---.⊥一一1 .--'T巾.
訂J∼ l
- 峯
言=-←「一←サー 兩.LLJ.Ah.._L.,.J_LL^.Lムー.∫...▲.▲一▲ALJlL_.■山..山J山._.山一__L..A.LAJ.▲L_LJ▲LdL^‥▲▲.一.
三二二宝二二藍\. 冤11r1.r-.lY¶'▼ーIT.I-TuPrT-,'TTーⅥ'''ーVIl--r■■ ー'Tlql'一.-IT.1rYTII‥-VY▼-l■▼
Fig. 3 Raw Data of Reciprocal FingerTapplng
Table 1 Changes of Motor Performance and
Electrophysiologlcal Properties
W 2W 3W 4W
RateofTapplng 40.1±16.1 51.9±15.7 56.8±17.6 62.8±16.3
MeanMdPFinEEG(ALL) 9.9± 2.5 ll.4± 1.9 ll.7± 1.9 12.6± 2.6
MeanMdPFinEEG(RHM) 9.7± 2.1 ll.6± 2.0 12.8± 2.9 12.8± 2.8
MeanMdPFinEMG(ExtL) 96.8± 5.9 101.1± 7.6 98.8±11.1 98.7± 9.9
MeanMdPFinEMG(Flex.) 96.9±16.3 91.5±14.0 85.8±18.0 90.0±18.8
I-mg0807060504030201 0 朗.0 0 0O O o
1W 2W 3W 4W
1 W 2W 3W 4W
Fig.4 Modification of Rate of Tapplng
data. To test for difference among the4-
Fig. 5 lndividual Difference in Rate ofTapplng
weeks, Dunn test was used as a post hoc test.
Although the mean rate of reciprocal tap-
Statistical significance was accepted at less
ping revealed an almost linear increase.indi-
than 0.05
vidual differences were observed in the
learning curve of individual subjects (Fig.
4. Results
1 ) Modi丘cation of Motor Perfわrmance
The mean number of times in reciprocal
tapping of ⅣandV finger were increased in
all subjects every week (Table 1, Fig. 4).
Statistical significance was observed
among weeks by Friedman test (p<0.01)
2 ) Changes in Median Power Frequency
Properties in EEC
Since subjects could keep their eyes open
during data acquisition. prevention of arti-
facts by blinking were confirmed by EOG.
Mean value of MdPF on all electrodes in
(Table 2). Dumm post hoc testing indicated
EEC reveals a tendency to increase every
significant difference between the 1 st week
week (Table 1, Fig. 6). Friedman test showed
and4thweek (p<0.01)
significant increase of MdPF on all elec-
Changes in kequency properties of electroencephalograph and clectromyography in motor learning process
HZ6・04mo8 2O
W 2W 3W 4W
1 W 2W 3W 4W
Fig.7 Changes of SpatlLal Distribution in EEG
Fig.6 Changesof Mean MdPF in EPG on ALL
on aH Electrode
E lectrodes
trodes in EEC (p<0.05) (Table2). Dunn
H.0 Z
n"i nHi Ji JHi
6 4 2 ∩) 8 6 4 2 0
post hoe testing indicated significant differ-
ence between the 1 st week and4th week
(p< 0.05)
In the change of MdPF among the place一
ment of electrodes, RHM revealed a steep increase of MdPF compared with other sites
(Table 1, Fig.7).
Friedman test showed signi丘cant increase
1 W 2W 3W 4W
of MdPF on RHM (p<0.05) (Table2. Fig.
Fig.8 Changes of Mean MdPF in EEG on
8). Dunn post hoe testing indicated signi丘-
cant difference between the 1 st week and 3
Table 2 ResultsofFriedman Test
r2 k
1) Rate of tapping
between conditions 17.00 4 6 3 0.01
subjects 16.00 6 4 5 0・001
2) MdPF in EEG (ATl)
between conditjons 9.80 4 6 3 0[05
1.29 6 4 5 0.05
subjects 1
3) MdPF in EEC (RHM)
L00 4 6 3 0.05
4.29 6 4 5 0.02
nificant increase every week.
0 0 ∩) 0 0 0 0 0 8 7 6 5 4 3 2 7 1 0
The motor task of present study was selected based on the comments that recipro-
cal tapping of Ⅳ and V丘nger in non-domi-
nant hand is difficult to trainforflute players. However, the motor task in present
W 2W 3W 4W
study does not have great benefit for the
subjects. Also. knowledge of results were not
Fig, 9 Changes of Coefficient of Corre一ation
between Rate of Tapplng and MdPF of
rdweek (p<0.05).
3 ) Changes in Median Power Frequency
Properties in EMG
Friedman test indicated no significant
informed to the subjects till completion of
their practice period.
Nevertheless, motor performance of the
maximum reciprocaltapping of N and V finger in non-dominant hand were improved by
practices during 4 weeks, as we expected.
change of MdPF in EMG for both extensor
This result indicates that new skilled mo-
and flexor during4weeks (P>0.05) (Ta-
tor task which has not been experienced pre-
ble. 1). MdPF of EMGinfinger extensor
viously has the possibility of modification in
muscles were higher than flexor muscles at
motor learnlng.
the2nd, 3 rdand4thweeks. Change of dis-
However, since we did not evaluate the
charge pattern of both EMG was not ob-
long term effect after completion of practice.
it seemed that山rther study should be con-
4 ) Correlation between motor performance
and electrophysiologlCal properties
With regards to the phases of motor learn-
Correlation coefficient between rate of tap-
ing, Fitts16) classified cognitive phase, asso-
ping and median power frequency proper-
ciative phase and autonomous phase. Since
ties on RHM in EEC were gradually in-
content of task in the present study was sim-
creased every week up to the 3 rd week
ple reciprocal finger tapping, it seemed that
the degree of participation of cognition may
(Fig. 9).
There were no slgnificant changes be-
be small. But the discrimination between as-
tween motor performance and EMG signals
sociative phase and autonomous phase in
or between extensor and flexor.
present study may be difficult because we
did not check the transfer trials.
5 l Discussion
The eflects of knowledge of results (KR)
ln the present study, it was demonstrated
should also be investigated since it is consid-
that motor performance in motor skill learn-
ered that import, timing and frequency of
ing was improved by daily practices, and
KR may affect the motor performance7) 17)
MdPF in EEC on all electrodes revealed sig-
Although Busk and Galbraith18) pointed
Changes in frequency properties of electroencephalograph and electromyography in motor learning process 33
out that skilled movements were not likely to
seemed to be necessary to analyze the
be reflected in the EEC, MdPF of EEC inall
change of frequency properties in EEC dur-
electrodes revealed significant increase, ac-
ing motor learnlng Process.
companied by the improvement of motor
As a spatial property of frequency on all
performance in present study. In particular,
electrodes in EEC, EEC on RHM revealed
RHM which lies over the prlmary motor cor-
linear increases of median power frequency
tex related to left finger movement revealed
in motor learning process. This tendency
on acute increase. Furthermore it was note-
may suggest the increase of neuronal activi-
worthy that the coefficient of correlation be-
ties in the cortex included RHM.
tween the rate of reclprOCal finger tapplng
Tanji and Mushiake20) reported that pri一
and frequency properties of EEC on R班M
mary motor cortex is mostly related to exe-
were increased during motor learnlng prOC-
cution of motor task, while neuronal actlVlty
in the supplementary motor area exhibits a
Gliner et al. 19) investigated changes in the
variety of complex relationship to many dif-
EEC power spectrum during repeated per-
ferent aspects of motor task. Thach21) re-
formances of the same task to determine
ported that there are close relationship be-
whether neurophysiologlCal correlates of at-
tween neuronal actlVlty ln prlmary motor
tention changed as a function of learnlng the
cortex and motor parameters such as JOint
task. His results indicated that mean alpha
position, force, Velocity and acceleration of
frequency increased slightly during the task
movement generated by muscle contraction.
while delta frequency decreased in electro-
From these evidence, it is considered that
encephalogram during perceptuaトmotor
neuronal actlVlty ln prlmary motor cortex
learning. He also reported that a decrease of
may correlate with evolution of rate of reclp-
duration in execution and decrease of error
rocal finger tapplng aS a motor learning task
in task were observed by motor learning.
in present study, because simple reciprocal
Since his report aimed to clarify the influ-
finger tapplng Was used as a motor task in
ence of attention in perceptual motor learn-
this study. This speculation may be sup-
ing based on the EEC, it is considered that di-
ported by increase of coefficient of correla-
rect comparison to the present study may be
tion between rate of reciprocal finger tap-
difiic ult.
ping and median power frequency of EEC on
Lang et al. 10) studied the changes of corti-
cal activity when executing learned motor
It is also suggested that correlation analy-
sequence, and he concluded that the execu-
sis between motor performance and cortical
tion of a learned motor sequence task cannot
activity seemed to be useful in quantitative
be associated with a particular size and pat-
evaluation of electrophysiological change in
tern of cortical activity. Since they did not
motor learnlng.
analyze the frequency property in EEC, it
In the present study, EMG of finger exten-
Sor revealed dominant activity ln late stage
trot theories and motor learnlng theories.
of motor learning process. However, there
Furthermore he indicated that incorrect
were no significant changes of MdPF in
model of treatment can lead to inappropriate
EMG on the finger muscles during motor
expectations of the results.
learnlng process.
From these valuable information, it is con-
Basmajian22) reported that properties of
sidered that appropriate treatment program
EMG activity in motor learning process is se-
for motor learning in physical therapy should
lective inhibition of unnecessary muscular
be constructed by integration of both the sci-
actlVlty rather than the activation of addi-
entific investigation and clinical data in mo-
tional motor units. He cited evidence that the
tor learning.
development of skill is accompanied by pro-
The findings of the present study are lim-
gressively more successful repression of un-
ited in that the sample was small and only
desired contractions and by a gradual in-
short term effects were studied. Further-
crease in the average duration and a de-
more, since we did not analyze the ratio in
crease in average frequency of potentials in
frequency band in EEC, it is considered that
the specific muscle under trainlng.
analysis of the detailed frequency component
Payton and Kelley23) investigated the
of EEC in motor learnlng lS necessary. Fur-
change of EMG activlty ln motor learnlng
ther research is necessary to analyze motor
process. They concluded that prlme muscles
learning process in actual patients with mo-
demonstrated significantly less totalelectri-
tor disorders.
cal activity during skilled movements than
during unskilled movements.
From these evidence, dominant change of
We would especially like to thank the col-
frequency properties in extensor muscles
lege students for their participation in this
may suggest inhibition of the activity of
study. We gratefully acknowledge the coop-
nexor muscles. However. since integrated
eration of the staff at Shinshu University,
EMG was not calculated in present study, we
School of Allied Medical Sciences. We also
could not conclude the change of total electri-
wish to express our thanks to Mr. Kazuo Ku-
cal activity in EMG in motor learnlng prOC-
doh, the staff of the ME corporation, for sup-
porting this study.
From above findings, it seemed that fur-
ther electromyographical study in motor
learning process should be continued.
Crutchfield2) indicated that considerable
portion of therapy is to teach the clients. It is
also described that a large body of knowl-
1 ) Schmidt, R. A. : Motor learning and performance : from prlnClples to practice. Human Kinetics Books, Illinois, 1991.
2 ) Crutchfield. C.A.. Barnes, M. R.:Motor
control and motor learning in rehabilitation.
edge in teaching is based on the motor con-
Changes in frequency properties Of electroencephalograph and electromyography in motor learnlng prOCeSS 35
Stocksville, Atlanta,1993.
3 ) Adam, ∫. A∴ A closed-loop theory of mo-
task. Journal of Physiology, 421 : 553-568,
tor learning. Journal of Motor Behavior, 3 :
13) Pets E.E. and Cheney P.D. : Postspike fa-
111-150. 1971.
cilitation of forelimb muscle activity by prト
4 ) Schmidt. R. A∴A schema theory of dis-
mate corticomotoneuronal cells. ∫.Neuro-
crete motor skill learnlng. PsychologlCal Re-
physiol., 44 : 7511772. 1980.
view, 82 : 225-260, 1975.
14) Orfanidis, S. J∴ Optimal signal process-
5 ) Schmidt, R. A., Zelaznik, H. N∴ Sources of
inaccuracyinrapid movement. In : G.E Stelmach, eds., Information processing in motor
control and learning. Academic Press, New
ing. 2 nd. ed., McGrawIHill. Singapore. 1988.
15) Shibasaki. HりBarrett, G., Halliday, E. &
Halliday, A. M∴Components of the move-
ment-related cortical potential and their
scalp topography. Electro-encephalogr. Clin.
York, pp. 183-203, 1978.
Neurophysio1., 49 : 213-226, 1980.
6 ) Schmidt, R. A∴ Motor control and learn16) Fitts, P. M∴ PerceptuaLmotor skill learn-
ing :A behavioral emphasis (2nd ed.). Hu-
manKinetics, Illinois. 198&
ing. in Melton A.W. (ed) : Categories of human learning. Academic Press, New York,
7 ) Salmoni, A. W., et al∴Knowledge of Re-
sults and motor learnlng : A review and criti-
17) Newell, K. M∴ Knowledge of results and
cal reappraisaL PsychologlCal Bulletin. 95 :
motor learning. Journal of Motor Behavior,
355-386, 1984.
8 ) Murthy, V. N., Pets, E. E. : Coherent 25-35
6 : 235-244, 1974
18) Busk, ∫. and Galbraith, G. C. : EEC corre-
Hz oscillations in the sensorimotor cortex of
the awake behaving monkey. Proc. Natl.
Acad. S°i., 89 : 5670-5674, 1992.
9 ) Karni, A.. Meyer, G., Jezzard, P., Åams,
M.M., Turner, R.&Ungerleider, L G. : Functional MRI evidence for adult motor cortex
plasticity during motor skill learnlng. Nature,
377 : 155-158,1995.
10) Lang,W., Beisteiner, R., Lindinger, G.,
Deecke, L∴ Changes of cortical activitiy
when executing learned motor sequences.
Experimental Brain Research, 89 (2) : 435440,1992.
ll) Sadoyama. T., Masuda, T.:A quantitative evaluation of myoelectric signals in mo-
tor learning. Japanese Journal of Electroen-
lates of visual motor practice in man. Electro-
encephalography and Clinical NeurophysioL
ogy, 38 : 415-422, 1975.
19) Gliner, ∫. A., Mihevic, P. M., Horvath, S.
M∴ Spectral analys上s Of electro-encephal0-
gram during perceptuaLmotor learning. Bi0loglCal Psychology, 16 : 1-13, 1983.
20) Tanji, ∫. and Mushiake, H∴ Comparison
of neuronal activlty in the supplementary
motor area and primary motor cortex. Cognト
tive Brain Research, 3 : 143-150, 1996.
21) Thach, W. T∴ Correlation of neural dis-
charge with pattern and force of muscular
activity. joint position, and direction of in-
tended next movement in motor cortex and
cerebellum. ∫. Neurophysio1., 41 : 654-676,
cephalography and Electr0-myography, 10
(2) : 139-142. 1982.
22) Basmajian, ∫. Ⅴ∴Muscle Alive2nd ed.
12) vallbo, A, B., A1-Falahe, N. A∴Human
muscle spindle response in a motor learnlng
Baltimore. The Williams&Willkins C0., 1967.
23) Payton, 0.D. and Kelley, D. L∴ Electro-
myographic Evidence of the Acquisition of a
Motor Skill. PhysicalTherapy, 52 (3) : 261266, 1 972.
受付日: 1997年9月30日
受理日: 1997年11月18日