- Email: [email protected]
Movement-related cortical potentials associated with voluntary muscle relaxation
ELSEVIER
Electroencephalography
and clinical Neurophysiology 9.5 (1995) 335-345
Movement-related cortical potentials associated with voluntary muscle relaxation * K. Terada *, A. Ikeda, T. Nagamine, H. Shibasaki Depment of Brain
P~tt/zc)physiolo~y, Kyoto
School of’Medicine. Sho~oinknwukurumnclzi
Unil;ersity
54, Sakyo-ku.
Kyoto 606.01,
Jqmn
Accepted for publication: 8 May 1995
Abstract
We recorded movement-related cortical potentials (MRCPS) associated with voluntary muscle relaxation, which was not accompanied by contraction of the antagonist or any other muscles, in 10 normal subjects. Voluntary, self-paced relaxation of the wrist extensors from the extended position was employed as the relaxation task, and wrist flexion by muscle contraction was employed as the contraction task. The accelerogram was used to obtain the trigger signals for both tasks. The electromyograms of the ipsilateral agonist and antagonist, the proximal muscles and the contralateral corresponding muscles were monitored to confirm the absence of muscle contraction for the relaxation task. All MRCP components were identified in both tasks; Bereitschaftspotential (BP), negative slope (NS’), parietal peak of motor potential (ppMP) and frontal peak of motor potential (fpMP). BP started earlier and was larger at the contralateral parietotemporal electrodes for the relaxation than for the contraction task, and the slow positive shift at the bilateral frontopolar electrodes was seen more often in the relaxation task. It is concluded that the voluntary muscle relaxation needs a cortical preparatory process similar to voluntary muscle contraction, and needs a more extensive and longer preparation process in the primary motor area and possibly in other motor areas as compared with the contraction. Kry\~or&:
Movement-related
cortical potentials; Voluntary
muscle relaxation; Negative motor phenomenon; Motor control
1. Introduction Movement-related cortical potentials (MRCPS) are defined as brain potentials associated with self-paced, voluntary movements (Kornhuber and Deecke, 1965; Vaughan et al., 1968; Shibasaki et al., 1980). MRCPs are of great interest because they are expected to help elucidating the neuronal mechanisms participating in the movement preparation, initiation, execution and feedback control (Benecke et al., 1985; Barrett et al., 1985, 1986; Tarkka and Hallett, 1990, 199 1). Currently it is accepted that pre-movement components of MRCPs consist of at least two subcomponents: the earlier slow negativity, Bereitschaftspotential (BP) (Kornhuber and Deecke, 1965) and the late negative
” This study was supported by Grant-in-Aid 0640403
I,
Grant-in-Aid
Grant-in-Aid
for
Priority
New Program 06NP0101,
search Grant 07044258
Scientific
for Scientific Research
and International
Research 06260225.
ReScience
Scientific
from the Japan Ministry of Education.
and Culture for H.S.
’ Corresponding 0013-4694/95/$09.50
author. Tel.: 8 I-75-751 -3603; Fax: 8 l-75-751-3202.
0
SSDI 0013-4694(95)00098-4
slope (NS’) (Shibasaki et al., 1980). These potentials are most likely to represent the function of the primary and supplementary motor areas, as shown by subdural recording in patients with intractable epilepsy (Lee et al., 1986; Neshige et al., 1988; Ikeda et al., 1992, 1993). In all of the previous studies on MRCPs, those potentials were recognized in association with the initiation of self-paced voluntary muscle contraction, i.e., “positive motor phenomenon,” and no studies have been reported in association with voluntary, pure relaxation of the ongoing muscle contraction, i.e., “negative motor phenomenon.” Although Dimitrov (1985) reported the presence of BP associated with the termination of the voluntary finger movement, he did not monitor the antagonist muscles and concluded that the BP was related to contraction of the antagonists rather than to pure relaxation. Recently, negative motor phenomena have attracted the attention of many investigators: negative myoclonus (Ugawa et al., 1989, 1990; Artieda et al., 1992; Shibasaki et al., 1994), negative motor response accompanying the motor evoked potential produced by transcranial magnetic stimulation of the motor cortex (Calancie et al., 1987; Fuhr et al., 199 1; Roick et
1995 Elsevier Science Ireland Ltd. All rights reserved EEG 94 I93
336
K. Teruda et ul. / EIectroencephaloRruphy
und clinical
Neurophysiology
95 (I 995) 335-345
al., 1993; Uncini et al., 1993; Wassermann et al., 1993; Wilson et al., 1993) and negative motor response to direct cortical electrical stimulation (Penfield and Jasper, 1954; Liiders et al., 1987, 1992). Since the mechanisms underlying voluntary negative motor phenomena are still unclear, we analyzed MRCPs associated with voluntary muscle rel;vation which was carefully controlled to avoid associated contractions of other muscles, especially those of the antagonists.
board
Paper
Fig. 1. Position of the subject’s right hand before each relaxation movement. Four digits were attached to a light paper plate on the dorsum side
2. Methods
and were wrapped together by an adhesive tape. The subject kept the
2.1. Subjects
onset, an accelerometer was attached to the distal tip of the plate.
right wrist in extended position to 150-160”.
Ten healthy volunteers (8 males and 2 females), aged 25-45 years (mean 30.3), served as subjects for the present study. Nine of them were right-handed and one was left-handed. None of them had a previous history of neurological illness. The purpose of the experiment was explained to each subject beforehand, and an informed consent was obtained. 2.2. Movement
paradigm
The subject was seated in an arm-chair in a quiet room with eyes kept open and fixating forward at the target in order to eliminate blinks and slow eye movements. Each subject performed two kinds of tasks both with the right hand, i.e., “relaxation task” and “contraction task” as described below. Each recording session consisted of 50 trials of the same task, and each subject had 3 sessions for the relaxation and contraction tasks which were performed alternately. For all subjects, the initial session was the relaxation task because a longer training period was needed for the relaxation than for the contraction task.
Contraction task The forearm was placed perpendicular to the paper board, ulnar side down, and the subject was instructed to keep the wrist in complete relaxation for more than 5 sec. The relaxation was confirmed by monitoring the polygraphic surface EMG records. Then the subject flexed the wrist voluntarily by contracting the wrist flexor muscle as quickly as possible. The subject was trained to make the wrist flexion to 150-160”. After having maintained the flexed position for more than 5 set, the subject returned the wrist to the relaxed, neutral position for the next trial. The rate of contraction movement was the same as that for the relaxation task. The performance was also monitored by the video camera system as described above. 2.3. Data acquisition For recording the electroencephalogram (EEG), 2 1 Ag/AgCl cup electrodes (FPl, FP2, F7, F3, Fz, F4, F8, Voluntary
Relaxation task The forearm was positioned, palm down, on a paper board placed on the arm-rest. Before each relaxation movement, the subject kept the right wrist in extension to 150-160” for at least 5 set (Fig. 1). The subject was trained so that the electromyographic (EMG) activity was restricted only to the extensor carpi radialis muscle (ECR) throughout the extended position (Fig. 2). Then, the subject relaxed the ECR muscle by terminating the contraction voluntarily (without any cue signal) and as quickly as possible. During the training, great caution was given by monitoring the polygraphic EMG as described below, so that the hand could drop without any associated contraction of other muscles. Once the complete relaxation was achieved, the subject kept the EMG silent for more than 5 set and then resumed the extended position of the wrist for the next trial. The rate of relaxation movement was once per 15 set on average for every subject. The performance was monitored with the aid of a video camera system throughout the experiment.
To detect the movement
4
Wrist
in
extension
relaxation * ‘I
rt. DELT rt. BR rt. ECR rt. FCU
-i---
rt. BIG rt. TRIG
-A-
It.ECR ----.
It.FCU
a-_l__._-._ -- ~- ‘1
?-
__--._._.-_ T
.____
,/=
I-
._
A
_ 1OO~b I see
Tr iggsr
Fig. 2. A sample of relaxation trials showing typical, acceptable EMG activities.
DELT
= deltoid. BR = brachioradialis,
ECR = extensor carpi
radialis, FCU = flexor carpi ulnaris, BIC = biceps brachii, TRIG = triceps brachii muscle, rt. = right, It. = left, Accel. = accelerogram. EMG
Note that the
activity is completely absent during the relaxation not only in the
ECR but also in other muscles of both upper extremities.
K. Terudu et al. / ElectroencephaloKraph~
T3, C3, C 1, Cz, C2, C4, T4, T5, P3, Pz, P4, T6, 01 and 02) (American EEG Society, 1992) were affixed on the scalp with collodion, and the impedance of all electrodes was kept less than 5 k 0. All electrodes were referenced to linked earlobe electrodes. The time constant (TC) was set to 3 sec. and the high frequency filter (HFF) used was 100 Hz. A 60 Hz notch filter was applied for all channels. EOG was monitored by an Ag/AgCl cup electrode placed at the right inferior lateral canthus referenced to FPz. TC and HFF for the EOG recording were set to the same values as for the EEG recording. In order to monitor the contraction and relaxation of the corresponding muscle and to detect concomitant contraction of other muscles (see above), EMGs were recorded by a pair of Ag/AgCl cup electrodes placed 3 cm apart on the skin overlying each of the right ECR, flexor carpi ulnaris (FCU), deltoid (DELT), biceps brachii (BIG), triceps brachii (TRIC) and brachioradialis (BR) muscles, and each of the left ECR and FCU muscles (Fig. 2). TC for the EMG recording was set to 0.075 set, and HFF to 100 Hz. EMGs were not rectified due to the limited capacity of the instrument, but the waveform of the averaged EMG was not important in the present study because the accelerogram was used to detect the movement onset (see below). To determine the movement onset, an accelerometer (AS-2GA, Kyowa Electronic Instruments) was fixed to the light paper plate which was bound on the dorsal aspect of the digits II-V by tapes (Fig. 1). Low frequency filter for the accelerometer was set to 0 Hz (DC) and HFF to 100 Hz. The accelerogram was rectified and used to obtain the fiducial point for averaging the EEGs and EOGs as described below. All segments of the amplified signals covering 2.4 set before and 1.2 set after the movement onset were stored on a computer (DPl 100, NEC-Sanei) at a sampling rate of 333 Hz for each channel for the subsequent off-line analysis. 2.4. Duta analysis After the completion of all sessions for each subject, the EEGs, EOG, EMGs and the accelerogram for each trial were displayed on the computer screen, and the precise onset of the rectified accelerogram was visually determined which was then used as the fiducial point for back-averaging the EEGs and EOG (Barrett et al., 1985). If there were any artifacts arising from eye movements, blinks or other sources, or if EMG activities were recognized in inappropriate muscles, that trial was eliminated from the subsequent analysis. After confirming the reproducibility of the averaged waveforms among different sessions of the same movement task, a group average for each task was obtained for each subject. Then, a grand average waveform for each task was obtained by averaging the group average data of like sessions across all the subjects. Although one of the
and clinical
Neurophy.Golog
95 Cl9951 335-345
337
amplitude
anpl i tude
CZbt?\
amp1itude
baseline
ppMP latency A =I ppMP amplitude
P3 \ -2sec
-1sec
movement onset
+
Fig. 3. Schematic diagrams illustrating BP (Bereitschaftspotential),
nega-
tive slope (NS’), parietal peak of motor potential (ppMP) and frontal peak of motor potential (fpMP),
and the measurements of amplitude, duration
and time interval of those MRCP components. The upward deflection for EEG data indicates the negativity at the exploring electrodes in this and all other figures.
male subjects was left-handed, his data were included because his group average waveforms and scalp distribution appeared similar to those of other subjects. Measurements of various time parameters with respect to the movement onset and amplitudes of identifiable components were computer-assisted. Amplitude measurements were based on the calibration data collected just before each experiment. Baseline was det.ermined by averaging the segment from 2.1 set to 1.6 set before the movement onset for each channel. The onset of BP and NS’ was visually determined and measured according to the method described previously (Kitamura et al., 1993a,b) (Fig. 3). The onset of BP was defined as the time when the baseline began deflecting toward negativity at electrode Cz, and its largest negative peak was defined as the BP peak. The BP duration was measured from its onset to peak. The amplitude of the BP peak at each electrode was measured from the above-described baseline. Likewise, the onset of NS’ was visually determined as the time when the following steeper slope started at Cl, and its largest peak was defined as the NS’ peak, regardless of whether it occurred before or after the movement onset. Consequently, the BP peak and the NS’ onset did not necessarily coincide. The NS’ duration was measured from its onset to peak. The amplitude of the NS’ peak was measured as the amplitude difference from its onset at each electrode. Although another negative slope corresponding to the intermediate shift (IS) (Barrett et al., 1986) was identified between BP and NS’ in the grand
338
K. Terada et al. / Elecfroencephalography
average waveform of the contraction data, it was identified in only 3 of the 10 individual data. Therefore, we dealt with that slope as a part of BP in this study. The parietal peak of motor potential (ppMP) and the frontal peak of motor potential (fpMP) were determined and measured using the definition proposed by Tarkka and Hallett (1990, 1991) (Fig. 3). The ppMP was defined as the first turn toward positivity after the movement onset at P3 electrode. The ppMP amplitude was measured from the immediately preceding positive peak (premotion positivity). The fpMP was defined as the largest negativity of the entire analysis window at Fz electrode. The fpMP amplitude was measured as the amplitude difference from the value at the movement onset. Differences in the time parameters of each MRCP component between the two tasks were statistically analyzed by the Wilcoxon signed-rank test. Differences in the amplitudes of each MRCP component between the two tasks were statistically analyzed by an analysis of variance for repeated measures (ANOVA) with factors of “task” and “electrode position” followed by the assessment using the conservative correction of Greenhouse and Geisser
(a>
Acce I Tri&er (b) rt.ECR
rt. BIC ...-~--__.____ TR ,c
Neurophysiology
_.__----______._____.___+
Accel. ___~
_.__.
_._,c i
__-_
._~. ._
+_.__ Trigger
.---
95 (1995) 335-345
which is designed to avoid non-replicable positive results by restricting the theoretically appropriate degrees of freedom. Furthermore, in order to evaluate the statistical significance of the possible asymmetry of each potential over the scalp for each task, the amplitudes between each pair of homologous electrodes, i.e., FPl and FP2, were also assessed using the conservative correction of Greenhouse and Geisser for the both relaxation and contraction tasks. The amplitudes of BP, NS’, ppMP and fpMP measured in the grand average waveforms were displayed as topographic maps by using Nicolet Pathfinder TMAP software, which used a triangular interpolation matrix. Calibration of the color scale was set to the maximal value of the absolute amplitude in the both relaxation and contraction tasks for each component.
3. Results The number of trials accepted for the analysis per subject was 20-106 (mean 78.3) for the relaxation task and 51-122 (mean 89.3) for the contraction task. Fig. 2 demonstrates a trial for the relaxation task showing well controlled EMG activities. We accepted only those trials in which EMG activities of the right ECR muscle completely disappeared just before or after the movement onset and in which no EMG activity was observed in any of the FCU and other muscles on either side throughout the trial. We rejected the trials in which inappropriate EMG activities, especially those in FCU, were observed before or after the movement onset (Fig. 4b). Fig. 4a shows an acceptable sample in which a phasic activity in the ECR muscle was observed 130 msec after the movement onset. This trial was accepted because the phasic EMG activity was interpreted to be a stretch reflex. As the result of averaging those trials containing the stretch reflex and also due to variable patterns of EMG offset, the activity of ECR in the averaged waveform demonstrated a rather gradual decrease in activity instead of abrupt cessation (Fig. 5a compared to Fig. 5b). 3.1. MRCPs relaxation
rt.FCU
rt,
and clinical
associated
with ooluntary
self-paced
muscle
--
. ..__._
J-w 5w mm+
Fig. 4. Two samples of EMG activities during the relaxation task which were accepted for (a) and rejected from (b) the analysis, respectively. a: a brief EMG activity is observed in the ECR muscle 130 mse.c after the onset of relaxation, which is most likely a stretch reflex. b: an inappropriate EMG activity is observed in the FCU muscle just after the fiducial point, suggesting a purposeful contraction of the wrist flexor muscle.
Fig. 5a demonstrates the grand average waveforms of MRCPs for the relaxation task across all the subjects. A slow negative shift, which corresponds to BP for the contraction task, started at 1600 msec before the movement onset and was maximal at the vertex (Cz) and the left central area (Cl). The mean onset time of BP in 10 subjects was 1521 msec before the movement onset (Table 1). At 220 msec before the movement onset (mean value in 10 subjects 275 msec; Table l>, the slow negative potential became steeper in the left hemisphere, maximal at the left central area (Cl). This steeper negative shift corresponds to NS’ for the contraction task. The small ppMP was
K. Terada et al. / Electroencephalography
md clinical
Neurophysiology
95 (1995) 335-345
339
n FE F3
T3
F4
T4
c3
T5
c4
T6
Cl
Fz
C2
cz
P3
PZ
P4 rt.ECR
01
rt.FCU
02
rt.BIC
EOG
rt.TRIC ACC IEEG, EOG / lO/.iV -2 set
lb)
-1
trigger
n
FP2
F8
F3
T3
F4
T4
c3
T5
c4
T6
Cl
Fz
c2
cz
P3
PZ
/-
20,uv
ACC j
0.25
G
I
I+
‘+
FPl
EMI
P4 01
rt.ECR
02
rt.FCU
EOG
rt.BIC rt.TRIC/------------ y--b"++-__. _._
.-. -2 set
trigger
EEG, EOG -10/1V +
EMG
2O/.tV
__~ ACC
0.25
G
+
Fig. 5. Grand average waveforms of cortical potentials across 10 normal subjects, associated with voluntary, self-paEd relaxation of the right ECR muscle (a) and voluntary, self-paced contraction of the right FCU muscle (b), both resulting in flexion movement of the right wrist (average of 783 and 893 trials, respectively). ACC: averaged, rectified accelerogram. In (a), BP struts at 1600 msec before the movement onset, is maximal at Cz and Cl, and becomes steeper at 220 msec before the movement onset on the left hemisphere with the maximum at Cl (NS’). At the bilateral frontopolar areas @PI and FP2), a small slow positive shift can be recognized. After the movement onset, a small negative peak is identified at 90 msec at the contralateral centroparietal electrodes (ppMP) and another negative peak at 210 msec at the frontocentral electrodes (fpMP). In (b), BP sttis at 1500 msec before the movement onset, is maximal equally at Cl and Cz, and becomes steeper at 210 msec before the movement onset on the left hemisphere with the maximum at Cl (NS’). After the movement onset, a small negative peak is identified at 24 msec at the contralateral centroparietal electrodes (ppMP) and another negative peak at I IO msec at the frontocentral electrodes (fpMP).
340
K. Terada et al. / Electroencephdography
identified at 90 msec after the movement onset at the contralateral centroparietal electrodes. The fpMP was identified at 210 msec after the movement onset maximally at the midline frontocentral electrodes (Fz and Cz). At the
and clinical Neurophysiology 95 (1995) 335-34.5
.bilateral started topolar The
frontopolar regions (FPl and FP2), a positive shift 1500 msec before the movement onset. This fronpositivity was seen in 6 out of the 10 subjects. topographic map of BP for the relaxation task
Fig 6. Topographic maps of BP, NS’, ppMP and fpMP obtained from the grand average waveforms for the relaxation (a) and the contraction (b) tasks BP disctribution is clearly different between the two tasks, i.e., the negative field at the parietal area is more prominent for the relaxation task as compare .d to the contraction task, and the frontal positivity is seen only for the relaxation task. Other components show similar distribution in the two tasks.
K. Terada et al. / Electroencephalography
and clinical
Table I Temporal parameters of each recognizable component of MRCPs associated with voluntary, self-paced relaxation and contraction tasks with respect to the movement onset in 10 normal subjects (msec, mean k SD.) Relaxation
Contraction
- 152I .O+ 233.6 -1216.8k334.7 BP onset time * 811.5k320.3 1160.1 f 177.5 BP duration * - 274.8 + 143.9 -235.2 f 142.8 NS’onset time * * 213.8f 138.2 242.4 k 149.5 NS’duration * * t 52.5+ 30.1 + 43.8 &- 24.9 ppMP peak time * * * t 150.6f 46.7 foMP peak time * * * ’ t 229.8 f 72.6
P
CO.05 <0.05 ns. n.s. n.s. < 0.05
“-” and “t” indicate “before” and “after” the movement onset, respectively. ns. = not significant in this and all other tables. measuredatCz*,CI**,P3”*,andFz*‘*’.
a relatively symmetric distribution of the negativity at the centroparietal areas with the maximum at Cz and Cl (Fig. 6a). The associated positive field was localized symmetrically at the frontopolar midline area. The NS’ was distributed at the left frontal, central and parietal areas with the maximum at Cl. The ppMP appeared as a negative field at the left central area, and the fpMP was seen as a relatively large negative field at the midline frontocentral area and distributed symmetrically.
demonstrated
Neurophysiology
95 (1995) 335-345
341
3.2. MRCPs associated with voluntary self-paced muscle contraction
Fig. 5b demonstrates the grand average waveforms of MRCPs for the contraction task across all the subjects. BP started at 1500 msec before the movement onset and was maximal at the vertex (Cz) and the left central area (Cl) with equal amplitude. The mean onset time of BP in IO subjects was 1217 msec before the movement onset (Table 1). NS’ started at 210 msec before the movement onset (mean value in 10 subjects 235 msec; Table I) and was maximal at the left central area (Cl). The small ppMP was seen at the centroparietal area at 24 msec after the movement onset. The fpMP was clearly observed at 110 msec after the movement onset at the frontocentrai electrodes with the maximum at Fz. Unlike the relaxation task, there was no slow positive shift seen at the frontal area in the grand average waveform. In the individual data, however, 3 subjects demonstrated a small positive shift at the frontopolar area, who had the positive shift also in the relaxation task. The topographic map of BP for the contraction task demonstrated a localized negative field at the midline central area with the maximal amplitude at Cl (Fig. 6b). The NS’ was maximal at Cl and distributed at the cen-
Table 2 Amplitudes of BP and NS’ of MRCPs associated with voluntary, self-paced relaxation and contraction tasks in 10 normal subjects, measured at each electrode ( pV, mean f S.D.) Electrode
FPI FP2 F3 F4 c3 c4 Cl c2 P3 P4 01 02 Fl F8 T3 T4 TS T6 FZ CZ PZ
NS’b
BP = Relaxation
Contraction
PC
Relaxation
Contraction
PC
0.69 + I .78 1.01 f 1.35 - 0.84 + 0.96 - 0.77 f I .oo -2.21* 1.07 - I .73 + 0.85 -3.16+ 1.34 -2.41 kO.61 -2.11 f 1.73 - 1.s7+ 1.39 - 0.97 f 1.69 -0.87* 1.69 -0.19*0.19 0.21 +0.70 -0.86+ 1.13 - 0.39 + 0.87 -0.73+ 1.19 -0.82* 1.01 - 1.44+0.88 -3.71rto.53 - 2.43 f I .28
-0.55+2.50 -0.58 + 2.70 -1.31*1.41 -1.11~1.30 -2.04f 1.31 - I .60 + 0.89 -2.62& 1.06 - I .98 + 0.97 - 1.19+ 1.32 - I .42 f 1.24 - 0.42 f 1.67 - 0.73 f 1.65 -0.19* 1.32 - 0.68 f I .22 0.18& 1.39 -0.53+ 1.00 -0.13f 1.26 -o.s2+ 1.21 - 1.50* 1.51 -2.71 + 1.12 -1.25dzl.27
< 0.05 < 0.05 n.s. n.s. n.s. ns. ns. ns. n.s. n.s. n.s. n.s. n.s. n.s. < 0.05 n.s. ns. n.s. ns. n.s. < 0.05
-0.98+2.18 - 1.30+ 1.50 - 3.07 f 1.50 - 2.43 + 1.20 -4.43+1.24+ - 2.62 f I .3 I + -4.87+ 1.53 + - 3.67 f I .45 + - 3.25 + I .29 - 2.57 + I .67 - 1.42f 1.63 - 1.54* 1.52 - 1.42f 1.10 - 1.19+ 1.05 - 1.52zb2.06 -0.81 *0.71 - 1.24f 1.00 -0.71 + 1.27 - 3.32 L- I .61 - 4.67 f 1.70 - 3.48 f 1.62
-0.81 f I.51 - 0.66 f I .93 - 1.90* 1.60 - 1.83* 1.88 -3.18* 1.83 + - I.955 1.44+ -3.62+2.25 - 2.75 f I .83 - 2.94 * 2.04 - 2.40 f I.56 - l.84* 1.63 - 1.72+ 1.41 -0.95* 1.00 - 0.48 * I .45 - 1.2 I + I .06 - 1.06+ 1.65 - 1.52* 1.55 - I .27 f 0.86 -2.501 1.87 -3.39+2.19 - 2.87 f I .92
ns. ns. n.s. ns. < 0.05 n.s. < 0.05 n.s. n.s. n.s. n.s. ns. ns. n.s. n.s. n.s. ns. ns. n.s. < 0.05 n.s.
a Measured from the basehne. h Measured as the amplitude difference from its onset. ’ Comparison of relaxation vs. contraction at each electrode. ‘*- ” indicates negativity at each exploring electrode with respect to the corresponding reference level. ./_P < 0.05.
342
K. Terada et al. / Electroencephalography and clinical Neurophysiology 95 (1995) 335-345
troparietal area, extending more to the left hemisphere. The ppMP was localized at the left central area. The fpMP was seen as a negative field at the frontocentral area, being distributed more predominantly to the left. 3.3. Comparison traction tasks
of
A4RCPs between
relaxation
and con-
Comparing the amplitudes between tasks and among electrode positions, the ANOVA demonstrated a signifi(P < 0.05) but no cant effect of “electrode position” significant effect of “task” for all MRCP components. A significant interaction between “task” and “electrode position” (P < 0.05) factors was seen only for the BP. The BP associated with the relaxation task started earlier and its duration was longer as compared with the contraction task (P < 0.05; Table 1). The BP amplitude at T3 and Pz was significantly larger in the relaxation task than in the contraction task (P < 0.05>, and the statistically significant positive activity at FPl and FP2 was observed for the relaxation task (Table 2 and Fig. 6). As for the NS’, there was no statistically significant difference in onset time and duration between the two tasks (Table 1). The amplitude of NS’ at C3 and Cl was significantly larger in the relaxation task than in the contraction task (P < 0.05; Table 2). With regard to the ppMP, there was no statistically significant difference in its time parameters between
Table 3 Amplitudes of ppMP and fpMP of MRCPs associated electrode ( WV, mean f S.D.) Electrode
FPI FP2 F3 F4 c3 c4 Cl c2 P3 P4 01 02 F7 F8 T3 T4 T5 T6 FZ CZ Pz
with voluntary,
self-paced
the two tasks (Table 1). However, the ppMP amplitude at Pz was significantly larger for the relaxation than for the contraction task (P < 0.05; Table 3). The peak time of fpMP associated with the relaxation task was longer than that with the contraction (P < 0.05; Table 1). No statistically significant difference was seen for the amplitude of fpMP between the two tasks (Table 3). As the result of comparison between the two sides, there was no statistically significant difference in the amplitude of BP, ppMP and fpMP between either pairs of the homologous electrodes for both tasks (Tables 2 and 3). The amplitude of NS’ was larger at C3 and Cl than at C4 and C2, respectively, for the relaxation task (P < 0.05) and was larger at C3 than at C4 for the contraction task (P < 0.05; Table 2).
4. Discussion 4.1. Comparison
between relaxation and contraction
tasks
In the present study, the slow negative potentials corresponding to BP and NS’ were successfully recorded for the first time prior to voluntary muscle relaxation which was confirmed by the EMG silence of antagonist and other proximal muscles. Thus the present findings suggest that, at least from an electrophysiological point of view, volun-
relaxation
and contraction
tasks in IO normal subjects,
measured
fpMP h
ppMP a Relaxation
Contraction
PC
Relaxation
Contraction
PC
- 1.00+1.41 - 0.75 f I .64 -2.14+2.04 - 1.94+ 1.82 -2.45 f 1.54 -2.06f 1.67 -2.83+ 1.93 - 2.52 f I .89 -2.16? 1.43 -2.01 f 1.73 - I .09 * 1.02 -1.23+1.16 - 1.00+ 1.12 - 0.86 f 0.85 - I .43 + 1.23 - 0.68 + 0.9 I -0.78+0.91 - 0.55 + I .26 -2.08 ir 1.79 -2.88 +2.06 - 2.47 + I .62
- 1.60+ 1.16 - 1.46+ 1.43 - 1.79* 1.16 - I.175 1.08 -2.06& 1.03 - I .42 + 0.98 -2.07+0.83 - 1.62+0.82 - 1.26kO.87 - 1.10+0.82 -0.80+0.88 -0.53* 1.11 -0.69, 1.74 - 0.62 f 0.55 - 1.06+ 1.65 -0.51* 1.00 -0.66*0.81 0.27 f 0.9 1 - 1.74f 1.02 - 1.82 k 0.75 -0.35)0.69
ns. ns. n.s. n.s. n.s. n.s. ns. ns. n.s. n.s. ns. ns. n.s. ns. n.s. n.s. ns. ns. ns. n.s. < 0.05
- 2.40 f 1.79 -2.40+ 1.91 -3.59+ 1.39 -2.93+ 1.46 -2.68+ 1.56 -2.38, 1.54 -3.44+ 1.82 -2.97f 1.66 - 0.84 f 2.07 - 1.36* 2.04 0.26 f I .48 0.05~1.51 - 1.81 + 1.12 -2.14i0.76 - I .56 + 2.35 - I .68 k 0.88 0.49 * 1.07 -0.53+ 1.18 - 3.94 f I .23 - 3.55 & I .78 - I .46 + 2.17
-2.15+1.68 - 2.24 f I .28 - 2.98 k 2.07 - 2.34 * 1.87 -2.13+ 1.21 - I .95 * I .04 -3.14+ 1.62 - 2.56 + I .39 - 0.67 + I .03 - 1.47+ 1.18 -0.20* 1.41 -0.34+1.17 - 2.10 + 1.09 - I .29 + I .24 - 0.90 * I.59 -0.71 + 1.66 - 0.29 + I .08 - 0.40 * 2.00 - 3.45 + 1.77 -3.16+1.72 - 1.38* 1.07
n.s. n.s.
’ Measured from the immediately preceding positive peak. h Measured from the movement onset. ’ Comparison of relaxation vs. contraction at each electrode. “ - ” indicates negativity at each exploring electrode with respect to the corresponding
reference
point
“3.
ns. ns. us. us. n.s. n.s. “.S.
n.s. ns. “3.
n.s. n.s. ns. ns. n.s. n.s. n.s. ns.
at each
K. Terudu et al. / Electroencep~logrph!,
und clinical Neurophpiolo:v
tary muscle relaxation has similar preparatory mechanisms to voluntary contraction at the cortical level. Furthermore, the overall similarity of waveforms to those of BP and NS’ associated with voluntary muscle contraction suggests that the main cortical generators of those potentials might be common for both tasks, namely the primary and supplementary motor areas according to the results of subdural recording (Neshige et al., 1988; Ikeda et al., 1992, 1993). However. as the result of quantitative and topographical comparisons of MRCPs, some differences were observed between the two tasks. First, the onset of BP was earlier and its duration was longer for the relaxation than for the contraction task, suggesting that the voluntary muscle relaxation needs longer preparation than the voluntary contraction. However, it is also possible that the different movement velocities, slower in the relaxation task than in the contraction task, might have resulted in a different onset time (Becker et al., 1976). Secondly, the BP at the contralateral temporal and midline parietal areas (T3 and Pz) and the NS’ at the contralateral central area (Cl, C3 and Cz) were larger for the relaxation than for the contraction task, and the slow positive activity was observed at the ipsilateral frontal area (FPI and FP2) only in the relaxation task. This anterior slow positive shift, however, was also observed for the contraction task in 3 out of 10 subjects in this study, and was previously observed in association with the contraction task (Shibasaki et al., 1980), suggesting that this is not an absolute difference between the two tasks. Since the amplitudes of BP and NS’ were larger at the contralateral centroparietal area for the relaxation task than for the contraction task, it is possible that the primary motor area is more active in the preparation for the voluntary muscle relaxation than for the contraction task, adding the contralateral parietal negativity and the frontopolar positivity as the result of a tangentially oriented dipole sitting in the anterior bank of the central sulcus contralateral to the movement. However, the other possibility that two independent activities might be present: one in the midline frontal region and the other in the parietal region, more predominantly for the relaxation than for the contraction task cannot be excluded. With regard to the frontal positivity, its possibility being the contamination by eye movement artifacts was unlikely because the EOG recorded bipolarly with the same sensitivity as for EEGs did not show any corresponding activity. The ppMP is thought to correspond to the terminal phase of the motor cortex firing which is related to either the preparation or execution of the movement (Tarkka and Hallett, 1991). While the onset time of ppMP was not statistically different between the two tasks, its amplitude at Pz was larger for the relaxation task than for the contraction. It is therefore suggested again that the motor cortex might be more active also in the actual execution of the muscle relaxation than that of the muscle contraction. The fpMP was suggested to be generated in the supplementary motor area and to reflect the kinesthetic sensory
95 11995) 335-345
343
input from the peripheral nervous system (Tarkka and Hallett, 1991). A clear fpMP peaked at + I IO msec in the contraction task and at +210 msec in the relaxation task in the present study. The time difference of fpMP in the two conditions could be partially explained by the larger jitter of the speed in the relaxation task as demonstrated by the waveform of accelerogram (Fig. 5a and b). The distribution of BP was symmetric with the midline vertex maximum in previous studies (Shibasaki et al., 1980; Barrett et al., 1986). In the present study, BP appeared larger in the central area slightly contralateral to the movement in both tasks (Fig. 6a and b). However, there was no statistical significance between the two sides for both the relaxation and contraction tasks. The contralatera1 predominance of NS’ appeared more prominent for the contraction than for the relaxation task on the topographic maps (Fig. 6a and b), but the statistical difference in the transverse distribution between the two sides demonstrated more prominent asymmetry for the relaxation than for the contraction task (Table 2). It was suggested that the primary motor cortex was more activated for the relaxation than for the contraction task, even in the NS’ period.
4.2. Mechanism
of negative
motor
phenomena
Negative motor response, that was defined as the inability to perform voluntary movement or to sustain voluntary muscle contraction, was seen when a part of the premotor cortex just above the sylvian fissure or the anterior part of the supplementary motor area was electrically stimulated (Penfield and Jasper, 1954). Hence these areas were called “primary and supplementary negative motor areas,” rcspectively (Liiders et al., 1987, 1992). In the present study, since the positive field of BP for the relaxation task was located in the bilateral frontal area, it might be possible that the positive field was generated in the supplementary negative motor area. In experimental studies, it was reported that ablation of the cerebellum or cooling of the dentate nucleus diminished the surface-negative, depth-positive field potential recorded in the motor cortex of monkey preceding selfpaced movement (Sasaki et al., 1979; Tsujimoto et al., 1993). Those experimental data suggest that BP and NS’ represent the excitatory postsynaptic potentials (EPSPs) of the corticomotor neurons as a result of the facilitatory input through the cerebella-thalamo-cortical pathway. Furthermore, disynaptic inhibition of the spinal motor neurons mediated by group Ia interneurons and pure inhibition of motor unit discharge without additional increase of motor units were demonstrated by electrical stimulation of the corticomotor neurons in monkey (Jankowska et al., 1976). It was also suggested that some corticomotor neurons might have an inhibitory effect on spinal motor neurons: “inhibitory corticomotor neurons” (Cheney et al., 1985; Lemon et al., 1987; Wassermann et al., 1993).
344
K. Terada et al. / Electroencephulography
and clinical
In the present study, the waveforms of BP and NS’ for the relaxation task were, as a whole, similar to those for the contraction task with the maximum at the centroparieta1 area. Taken together with the above experimental data, it is suggested that BP and NS’ preceding voluntary, self-paced relaxation may represent EPSPs on the inhibitory corticomotor neurons. The evidence to support that positive and negative motor phenomena might occur through common mechanisms at the cortical level was also reported in patients suffering from cortical negative myoclonus (Guerrini et al., 1993; Shibasaki et al., 1994). The precise localization of inhibitory motor centers requires further studies by using other non-invasive studies such as magnetoencephalography and measurements of cerebral blood flow by positron emission tomography.
Neurophysiolo,gy
Jankowska,
95 (1995) 335-345
E., Padel, Y. and Tanaka,
R. (1976)
Disynaptic
inhibition
spinal motor neurons from the motor cortex in the monkey. (Land.),
258: 467-487.
Kitamura, J., Shibasaki, potential
H. and
Kondo,
T. (1993a)
is larger before an isolated movement
simultaneous physiol.,
movement
of two
fingers.
A
cortical
slow
of a single finger than
Electroenceph.
clin.
Neuro-
86: 252-258.
Kitamura, A.
of
J. Physiol.
J., Shibasaki,
(1993b)
sequential
A., Nabeshima,
negative
as compared
Electroenceph. Komhuber,
H., Takagi,
Enhanced
with
slope
and Deecke,
L.
H. and Yamaguchi,
cortical
simultaneous
clin. Neurophysiol.,
H.H.
of
potentials
extensions
before
of two fingers.
86: 176-182. (1965)
Himpotentiallnderungen
Willkiirbewegungen
und passiven
Bewegungen
reitschaftspotential
und reafferente
Potentiale.
bei
des Menschen: Pfliigers
Be-
Arch.,
284:
l-17. Lee, B.I., Liiders, Cortical
H., Lesser, R.P., Dinner,
potentials
recorded
from
D.S. and Morris,
related to voluntary
subdural
H.H. (1986)
and passive finger
electrodes
in
humans.
Ann.
movements Neural.,
20:
32-37. Lemon,
R.N., Muir,
activity
the vicinity Brain
References
of corticomotor
H., Lesser,
(1987) Negative J., Muruzabal,
J., Larumbe,
Obeso, J.A. (1992) Dis., 7: 209-2 Barrett,
Cortical
H. and Neshige,
for averaging
G.,
W., Iwase,
movement:
In: W.C.
preceding
McCallum Bristol,
Benecke,
Calancie,
of
B., Nordin,
the
Fetz,
facilitation
J.C., Day, B.L.
Neurosci.
Lett.,
BereBrain.
and
and
K.E. (1987)
of antagonistic monkey.
Motor
185.
forelimb
Patterns muscles
J. Neurophysiol.,
of
from
53: 805-
Dimitrov,
B. (1985)
Brain
potentials
termination
of voluntary
chophysiol.,
3: 13-22.
P.. Agostino,
excitability Guerrini,
flexion Hallett,
the silent
and Tassinari,
C., Genton,
C.A. (1993)
M.
period
clin. Neurophysiol.,
R., Dravet,
and extension (1991)
and to the
in man. Int. J. PsySpinal
after cortical
motor
stimulation.
neuron Elec-
Epileptic
M., Roger, J., Rubboli,
negative
myoclonus.
G.
Neurology,
H.O.,
ment-related
potentials
primary
motor
movements.
Burgess,
Brain,
recorded
from
supplementary motor
H. (1992)
Move-
motor area and area in voluntary
H.O., potentials from
Burgess,
R.C. and Shibasaki,
associated human
ceph. clin. Neurophysiol.,
with
single
supplementary
89: 269-277.
H. (1993)
Move-
and repetitive
move-
motor
in
monkey.
Exp.
D.S.
and Hahn,
J.H.
by stimulation
D.S.,
Morris,
A negative
H.H.,
motor
of the human frontal E. Halgren
H. and Shibasaki, from
Wyllie,
E.,
response elicited
cortex.
and J. Bancaud
57. Raven Press, New York,
potentials
of the human
I.
Dinner,
In: P. Chauvel, (Eds.),
Advances
pp. 149-157.
H. (1988)
Recording
scalp and cortex
area. Electroen-
Brain.
Little,
Brown
in man.
of moveBrain,
and the Functional
and Company,
H., Von Giesen, H.J. and Benecke, inhibition
R. (1993)
seen after transcranial
lation in awake human subjects. Exp. Brain Sasaki, K., Gemba, H., Hashimoto,
I I I:
MA.
On the origin
of the
brain stimu-
Res., 94: 489-498. N. (1979)
on slow potentials
hand movements
Anatomy
Boston,
magnetic
S. and Mizuno,
hemispherectomy
preceding self-paced
Influences
in the motor cottex
in the monkey.
Neurosci.
Lett.,
15: 23-28. H.,
Barrett,
G.,
Halliday,
E.
Components
of the movement-related
topography.
Electroenceph.
Shibasaki, Kimura.
M., Takano,
J. (1994)
and
Halliday,
cortical
potential
clin. Neurophysiol.,
H., Ikeda, A., Nagamine,
T., Mima,
S., Hanazono,
Cortical
reflex
A.M.
49: 213-226.
T., Terada,
T., Kohara,
negative
(1980)
and their scalp K., Nishitani,
N.. Kaji,
myoclonus.
R. and
Brain,
117:
477-486. I.M.
and Hallett,
potentials physiol.,
M. (1990) Topography
in human
finger
motor
movements.
of scalp-recorded
Electroenceph.
clin.
motor Neuro-
75: 36-43.
I.M. and Hallett, potentials
M. (1991)
preceding
nant and non-dominant T., Gemba,
Cortical
self-paced
topography
voluntary K. (1993)
of premotor
movement
hands. J. Clin. Neurophysiol.,
H. and Sasaki,
dentate nucleus of the cerebellum
Effect
on hand movement
and
of domi-
8: 331-341. of cooling
the
of the monkey.
Brain Res., 629: I-9. Ugawa,
Y., Shimpo,
asterixis:
115: 1017-1043.
ment-related
recorded
R.C. and Shibasaki,
area: role of supplementary
Ikeda, A., Liiders, ments
Roick,
Tsujimoto,
43: 1078-1083. Ikeda, A., Liiders,
Dinner,
W. and Jasper, H. (1954) Epilepsy
Tarkka,
81: 257-262. P., Bureau,
Vol.
R., Liiders,
Penfield,
Tarkka,
R. and
during
troenceph.
related to the beginning
stimulation
ment-related
N., Kanda,
820.
Fuhr,
Neshige,
Shibasaki,
S. (1985)
stimulation
in the conscious
H.H.,
16: 229-23
R.P.,
Delgado-Escueta,
of cerebellar
muscles to transcra-
58: 1168-l
Sawyer-Palmer,
sites in the awake
C.D.
simultaneous
62: 347-352.
and extensor
Morris,
J. and Hahn, J.H. (1992)
of the Human
and Marsden,
in
U. and Hagbarth,
J. Neurophysiol., E.E.
and suppression
motor cortex
(1976)
(Ed%), The Responsive
Bereitschaftspotential
(1987) The effect upon the
719-736
63: 327-339. H.H.
A.V.
R.P.,
Lesser,
postexcitatory
M., Wallin,
stimuli.
P.D.,
potentials
neurons
motor responses elicited
Adv. Epileptol.,
H.O.,
in Neurology,
slow and rapid hand movement
and J.R. Knotts
movements.
nial cortical
60: 276-281.
Cortical
R. and Komhuber,
voluntary
unit responses in human wrist flexor Cheney,
cortex. Liiders,
by electrical
re-
for three periods of prepara-
Neurophysiol.,
J.P.R., Rothwell,
Increase
sequential
R. (1986)
evidence
with
pp. 99-102.
R., Dick,
(198.5)
potentials
clin. Neurophysiol.,
clin.
K., Jurgens,
itschaftspotential Wright,
Move.
A computer-assisted
cortical
H. and Neshige,
voluntary
tion in man. Electroenceph. Becker.
R. (1985)
movement-related
Shibasaki,
preceding
C. and
asterixis.
Godoy,
spect to EMG onset. Electroenceph. Barrett,
de Casasola,
mediating
16.
G., Shibasaki,
method
R., Garcia
mechanisms
G.W.H.
muscles of intra-cortical
Res.. 66: 621-637.
Liiders, Artieda,
R.B. and Mantel,
of hand and forearm
chiat., Ugawa, offset
T. and Mannen,
silent period
locked
T. (1989)
avenging.
Physiological
J. Neural.
analysis
Neurosurg.
of
Psy-
52: 89-93. Y., Genba,
K., Shimpo,
of electromyographic
Neurosurg.
Psychiat.,
T. and Mannen. (EMG)
53: 260-262.
silence
T. (1990) in asterixis.
Onset and J. Neural.
K. Terndu et al. / Electrr,encephulography
Uncini,
A.,
(1993)
Treviso,
M.,
Physiological
transcranial
cortical
Di
Muzio,
A.,
basis of voluntary stimulation.
Simone, activity
Electroenceph.
P. and Pullman, inhibition clin.
and clinical
S.
induced by
Neurophysiol.,
89: 2 I I-220. Vaughan.
H.G.,
L.G. and Hallett,
Costa,
L.D.
E.M.,
and Ritter,
Electroenceph.
Pascual-Leone,
W. (1968)
Topography
clin. Neurophysiol.,
A., Valls-Sole,
.I., Toro,
of the
25: I-IO. C., Cohen,
95 (1995) 335-345
M. (1993)
tory responses to transcranial Electroenceph. Wilson,
human motor potential. Wassermann.
Neurophysiolo~y
magnetic 52-56.
magnetic
clin. Neurophysiol..
S.A., Thickbroom,
of excitatory
Topography
of the inhibitory stimulation
and excita-
in a hand muscle.
89: 424-433.
G.W. and Mastaglia,
and inhibitory
stimulation
34s
F.L. (199.3) Topography
muscle responses evoked
in the human motor cortex.
by transcranial
Neurosci.
Lett.,
154:
Electroencephalography
and clinical Neurophysiology 9.5 (1995) 335-345
Movement-related cortical potentials associated with voluntary muscle relaxation * K. Terada *, A. Ikeda, T. Nagamine, H. Shibasaki Depment of Brain
P~tt/zc)physiolo~y, Kyoto
School of’Medicine. Sho~oinknwukurumnclzi
Unil;ersity
54, Sakyo-ku.
Kyoto 606.01,
Jqmn
Accepted for publication: 8 May 1995
Abstract
We recorded movement-related cortical potentials (MRCPS) associated with voluntary muscle relaxation, which was not accompanied by contraction of the antagonist or any other muscles, in 10 normal subjects. Voluntary, self-paced relaxation of the wrist extensors from the extended position was employed as the relaxation task, and wrist flexion by muscle contraction was employed as the contraction task. The accelerogram was used to obtain the trigger signals for both tasks. The electromyograms of the ipsilateral agonist and antagonist, the proximal muscles and the contralateral corresponding muscles were monitored to confirm the absence of muscle contraction for the relaxation task. All MRCP components were identified in both tasks; Bereitschaftspotential (BP), negative slope (NS’), parietal peak of motor potential (ppMP) and frontal peak of motor potential (fpMP). BP started earlier and was larger at the contralateral parietotemporal electrodes for the relaxation than for the contraction task, and the slow positive shift at the bilateral frontopolar electrodes was seen more often in the relaxation task. It is concluded that the voluntary muscle relaxation needs a cortical preparatory process similar to voluntary muscle contraction, and needs a more extensive and longer preparation process in the primary motor area and possibly in other motor areas as compared with the contraction. Kry\~or&:
Movement-related
cortical potentials; Voluntary
muscle relaxation; Negative motor phenomenon; Motor control
1. Introduction Movement-related cortical potentials (MRCPS) are defined as brain potentials associated with self-paced, voluntary movements (Kornhuber and Deecke, 1965; Vaughan et al., 1968; Shibasaki et al., 1980). MRCPs are of great interest because they are expected to help elucidating the neuronal mechanisms participating in the movement preparation, initiation, execution and feedback control (Benecke et al., 1985; Barrett et al., 1985, 1986; Tarkka and Hallett, 1990, 199 1). Currently it is accepted that pre-movement components of MRCPs consist of at least two subcomponents: the earlier slow negativity, Bereitschaftspotential (BP) (Kornhuber and Deecke, 1965) and the late negative
” This study was supported by Grant-in-Aid 0640403
I,
Grant-in-Aid
Grant-in-Aid
for
Priority
New Program 06NP0101,
search Grant 07044258
Scientific
for Scientific Research
and International
Research 06260225.
ReScience
Scientific
from the Japan Ministry of Education.
and Culture for H.S.
’ Corresponding 0013-4694/95/$09.50
author. Tel.: 8 I-75-751 -3603; Fax: 8 l-75-751-3202.
0
SSDI 0013-4694(95)00098-4
slope (NS’) (Shibasaki et al., 1980). These potentials are most likely to represent the function of the primary and supplementary motor areas, as shown by subdural recording in patients with intractable epilepsy (Lee et al., 1986; Neshige et al., 1988; Ikeda et al., 1992, 1993). In all of the previous studies on MRCPs, those potentials were recognized in association with the initiation of self-paced voluntary muscle contraction, i.e., “positive motor phenomenon,” and no studies have been reported in association with voluntary, pure relaxation of the ongoing muscle contraction, i.e., “negative motor phenomenon.” Although Dimitrov (1985) reported the presence of BP associated with the termination of the voluntary finger movement, he did not monitor the antagonist muscles and concluded that the BP was related to contraction of the antagonists rather than to pure relaxation. Recently, negative motor phenomena have attracted the attention of many investigators: negative myoclonus (Ugawa et al., 1989, 1990; Artieda et al., 1992; Shibasaki et al., 1994), negative motor response accompanying the motor evoked potential produced by transcranial magnetic stimulation of the motor cortex (Calancie et al., 1987; Fuhr et al., 199 1; Roick et
1995 Elsevier Science Ireland Ltd. All rights reserved EEG 94 I93
336
K. Teruda et ul. / EIectroencephaloRruphy
und clinical
Neurophysiology
95 (I 995) 335-345
al., 1993; Uncini et al., 1993; Wassermann et al., 1993; Wilson et al., 1993) and negative motor response to direct cortical electrical stimulation (Penfield and Jasper, 1954; Liiders et al., 1987, 1992). Since the mechanisms underlying voluntary negative motor phenomena are still unclear, we analyzed MRCPs associated with voluntary muscle rel;vation which was carefully controlled to avoid associated contractions of other muscles, especially those of the antagonists.
board
Paper
Fig. 1. Position of the subject’s right hand before each relaxation movement. Four digits were attached to a light paper plate on the dorsum side
2. Methods
and were wrapped together by an adhesive tape. The subject kept the
2.1. Subjects
onset, an accelerometer was attached to the distal tip of the plate.
right wrist in extended position to 150-160”.
Ten healthy volunteers (8 males and 2 females), aged 25-45 years (mean 30.3), served as subjects for the present study. Nine of them were right-handed and one was left-handed. None of them had a previous history of neurological illness. The purpose of the experiment was explained to each subject beforehand, and an informed consent was obtained. 2.2. Movement
paradigm
The subject was seated in an arm-chair in a quiet room with eyes kept open and fixating forward at the target in order to eliminate blinks and slow eye movements. Each subject performed two kinds of tasks both with the right hand, i.e., “relaxation task” and “contraction task” as described below. Each recording session consisted of 50 trials of the same task, and each subject had 3 sessions for the relaxation and contraction tasks which were performed alternately. For all subjects, the initial session was the relaxation task because a longer training period was needed for the relaxation than for the contraction task.
Contraction task The forearm was placed perpendicular to the paper board, ulnar side down, and the subject was instructed to keep the wrist in complete relaxation for more than 5 sec. The relaxation was confirmed by monitoring the polygraphic surface EMG records. Then the subject flexed the wrist voluntarily by contracting the wrist flexor muscle as quickly as possible. The subject was trained to make the wrist flexion to 150-160”. After having maintained the flexed position for more than 5 set, the subject returned the wrist to the relaxed, neutral position for the next trial. The rate of contraction movement was the same as that for the relaxation task. The performance was also monitored by the video camera system as described above. 2.3. Data acquisition For recording the electroencephalogram (EEG), 2 1 Ag/AgCl cup electrodes (FPl, FP2, F7, F3, Fz, F4, F8, Voluntary
Relaxation task The forearm was positioned, palm down, on a paper board placed on the arm-rest. Before each relaxation movement, the subject kept the right wrist in extension to 150-160” for at least 5 set (Fig. 1). The subject was trained so that the electromyographic (EMG) activity was restricted only to the extensor carpi radialis muscle (ECR) throughout the extended position (Fig. 2). Then, the subject relaxed the ECR muscle by terminating the contraction voluntarily (without any cue signal) and as quickly as possible. During the training, great caution was given by monitoring the polygraphic EMG as described below, so that the hand could drop without any associated contraction of other muscles. Once the complete relaxation was achieved, the subject kept the EMG silent for more than 5 set and then resumed the extended position of the wrist for the next trial. The rate of relaxation movement was once per 15 set on average for every subject. The performance was monitored with the aid of a video camera system throughout the experiment.
To detect the movement
4
Wrist
in
extension
relaxation * ‘I
rt. DELT rt. BR rt. ECR rt. FCU
-i---
rt. BIG rt. TRIG
-A-
It.ECR ----.
It.FCU
a-_l__._-._ -- ~- ‘1
?-
__--._._.-_ T
.____
,/=
I-
._
A
_ 1OO~b I see
Tr iggsr
Fig. 2. A sample of relaxation trials showing typical, acceptable EMG activities.
DELT
= deltoid. BR = brachioradialis,
ECR = extensor carpi
radialis, FCU = flexor carpi ulnaris, BIC = biceps brachii, TRIG = triceps brachii muscle, rt. = right, It. = left, Accel. = accelerogram. EMG
Note that the
activity is completely absent during the relaxation not only in the
ECR but also in other muscles of both upper extremities.
K. Terudu et al. / ElectroencephaloKraph~
T3, C3, C 1, Cz, C2, C4, T4, T5, P3, Pz, P4, T6, 01 and 02) (American EEG Society, 1992) were affixed on the scalp with collodion, and the impedance of all electrodes was kept less than 5 k 0. All electrodes were referenced to linked earlobe electrodes. The time constant (TC) was set to 3 sec. and the high frequency filter (HFF) used was 100 Hz. A 60 Hz notch filter was applied for all channels. EOG was monitored by an Ag/AgCl cup electrode placed at the right inferior lateral canthus referenced to FPz. TC and HFF for the EOG recording were set to the same values as for the EEG recording. In order to monitor the contraction and relaxation of the corresponding muscle and to detect concomitant contraction of other muscles (see above), EMGs were recorded by a pair of Ag/AgCl cup electrodes placed 3 cm apart on the skin overlying each of the right ECR, flexor carpi ulnaris (FCU), deltoid (DELT), biceps brachii (BIG), triceps brachii (TRIC) and brachioradialis (BR) muscles, and each of the left ECR and FCU muscles (Fig. 2). TC for the EMG recording was set to 0.075 set, and HFF to 100 Hz. EMGs were not rectified due to the limited capacity of the instrument, but the waveform of the averaged EMG was not important in the present study because the accelerogram was used to detect the movement onset (see below). To determine the movement onset, an accelerometer (AS-2GA, Kyowa Electronic Instruments) was fixed to the light paper plate which was bound on the dorsal aspect of the digits II-V by tapes (Fig. 1). Low frequency filter for the accelerometer was set to 0 Hz (DC) and HFF to 100 Hz. The accelerogram was rectified and used to obtain the fiducial point for averaging the EEGs and EOGs as described below. All segments of the amplified signals covering 2.4 set before and 1.2 set after the movement onset were stored on a computer (DPl 100, NEC-Sanei) at a sampling rate of 333 Hz for each channel for the subsequent off-line analysis. 2.4. Duta analysis After the completion of all sessions for each subject, the EEGs, EOG, EMGs and the accelerogram for each trial were displayed on the computer screen, and the precise onset of the rectified accelerogram was visually determined which was then used as the fiducial point for back-averaging the EEGs and EOG (Barrett et al., 1985). If there were any artifacts arising from eye movements, blinks or other sources, or if EMG activities were recognized in inappropriate muscles, that trial was eliminated from the subsequent analysis. After confirming the reproducibility of the averaged waveforms among different sessions of the same movement task, a group average for each task was obtained for each subject. Then, a grand average waveform for each task was obtained by averaging the group average data of like sessions across all the subjects. Although one of the
and clinical
Neurophy.Golog
95 Cl9951 335-345
337
amplitude
anpl i tude
CZbt?\
amp1itude
baseline
ppMP latency A =I ppMP amplitude
P3 \ -2sec
-1sec
movement onset
+
Fig. 3. Schematic diagrams illustrating BP (Bereitschaftspotential),
nega-
tive slope (NS’), parietal peak of motor potential (ppMP) and frontal peak of motor potential (fpMP),
and the measurements of amplitude, duration
and time interval of those MRCP components. The upward deflection for EEG data indicates the negativity at the exploring electrodes in this and all other figures.
male subjects was left-handed, his data were included because his group average waveforms and scalp distribution appeared similar to those of other subjects. Measurements of various time parameters with respect to the movement onset and amplitudes of identifiable components were computer-assisted. Amplitude measurements were based on the calibration data collected just before each experiment. Baseline was det.ermined by averaging the segment from 2.1 set to 1.6 set before the movement onset for each channel. The onset of BP and NS’ was visually determined and measured according to the method described previously (Kitamura et al., 1993a,b) (Fig. 3). The onset of BP was defined as the time when the baseline began deflecting toward negativity at electrode Cz, and its largest negative peak was defined as the BP peak. The BP duration was measured from its onset to peak. The amplitude of the BP peak at each electrode was measured from the above-described baseline. Likewise, the onset of NS’ was visually determined as the time when the following steeper slope started at Cl, and its largest peak was defined as the NS’ peak, regardless of whether it occurred before or after the movement onset. Consequently, the BP peak and the NS’ onset did not necessarily coincide. The NS’ duration was measured from its onset to peak. The amplitude of the NS’ peak was measured as the amplitude difference from its onset at each electrode. Although another negative slope corresponding to the intermediate shift (IS) (Barrett et al., 1986) was identified between BP and NS’ in the grand
338
K. Terada et al. / Elecfroencephalography
average waveform of the contraction data, it was identified in only 3 of the 10 individual data. Therefore, we dealt with that slope as a part of BP in this study. The parietal peak of motor potential (ppMP) and the frontal peak of motor potential (fpMP) were determined and measured using the definition proposed by Tarkka and Hallett (1990, 1991) (Fig. 3). The ppMP was defined as the first turn toward positivity after the movement onset at P3 electrode. The ppMP amplitude was measured from the immediately preceding positive peak (premotion positivity). The fpMP was defined as the largest negativity of the entire analysis window at Fz electrode. The fpMP amplitude was measured as the amplitude difference from the value at the movement onset. Differences in the time parameters of each MRCP component between the two tasks were statistically analyzed by the Wilcoxon signed-rank test. Differences in the amplitudes of each MRCP component between the two tasks were statistically analyzed by an analysis of variance for repeated measures (ANOVA) with factors of “task” and “electrode position” followed by the assessment using the conservative correction of Greenhouse and Geisser
(a>
Acce I Tri&er (b) rt.ECR
rt. BIC ...-~--__.____ TR ,c
Neurophysiology
_.__----______._____.___+
Accel. ___~
_.__.
_._,c i
__-_
._~. ._
+_.__ Trigger
.---
95 (1995) 335-345
which is designed to avoid non-replicable positive results by restricting the theoretically appropriate degrees of freedom. Furthermore, in order to evaluate the statistical significance of the possible asymmetry of each potential over the scalp for each task, the amplitudes between each pair of homologous electrodes, i.e., FPl and FP2, were also assessed using the conservative correction of Greenhouse and Geisser for the both relaxation and contraction tasks. The amplitudes of BP, NS’, ppMP and fpMP measured in the grand average waveforms were displayed as topographic maps by using Nicolet Pathfinder TMAP software, which used a triangular interpolation matrix. Calibration of the color scale was set to the maximal value of the absolute amplitude in the both relaxation and contraction tasks for each component.
3. Results The number of trials accepted for the analysis per subject was 20-106 (mean 78.3) for the relaxation task and 51-122 (mean 89.3) for the contraction task. Fig. 2 demonstrates a trial for the relaxation task showing well controlled EMG activities. We accepted only those trials in which EMG activities of the right ECR muscle completely disappeared just before or after the movement onset and in which no EMG activity was observed in any of the FCU and other muscles on either side throughout the trial. We rejected the trials in which inappropriate EMG activities, especially those in FCU, were observed before or after the movement onset (Fig. 4b). Fig. 4a shows an acceptable sample in which a phasic activity in the ECR muscle was observed 130 msec after the movement onset. This trial was accepted because the phasic EMG activity was interpreted to be a stretch reflex. As the result of averaging those trials containing the stretch reflex and also due to variable patterns of EMG offset, the activity of ECR in the averaged waveform demonstrated a rather gradual decrease in activity instead of abrupt cessation (Fig. 5a compared to Fig. 5b). 3.1. MRCPs relaxation
rt.FCU
rt,
and clinical
associated
with ooluntary
self-paced
muscle
--
. ..__._
J-w 5w mm+
Fig. 4. Two samples of EMG activities during the relaxation task which were accepted for (a) and rejected from (b) the analysis, respectively. a: a brief EMG activity is observed in the ECR muscle 130 mse.c after the onset of relaxation, which is most likely a stretch reflex. b: an inappropriate EMG activity is observed in the FCU muscle just after the fiducial point, suggesting a purposeful contraction of the wrist flexor muscle.
Fig. 5a demonstrates the grand average waveforms of MRCPs for the relaxation task across all the subjects. A slow negative shift, which corresponds to BP for the contraction task, started at 1600 msec before the movement onset and was maximal at the vertex (Cz) and the left central area (Cl). The mean onset time of BP in 10 subjects was 1521 msec before the movement onset (Table 1). At 220 msec before the movement onset (mean value in 10 subjects 275 msec; Table l>, the slow negative potential became steeper in the left hemisphere, maximal at the left central area (Cl). This steeper negative shift corresponds to NS’ for the contraction task. The small ppMP was
K. Terada et al. / Electroencephalography
md clinical
Neurophysiology
95 (1995) 335-345
339
n FE F3
T3
F4
T4
c3
T5
c4
T6
Cl
Fz
C2
cz
P3
PZ
P4 rt.ECR
01
rt.FCU
02
rt.BIC
EOG
rt.TRIC ACC IEEG, EOG / lO/.iV -2 set
lb)
-1
trigger
n
FP2
F8
F3
T3
F4
T4
c3
T5
c4
T6
Cl
Fz
c2
cz
P3
PZ
/-
20,uv
ACC j
0.25
G
I
I+
‘+
FPl
EMI
P4 01
rt.ECR
02
rt.FCU
EOG
rt.BIC rt.TRIC/------------ y--b"++-__. _._
.-. -2 set
trigger
EEG, EOG -10/1V +
EMG
2O/.tV
__~ ACC
0.25
G
+
Fig. 5. Grand average waveforms of cortical potentials across 10 normal subjects, associated with voluntary, self-paEd relaxation of the right ECR muscle (a) and voluntary, self-paced contraction of the right FCU muscle (b), both resulting in flexion movement of the right wrist (average of 783 and 893 trials, respectively). ACC: averaged, rectified accelerogram. In (a), BP struts at 1600 msec before the movement onset, is maximal at Cz and Cl, and becomes steeper at 220 msec before the movement onset on the left hemisphere with the maximum at Cl (NS’). At the bilateral frontopolar areas @PI and FP2), a small slow positive shift can be recognized. After the movement onset, a small negative peak is identified at 90 msec at the contralateral centroparietal electrodes (ppMP) and another negative peak at 210 msec at the frontocentral electrodes (fpMP). In (b), BP sttis at 1500 msec before the movement onset, is maximal equally at Cl and Cz, and becomes steeper at 210 msec before the movement onset on the left hemisphere with the maximum at Cl (NS’). After the movement onset, a small negative peak is identified at 24 msec at the contralateral centroparietal electrodes (ppMP) and another negative peak at I IO msec at the frontocentral electrodes (fpMP).
340
K. Terada et al. / Electroencephdography
identified at 90 msec after the movement onset at the contralateral centroparietal electrodes. The fpMP was identified at 210 msec after the movement onset maximally at the midline frontocentral electrodes (Fz and Cz). At the
and clinical Neurophysiology 95 (1995) 335-34.5
.bilateral started topolar The
frontopolar regions (FPl and FP2), a positive shift 1500 msec before the movement onset. This fronpositivity was seen in 6 out of the 10 subjects. topographic map of BP for the relaxation task
Fig 6. Topographic maps of BP, NS’, ppMP and fpMP obtained from the grand average waveforms for the relaxation (a) and the contraction (b) tasks BP disctribution is clearly different between the two tasks, i.e., the negative field at the parietal area is more prominent for the relaxation task as compare .d to the contraction task, and the frontal positivity is seen only for the relaxation task. Other components show similar distribution in the two tasks.
K. Terada et al. / Electroencephalography
and clinical
Table I Temporal parameters of each recognizable component of MRCPs associated with voluntary, self-paced relaxation and contraction tasks with respect to the movement onset in 10 normal subjects (msec, mean k SD.) Relaxation
Contraction
- 152I .O+ 233.6 -1216.8k334.7 BP onset time * 811.5k320.3 1160.1 f 177.5 BP duration * - 274.8 + 143.9 -235.2 f 142.8 NS’onset time * * 213.8f 138.2 242.4 k 149.5 NS’duration * * t 52.5+ 30.1 + 43.8 &- 24.9 ppMP peak time * * * t 150.6f 46.7 foMP peak time * * * ’ t 229.8 f 72.6
P
CO.05 <0.05 ns. n.s. n.s. < 0.05
“-” and “t” indicate “before” and “after” the movement onset, respectively. ns. = not significant in this and all other tables. measuredatCz*,CI**,P3”*,andFz*‘*’.
a relatively symmetric distribution of the negativity at the centroparietal areas with the maximum at Cz and Cl (Fig. 6a). The associated positive field was localized symmetrically at the frontopolar midline area. The NS’ was distributed at the left frontal, central and parietal areas with the maximum at Cl. The ppMP appeared as a negative field at the left central area, and the fpMP was seen as a relatively large negative field at the midline frontocentral area and distributed symmetrically.
demonstrated
Neurophysiology
95 (1995) 335-345
341
3.2. MRCPs associated with voluntary self-paced muscle contraction
Fig. 5b demonstrates the grand average waveforms of MRCPs for the contraction task across all the subjects. BP started at 1500 msec before the movement onset and was maximal at the vertex (Cz) and the left central area (Cl) with equal amplitude. The mean onset time of BP in IO subjects was 1217 msec before the movement onset (Table 1). NS’ started at 210 msec before the movement onset (mean value in 10 subjects 235 msec; Table I) and was maximal at the left central area (Cl). The small ppMP was seen at the centroparietal area at 24 msec after the movement onset. The fpMP was clearly observed at 110 msec after the movement onset at the frontocentrai electrodes with the maximum at Fz. Unlike the relaxation task, there was no slow positive shift seen at the frontal area in the grand average waveform. In the individual data, however, 3 subjects demonstrated a small positive shift at the frontopolar area, who had the positive shift also in the relaxation task. The topographic map of BP for the contraction task demonstrated a localized negative field at the midline central area with the maximal amplitude at Cl (Fig. 6b). The NS’ was maximal at Cl and distributed at the cen-
Table 2 Amplitudes of BP and NS’ of MRCPs associated with voluntary, self-paced relaxation and contraction tasks in 10 normal subjects, measured at each electrode ( pV, mean f S.D.) Electrode
FPI FP2 F3 F4 c3 c4 Cl c2 P3 P4 01 02 Fl F8 T3 T4 TS T6 FZ CZ PZ
NS’b
BP = Relaxation
Contraction
PC
Relaxation
Contraction
PC
0.69 + I .78 1.01 f 1.35 - 0.84 + 0.96 - 0.77 f I .oo -2.21* 1.07 - I .73 + 0.85 -3.16+ 1.34 -2.41 kO.61 -2.11 f 1.73 - 1.s7+ 1.39 - 0.97 f 1.69 -0.87* 1.69 -0.19*0.19 0.21 +0.70 -0.86+ 1.13 - 0.39 + 0.87 -0.73+ 1.19 -0.82* 1.01 - 1.44+0.88 -3.71rto.53 - 2.43 f I .28
-0.55+2.50 -0.58 + 2.70 -1.31*1.41 -1.11~1.30 -2.04f 1.31 - I .60 + 0.89 -2.62& 1.06 - I .98 + 0.97 - 1.19+ 1.32 - I .42 f 1.24 - 0.42 f 1.67 - 0.73 f 1.65 -0.19* 1.32 - 0.68 f I .22 0.18& 1.39 -0.53+ 1.00 -0.13f 1.26 -o.s2+ 1.21 - 1.50* 1.51 -2.71 + 1.12 -1.25dzl.27
< 0.05 < 0.05 n.s. n.s. n.s. ns. ns. ns. n.s. n.s. n.s. n.s. n.s. n.s. < 0.05 n.s. ns. n.s. ns. n.s. < 0.05
-0.98+2.18 - 1.30+ 1.50 - 3.07 f 1.50 - 2.43 + 1.20 -4.43+1.24+ - 2.62 f I .3 I + -4.87+ 1.53 + - 3.67 f I .45 + - 3.25 + I .29 - 2.57 + I .67 - 1.42f 1.63 - 1.54* 1.52 - 1.42f 1.10 - 1.19+ 1.05 - 1.52zb2.06 -0.81 *0.71 - 1.24f 1.00 -0.71 + 1.27 - 3.32 L- I .61 - 4.67 f 1.70 - 3.48 f 1.62
-0.81 f I.51 - 0.66 f I .93 - 1.90* 1.60 - 1.83* 1.88 -3.18* 1.83 + - I.955 1.44+ -3.62+2.25 - 2.75 f I .83 - 2.94 * 2.04 - 2.40 f I.56 - l.84* 1.63 - 1.72+ 1.41 -0.95* 1.00 - 0.48 * I .45 - 1.2 I + I .06 - 1.06+ 1.65 - 1.52* 1.55 - I .27 f 0.86 -2.501 1.87 -3.39+2.19 - 2.87 f I .92
ns. ns. n.s. ns. < 0.05 n.s. < 0.05 n.s. n.s. n.s. n.s. ns. ns. n.s. n.s. n.s. ns. ns. n.s. < 0.05 n.s.
a Measured from the basehne. h Measured as the amplitude difference from its onset. ’ Comparison of relaxation vs. contraction at each electrode. ‘*- ” indicates negativity at each exploring electrode with respect to the corresponding reference level. ./_P < 0.05.
342
K. Terada et al. / Electroencephalography and clinical Neurophysiology 95 (1995) 335-345
troparietal area, extending more to the left hemisphere. The ppMP was localized at the left central area. The fpMP was seen as a negative field at the frontocentral area, being distributed more predominantly to the left. 3.3. Comparison traction tasks
of
A4RCPs between
relaxation
and con-
Comparing the amplitudes between tasks and among electrode positions, the ANOVA demonstrated a signifi(P < 0.05) but no cant effect of “electrode position” significant effect of “task” for all MRCP components. A significant interaction between “task” and “electrode position” (P < 0.05) factors was seen only for the BP. The BP associated with the relaxation task started earlier and its duration was longer as compared with the contraction task (P < 0.05; Table 1). The BP amplitude at T3 and Pz was significantly larger in the relaxation task than in the contraction task (P < 0.05>, and the statistically significant positive activity at FPl and FP2 was observed for the relaxation task (Table 2 and Fig. 6). As for the NS’, there was no statistically significant difference in onset time and duration between the two tasks (Table 1). The amplitude of NS’ at C3 and Cl was significantly larger in the relaxation task than in the contraction task (P < 0.05; Table 2). With regard to the ppMP, there was no statistically significant difference in its time parameters between
Table 3 Amplitudes of ppMP and fpMP of MRCPs associated electrode ( WV, mean f S.D.) Electrode
FPI FP2 F3 F4 c3 c4 Cl c2 P3 P4 01 02 F7 F8 T3 T4 T5 T6 FZ CZ Pz
with voluntary,
self-paced
the two tasks (Table 1). However, the ppMP amplitude at Pz was significantly larger for the relaxation than for the contraction task (P < 0.05; Table 3). The peak time of fpMP associated with the relaxation task was longer than that with the contraction (P < 0.05; Table 1). No statistically significant difference was seen for the amplitude of fpMP between the two tasks (Table 3). As the result of comparison between the two sides, there was no statistically significant difference in the amplitude of BP, ppMP and fpMP between either pairs of the homologous electrodes for both tasks (Tables 2 and 3). The amplitude of NS’ was larger at C3 and Cl than at C4 and C2, respectively, for the relaxation task (P < 0.05) and was larger at C3 than at C4 for the contraction task (P < 0.05; Table 2).
4. Discussion 4.1. Comparison
between relaxation and contraction
tasks
In the present study, the slow negative potentials corresponding to BP and NS’ were successfully recorded for the first time prior to voluntary muscle relaxation which was confirmed by the EMG silence of antagonist and other proximal muscles. Thus the present findings suggest that, at least from an electrophysiological point of view, volun-
relaxation
and contraction
tasks in IO normal subjects,
measured
fpMP h
ppMP a Relaxation
Contraction
PC
Relaxation
Contraction
PC
- 1.00+1.41 - 0.75 f I .64 -2.14+2.04 - 1.94+ 1.82 -2.45 f 1.54 -2.06f 1.67 -2.83+ 1.93 - 2.52 f I .89 -2.16? 1.43 -2.01 f 1.73 - I .09 * 1.02 -1.23+1.16 - 1.00+ 1.12 - 0.86 f 0.85 - I .43 + 1.23 - 0.68 + 0.9 I -0.78+0.91 - 0.55 + I .26 -2.08 ir 1.79 -2.88 +2.06 - 2.47 + I .62
- 1.60+ 1.16 - 1.46+ 1.43 - 1.79* 1.16 - I.175 1.08 -2.06& 1.03 - I .42 + 0.98 -2.07+0.83 - 1.62+0.82 - 1.26kO.87 - 1.10+0.82 -0.80+0.88 -0.53* 1.11 -0.69, 1.74 - 0.62 f 0.55 - 1.06+ 1.65 -0.51* 1.00 -0.66*0.81 0.27 f 0.9 1 - 1.74f 1.02 - 1.82 k 0.75 -0.35)0.69
ns. ns. n.s. n.s. n.s. n.s. ns. ns. n.s. n.s. ns. ns. n.s. ns. n.s. n.s. ns. ns. ns. n.s. < 0.05
- 2.40 f 1.79 -2.40+ 1.91 -3.59+ 1.39 -2.93+ 1.46 -2.68+ 1.56 -2.38, 1.54 -3.44+ 1.82 -2.97f 1.66 - 0.84 f 2.07 - 1.36* 2.04 0.26 f I .48 0.05~1.51 - 1.81 + 1.12 -2.14i0.76 - I .56 + 2.35 - I .68 k 0.88 0.49 * 1.07 -0.53+ 1.18 - 3.94 f I .23 - 3.55 & I .78 - I .46 + 2.17
-2.15+1.68 - 2.24 f I .28 - 2.98 k 2.07 - 2.34 * 1.87 -2.13+ 1.21 - I .95 * I .04 -3.14+ 1.62 - 2.56 + I .39 - 0.67 + I .03 - 1.47+ 1.18 -0.20* 1.41 -0.34+1.17 - 2.10 + 1.09 - I .29 + I .24 - 0.90 * I.59 -0.71 + 1.66 - 0.29 + I .08 - 0.40 * 2.00 - 3.45 + 1.77 -3.16+1.72 - 1.38* 1.07
n.s. n.s.
’ Measured from the immediately preceding positive peak. h Measured from the movement onset. ’ Comparison of relaxation vs. contraction at each electrode. “ - ” indicates negativity at each exploring electrode with respect to the corresponding
reference
point
“3.
ns. ns. us. us. n.s. n.s. “.S.
n.s. ns. “3.
n.s. n.s. ns. ns. n.s. n.s. n.s. ns.
at each
K. Terudu et al. / Electroencep~logrph!,
und clinical Neurophpiolo:v
tary muscle relaxation has similar preparatory mechanisms to voluntary contraction at the cortical level. Furthermore, the overall similarity of waveforms to those of BP and NS’ associated with voluntary muscle contraction suggests that the main cortical generators of those potentials might be common for both tasks, namely the primary and supplementary motor areas according to the results of subdural recording (Neshige et al., 1988; Ikeda et al., 1992, 1993). However. as the result of quantitative and topographical comparisons of MRCPs, some differences were observed between the two tasks. First, the onset of BP was earlier and its duration was longer for the relaxation than for the contraction task, suggesting that the voluntary muscle relaxation needs longer preparation than the voluntary contraction. However, it is also possible that the different movement velocities, slower in the relaxation task than in the contraction task, might have resulted in a different onset time (Becker et al., 1976). Secondly, the BP at the contralateral temporal and midline parietal areas (T3 and Pz) and the NS’ at the contralateral central area (Cl, C3 and Cz) were larger for the relaxation than for the contraction task, and the slow positive activity was observed at the ipsilateral frontal area (FPI and FP2) only in the relaxation task. This anterior slow positive shift, however, was also observed for the contraction task in 3 out of 10 subjects in this study, and was previously observed in association with the contraction task (Shibasaki et al., 1980), suggesting that this is not an absolute difference between the two tasks. Since the amplitudes of BP and NS’ were larger at the contralateral centroparietal area for the relaxation task than for the contraction task, it is possible that the primary motor area is more active in the preparation for the voluntary muscle relaxation than for the contraction task, adding the contralateral parietal negativity and the frontopolar positivity as the result of a tangentially oriented dipole sitting in the anterior bank of the central sulcus contralateral to the movement. However, the other possibility that two independent activities might be present: one in the midline frontal region and the other in the parietal region, more predominantly for the relaxation than for the contraction task cannot be excluded. With regard to the frontal positivity, its possibility being the contamination by eye movement artifacts was unlikely because the EOG recorded bipolarly with the same sensitivity as for EEGs did not show any corresponding activity. The ppMP is thought to correspond to the terminal phase of the motor cortex firing which is related to either the preparation or execution of the movement (Tarkka and Hallett, 1991). While the onset time of ppMP was not statistically different between the two tasks, its amplitude at Pz was larger for the relaxation task than for the contraction. It is therefore suggested again that the motor cortex might be more active also in the actual execution of the muscle relaxation than that of the muscle contraction. The fpMP was suggested to be generated in the supplementary motor area and to reflect the kinesthetic sensory
95 11995) 335-345
343
input from the peripheral nervous system (Tarkka and Hallett, 1991). A clear fpMP peaked at + I IO msec in the contraction task and at +210 msec in the relaxation task in the present study. The time difference of fpMP in the two conditions could be partially explained by the larger jitter of the speed in the relaxation task as demonstrated by the waveform of accelerogram (Fig. 5a and b). The distribution of BP was symmetric with the midline vertex maximum in previous studies (Shibasaki et al., 1980; Barrett et al., 1986). In the present study, BP appeared larger in the central area slightly contralateral to the movement in both tasks (Fig. 6a and b). However, there was no statistical significance between the two sides for both the relaxation and contraction tasks. The contralatera1 predominance of NS’ appeared more prominent for the contraction than for the relaxation task on the topographic maps (Fig. 6a and b), but the statistical difference in the transverse distribution between the two sides demonstrated more prominent asymmetry for the relaxation than for the contraction task (Table 2). It was suggested that the primary motor cortex was more activated for the relaxation than for the contraction task, even in the NS’ period.
4.2. Mechanism
of negative
motor
phenomena
Negative motor response, that was defined as the inability to perform voluntary movement or to sustain voluntary muscle contraction, was seen when a part of the premotor cortex just above the sylvian fissure or the anterior part of the supplementary motor area was electrically stimulated (Penfield and Jasper, 1954). Hence these areas were called “primary and supplementary negative motor areas,” rcspectively (Liiders et al., 1987, 1992). In the present study, since the positive field of BP for the relaxation task was located in the bilateral frontal area, it might be possible that the positive field was generated in the supplementary negative motor area. In experimental studies, it was reported that ablation of the cerebellum or cooling of the dentate nucleus diminished the surface-negative, depth-positive field potential recorded in the motor cortex of monkey preceding selfpaced movement (Sasaki et al., 1979; Tsujimoto et al., 1993). Those experimental data suggest that BP and NS’ represent the excitatory postsynaptic potentials (EPSPs) of the corticomotor neurons as a result of the facilitatory input through the cerebella-thalamo-cortical pathway. Furthermore, disynaptic inhibition of the spinal motor neurons mediated by group Ia interneurons and pure inhibition of motor unit discharge without additional increase of motor units were demonstrated by electrical stimulation of the corticomotor neurons in monkey (Jankowska et al., 1976). It was also suggested that some corticomotor neurons might have an inhibitory effect on spinal motor neurons: “inhibitory corticomotor neurons” (Cheney et al., 1985; Lemon et al., 1987; Wassermann et al., 1993).
344
K. Terada et al. / Electroencephulography
and clinical
In the present study, the waveforms of BP and NS’ for the relaxation task were, as a whole, similar to those for the contraction task with the maximum at the centroparieta1 area. Taken together with the above experimental data, it is suggested that BP and NS’ preceding voluntary, self-paced relaxation may represent EPSPs on the inhibitory corticomotor neurons. The evidence to support that positive and negative motor phenomena might occur through common mechanisms at the cortical level was also reported in patients suffering from cortical negative myoclonus (Guerrini et al., 1993; Shibasaki et al., 1994). The precise localization of inhibitory motor centers requires further studies by using other non-invasive studies such as magnetoencephalography and measurements of cerebral blood flow by positron emission tomography.
Neurophysiolo,gy
Jankowska,
95 (1995) 335-345
E., Padel, Y. and Tanaka,
R. (1976)
Disynaptic
inhibition
spinal motor neurons from the motor cortex in the monkey. (Land.),
258: 467-487.
Kitamura, J., Shibasaki, potential
H. and
Kondo,
T. (1993a)
is larger before an isolated movement
simultaneous physiol.,
movement
of two
fingers.
A
cortical
slow
of a single finger than
Electroenceph.
clin.
Neuro-
86: 252-258.
Kitamura, A.
of
J. Physiol.
J., Shibasaki,
(1993b)
sequential
A., Nabeshima,
negative
as compared
Electroenceph. Komhuber,
H., Takagi,
Enhanced
with
slope
and Deecke,
L.
H. and Yamaguchi,
cortical
simultaneous
clin. Neurophysiol.,
H.H.
of
potentials
extensions
before
of two fingers.
86: 176-182. (1965)
Himpotentiallnderungen
Willkiirbewegungen
und passiven
Bewegungen
reitschaftspotential
und reafferente
Potentiale.
bei
des Menschen: Pfliigers
Be-
Arch.,
284:
l-17. Lee, B.I., Liiders, Cortical
H., Lesser, R.P., Dinner,
potentials
recorded
from
D.S. and Morris,
related to voluntary
subdural
H.H. (1986)
and passive finger
electrodes
in
humans.
Ann.
movements Neural.,
20:
32-37. Lemon,
R.N., Muir,
activity
the vicinity Brain
References
of corticomotor
H., Lesser,
(1987) Negative J., Muruzabal,
J., Larumbe,
Obeso, J.A. (1992) Dis., 7: 209-2 Barrett,
Cortical
H. and Neshige,
for averaging
G.,
W., Iwase,
movement:
In: W.C.
preceding
McCallum Bristol,
Benecke,
Calancie,
of
B., Nordin,
the
Fetz,
facilitation
J.C., Day, B.L.
Neurosci.
Lett.,
BereBrain.
and
and
K.E. (1987)
of antagonistic monkey.
Motor
185.
forelimb
Patterns muscles
J. Neurophysiol.,
of
from
53: 805-
Dimitrov,
B. (1985)
Brain
potentials
termination
of voluntary
chophysiol.,
3: 13-22.
P.. Agostino,
excitability Guerrini,
flexion Hallett,
the silent
and Tassinari,
C., Genton,
C.A. (1993)
M.
period
clin. Neurophysiol.,
R., Dravet,
and extension (1991)
and to the
in man. Int. J. PsySpinal
after cortical
motor
stimulation.
neuron Elec-
Epileptic
M., Roger, J., Rubboli,
negative
myoclonus.
G.
Neurology,
H.O.,
ment-related
potentials
primary
motor
movements.
Burgess,
Brain,
recorded
from
supplementary motor
H. (1992)
Move-
motor area and area in voluntary
H.O., potentials from
Burgess,
R.C. and Shibasaki,
associated human
ceph. clin. Neurophysiol.,
with
single
supplementary
89: 269-277.
H. (1993)
Move-
and repetitive
move-
motor
in
monkey.
Exp.
D.S.
and Hahn,
J.H.
by stimulation
D.S.,
Morris,
A negative
H.H.,
motor
of the human frontal E. Halgren
H. and Shibasaki, from
Wyllie,
E.,
response elicited
cortex.
and J. Bancaud
57. Raven Press, New York,
potentials
of the human
I.
Dinner,
In: P. Chauvel, (Eds.),
Advances
pp. 149-157.
H. (1988)
Recording
scalp and cortex
area. Electroen-
Brain.
Little,
Brown
in man.
of moveBrain,
and the Functional
and Company,
H., Von Giesen, H.J. and Benecke, inhibition
R. (1993)
seen after transcranial
lation in awake human subjects. Exp. Brain Sasaki, K., Gemba, H., Hashimoto,
I I I:
MA.
On the origin
of the
brain stimu-
Res., 94: 489-498. N. (1979)
on slow potentials
hand movements
Anatomy
Boston,
magnetic
S. and Mizuno,
hemispherectomy
preceding self-paced
Influences
in the motor cottex
in the monkey.
Neurosci.
Lett.,
15: 23-28. H.,
Barrett,
G.,
Halliday,
E.
Components
of the movement-related
topography.
Electroenceph.
Shibasaki, Kimura.
M., Takano,
J. (1994)
and
Halliday,
cortical
potential
clin. Neurophysiol.,
H., Ikeda, A., Nagamine,
T., Mima,
S., Hanazono,
Cortical
reflex
A.M.
49: 213-226.
T., Terada,
T., Kohara,
negative
(1980)
and their scalp K., Nishitani,
N.. Kaji,
myoclonus.
R. and
Brain,
117:
477-486. I.M.
and Hallett,
potentials physiol.,
M. (1990) Topography
in human
finger
motor
movements.
of scalp-recorded
Electroenceph.
clin.
motor Neuro-
75: 36-43.
I.M. and Hallett, potentials
M. (1991)
preceding
nant and non-dominant T., Gemba,
Cortical
self-paced
topography
voluntary K. (1993)
of premotor
movement
hands. J. Clin. Neurophysiol.,
H. and Sasaki,
dentate nucleus of the cerebellum
Effect
on hand movement
and
of domi-
8: 331-341. of cooling
the
of the monkey.
Brain Res., 629: I-9. Ugawa,
Y., Shimpo,
asterixis:
115: 1017-1043.
ment-related
recorded
R.C. and Shibasaki,
area: role of supplementary
Ikeda, A., Liiders, ments
Roick,
Tsujimoto,
43: 1078-1083. Ikeda, A., Liiders,
Dinner,
W. and Jasper, H. (1954) Epilepsy
Tarkka,
81: 257-262. P., Bureau,
Vol.
R., Liiders,
Penfield,
Tarkka,
R. and
during
troenceph.
related to the beginning
stimulation
ment-related
N., Kanda,
820.
Fuhr,
Neshige,
Shibasaki,
S. (1985)
stimulation
in the conscious
H.H.,
16: 229-23
R.P.,
Delgado-Escueta,
of cerebellar
muscles to transcra-
58: 1168-l
Sawyer-Palmer,
sites in the awake
C.D.
simultaneous
62: 347-352.
and extensor
Morris,
J. and Hahn, J.H. (1992)
of the Human
and Marsden,
in
U. and Hagbarth,
J. Neurophysiol., E.E.
and suppression
motor cortex
(1976)
(Ed%), The Responsive
Bereitschaftspotential
(1987) The effect upon the
719-736
63: 327-339. H.H.
A.V.
R.P.,
Lesser,
postexcitatory
M., Wallin,
stimuli.
P.D.,
potentials
neurons
motor responses elicited
Adv. Epileptol.,
H.O.,
in Neurology,
slow and rapid hand movement
and J.R. Knotts
movements.
nial cortical
60: 276-281.
Cortical
R. and Komhuber,
voluntary
unit responses in human wrist flexor Cheney,
cortex. Liiders,
by electrical
re-
for three periods of prepara-
Neurophysiol.,
J.P.R., Rothwell,
Increase
sequential
R. (1986)
evidence
with
pp. 99-102.
R., Dick,
(198.5)
potentials
clin. Neurophysiol.,
clin.
K., Jurgens,
itschaftspotential Wright,
Move.
A computer-assisted
cortical
H. and Neshige,
voluntary
tion in man. Electroenceph. Becker.
R. (1985)
movement-related
Shibasaki,
preceding
C. and
asterixis.
Godoy,
spect to EMG onset. Electroenceph. Barrett,
de Casasola,
mediating
16.
G., Shibasaki,
method
R., Garcia
mechanisms
G.W.H.
muscles of intra-cortical
Res.. 66: 621-637.
Liiders, Artieda,
R.B. and Mantel,
of hand and forearm
chiat., Ugawa, offset
T. and Mannen,
silent period
locked
T. (1989)
avenging.
Physiological
J. Neural.
analysis
Neurosurg.
of
Psy-
52: 89-93. Y., Genba,
K., Shimpo,
of electromyographic
Neurosurg.
Psychiat.,
T. and Mannen. (EMG)
53: 260-262.
silence
T. (1990) in asterixis.
Onset and J. Neural.
K. Terndu et al. / Electrr,encephulography
Uncini,
A.,
(1993)
Treviso,
M.,
Physiological
transcranial
cortical
Di
Muzio,
A.,
basis of voluntary stimulation.
Simone, activity
Electroenceph.
P. and Pullman, inhibition clin.
and clinical
S.
induced by
Neurophysiol.,
89: 2 I I-220. Vaughan.
H.G.,
L.G. and Hallett,
Costa,
L.D.
E.M.,
and Ritter,
Electroenceph.
Pascual-Leone,
W. (1968)
Topography
clin. Neurophysiol.,
A., Valls-Sole,
.I., Toro,
of the
25: I-IO. C., Cohen,
95 (1995) 335-345
M. (1993)
tory responses to transcranial Electroenceph. Wilson,
human motor potential. Wassermann.
Neurophysiolo~y
magnetic 52-56.
magnetic
clin. Neurophysiol..
S.A., Thickbroom,
of excitatory
Topography
of the inhibitory stimulation
and excita-
in a hand muscle.
89: 424-433.
G.W. and Mastaglia,
and inhibitory
stimulation
34s
F.L. (199.3) Topography
muscle responses evoked
in the human motor cortex.
by transcranial
Neurosci.
Lett.,
154: