- Email: [email protected]
Noninvasive mapping of muscle representations in human motor cortex
Electroencephalography and clinical Neurophysiology, 85 (1992) 1-8 © 1992 Elsevier Scientific Publishers Ireland, Ltd. 0924-980X/92/$05.00
1
ELMOCO 91067
Noninvasive mapping of muscle representations in human motor cortex Eric M. Wassermann, Lisa M. McShane, Mark Hallett and Leonardo G. Cohen Human CorticalPhysiology Unit, Human Motor Control Section, Medical Neurology Branch, Office of the Clinical Director, and Biometry and Field Studies Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD (U.S.4.)
(Accepted for publication: 30 August 1991)
Summary We used transcraniai magnetic stimulation to map the cortical representations of 4 upper extremity muscles (abductor pollicis brevis, flexor carpi radialis, biceps, and deltoid) of 10 normal subjects. Three stimuli were delivered to scalp positions 1 cm apart, and the amplitude and latency of the motor evoked potentials (MEPs) were averaged for each position. Maps were described in terms of number of excitable scalp positions, amplitude of MEPs, scalp positions for evoking largest amplitude MEPs, and threshold for producing MEPs. We compared different muscles across subjects and the same muscles on the left and right sides in individual subjects. Distal muscles had larger representations with higher amplitude MEPs and lower thresholds. Biceps and deltoid on the left had larger representations and higher MEP amplitudes than on the right. Maps showed a somatotopic progression on the scalp of proximal to distal muscles along a posteromedial to anterolateral axis. Key words: Magnetic stimulation; Mapping; Human motor cortex; Motor representations
Until recently, stimulation of the h u m a n cerebral cortex was confined to the operating room where ex-
Subjects and methods
perimentation was subject to limitations of time and the extent of cortex that could be investigated in an individual subject (Penfield and Boldrey 1937; Woolsey et al. 1979). H u m a n studies have also b e e n restricted to subjects with disorders of the central nervous system, Modern cortical stimulation e x p e r i m e n t s in animals have focused on the details of organization of m o t o r
Subjects were 10 n o r m a l volunteers, including one of the authors (7 m e n and 3 women), aged 20-49 years. H a n d e d n e s s was assessed by a questionnaire (Oldfield 1971). All subjects gave written informed consent for the study, and the protocol was approved by the institute's clinical research review committee.
representation areas and pathways (Phillips and Porter 1977; Phillips 1987), but how well these findings apply to the h u m a n m o t o r system is unknown. T h e advent of transcranial stimulation has made it possible safely and painlessly to investigate the h u m a n motor cortex and its descending connections in normal subjects (Merton and Morton 1980; Barker et al. 1985). Electrical (Rossini et al. 1987, Cohen and Hallett 1988) and magnetic (Cohen et al. 1990a) transcranial stimulation have b e e n used to m a p the representations of muscles in normal humans. These studies, however, have used a small n u m b e r of stimulation sites. In the present study, we used an 8-shaped coil to deliver
The 10-20 international system for electrode placem e n t was used to locate Cz (a point on the midsagittal line, midway between the nasion and inion), and a grid of positions 1 cm apart was marked on the scalp with reference to Cz. Subjects lay supine on a comfortable bed in a quiet room and were instructed to keep their arms relaxed in extension and pronation at their sides. Subjects were kept alert by the investigator. Surface electrodes were applied to the skin over
multiple focal magnetic stimuli to closely spaced positions on the scalp to m a p and compare the representations of 4 u p p e r extremity muscles in normal subjects,
neously and printed on p a p e r with a conventional 4-channel E M G machine ( D I S A 1500 or Dantec Counterpoint) with filter settings of 100 Hz a n d 10 kHz. The maximal amplitude of the compound motor action potential (CMAP) to peripheral nerve stimulation was
Correspondence to: Dr. Eric M. Wassermann, Building 10, Room 5N226, NINDS, NIH, Bethesda, MD 20892 (U.S.A.).
determined for biceps and d e l t ~ with supramaximal electrical stimulation at Erb's point and for F C R and APB with median nerve stimulation at the elbow and wrist, respectively. Muscle relaxation was assessed by
2
E.M. WASSERMANN ET AL.
Anterior
-
-
,
~
~
~
,..,..,~
~
~
.
.
-.
~
-
.
.
,
_
_
~
~
,,,------
. _ - - . ~ . . . ~
~
~-,--"-"
CZ
_
L
.
Posterior
Left Hemisphere
Right Hemisphere
L 20 msec Fig. 1. Superimposed MEPs produced in right and left deltoid by 3 trials of magnetic stimulation at each scalp position in one subject. The locations of the EMG traces correspond to the scalp positions where they were evoked. Data from the contralateral muscle are shown on each hemisphere.
EMG monitoring before and several times during the experiment. Relaxation was defined as the absence of apparent EMG activity at a display gain of 100/~V/division, Magnetic stimulation was delivered with a Cadwell MES-10 magnetoelectric stimulator through an 8shaped coil whose characteristics are described elsewhere (Cohen et al. 1990b; Roth et al. 1991). The coil was positioned with t h e handle pointing backward in the sagittal axis and the center(intersection of the coil's wings) in contact with, and tangential to, the scalp at the point of stimulation,
Mapping Mapping experiments were divided into two sessions, one for each arm, 1-39 days apart, except for 3 subjects who had both arms tested on the same day. The experiment took 2-3 h for each arm. At 100% of stimulator output, 3 stimuli were delivered to positions 1 cm apart over the hemisphere contralateral to the muscles being recorded. Stimulation was started at different positions and continued in various directions in different subjects and sessions. Successive positions were stimulated until the area where MEPs were produced was surrounded by inactive positions. The triterion for the existence of a MEP was a deflection of 10 /zV or greater with an appropriate latency occurring on any of the 3 trials. The peak-to-peak amplitudes of the
3 MEPs produced in each muscle at each scalp position were averaged off-line and expressed as a percentage of the CMAP for that muscle (%M). Positions where no MEPs could be produced were assigned a value of zero. Latencies of the MEPs were determined for 7 subjects. The MEP latencies obtained in the 3 trials at each scalp position were averaged, and the inverse of that average (1/mean latency) was computed. (Trials in which no MEP occurred were excluded from the analysis.) Each muscle representation could then be described by a 2-dimensional matrix in which the cells contained the %M recruited or the inverse latency of the MEP. This matrix was defined as the map of the muscle. Each map was characterized by maximal amplitude (the highest %M recorded), volume (the sum of the •%Ms at all scalp positions), area (the number of excitable scalp positions), center of gravity (a map position representing the amplitude-weighted or inverse latency-weighted center of the excitable area; see Statistical analysis), optimal position (the scalp position where the maximal MEP amplitude was evoked), and threshold (the lowest intensity stimulus able to produce a 10/zV MEP in at least 5 out of 10 consecutive trials of stimulation delivered to the optimal position (Cohen et al. 1991a)). The threshold for each muscle was determined independently by positioning the coil at the optimal position, starting stimulation at a subthreshold
MAPPING WITH MAGNETIC STIMULATION
3
\
• 35-41% FCR
• 25-3S~
i::i:i::. 0
0 ~:i:il
•
1s - 2 5 %
t
5- 1 5 %
ii#~i 2 . 5 - 5 % >0-2s~
~::+ii:!i:~!ii
0 CZ
APB
Biceps Llcm ~i
+
.......i?,+~ ............
i
i i ~:!!i~i+~@i i
i
.........
Deltoid Fig. 2. Maps of M E P amplitudes from all 4 muscles of one subject.Squares represent stimulationsiteslocated I cm apart on the scalp.The shading of each square representsthe % M recruited at that position.Maps from prmdmal muscles show lower amplitudes and fewer excitable positions than those from distal muscles. Note the larger representation of biceps and deltoid on the right hemisphere.
[]
APB
[ ] FCR
[ ] Biceps
[ ] Deltoid
1400 1200
•
I
F, o 1000 E~ ~ ~
IIL
80o
+++oo
+ +
~- E
400
+ if)
+
O-
i°.
5
III
HB, IH i._ [n
I, I
La
lm~l~, k
Subject Fig. 3. Averages of left and right map volumes for all 4 muscles of 10 subjects.
[.
4
E.M. W A S S E R M A N N E T AL.
intensity and gradually increasing the intensity until the criteria for threshold were met.
Anterior L e f t hemisphere
Right hemisphere
Statistical analysis Repeated measures analysis of variance techniques using t tests on contrasts (Fleiss 1986) were used to compare differences in response between distal (APB, FCR) and proximal (biceps, deltoid) muscles. Results for adjacent muscle pairs (APB-FCR, FCR-biceps, biceps-deltoid) were also compared by t tests. Map location was measured in 2 ways: optimal position and center of gravity. Each was a bivariate measurement (lateral coordinate, anteroposterior coordinate). The center of gravity w a s a n amplitude-weighted or inverse latency-weighted mean position computed as follows: for each scalp position on a map, the amplitude-weight was computed as the amplitude at that position divided by the sum of the %Ms recorded for the map (volume). The weight at any scalp position can be interpreted as the proportion of the total map amplitude contributed by that location. The weights must sum to 1 over the map. The inverse latency-weight was computed by replacing the amplitude at each location by the inverse latency. The lateral coordinate of the center of gravity was computed by multiplying the lateral coordinate at each position by its amplitude-weight or inverse latency-weight and summing over all positions. The anteroposterior coordinate was computed by an analogous method. The 95% confidence regions were calculated for the mean locations of the optimal position and amplitude center of gravity for each of the 4 muscles (Johnson and Wichern 1988). Hemispheric differences within each muscle were assessed by paired t tests. The significance level was P < 0.05.
Results
Motor evoked potentials to transcranial magnetic stimulation of the contralateral hemisphere were obtained in resting muscles of all subjects. Responses to stimulation from one muscle pair are shown in Fig. 1, and maps of average %M from 3 trials of stimulation in 4 muscle pairs are shown in Fig. 2. The highest MEP amplitudes were clustered near the middle of the maps.
cz
~ APB
0 FCR
~I~ Biceps
0 Deltoid
I
cm
Fig. 4. Superimposed smoothed plots of the excitable scalp areas for all 4 muscles of one subject. Areas for different muscles overlap, but
more distal musclesare morelateral.
Amplitudes fell off gradually around this peak, and there was generally a large surrounding region of low amplitude responses. The following relationships were found to hold for the averages across subjects. The P values reported for differences between distal and proximal muscles result from 1-sided t tests on the means. All other P values reported result from 2-sided tests. Area, maximal amplitude, and volume were greater in all 10 subjects for distal (APB and FCR) than for proximal (biceps and deltoid) muscles (all P < 0.005), and threshold was lower in distal muscles than in proximal muscles (P < 0.001; Table I). Volume and maximal amplitude were greater in the more distal muscle of each adjacent muscle pair (P < 0.05). Fig. 3 shows this relation for the map volumes of the 10 subjects. Area was larger and threshold lower in biceps than in deltoid and lower in FCR than in biceps (P < 0.05). Area and threshold differences were not significant between FCR and APB. Amplitude-dependent measures (maximal amplitude and volume) exhibited much greater variability between muscles and subjects than did threshold and area (Table I). Large amplitudes were always associated with large areas and low thresholds. Although, in individual subjects, the maps of different muscles overlapped, there was a somatotopic shift in their location (Fig. 4). Maps of deltoid were centered posteromedially and APB maps anterolaterally,
TABLE I Summary of m a p m e a s u r e s by muscle for right and left arms of 10 subjects. Measure
Area Maximal amplitude Volume Threshold
APB
FCR
Biceps
Deltoid
Range
M e a n + S.D.
Range
M e a n + S.D.
Range
M e a n + S.D.
Range
M e a n 5: S.D.
16 - 66 8,5- 67.3 35-2043 36 - 78
45.0+ 15.0 38.9-1-14.2 636 + 3 7 2 52.3+ 10.1
13 - 66 2.2- 58.3 8-757 36 - 8 0
45 + 14.9 22.7+ 14.3 325 + 2 2 6 53.0+ 10.0
6 - 59 0~2- 53.3 1 -460 43 -100
34.8+13.0 10.8+ 8.9 97 +81 62.7+15.4
6 - 49 0.2- 14.8 1 -144 45 -100
27.2+11.2 4 . 7 + 2.8 40 + 2 7 71.2+15.8
Standard deviations are calculated for the m e a n s of the right and left arms of each subject. Values are expressed as follows: area = n u m b e r of excitable scalp positions; maximal amplitude = % M (see text for definition); volume = % M x area; threshold = % stimulator output.
M A P P I N G WITH MAGNETIC STIMULATION
5
TABLE II Grid coordinates of average optimal positions and centers of gravity for muscles of 10 subjects. APB
Optimal position Center of gravity
FCR
Biceps
Deltoid
Left
Right
Left
Right
Left
Right
Left
Right
5.0, 1.6 5.0, 1.0
- 5.6, 1.8 - 5.7, 0.7
4.8, 1.0 4.6, 0.9
- 5.3, 1.4 - 5.2, 0.6
3.3, 0.0 3.9, 0.5
- 4.9, 0.9 - 4.7, 0.2
3.5, 0.2 3.6, 0.5
- 4.4, 0.4 - 4.4, 0.1
The first term in the vector is the lateral distance (in cm) from Cz, with positive to the right. The second term is the anteroposterior distance (in cm) from Cz, with positive anterior.
Left h e m i s p h e r e
Right hemisphere
Anterior
A
t
3'
1
o
CZ
1
2
i
i
6
4
i
i
i
i
i
i
2
0
2
4
6
[] APB
i
8
~
B
FCR
.~ Biceps
4
[ ] Deltoid
I
tiE
2
0
I
I
I
C 2
Latency
' S@ o
A&~I~ ~
CZ
Amplitude
[]
APB
•
o
FCR
•
A
Biceps
•
Deltoid
•
Fig. 5. Average positions and 95% confidence regions for amplitude center of gravity (A) and optimal position (B) for 10 subjects and average locations of inverse latency centers of gravity and amplitude centers of gravity for 7 subjects (C). Scale is in centimeters from Cz. Note the posteromedial to anterolateral progression in the location of more distal muscles. Variability in both center of gravity and optimal position is greater on the left than on the right hemisphere. The similarity between amplitude and inverse latency locations reflects the fact that amplitudes tended to be larger and latencies shorter near the center of the map.
6
E.M. WASSERMANN ET AL.
A tO
~:~
300
>~
200,
-
_
IK > o
gg
,oo
o E > ~m 4:
0
~i_
T
LBiceps
-'F R Biceps L Deltoid RDeltoid
v
B [ ] APB
[ ] FCR
~JBiceps
[ ] Deltoid
20~
E 15
I
0
~=
io
i 1 21
41
I BI
Right handed
Ambidextrous
li01 Left handed
Subject Fig. 6. A: average volumes for left and right biceps and deltoid for all 10 subjects. Error bars represent 1 S.D. B: comparison of volume in all 4 muscles of left and right arms of each subject. Column height represents the ratio of left to right volume. For all subjects, except subject 6, the volume is greater (ratio > 1) in the left deltoid and biceps. There is a tendency for the volume in the right APB to be larger (ratio < 1) for right handed subjects. Left handed subjects showed the inverse tendency.
with biceps and FCR maps distributed in between. The optimal positions and centers of gravity for amplitude and inverse latency averaged across subjects clearly show this progression (Fig. 5 and Table II). Variability was greater on the left than on the right hemisphere. Volumes for biceps and deltoid were larger on the right hemisphere (left arm) than on the left (P < 0.05, all subjects except no. 6; Fig. 6). Area for biceps was larger in 8 of 10 subjects on the right hemisphere than on the left (P < 0.01), and area for deltoid showed a strong tendency in this direction (P < 0.06). Threshold tended to be lower for the proximal muscles on the right hemisphere as well, but this difference was only significant for the deltoid (P < 0.01). Right handed subjects tended to have larger representations of APB on the left hemisphere, whereas in the two left handed
subjects these representations were larger on the right hemisphere (Fig. 6B). There were no significant differences between CMAPs on the two sides across subjects. Latencies of the MEPs varied according to the scalp position stimulated, with a tendency to be shorter near the center of the map (Fig. 7). Centers of gravity for inverse latency were close to the centers of gravity for amplitude (Fig. 5C).
Discussion Magnetic stimulation is useful for exploring the cortical representations of muscles in humans (Cohen and Hallett 1988; Cohen et al. 1990a, 1991a-c). The pres-
MAPPING WITH MAGNETIC STIMULATION
(~
• •
>1 SO shorter
i
<1 SD l o n g e r >l SD longer
O .............
~
7
~~:~
J
© cz L_ 1 cm
Fig. 7. Maps of MEP latencies for the APBs of the subject whose MEP amplitude maps are shown in Fig. 2. The shading of each square represents the difference (in S.D.s) between the average latency of the MEP evoked at that scalp position on 3 trials, and the mean latency for the entire map of the muscle. Latencies tend to be shorter near the center of the map.
ent study shows that it can define the 2-dimensional shapes of muscle representations in detail. The overlapping representations that we observed are consistent with the results of direct cortical stimulation studies in both humans (Penfield and Boldrey 1 9 3 7 ; Woolsey et al. 1979; Liiders et al. 1987) and animals (Chang et al. 1947; Woolsey et al. 1952; Phillips and Porter 1977). Our maps generally showed discrete amplitude peaks, or "hot spots," within the more diffuse overlapping representations. The hot spots, although closely spaced, fell in the expected somatotopic pattern. These points may represent low threshold areas where corticospinal neurons projecting to the particular muscle are most concentrated. This was the hypothesis advanced by Ruch and colleagues in 1946 (see Ruch and Fetz 1979). The tendency for latencies of the MEPs to be shorter in the center of the maps than in the periphery was also observed by Fuhr e{ al. (1991), who suggested that the periphery of the muscle representation, with its lower density of corticospinal neurons, may generate fewer descending impulses in r~sponse to stimulation and require a longer time to achieve the temporal summation necessary for activation of spinal motoneurons. Alternatively, attenuated stimulation reaching the corticospinal tract cells at the center transsynaptically or by spread of the stimulus would excite fewer cells and require more time to generate sufficient temporal summation.
Rothwell et al. (1987) found the threshold to transcranial electrical stimulation to be higher in proximal than in distal arm muscles. Our maps of distal and proximal muscles differed significantly with respect to area and threshold and, most markedly, with respect to volume. This presumably reflects the greater density and absolute number of corticospinal neurons in the cortical representations of distal muscles (Phillips and Porter 1977; Kuypers 1981). We expected that the distal muscles might be more accessible to stimulation in the hemisphere contralateral to the preferred hand (Shapiro et al. 1990). AIthough there was a trend toward the finding of a larger representation of distal muscles in the hand-dominant hemisphere, there was a more dramatic difference in the proximal muscles, which were more strongly represented on the right hemisphere in all but one of our subjects, with no apparent effect of handedness. Sub-
ject 6, who did not show this tendency, was also atypical in that all of his muscles w e r e v e r y difficult to s t i m u l a t e ( s e e Fig. 3). T h e s m a l l e r r e p r e s e n t a t i o n o f p r o x i m a l muscles o n t h e l e f t h e m i s p h e r e may reflect
encroachment upon the arm area by an expanded representation of the speech apparatus. Cortical surface stimulation showed significant interhemispheric and interindividual differences in the topography of the motor cortex of monkeys, but the differences were not correlated with any behavioral measures such as arm preference (Franz 1915). The differences in topography and response amplitude among our subjects may be due to topological factors such as the thickness and shape of the extracerebral structures or curvature of the cortex itself. Alternatively these differences may reflect the intrinsic excitability of the neural substrate. Focal transcranial magnetic stimulation provides a painless, noninvasive method of mapping the cortical representation of muscles in humans. It can distinguish the representations of different muscles in the same limb, demonstrate their somatotopic relation, and identify focal low threshold regions within these areas. It is sensitive to differences in the accessibility of distal and proximal muscles and also to hemispheric asymmetries. There is evidence that the characteristics of motor maps obtained noninvasively can change systematically in pathological states, for example, following amputation (Cohen et al. 1991a), or spinal cord injury (Levy et al. 1990; Topka et al. 1991). The technique has also been combined with positron emission tomography and evoked cortical potentials to reveal reorganization following hemispherectomy (Cohen et al. 1991c) and abnormal organization in congenital mirror movements (Cohen et al. 1991b).
The editorial assistance of Ms. B.J. Hessie is greatly appreciated.
8
References Barker, A.T., Jalinous, R. and Freeston, I.L. Noninvasive magnetic stimulation of human motor cortex. Lancet, 1985, i: 1106-1107. Chang, H.-T., Ruch, T.C. and Ward, Jr., A.A. Topographical representation of muscles in motor cortex in monkeys. J. Neurophysiol., 1947, 10: 39-56. Cohen, L.G. and Hallett, M. Noninvasive mapping of human motor cortex. Neurology, 1988, 38: 904-909. .... ; Cohen, L.G., Hallett, M. and Lelli, S: Noninvasive mapping of human motor cortex with transcranial magnetic stimulation. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Stoneham, MA, 1990a: 113-119. Cohen, L.G., Roth, B., Nilsson, J., Dang, N., Panizza, M., Bandinelli, S., Friauf, W. and Hallett, M. Effects of coil design on delivery of focal magnetic stimulation. Technical considerations. Electroenceph. clin. Neurophysiol., 1990b, 75: 350-357. Cohen, L.G., Bandinelli, S., Findlay, T.W. and Hallett, M. Motor reorganization after upper limb amputation in man. Brain, 1 9 9 1 a , 114: 615-627. Cohen, L.G., Meer, J., Tarkka, I., Bierner, S., Leiderman, D.B., Dubinsky, R.M., Sanes, J.N., Jabbari, B., Branscum, B. and Hallett, M. Congenital mirror movements. Abnormal motor pathways in two patients. Brain, 1991b, 114: 381-403. Cohen, L.G., Zeffiro, T., Bookheimer, S., Wassermann, E.M., Fuhr, P., Matsumoto, J,, Toro, C. and Hallett, M. Reorganization in motor pathways following a large congenital hemispheric lesion: different motor representation areas for ipsi- and contralateral muscles. J. Physiol. (Lond.), 1991c, 438: 33P. Fleiss, J. The Design and Analysis of Clinical Experiments. Wiley, New York, 1986: 232-235. Franz, S.I. Variations in the distribution of the motor centers. Psychol. Monogr., 1915, 19: 80-162. Fuhr, P., Cohen, L.G., Roth, B. and Hallett, M. Latency of motor evoked potentials to focal transcranial stimulation varies as a function of scalp positions stimulated. Electroenceph. clin. Neurophysiol., 1991, 81: 81-89. Johnson, R.A. and Wichern, D.W. Applied Multivariate Statistical Analysis. Prentice Hall, Englewood Cliffs, NJ, 1988: 179-180. Kuypers, H.G.J.M. Anatomy of the descending pathways. In: J.M. Brookhart and V.B. Mountcastle (Eds.), Handbook of Physiology, Section I: the Nervous System. Vol. II. Motor Control, Part l. American Physiological Society, Bethesda, MD, 1981: 597-666. Levy, W,J., Amassian, V., Traad, M. and Cadwell, J. Plasticity of the motor system demonstrated after spinal cord injury (abstract). Mouv. Disord., 1990, 5 (Suppl. 1): 34. Liiders, H., Lesser, R.P., Dinner, D.S., Morris, H.H., Hahn, J.F.,
E.M. WASSERMANN ET AL. Friedman, L., Skipper, G., Wyllie, E. and Friedman, D. Commentary: Chronic intracranial recording and stimulation with subdural electrodes. In: J.J. Engel (Ed.), Surgical Treatment of the Epilepsies. Raven Press, New York, 1987: 297-321. Merton, P.A. and Morton, H.B. Stimulation of the cerebral cortex in the intact human subject. Nature, 1980, 285: 277. Oldfield, R.C. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia, 1971, 9: 97-113. Penfield, W. and Boldrey, E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain, 1937, 60: 389-443. Phillips, C.G. Epicortical mapping of motor areas in primates. In: Motor Areas of the Motor Cortex. Ciba Foundation Symposium 132. Wiley, Chichester, 1987: 5-16. Phillips, C.G. and Porter, R. Corticospinal Neurones. Academic Press, New York, 1977. Rossini, P.M., Gigli, G.L., Marciani, M.G., Zarola, F. and Caramia, M. Non-invasive evaluation of input-output characteristics of sensorimotor areas in healthy humans. Electroenceph. din. Neurophysiol., 1987, 68: 88-100. Roth, B.J., Saypol, J.M., Hallett, M. and Cohen, L.G. A theoretical calculation of the electrical field induced in the cortex during magnetic stimulation. Electroenceph. clin. Neurophysiol., 1991, 81: 47-56. Rothwell, J.C., Thompson, P.D., Day, B.L., Dick, J.P.R., Kachi, T., Cowan, J.M.A. and Marsden, C.D. Motor cortex stimulation in intact man. 1. General characteristics of EMG responses in different muscles. Brain, 1987, 110: 1173-1190. Ruch, T.C. and Fetz, E.E. The cerebral cortex: its structure and functions. In: T.C. Ruch and H.D. Patton (Eds.), Physiology and Biophysics, Vol. II. Saunders, Philadelphia, PA, 1979: 122. Shapiro, B.E., Macdonell, R.A.L., Day, B.D., Cros, D., Chiappa, K.H. and Helmers, S.L. Interhemispheric differences in stimulus thresholds to transcranial magnetic stimulation. Neurology, 1990, 40 (Suppl 1): 215. Topka, H., Cohen, L.G., Cole, R.A. and Hallett, M. Reorganization of corticospinal pathways following spinal cord injury. Neurology, 1991, 41: 1226-1283. Woolsey, C.N., Settlage, P.H., Meyer, D.R., Sencer, W., Hamuy, T.P. and Travis, A.M. Patterns of localization in precentral and supplementary motor areas and their relation to the concept of a premotor area. Res. Publ. Ass. Nerv. Ment. Dis., 1952, 30: 238-264. Woolsey, C.N., Erickson, T.C. and Gilson, W.E. Localization in somatic sensory and motor areas of human cerebral cortex as determined by direct recording of evoked potentials and electrical stimulation. J. Neurosurg., 1979, 51: 476-506.
1
ELMOCO 91067
Noninvasive mapping of muscle representations in human motor cortex Eric M. Wassermann, Lisa M. McShane, Mark Hallett and Leonardo G. Cohen Human CorticalPhysiology Unit, Human Motor Control Section, Medical Neurology Branch, Office of the Clinical Director, and Biometry and Field Studies Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD (U.S.4.)
(Accepted for publication: 30 August 1991)
Summary We used transcraniai magnetic stimulation to map the cortical representations of 4 upper extremity muscles (abductor pollicis brevis, flexor carpi radialis, biceps, and deltoid) of 10 normal subjects. Three stimuli were delivered to scalp positions 1 cm apart, and the amplitude and latency of the motor evoked potentials (MEPs) were averaged for each position. Maps were described in terms of number of excitable scalp positions, amplitude of MEPs, scalp positions for evoking largest amplitude MEPs, and threshold for producing MEPs. We compared different muscles across subjects and the same muscles on the left and right sides in individual subjects. Distal muscles had larger representations with higher amplitude MEPs and lower thresholds. Biceps and deltoid on the left had larger representations and higher MEP amplitudes than on the right. Maps showed a somatotopic progression on the scalp of proximal to distal muscles along a posteromedial to anterolateral axis. Key words: Magnetic stimulation; Mapping; Human motor cortex; Motor representations
Until recently, stimulation of the h u m a n cerebral cortex was confined to the operating room where ex-
Subjects and methods
perimentation was subject to limitations of time and the extent of cortex that could be investigated in an individual subject (Penfield and Boldrey 1937; Woolsey et al. 1979). H u m a n studies have also b e e n restricted to subjects with disorders of the central nervous system, Modern cortical stimulation e x p e r i m e n t s in animals have focused on the details of organization of m o t o r
Subjects were 10 n o r m a l volunteers, including one of the authors (7 m e n and 3 women), aged 20-49 years. H a n d e d n e s s was assessed by a questionnaire (Oldfield 1971). All subjects gave written informed consent for the study, and the protocol was approved by the institute's clinical research review committee.
representation areas and pathways (Phillips and Porter 1977; Phillips 1987), but how well these findings apply to the h u m a n m o t o r system is unknown. T h e advent of transcranial stimulation has made it possible safely and painlessly to investigate the h u m a n motor cortex and its descending connections in normal subjects (Merton and Morton 1980; Barker et al. 1985). Electrical (Rossini et al. 1987, Cohen and Hallett 1988) and magnetic (Cohen et al. 1990a) transcranial stimulation have b e e n used to m a p the representations of muscles in normal humans. These studies, however, have used a small n u m b e r of stimulation sites. In the present study, we used an 8-shaped coil to deliver
The 10-20 international system for electrode placem e n t was used to locate Cz (a point on the midsagittal line, midway between the nasion and inion), and a grid of positions 1 cm apart was marked on the scalp with reference to Cz. Subjects lay supine on a comfortable bed in a quiet room and were instructed to keep their arms relaxed in extension and pronation at their sides. Subjects were kept alert by the investigator. Surface electrodes were applied to the skin over
multiple focal magnetic stimuli to closely spaced positions on the scalp to m a p and compare the representations of 4 u p p e r extremity muscles in normal subjects,
neously and printed on p a p e r with a conventional 4-channel E M G machine ( D I S A 1500 or Dantec Counterpoint) with filter settings of 100 Hz a n d 10 kHz. The maximal amplitude of the compound motor action potential (CMAP) to peripheral nerve stimulation was
Correspondence to: Dr. Eric M. Wassermann, Building 10, Room 5N226, NINDS, NIH, Bethesda, MD 20892 (U.S.A.).
determined for biceps and d e l t ~ with supramaximal electrical stimulation at Erb's point and for F C R and APB with median nerve stimulation at the elbow and wrist, respectively. Muscle relaxation was assessed by
2
E.M. WASSERMANN ET AL.
Anterior
-
-
,
~
~
~
,..,..,~
~
~
.
.
-.
~
-
.
.
,
_
_
~
~
,,,------
. _ - - . ~ . . . ~
~
~-,--"-"
CZ
_
L
.
Posterior
Left Hemisphere
Right Hemisphere
L 20 msec Fig. 1. Superimposed MEPs produced in right and left deltoid by 3 trials of magnetic stimulation at each scalp position in one subject. The locations of the EMG traces correspond to the scalp positions where they were evoked. Data from the contralateral muscle are shown on each hemisphere.
EMG monitoring before and several times during the experiment. Relaxation was defined as the absence of apparent EMG activity at a display gain of 100/~V/division, Magnetic stimulation was delivered with a Cadwell MES-10 magnetoelectric stimulator through an 8shaped coil whose characteristics are described elsewhere (Cohen et al. 1990b; Roth et al. 1991). The coil was positioned with t h e handle pointing backward in the sagittal axis and the center(intersection of the coil's wings) in contact with, and tangential to, the scalp at the point of stimulation,
Mapping Mapping experiments were divided into two sessions, one for each arm, 1-39 days apart, except for 3 subjects who had both arms tested on the same day. The experiment took 2-3 h for each arm. At 100% of stimulator output, 3 stimuli were delivered to positions 1 cm apart over the hemisphere contralateral to the muscles being recorded. Stimulation was started at different positions and continued in various directions in different subjects and sessions. Successive positions were stimulated until the area where MEPs were produced was surrounded by inactive positions. The triterion for the existence of a MEP was a deflection of 10 /zV or greater with an appropriate latency occurring on any of the 3 trials. The peak-to-peak amplitudes of the
3 MEPs produced in each muscle at each scalp position were averaged off-line and expressed as a percentage of the CMAP for that muscle (%M). Positions where no MEPs could be produced were assigned a value of zero. Latencies of the MEPs were determined for 7 subjects. The MEP latencies obtained in the 3 trials at each scalp position were averaged, and the inverse of that average (1/mean latency) was computed. (Trials in which no MEP occurred were excluded from the analysis.) Each muscle representation could then be described by a 2-dimensional matrix in which the cells contained the %M recruited or the inverse latency of the MEP. This matrix was defined as the map of the muscle. Each map was characterized by maximal amplitude (the highest %M recorded), volume (the sum of the •%Ms at all scalp positions), area (the number of excitable scalp positions), center of gravity (a map position representing the amplitude-weighted or inverse latency-weighted center of the excitable area; see Statistical analysis), optimal position (the scalp position where the maximal MEP amplitude was evoked), and threshold (the lowest intensity stimulus able to produce a 10/zV MEP in at least 5 out of 10 consecutive trials of stimulation delivered to the optimal position (Cohen et al. 1991a)). The threshold for each muscle was determined independently by positioning the coil at the optimal position, starting stimulation at a subthreshold
MAPPING WITH MAGNETIC STIMULATION
3
\
• 35-41% FCR
• 25-3S~
i::i:i::. 0
0 ~:i:il
•
1s - 2 5 %
t
5- 1 5 %
ii#~i 2 . 5 - 5 % >0-2s~
~::+ii:!i:~!ii
0 CZ
APB
Biceps Llcm ~i
+
.......i?,+~ ............
i
i i ~:!!i~i+~@i i
i
.........
Deltoid Fig. 2. Maps of M E P amplitudes from all 4 muscles of one subject.Squares represent stimulationsiteslocated I cm apart on the scalp.The shading of each square representsthe % M recruited at that position.Maps from prmdmal muscles show lower amplitudes and fewer excitable positions than those from distal muscles. Note the larger representation of biceps and deltoid on the right hemisphere.
[]
APB
[ ] FCR
[ ] Biceps
[ ] Deltoid
1400 1200
•
I
F, o 1000 E~ ~ ~
IIL
80o
+++oo
+ +
~- E
400
+ if)
+
O-
i°.
5
III
HB, IH i._ [n
I, I
La
lm~l~, k
Subject Fig. 3. Averages of left and right map volumes for all 4 muscles of 10 subjects.
[.
4
E.M. W A S S E R M A N N E T AL.
intensity and gradually increasing the intensity until the criteria for threshold were met.
Anterior L e f t hemisphere
Right hemisphere
Statistical analysis Repeated measures analysis of variance techniques using t tests on contrasts (Fleiss 1986) were used to compare differences in response between distal (APB, FCR) and proximal (biceps, deltoid) muscles. Results for adjacent muscle pairs (APB-FCR, FCR-biceps, biceps-deltoid) were also compared by t tests. Map location was measured in 2 ways: optimal position and center of gravity. Each was a bivariate measurement (lateral coordinate, anteroposterior coordinate). The center of gravity w a s a n amplitude-weighted or inverse latency-weighted mean position computed as follows: for each scalp position on a map, the amplitude-weight was computed as the amplitude at that position divided by the sum of the %Ms recorded for the map (volume). The weight at any scalp position can be interpreted as the proportion of the total map amplitude contributed by that location. The weights must sum to 1 over the map. The inverse latency-weight was computed by replacing the amplitude at each location by the inverse latency. The lateral coordinate of the center of gravity was computed by multiplying the lateral coordinate at each position by its amplitude-weight or inverse latency-weight and summing over all positions. The anteroposterior coordinate was computed by an analogous method. The 95% confidence regions were calculated for the mean locations of the optimal position and amplitude center of gravity for each of the 4 muscles (Johnson and Wichern 1988). Hemispheric differences within each muscle were assessed by paired t tests. The significance level was P < 0.05.
Results
Motor evoked potentials to transcranial magnetic stimulation of the contralateral hemisphere were obtained in resting muscles of all subjects. Responses to stimulation from one muscle pair are shown in Fig. 1, and maps of average %M from 3 trials of stimulation in 4 muscle pairs are shown in Fig. 2. The highest MEP amplitudes were clustered near the middle of the maps.
cz
~ APB
0 FCR
~I~ Biceps
0 Deltoid
I
cm
Fig. 4. Superimposed smoothed plots of the excitable scalp areas for all 4 muscles of one subject. Areas for different muscles overlap, but
more distal musclesare morelateral.
Amplitudes fell off gradually around this peak, and there was generally a large surrounding region of low amplitude responses. The following relationships were found to hold for the averages across subjects. The P values reported for differences between distal and proximal muscles result from 1-sided t tests on the means. All other P values reported result from 2-sided tests. Area, maximal amplitude, and volume were greater in all 10 subjects for distal (APB and FCR) than for proximal (biceps and deltoid) muscles (all P < 0.005), and threshold was lower in distal muscles than in proximal muscles (P < 0.001; Table I). Volume and maximal amplitude were greater in the more distal muscle of each adjacent muscle pair (P < 0.05). Fig. 3 shows this relation for the map volumes of the 10 subjects. Area was larger and threshold lower in biceps than in deltoid and lower in FCR than in biceps (P < 0.05). Area and threshold differences were not significant between FCR and APB. Amplitude-dependent measures (maximal amplitude and volume) exhibited much greater variability between muscles and subjects than did threshold and area (Table I). Large amplitudes were always associated with large areas and low thresholds. Although, in individual subjects, the maps of different muscles overlapped, there was a somatotopic shift in their location (Fig. 4). Maps of deltoid were centered posteromedially and APB maps anterolaterally,
TABLE I Summary of m a p m e a s u r e s by muscle for right and left arms of 10 subjects. Measure
Area Maximal amplitude Volume Threshold
APB
FCR
Biceps
Deltoid
Range
M e a n + S.D.
Range
M e a n + S.D.
Range
M e a n + S.D.
Range
M e a n 5: S.D.
16 - 66 8,5- 67.3 35-2043 36 - 78
45.0+ 15.0 38.9-1-14.2 636 + 3 7 2 52.3+ 10.1
13 - 66 2.2- 58.3 8-757 36 - 8 0
45 + 14.9 22.7+ 14.3 325 + 2 2 6 53.0+ 10.0
6 - 59 0~2- 53.3 1 -460 43 -100
34.8+13.0 10.8+ 8.9 97 +81 62.7+15.4
6 - 49 0.2- 14.8 1 -144 45 -100
27.2+11.2 4 . 7 + 2.8 40 + 2 7 71.2+15.8
Standard deviations are calculated for the m e a n s of the right and left arms of each subject. Values are expressed as follows: area = n u m b e r of excitable scalp positions; maximal amplitude = % M (see text for definition); volume = % M x area; threshold = % stimulator output.
M A P P I N G WITH MAGNETIC STIMULATION
5
TABLE II Grid coordinates of average optimal positions and centers of gravity for muscles of 10 subjects. APB
Optimal position Center of gravity
FCR
Biceps
Deltoid
Left
Right
Left
Right
Left
Right
Left
Right
5.0, 1.6 5.0, 1.0
- 5.6, 1.8 - 5.7, 0.7
4.8, 1.0 4.6, 0.9
- 5.3, 1.4 - 5.2, 0.6
3.3, 0.0 3.9, 0.5
- 4.9, 0.9 - 4.7, 0.2
3.5, 0.2 3.6, 0.5
- 4.4, 0.4 - 4.4, 0.1
The first term in the vector is the lateral distance (in cm) from Cz, with positive to the right. The second term is the anteroposterior distance (in cm) from Cz, with positive anterior.
Left h e m i s p h e r e
Right hemisphere
Anterior
A
t
3'
1
o
CZ
1
2
i
i
6
4
i
i
i
i
i
i
2
0
2
4
6
[] APB
i
8
~
B
FCR
.~ Biceps
4
[ ] Deltoid
I
tiE
2
0
I
I
I
C 2
Latency
' S@ o
A&~I~ ~
CZ
Amplitude
[]
APB
•
o
FCR
•
A
Biceps
•
Deltoid
•
Fig. 5. Average positions and 95% confidence regions for amplitude center of gravity (A) and optimal position (B) for 10 subjects and average locations of inverse latency centers of gravity and amplitude centers of gravity for 7 subjects (C). Scale is in centimeters from Cz. Note the posteromedial to anterolateral progression in the location of more distal muscles. Variability in both center of gravity and optimal position is greater on the left than on the right hemisphere. The similarity between amplitude and inverse latency locations reflects the fact that amplitudes tended to be larger and latencies shorter near the center of the map.
6
E.M. WASSERMANN ET AL.
A tO
~:~
300
>~
200,
-
_
IK > o
gg
,oo
o E > ~m 4:
0
~i_
T
LBiceps
-'F R Biceps L Deltoid RDeltoid
v
B [ ] APB
[ ] FCR
~JBiceps
[ ] Deltoid
20~
E 15
I
0
~=
io
i 1 21
41
I BI
Right handed
Ambidextrous
li01 Left handed
Subject Fig. 6. A: average volumes for left and right biceps and deltoid for all 10 subjects. Error bars represent 1 S.D. B: comparison of volume in all 4 muscles of left and right arms of each subject. Column height represents the ratio of left to right volume. For all subjects, except subject 6, the volume is greater (ratio > 1) in the left deltoid and biceps. There is a tendency for the volume in the right APB to be larger (ratio < 1) for right handed subjects. Left handed subjects showed the inverse tendency.
with biceps and FCR maps distributed in between. The optimal positions and centers of gravity for amplitude and inverse latency averaged across subjects clearly show this progression (Fig. 5 and Table II). Variability was greater on the left than on the right hemisphere. Volumes for biceps and deltoid were larger on the right hemisphere (left arm) than on the left (P < 0.05, all subjects except no. 6; Fig. 6). Area for biceps was larger in 8 of 10 subjects on the right hemisphere than on the left (P < 0.01), and area for deltoid showed a strong tendency in this direction (P < 0.06). Threshold tended to be lower for the proximal muscles on the right hemisphere as well, but this difference was only significant for the deltoid (P < 0.01). Right handed subjects tended to have larger representations of APB on the left hemisphere, whereas in the two left handed
subjects these representations were larger on the right hemisphere (Fig. 6B). There were no significant differences between CMAPs on the two sides across subjects. Latencies of the MEPs varied according to the scalp position stimulated, with a tendency to be shorter near the center of the map (Fig. 7). Centers of gravity for inverse latency were close to the centers of gravity for amplitude (Fig. 5C).
Discussion Magnetic stimulation is useful for exploring the cortical representations of muscles in humans (Cohen and Hallett 1988; Cohen et al. 1990a, 1991a-c). The pres-
MAPPING WITH MAGNETIC STIMULATION
(~
• •
>1 SO shorter
i
<1 SD l o n g e r >l SD longer
O .............
~
7
~~:~
J
© cz L_ 1 cm
Fig. 7. Maps of MEP latencies for the APBs of the subject whose MEP amplitude maps are shown in Fig. 2. The shading of each square represents the difference (in S.D.s) between the average latency of the MEP evoked at that scalp position on 3 trials, and the mean latency for the entire map of the muscle. Latencies tend to be shorter near the center of the map.
ent study shows that it can define the 2-dimensional shapes of muscle representations in detail. The overlapping representations that we observed are consistent with the results of direct cortical stimulation studies in both humans (Penfield and Boldrey 1 9 3 7 ; Woolsey et al. 1979; Liiders et al. 1987) and animals (Chang et al. 1947; Woolsey et al. 1952; Phillips and Porter 1977). Our maps generally showed discrete amplitude peaks, or "hot spots," within the more diffuse overlapping representations. The hot spots, although closely spaced, fell in the expected somatotopic pattern. These points may represent low threshold areas where corticospinal neurons projecting to the particular muscle are most concentrated. This was the hypothesis advanced by Ruch and colleagues in 1946 (see Ruch and Fetz 1979). The tendency for latencies of the MEPs to be shorter in the center of the maps than in the periphery was also observed by Fuhr e{ al. (1991), who suggested that the periphery of the muscle representation, with its lower density of corticospinal neurons, may generate fewer descending impulses in r~sponse to stimulation and require a longer time to achieve the temporal summation necessary for activation of spinal motoneurons. Alternatively, attenuated stimulation reaching the corticospinal tract cells at the center transsynaptically or by spread of the stimulus would excite fewer cells and require more time to generate sufficient temporal summation.
Rothwell et al. (1987) found the threshold to transcranial electrical stimulation to be higher in proximal than in distal arm muscles. Our maps of distal and proximal muscles differed significantly with respect to area and threshold and, most markedly, with respect to volume. This presumably reflects the greater density and absolute number of corticospinal neurons in the cortical representations of distal muscles (Phillips and Porter 1977; Kuypers 1981). We expected that the distal muscles might be more accessible to stimulation in the hemisphere contralateral to the preferred hand (Shapiro et al. 1990). AIthough there was a trend toward the finding of a larger representation of distal muscles in the hand-dominant hemisphere, there was a more dramatic difference in the proximal muscles, which were more strongly represented on the right hemisphere in all but one of our subjects, with no apparent effect of handedness. Sub-
ject 6, who did not show this tendency, was also atypical in that all of his muscles w e r e v e r y difficult to s t i m u l a t e ( s e e Fig. 3). T h e s m a l l e r r e p r e s e n t a t i o n o f p r o x i m a l muscles o n t h e l e f t h e m i s p h e r e may reflect
encroachment upon the arm area by an expanded representation of the speech apparatus. Cortical surface stimulation showed significant interhemispheric and interindividual differences in the topography of the motor cortex of monkeys, but the differences were not correlated with any behavioral measures such as arm preference (Franz 1915). The differences in topography and response amplitude among our subjects may be due to topological factors such as the thickness and shape of the extracerebral structures or curvature of the cortex itself. Alternatively these differences may reflect the intrinsic excitability of the neural substrate. Focal transcranial magnetic stimulation provides a painless, noninvasive method of mapping the cortical representation of muscles in humans. It can distinguish the representations of different muscles in the same limb, demonstrate their somatotopic relation, and identify focal low threshold regions within these areas. It is sensitive to differences in the accessibility of distal and proximal muscles and also to hemispheric asymmetries. There is evidence that the characteristics of motor maps obtained noninvasively can change systematically in pathological states, for example, following amputation (Cohen et al. 1991a), or spinal cord injury (Levy et al. 1990; Topka et al. 1991). The technique has also been combined with positron emission tomography and evoked cortical potentials to reveal reorganization following hemispherectomy (Cohen et al. 1991c) and abnormal organization in congenital mirror movements (Cohen et al. 1991b).
The editorial assistance of Ms. B.J. Hessie is greatly appreciated.
8
References Barker, A.T., Jalinous, R. and Freeston, I.L. Noninvasive magnetic stimulation of human motor cortex. Lancet, 1985, i: 1106-1107. Chang, H.-T., Ruch, T.C. and Ward, Jr., A.A. Topographical representation of muscles in motor cortex in monkeys. J. Neurophysiol., 1947, 10: 39-56. Cohen, L.G. and Hallett, M. Noninvasive mapping of human motor cortex. Neurology, 1988, 38: 904-909. .... ; Cohen, L.G., Hallett, M. and Lelli, S: Noninvasive mapping of human motor cortex with transcranial magnetic stimulation. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Stoneham, MA, 1990a: 113-119. Cohen, L.G., Roth, B., Nilsson, J., Dang, N., Panizza, M., Bandinelli, S., Friauf, W. and Hallett, M. Effects of coil design on delivery of focal magnetic stimulation. Technical considerations. Electroenceph. clin. Neurophysiol., 1990b, 75: 350-357. Cohen, L.G., Bandinelli, S., Findlay, T.W. and Hallett, M. Motor reorganization after upper limb amputation in man. Brain, 1 9 9 1 a , 114: 615-627. Cohen, L.G., Meer, J., Tarkka, I., Bierner, S., Leiderman, D.B., Dubinsky, R.M., Sanes, J.N., Jabbari, B., Branscum, B. and Hallett, M. Congenital mirror movements. Abnormal motor pathways in two patients. Brain, 1991b, 114: 381-403. Cohen, L.G., Zeffiro, T., Bookheimer, S., Wassermann, E.M., Fuhr, P., Matsumoto, J,, Toro, C. and Hallett, M. Reorganization in motor pathways following a large congenital hemispheric lesion: different motor representation areas for ipsi- and contralateral muscles. J. Physiol. (Lond.), 1991c, 438: 33P. Fleiss, J. The Design and Analysis of Clinical Experiments. Wiley, New York, 1986: 232-235. Franz, S.I. Variations in the distribution of the motor centers. Psychol. Monogr., 1915, 19: 80-162. Fuhr, P., Cohen, L.G., Roth, B. and Hallett, M. Latency of motor evoked potentials to focal transcranial stimulation varies as a function of scalp positions stimulated. Electroenceph. clin. Neurophysiol., 1991, 81: 81-89. Johnson, R.A. and Wichern, D.W. Applied Multivariate Statistical Analysis. Prentice Hall, Englewood Cliffs, NJ, 1988: 179-180. Kuypers, H.G.J.M. Anatomy of the descending pathways. In: J.M. Brookhart and V.B. Mountcastle (Eds.), Handbook of Physiology, Section I: the Nervous System. Vol. II. Motor Control, Part l. American Physiological Society, Bethesda, MD, 1981: 597-666. Levy, W,J., Amassian, V., Traad, M. and Cadwell, J. Plasticity of the motor system demonstrated after spinal cord injury (abstract). Mouv. Disord., 1990, 5 (Suppl. 1): 34. Liiders, H., Lesser, R.P., Dinner, D.S., Morris, H.H., Hahn, J.F.,
E.M. WASSERMANN ET AL. Friedman, L., Skipper, G., Wyllie, E. and Friedman, D. Commentary: Chronic intracranial recording and stimulation with subdural electrodes. In: J.J. Engel (Ed.), Surgical Treatment of the Epilepsies. Raven Press, New York, 1987: 297-321. Merton, P.A. and Morton, H.B. Stimulation of the cerebral cortex in the intact human subject. Nature, 1980, 285: 277. Oldfield, R.C. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia, 1971, 9: 97-113. Penfield, W. and Boldrey, E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain, 1937, 60: 389-443. Phillips, C.G. Epicortical mapping of motor areas in primates. In: Motor Areas of the Motor Cortex. Ciba Foundation Symposium 132. Wiley, Chichester, 1987: 5-16. Phillips, C.G. and Porter, R. Corticospinal Neurones. Academic Press, New York, 1977. Rossini, P.M., Gigli, G.L., Marciani, M.G., Zarola, F. and Caramia, M. Non-invasive evaluation of input-output characteristics of sensorimotor areas in healthy humans. Electroenceph. din. Neurophysiol., 1987, 68: 88-100. Roth, B.J., Saypol, J.M., Hallett, M. and Cohen, L.G. A theoretical calculation of the electrical field induced in the cortex during magnetic stimulation. Electroenceph. clin. Neurophysiol., 1991, 81: 47-56. Rothwell, J.C., Thompson, P.D., Day, B.L., Dick, J.P.R., Kachi, T., Cowan, J.M.A. and Marsden, C.D. Motor cortex stimulation in intact man. 1. General characteristics of EMG responses in different muscles. Brain, 1987, 110: 1173-1190. Ruch, T.C. and Fetz, E.E. The cerebral cortex: its structure and functions. In: T.C. Ruch and H.D. Patton (Eds.), Physiology and Biophysics, Vol. II. Saunders, Philadelphia, PA, 1979: 122. Shapiro, B.E., Macdonell, R.A.L., Day, B.D., Cros, D., Chiappa, K.H. and Helmers, S.L. Interhemispheric differences in stimulus thresholds to transcranial magnetic stimulation. Neurology, 1990, 40 (Suppl 1): 215. Topka, H., Cohen, L.G., Cole, R.A. and Hallett, M. Reorganization of corticospinal pathways following spinal cord injury. Neurology, 1991, 41: 1226-1283. Woolsey, C.N., Settlage, P.H., Meyer, D.R., Sencer, W., Hamuy, T.P. and Travis, A.M. Patterns of localization in precentral and supplementary motor areas and their relation to the concept of a premotor area. Res. Publ. Ass. Nerv. Ment. Dis., 1952, 30: 238-264. Woolsey, C.N., Erickson, T.C. and Gilson, W.E. Localization in somatic sensory and motor areas of human cerebral cortex as determined by direct recording of evoked potentials and electrical stimulation. J. Neurosurg., 1979, 51: 476-506.