Corticomotor representation of the sternocleidomastoid muscle

Transcription

1 braini0203 Corticomotor representation of the sternocleidomastoid muscle Brain (1997), 120, M. L. Thompson, 1,2 G. W. Thickbroom 1,2 and F. L. Mastaglia 1,2,3 1 Australian Neuromuscular Research Institute, 2 University Correspondence to: Dr G. Thickbroom, Australian Department of Medicine and 3 Department of Neurology Neuromuscular Research Institute, Queen Elizabeth II and Clinical Neurophysiology, QEII Medical Centre, Medical Centre, Nedlands, WA 6009, Australia Nedlands, Australia Summary The topography of the cortical motor projection to the sternocleidomastoid (SCM) muscles was investigated in 15 normal subjects using the technique of transcranial magnetic stimulation. Contrary to the long-held view that the representation of the neck muscles in the motor strip is close to that of the face, our findings indicate that the projection to both the ipsilateral and contralateral SCM muscles arises from an area of cortex high up on the cerebral convexity close to the trunk representation and at a comparable level to the sensory representation of the neck in the post-central cortex. The use of shielded intramuscular recordings showed that surface recordings detect activity not only from the SCM muscle but also from the overlying platysma muscle whose cortical representation is lateral to that of the hand muscles. Our findings confirm that the corticomotor projection to the SCM muscle follows both a contralateral monosynaptic pathway, and an ipsilateral pathway, and indicate that the latter pathway may be disynaptic. Keywords: transcranial magnetic stimulation; motor cortex; mapping; sternocleidomastoid muscle Abbreviations: APB abductor pollicis brevis; MEP motor (electromyographic) evoked potential; SCM sternocleidomastoid (muscle) Introduction The sternocleidomastoid (SCM) muscle is a functionally important muscle involved in a range of head turning, tilting and flexing manoeuvres; however, little is known of its cortical representation, and there is uncertainty concerning the course of the corticomotor pathway to this muscle. A number of studies have provided evidence for an ipsilateral projection to the SCM muscle, and have furthermore proposed that the pathway may undergo a double decussation, firstly in the pons and then again in the upper cervical cord (Bender et al., 1964; Geschwind, 1981; Iannone and Gerber, 1982; Mastaglia et al., 1986; Mazzini and Schieppati, 1992; Odergren and Rimpiläinen, 1996). However, there is no direct neuroanatomical evidence for such a course in the human. An alternative hypothesis is that the SCM muscle is controlled primarily by a contralateral corticomotor projection. This view has received some support from recent transcranial stimulation studies, in which the largest amplitude motor evoked potentials (MEPs) were recorded on the side contralateral to the stimulated hemisphere (Benecke et al., 1988; Gandevia and Applegate, 1988; Berardelli et al., 1991). Oxford University Press 1997 The current understanding of the cerebral origins of the corticomotor projection to the SCM is derived mainly from observations of head movements following direct cortical stimulation (Penfield and Rasmussen, 1952; Woolsey et al., 1979). According to these studies, the cortical sites at which stimulation could evoke head movement fell into two separate groups (Penfield and Rasmussen, 1952). First, sites at which stimulation resulted almost exclusively in head turning to the opposite side were located anterior to the motor strip, at sites often associated with eye turning to the contralateral side. It was concluded that these anterior sites probably represented a secondary motor area associated with coordinated head and eye turning movements. A second group of cortical sites at which stimulation could lead to head rotation to either side, and at which neck flexion and extension movements could also be evoked, was located within the motor strip at the level of the face representation, and it was on the basis of these findings that the neck muscles were entered into the motor homunculus at this level. A more recent direct cortical stimulation study made some observations on head turning.

2 246 M. L. Thompson et al. However, these were not detailed, and did not challenge the earlier findings (Woolsey et al., 1979). Consequently, it is generally considered that the SCM muscle representation is located near the face area of the motor cortex, lateral to the representation of the upper extremity. In the present study we have mapped the topography of the cortical projection to the SCM muscles using transcranial magnetic stimulation. Our findings show that, contrary to the conventional view, the representation of the SCM muscle in the motor cortex is situated high up on the convexity of the cerebral hemisphere close to the trunk area and on a level equivalent to the sensory representation of the neck in the post-central cortex. Methods Subjects With the approval of the Ethics Committee of the University of Western Australia, studies were performed on 15 healthy adult subjects (14 male aged years and one female aged 28 years) who gave written, informed consent. Electromyographic (EMG) recordings Surface EMG activity was recorded from the SCM muscles on both sides. The active electrode was placed between the upper third and the lower two thirds of the SCM muscle and the reference electrode was 2 cm below this position. Intramuscular recordings were also made in two subjects using dual monopolar shielded needles inserted into either the SCM or the platysma muscle near the motor point. Surface EMG activity was recorded from the left abductor pollicis brevis (APB) muscle, with the active electrode over the motor point and the reference electrode over the metacarpophalangeal joint. The EMG signal was amplified 1000 with high and low pass filtering at 10 Hz and 2 khz, respectively, and was digitized at intervals of 0.5 ms for 500 ms following each stimulus. of the APB muscle (10 3% of maximum root-meansquared EMG). Magnetic stimulation A Magstim 200 with a 5 cm diameter figure-of-eight coil (Magstim Company, Sheffield, UK) was used to deliver stimuli at multiple scalp sites overlying the motor cortex, at an intensity of 100% of stimulator output in the case of SCM recordings, and 20% above motor threshold in the case of the APB. The stimulator coil was held tangential to the skull with the handle in an antero-posterior orientation, and with the centre of the figure-of-eight over the site to be stimulated. Stimulus sites were located using a latitude/longitude based coordinate system. Latitude was defined as the distance over the scalp from the vertex, and longitude as the distance along a line of constant latitude from a reference line passing through the vertex and joining the left and right pre-auricular crease (the interaural line). A snugly fitting, flexible cap with markings at spacings of 1 cm in latitude and longitude was used to identify stimulus sites. During mapping, four stimuli were presented s apart at each scalp site. The first site stimulated was close to the estimated centre of the motor area, at which the MEP was of the greatest amplitude. Scalp sites were stimulated by increasing and then decreasing the longitude in 2 cm steps until there was no discernible MEP. This process was repeated at latitudes 1 cm apart until the map borders had been defined. Voluntary muscle activation Subjects lay in a reclinable chair, in a semi-supine position, with the head resting on a support. Recordings were made while the subject maintained a controlled low-level voluntary forward flexion of the head of 10 3% of the maximum rootmean-squared EMG activity, which usually corresponded to the effort necessary to just raise the head off the head support. For the intramuscular studies, the EMG could not be monitored in the same manner as with surface recordings, due to problems with electrode movement during measurement of maximum voluntary contraction, and recordings were therefore made with the head just raised above the head support. For the hand studies, recordings were made during a controlled low-level voluntary contraction MEP analysis and map derivation Peak-to-peak MEP amplitude and duration of the post-mep silent period were measured. Maps were generated by plotting MEP amplitude as a function of scalp site stimulated. This technique has been used to generate topographic maps of the cortical motor output to individual upper limb muscles (Wassermann et al., 1992; Wilson et al., 1993; Mortifee et al., 1994), and these maps have been shown to correspond to activation of primary motor cortex by co-registration with MRIs and PET of the cerebral convexity (Wassermann et al., 1996). In addition, using functional MRI (Siemens 1.5 Tesla MRI system), we have shown that there is a close relationship between the motor output maps for individual hand muscles and the area of precentral cortex activated during a thumb index finger pinch (G. W. Thickbroom, M. L. Thompson and F. L. Mastaglia, unpublished data). In the present study we used a mapping technique which has been described in detail elsewhere (Wilson et al., 1993). In summary, a continuous spline function was fitted to the MEP amplitude versus scalp- site data, and map quantification was based upon this function. The centre of the map was determined from the location of the maximum of the fitted function, and expressed in centimetres based upon a standardized inter-aural distance of 37 cm. Results MEPs could be recorded from surface electrodes located on both sides of the neck in eight out of 15 subjects. In the

3 Corticomotor representation in sternocleidomastoid muscle 247 Fig. 1 Contralateral MEP waveforms (averaged from four stimuli per site) obtained following stimulation at multiple sites over the right motor cortex in one subject. The waveforms are arranged with increasing distance of the stimulus from the vertex (left to right from 2 cm to 11 cm, in 1 cm steps), and with distance from the interaural line (base to top from 2 cm posterior to the interaural line to 4 cm anterior to this line, in 2 cm steps). The time-origin of the waveforms has been offset by 5 ms to display the MEP optimally. remaining seven subjects, MEPs were not recorded from either side of the neck, although it was determined that in some subjects MEPs could be recorded when using a stimulating coil with a greater peak field strength (14 cm diameter figure-of-eight coil). There was an extensive area over which stimulation could yield a MEP on the contralateral side of the neck, extending from near the midline to 12 cm lateral to the vertex in some subjects. However, with stimulation of different sites, the MEP waveforms appeared to fall into two distinguishable groups. This is illustrated in Fig. 1, where waveforms of a consistent morphology were obtained with stimulation of sites 2 6 cm lateral to the vertex, while a second group of responses were obtained with stimulus sites 7 11 cm lateral to the vertex. This was the case in six out of eight subjects, for whom it was possible to identify two MEP waveform morphologies associated with stimulation of medial or lateral areas of the scalp. In four of these subjects, there was a clear medio-lateral spatial separation between these two areas, while in the other two subjects, the waveforms merged between areas. In the remaining two out of eight subjects, MEPs of only one morphology were obtained with stimulation of lateral sites, and MEPs were not obtained for medially located stimulus sites. On the side of the neck ipsilateral to stimulation, the amplitude of the MEP was generally smaller, the latency longer and the waveform less well defined than was the case contralaterally (Fig. 2). Nevertheless, the topography of the stimulus sites giving rise to ipsilateral responses was similar to that giving rise to responses on the contralateral side. While it was not possible to discriminate distinct separable cortical areas clearly, as was the case contralaterally; in each subject, the sites for which stimulation yielded an ipsilateral response were similar in extent to those which yielded contralateral responses. In three subjects, a third group of responses were obtained on the ipsilateral side at a very short latency (mean 3.3 ms, see Fig. 2) with stimulation of very

4 248 M. L. Thompson et al. Fig. 2 Ipsilateral averaged MEP waveforms obtained following stimulation at multiple sites over the right motor cortex (same subject as in Fig. 1). The star marks the response obtained from stimulation of the facial nerve. lateral sites, probably as a result of activation of the platysma period was ms on the contralateral side, and muscle due to direct stimulation of the facial nerve (see ms on the ipsilateral side. below). The MEP associated with stimulation of the lateral area On the basis of the morphology and topographic had a mean latency of ms on the side contralateral distribution of the MEPs, a distinction was made between to stimulation, and a latency of ms on the ipsilateral the MEPs obtained from stimulation of medial and lateral side (Table 1). The difference in latency between the MEPs areas of the scalp, and maps were generated separately for on the contralateral and ipsilateral sides was ms. the medial and lateral distributions (see Fig. 3). MEP latency The ratio of the mean amplitude of the contralateral to the and amplitude parameters were also analysed separately ipsilateral MEP was 3.1:1. The mean duration of the silent for the medial and lateral stimulation areas. On the side period was ms on the contralateral side, and contralateral to stimulation, the MEPs with the largest ms on the ipsilateral side. amplitude associated with the medial and lateral stimulus The mean location of the medial map was cm areas were measured. On the side ipsilateral to stimulation, from the vertex and cm posterior to the interaural the MEPs associated with stimulation of sites homologous line (n 4). The location of the lateral map was to those chosen for the contralateral MEP were used cm from the vertex and cm anterior to the The MEP associated with stimulation of the medial area interaural line (n 8). In the case of the APB muscle, had a mean latency of ms on the side contralateral the maps were located at a mean of cm from the to stimulation, and ms on the ipsilateral side (Table vertex and cm anterior to the interaural line (n 1). The difference in latency between the MEPs on the 6). The relationship between the map centres for the medial contralateral and ipsilateral sides was ms. The ratio and lateral neck maps and the hand maps are presented for of the mean amplitude of the contralateral to the ipsilateral each subject in Table 2 and illustrated for one subject in Fig. 4. MEP was 1.9:1. The mean duration of the post-mep silent It was hypothesized that the presence of MEPs associated

5 Corticomotor representation in sternocleidomastoid muscle 249 Fig. 3 Topographic maps generated from the waveforms shown in Fig. 1. The view is from above the vertex with the nasion towards the top. The circular dotted lines are at spacings of 30 in latitude, and the cross marks are at 30 spacings in longitude. The key indicates the portion of the head displayed. Scaling was performed separately for the medial and the lateral maps. The two optimal sites are indicated by crosses. Table 1 Contralateral and ipsilateral MEP responses with cortical stimulation at medial and lateral sites Subject Latency (ms) Amplitude (mv) Medial Lateral Medial Lateral Contralateral Ipsilateral Contralateral Ipsilateral Contralateral Ipsilateral Contralateral Ipsilateral FN FN NR NR NR NR NR NR 9 11 NR NR Mean SD FN contamination by the 3 ms short-latency response due to direct stimulation of the facial nerve; NR no MEP response recorded.

6 250 M. L. Thompson et al. Table 2 Location of medial, abductor pollicis brevis and lateral maps for each subject relative to the vertex and interaural line Subject Distance (cm) between medial map and APB map and lateral map and vertex IAL vertex IAL vertex IAL NS NS M M M M NR NR NS NS NR NR Mean SD APB abductor pollicis brevis; IAL inter-aural line; NS muscle not studied; NR no MEP responses recorded; M medial and lateral areas merged, therefore only one map with a lateral optimal site was produced. Fig. 4 The maps from Fig. 3 with the map of the APB muscle from the same subject superimposed. Each map has been scaled separately, and the optimal sites indicated by a cross.

7 Corticomotor representation in sternocleidomastoid muscle 251 Fig. 5 Dual monopolar shielded needle recordings from the left SCM muscle following stimulation at multiple sites over the right motor cortex (averaged from four stimuli per site, same subject as for Fig. 1). The time-origin of the waveforms has been offset by 1 ms to display the MEP optimally. with medial and lateral areas of stimulation may be a result of the surface electrodes detecting activity from two separate muscles with different corticomotor representations. In particular, activity from the platysma, a broad thin muscle which overlies the SCM muscle, could be contributing to the recordings. Therefore in two subjects (1 and 5, Table 1), dual monopolar shielded needle recordings were made from the SCM and platysma muscles. For Subject 1, for whom surface recordings showed both medial and lateral areas, only stimulation of the cortical area between 2 cm and 5 cm from the vertex gave rise to a MEP in the contralateral SCM muscle (Fig. 5). Recordings made from the platysma muscle showed that only stimulation of the cortical area between 7 cm and 11 cm lateral from the vertex gave rise to a MEP in the contralateral platysma (Fig. 6). Figure 7 shows the two separate topographic maps generated from the waveforms presented in Figs 5 and 6 (note the similarity with Fig. 3). In the case of Subject 5, for whom surface recordings were most predominant with stimulation over the lateral scalp area, the needle recordings showed that this response was originating from the platysma muscle and not the SCM muscle. Discussion In the present study, surface responses were recorded over the SCM muscle following stimulation of either of two areas of motor cortex, one of which was medial and the other lateral to the hand area. Based on the results of intramuscular recordings, it was concluded that the area medial to the hand representation corresponded to the corticomotor representation of the SCM muscle, while the region lateral to the hand area represented the platysma muscle. The corticomotor projection to the SCM muscle is conventionally believed to arise from a region of cortex on the lateral convexity of the cerebral hemisphere, located between the representations of the upper limb and face (Penfield and Rasmussen, 1952). The present mapping findings do not concur with this, and suggest that the representation of the SCM muscle is in fact located high up

8 252 M. L. Thompson et al. Fig. 6 Dual monopolar shielded needle recordings from the left platysma muscle following stimulation at multiple sites over the right motor cortex (averaged from four stimuli per site, same subject as for Fig. 1). The time-origin of the waveforms has been offset by 1 ms to display the MEP optimally. on the cerebral convexity in a region of cortex that is medial findings in regard to head movements were variable. In their to the upper limb representation. While the most convincing earliest study (1937), neck movements were omitted from mapping data is available for the contralateral projection, the motor sequence, due to the inconclusive localizing value which yielded the largest amplitude MEPs, the data are also of the small number of points found. In later reports (Penfield consistent with a similar origin for the ipsilateral projection. and Rasmussen, 1952), it was found that non-specific It is therefore concluded that the human neck representation extension, flexion, retraction and jerking movements could lies in the motor sequence at a level between that of the be elicited from stimulation of cortical areas near the face trunk and the upper limb. This is also the level at which, representation; however, the number of cases was still very according to conventional understanding, the representation small (12 times in a series of ~400 patients). We were unable of the neck is located in the sensory sequence (Penfield and to activate the SCM muscle by stimulation near the face Rasmussen, 1952), and it is in keeping with the close area; however, it is conceivable that contraction of the correspondence between the sensory and motor platysma muscle from stimulation of this region may lead to representations of other body regions. head retraction or flexion. While Penfield and Rasmussen While it is conventionally thought that the neck muscles were able to demonstrate head turning in conjunction with are represented in contiguity with the face muscles, the early eye movements with stimulation of an area anterior to the stimulation data of Penfield are based upon relatively limited motor strip, we were unable to activate the SCM muscle observations under difficult conditions. Penfield himself notes with stimulation at this site. This may be because singlepulse that the patients were lying with the head in a ring and transcranial magnetic stimulation preferentially supported by towels, which may have restricted the ease of activates direct, monosynaptic corticospinal pathways, head turning (Penfield and Boldrey, 1937). Furthermore, the whereas this anterior region of cortex may represent a

9 Corticomotor representation in sternocleidomastoid muscle 253 Fig. 7 Maps from the needle recordings of Figs 5 and 6. The medial map is based on the data from the left SCM needle waveforms, and the lateral map is from the platysma waveforms. Each map has been scaled separately. Note the similarity with the medial and lateral maps obtained from surface recordings in Fig. 3. secondary motor area associated with looking and turning compared with magnetic stimulation (Day et al., 1989). There movements (Penfield and Rasmussen, 1952) without a direct was also a greater variability in all response latencies, possibly projection to spinal motoneurons. as a result of the use of intramuscular recordings. In the In previous transcranial magnetic stimulation studies, the present study it was feasible to use surface recording to use of surface EMG recordings or conventional concentric measure MEP parameters of the SCM or platysma muscles, needle electrodes with an unshielded shaft has meant that a by taking advantage of a difference in the scalp sites at contribution to the MEPs from the platysma muscle could which responses could be elicited in these muscles. not be excluded (Benecke et al., 1988; Berardelli et al., The largest amplitude and shortest latency SCM muscle 1991; Werhahn et al., 1995). In the studies of Benecke et al. MEPs were recorded on the side contralateral to the stimulated (1988) and Werhahn et al. (1995), the placement of the coil hemisphere, which is contrary to the findings of Odergren was consistent with stimulation of a region of cortex which and Rimpiläinen (1996), but in agreement with other previous we have found projects to the platysma muscle. A transcranial transcranial stimulation studies (Benecke et al., 1988; stimulation study which did use shielded intra- Gandevia and Applegate, 1988; Berardelli et al., 1991). The muscular recording was carried out using electrical latency of ms to the contralateral SCM suggests a stimulation with the anode 6 cm lateral to the vertex and the fast conducting monosynaptic projection, with a central cathode 3 6 cm anterior to the anode (Gandevia and conduction time of ~5 6 ms (2 ms longer than that estimated Applegate, 1988). In that study, the mean conduction time with electrical stimulation; Gandevia and Applegate, 1988) to both sides was 2 3 ms shorter than in the present study, and a slightly shorter peripheral conduction time. On this which would be expected with electrical anodal stimulation basis, it would appear that the major monosynaptic projection

10 254 M. L. Thompson et al. from motor cortex to the SCM muscle follows a contralateral Acknowledgements course, in keeping with the cortical control of the limb We wish to thank Dr R. Stell, Dr B. Taylor, Ms B. A. Phillips muscles. However, it does not necessarily follow that the and Dr T. Day for their contributions to this study, which contralateral projection is the functionally more important was supported by the Neuromuscular Foundation and the projection to the SCM muscle. Given the complexity of Medical Research Fund of Western Australia. possible head turning, tilting and flexing manoeuvres involving multiple neck muscles on both sides, it seems more reasonable to conclude that head movements are mediated References by contralateral and ipsilateral pathways involving both Amassian VE, Stewart M, Quirk GJ, Rosenthal JL. Physiological cerebral hemispheres. basis of motor effects of a transient stimulus to cerebral cortex. The latency of the ipsilateral MEP was consistently longer [Review]. Neurosurgery 1987; 20: (mean ms) than that of the contralateral MEP. The Bender MB, Shanzer S, Wagman IH. On the physiologic decussation magnitude of this difference is suggestive of an additional concerned with head turning. Confin Neurol 1964; 24: synaptic delay in the ipsilateral pathway. There are at least two possible mechanisms which could be responsible for Benecke R, Meyer B-U, Schonle P, Conrad B. Transcranial magnetic such a delay. First, there may actually be an additional stimulation of the human brain: responses in muscles supplied by synapse in the ipsilateral pathway. This would be consistent cranial nerves. Exp Brain Res 1988; 71: with the smaller ipsilateral MEP amplitude, as transcranial Berardelli A, Priori A, Inghilleri M, Cruccu G, Mercuri B, Manfredi magnetic stimulation preferentially activates monosynaptic M. Corticobulbar and corticospinal projections to neck muscle pathways. Although there is no neuroanatomical data on the motoneurons in man. Exp Brain Res 1991; 87: course of the ipsilateral cortical projection to the SCM muscle Day BL, Dressler D, Maertens de Noordhout A, Marsden CD, in human or non-human primates, it is possible that there Nakashima K, Rothwell JC, et al. Electric and magnetic stimulation could be an additional synapse in the reticular formation as of human motor cortex: surface EMG and single motor unit has been demonstrated for the muscles of the upper face in responses [published erratum appears in J Physiol (Lond) 1990; the monkey (Jenny and Saper, 1987). Alternatively, it is 430: 617]. J Physiol (Lond) 1989; 412: possible that a lower fibre density or a weaker activation of Gandevia SC, Applegate C. Activation of neck muscles from the the ipsilateral pathway results in a longer synaptic summation human motor cortex. Brain 1988; 111: time at the motor nucleus, and that the SCM motor neurons would therefore be brought to firing threshold later in the Geschwind N. Nature of the decussated innervation of the sequence of descending I-wave volleys. If this were the case, sternocleidomastoid muscle [letter]. Ann Neurol 1981; 10: 495. we might have expected to see more than one synaptic delay Iannone AM, Gerber AM. Brown-Sequard syndrome with paralysis between the contralateral and ipsilateral pathways in some of head turning [abstract]. Ann Neurol 1982; 12: 116. circumstances, however this was not the case. While the Jenny AB, Saper CB. Organization of the facial nucleus and delay is ~0.5 ms longer than the usual estimate of a synaptic corticofacial projection in the monkey: a reconsideration of the delay (1 2 ms; Day et al., 1989), the additional 0.5 ms could upper motor neuron facial palsy. Neurology 1987; 37: be accounted for by a slightly longer ipsilateral path length. It is less likely that the whole 2.2 ms delay is due simply to Mastaglia FL, Knezevic W, Thompson PD. Weakness of head a longer monosynaptic ipsilateral pathway, as the pathway turning in hemiplegia: a quantitative study. J Neurol Neurosurg would need to be very much longer than the contralateral Psychiatry 1986; 49: pathway (assuming that contralateral central conduction time Mazzini L, Schieppati M. Activation of the neck muscles from the is of the order of 5 6 ms), or that it results from activation ipsi- or contralateral hemisphere during voluntary head movements of more slowly conducting fibres, as transcranial magnetic in humans. A reaction-time study. Electroencephalogr Clin stimulation preferentially activates fast conducting fibres Neurophysiol 1992; 85: (Amassian et al., 1987). Our data indicate that the cortical Mortifee P, Stewart H, Schulzer M, Eisen A. Reliability of projection to the platysma muscle also follows both a transcranial magnetic stimulation for mapping the human motor contralateral and ipsilateral course with a relative delay of cortex. Electroencephalogr Clin Neurophysiol 1994; 93: ms in the ipsilateral response, again suggesting the Odergren T, Rimpiläinen I. Activation and suppression of the possibility that the ipsilateral pathway is disynaptic. sternocleidomastoid muscle induced by transcranial magnetic It is concluded therefore that the corticomotor projection to stimulation. Electroencephalogr Clin Neurophysiol 1996; 101: the SCM muscle follows both a fast conducting monosynaptic contralateral pathway and an ipsilateral pathway which may be disynaptic. The origin of the corticomotor projection to Penfield W, Boldrey E. Somatic motor and sensory representation this muscle on both sides lies in a region of cortex located in the cerebral cortex of man as studied by electrical stimulation. between the representations of the trunk and upper limb, Brain 1937; 60: at a level corresponding to the sensory representation of the neck. Penfield W, Rasmussen TL. The cerebral cortex of man. A clinical study of localization of function. New York: Macmillan, 1950.

11 Corticomotor representation in sternocleidomastoid muscle 255 Wassermann EM, McShane LM, Hallett M, Cohen LG. Noninvasive Wilson SA,Thickbroom GW, Mastaglia FL. Transcranial magnetic mapping of muscle representations in human motor cortex. stimulation mapping of the motor cortex in normal subjects: the Electroencephalogr Clin Neurophysiol 1992; 85: 1 8. representation of two intrinsic hand muscles. J Neurol Sci 1993; 118: Wassermann EM, Wang B, Zeffiro TA, Sadato N, Pascual-Leone A, Toro C, et al. Locating the motor cortex on the MRI with Woolsey CN, Erickson TC, Gilson WE. Localization in somatic transcranial magnetic stimulation and PET. Neuroimage 1996; 3: sensory and motor areas of human cerebral cortex as determined 1 9. by direct recording of evoked potentials and electrical stimulation. J Neurosurg 1979; 51: Werhahn KJ, Classen J, Benecke R. The silent period induced by transcranial magnetic stimulation in muscles supplied by cranial nerves: normal data and changes in patients. J Neurol Neurosurg Received March 22, Revised October 9, Psychiatry 1995; 59: Accepted October 24, 1996.