- Email: [email protected]
Somatosensory Homunculus as Drawn by MEG
7, 377–386 (1998) NI980332
NEUROIMAGE ARTICLE NO.
Somatosensory Homunculus as Drawn by MEG Akinori Nakamura,*,1 Takako Yamada,† Atsuko Goto,† Takashi Kato,* Kengo Ito,* Yuji Abe,† Teruhiko Kachi,† and Ryusuke Kakigi‡ *Department of Biofunctional Research, National Institute for Longevity Sciences, 36-3 Gengo, Morioka-cho, Obu, Aichi 474, Japan; †Department of Neurology, Chubu National Hospital, 36-3 Gengo, Morioka-cho, Obu, Aichi 474, Japan; and ‡Department of Integrative Physiology, National Institute for Physiological Sciences, Myodaiji, Okazaki 444, Japan Received November 12, 1997
We studied a detailed somatosensory representation map of the human primary somatosensory cortex using magnetoencephalography. Somatosensory-evoked magnetic fields following tactile stimulation of multiple points in the right hemibody (including the tongue, lips, fingers, arm, trunk, leg, and foot) were analyzed in five normal subjects. We were able to estimate equivalent current dipoles (ECDs) following stimulation of the tongue, lips, fingers, palm, forearm, elbow, upper arm, and toes in most subjects and those following the stimulation of the chest, ankle, and thigh in one subject. The ECDs were located in the postcentral gyrus and generally arranged in order along the central sulcus, which is compatible with the somatosensory ‘‘homunculus.’’ Linear distances, averaged in five subjects, from the receptive area of the thumb to that of the tongue, little finger, forearm, upper arm, and toes were estimated to be 2.42 6 0.28, 1.25 6 0.28, 2.21 6 0.72, 2.75 6 0.63, and 5.29 6 0.48 cm, respectively. The moment of each ECD, which suggested the size of the cortical areas responsive to the stimulation, was also compatible with the bizarre proportion of the homunculus with a large tongue, lips, and fingers. According to these results, we were able to reproduce a large part of the somatosensory homunculus quantitatively on an individual brain MRI. r 1998 Academic Press
INTRODUCTION Since the landmark study of Penfield and Boldrey (1937), various extensive studies have revealed the one-to-one somatotopic representation of the contralateral body surface in the human postcentral cortex, which has been known as the somatosensory ‘‘homunculus.’’ Most studies were performed by means of invasive techniques under neurosurgical conditions, such as direct electrical stimulation of the cortical surface 1 To whom correspondence should be addressed. Fax: 81-562-446596. E-mail: [email protected].
(Penfield and Boldrey, 1937; Penfield and Rasmussen, 1950) and somatosensory-evoked responses recorded on the cortical surface (Woolsey et al., 1979; Wood et al., 1988; Baumgartner et al., 1991a; Allison et al., 1989). However, these techniques are very restricted in application. It is impossible to use these techniques for evaluation of nonneurosurgical patients or normal subjects. A large number of noninvasive studies using scalprecorded somatosensory evoked potentials (SEPs) have supported the somatotopic organization in the somatosensory cortex (Duff, 1980; Desmedt and Bourguet, 1985; Kakigi and Shibasaki, 1991, 1992). However, the scalp-recorded electroencephalogram (EEG) could not provide enough resolution to estimate the precise electrical source location in the brain. This is because the influence of volume currents on scalp-recorded EEG severely affects precise source identification. Furthermore, large interindividual variability of intervening tissues, including scalp, skull, and cerebrospinal fluid, makes it difficult to estimate the dipole location. Another noninvasive technique for investigating the bioelectrical functions of the brain, magnetoencephalography (MEG), has been developed during the past 25 years. MEG has several advantages over EEG in localizing cortical sources, because the magnetic fields recorded on the scalp are less affected by volume currents and anatomical inhomogeneities (Ha¨ma¨la¨inenet al., 1993). Based on its spatial resolution of a few millimeters, several reports have demonstrated the somatotopic projection of the hand (Brenner et al., 1978; Sutherling et al., 1988; Bamgartner et al., 1991b; Suk et al., 1991; Hari et al., 1984; Kakigi et al., 1994), foot (Kakigi et al., 1995; Shimojo et al., 1996), lips (Mogilner et al., 1994; Hoshiyama, 1996), and scalp (Hoshiyama, 1995) in the human cortex using somatosensory-evoked magnetic fields (SEFs). These reports partially reconfirmed the homunculus. There have been several reports which covered a wider area of the homunculus. Narici et al. (1991) studied multiple receptive fields using SEF following median, femoral, tibial,
377
1053-8119/98 $25.00 Copyright r 1998 by Academic Press All rights of reproduction in any form reserved.
378
NAKAMURA ET AL.
and pudendal nerve stimulations. Hari et al. (1993) demonstrated representations of the leg, hand, fingers, lips, and tongue which agreed with the homunculus. Yang et al. (1993) extensively studied somatosensory mapping of various points in the region of the face and upper limb. However, there are no reports which cover the ‘‘whole body’’ of the homunculus. The objective of our study was to investigate the quantitative, wholebody representation map of the human primary somatosensory cortex and to evaluate the capacities and limitations of today’s multichannel MEG system. We used tactile stimulation for this study. Most of the previous studies were performed using electric stimulation. However, these unnatural stimuli activate a large number of fibers from both deep and superficial receptors, bypassing the peripheral receptors. Tactile stimuli are more physiological than electric stimuli and more suitable for studying the functional map of the cortex involved in single modality of sensation. A proper tactile stimulus activates cutaneous mechanoreceptors selectively and causes more natural sensation without pain or discomfort. METHODS The subjects were five male volunteers ages 21 to 36 years. All subjects were healthy right-handed Japanese. Informed consent was obtained from all five subjects prior to the experiments. An air-puff-derived tactile stimulator (Hoshiyama et al., 1995), which provides a light, superficial pressure stimulus to the skin surface, was used for the stimulation. The skin contact area was a circular rubber bladder 1 cm in diameter, and the intensity of the mechanical stimulation was 40 g/cm2. The rise time was 20 ms as measured from 10 to 90% of the intensity increment. No joint movement was observed in this stimulation. About 40 points in the right hemibody (including the tongue, lips, hand, arm, trunk, leg, and foot) were stimulated in a randomly determined order (Fig. 1). The stimulation device was fixed to each stimulation point at the optimal position at which maximal subjective sensation could be obtained. This point was determined through communication with the subject in each session. The interstimulus interval was randomly jittered, from 400 to 600 ms. SEFs were recorded with a 37-channel biomagnetometer (Magnes, Biomagnetic Technologies, Inc., San Diego, CA). The detection coils of the biomagnetometers were arranged in a uniformly distributed array of concentric circles over a spherically concave surface. The device was 144 mm in diameter. Each coil was 20 mm in diameter and the centers of the coils were 22 mm apart. Each coil was connected to a superconducting quantum interference device. The intrinsic noise level in each channel was ,10 fT/Hz1/2.
FIG. 1. Forty-three stimulated points of the right hemibody. The gray circles indicate stimulus points on the frontside of the body, and black circles indicate stimulus points on the backside of the body. In the cases of arm and leg stimulations, lateral, medial, anterior, and posterior points were stimulated.
When the points in the face and upper limb were stimulated, the recording probe was centered around the C3 (international 10–20 system), and when the points of the lower limb were stimulated, the probe was centered around the Cz. Both probe positions were applied when the points of the trunk were stimulated. Responses were recorded with a 0.1- to 200-Hz bandpass filter, followed by 1- to 100-Hz, and digitized at a sampling rate of 512 Hz. The analysis time was 100 ms before and 200 ms after the stimuli. DC was offset using a prestimulus period (100 ms) as the baseline. At least 150 epochs for each stimulus point were collected and averaged in each session. At least two averages were obtained to ensure reproducibility. Peak latency of the wave component was determined according to the maximum value of the root mean square (RMS) field strength taken over the 37 sensors.
379
SOMATOSENSORY HOMUNCULUS AS DRAWN BY MEG
TABLE 1 Detailed Value of Each ECD
Tongue Lower lip Upper lip Thumb Index finger Middle finger Ring finger Little finger Radial palm Ulnar palm Forearm Elbow Upper arm Chestd Thighd Ankled Toes
Latency
Distancea (cm)
Dipole moment, (nA · m)
Ratiob
Activated areac (mm2 )
35.7 6 5.5 34.5 6 6.9 33.6 6 5.4 46.2 6 3.9 47.2 6 3.0 45.4 6 4.3 46.4 6 5.4 47.5 6 6.0 44.3 6 6.9 46.7 6 4.9 42.1 6 6.9 40.6 6 7.8 38.5 6 4.1 36.4 57.5 71.4 72.5 6 10.5
2.42 6 0.28 1.46 6 0.70 1.17 6 0.53 0 1.05 6 0.58 1.17 6 0.36 1.22 6 0.23 1.25 6 0.28 0.93 6 0.51 1.58 6 0.45 2.21 6 0.72 2.44 6 0.33 2.75 6 0.63 3.89 4.16 5.43 5.29 6 0.48
20.5 6 8.0 17.1 6 3.3 14.9 6 4.8 23.3 6 5.2 16.9 6 5.2 16.5 6 3.8 14.4 6 4.9 15.3 6 4.7 12.0 6 1.9 11.0 6 3.6 9.7 6 2.0 8.4 6 1.6 9.8 6 2.8 5.7 4.4 6.3 12.3 6 3.0
0.880 0.734 0.642 1 0.728 0.709 0.620 0.659 0.516 0.473 0.419 0.361 0.422 0.245 0.189 0.271 0.530
82.0 6 32.0 68.4 6 13.2 59.6 6 19.2 93.2 6 20.8 67.6 6 20.8 66.0 6 15.2 57.6 6 19.6 61.2 6 18.8 48 6 7.6 44.0 6 14.4 38.8 6 8.0 33.6 6 6.4 39.2 6 11.2 22.8 17.6 25.2 49.2 6 12.0
The mean linear distances (mean 6 SD) from the thumb ECD to the other ECDs. Relative ratio of the dipole moment compared to the thumb. c Activated cortical area, estimated as 40 mm2/10 nA · m. d Data obtained from one subject. a b
Source analyses, based on a single moving equivalent current dipole (ECD) model (Sarvas, 1987) in a spherical volume conduction, were applied to magnetic field distribution studies. A local sphere, which was digitized from the parietal portion of each subject’s head, was used for the model. The location (x, y, and z positions), orientation, and dipole moment of the best-fitting single ECD were estimated for each time point. The origin of the head-based coordinate system was the point exactly halfway between the preauricular points. The x axis indicated the coronal plane with positive values to the nasion, the y axis the sagittal plane with positive values in the left, and the z axis the transaxial plane with positive values up. Correlations between the recorded measurements and the values expected from the ECD estimate were calculated. To ensure a strict criterion for dipole fitting, only estimates with a correlation above 0.97 were analyzed. Dipole strength at the peak latency was used as the representative dipole moment of each ECD. In most cases, latencies of the RMS peak and correlation peak were the same or very close. Magnetic resonance images (MRIs) were obtained using a 1.5-T system (Toshiba Medical). T1-weighted images with a contiguous 1-mm slice thickness were used for overlays, with ECD sources determined by MEG. The same anatomical landmarks used to create the MEG head-based 3D coordinate system (the nasion and bilateral preauricular points) were visualized in the MR images by affixing to these points high-contrast cod liver oil capsules (3-mm diameter). The MEG source locations were converted into pixels and slice
values using the MRI transformation matrix, then inserted onto the corresponding MRIs. In order to compare the ECD maps obtained from the different subjects, the coordinate differences from the ECD of the thumb to the ECD of the otherpoints (Dx, Dy, Dz) were calculated. Linear distances between the ECD of the thumb and the ECD of the other points were calculated by the formula ((Dx)2 1 (Dy)2 1 (Dz )2)1/2. RESULTS SEF waveform following tactile stimulation typically showed two major components. We termed them M1 for the earlier in latency and M2 for the later. The M1 was often combined with its subcomponent M18, which was delayed about 10–30 ms from the M1. A small deflection preceding the M1, which we termed M0, was also frequently detected (Fig. 2). We analyzed the M1 in the present study, because this first major component was the most reproducible wave and was thought to originate mainly in area 3b (see Discussion). In all five subjects, high-amplitude M1s were evoked following tactile stimulation of the points on the tongue, lips, and hand. Relatively low-amplitude, but reproducible M1s were recorded following stimulation of the palm, forearm, upper arm, and toes. It was hard to record clear and reproducible M1s following stimulation of the points on the trunk and leg (Fig. 2). The peak latency of the M1 was shortest when the tongue, lower lip, or upper lip was stimulated (,34 ms) and was longest when the toes were stimulated (,72 ms), reflecting the distances from the brain (Fig. 2, Table 1).
380
NAKAMURA ET AL.
FIG. 2. Representative SEF waveforms following tactile stimulation of each point. Waves recorded at each of the 37 sensor points were superimposed on the corrected baseline. The isocontour map of the M1 (right) showed distribution of the magnetic fields. Outward flux is marked by a red gradation, inward flux by green, and the arrow shows the estimated ECD.
Following the tongue, lips, and fingers stimulation, most of the M1 ECDs could be estimated very reliably (more than 99% of correlation). Following the palm, forearm, upper arm, and toes stimulation, most of the ECDs could also be estimated reliably (more than 97%
of correlation). However, in most subjects, reliable M1 ECDs could not be estimated when the trunk or leg was stimulated. Only ECDs following stimulation of the anterior part of the thigh and medial part of the ankle could be estimated with a more than 97% correlation in subject 1. Additionally, one ECD following chest stimulation with a 96.3% correlation in subject 1 was accepted in order to complete the homunculus. These estimated ECDs were located in the left postcentral gyrus. The ECD location to the tongue stimulation was the most inferior, followed by the lips, fingers, and arm, which were gradually shifted to superior and medial, along the central sulcus (Figs. 3A and 3B). The ECDs to the ankle and toes were localized on the midline (Fig. 3C). These findings are more clearly demonstrated by Fig. 4A, which illustrates the averaged coordinate differences between the receptive area of the thumb and that of the other points in the five subjects. Averaged linear distances between the locations of each source were also calculated (Fig. 4B and Table 1). The tongue source was located 2.42 6 0.28 cm (mean 6 SD) from the thumb source. Distances from the area of the thumb to the little finger, forearm, and upper arm were 1.25 6 0.28, 2.21 6 0.72, and 2.75 6 0.63 cm, respectively. The ECD to the radial side of the palm stimulation was estimated close to that of the thumb stimulation, and the ECD to the ulnar side of the palm stimulation was estimated close to the little finger stimulation. No apparent differences could be detected in source locations between the radial, ulnar, anterior, and posterior sides of the forearm stimulation or between the medial, lateral, anterior, and posterior sides of the upper arm stimulation. Similarly, the source locations of toes 1, 2, and 5 were very close and no consistency could be found (e.g., the toe 1 ECD was estimated above the toe 5 ECD in three subjects, but it was estimated below the toe 5 ECD in two subjects). ECD data of the four points around the forearm, elbow, and upper arm, as well as the three points of the toes, were averaged in each subject and used for analyses. The dipole moment of each ECD was also averaged (Table 1). The thumb ECD showed the largest dipole moment (23.3 6 5.2 nA · m) and the tongue ECD was the second (20.5 6 8.0 nA · m). ECDs of the lips and digits showed relatively large dipole moments, while ECDs of the arm, trunk, and leg showed smaller dipole moments. The ECD model approximates a summation of the postsynaptic potentials in the axon dendrites of the pyramidal cells. It is thought that a dipole moment of 10 nA · m would be caused by the synchronous activity of tens of thousands of neurons (Ha¨ma¨la¨inen et al., 1993) and would correspond to 40 mm2 of active cortex (Chapman et al., 1984). According to this hypothesis, about 93 mm2 of the SI cortex was supposed to be shared for the stimulated skin area (about 78.5 mm2) of the thumb, which showed 23.3 nA · m of averaged
SOMATOSENSORY HOMUNCULUS AS DRAWN BY MEG
381
FIG. 3. Estimated M1 ECDs of subject 1 overlapped on his MRI. Displayed MR images are sliced at the position of the ECD of the lower lip (A), thumb (B), and toe 2 (C). The other ECDs were projected onto each slice.
382
NAKAMURA ET AL.
FIG. 4. (A) The coordinate differences between the ECD of the thumb and the ECD of the other points averaged in five subjects. X-position indicates anterior–posterior coordinates with the positive value anteriorly, Y-position indicates lateral–medial coordinates with the positive value laterally, and Z–position indicates vertical coordinates with the positive value superiorly. (B) The linear distances from the ECD of the thumb to the ECD of each other point, averaged in five subjects. The ECDs which locate more inferior than the thumb ECD were demonstrated by negative values (tongue and lips).
dipole moment. Similarly, the presumed cortical area responsive to the stimulation of each part of the body was calculated (Table 1). Taking these estimations into account, we finally drew a somatosensory representation map on subject 1’s 3D brain image (Fig. 5). The size of each ellipse reflected the presumed size of each activated area to the stimulation, demonstrating the weight of each part of the body in the human SI cortex.
DISCUSSION There have been several reports studying SEP or SEF following mechanical stimulation, including tactile and/or vibratory (Ha¨ma¨la¨inen et al., 1990; Yang, 1993; Gallen et al., 1994; Hoshiyama et al., 1995) and air-puff (Hashimoto, 1987; Rossini et al., 1996) stimulation. Forss et al. (1994) studied detailed characteristics
SOMATOSENSORY HOMUNCULUS AS DRAWN BY MEG
of SEF following air-puff stimulation of the middle finger. They demonstrated small deflection N26m (latency around 26 ms), large deflections P42m (43–46 ms) and P66m (66–71 ms), and late components (latency around 100 ms). They estimated ECD locations of earlier components N26m, P42m, and P66m in the SI and late components in bilateral SII and posterior parietal cortices. The N26m, P42m, P66m, and late components were thought to correspond with the M0, M1, M18, and M2 of our results according to latency and morphology. It is thought that tactile stimulation activates mainly area 3b (Iwamura et al., 1983), and the first major component, the M1 in our study, is considered to originate mainly in area 3b (Suk et al., 1991; Forss et al., 1994). MEG analysis is favorable for the sources in area 3b, which is located on the posterior bank of the central sulcus. This is because MEG is most sensitive to brain current tangential to the skull (Ha¨ma¨la¨inen et al., 1993), and dipoles generated in area 3b are mainly tangentially oriented. Additionally, it is known that hierarchical information processing occurs in the rostrocaudal direction within the SI. The neurons in areas 1 and 2 have complex receptive fields, whereas one-to-one somatotopic representation is clearly preserved in area 3b (Iwamura et al., 1983, 1993). Thus, the homuncular organization should be more clearcut in area 3b than in areas 1 and 2. Using MEG and tactile stimulation, we could map a large part of the somatosensory receptive fields on SI, including tongue, lips, fingers, palm, arm, and toes. These representation areas were generally arranged in the above order from inferior to superior, lateral to medial, and anterior to posterior (Figs. 3 and 4). These changes in the coordinates were compatible with the anatomy of the central sulcus and the homunculus. ECD location to the upper lip could be distinguished from that to the lower lip, the former located more superior than the latter in all subjects. This finding also agreed with Penfield’s observations and previous MEG studies (Moligner et al., 1994; Hoshiyama, 1996). Each finger representation area of the thumb, index finger, middle finger, ring finger, and little finger was also distinguishable from the others. They were represented sequentially from thumb to little finger, ascending the postcentral sulcus as noted in previous reports (Baumgartner et al., 1991b; Suk et al., 1991; Hari et al., 1993). The mean linear distances from the thumb ECD to each of the other finger ECDs had very similar numeric values (Fig. 4B, Table 1). We considered that this was because the finger representation area of the somatosensory cortex is convexly curved anteriorly and that the measurement of linear distance cannot reflect precise distance along such a curved portion. This curvature was shown in the MR images (Figs. 3B and 5) and was supported by Fig. 4A, which showed that the index finger ECD was located more anteriorly than the
383
thumb ECD. However, the linear distance between the source of the thumb and that of the little finger (1.25 6 0.28) was well consistent with previous reports, in which the distance was 1.12 to 1.45 cm (Suk et al., 1991), 1.1 6 0.16 cm (Mogilner et al., 1993), or up to 2 cm (Hari et al., 1993). These values were also compatible with the value of 1.5–2 cm, which was studied by direct cortical recording of SEP (Wood et al., 1988). Yang et al. (1993) reported in their MEG study that the ECDs of the palmar pads stimulation were located slightly superior and medial to those of the digits stimulation in one subject. They found converse results in another subject. However, our results showed that the receptive areas of the radial and ulnar side of the palm were estimated close to those of the thumb and little finger, respectively (Figs. 3 and 4, Table 1). These findings were consistent with studies in monkeys (Iwamura et al., 1983). We could map receptive areas to forearm, elbow, and upper arm stimulation and showed that they were located in the above order from inferior to superior and lateral to medial. However, ECD locations following the stimulation of radial, ulnar, anterior, and posterior points of the forearm, which were in a different dermatome, were indistinguishable. Similarly, no apparent difference in ECD location was found following the stimulation of four points around the upper arm. These results suggested that the representation area of the arm might be organized mainly along the distal–proximal axis (from the forearm and elbow to the upper arm) rather than the dermatome. These results were also consistent with studies in monkeys (Werner and Whitsel, 1973). It was difficult to estimate ECDs following the trunk and leg stimulation by the methods presented, because the magnetic fields elicited by these stimulations were, in most cases, too low to obtain a desirable signal to noise ratio. This is probably because the trunk and leg have small representation areas. We considered that these results showed the sensitivity limit of our MEG system, which has an intrinsic noise level of ,10 fT/Hz1/2 in each channel. However, ECD locations following the chest, thigh, and ankle stimulation, which could be estimated only in subject 1, were compatible with the homunculus. Receptive areas to toe stimulation were estimated to be located in the mesial side of the hemisphere. We could not separate the sources of toes 1, 2, and 5 clearly; however, the receptive area for toe stimulation was located deeper than that of the ankle. Penfield demonstrated large cortical representation areas for the tongue, lips, and fingers by bizarre proportions of the homunculus with a large tongue, lips, and fingers. We could confirm these observations by estimating the activated cortical area using the dipole moment of the MEG (Table 1, Fig. 5). The activated areas of the tongue, lips, and fingers were supposed to be larger than those of the palm, arm, body,
FIG. 5. Detailed somatosensory receptive map represented by MEG. The 3D brain image was reconstructed using MRI of subject 1. Each receptive area, which was estimated to be located in the posterior bank of the central sulcus, was projected onto the cortical surface. The size of each ellipse reflects the presumed size of activated cortical area. The ellipses illustrated on the brain image are three times smaller than the actual sizes predicted in Table 1. Note that the receptive area for the toes is in the mesial side of the left hemisphere.
384 NAKAMURA ET AL.
SOMATOSENSORY HOMUNCULUS AS DRAWN BY MEG
and leg. The presumed responsible area of the thumb was about 2, 2.4, and 5.3 times larger than those of the palm, arm, and thigh, respectively. Although there is considerable interindividual variation in the size and shape of the human brain, standardization of the functional somatosensory map will contribute to such studies as plastic reorganization of the human somatosensory cortex (Mogilner et al., 1993; Yang et al., 1994; Elbert et al., 1995). Today’s technologies (just 60 years after the study of Penfield and Boldrey), including MEG, high-resolution MRI, and computer systems, enable us to reproduce the functional somatosensory homunculus on an individual brain image quantitatively by completely noninvasive techniques. ACKNOWLEDGMENT This work was supported by Funds for Comprehensive Research on Aging and Health from the Ministry of Health and Welfare, Japan.
REFERENCES Allison, T., McCarthy, G., Wood, C. C., Darcey, T. M., Spencer, D. D., and Williamson, P. D. 1989. Human cortical potentials evoked by stimulation of the median nerve. I. Cytoarchitectonic areas generating short-latency activity. J. Neurophysiol. 62:694–710. Baumgartner, C., Barth, D. S., Levesque, M. F., and Sutherling, W. W. 1991a. Functional anatomy of human hand sensorimotor cortex from spatiotemporal analysis of electrocorticography. Electroencephalogr. Clin. Neurophysiol. 78:56–65. Baumgartner, C., Doppelbauer, A., Sutherling, W. W., Zeitlhofer, J., Lindinger, G., Lind, C., and Deecke, L. 1991b. Human somatosensory cortical finger representation as studied by combined neuromagnetic and neuroelectric measurements. Neurosci. Lett. 134:103– 108. Brenner, D., Lipton, J., Kaufman, L., and Williamson, S. J. 1978. Somatic evoked magnetic fields of the human brain. Science 199:81–83. Chapman, R. M., Ilmoniemi, R. J., Barbanera, S., and Romani, G. L. 1984. Selective localization of alpha brain activity with neuromagnetic measurements. Electroencephalogr. Clin. Neurophysiol. 58: 569–572. Desmedt, J. E., and Bourguet, M. 1985. Color imaging of parietal and frontal somatosensory potential fields evoked by stimulation of median or posterior tibial nerve in man. Electroencephalogr. Clin. Neurophysiol. 62:1–17. Duff, T. A. 1980. Topography of scalp recorded potentials evoked by stimulation of the digits. Electroencephalogr. Clin. Neurophysiol. 49:452–460. Elbert, T., Pantev, C., Wienbruch, C., Rockstroh, B., and Taub, E. 1995. Increased cortical representation of the fingers of the left hand in string players. Science 270:305–307. Forss, N., Salmelin, R., and Hari, R. 1994. Comparison of somatosensory evoked fields to airpuff and electric stimuli. Electroencephalogr. Clin. Neurophysiol. 92:510–517. Gallen, C. C., Schwartz, B., Rieke, K., Pantev, C., Sobel, D., Hirschkoff, E., and Bloom, F. E. 1994. Intrasubject reliability and validity of somatosensory source localization using a large array biomagnetometer. Electroencephalogr. Clin. Neurophysiol. 90:145–156. Ha¨ma¨la¨inen, H., Kekoni, J., Sams, M., Reinikainen, K., and Na¨a¨ta¨nen,
385
R. 1990. Human somatosensory evoked potentials to mechanical pulses and vibration: Contributions of SI and SII somatosensory cortices to P50 and P100 components. Electroencephalogr. Clin. Neurophysiol. 75:13–21. Ha¨ma¨la¨inen, M., Hari, R., Ilmoniemi, R. J., Knuutila, J., and Lounasmaa, O. V. 1993. Magnetoencephalography. Rev. Mod. Phys. 65: 413–496. Hari, R., Reinikainen, K., Kaukoranta, E., Ha¨ma¨la¨inen, M., Ilmoniemi, R., Penttinen, A., Salminen, J., and Teszner, D. 1984. Somatosensory evoked cerebral magnetic fields from SI and SII in man. Electroencephalogr. Clin. Neurophysiol. 57:254–263. Hari, R., Karhu, J., Ha¨ma¨la¨inen, M., Knuutila, J., Salonen, O., Sams, M., and Vilkman, V. 1993. Functional organization of the human first and second somatosensory cortices: A neuromagnetic study. Eur. J. Neurosci. 5:724–734. Hashimoto, I. 1987. Somatosensory evoked potentials elicited by air-puff stimuli generated by a new high-speed air control system. Electroencephalogr. Clin. Neurophysiol. 67:231–237. Hoshiyama, M., Kakigi, R., Koyama, S., Kitamura, Y., Shimoio, M., and Watanabe, S. 1995. Somatosensory evoked magnetic fields after mechanical stimulation of the scalp in humans. Neurosci. Lett. 195:29–32. Hoshiyama, M., Kakigi, R., Koyama, S., Kitamura, Y., Shimojo, M., and Watanabe, S. 1996. Somatosensory evoked magnetic fields following stimulation of the lip in humans. Electroencephalogr. Clin. Neurophysiol. 100:96–104. Ishiko, N., Hanamori, T., and Murayama, N. 1980. Spatial distribution of somatosensory responses evoked by tapping the tongue and finger in man. Electroencephalogr. Clin. Neurophysiol. 50:1–10. Iwamura, Y., Tanaka, M., Sakamoto, M., and Hikosaka, O. 1983. Functional subdivisions representing different finger regions in area 3 of the first somatosensory cortex of the conscious monkey. Exp. Brain Res. 51:327–337. Iwamura, Y., Tanaka, M., Sakamoto, M., and Hikosaka, O. 1993. Rostrocaudal gradients in the neuronal receptive field complexity in the finger region of the alert monkey’s postcentral gyrus. Exp. Brain Res. 92:360–368. Kakigi, R. 1994. Somatosensory evoked magnetic fields following median nerve stimulation. Neurosci. Res. 20:165–174. Kakigi, R., and Shibasaki, H. 1991. Effects of age, gender, and stimulus side on scalp topography of somatosensory evoked potentials following median nerve stimulation. J. Clin. Neurophysiol. 8:320–330. Kakigi, R., and Shibasaki, H. 1992. Effects of age, gender, and stimulus side on the scalp topography of somatosensory evoked potentials following posterior tibial nerve stimulation. J. Clin. Neurophysiol. 9:431–440. Kakigi, R., Koyama, S., Hoshiyama, M., Shimojo, M., Kitamura, Y., and Watanabe, S. 1995. Topography of somatosensory evoked magnetic fields following posterior tibial nerve stimulation. Electroencephalogr. Clin. Neurophysiol. 95:127–134. Mogilner, A., Grossman, J. A., Ribary, U., Joliot, M., Volkmann, J., Rapaport, D., Beasley, R. W., and Llinas, R. R. 1993. Somatosensory cortical plasticity in adult humans revealed by magnetoencephalography. Proc. Natl. Acad. Sci. USA 90:3593–3597. Mogilner, A., Nomura, M., Ribary, U., Jagow, R., Lado, F., Rusinek, H., and Llinas, R. 1994. Neuromagnetic studies of the lip area of primary somatosensory cortex in humans: Evidence for an oscillotopic organization. Exp. Brain Res. 99:137–147. Narici, L., Modena, I., Opsomer, R. J., Pizzella, V., Romani, G. L., Torrioli, G., Traversa, R., and Rossini, P. M. 1991. Neuromagnetic somatosensory homunculus: A non-invasive approach in humans. Neurosci. Lett. 121:51–54. Penfield, W., and Boldrey, E. 1937. Somatic motor and sensory
386
NAKAMURA ET AL.
representation in the cerebral cortex of man as studied by electrical stimulation. Brain 60:389–443. Penfield, W., and Rasmussen, T. 1957. The Cerebral Cortex in Man. Macmillan, New York. Rossini, P. M., Deuschl, G., Pizzella, V., Tecchio, F., Pasquarelli, A., Feifel, E., Romani, G. L., and Lu¨cking, C. H. 1996. Topography and sources of electromagnetic cerebral responses to electrical and air-puff stimulation of the hand. Electroencephalogr. Clin. Neurophysiol. 100:229–239. Sarvas, J. 1987. Basic mathematical and electromagnetic concepts of the biomagnetic inverse problem. Phys. Med. Biol. 32:11–22. Shimojo, M., Kakigi, R., Hoshiyama, M., Koyama, S., Kitamura, Y., and Watanabe, S. 1996. Differentiation of receptive fields in the sensory cortex following stimulation of various nerves of the lower limb in humans: A magnetoencephalographic study. J. Neurosurg. 85:255–262. Suk, J., Ribary, U., Cappell, J., Yamamoto, T., and Llina´s, R. 1991. Anatomical localization revealed by MEG recordings of the human somatosensory system. Electroencephalogr. Clin. Neurophysiol. 78:185–196. Sutherling, W. W., Crandall, P. H., Darcey, T. M., Becker, D. P., Levesque, M. F., and Barth, D. S. 1988. The magnetic and electric
fields agree with intracranial localizations of somatosensory cortex. Neurology 38:1705–1714. Werner, G., and Whistel, B. L. 1973. Functional organization of the somatosensory cortex. In Handbook of Sensory Physiology (H. Autrum, R. Jung, W. R. Loewenstein, D. M. Mackay, and H. Teuber, Eds.), pp. 621–700. Springer-Verlag, Berlin. Wood, C. C., Spencer, D. D., Allison, T., McCarthy, G., Williamson, P. D., and Goff, W. R. 1988. Localization of human sensorimotor cortex during surgery by cortical surface recording of somatosensory evoked potentials. J. Neurosurg. 68:99–111. Woolsey, C. N., Erickson, T. C., and Gilson, W. E. 1979. Localization in somatic sensory and motor areas of human cerebral cortex as determined by direct recording of evoked potentials and electrical stimulation. J. Neurosurg. 51:476–506. Yang, T. T., Gallen, C. C., Schwartz, B. J., and Bloom, F. E. 1993. Noninvasive somatosensory homunculus mapping in humans by using a large-array biomagnetometer. Proc. Natl. Acad. Sci. USA 90:3098–3102. Yang, T. T., Gallen, C., Schwartz, B., Bloom, F. E., Ramachandran, V. S., and Cobb, S. 1994. Sensory maps in the human brain [letter]. Nature 368: 592–593.
NEUROIMAGE ARTICLE NO.
Somatosensory Homunculus as Drawn by MEG Akinori Nakamura,*,1 Takako Yamada,† Atsuko Goto,† Takashi Kato,* Kengo Ito,* Yuji Abe,† Teruhiko Kachi,† and Ryusuke Kakigi‡ *Department of Biofunctional Research, National Institute for Longevity Sciences, 36-3 Gengo, Morioka-cho, Obu, Aichi 474, Japan; †Department of Neurology, Chubu National Hospital, 36-3 Gengo, Morioka-cho, Obu, Aichi 474, Japan; and ‡Department of Integrative Physiology, National Institute for Physiological Sciences, Myodaiji, Okazaki 444, Japan Received November 12, 1997
We studied a detailed somatosensory representation map of the human primary somatosensory cortex using magnetoencephalography. Somatosensory-evoked magnetic fields following tactile stimulation of multiple points in the right hemibody (including the tongue, lips, fingers, arm, trunk, leg, and foot) were analyzed in five normal subjects. We were able to estimate equivalent current dipoles (ECDs) following stimulation of the tongue, lips, fingers, palm, forearm, elbow, upper arm, and toes in most subjects and those following the stimulation of the chest, ankle, and thigh in one subject. The ECDs were located in the postcentral gyrus and generally arranged in order along the central sulcus, which is compatible with the somatosensory ‘‘homunculus.’’ Linear distances, averaged in five subjects, from the receptive area of the thumb to that of the tongue, little finger, forearm, upper arm, and toes were estimated to be 2.42 6 0.28, 1.25 6 0.28, 2.21 6 0.72, 2.75 6 0.63, and 5.29 6 0.48 cm, respectively. The moment of each ECD, which suggested the size of the cortical areas responsive to the stimulation, was also compatible with the bizarre proportion of the homunculus with a large tongue, lips, and fingers. According to these results, we were able to reproduce a large part of the somatosensory homunculus quantitatively on an individual brain MRI. r 1998 Academic Press
INTRODUCTION Since the landmark study of Penfield and Boldrey (1937), various extensive studies have revealed the one-to-one somatotopic representation of the contralateral body surface in the human postcentral cortex, which has been known as the somatosensory ‘‘homunculus.’’ Most studies were performed by means of invasive techniques under neurosurgical conditions, such as direct electrical stimulation of the cortical surface 1 To whom correspondence should be addressed. Fax: 81-562-446596. E-mail: [email protected].
(Penfield and Boldrey, 1937; Penfield and Rasmussen, 1950) and somatosensory-evoked responses recorded on the cortical surface (Woolsey et al., 1979; Wood et al., 1988; Baumgartner et al., 1991a; Allison et al., 1989). However, these techniques are very restricted in application. It is impossible to use these techniques for evaluation of nonneurosurgical patients or normal subjects. A large number of noninvasive studies using scalprecorded somatosensory evoked potentials (SEPs) have supported the somatotopic organization in the somatosensory cortex (Duff, 1980; Desmedt and Bourguet, 1985; Kakigi and Shibasaki, 1991, 1992). However, the scalp-recorded electroencephalogram (EEG) could not provide enough resolution to estimate the precise electrical source location in the brain. This is because the influence of volume currents on scalp-recorded EEG severely affects precise source identification. Furthermore, large interindividual variability of intervening tissues, including scalp, skull, and cerebrospinal fluid, makes it difficult to estimate the dipole location. Another noninvasive technique for investigating the bioelectrical functions of the brain, magnetoencephalography (MEG), has been developed during the past 25 years. MEG has several advantages over EEG in localizing cortical sources, because the magnetic fields recorded on the scalp are less affected by volume currents and anatomical inhomogeneities (Ha¨ma¨la¨inenet al., 1993). Based on its spatial resolution of a few millimeters, several reports have demonstrated the somatotopic projection of the hand (Brenner et al., 1978; Sutherling et al., 1988; Bamgartner et al., 1991b; Suk et al., 1991; Hari et al., 1984; Kakigi et al., 1994), foot (Kakigi et al., 1995; Shimojo et al., 1996), lips (Mogilner et al., 1994; Hoshiyama, 1996), and scalp (Hoshiyama, 1995) in the human cortex using somatosensory-evoked magnetic fields (SEFs). These reports partially reconfirmed the homunculus. There have been several reports which covered a wider area of the homunculus. Narici et al. (1991) studied multiple receptive fields using SEF following median, femoral, tibial,
377
1053-8119/98 $25.00 Copyright r 1998 by Academic Press All rights of reproduction in any form reserved.
378
NAKAMURA ET AL.
and pudendal nerve stimulations. Hari et al. (1993) demonstrated representations of the leg, hand, fingers, lips, and tongue which agreed with the homunculus. Yang et al. (1993) extensively studied somatosensory mapping of various points in the region of the face and upper limb. However, there are no reports which cover the ‘‘whole body’’ of the homunculus. The objective of our study was to investigate the quantitative, wholebody representation map of the human primary somatosensory cortex and to evaluate the capacities and limitations of today’s multichannel MEG system. We used tactile stimulation for this study. Most of the previous studies were performed using electric stimulation. However, these unnatural stimuli activate a large number of fibers from both deep and superficial receptors, bypassing the peripheral receptors. Tactile stimuli are more physiological than electric stimuli and more suitable for studying the functional map of the cortex involved in single modality of sensation. A proper tactile stimulus activates cutaneous mechanoreceptors selectively and causes more natural sensation without pain or discomfort. METHODS The subjects were five male volunteers ages 21 to 36 years. All subjects were healthy right-handed Japanese. Informed consent was obtained from all five subjects prior to the experiments. An air-puff-derived tactile stimulator (Hoshiyama et al., 1995), which provides a light, superficial pressure stimulus to the skin surface, was used for the stimulation. The skin contact area was a circular rubber bladder 1 cm in diameter, and the intensity of the mechanical stimulation was 40 g/cm2. The rise time was 20 ms as measured from 10 to 90% of the intensity increment. No joint movement was observed in this stimulation. About 40 points in the right hemibody (including the tongue, lips, hand, arm, trunk, leg, and foot) were stimulated in a randomly determined order (Fig. 1). The stimulation device was fixed to each stimulation point at the optimal position at which maximal subjective sensation could be obtained. This point was determined through communication with the subject in each session. The interstimulus interval was randomly jittered, from 400 to 600 ms. SEFs were recorded with a 37-channel biomagnetometer (Magnes, Biomagnetic Technologies, Inc., San Diego, CA). The detection coils of the biomagnetometers were arranged in a uniformly distributed array of concentric circles over a spherically concave surface. The device was 144 mm in diameter. Each coil was 20 mm in diameter and the centers of the coils were 22 mm apart. Each coil was connected to a superconducting quantum interference device. The intrinsic noise level in each channel was ,10 fT/Hz1/2.
FIG. 1. Forty-three stimulated points of the right hemibody. The gray circles indicate stimulus points on the frontside of the body, and black circles indicate stimulus points on the backside of the body. In the cases of arm and leg stimulations, lateral, medial, anterior, and posterior points were stimulated.
When the points in the face and upper limb were stimulated, the recording probe was centered around the C3 (international 10–20 system), and when the points of the lower limb were stimulated, the probe was centered around the Cz. Both probe positions were applied when the points of the trunk were stimulated. Responses were recorded with a 0.1- to 200-Hz bandpass filter, followed by 1- to 100-Hz, and digitized at a sampling rate of 512 Hz. The analysis time was 100 ms before and 200 ms after the stimuli. DC was offset using a prestimulus period (100 ms) as the baseline. At least 150 epochs for each stimulus point were collected and averaged in each session. At least two averages were obtained to ensure reproducibility. Peak latency of the wave component was determined according to the maximum value of the root mean square (RMS) field strength taken over the 37 sensors.
379
SOMATOSENSORY HOMUNCULUS AS DRAWN BY MEG
TABLE 1 Detailed Value of Each ECD
Tongue Lower lip Upper lip Thumb Index finger Middle finger Ring finger Little finger Radial palm Ulnar palm Forearm Elbow Upper arm Chestd Thighd Ankled Toes
Latency
Distancea (cm)
Dipole moment, (nA · m)
Ratiob
Activated areac (mm2 )
35.7 6 5.5 34.5 6 6.9 33.6 6 5.4 46.2 6 3.9 47.2 6 3.0 45.4 6 4.3 46.4 6 5.4 47.5 6 6.0 44.3 6 6.9 46.7 6 4.9 42.1 6 6.9 40.6 6 7.8 38.5 6 4.1 36.4 57.5 71.4 72.5 6 10.5
2.42 6 0.28 1.46 6 0.70 1.17 6 0.53 0 1.05 6 0.58 1.17 6 0.36 1.22 6 0.23 1.25 6 0.28 0.93 6 0.51 1.58 6 0.45 2.21 6 0.72 2.44 6 0.33 2.75 6 0.63 3.89 4.16 5.43 5.29 6 0.48
20.5 6 8.0 17.1 6 3.3 14.9 6 4.8 23.3 6 5.2 16.9 6 5.2 16.5 6 3.8 14.4 6 4.9 15.3 6 4.7 12.0 6 1.9 11.0 6 3.6 9.7 6 2.0 8.4 6 1.6 9.8 6 2.8 5.7 4.4 6.3 12.3 6 3.0
0.880 0.734 0.642 1 0.728 0.709 0.620 0.659 0.516 0.473 0.419 0.361 0.422 0.245 0.189 0.271 0.530
82.0 6 32.0 68.4 6 13.2 59.6 6 19.2 93.2 6 20.8 67.6 6 20.8 66.0 6 15.2 57.6 6 19.6 61.2 6 18.8 48 6 7.6 44.0 6 14.4 38.8 6 8.0 33.6 6 6.4 39.2 6 11.2 22.8 17.6 25.2 49.2 6 12.0
The mean linear distances (mean 6 SD) from the thumb ECD to the other ECDs. Relative ratio of the dipole moment compared to the thumb. c Activated cortical area, estimated as 40 mm2/10 nA · m. d Data obtained from one subject. a b
Source analyses, based on a single moving equivalent current dipole (ECD) model (Sarvas, 1987) in a spherical volume conduction, were applied to magnetic field distribution studies. A local sphere, which was digitized from the parietal portion of each subject’s head, was used for the model. The location (x, y, and z positions), orientation, and dipole moment of the best-fitting single ECD were estimated for each time point. The origin of the head-based coordinate system was the point exactly halfway between the preauricular points. The x axis indicated the coronal plane with positive values to the nasion, the y axis the sagittal plane with positive values in the left, and the z axis the transaxial plane with positive values up. Correlations between the recorded measurements and the values expected from the ECD estimate were calculated. To ensure a strict criterion for dipole fitting, only estimates with a correlation above 0.97 were analyzed. Dipole strength at the peak latency was used as the representative dipole moment of each ECD. In most cases, latencies of the RMS peak and correlation peak were the same or very close. Magnetic resonance images (MRIs) were obtained using a 1.5-T system (Toshiba Medical). T1-weighted images with a contiguous 1-mm slice thickness were used for overlays, with ECD sources determined by MEG. The same anatomical landmarks used to create the MEG head-based 3D coordinate system (the nasion and bilateral preauricular points) were visualized in the MR images by affixing to these points high-contrast cod liver oil capsules (3-mm diameter). The MEG source locations were converted into pixels and slice
values using the MRI transformation matrix, then inserted onto the corresponding MRIs. In order to compare the ECD maps obtained from the different subjects, the coordinate differences from the ECD of the thumb to the ECD of the otherpoints (Dx, Dy, Dz) were calculated. Linear distances between the ECD of the thumb and the ECD of the other points were calculated by the formula ((Dx)2 1 (Dy)2 1 (Dz )2)1/2. RESULTS SEF waveform following tactile stimulation typically showed two major components. We termed them M1 for the earlier in latency and M2 for the later. The M1 was often combined with its subcomponent M18, which was delayed about 10–30 ms from the M1. A small deflection preceding the M1, which we termed M0, was also frequently detected (Fig. 2). We analyzed the M1 in the present study, because this first major component was the most reproducible wave and was thought to originate mainly in area 3b (see Discussion). In all five subjects, high-amplitude M1s were evoked following tactile stimulation of the points on the tongue, lips, and hand. Relatively low-amplitude, but reproducible M1s were recorded following stimulation of the palm, forearm, upper arm, and toes. It was hard to record clear and reproducible M1s following stimulation of the points on the trunk and leg (Fig. 2). The peak latency of the M1 was shortest when the tongue, lower lip, or upper lip was stimulated (,34 ms) and was longest when the toes were stimulated (,72 ms), reflecting the distances from the brain (Fig. 2, Table 1).
380
NAKAMURA ET AL.
FIG. 2. Representative SEF waveforms following tactile stimulation of each point. Waves recorded at each of the 37 sensor points were superimposed on the corrected baseline. The isocontour map of the M1 (right) showed distribution of the magnetic fields. Outward flux is marked by a red gradation, inward flux by green, and the arrow shows the estimated ECD.
Following the tongue, lips, and fingers stimulation, most of the M1 ECDs could be estimated very reliably (more than 99% of correlation). Following the palm, forearm, upper arm, and toes stimulation, most of the ECDs could also be estimated reliably (more than 97%
of correlation). However, in most subjects, reliable M1 ECDs could not be estimated when the trunk or leg was stimulated. Only ECDs following stimulation of the anterior part of the thigh and medial part of the ankle could be estimated with a more than 97% correlation in subject 1. Additionally, one ECD following chest stimulation with a 96.3% correlation in subject 1 was accepted in order to complete the homunculus. These estimated ECDs were located in the left postcentral gyrus. The ECD location to the tongue stimulation was the most inferior, followed by the lips, fingers, and arm, which were gradually shifted to superior and medial, along the central sulcus (Figs. 3A and 3B). The ECDs to the ankle and toes were localized on the midline (Fig. 3C). These findings are more clearly demonstrated by Fig. 4A, which illustrates the averaged coordinate differences between the receptive area of the thumb and that of the other points in the five subjects. Averaged linear distances between the locations of each source were also calculated (Fig. 4B and Table 1). The tongue source was located 2.42 6 0.28 cm (mean 6 SD) from the thumb source. Distances from the area of the thumb to the little finger, forearm, and upper arm were 1.25 6 0.28, 2.21 6 0.72, and 2.75 6 0.63 cm, respectively. The ECD to the radial side of the palm stimulation was estimated close to that of the thumb stimulation, and the ECD to the ulnar side of the palm stimulation was estimated close to the little finger stimulation. No apparent differences could be detected in source locations between the radial, ulnar, anterior, and posterior sides of the forearm stimulation or between the medial, lateral, anterior, and posterior sides of the upper arm stimulation. Similarly, the source locations of toes 1, 2, and 5 were very close and no consistency could be found (e.g., the toe 1 ECD was estimated above the toe 5 ECD in three subjects, but it was estimated below the toe 5 ECD in two subjects). ECD data of the four points around the forearm, elbow, and upper arm, as well as the three points of the toes, were averaged in each subject and used for analyses. The dipole moment of each ECD was also averaged (Table 1). The thumb ECD showed the largest dipole moment (23.3 6 5.2 nA · m) and the tongue ECD was the second (20.5 6 8.0 nA · m). ECDs of the lips and digits showed relatively large dipole moments, while ECDs of the arm, trunk, and leg showed smaller dipole moments. The ECD model approximates a summation of the postsynaptic potentials in the axon dendrites of the pyramidal cells. It is thought that a dipole moment of 10 nA · m would be caused by the synchronous activity of tens of thousands of neurons (Ha¨ma¨la¨inen et al., 1993) and would correspond to 40 mm2 of active cortex (Chapman et al., 1984). According to this hypothesis, about 93 mm2 of the SI cortex was supposed to be shared for the stimulated skin area (about 78.5 mm2) of the thumb, which showed 23.3 nA · m of averaged
SOMATOSENSORY HOMUNCULUS AS DRAWN BY MEG
381
FIG. 3. Estimated M1 ECDs of subject 1 overlapped on his MRI. Displayed MR images are sliced at the position of the ECD of the lower lip (A), thumb (B), and toe 2 (C). The other ECDs were projected onto each slice.
382
NAKAMURA ET AL.
FIG. 4. (A) The coordinate differences between the ECD of the thumb and the ECD of the other points averaged in five subjects. X-position indicates anterior–posterior coordinates with the positive value anteriorly, Y-position indicates lateral–medial coordinates with the positive value laterally, and Z–position indicates vertical coordinates with the positive value superiorly. (B) The linear distances from the ECD of the thumb to the ECD of each other point, averaged in five subjects. The ECDs which locate more inferior than the thumb ECD were demonstrated by negative values (tongue and lips).
dipole moment. Similarly, the presumed cortical area responsive to the stimulation of each part of the body was calculated (Table 1). Taking these estimations into account, we finally drew a somatosensory representation map on subject 1’s 3D brain image (Fig. 5). The size of each ellipse reflected the presumed size of each activated area to the stimulation, demonstrating the weight of each part of the body in the human SI cortex.
DISCUSSION There have been several reports studying SEP or SEF following mechanical stimulation, including tactile and/or vibratory (Ha¨ma¨la¨inen et al., 1990; Yang, 1993; Gallen et al., 1994; Hoshiyama et al., 1995) and air-puff (Hashimoto, 1987; Rossini et al., 1996) stimulation. Forss et al. (1994) studied detailed characteristics
SOMATOSENSORY HOMUNCULUS AS DRAWN BY MEG
of SEF following air-puff stimulation of the middle finger. They demonstrated small deflection N26m (latency around 26 ms), large deflections P42m (43–46 ms) and P66m (66–71 ms), and late components (latency around 100 ms). They estimated ECD locations of earlier components N26m, P42m, and P66m in the SI and late components in bilateral SII and posterior parietal cortices. The N26m, P42m, P66m, and late components were thought to correspond with the M0, M1, M18, and M2 of our results according to latency and morphology. It is thought that tactile stimulation activates mainly area 3b (Iwamura et al., 1983), and the first major component, the M1 in our study, is considered to originate mainly in area 3b (Suk et al., 1991; Forss et al., 1994). MEG analysis is favorable for the sources in area 3b, which is located on the posterior bank of the central sulcus. This is because MEG is most sensitive to brain current tangential to the skull (Ha¨ma¨la¨inen et al., 1993), and dipoles generated in area 3b are mainly tangentially oriented. Additionally, it is known that hierarchical information processing occurs in the rostrocaudal direction within the SI. The neurons in areas 1 and 2 have complex receptive fields, whereas one-to-one somatotopic representation is clearly preserved in area 3b (Iwamura et al., 1983, 1993). Thus, the homuncular organization should be more clearcut in area 3b than in areas 1 and 2. Using MEG and tactile stimulation, we could map a large part of the somatosensory receptive fields on SI, including tongue, lips, fingers, palm, arm, and toes. These representation areas were generally arranged in the above order from inferior to superior, lateral to medial, and anterior to posterior (Figs. 3 and 4). These changes in the coordinates were compatible with the anatomy of the central sulcus and the homunculus. ECD location to the upper lip could be distinguished from that to the lower lip, the former located more superior than the latter in all subjects. This finding also agreed with Penfield’s observations and previous MEG studies (Moligner et al., 1994; Hoshiyama, 1996). Each finger representation area of the thumb, index finger, middle finger, ring finger, and little finger was also distinguishable from the others. They were represented sequentially from thumb to little finger, ascending the postcentral sulcus as noted in previous reports (Baumgartner et al., 1991b; Suk et al., 1991; Hari et al., 1993). The mean linear distances from the thumb ECD to each of the other finger ECDs had very similar numeric values (Fig. 4B, Table 1). We considered that this was because the finger representation area of the somatosensory cortex is convexly curved anteriorly and that the measurement of linear distance cannot reflect precise distance along such a curved portion. This curvature was shown in the MR images (Figs. 3B and 5) and was supported by Fig. 4A, which showed that the index finger ECD was located more anteriorly than the
383
thumb ECD. However, the linear distance between the source of the thumb and that of the little finger (1.25 6 0.28) was well consistent with previous reports, in which the distance was 1.12 to 1.45 cm (Suk et al., 1991), 1.1 6 0.16 cm (Mogilner et al., 1993), or up to 2 cm (Hari et al., 1993). These values were also compatible with the value of 1.5–2 cm, which was studied by direct cortical recording of SEP (Wood et al., 1988). Yang et al. (1993) reported in their MEG study that the ECDs of the palmar pads stimulation were located slightly superior and medial to those of the digits stimulation in one subject. They found converse results in another subject. However, our results showed that the receptive areas of the radial and ulnar side of the palm were estimated close to those of the thumb and little finger, respectively (Figs. 3 and 4, Table 1). These findings were consistent with studies in monkeys (Iwamura et al., 1983). We could map receptive areas to forearm, elbow, and upper arm stimulation and showed that they were located in the above order from inferior to superior and lateral to medial. However, ECD locations following the stimulation of radial, ulnar, anterior, and posterior points of the forearm, which were in a different dermatome, were indistinguishable. Similarly, no apparent difference in ECD location was found following the stimulation of four points around the upper arm. These results suggested that the representation area of the arm might be organized mainly along the distal–proximal axis (from the forearm and elbow to the upper arm) rather than the dermatome. These results were also consistent with studies in monkeys (Werner and Whitsel, 1973). It was difficult to estimate ECDs following the trunk and leg stimulation by the methods presented, because the magnetic fields elicited by these stimulations were, in most cases, too low to obtain a desirable signal to noise ratio. This is probably because the trunk and leg have small representation areas. We considered that these results showed the sensitivity limit of our MEG system, which has an intrinsic noise level of ,10 fT/Hz1/2 in each channel. However, ECD locations following the chest, thigh, and ankle stimulation, which could be estimated only in subject 1, were compatible with the homunculus. Receptive areas to toe stimulation were estimated to be located in the mesial side of the hemisphere. We could not separate the sources of toes 1, 2, and 5 clearly; however, the receptive area for toe stimulation was located deeper than that of the ankle. Penfield demonstrated large cortical representation areas for the tongue, lips, and fingers by bizarre proportions of the homunculus with a large tongue, lips, and fingers. We could confirm these observations by estimating the activated cortical area using the dipole moment of the MEG (Table 1, Fig. 5). The activated areas of the tongue, lips, and fingers were supposed to be larger than those of the palm, arm, body,
FIG. 5. Detailed somatosensory receptive map represented by MEG. The 3D brain image was reconstructed using MRI of subject 1. Each receptive area, which was estimated to be located in the posterior bank of the central sulcus, was projected onto the cortical surface. The size of each ellipse reflects the presumed size of activated cortical area. The ellipses illustrated on the brain image are three times smaller than the actual sizes predicted in Table 1. Note that the receptive area for the toes is in the mesial side of the left hemisphere.
384 NAKAMURA ET AL.
SOMATOSENSORY HOMUNCULUS AS DRAWN BY MEG
and leg. The presumed responsible area of the thumb was about 2, 2.4, and 5.3 times larger than those of the palm, arm, and thigh, respectively. Although there is considerable interindividual variation in the size and shape of the human brain, standardization of the functional somatosensory map will contribute to such studies as plastic reorganization of the human somatosensory cortex (Mogilner et al., 1993; Yang et al., 1994; Elbert et al., 1995). Today’s technologies (just 60 years after the study of Penfield and Boldrey), including MEG, high-resolution MRI, and computer systems, enable us to reproduce the functional somatosensory homunculus on an individual brain image quantitatively by completely noninvasive techniques. ACKNOWLEDGMENT This work was supported by Funds for Comprehensive Research on Aging and Health from the Ministry of Health and Welfare, Japan.
REFERENCES Allison, T., McCarthy, G., Wood, C. C., Darcey, T. M., Spencer, D. D., and Williamson, P. D. 1989. Human cortical potentials evoked by stimulation of the median nerve. I. Cytoarchitectonic areas generating short-latency activity. J. Neurophysiol. 62:694–710. Baumgartner, C., Barth, D. S., Levesque, M. F., and Sutherling, W. W. 1991a. Functional anatomy of human hand sensorimotor cortex from spatiotemporal analysis of electrocorticography. Electroencephalogr. Clin. Neurophysiol. 78:56–65. Baumgartner, C., Doppelbauer, A., Sutherling, W. W., Zeitlhofer, J., Lindinger, G., Lind, C., and Deecke, L. 1991b. Human somatosensory cortical finger representation as studied by combined neuromagnetic and neuroelectric measurements. Neurosci. Lett. 134:103– 108. Brenner, D., Lipton, J., Kaufman, L., and Williamson, S. J. 1978. Somatic evoked magnetic fields of the human brain. Science 199:81–83. Chapman, R. M., Ilmoniemi, R. J., Barbanera, S., and Romani, G. L. 1984. Selective localization of alpha brain activity with neuromagnetic measurements. Electroencephalogr. Clin. Neurophysiol. 58: 569–572. Desmedt, J. E., and Bourguet, M. 1985. Color imaging of parietal and frontal somatosensory potential fields evoked by stimulation of median or posterior tibial nerve in man. Electroencephalogr. Clin. Neurophysiol. 62:1–17. Duff, T. A. 1980. Topography of scalp recorded potentials evoked by stimulation of the digits. Electroencephalogr. Clin. Neurophysiol. 49:452–460. Elbert, T., Pantev, C., Wienbruch, C., Rockstroh, B., and Taub, E. 1995. Increased cortical representation of the fingers of the left hand in string players. Science 270:305–307. Forss, N., Salmelin, R., and Hari, R. 1994. Comparison of somatosensory evoked fields to airpuff and electric stimuli. Electroencephalogr. Clin. Neurophysiol. 92:510–517. Gallen, C. C., Schwartz, B., Rieke, K., Pantev, C., Sobel, D., Hirschkoff, E., and Bloom, F. E. 1994. Intrasubject reliability and validity of somatosensory source localization using a large array biomagnetometer. Electroencephalogr. Clin. Neurophysiol. 90:145–156. Ha¨ma¨la¨inen, H., Kekoni, J., Sams, M., Reinikainen, K., and Na¨a¨ta¨nen,
385
R. 1990. Human somatosensory evoked potentials to mechanical pulses and vibration: Contributions of SI and SII somatosensory cortices to P50 and P100 components. Electroencephalogr. Clin. Neurophysiol. 75:13–21. Ha¨ma¨la¨inen, M., Hari, R., Ilmoniemi, R. J., Knuutila, J., and Lounasmaa, O. V. 1993. Magnetoencephalography. Rev. Mod. Phys. 65: 413–496. Hari, R., Reinikainen, K., Kaukoranta, E., Ha¨ma¨la¨inen, M., Ilmoniemi, R., Penttinen, A., Salminen, J., and Teszner, D. 1984. Somatosensory evoked cerebral magnetic fields from SI and SII in man. Electroencephalogr. Clin. Neurophysiol. 57:254–263. Hari, R., Karhu, J., Ha¨ma¨la¨inen, M., Knuutila, J., Salonen, O., Sams, M., and Vilkman, V. 1993. Functional organization of the human first and second somatosensory cortices: A neuromagnetic study. Eur. J. Neurosci. 5:724–734. Hashimoto, I. 1987. Somatosensory evoked potentials elicited by air-puff stimuli generated by a new high-speed air control system. Electroencephalogr. Clin. Neurophysiol. 67:231–237. Hoshiyama, M., Kakigi, R., Koyama, S., Kitamura, Y., Shimoio, M., and Watanabe, S. 1995. Somatosensory evoked magnetic fields after mechanical stimulation of the scalp in humans. Neurosci. Lett. 195:29–32. Hoshiyama, M., Kakigi, R., Koyama, S., Kitamura, Y., Shimojo, M., and Watanabe, S. 1996. Somatosensory evoked magnetic fields following stimulation of the lip in humans. Electroencephalogr. Clin. Neurophysiol. 100:96–104. Ishiko, N., Hanamori, T., and Murayama, N. 1980. Spatial distribution of somatosensory responses evoked by tapping the tongue and finger in man. Electroencephalogr. Clin. Neurophysiol. 50:1–10. Iwamura, Y., Tanaka, M., Sakamoto, M., and Hikosaka, O. 1983. Functional subdivisions representing different finger regions in area 3 of the first somatosensory cortex of the conscious monkey. Exp. Brain Res. 51:327–337. Iwamura, Y., Tanaka, M., Sakamoto, M., and Hikosaka, O. 1993. Rostrocaudal gradients in the neuronal receptive field complexity in the finger region of the alert monkey’s postcentral gyrus. Exp. Brain Res. 92:360–368. Kakigi, R. 1994. Somatosensory evoked magnetic fields following median nerve stimulation. Neurosci. Res. 20:165–174. Kakigi, R., and Shibasaki, H. 1991. Effects of age, gender, and stimulus side on scalp topography of somatosensory evoked potentials following median nerve stimulation. J. Clin. Neurophysiol. 8:320–330. Kakigi, R., and Shibasaki, H. 1992. Effects of age, gender, and stimulus side on the scalp topography of somatosensory evoked potentials following posterior tibial nerve stimulation. J. Clin. Neurophysiol. 9:431–440. Kakigi, R., Koyama, S., Hoshiyama, M., Shimojo, M., Kitamura, Y., and Watanabe, S. 1995. Topography of somatosensory evoked magnetic fields following posterior tibial nerve stimulation. Electroencephalogr. Clin. Neurophysiol. 95:127–134. Mogilner, A., Grossman, J. A., Ribary, U., Joliot, M., Volkmann, J., Rapaport, D., Beasley, R. W., and Llinas, R. R. 1993. Somatosensory cortical plasticity in adult humans revealed by magnetoencephalography. Proc. Natl. Acad. Sci. USA 90:3593–3597. Mogilner, A., Nomura, M., Ribary, U., Jagow, R., Lado, F., Rusinek, H., and Llinas, R. 1994. Neuromagnetic studies of the lip area of primary somatosensory cortex in humans: Evidence for an oscillotopic organization. Exp. Brain Res. 99:137–147. Narici, L., Modena, I., Opsomer, R. J., Pizzella, V., Romani, G. L., Torrioli, G., Traversa, R., and Rossini, P. M. 1991. Neuromagnetic somatosensory homunculus: A non-invasive approach in humans. Neurosci. Lett. 121:51–54. Penfield, W., and Boldrey, E. 1937. Somatic motor and sensory
386
NAKAMURA ET AL.
representation in the cerebral cortex of man as studied by electrical stimulation. Brain 60:389–443. Penfield, W., and Rasmussen, T. 1957. The Cerebral Cortex in Man. Macmillan, New York. Rossini, P. M., Deuschl, G., Pizzella, V., Tecchio, F., Pasquarelli, A., Feifel, E., Romani, G. L., and Lu¨cking, C. H. 1996. Topography and sources of electromagnetic cerebral responses to electrical and air-puff stimulation of the hand. Electroencephalogr. Clin. Neurophysiol. 100:229–239. Sarvas, J. 1987. Basic mathematical and electromagnetic concepts of the biomagnetic inverse problem. Phys. Med. Biol. 32:11–22. Shimojo, M., Kakigi, R., Hoshiyama, M., Koyama, S., Kitamura, Y., and Watanabe, S. 1996. Differentiation of receptive fields in the sensory cortex following stimulation of various nerves of the lower limb in humans: A magnetoencephalographic study. J. Neurosurg. 85:255–262. Suk, J., Ribary, U., Cappell, J., Yamamoto, T., and Llina´s, R. 1991. Anatomical localization revealed by MEG recordings of the human somatosensory system. Electroencephalogr. Clin. Neurophysiol. 78:185–196. Sutherling, W. W., Crandall, P. H., Darcey, T. M., Becker, D. P., Levesque, M. F., and Barth, D. S. 1988. The magnetic and electric
fields agree with intracranial localizations of somatosensory cortex. Neurology 38:1705–1714. Werner, G., and Whistel, B. L. 1973. Functional organization of the somatosensory cortex. In Handbook of Sensory Physiology (H. Autrum, R. Jung, W. R. Loewenstein, D. M. Mackay, and H. Teuber, Eds.), pp. 621–700. Springer-Verlag, Berlin. Wood, C. C., Spencer, D. D., Allison, T., McCarthy, G., Williamson, P. D., and Goff, W. R. 1988. Localization of human sensorimotor cortex during surgery by cortical surface recording of somatosensory evoked potentials. J. Neurosurg. 68:99–111. Woolsey, C. N., Erickson, T. C., and Gilson, W. E. 1979. Localization in somatic sensory and motor areas of human cerebral cortex as determined by direct recording of evoked potentials and electrical stimulation. J. Neurosurg. 51:476–506. Yang, T. T., Gallen, C. C., Schwartz, B. J., and Bloom, F. E. 1993. Noninvasive somatosensory homunculus mapping in humans by using a large-array biomagnetometer. Proc. Natl. Acad. Sci. USA 90:3098–3102. Yang, T. T., Gallen, C., Schwartz, B., Bloom, F. E., Ramachandran, V. S., and Cobb, S. 1994. Sensory maps in the human brain [letter]. Nature 368: 592–593.