Electroencephalography and clinical Neurophysiology, 84 (1992) 115- 126

115

© 1992 Elsevier Scientific Publishers Ireland, Ltd. 0168-5597/92/$05.00

EVOPOT 91519

H u m a n h a n d a n d lip s e n s o r i m o t o r cortex as studied on electrocorticography * Christoph Baumgartner a,c, Daniel S. Barth d, Michel F. Levesque b and William W. Sutherling a Departments of a Neurology and h Neurosurgery, University of California, Los Angeles, CA 90024 (U.S.A.), c Neurological University Clinic, Vienna (Austria), and d Department of Psychology, University of Colorado, Boulder, CO (U.S.A.) (Accepted for publication: 16 September 1991)

Summary We investigated functional topography of human hand and lip sensorimotor cortex using somatosensory evoked potentials (SEPs) from chronically indwelling subdural grid electrodes (ECoG) in 3 epilepsy patients during stimulation of median nerve, ulnar nerve, and lower lip. We used dipole modeling to determine the cortical location of each peripheral sensory field. The cortical locations were in the postcentral gyrus and showed a clear somatotopic organization from medial superior to lateral inferior in the order: ulnar nerve, median nerve, and lip. The source Iocalizations agreed with the results of cortical stimulations and anatomical features on intraoperative photographs. The cortical regions of median and ulnar nerve each could be modeled by sequential tangential and radial dipoles. The cortical region of lip was different and could be explained mostly by tangential dipoles. These findings suggest a difference in the cortical organization of human lip and hand sensory cortex and are consistent with a larger representation of lip in the posterior bank of central fissure in area 3b than on the gyral surface in area l, similar to findings in macaque. Further studies in a larger population of patients with ECoG or normal subjects with scalp-EEG and MEG are warranted to test this hypothesis. Key words: Sensorimotor cortex; Somatotopic organization; Evoked potential; Etectrocorticogram; Functional anatomy

Since Penfield and Boldrey's landmark study on motor and sensory representations in human cerebral cortex based on direct electrical stimulations of the cortical surface (Penfield and Boldrey 1937), it is known that human primary sensorimotor cortex (SI) is organized in an orderly somatotopic way which has been termed "homunculus" representation of the cutaneous body surface. Direct cortical stimulations are considered the most accurate method for localizing essential cortical functions, have been widely used in neurosurgical patients to delineate focal excisional surgery, and have generated detailed somatotopic maps of sensorimotor cortex (Lesser et al. 1987; Ojemann 1987; Ojemann and Engel 1987). Direct cortical stimulations, however, have several disadvantages. First, they rely on the patient's subjective experiences, require a high degree of cooperation and sometimes yield ambiguous results as stim-

* This research was supported by the Fonds zur F6rderung der wissenschaftlichen Forschung Osterreichs (Erwin Sch(6dinger Stipendium J246M and J334MED and Forschungsprojekt P7434) and by USPHS Grant 1-R01-NS20806.

Correspondence to: Dr. Christoph Baumgartner, Neurological University Clinic, W~ihringer Giirtel 18-20, A-1090 Vienna (Austria). Tel: (0222) 40400-3107.

ulation of motor cortex sometimes elicits sensory experiences and, conversely, stimulation of sensory cortex sometimes elicits movements. Second, cortical stimulations can be performed either intraoperatively under local anesthesia which is inconvenient for the patient and the surgical team or extraoperatively from chronically indwelling subdural grid electrodes which requires an additional craniotomy and is a time consuming procedure. Finally, direct cortical stimulations are non-physiological and may produce sensations in a wide area of the body surface. Somatosensory evoked potentials (SEPs) are an alternative method of studying somatotopic organization of sensorimotor cortex. SEPs have the advantages that they are objective and do not rely on patients' subjective reports, that they can be performed under general anesthesia, and that different parts of the cutaneous body surface can be stimulated selectively. Woolsey et al. demonstrated a topographic relationship between the cutaneous body surface and its cortical representation using SEPs in animals and man (Woolsey 1958, 1981; Woolsey et al. 1979). SEPs recorded on electrocorticography (ECoG) are considered most accurate for localization because ECoG neither is distorted by the skull like scalp-EEG nor loses spatial resolution due to distance like EEG and MEG. SEPs on ECoG have been used mainly to localize central sulcus by

116

stimulating the median nerve (Jasper et al. 1960; Kelly et al. 1965; St6hr and Goldring 1969; Goldring et al. 1970; Celesia 1979; Allison et al. 1980, 1989; Broughton et al. 1981; Lueders et al. 1983; Goldring and Gregorie 1984; Gregorie and Goldring 1984; Sutherling et al. 1988; Wood et al. 1988; Baumgartner et al. 1991a) and only few studies dealt with somatotopic features of SEPs (Woolsey 1958; Woolsey et al. 1979). Exact delineation of cortical hand and face representations is important in neurosurgical patients undergoing surgery adjacent to central sulcus in order to avoid neur01ogic deficits. Despite the wide representation of the face in sensorimotor cortex, reports on SEPs during stimulation of the lips have been scarce (Liiders et al. 1986). Furthermore, an ultimate understanding of SEPs requires knowledge of their underlying neuronal sources which have been under considerable debate (St6hr and Goldring 1969; Celesia 1979; Allison et al. 1980, 1989; Desmedt and Brunko 1980; Desmedt and Cheron 1980, 1982; Papakostopoulos and Crow 1980; Broughton et al. 1981; Chiappa 1983; Lueders et al. 1983; Maugui~re et al. 1983; Desmedt and Bourguet 1985; Wood et al. 1985, 1988; Deiber et al. 1986; Liiders et al. 1986; Slimp et al. 1986; Desmedt et al. 1987, 1990; Sutherling et al. 1988; Baumgartner et al. 1989a, 1991a,b). Therefore, we studied functional topography of human sensorimotor cortex using SEPs from chronically indwelling subdural grid electrodes during stimulation of median nerve, ulnar nerve, and lower lip in 3 epilepsy patients undergoing presurgical evaluation. We used dipole models to study somatotopic organization of the neuronal sources underlying SEPs. We validated our findings by cortical stimulations and by intraoperative photographs using a grid recording matrix.

Methods

Patients and procedures We measured evoked potentials on ECoG during the first 60 msec after shock stimulation of the median nerve, the ulnar nerve, and the lower lip in 3 epilepsy patients. All patients had partial seizures defined by EEG recordings and behavior during seizures. The seizures were medically intractable and the patients were evaluated with subdural grids for definitive localization of the seizure focus. Median and ulnar nerve were stimulated at the wrist. The lower lip was stimulated by two disk electrodes, the cathode was placed at the corner of the mouth, the anode paramedian. The stimulus was delivered by a Grass $88 stimulator (Grass Instrument Company, Quincy, MA) and consisted of monophasic, constant current 0.3 msec pulses. Stimulus

C. B A U M G A R T N E R ET AL.

intensity was above motor threshold for stimulation of median and ulnar nerve and twice sensory threshold for stimulation of the lower lip. ECoG was recorded simultaneously from a rectangular array of 48 chronically indwelling subdural grid electrodes consisting of platinum iridium disks with a diameter of 6 mm and 10 mm spacing center-to-center (PMT Corporation, Minneapolis, MN). ECoG was referenced to a scalp-EEG needle electrode at central vertex. Data were amplified (xl0,000) and filtered (bandpass 1-1000 Hz) using Grass 12A5 amplifiers (Grass Instruments Company, Quincy, MA), digitized at a sampling rate of 4096 Hz (12 bits), and stored digitally for off-line data analysis. Two runs of 250 trials each were superimposed to assess reproducibility. Cortical stimulations were performed in all patients to localize motor, sensory, and language areas according to standard procedures (Lesser et al. 1987; Sutherling et al. 1988).

Data analysis Isopotential maps were calculated at selected latencies. We applied dipole models to study the 3-dimensional location of the neuronal sources underlying the SEPs. Each dipole was uniquely given by 6 parameters i 3 location parameters representing its 3-dimensional location, 2 orientation parameters representing its orientation, and an amplitude parameter representing its strength. The potential distribution generated by a dipole was calculated from these 6 parameters according to standard formulas assuming the brain as a homogeneous sphere (Darcey et al. 1980; Stok et al. 1987). We assumed that inhomogeneities would have little effect on modeling since recordings were made directly from the surface of the brain. Furthermore, the surface of the brain is similar to a sphere in the parieto-frontal region. The shape of the central sulcus was not considered in the model as the effect of fissures has been shown to produce only small distortions of the electric field outside a spherical volume conductor (Cuffin 1985). The goal of the modeling procedure thus was to determine which set of dipole parameters yielded a potential distribution that most closely reproduced the data. The dipole parameters were varied iteratively using the simplex algorithm (Press et al. 1986) until the optimum combination of source location, orientation, and strength parameters was obtained. The dipole was allowed to move within the sphere to minimize the difference between the model and data. Variance accounted for was computed as the squared difference between the model and the data divided by the sum of squares of the data across all recording positions. Dipole locations were calculated at the peak latencies. Finally, we compared the source localizations with the results of direct cortical stimulations and intraoperative photographs using a grid recording matrix.

SOMATOTOPY ON ECoG

Results

Data and isopotential maps The data were reproducible within patients showing a run-to-run variability of 5-10%. Between patients there was some variability in wave forms and peak latencies which could be attributed to age differences and differences in lesion sites (compare Figs. 1, 2, and 3). Median (Figs. 1B, 2B, and 3B) and ulnar SEPs (Figs. 1C, 2C, and 3C) showed prominent peaks at latencies of about 20 and 30 msec with P20-N30 wave forms at the anterior recording electrodes and N20-P30 wave forms at the posterior recording electrodes. These components will be referred to as N20 and P30 in the following. In addition, positive peaks were recorded at about 25 msec and negative peaks at about 35 msec. These components will be referred to as P25 and N35.

All

117

Lower lip SEPs (Figs. 1D, 2D, and 3D) showed prominent peaks at latencies of about 15 and 25 msec, with P15-N25 wave forms at the anterior recording electrodes and N15-P25 wave forms at the posterior recording electrodes, analogous to the N20 and P30 for median and ulnar nerve. Positive P20 and negative N30 components also were seen, analogous to the P25 and N35 for median and ulnar nerve. Peak latencies for the individual components for all patients are shown in Table I. Fig. 4 shows the isocontour maps for patient no. 1. Whereas the median and ulnar N20 and P30 showed tangential dipolar patterns with a phase reversal across central fissure (Fig. 4A, C, E, and G), the P25 and N35 showed radial patterns (Fig. 4B, D, F, and H). The lip N15 and P25 maps had tangential dipolar patterns (Fig. 4I and K). In this patient the lip P20 component showed a map which was similar to the N15 (compare

am

0 sensory

U

SKULL X-RAY WITH GRID Ci

MEDIAN NERVE DI

L

L

ULNAR NERVE

LOWER LIP

Fig. 1. Skull X-ray with electrode grid (A) and somatosensory evoked potentials in response to stimulation of the median nerve (B), the ulnar nerve (C), and the lower lip (D) for patient no. 1. Calibration: horizontal = 10 msec, time scale begins with stimulus onset; vertical = 25 ~V (B, C), and 5 ~V (D); polarities: positive ( + ) up, negative ( - ) down.

118

C. BAUMGARTNER ET AL,

AI

BI

O motor O =ensory

SKULL X-RAY WITH GRID

MEDIAN NERVE

CI

DI

L

L

ULNAR NERVE

LOWER LIP

Fig. 2. Skull X-ray with electrode grid CA) and somatosensory evoked potentials in response to stimulation of the median nerve (BI, the ulnar nerve (C), and the lower lip (D) for patient no. 2. Calibration: horizontal = 10 msec, time scale begins with stimulus onset; vertical = 25 #V (B, C), and 5 gV (D); polarities: positive ( + ) up, negative ( - ) down. TABLE I Results of dipole modeling. Patient

Median nerve

(no.)

Latency (msec)

1

19(N20) 23 (P25) 28 (P30) 34 (N35)

2

3

Dist (ram)

Ulnar nerve

Lower lip

Depth (ram)

Orient (deg)

Var (%)

Latency (msec)

Dist (ram)

Depth (ram)

Orient (deg)

Vat (%)

3 2 3 5

8 2 8 2

73 16 102 160

90.2 73.4 94.5 74.8

18 (N20) 24 (P25) 29 (P30) 36 (N35)

7 7 7 5

6 1 8 3

58 9 122 165

22 (N20) 27 (P25) 31(P30) 37 (N35)

11 11 11 9

5 2 6 2

117 25 56 162

97.6 20 (N20) 79.8 25 (P25) 99.5 32(P30) 82.6 37(N35)

4 1 3 1

8 4 8 1

19 (N20) 26 (P25) 31 (P30) 35 (N35)

6 1 7 3

8 2 9 3

87 28 89 160

82,6 75.2 83.6 91.6

8 5 5 6

8 2 7 3

20 (N20) 26 (P25) 30 (P30) 36(N35)

Latency (msec)

Dist (ram)

Depth (ram)

Orient (deg)

Var (%)

93.7 14(N15) 92.6 20 (P20) 90.6 23 (P25) 95.0 30 (N30)

10 4 5 7

7 14 13 14

61 99 108 72

48.9 69.9 75.0 69.8

110 25 54 154

96.7 16 (N15) 97.0 21 (P20) 96.7 25(P25) 87.0 32(N30)

6 2 4 2

8 5 ll 8

93 129 73 131

93.5 60.7 85.7 70,0

78 11 83 144

89.8 13 (N15) 86.0 19 (P20) 76.7 22 (P25) 89.6 29(N30)

4 3 6 1

15 5 8 2

106 106 102 129

66.6 54.0 45.1 41.1

Dist = distance from central sulcus; depth = depth below cortical surface; var = variance accounted for; orient = orientation: azimuth angie versus radial (radial: 0 < theta < 45, 135 < theta < 180; tangential: 45 < theta < 135).

SOMATOTOPY ON ECoG

Fig. 41 and J), whereas the lip N30 component was similar to the P25 (compare Fig. 4K and 4L). The lip N15 and P20 map peaks also were more elongated than those of the median and ulnar nerve. Figs. 5 and 6 show the isocontour plots for patient no. 2 and patient no. 3, respectively. Despite some differences in the original traces, isopotential maps for median and ulnar nerve exhibited similar features in all 3 patients, namely tangential patterns for N20 and P30 as well as radial patterns for P25 and N35 components. On the contrary, lip isocontour maps showed some differences between patients. In contrast to patient no. 1, patient no. 2 had a different map for the lip P20 than for the lip N15, but the map still had a tangential dipolar pattern (Fig. 51 and J). Analogous to patient no. 1, lip P25 and N30 were similar (Fig. 5K and L). The findings in patient no. 3 were in good agreement to those for patient no. 1, namely similar maps for lip

A.

119

N15 and P20 (Fig. 61 and J) as well as for lip P25 and N30 (Fig. 6K and L).

Source localizations N20 and P30 components of median and ulnar nerve SEPs could be modeled by primarily tangentially oriented dipoles. Dipole orientations were opposite to each other at these latencies. For the P25 and N35 components, we found primarily radially oriented sources which also were of opposite polarity at these latencies. The sources underlying the N20 and P30 components were located deeper compared to those underlying the P25 and N35 components (N20 and P30 components: mean = 7.4 mm, standard error = 0.3 mm; 7.4 + 0.3 mm; P25 and N35 component: 2.6 + 0.3 mm; P < 0.001). In contrast to median and ulnar nerve SEPs, all components of lip SEPs could be modeled by primarily tangential dipoles. Similar to hand SEPs,

a.

o" SKULL X-RAY WITH GRID

Cm

MEDIAN NERVE Oll

_-..----~ ~

L

_..-~--~ ~

~

~

, ~

- ~

L

ULNAR NERVE

LOWER LIP

Fig. 3. Skull X-ray with electrode grid (A) and somatosensory evoked potentials in response to stimulation of the median nerve (B), the ulnar nerve (C), and the lower lip (D) for patient no. 3. Calibration: horizontal = 10 msec, time scale begins with stimulus onset; vertical = 25/zV (B, C), and 5/zV (D); polarities: positive ( + ) up, negative ( - ) down.

120

C. B A U M G A R T N E R E T AL.

dipoles reversed orientations for the N15 and P25 components. There was no consistent difference concerning depth between the sources at different latencies of lip SEPs. Table I shows distance from central sulcus, orientation, depth, and variance accounted for by the sources at the peak latencies. All sources were located within 11 mm of central sulcus, with an average distance of 6.0 + 1.0 mm for the median nerve, of 4.9 + 2.3 mm for the ulnar nerve, and of 4.5 + 0.7 mm for the lip. The lip SEP sources were deeper (9.2 + 1.2 mm) than either median SEP (4.8 + 0.8 mm) or ulnar SEP sources (4.8 + 0.9 mm; P < 0.001). The dipole fits for lip SEPs were worse with more variability (65.0 + 4.6%) than those of either median SEPs (85.5 + 2.6%) or ulnar SEPs (91.0 + 1.7%; P < 0.001). The dipole locations for the individual peak latencies and the results of cortical stimulations for all patients are shown in Figs. 7-9. Source locations are depicted as surface projections, i.e., radial projection of the 3-dimensional intracerebral location of the source from the center of the sphere onto the cortical surface. The sources underlying stimulation of different peripheral fields were well separated from each

other and showed a somatotopic organization with the sources during ulnar nerve stimulation located medial superior, those during stimulation of the lower lips lateral inferior, and those during stimulation of the median nerve in between. Source localizations obtained at different latencies during stimulation of the same peripheral receptive field (N20, P25, P30, and N35 for median and ulnar nerve, as well as N15, P20, P25, and N30 for lower lip) were close to each other and showed no consistent differences concerning the anterior-posterior or the medio-lateral direction. Source localizations showed good agreement with the results of cortical stimulations as the sources underlying median and ulnar SEPs were near electrodes where cortical stimulations elicited sensory responses in the hand and the sources of lip SEPs were near electrodes where cortical stimulations produced lip or face sensations.

Discussion Our findings confirm a somatotopic organization of human sensorimotor cortex. All sources were located

MEDIAN NERVE B.

A.

N20-COMPONENT

C.

P25-COMPONENT

D

P30-COMPONENT

.

~

N35-COMPONENT

ULNAR NERVE E.

F.

N20-COMPONENT

G.

P25-COMPONENT

H.

P30-COMPONENT

N35-COMPONENT

LOWER LIP J.

N15-COMPONENT

K.

P20-COMPONENT

L.

P25-COMPONENT

N30-COMPONENT

Fig. 4. Isopotential m a p s for patient no. 1. N20 and P30 m a p s for median nerve (A, C) and ulnar nerve SEPs (E, G), as well as N15 and P25 maps for lower lip (I, K) show tangential dipolar patterns with phase reversals across central sulcus. P25 and N35 maps for median (B, D) and ulnar nerve (F, H) show a more radial pattern. Lip P20 (J) shows a pattern similar to N15 (I), and lip N30 (L) a pattern similar to P25 (K). Isopotential maps: 10% isocontour lines; amplitudes scaled to m a x i m u m for each map; polarities: positive ( + ) white, negative ( - ) black.

SOMATOTOPY ON ECoG

121

in postcentral gyrus within 11 mm of central sulcus. The localization estimates agreed with the results of cortical stimulations and with anatomical features on intraoperative photographs. Although there was intersubject variability like in previous studies (Goldring et al. 1970; Goldring and Gregorie 1984; Gregorie and Goldring 1984; Wood et al. 1988), data were reproducible within patients with low run-to-run variability and similar source localizations across patients. These findings of a somatotopic organization of human sensorimotor cortex agree with previous findings in primates and humans using direct cortical stimulations and SEPs (Penfield and Boldrey 1937; Woolsey 1958; Kaas et al. 1979; Woolsey et al. 1979; Woolsey 1981; I_~sser et al. 1987). SEPs during lip stimulation have been recorded on scalp EEG in normal subjects and in patients with trigeminal neuralgia (St/Shr and Petruch 1979; Bennett and Jannetta 1980; St~Shr et al. 1981; Buettner et al. 1982; Findler and Feinsod 1982; Salar et al. 1982). Different techniques have produced different results (Chiappa 1983). Reports of lip SEPs on ECoG are

scarce (Liiders et al. 1986). SEP wave forms and latencies of our patients were similar to those of previous large series on scalp-EEG (St/Shr and Petruch 1979; Buettner et al. 1982) and to those of the few reports on ECoG (Liiders et al. 1986). Previous authors, however, studied only wave morphologies and latencies with few channels, without mapping or localization of the underlying neuronal sources. We mapped SEPs using 48 simultaneous channels, applied source localization techniques, and compared the 3-dimensiorlal location of the neuronal sources of lip SEPs to those of median and ulnar nerve SEPs. The findings here on median nerve SEPs are similar to previous studies which suggested that early median SEPs are generated by two neuronal sources, one in the posterior bank of the central fissure corresponding to Brodmann's area 3b producing the N20-P30 components and the other in the anterior crown of the postcentral gyrus corresponding to Brodmann's area 1 producing the P25-N35 components (Allison et al. 1980, 1989; Broughton et al. 1981; Sutherling et al. 1988; Wood et al. 1988; Baumgartner et al. 1991a). In the

MEDIAN NERVE A.

C.

B.

N20-COMPONENT

P25-COMPONENT

D.

P ~ N E ~ r

~ ~ N E N T

ULNAR NERVE °"

F.

E.

N20-COMPONENT

P25-COMPONENT

H.

P30-COMPONENT

N3S-COMPONENT

P25-COMPONENT

~-OOMPONEICr

LOWER LIP J.

N15-COMPONENT

K.

P'a)-COMPONENT

Fig. 5. Isopotential maps for patient no. 2. Median and ulnar nerve maps show similar features as those for patient no. 1 presented in Fig. 4. In contrast to patient no. 1, the lip P20 m a p shows a different pattern and different orientation than the N15 map. In agreement to patient no. 1, lip N30 (L) is similar to P25 (K). Isopotential maps: 10% isocontour lines; amplitudes scaled to m a x i m u m for each map; polarities: positive ( + ) white, negative ( - ) black.

122

C. B A U M G A R T N E R E T AL.

present study, sources underlying N20 and P30 components were about 7.4 mm below the cortical surface which is compatible with activity in the central sulcus. Sources underlying P25 and N35 components were more superficial - - around 2.6 mm - - which is compatible with activity in the crown of postcentral gyrus. Sources were not strictly tangential or radial. This finding agrees with results of other studies (Allison et al. 1980, 1989; Wood et al. 1988; Baumgartner et al. 1989a, 1991a,b) and probably is due to the fact the gyri and sulci are convoluted rather than strictly tangential or radial. The dipole locations and orientations of N20 and P30 were similar suggesting that the neuronal populations generating the N20 and P30 components are similar. This interpretation is different from that of Lueders et al. (1983) who suggested the possibility that N1 (N20) is generated by a horizontal dipole in area 3b, but argued that P2 (P30) is generated in part in areas 1 and 2 (postcentral gyrus), and in part in areas 3a and 4 (precentral gyrus). The findings using magnetic recordings support the interpretation of a similar, tangential source for N20 and P30 (Wood et al. 1985;

Sutherling et al. 1988; Baumgartner et al. 1989a, 1991b). The exact anatomical location of the P25-N35 source is still under considerable debate. Some studies suggested a dipolar source in the anterior crown of the postcentral gyrus generating the P25-N35 component (Allison et al. 1980, 1989; Broughton et al. 1981; Sutherling et al. 1988; Wood et al. 1988; Baumgartner et al. 1991a). Other authors proposed and documented evidence for a precentral P22 radial equivalent dipole (Lueders et al. 1983; Maugui~re et al. 1983; Desmedt and Bourguet 1985; Deiber et al. 1986; Desmedt et al. 1987, 1990; Tiihonen et al. 1989) which persisted in patients with a complete postcentral surgical destruction (Maugui~re et al. 1983). Although in the present study we could localize all sources in postcentral gyrus, we think that attribution of neuronal sources to specific cytoarchitectonic structures is beyond the accuracy limits of our procedure. Recordings of SEPs on ECoG with closer spaced electrodes and most important transcortical recordings during clinically indicated procedures which physically separate the precentral gyrus from the postcentral gyrus may be able to resolve this

MEDIAN NERVE A.

~ N E N T

N20-COMPONENT

D.

C.

B.

N3S-COMPONENT

P30-COMPONENT

ULNAR NERVE F.

N20-COMPONENT

H.

G.

P25-COMPONENT

P30-COMPONENT

N35-COMPONENT

P'~-~PONEICr

~J0-COMOONEKr

LOWER LIP J.

N15-COMPONENT

K.

P20.COMPONENT

Fig. 6. lsopotential maps for patient no. 3. Median and ulnar nerve maps show similar features as those for patient no. 1 and patient no. 2 presented in Figs. 4 and 5, respectively. In agreement to patient no. 1, lip P20 (J) shows a pattern similar to N15 (I), and lip N30 (L) a pattern similar to P25 (K). Isopotential maps: 10% isocontour lines; amplitudes scaled to m a x i m u m for each map; polarities: positive ( + ) white, negative ( - ) black.

S O M A T O T O P Y ON E C o G

123

for median nerve. This suggests a common cortical organization for different receptive fields of the hand. The present findings on lip, however, are different from median and ulnar nerve. Lip SEPs could be explained mostly by tangential sources without the requirement for additional radial sources. The lip SEP peaks which are analogous in latency to the P25 and N35 appeared mostly easily explained by tangential sources. This finding was unexpected. It suggests that lip cortical organization follows a different pattern than that of the hand. This finding is consistent with the presence of a large cortical representation of lower lip in area 3b and a vestigial representation in area 1 of monkey (Kaas et al. 1979). The contribution of a small radially oriented cortical region could have been obscured by our use of large macroelectrodes at 10 mm spacing. Recordings from closer spaced electrodes or transcortical recordings would be useful to further investigate lower lip cortical representation in area 1.

Am

•

sensory

Be Ha

Ar

Ar

Ha

Ha

ulnar nerve

0 0 0

A.

median nerve lower lip Sh Shoulder

Ar Arm Ha Hand

motor

sensory

Fa Face

Fig. 7. Results of source localizations and of cortical stimulations for patient no. 1. A: skull X-rays with electrode grids. Electrodes were 6 m m in diameter and 10 m m center-to-center. Region inside dotted rectangle is enlarged for clarity at the bottom in B. B: enlargements show results of source localizations and cortical stimulations. Motor and sensory responses elicited by cortical stimulation as well as central sulcus are shown. Source locations at the peak latencies are depicted as surface projections (radial projection of the 3-dimensional intracerebral location of the source from the center of the sphere onto the cortical surface) are displayed as circled dots. The sources were located in postcentral gyrus and showed a clear somatotopic organization with the sources underlying ulnar nerve stimulation located medial superior, those underlying stimulation of the lower lip located lateral inferior, and those underlying median nerve stimulation in between. Source localizations obtained at different latencies during stimulation of the same peripheral receptive field (N20, P25, P30, and N35 for median and ulnar nerve, as well as N15, P20, P25, and P30 for lower lip) were close to each other and showed no consistent differences concerning the antero-posterior or the medio-lateral direction.

ambiguity (Allison et al. 1989). This study extends the findings of previous studies which did not investigate functional anatomy of the ulnar nerve cortex. The maps and sources for ulnar nerve were similar to those

S. To

~

Ha

Ha

ulnar nerve

o O.o

median nerve lower lip

eeeo motor

•

sensory

Ha Hand Fa Face

To Tongue

Fig. 8. Results of source localizations and of cortical stimulations for patient no. 2, presented like for patient no. 1 in Fig. 7.

124

C. BAUMGARTNER ET AL.

A.

• motor

(

O sensory

B.

., 0

N, I P a

~

~)

( i - - \ ..

Q

O

•

Ho

.,

~ I~,,~,

O P25~O~O

motor (~) sensory

ulnar nerve medIannerve lowerlip

HaHand FaFace ToTongue

Fig. 9. Results of source localizations and of cortical s,imulations for patient no. 3, presented like for patient no. 1 in Fig. 7.

The sources in lip cortex were deeper than the sources in median or ulnar cortex and the dipole model accounted for less data variance. Furthermore, dipole locations at different latencies showed more variability for lip than for hand SEPs. These findings are not surprising. Cortical stimulations have shown that the lips are represented in a wide area of sensorimotor cortex (Penfield and Boldrey 1937). SEPs during lip stimulation probably arise from an extended cortical area rather than a focal cortical region. The widespread peaks on the lip isopotential maps suggest an extended source (Figs. 4I-L, 5I-L, and 6I-L). Modeling an extended cortical area by an equivalent dipolar source results in a deeper location estimate and in a poorer dipole fit. Although more complex sources like dipole layers or sheets can be represented reasonably well by simple dipoles without introducing significant errors in localization on MEG and E E G (De Munck et al. 1988),

the dipole approximation is worse when recording near an extended source. We recorded directly from cortex on ECoG. These factors probably accounted for the poor dipole fit and deeper localization estimates. Our results, however, should be viewed within the limitations of the procedure. Specifically, it should be noted that we applied single equivalent dipole models as source localization techniques. Early SEPs have been shown to be generated by multiple, simultaneously active sources overlapping both in space and time (Allison et al. 1989; Baumgartner et al. 1991a,b). Modeling activity generated by multiple brain regions with a single dipole can lead to erroneous conclusions (Nunez 1986; Baumgartner et al. 1989b). In the context of the present study, however, the application of a single dipole model at the peak latencies seems to be only a minor limitation for the following reasons. Our previous work demonstrated that there is predominantly a single source active during the N20 component representing the arrival of the sensory volley at the primary sensorimotor cortex, while the subsequent components are generated by multiple sources overlapping both in space and time (Baumgartner et al. 1991a,b). Modeling the N20 component with a single equivalent dipole therefore can be considered as a reliable localization marker for this cortical activity. In the present study, source localizations at the subsequent peaks were close to each other, were clearly separated from the sources underlying stimulation of other peripheral receptive fields and, thus, demonstrated a somatotopic arrangement. Therefore, we believe that our results are real and not due to modeling errors. From a clinical standpoint, SEP localization of lip cortex could be useful in surgical procedures. Hand primary cortex must be avoided to prevent permanent loss of fine finger movements. Although central fissure near hand motor cortex is best identified using median nerve SEPs, the inferior border of hand area could be more accurately delineated by mapping a SEP homunculus which includes structures below thumb area as face and lip (Penfield and Boldrey 1937). This could be useful in planning surgical resections for medically intractable seizures arising from operculum. Furthermore, large lesions may displace common anatomical landmarks and also may cause functional reorganization of sensorimotor cortex (Gregorie and Goldring 1984). These findings also may be useful to study functional reorganization of sensorimotor cortex. However, our results should be viewed within the limitations of the procedure and further studies are warranted to investigate their utility in difficult cases.

We thank Dr. Donald Olson for help with the cortical stimulations. We thank Dr. Charles Wilson, Eric Behnke, and George Mutafyan for technical assistance.

SOMATOTOPY ON ECoG

References Allison, T., Goff, W.R., Williamson, P.D. and Van Gilder, J.C. On the neural origin of early components of the human somatosenspry evoked potential. In: J.E. Desmedt (Ed.), Clinical Uses of Cerebral, Brainstem and Spinal Somatosensory Evoked Potentials. Prog. Clin. Neurophysiol., Vol. 7. Karger, Basel, 1980: 51-68. Allison, T., McCarthy, G., Wood, C.C., Darcey, T.M., Spencer, D.D. and Williamson, P.D. Human cortical potentials evoked by stimulation of the median nerve. I. Cytoarchitectonic areas generating short-latency activity. J. Neurophysiol., 1989, 62: 694-710. Baumgartner, C., Barth, D.S. and Sutherling, W.W. Spatiotemporal modeling of somatosensory evoked magnetic fields. In: S.J. Williamson, M. Hoke, G. Stroink and M. Kotani (Eds.), Advances in Biomagnetism. Plenum Press, New York, 1989a: 161-164. Baumgartner, C., Sutherling, W.W., Di, S. and Barth, D.S. Investigation of multiple simultaneously active brain sources in the electroencephalogram. J. Neurosci. Meth., 1989b, 30: 175-184. Baumgartner, C., Barth, D.S., Levesque, M.F. and Sutherling, W.W. Functional anatomy of human hand sensorimotor cortex from spatiotemporal analysis of electrocorticography. Electroenceph. clin. Neurophysiol., 1991a, 78: 56-65. Baumgartner, C., Sutherling, W.W., Di, S. and Barth, D.S. Spatiotemporal modeling of cerebral evoked magnetic fields to median nerve stimulation. Electroenceph. clin. Neurophysiol., 1991b, 79: 27-35. Bennett, M.H. and Jannetta, P.J. Trigeminal evoked potentials in humans. Electroenceph. clin. Neurophysiol., 1980, 48: 517-526. Broughton, R., Rasmussen, T. and Branch, C. Scalp and direct cortical recordings of somatosensory evoked potentials in man (circa 1967). Can. J. Psychol., 1981, 35: 136-158. Buettner, U.W., Petruch, F., Scheglmann, K. and Stfhr, M. Diagnostic significance of cortical somatosensory evoked potentials following trigeminal nerve stimulation. In: J. Courjon, F. Maugui~re and M. Revol (Eds.), Clinical Applications of Evoked Potentials in Neurology. Raven Press, New York, 1982: 339-345. Celesia, G. Somatosensory evoked potentials recorded directly from human thalamus and Sm I cortical area. Arch. Neurol., 1979, 36: 399-405. Chiappa, K.H. Evoked Potentials in Clinical Medicine. Raven Press, New York, 1983: 203-250. Cuffin, B.N. Effects of fissures in the brain on electroencephalograms and magnetoencephalo-grams. J. Appl. Phys., 1985, 57: 146-153. Darcey, T.M., Ary, J.P. and Fender, D.H. Methods for localization of electrical sources in the human brain. In: H.H. Kornhuber and L. Deecke (Eds.), Progress in Brain Research, Vol. 54. Elsevier, Amsterdam, 1980: 128-134. Deiber, M.P., Giard, M.H. and Maugui~re, F. Separate generators with distinct orientations for N20 and P22 somatosensory evoked potentials to finger stimulation? Electroenceph. clin. Neurophysiol., 1986, 65: 321-334. De Munck, J.C., Van Dijk, B.W. and Spekreijse, H. Mathematical dipoles are adequate to describe realistic generators of human brain activity. IEEE Trans. Biomed. Eng., 1988, BME-25: 421429. Desmedt, J.E. and Bourguet, M. Color imaging of parietal and frontal somatosensory potential fields evoked by stimulation of median and posterior tibial nerve in man. Electroenceph. clin. Neurophysiol., 1985, 62: 1-17. Desmedt, J.E. and Brunko, E. Functional organization of far-field and cortical components of somatosensory evoked potentials in normal adults. In: J.E. Desmedt (Ed.), Clinical Uses of Cerebral, Brainstem and Spinal Somatosensory Evoked Potentials. Prog. Clin. Neurophysiol., Vol. 7. Karger, Basel, 1980: 27-50.

125 Desmedt, J.E. and Cheron, G. Somatosensory evoked potentials to finger stimulation in healthy octogenarians and in young adults: wave forms, scalp topography and transit times of parietal and frontal components. Electroenceph. clin. Neurophysiol., 1980, 50: 404-425. Desmedt, J.E. and Cheron, G. Somatosensory evoked potentials in man: subcortical and cortical components and their neural basis. Ann. NY Acad. Sci., 1982, 388: 388-411. Desmedt, J.E., Nguyen, T.H. and Bourguet, M. Bit-mapped color imaging of human evoked potentials with reference to the N20, P22, P27 and N30 somatosensory responses. Eiectroenceph. clin. Neurophysiol., 1987, 68: 1-19. Desmedt, J.E., Chalklin, V. and Tomberg, C. Emulation of somatosensory evoked potential (SEP) components with the 3-shell head model and the problem of "ghost potential fields" when using an average reference in brain mapping. Electroenceph. clin. Neurophysiol., 1990, 77: 243-258. Findler, G. and Feinsod, M. Sensory evoked response to electrical stimulation of the trigeminal nerve in humans. J. Neurosurg., 1982, 56: 545-549. Goldring, S. and Gregorie, E.M. Surgical management of epilepsy using epidural recordings to localize the seizure focus. J. Neurosurg., 1984, 60: 457-466. Goldring, S., Aras, E. and Weber, P.C. Comparative study of sensory input to motor cortex in animals and man. Electroenceph. clin. Neurophysiol., 1970, 29: 537-550. Gregorie, E.M. and Goldring, S. Localization of function in the excision of lesions from sensorimotor region. J. Neurosurg., 1984, 61: 1047-1054. Jasper, H., Lende, R. and Rasmussen, T. Evoked potentials from the exposed somato-sensory cortex in man. J, Nerv. Ment. Dis., 1960, 130: 526-537. Kaas, J.H., Nelson, R.J., Sur, M., Lin, C.S. and Merzenich, M.M. Multiple representations of the body within primary somatosenspry cortex of primates. Science, 1979, 204: 521-523. Kelly, D.L., Goldring, S. and O'Leary, J.L. Averaged evoked somatosensory responses from exposed cortex in man. Arch. Neurol., 1965, 13: 1-9. Lesser, R.P., Liiders, H., Klem, G., Dinner, D.S., Morris, H.H., Hahn, J.F. and Wyllie, E. Extraoperative cortical functional localization in patients with epilepsy. J. Clin. Neurophysiol., 1987, 4: 27-53. Liiders, H., Dinner, D.S., Lesser, R.P. and Morris, H.H. Evoked potentials in cortical localization. J. Clin. Neurophsiol., 1986, 3: 75-84. Lueders, H., Lesser, R.P., Hahn, J., Dinner, D.S. and Klem, G. Cortical somatosensory evoked potentials in response to hand stimulation. J. Neurosurg., 1983, 58: 885-894. Maugui~re, F., Desmedt, J.E. and Courjon, J. Astereognosis and dissociated loss of frontal and parietal components of somatosenspry evoked potentials in hemispheric lesions. Brain, 1983, 106: 271-311. Nunez, P.L. The brain's magnetic field: some effects of multiple sources on localization methods. Electroenceph. clin. Neurophysiol., 1986, 63: 75-82. Ojemann, G.A. Intraoperative functional mapping at the University of Washington, Seattle. In: J.J. Engel (Ed.), The Surgical Treatment of the Epilepsies. Raven Press, New York, 1987: 635-640. Ojemann, G.A. and Engel, J.J. Acute and chronic intracranial recording and stimulation. In: J.J. Engel (Ed.), Surgical Treatment of the Epilepsies. Raven Press, New York, 1987: 263-288. Papakostopoulos, D. and Crow, H.J. Direct recording of the somatosensory evoked potentials from the cerebral cortex of man and the difference between precentral and postcentral potentials. In: J.E. Desmedt (Ed.), Clinical Uses of Cerebral, Brainstem and Spinal Somatosensory Evoked Potentials. Karger, Basel, 1980: 15-26.

126 Penfield, W. and Boldrey, E. Somatic and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain, 1937, 60: 389-443. Press, W.R., Flannery, B.P., Teukolsky, S.A. and Vetterling, W.T. Numerical Recipes. Cambridge University Press, Cambridge, 1986. Salar, G., lob, I. and Mingrino, S. Somatosensory evoked potentials before and after percutaneous thermocoagulation of the Gasserian ganglion for trigeminal neuralgia. In: J. Courjon, F. Maugui~re and M. Revol (Eds.), Clinical Applications of Evoked Potentials in Neurology. Raven Press, New York, 1982: 359-365. Slimp, J.C., Tamas, L.B., Stolov, W.C. and Wyler, A.R~ Somatosensory evoked potentials after removal of somatosensory cortex in man. Electroenceph. clin. Neurophysiol., 1986, 65: 11-117. St6hr, P.E. and Goldring, S. Origin of somatosensory evoked scalp responses in man. J. Neurosurg., 1969, 31: 117-127. St6hr, M. and Petruch, F. Somatosensory evoked potentials following stimulation of the trigeminal nerve in man. J. Neurol., 1979, 220: 95-98. St6hr, M., Petruch, F. and Scheglmann, K. Somatosensory evoked potentials following trigeminal nerve stimulation in trigeminal neuralgia. Ann. Neurol., 1981, 9: 63-66. Stok, C.J., Meijs, W.H. and Peters, M.J. Inverse solutions based on MEG and EEG applied to volume conductor analysis. Phys. Med. Biol., 1987, 32: 99-104.

C. BAUMGARTNER ET AL. Sutherling, W.W., Crandall, P.H., Darcey, T.M., Becker, D.P., Levesque, M.F. and Barth, D.S. The magnetic and electric fields agree with intracranial localizations of somatosensory cortex. Neurology, 1988, 38: 1705-1714. Tiihonen, J., Hari, R. and H~im~il~iinen, M. Early deflections of cerebral magnetic responses to median nerve stimulation. Electroenceph, clin. Neurophysiol., 1989, 74: 290-296. Wood, C.C., Cohen, D., Cuffin, B.N., Yarita, M. and Allison, T. Electrical sources in human somatosensory cortex: identification by combined magnetic and potential recordings. Science, 1985, 227: 1051-1053. Wood, C.C., Spencer, D.D., Allison, T., McCarthy, G., Williamson, P.D. and Goff, W.R. Localization of human sensorimotor cortex during surgery by cortical surface recording of somatosensory evoked potentials. J. Neurosurg., 1988, 68: 99-111. Woolsey, C.N. Organization of somatic sensory and motor areas of the cerebral cortex. In: H.F. Harlow and C.N. Woolsey (Eds.), Biological and Biochemical Bases of Behaviour. University of Wisconsin Press, Madison, 1958: 63-81. Woolsey, C.N. Cortical Sensory Organization. Humana Press, Clifton, NJ, 1981. Woolsey, C.N., Erickson, T.C. and Gilson, W.E. Localization in somatic sensory and motor areas of human cortex determined by direct recording of evoked potentials and electrical stimulation. J. Neurosurg., 1979, 51: 476-506.