Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Brain (1994),117,117-132 Cortical tongue area studied by chronically implanted subdural electrodes - with special reference to parietal motor and frontal sensory responses Eiichirou Urasaki,l tSumio Uematsu,l Barry Gordon2 ,3 and Ronald P. Lesser 1,2,3 Departments of INeurosurgery and 2Neurology, The Johns Hopkins University School of Medicine and 3The Zanvyl Krieger Mind/Brain Institute, The Johns Hopkins University, Baltimore, Maryland, USA Correspondence to: Ronald P. Lesser, Department of Neurology, Meyer 2-147, The Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, MD 21287-7247, USA Summary Motor and sensory cortical tongue representations were ex amined in 40 patients }vith intractable seizures who underwent chronic subdural electrode grid implantation. Tongue responses were observed in a wide area 4.5 cm anterior and 3 cm posterior to the central sulcus. The distribution of the responses was not influenced by whether the responses were unilateral or bilateral. In patients with fronto-parietal lesions, the tongue motor area was located significantly more superior to the Sylvian fissure and more anterior to the central sulcus than was the tongue motor area ofpatients without organic lesion. Both motor and sensory responses were found outside of the classic precentral or postcentral area on the lateral surface of the cortex. Motor responses (parietal motor responses ') could occur posterior to the central sulcus and, rarely, sensory responses ('frontal sensory responses ') were identified anterior to the central sulcus. These paradoxical parietal motor and frontal sensory responses were seen in 17 out of 40 (42.5 %) patients. Nine of these 17 patients had no organic brain lesion on MRI. Clinical factors, such as patient's age, duration of seizures and cognitive functions (IQ, word fluency score), did not influence the frequency of the paradoxical responses. However, patients with brain lesions showed a tendency to have associated paradoxical responses (P < 0.05). In conclusion, paradoxical responses are not uncommon in epilepsy patients, particularly in those with organic lesions. The physiological and clinical implications of the paradoxical responses are discussed. Key words: epilepsy; cerebral cortex; brain mapping; subdural electrode; tongue Introduction Material and methods Classically, motor representation for the tongue is regarded as being fairly localized and centred -- 1 cm anterior to the central sulcus, with little spread further anterior or posterior to the central sulcus (Foerster, 1936a,b; Penfield and Boldrey, 1937; Penfield and Rasmussen, 1950; Libet et al., 1964; Libet, 1973; Uematsu et al., 1992b) . We utilized chronic subdural recording electrodes to investigate whether this localization is affected in patients with brain lesions, and whether more widespread distribution of responses could be seen in normal behaving humans. The present study documents (i) the wide distribution of tongue motor and sensory responses on the lateral surface of the brain; (ii) the low rate of sensory response to electrical stimulation; (iii) the existence of paradoxical responses, such as parietal nlotor and frontal sensory responses. Ethical approval was received from the joint committee of clinical investigations of the Johns Hopkins Medical Institute for this study and informed consent was given by all patients participating in this study. Details of the surgical technique, grid placement and cortical stimulations are described elsewhere (Lesser et al., 1987; Uematsu, 1992). © Oxford University Press 1994 Patients and grid implant Forty patients were selected fronl 58 undergoing stimulation mapping as part of epilepsy surgery undertaken at the Johns Hopkins Hospital. Eighteen patients were excluded because of uncertain identification of cerebral sulci (four patients) or no tongue responses obtained (14 patients). The 40 patients studied 118 E. Urasaki et al. included 22 males and 18 fen1ales, ranging in age from 9 to 57 years (mean 25.4 ± 10.2 years). The patients were divided into three groups: (i) patients without any organic lesion; (ii) patients with a fronto-parietal lesion; (iii) patients with a temporal lesion. There were 28 patients without organic lesions (24 left, four right) and 12 patients with organic lesions (10 left, two right). The patients with organic lesion included seven fronto-parietal lesions (four astrocytomas, two angiomas and one cortical dysplasia) and five temporal lesions (two astro cytomas, two angiomas and one arachnoid cyst). Lesions were judged present or absent based on the results of MRI and direct intra-operative visualization. Grid and stimulation study The electrode grid consists of multiple electrodes, each with a diameter of 3 mm with an area 2 mm in diameter on each electrode exposed for contact with cortex. The centre-to-centre inter-electrode distance is 10 Illill. Stimulation is delivered with 50 Hz, 0.3 ms duration pulses in 3 - 5 strains. Pulses were generated by a Grass 12 or a Grass S88 dual channel stimulator. In each stimulation procedure, the beginning stimulus intensity is 1 mA and stimulus intensity increases by 0.5 -1 rnA increments to a maximum of 15 rnA or to a level producing after-discharges. Maximum stimulation intensity depends on the observation of motor or sensory phenomena by testing personnel or reported by the patient, or the occurrence of EEG after discharges, which were monitored during the test (Lesser et al., 1987). For a typical electrode stimulation study (in 35 out of 40 patients), an electrode pair was formed between two adj acent electrode contacts in the anterior to posterior direction, usually parallel to the Sylvian fissure (Figs 1 and 2). Stimulation was delivered initially to consecutive adjacent pairs of electrodes. When precise delineation of a critical sensorimotor area was needed, additional testing was performed, using additional electrode combinations in the region of interest (Fig. 2A). In five patients, stimulation was given approximately in the orientation of the central sulcus, i. e. in a superior-to-inferior direction (Fig. 3A). Location of electrodes The location of the electrodes in relation to the main sulci and cortical vessels was noted and photographed intra-operatively. A con1posite brain diagram was made based on the intra operative photographs and postoperative radiological studies for each individual patient. The central sulcus was determined by choosing among the cortical fissures along a modified Rolandic line. First, the conventional Rolandic line as described by Taylor and Haughton (1900) was drawn and then the line was modified by shifting 100 anterior to the conventional line (Fig. 1), since the conventional Rolandic line was established using cadaver brain, which has been shown to displace posteriorly -- 100 from the pre-mortem position of the central sulcus artery seen in angiography (Ring and Waddington, 1967). Using these two lines as gross landmarks of the central sulcus, and with special attention paid to pre- and postcentral sulci, which are often interrupted or connected to other sulci, while the majority of the central sulcus is continuous, the central sulcus was drawn for each patient (Missir et al., 1989; Ono et al., 1990). Compiling data For each pair of electrodes stimulated, the cortical location of a given response was chosen to be the electrode closest to the central sulcus and Sy1vian fissure. When only part of a disc electrode of the pair contacted (but did not cross) the central sulcus, the closest electrode was said to be the location of the response. However, when the electrode pair crossed the central sulcus, the electrode located further away from the central sulcus was used for statistical calculation. This electrode was used because, when a tongue response was obtained by stimulation of the pair of bipolar electrodes crossing the central sulcus, there was a greater possibility that the point eliciting the response was included in the stimulated area of larger volume. We separated these two categories of electrodes (touching or crossing the central sulcus) and did not use them for analysis of sensory responses in the frontal region and motor responses in the parietal region, because they could possibly stimulate both precentral and postcentral gyri. The cortical area was divided into 1 x 1.5 em zones, delineated by lines paralleling the Sylvian fissure and falling 1 em apart, as well as lines paralleling the central sulcus and falling 1.5 em apart. The number of responses was counted in each zone. Stin1ulation using a distant inactive reference was not utilized because of time considerations, but we recognize the possibility of some ambiguity concerning the electrode from which the actual response was obtained in certain cases. In some patients, individual electrodes were combined into more than one bipolar pair (Figs 2A and 3A). If there was uncertainty about stimulus response on a particular electrode contact during serial additional stimulation, the response was not counted in the statisticentral sulcus. For example, stimuli to electrodes BI-Cl, Cl- Dl, Dl- El and El- Fl in Fig. 2A elicited motor responses. Since no response occurred with stimulation of Al - Bland F 1- G 1, we could conclude that responses to stimulation of B1- C 1 and E 1- F 1 had to be due to effects at C 1 and E 1, respectively. The D 1 electrode was not counted as a response site, however, because we could neither prove nor disprove D 1 as an origin of the motor response elicited. These procedures certainly underestin1ate the number of responses and the extent of the response distribution, but they can be compared with the results obtained from mono polar stimulation (Penfield and Boldrey, 1937; Penfield and Ras mussen, 1950; Picard and Olivier, 1983; Luders et al., 1987) and bipolar electrodes separated by < 1 cn1 (Libet et al., 1964; Libet, 1973). Responses were defined as (i) motor arrest (stopping of voluntary rapid alternating tongue n10vements); (ii) motor response (localized involuntary movement, pulling back or twitching of the tongue, in response to cortical stimulation); Sensory-motor cortical tongue representation 1 Fig. 1 Location of the electrodes on the implanted grid is shown. After tracing cortical sulci and gyri, as determined from intraoperative photographs, on lateral skull X-ray films, the Rolandic line is measured by the original method of Taylor and Haughton (1900), and a line 10° anterior to the original line also is drawn. In this patient, the central sulcus is located between the two lines. (Electrodes on the inferior temporal lobe and temporal base are not shown.) The patient was a 36-year-old, right-handed man, with left-dominant hemisphere as determined by the Wada test, a verbal IQ of 111, performance IQ of 92, full-scale IQ of 101, and word fluency score of 7.7. Seizure onset was at age 15 years. His MRI and CT scan were normal. Left temporal seizure focus was detected. Open circles indicate disc electrodes that did not elicit tongue responses; filled circles indicate electrode locations where motor responses were elicited. A small filled circle in a square means a mixed response. Thin lines connect two electrodes that evoked no tongue responses. Thick lines between electrodes indicate an electrode pair that evoked some tongue responses. The location of a given response is chosen to be the electrode closest to the central sulcus, e.g. G I in G 1- HI, 02 in C2 - 02, 03 in C3 - 03, and E3 in E3 - F3 of bipolar electrodes. Electrode pairs E I - F I and E2 - F2 cross the central sulcus (CS) and because the division of the inter-electrode line, as divided by the central sulcus, is larger from the central sulcus to E I and E2, these sites were selected as the stimulus locations. Results from cortical stimulations are as follows. AI-BI (15 rnA), no response; CI-DI (14 rnA), deficits in comprehension, reading passages, reading single words, responsive naming and spontaneous speech; E 1- F I (3 rnA), tongue pulled in and unable to speak; GI-HI (10 rnA), tongue pulled in and jaw twitch to right side, deficit in reading a passage; A2-B2 (10 rnA), no response; C2-D2 (14 rnA), tongue pulled in; E2-F2 (6 rnA), tongue pulled in, jaw pulled to right and unable to speak; G2 - H2 (12 rnA), no response; A3 - B3 (15 rnA), no response; C3-D3 (15 rnA), tip of tongue curled up and deficit in reading passages; E3-F3 (II rnA), tongue pulled in and bilateral tongue sensation; G3-H3 (15 rnA), deficits in all speech and language functions; from A4- B4 to G6- H6 (10-15 rnA), no response. 119 120 E. Urasaki et al. Fig. 2 Tongue responses obtained from four patients without organic lesion. A square indicates an electrode where a sensory response was elicited. An X-n1ark in an open circle means tongue motor arrest response. Other symbols are the same as in Fig. 1. A~ example of parietal motor response is shown. Stimulation of electrode pairs D 1 - Eland E 1- F 1 produced tongue motor responses (pulling to right) ~ as well as right thumb twitching and a tingling sensation behind the right side of the lips and teeth. In this case, grid sizes were 6 x 8 over the lateral convexity ~ 2 x 8 underneath on the basiomesial temporal lobe and 4 x 5 over the frontal pole of the left language-dolninant hemisphere. The patient was a 19-year-old left-handed man with complex partial seizures. The Wada test showed that the language areas were represented bilaterally. The MRI scan was negative for the organic brain lesion. B l-C 1, tongue and jaw pulls to right~ and right hand finger nloven1ent stops~ C 1- D 1 ~ same as above; A2 - B2, tongue pulls in, right fingers bend at first knuckle, deficit in reading phrase and comprehension; B2 -C2~ tongue n10ven1ent stops, mouth and jaw open, right finger movement stops; C2 - D2~ tingling on tip of tongue, tongue pulls to right. jaw pulls to right~ right finger movement stops; D2 - E2, tongue pulls to right, jaw pulls to right, tingling on right side of jaw; E2 - F2, tongue pulls to right, mouth pulls to right, tingling in right side of mouth; C3 - D3. tongue pulls in; D3 - E3~ tongue pulls in~ jaw pulls down; E3 - F3, tingling in tongue on both sides. B~ example of frontal sensory response. Upon stimulation through electrode pair C3 - D3 ~ the patient reported tingling on the right side of the tongue. The D3 electrode contact is located just anterior to the central sulcus. (The diagram shows only the main 6 x 8 grid over the left language-dominant hemisphere.) The patient was a 23-year-old won1an with the diagnosis of refractory seizure disorders. The MRI scan was negative for this organic brain lesion. C 1- D 1, tongue retracts, mouth pulls to right, deficit in spontaneous speech; E2 - F2, tingling on the right side of tongue~ tongue pulls to right. c, a 24-year-old, right-handed woman with complex partial seizures. The Wada test showed bilateral representation of the language area. Mixed motor and sensory response was seen beyond the line 3 cm superior to Sylvian fissure (FI)~ which also implies the existence of frontal sensory response. El- FI, right side of tongue and mouth twitch and tingle, right thumb tingles; C2 - D2~ tongue pulls in, can~t talk; E2-F2~ tongue pulls in, can~t talk; C3-D3, tongue movement stops, mouth and jaw pull to right~ can't talk; E3 - F3. tongue pulls in and tingle occurs bilaterally. D~ a 22-year-old man with temporal lobe epilepsy. Left language-dominant hemisphere. In the temporal lobe ~ one motor arrest (D3) and one motor response (D4) were elicited. C 1- D 1, right tongue and facial twitching; El- Fl, tongue pulls back, right n10uth opens, right lip twitches, right finger movement stops, right hand tingles; C3 - D3. tongue movement stops~ mouth and jaw pull to right, right finger moven1ent stops, left throat tingles; E3 - F3, tongue pulls in~ C4 - D4. tongue pulls back. Sensory-motor cortical tongue representation 121 Fig. 3 Tongue responses obtained from two patients with organic lesions. Shaded areas indicate the lesions, as revealed by sagittal MRI scan. A. a 36-year-old woman. left hemisphere dominant for language, with seizures secondary to left prefrontal cavernous angioma. Although they were part of a mixed response, two frontal sensory responses were seen (E5 and E7). Tongue motor responses were also obtained more than 4 cm superior to the Sylvian fissure (E4, F4). E3-E4, tongue pulls to right; F3-F4, tongue pulls to right; E2-F4, tongue pulls to right. right side of face pulls to right; 05 - E5, tingling in right side of tongue, jaw and cheek, tongue pulls to right; E5-E6. tingling on right side of tongue, numbness on right side of jaw, tongue pulls to right; C6-06, tongue pulls to right. 05-06, lOngue and face pull to right; F5 - F6. tingling on the tip of right side of tongue. tongue pulls to right, right side of mouth twitches; G5 - G6. tongue movement stops, right side of jaw twitches; 07 - E7, tongue pulls to right, bilateral tongue tingle; E7 - F7, same as 07-E7: F7-G7. same as 07-E7. B, a 22-year-old man, left hemisphere dominant for language, with seizures secondary to a low grade astrocytoma in left franta-parietal region. Motor and sensory responses were seen above the line 4 cm superior to the Sylvian fissure. but no paradoxical responses were obtained in this case. EI-FI, tongue and mouth pull to right; C2-02, tongue pulls to right; G2-H2. tingling in right side of tongue and lip. (iii) sensory response (such as distinct tingling and numb ness): (iv) mixed response (simultaneous motor and sensory responses). Awareness of tongue movement was not included as a sensory response in this study. Discrete tongue sensation localized in the tongue ipsilateral to the side of stimulation was not included in this study since this is likely to be of trigeminal origin (Lesser et al., 1985). The location, number and rate of responses were examined and statistically analysed (Student's non-paired t test, XC test, Fisher's exact test, McNemar's test and multiple comparisons with Bonferroni's correction) (Armitage, 1971; Itho, 1984; Tanaka and Tarumi, 1986). Results Tongue response In the 40 patients, a total of 586 electrode pairs were placed over the lateral convexity, within the cortical area from Sylvian fissure to 5 cm superior to the Sylvian fissure and from 4.5 cm anterior to 4.5 cm posterior to the central sulcus. Patients included both those with and those without organic lesion. One hundred and thirty of 586 electrode pairs (22.2%) produced tongue responses. Seventy of the 130 (53.8%) were motor responses, 26 (20%) were motor arrest, 23 (17.7%) were mixed responses, and II (8.5%) were sensory responses. In patients without organic lesion, 88 tongue responses were obtained: 48 (54.5%) motor, 19 (21.6%) motor arrest, 15 (17.1 %) mixed and six (6.8%) sensory. The remaining 42 tongue responses were obtained from the patients with organic lesions in either fronto-parietal or temporal regions. Corlical tongue response in relation to central sulcus In patients both with and without organic lesion, 57 out of 70 (81.4 %) motor responses were observed anterior to the cen tral sulcus (frontal lobe). The remaining 13 (18.6%) motor responses were observed posterior to the central sulcus (parietal lobe) (Figs I and 2A). With regard to sensory responses, seven out of II (63.6%) were located posterior to the central sulcus and the remaining four (36.4 %) were observed in the frontal lobe, anterior to the central sulcus (Fig. 28). Seventeen (65%) of the 26 motor arrest responses and 14 (61 %) of the 23 mixed responses were observed in the frontal lobe anterior to the central sulcus. In patients without organic lesion, 39 out of 48 (8 1%) motor responses were anterior to the central sulcus and the remaining nine (19 %) were observed posterior to the central sulcus (Table IA). Of the sensory responses, five out of six (83%) were located posterior to the central sulcus and the remaining one (17 %) was anterior to the central sulcus. Twelve (63 %) of the 19 motor arrest responses and eight (53 %) of the 15 mixed responses were observed in the frontal lobe (Table IA). 122 E. Urasaki et al. Extent of cortical tongue representation (c) Tongue responses in patients with temporal lesions In all patients, with and without organic lesion, the tongue responses were distributed along a 7.5 cm anterior-to-posterior segment (4.5 cm anterior to the central sulcus and 3 cm posterior to the central sulcus) (Table lA -e). Frontal Parietal - ~- A M X S 0 8 0 0 A M X S 0 4 0 0 (40) (22) -- Table 1 (A - C) Distribution of tongue responses by type in patients without organic lesions (A), with fronto-parietal organic lesions (B) and with temporal lesions (C) - - - FRONTAL AREA +4.5an A ~ PARIETAL AREA +3.0 X S A +1.5 CS ~~-F-~ X -1.5 SAM X -3.0 S M X -4.5 ~ S X S (A) Tongue responses in patients without organic lesions Frontal 12 39 (1) Parietal (0) (1) 8 (148) - 1 - - - - - (0) - - - - - - - 1 - (3) 1---- b FRONTAL AREA PARIETAL AREA -----f----- - A It! +3.0 X S A cs +1.5 It! S X A M X S -1.5 A M S X -3.0 A M X -4.5 A S M X S (4) - 2 - - 3 - - 1 - (3) (3) +4.5an - - (4) (1) (0) (0) (2) (1) (1) (0) (3) (4) - (0) - 1 - - 2 - (7) (3) (4) (5) - (0) (3) - - - 1 (4) - (3) (0) - - - (II SF (0) (5) 1 (7) (0) (2) (10) (4) (4) 1 2 1 2 I 9 3 11 1 3 (9) (l) 1 (10) (1) ---~-- 1 3 2 3 I (22) (16) I 2 I 4 2 (27) (5) (22) (2) _ .. - (18) 3 (25) I (27) 2 (34) 2 I (30) 2 I 2 9 (16) I 2 (16) (2) (19) (3) 2 4 (33) (31) SF (B) Tongue responses in patients with fronto- parietal lesions Frontal A M 5 10 Parietal X S A M X S 6 3 2 0 2 2 (86) (37) ----~-- ~._-------- FRONTAL I AREA----~--+---- PARIETAL AREA b +4.5an A M +1. 5 +3.0 X SAM X - - I (6) - 1 I - I - 2 - - SF 1 - - - 1 ~---~ - - (3) 1 1 1 - - I 3 I - - - 1 - - - 1 M XS (1) (0) (1) (0) (4) (l) - (5) 1 -4.5 A 1 1 (8) 1 XS (3) (6) - -3.0 SAM (3) - (6) (5) - X (9) I 1 SAM I (7) (4) (5) 3 (4) (8) - I X (6) (3) - -1. 5 CS SAM - - - (4) - (6)......I... - - (6)"""' 1 (4) """" - - - (3) ---.J For each table: A = total number of tongue responses in patients without organic lesions, by type; B = cortical tongue responses, distance from central sulcus (CS) and Sylvian fissure (SF), in centimetres. Upper numbers in each column indicate the number of responses per electrode pair. Lower numbers in parentheses represent total electrode pairs at that cortical segment. For exan1ple, in A, at the cortical segment 3 -4.5 em anterior to the central sulcus and 3 -4 cm superior to the Sylvian fissure, there is a motor arrest (A) response upon stimulation of seven electrode pairs. In the inlffiediately posterior segment (1.5 - 3 cm anterior to the central sulcus and 3 -4 em superior to the Sylvian fissure), two motor arrest (A) responses and one mixed (X) response were produced out of nine electrode pairs. There were no motor (M) or sensory (S) responses at that location. The area outlined with a thick line indicates the zones where tongue responses were elicited. With regard to distribution in relation to the Sylvian fissure, the cortical area for tongue responses extended to as much as 4 cm superior to the Sy1vian fissure in patients without organic lesion and to as much as 5 cm superior to the Sylvian fissure in the patients with fronto-parietal lesions (Table 1A and B). In patients with temporal lesions, only motor tongue responses were obtained, and they were found within 3 cm superior to the Sylvian fissure (Table Ie). Tongue response rate and distribution Patients without organic lesion Rate of tongue responses (percentage of electrode pairs producing tongue responses) upon stimulation of all electrodes in each 1 x 1.5 cn1 area was calculated, as shown in Table 2. In patients without organic lesion, the highest rate (%) of motor response was found in the 1.5 cm area immediately anterior to the central sulcus and I - 2 cm superior to the Sy1vian fissure Sensory-motor cortical tongue representation (Table 2A). At that site, 11 out of 27 electrode pairs (41 %) produced tongue motor responses (Tables 1A and 2A). The second and third most frequent locations were immediately above and below the first site. Immediately above, 33 % of the electrode pairs (nine out of 27 pairs) produced motor response, and immediately below, 27 % (nine out of 33) produced motor response. Although the numbers are noticeably smaller, there were motor responses posterior to the central sulcus. For example, a 13 % response rate (four out of 30) was found at the area 0 - 1.5 em posterior to the central sulcus and 1- 2 em superior to the Sylvian fissure, and a 6% response rate (one out of 16) was found further posteriorly (Table 2A). A relatively higher sensory response rate was found posterior to the central sulcus and 0 - 1 or 2 - 3 em superior to the Sylvian fissure (Table 2A). Tongue motor arrest tended to be found toward the vertex, further away from the Sylvian fissure. The higher rate of response was noted 3 - 4 em superior to the SyIvian fissure. In relation to the central sulcus, the majority of the motor arrest responses were seen anterior to the central sulcus (Tables 1A and 2A). A higher rate of mixed responses appeared to be located around a 1.5 em zone both anterior and posterior to the central sulcus (Table 2A). Patients with organic lesion Compared with the patients without organic lesion, patients with fronto-parietallesions demonstrated tongue motor and sensory response rates that tended to be higher at sites more anterior to the central sulcus and more superior to the Sylvian fissure (Table 2B). In patients with temporal lesions, motor responses tended to be distributed 0-1.5 em anterior to the central sulcus, a distribution similar to that in patients without organic lesions (Table 2c). Statistical analysis of the distribution of responses Relation to central sulcus and Sylvian fissure. To find out whether there is a significant difference among cortical sites for the different types of tongue responses, particularly in relation to the central sulcus and/or SyIvian fissure, Student's t test with Bonferroni' s correction for multiple comparison was performed (Table 3A and B). In patients without organic lesion, the mean location of motor responses, 0.77 ± 0.93 em anterior to the central sulcus, was compared with the mean location of sensory responses, 0.83 ± 0.72 em posterior to the central sulcus (P < 0.001) (Table 3A and B). There was no significant difference between the motor and sensory sites in relation to the Sylvian fissure (Table 3B). For the motor and motor arrest sites, there was no statistical difference in relation to either central sulcus or Sylvian fissure. However, motor arrest sites were significantly nlore anterior to central sulcus when com pared with sensory sites (P < 0.005) (Table 3B). Cortical sites for nlixed responses tended to be closer to the 123 central sulcus than were motor response sites (P < 0.05) (Table 3A and B). There was no apparent statistically significant difference in the distance from the central sulcus or the Sylvian fissure for nlotor versus motor arrest, mixed versus sensory, or mixed versus motor arrest responses (P > 0.05) (Table 3A and B). Of the 48 nl0tor responses in patients without organic lesion, two types of responses were noted; namely, pulling back with no laterality on 32 occasions and lateral tongue deviation on 16 occasions. There was no apparent statistical difference in their relationship to the central sulcus or the SyIvian fissure (Table 4). Similarly, there was no significant difference between bilateral sensory and unilateral sensory responses in relation to the central sulcus or Sylvian fissure (Table 4). Response rate. Response rates upon stimulation of all the electrode pairs in a given square area of the frontal versus parietal lobes were compared in patients without organic lesion. There were 253 electrode pairs in the frontal lobe (anterior to central sulcus) and 148 electrode pairs in the parietal lobe (posterior to central sulcus), as shown in Table lA. Upon stimulation of the electrode pairs in the frontal lobe, 60 electrode pairs produced at least one response: motor, sensory, arrest or mixed, a response rate of 23.7 % (Table 5). In the parietal area, 28 out of 148 electrode pairs (18.9 %) produced some sort of response. Although the frontal area has a higher response rate than the parietal lobe, X2 testing showed no significant difference. When the motor response rate was analysed in frontal versus parietal areas, the response rate for motor responses alone was significantly higher in the frontal area (P < 0.01). Similarly, the sensory response rate was significantly higher in the parietal area (P < 0.05). For nlotor arrest and mixed responses, there were no statistically significant differences between the electrode pairs in frontal versus parietal lobes (Table 5). When the sensory response rate was compared with the motor response rate using McNenlar's statistical test (Armitage, 1971), the sensory response rate was significantly lower in the combined frontal and parietal area and in the area anterior to the central sulcus (frontal) alone (Table 5). However, in the parietal region, no statistically significant difference between sensory response and motor response was obtained (McNenlar's test) . Location of organic lesions Among the 12 patients with organic lesions, seven patients had lesions in fronto-parietal areas. Four of the seven patients had space-occupying lesions in the vicinity of the central sulcus. In one of these patients, tongue motor response was recorded 4-5 em superior to the Sylvian fissure (Fig. 3B). The remaining three with fronto-parietal lesions had non space-occupying lesions. Two of these were small (1.5 x 2 em and 0.5 X 1 em) prefrontal angiomas detected by MRI and confirmed histologically. In one of the patients with angioma, 124 E. Urasaki et al. two motor responses were observed in the area 4 - 5 cm superior to the Sylvian fissure (Fig. 3A). The remaining patient with fronto-parietal lesion had dysplastic change in the prefrontal lobe. Tongue motor response and arrest, one each, were found 4.2 and 4.5 cm superior to the Sylvian fissure, respectively. The renlaining five patients had organic changes in the temporal lobe away from the fronto-parietal area. Two patients had angiomas, another two had a previous history of anterior temporal lobectonly for low-grade astrocytoma, and the last patient had a postoperative pseudocyst caused by posterior temporal lobectomy for low-grade astrocytoma 10 years earlier. Tongue responses were found 1- 3 cm from the SyIvian fissure in all these patients. Distances from the central sulcus and Sylvian fissure of each tongue response in the patients with organic lesions were compared with those of the patients without organic lesions (Table 3A). Statistically, the tongue motor area of patients with fronto-parietal lesions was located significantly more superior to the Sylvian fissure (P < 0.001) and more anterior to the central sulcus (P < 0.05) than that of patients without organic lesion. There were no significant differences in the distances for motor arrest, sensory and mixed tongue responses between the patient groups. In patients with temporal lesions, only motor responses were obtained, but the distances of the responses from the Sylvian fissure and central sulcus were not signifi cantly different from those of patients without organic lesion (Table 3A). Tongue responses in the superior temporal gyrus There were almost no tongue responses of any type outside the brain area outlined in Table 1, namely > 5 cm superior to the Sylvian fissure (67 electrode pairs studied) and > 4.5 cm anterior or posterior to the central sulcus (56 electrode sites). The exceptions were the three tongue responses found outside the area; they occurred upon stimulation immediately inferior to the Sylvian fissure, in the vicinity of the superior temporal gyrus in two patients, one with a prefrontal angioma (0.5 x 1 cm) and the other with no MRI-demonstrated organic lesion (Fig. 2D). One patient, a 22-year-old male, had temporal lobe epilepsy in the left, language-dominant hemisphere. Pullback tongue movement was observed upon stimulation of electrode pair C4 - D4, 1.5 cm inferior to the Sylvian fissure and well within the superior temporal gyrus. Furthermore, upon stimu lation of electrode pair C3 - D3, at 0.3 cm inferior to the Sylvian fissure, side-to-side tongue movement was arrested (D - 3 in Fig. 2D). The remaining patient was a 30-year-old wonlan with complex partial and secondary generalized seizures in the left, language-dominant hemisphere. Numbness of the tip of tongue was observed upon stimulation of the electrode pair 1 cm inferior to the Sylvian fissure-central sulcus junction. Table 2 (A - C) Distribution of motor, sensory, arrest and mixed responses in patients without organic lesions (A), with fronto-parietal lesions (B) and with temporal lesions (C) (A) Patients without organic lesions Sensory Motor 5.-------,---------,----r--------.------,---------, 5......------------,-------,-----y-----r-------r-----, (em) (em) 20% 9% 14% 33% 12% 41% 13% 6% 27% 6?'6 6% 6% +4.5(em) +3.0 cs +1.5 6% 3% SFL...----------'---------'-----L...---------'------'-------' -1.5 -3.0 SF " - - _ - - - L_ _---'--_ _. L . . . - _ - - - - ' +4.5(em) +3.0 -4.5 cs +1.5 -1.5 -"--_--' -3.0 -4.5 Mixed Arrest 5.----,-------,----r--------.-----,-------, (em) 5......---------,----,.-----y--------r-----.--------, (em) 14% 10% 10% 3 0/0/ 11% 7% 6°0 6 0/0/ 13% 10% 30% 11% 5% 4% 5 5% 8% 11% 22% 3% 0 / /0 SF SF +4.5(em) +3.0 +1.5 CS -1.5 -3.0 -4.5 +4.5(cm) +3.0 +1.5 CS -1.5 -3.0 -4.5 Sensory-motor cortical tongue representation 125 (B) Patients with fronto-parietal lesions Motor Sensory 5 5 1 7% (CfIl) 33% (em) 25% 29% 25% 1 7% 33% 17% 13% 20% 17% SF 25% SF +4.5(em) +3.0 cs +1. 5 -1. 5 -3.0 -4.5 +4.5 (em) +3.0 cs +1.5 Arrest -1.5 -3.0 -4.5 Mixed 5.------r-----r-------y-----,----.,.--------, 11% (CfIl) 25% (em) 17% 17% 1 7% 20% 2:,% 20% 17% 20% 38% 25% 1 7% SF SF +4.5(em) +3.0 +1. 5 cs -1.5 -3.0 '---------'--------'---~---'---------'-----------' +4.5(cm) +3.0 -4.5 cs +1. 5 -1.5 '-3.0 -4.5 (c) Patients with temporal lesions Motor 5 (em) 67% 33% 25% 60% 25% 14% 50% 33% SF +4.5(cm) +3.0 +1.5 cs -1.5 -3.0 -4.5 The percentages in each case correspond to the numbers in A -c, respectively. Numbers in each box indicate percentage of responses to electrode stimulation to all electrode pairs stimulated in the cortical segn1ents. For example, in A, for motor responses, in the particular cortical segment 1.5 cm anterior to central sulcus and 1- 2 cm superior to Sylvian fissure, 41 % of electrode pairs showed responses (11 electrode pairs out of 27 total). CS = central sulcus; SF = Sylvian fissure. Parietal motor and frontal sensory responses For the analysis of paradoxical responses, all patients (with and without organic lesion) were included (Tables 6 and 7). Paradoxical nlotor or sensory cortical responses were found in 17 out of 40 patients. In six patients some of the motor responses were observed in the parietal lobe instead of the frontal lobe. In seven patients, some of the sensory responses were observed in the frontal lobe instead of the parietal lobe. They were reported as discrete tingling or numbness identical with post central sensory response (Libet et al., 1964; Libet, 1973; Amassian et al., 1991; Cohen et al., 1991). In four patients, both frontal sensory and parietal motor responses were ob served. Five of the 12 parietal motor responses were part of mixed responses. Eight of the 12 frontal sensory responses were part of mixed responses (Table 6). Age, sex, duration of illness and cognitive function level in patients having paradoxical responses were compared with the same factors in patients without such responses (Table 7). There 126 E. Urasaki et al. Table 3 (A and B) (A) Location expressed as distance (in centimetres) from the central sulcus (CS) or Sylvian fissure (SF) of each type of tongue response in patients without brain lesions compared with patients with lesions. Minus sign at mean value indicates posterior to the central sulcus. Abbreviations as in Table 1. Patients Motor arrest Motor Mixed Sensory From CS (em) From SF (em) From CS Patients without organic lesions 0.70 ± 1.21 (n = 19) 2.05 ± 1.0 (n = 19) 0.77 ±0.93* 1.60 ± 0.85** 0.03 ±0.89 1.56 ± 1.0 (n = 48) (n = 48) (n = 15) (n = 15) Patients with fronto- parietal lesions 0.39±0.97 (n = 7) 2.74 ± 1.26 (n = 7) (n = 10) From CS From SF From CS From SF -0.83 ±0.72 1.03 ±0.87 (n = 6) (n = 6) 0.34 ± 1.30 1. 76 ± 1.33 1.68 ±0.91 * 3.16 ± 1.09** 0.43 ±0.83 1.96 ±0.89 (n = 8) (n = 5) (n = 5) (n = 10) (n = 8) 0.40± 1.36 (n = 12) Patients with temporal lesions From SF 1.25 ±0.81 12) (n = (B) Statistical comparison (t test with Bonferroni's correction) of the cortical locations of each tongue response in patients without organic lesion. Abbreviations are as in Table 1. Distance M versus S M versus X S versus X A versus M A versus S From CS p < 0.001 P < 0.05 NS NS P < 0.005 NS From SF NS NS NS NS NS NS *p < 0.05~ A versus X **p <0.001 (t test with Bonferroni's correction for multiple comparisons) were no statistically significant differences between the two groups in relation to any of the factors. Table 4 Comparison of the cortical locations of unilateral and bilateral tongue responses in patients without organic lesion Paradoxical responses in patients with brain lesion versus responses in patients without brain lesion Eight of 12 (67 %) patients with brain lesion had paradoxical responses (Tables 6 and 8). Four patients showed frontal sensory and two showed parietal motor responses. The remaining two patients had both parietal motor and frontal sensory responses. Paradoxical responses were seen in six out of seven patients with a fronto-parietal lesion (86 %), and two out of five patients with a temporal lesion (40 %) (Table 8). Of the 28 patients without brain lesion, nine (32 %) had paradoxical responses. Thus the brain-lesion group had a higher chance of having paradoxical responses (P < 0.05), particularly those with fronto-parietal lesion (P < 0.025) (Table 8). In the patients with temporal lesions, there was no statistical significance to the presenting paradoxical responses (P = 0.55). There were 12 parietal motor responses. Ten of the 12 responses were localized in the first 1.5 em posterior to the central sulcus, and the remaining two responses were located within 3 em posterior to the central sulcus. Eight of 12 frontal sensory responses were within 1.5 em anterior to the central sulcus, and three responses were within 3 em anterior to the central sulcus (Table 6). Discussion Tongue representations on the corlex Our study indicated that the cortical tongue area extended beyond the line 1.5 em anterior to the central sulcus and also From central sulcus From Sylvian fissure From central sulcus From Sylvian fissure Bilateral M IT nilateral M (n = 32) (n = 16) 0.85 ±0.93 1.63 ±0.87 0.59±0.97 1.54 ±0.80 NS NS Bilateral S Unilateral S t (n = 4) (n = 2) -1.08 ±0.49 0.90±0.88 -0.35 ±0.85 1.30 ± 0.80 t test test NS NS Bilateral M = bilateral motor responses, pulling back of the tongue without laterality ~ unilateral M = unilateral motor responses, lateral tongue deviation~ bilateral S = bilateral sensory responses, bilateral tongue sensation~ unilateral S = unilateral sensory response, unilateral tongue sensation contralateral to the side of stimulation~ minus sign at nlean value indicates posterior to the central sulcus. extended posterior to the central sulcus (Tables 1 and 2 and Figs 1- 3). The mean distance of the pre- and postcentral sulcus from the central sulcus is known to be -- 10- 12 mm (Ono et al., 1990). The tongue area showed a wide distribution in patients without organic lesions (Tables 1 and 2, Figs 1 and 2) and this distribution was wider still in patients with evidence of an organic lesion. It must be noted that, in the patients without fronto-parietal lesions, the grid rarely covered the area 3 em away from the Sylvian fissure. Nonetheless, tongue responses Sensory-motor cortical tongue representation 127 Table 5 (A and B) Statistical comparison of tongue responses by type, in frontal region versus parietal region (A), and comparison of response rates, by location, in motor responses versus sensory responses (B) (A) Total response (A+M+X+S) Motor arrest (A) Motor (M) Mixed (X) Sensory (S) Response rates % in frontal versus parietal 23.7 18.9 4.7 4.7 15.4 6.1 3.2 4.7 0.4 3.4 x2 NS NS P < 0.01 NS P < 0.05* test or Fisher's exact test (B) Motor versus sensory McNemar's test Whole area Frontal Parietal P < 0.001 P < 0.001 NS Note that McNemar's test is applied to B, because the rates of occurrence of motor and sensory are compared with each other in the same area, such as frontal, parietal and whole (frontal and parietal) areas. *Fisher's exact test is only used for sensory response because the data contain a number below 5. Table 6 (A and B) Summary of the patients with paradoxical frontal sensory and parietal Inotor responses (A) No. of patients Patients with organic lesions Frontal sensory response alone (FS) Parietal motor response alone (PM) 7 1 2 1 6 Angioma* Astrocytoma*** Cortical dysplasia (frontal) Patients without organic lesions 3 Both FS and PM Angioma** Astrocytoma t 4 1 1 Post-resection of 2 temporal astrocytoma 4 2 (B) Frontal sensory response No. of responses Only sensory response Part of mixed response Total Location of responses Parietal nlotor response 4 8 Only motor response Part of mixed response 12 From central sulcus 0- 1.5 cnl 1.5 -3.0 cm 8 4 7 5 12 From central sulcus about -1.5 cm -1.5 to about - 3.0 cm o to 10 2 Note that A is nunlber of patients, B is number of responses. One patient could have more than one response. *Patient had 1.5 x 2 cm angioma. The lesion was located in prefrontal subcortex; **in this patient the size of angioma was 0.5 x 1 cm and it was located in the prefrontal cortex; ***in these two patients, the size and location of the lesions were 3 x 2.5 cm in the precentral area and 4 x 4 cm in the pericentral area, respectively, seen on MRI; tin this patient, a 4 x 3.5 em lesion was found in the prefrontal area on MR!. Histological study showed low-grade astrocytoma. were often obtained at the distal edge of the grid (see Fig. 2), iInplying that tongue responses could occur even more superior to the Sylvian fissure in these patients. Furthermore, even in the group with small frontal lesions without mass effect, tongue responses anterior to the line 1.5 cm from the central sulcus were frequently found (see D6 in Fig. 3A), a finding suggesting 128 E. Urasaki et al. Table 7 Statistical cornparisons of clinical factors benveen the patients with and without paradoxical responses Patients with paradoxical responses Patients without paradoxical responses Age (years) Duration of seizure (years) 24.7 ± 11.7 (n 14.7 ± 11.8 (n 25.8±9.4 (n = 23) 14.2 ± 9.0 (n = 23) NS NS VIQ PIQ Total IQ Word tl uency score 86.4 ± 11.8 (n 10) 85.2 ± 9.5 (n = 10) 88.2±15.3 (n = 11) 9.4±3.0 (n = 8) 87.1±16.2 (n 17) 83.8 ± 10.3 (n 17) 83.6± 13.6 (n 17) 11.0 ± 2.6 (n = 15) NS NS NS NS t 17) 17) test x2 Male/female 13/10 9/8 test NS VIQ, PIQ = verbal and perfornlance intelligent quotients. Table 8 Paradoxical responses in patients with and ~vithout brain lesion Negative MRI for brain lesion (N) Brain lesion fronto-parietal or temporal areas (FPT) Brain lesion limited to fronto-parietal area (FP) Brain lesion limited to temporal area (T) Patients with/without paradoxical responses With: 9 Without: 19 8 4 6 1 2 3 Rate (percentage) of paradoxical responses 32% (9/28) 67% (8/12) 86% (6/7) 40% (2/5) On statistical conlparison (Fisher's exact test) of the groups of patients, the occurrence of responses was significantly greater in all patients with lesions (FPT) (P = 0.047 < 0.05) and in patients with fronto-parietal lesions (FP) (P = 0.016 < 0.025) than in patients with no evidence of brain lesions (N). There was no significant difference between patients without lesions (N) and those with temporal lesions (T) (P = 0.05). that dislocation or displacement of the 'primary' tongue area by pericentral lesion is not simply a factor of far anterior displacement of the tongue area by a mass or space-occupying lesion. This concept was also supported by the findings in patients without organic lesion, who showed wide distribution of the tongue area in an antero-posterior direction (Tables 1 and 2). Penfield and Boldrey (1937), Penfield and Rasmussen (1950), Libet et al. (1964), and Picard and Olivier (1983) have not observed such a wide distribution of tongue motor responses, except in one case of infiltrative glioma reported by Penfield and Boldrey (1937). Comparisons between their findings and ours are difficulC because their reports do not give specific details about their patients, including whether or not they had organic lesions (Penfield and Boldrey, 1937 ~ Penfield and Rasmussen, 1950~ Picard and Olivier, 1983). However, sensory tongue response occurring with stimulation outside the Rolandic region was reported by Penfield and Boldrey (1937). In addition, stimulation of the superior tenlporal gyrus elicited two tongue sensory responses: one at the tip and one contralateral to the side of cortical stinlulation (Penfield and Boldrey, 1937). Our study is in accord with their report. The issue of current spread is always difficult to resolve in cases of surface stimulation of human cortex, whether it occurs intra- or extra-operatively (Penfield and Boldrey, 1937 ~ Penfield and Rasmussen, 1950~ Picard and Olivier, 1983 ~ Uematsu et al., 1992a), and it could be the reason for the apparent wide distribution of the cortical tongue area. It is, however, unlikely that all our results could be accounted for by current spread. If current spread affects the area adjacent to the stimulated electrodes, direct stimulation of the area would be expected to elicit similar tongue responses. However, we frequently observed a non-tongue-response area between the areas where apparent tongue responses were elicited (Fig. 2B and D). This indicates that current spread is not the sole explanation for the wide distribution. It is already known that particular body parts (e. g. finger, hand or head or face) may have nlultiple representation on the cortex, e.g. in the first somatosensory area (SmI) or first somatomotor area (MsI) (Dreyer et al., 1975 ~ Strick and Preston, 1979). Although it was described in studies using a finer micro unit recording (Dreyer et al., 1975~ Strick and Preston, 1979), this concept may also be applied to our observations obtained with grid brain mapping. These obser vations are supported by similar brain-mapping studies carried out by others (Murphy and Gellhorn, 1945). Sensory-motor cortical tongue representation Low rate of sensory response Sensory responses were less frequent than motor responses even in the postcentral region (Table 5). It is unlikely that this is due entirely to different levels of excitability between the cortical area anterior to the central sulcus and the cortical area posterior to the central sulcus, because there was no difference in the rate for total tongue responses between these sites (Table 5). In our assessments, we slowly increase the intensity of the stimulus at each site until the appearance of a response or until we reach the 15 rnA Iimit of our machines (Lesser et al., 1987 ~ Uen1atsu et al., 1992a), which optimizes the chance of detect ing functional changes. Libet et al. (1964) and Libet (1973) en1phasized that the electrical threshold for postcentral motor response was higher than that for postcentral sensory or precentral motor response. However, the present study showed that over half (58 %) of the parietal motor responses were detected without conscious sensation (Table 6). One of the reasons for the low rate of sensory response might be that the disc electrodes mainly stimulated the surface of exposed cortical sensory area 2 and part of area 1, but less of area 3, because it is buried deep in the posterior wall of the central sulcus (Williams et al., 1989). Current spread was n1easured to be about one-quarter of the original value when the recording site was 2 mm away from the stimulation site (Murphy and Gellhorn, 1945) and the mean depth of the central sulcus was -- 16 n1n1 (Ono et al., 1990) so that current intensity in areas 3 A and B should be nluch lower than that at the surface and would likely be subthreshold for producing functional changes. Compared with area 3, areas 1 and 2 are similar to association cortex both morphologically and physiologically (Iwamura, 1991). Area 3 contains nunlerous granular cells, which are particularly associated with afferent projections, while areas 1 and 2 have increased numbers of pyramidal cells, which are involved in information transmission (Williams et al., 1989 ~ Iwamura, 1991). Single neuron studies disclosed an increased number of movetTIent-related neurons in areas 1 and 2 (Schwarz and Fredrickson, 1971: Iwamura, 1991). Iwamura (1991) stated that in area 2 of the tTIonkey, 10.6 % of neurons fired only upon active movement and 31.3 % showed no clear receptive fields, while only 56.6% fired upon superficial and deep sensation. These results support our finding of relatively few sensory responses evoked by stimulation of the surface area of the postcentral gyrus. The differences in test conditions between our study and previous reports must also be considered. Many of the previous stimulation-based studies of functional- anatomical relationships in the hUITIan brain mapping have been carried out intra operatively (Cushing, 1909~ Foerster. 1936b~ Penfield and Boldrey, 1937: Penfield and Rasmussen, 1950~ Walshe, 1951~ Libet et al., 1964: Goldring and Ratcheson, 1972 ~ Libet, 1973 ~ Celesia et al., 1979: Woolsey, 1979~ Picard and Olivier, 1983). Obviously, intra-operative conditions include local anaesthesia, possibly with light sedatives or narcoticentral sulcus, all of which could influence cortical responsivity. 129 Mechanism of frontal sensory and parietal motor response Paradoxical response Debate continues concerning parietal motor and frontal sensory responses (Foerster, 1936a, b ~ Penfield and Boldrey, 1937 ~ Penfield and Rasmussen, 1950~ Libet et al., 1964~ Libet, 1973). Recently, Baumgartner et al. (1992) also noted these paradoxical responses in their study using a subdural grid. According to Libet et al. (1964) and Libet (1973), parietal motor responses were described as non-smooth, intermittent moven1ents in comparison with 'pyramidal type' movements from the frontal lobe. In our study it was difficult to distinguish these characterists in tongue motor responses. Libet et al. (1964) and Libet (1973) stated that no frontal sensory response could be elicited in portions of the brain which were not specifically epileptogenic, except for motion sensation accompanying movenlents. In the present study, paradoxical responses occurred in eight out of 12 patients with and nine out of 28 patients without organic lesions (Table 6). A recent study using magnetic coil stimulation detected a frontal sensory response (tingling) in a minority of normal subjects (Amassian et al., 1991). Both the Amassian et al. study and our own suggest that the normal brain has the potential for paradoxical responses. Cortico-cortical connections In a cortical stimulation study, it is almost impossible to determine whether the responses result from the stimulated sites or from cortico-cortical or cortico-subcortical connections, including thalamus and more caudal nuclei (Williams et al., 1989). Areas 1 and 2 have bidirectional projections to areas 4 and 6, and reciprocal fibre connections exist between areas 1, 2, 5 and 7. In contrast, area 3b has only unidirectional connections to areas 1 and 2 (Jones and Powell, 1969~ Pandya and Kuypers, 1969: Yamaguchi and Knight, 1990). Thus stimulation of areas 4 and 6 could activate areas 1 and 2 and vice versa. The present study showed that 67 % of frontal sensory responses were expressed as part of a nnxed response, which could be the result of these cortico-cortical connections. It should be noted, however, that cortico-cortical connections are not the sole factor in frontal sensory or parietal motor responses, because, as reported by Penfield and Rasmussen (1950), these responses still could be elicited after excision of the postcentral gyrus or precentral gyrus, respectively (Brodmann's areas 4 and 6 and areas 3, 1 and 2). As noted above, tongue representation extended beyond the precentral and postcentral gyri, perhaps therefore including areas 44 and 45, and areas 43 and 40. Efferents from areas 44 and 45 are shown to terminate in area 4 (Zilles, 1990). It is possible that area 40 and its adjacent area 43 in the parietal lobe also are connected to the inferior frontal gyrus, including areas 44 and 45. Our data are in good agreement with the theory that physiologically, parts of areas 8, 44 and 45, together with 130 E. Urasaki et al. area 6, are in the premotor area, and that the whole precentral area, including area 4, belongs in the first somatomotor area (Williams et al., 1989), although the cyto-architecture of Brodmann differentiates them (Williams et al., 1989; Zilles, 1990). Brodmann's map and numeration are widely used as a reference for cortical locations, but recent pigmento architectonic observations of the cortex indicate the difficulty of dividing areas 43 and 40 from areas 3, 1 and 2 at the level of the tongue area, because they can be included in a parietal magnopyramidal region (Zilles, 1990). Other physio-anatomical supports Sensory cells were found in motor cortex by single neuron studies (Goldring and Ratcheson, 1972; Asanuma et al., 1979). Areas 3, 1 and 2 also contribute motor (pyramidal) fibres in monkey and man, and some (a minority) are generated from temporal and even occipital lobes in cats (Walberg and Brodal, 1953; Williams et al., 1989). These findings might explain in part the paradoxical frontal sensory and parietal motor responses, as well as the motor responses obtained in the temporal lobe. Second sornatosensory area (SmII) The anterior part of area 40 and the caudal part of the perirolandic area comprise the region where the second somatosensory area (SmII) son1etimes appears (Penfield and Rasmussen, 1950; Celesia, 1979; Williams et al., 1989). Penfield and Rasmussen (1950) were able to identify SmII in both prerolandic and postrolandic areas adjacent to the Sylvian fissure. The SmII is predominant!y sensory but motor function has also been documented in this area (Penfield and Rasmussen, 1950; Williams et al., 1989). Therefore, some parietal motor responses might be evoked through activation of efferent pathways originating in SmII (Celesia, 1979; Luders et al., 1985; Williams et al., 1989), some frontal sensory responses might be activated by stimulation of cells receiving afferents from ventrobasal or posterior nucleus of the thalamus (Celesia, 1979; Luders et al., 1985; Williams et al., 1989) and para doxical responses might be evoked via cortico-cortical connec tions between SmII and SmI or MsI (Pandya and Kuypers, 1969). However, SmII is known to represent mainly the extremities, and it is usually buried in the Sylvian fissure. The latter fact explains the low appearance rate of SmII; it was shown in only 10 out of over 400 patients on cortical stimulation study by Penfield and Rasmussen (1950) and only one out of over 50 patients in the somatosensory evoked potential study of Luders et al. (1985). Therefore, the paradoxical responses seen in 17 cases in this study seen1 unlikely to be explained by stimulation of SmII. with drug injection in the brain (Alloway et al., 1989) and with the strength of cortical stimulation (Liddell and Phillips, 1950; Cure and Rasmussen, 1954) in both animals and humans-findings all indicating plasticity of brain (Green and Walker, 1938; Glees and Cole, 1950). These phenomena are considered to result from unmasking of existing but ordinarily silent connections or sprouting of new synaptic contacts (Merzenich et al., 1983a,b; Alloway et al., 1989; Cohen et al., 1991; Iwamura, 1991). Similar mechanisms also might explain some of the paradoxical responses in our study, perhaps activated by epilepsy or organic lesions. All of these data strongly suggest that brain is not constructed of •static mosaics' with invariable relationships to one another. Conclusion As indicated by the present study, paradoxical responses, such as frontal sensory and parietal motor responses, are relatively common findings upon stimulation of the cortical tongue area in patients with epilepsy. The surgeon should be aware of the existence of such paradoxical responses, particularly during cortical stimulation, to establish the junction of the central sulcus and Sylvian fissure, especially in cases with fronto-parietal lesions. The anatomist and physiologist should be aware of them when attempting to determine the overall anatomical - functional relationships of human cortex. Acknowledgements We wish to thank Mary Bare and Pamela Schwerdt for their technical assistance in the cortical stimulation study, Robert S. Fisher, Gregory L. Krauss, Eileen P. Vining and Robert W. Webber for their valuable assistance, and Sakae Yamamoto, Associate Professor, Department of Management Science. Dokkyo University, and Shinya Matsuda, Assistant Professor. Department of Public Hygiene, University of Occupational and Environn1ental Health, for their assistance with the statistical analyses. This study was supported by the Tanaka Memorial Scholarship Fund provided by his widow Mary Lorenc Tanaka, and by grants fron1 the National Institute of Deafness and other Conm1unicative Disorders R03-DCO 1181), The National Institute of Neurological Disorders and Stroke (RO l-l'JS26553), the Seaver Foundation, the Whittier Foundation and the McDonnell-Pew Program in Cognitive Neuroscience. References Alloway KD, Rosenthal P, Burton H. Quantitative measurenlents of receptive field changes during antagonisnl of GABAergic transmission in primary somatosensory cortex of cats. Exp Brain Res 1989 ~ 78: 514-32. Paradoxical responses and 'mosaic' theory Amassian VE. Somasundaram M. Rothwell lC, Britton T, Cracco lB. Cracco RQ et al. Paraesthesias are elicited by single pulse, magnetic coil stinlulation of motor cortex in susceptible hunlans. Brain 1991: 114: 2505 - 20. Motor and sensory areas are revealed to change upon peripheral nerve damage (Merzenich et al., 1983a,b; Cohen et al., 1991), Armitage P. Statistical methods in medical research. Oxford: Blackwell Scientific 1971: 99 - 146. Sensory-motor cortical tongue representation Asanuma H. Larsen KD. YUI11iya H. Direct sensory pathways to the n10tor cortex ill the I110nkey: a basis of cortical ret1exes. In: Asanuma H, Wilson VJ. editors. Integration in the nervous system: a symposium in honor of David PC Lloyd and Rafael Lorente de No. Tokyo, New York: Igaku-Shoin. 1979: 223-38. BaUI11gartner C. Barth OS. Levesque MF, Sutherling WW. Human hand and lip sensorin10tor cortex as studied on electrocorticography. Electroencephalogr clin Neurophysiol 1992~ 84: 115 - 26. Celesia GG. S0I11atosensory evoked potentials recorded directly from human thalan1us and Sm I cortical area. Arch Neurol 1979~ 36: 399-405. Cohen LG. Bandinelli S. Findley TW, Hallett M. Motor reorganization after upper liI11b amputation in man: a study with focal magnetic stinlulation. Brain 1991: 114: 615-27. Cure C, RaSI11USSen T. Effects of altering the paranleters of electrical stiI11ulating currents upon n10tor responses from the precentral gyrus of Macaca nlulatta. Brain 1954 ~ 77: 18 - 33. Cushing H. A note upon the faradic stiI11ulation of the postcentral gyrus in conscious patients. Brain 1909~ 32: 44-53. Dreyer DA, Loe PR, Metz CB, Whitsel BL. Representation of head and face in postcentral gyrus of the macaque. J Neurophysiol 1975 ~ 38: 714-33. Foerster O. Motorische Felder und Bahnen. In: Bunlke 0, Foerster 0, editors. Handbuch der Neurologie. Berlin: Springer, 1936a~ 6: 1-357. Foerster O. The motor cortex in I11an in the light of Hughlings Jackson's doctrines. Brain 1936h ~ 59: 135 - 59. Glees P, Cole J. Recovery of skilled motor functions after small repeated lesions of nlotor cortex in macaque. J Neurophysiol 1950; 13: 137 -48. Goldring S, Ratcheson R. Human motor cortex: sensory input data fronl single neuron recordings. Science 1972 ~ 175: 1493 - 5. Green HD, Walker AE. The effects of ablation of the cortical nlotor face area in monkeys. J Neurophysiol 1938~ 1: 262 - 80. !tho T. Medical statistical analysis by n1icrocomputer program. Tokyo: Igaku-Shoin, 1984: 125 - 64 (J pn). Iwamura Y. The role of the S0I11atosensory and association cortices in hand function. No To Shinkei 199 I: 43: 603 -11 (Jpn). Jones EG. Powell TPS. Connexions of the somatic sensory cortex of the rhesus nl0nkey. I. - Ipsilateral cortical connexions. Brain 1969; 92: 477 - 502. Lesser RP. Luders H. KleI11 G. Dinner OS, Morris HH, Hahn JF. Ipsilateral trigeI11inal sensory responses to cortical stimulation by subdural electrodes. Neurology 1985 ~ 35: 1760 - 3. Lesser RP, Luders H. Klem G. Dinner OS. Morris HH, Hahn JF et at. Extraoperative cortical functional localization in patients with epilepsy. J Clin Neurophysiol 1987 ~ 4: 27 - 53. Libet B. Electrical stiI11ulation of cortex in human subjects, and conscious sensory aspects. In: Iggo A. editor. Somatosensory system. Handbook of Sensory Physiology. Berlin: Springer-Verlag, 1973~ 2: 743 -90. Libet B. Alberts WW. Wright EW. Delattre LD, Levin G, Feinstein 131 B. Production of threshold levels of conscious sensation by electrical stimulation of human somatosensory cortex. J Neurophysiol 1964; 27: 546-78. Liddell EGT, Phillips CG. Thresholds of cortical representation. Brain 1950; 73: 125-40. Luders H, Lesser RP, Dinner DS, Hahn JF, Salanga V, Morris HH. The second sensory area in hunlans: evoked potential and electrical stimulation studies. Ann Neurol 1985; 17: 177 - 84. Luders H, Lesser RP, Dinner DS, Morris HH, Hahn JF, Friedman L et al. Commentary: chronic intracranial recording and stiI11ulation with subdural electrodes. In: Engel J, editor. Surgical treatment of the epilepsies . New York: Raven Press, 1987: 297 - 321. Merzenich MM. Kaas JH, Wall J, Nelson RJ, Sur M, Felleman D. Topographic reorganization of somatosensory cortical areas 3b and 1 in adult nl0nkeys following restricted deafferentation. Neuroscience 1983a; 8: 33-55. Merzenich MM, Kaas JH. Wall JT, Sur M, Nelson RJ, Felleman DJ. Progression of change following median nerve section in the cortical representation of the hand in areas 3b and 1 in adult owl and squirrel monkeys. Neuroscience 1983b; 10: 639-65. Missir 0, Dutheil-Desclercs C, Meder JF, Musolino A, Fredy D. Aspect du sillon central en IRM: Central sulcus patterns at MRI. J Neuroradiol 1989; 16: 133 -44. Murphy JP, Gellhorn E. Multiplicity of representation versus punctate localization in the motor cortex. Arch Neurol Psychiat 1945; 54: 256-73. Ono M, Kubik S, Abernathey CD. Atlas of the cerebral sulci. New York: Thieme Medical Publishers, 1990: 14-180. Pandya DN, Kuypers HGJM. Cortico-cortical connections in the rhesus monkey. Brain Res 1969~ 13: 13 -36. Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain 1937; 60: 389 -443. Penfield W, Rasmussen T. The cerebral cortex of man. A clinical study of localization of function. New York: Macmillan, 1950: 19-61, 109 - 34, 209-23. Picard C, Olivier A. Sensory cortical tongue representation in I11an. J Neurosurg 1983; 59: 781-9. Ring BA, Waddington MM. Angiographic identification of the motor strip. J Neurosurg 1967; 6: 249-54. Schwarz DWF, Fredrickson JM. Tactile direction sensitivity of area 2 oral neurons in the rhesus monkey cortex. Brain Res 1971: 27: 397 -401. Strick PL, Preston JB. Multiple representation in the motor cortex: a new concept of input-output organization for the forearnl representation. In: Asanuma H, Wilson VJ, editors. Integration in the nervous system: a symposium in honor of David P. C. Lloyd and Rafael Lorente de No. Tokyo: Igaku-Shoin, 1979: 205 - 21. Tanaka Y, Tarumi T. Handbook of statistical analysis by personal computer, III planning method for experiment. Tokyo: Kyoritsu Shuppan, 1986: 42 -160 (Jpn). Taylor EH, Haughton WS. Some recent researches on the topography 132 E. Urasaki et al. of the convolutions and fissures of the brain. Trans Royal Acad Ireland 1900 ~ 18: 511 - 22 . nlotor function: with special reference to the "operational view' of experimental procedures. Brain 1951; 74: 249 -66. Uenlatsu S. Detection of an epileptic focus and cortical mapping using a subdural grid. In: Rengachary SS, Wilkins RH, editors. Neurosurgical operative atlas. Park Ridge, IL: American Association of Neurological Surgeons, 1992~ 2: 71- 82. Williams PL, Warwick R, Dyson M, Bannister LH, editors. Gray's Anatomy. 37th English ed. Edinburgh: Churchill Livingstone, 1989: 1039-68. Uematsu S, Lesser RP, Fisher RS, Gordon B, Hara K, Krauss GL et al. Motor and sensory cortex in humans: topography studied with chronic subdural stimulation. Neurosurgery 1992a ~ 31: 59 - 72. Uematsu S, Lesser RP, Gordon B. Localization of sensorimotor cortex: the influence of Sherrington and Cushing on the modern concept. Neurosurgery 1992b~ 30: 904-13. Walberg F, Brodal A. Pyramidal tract fibres from temporal and occipital lobes: an experimental study in the cat. Brain 1953~ 76: 491-508. Walshe FMR. On the interpretation of experimental studies of cortical Woolsey CN, Erickson TC, Gilson WE. Localization in somatic sensory and nlotor areas of human cerebral cortex as determined by direct recording of evoked potentials and electrical stimulation. J Neurosurg 1979; 51: 476-506. Yamaguchi S, Knight RT. Gating of sOlllatosensory input by human prefrontal cortex. Brain Res 1990; 521: 281 - 8. Zilles K. Cortex. In: Paxinos G, editor. The human nervous system. San Diego: Academic Press, 1990: 757-802. Received August 18, 1992. Revised July 14, 1993. Accepted September 11, 1993