Endogenous potentials after anterior temporal lobectomy

Neuropsychologia, Printed in Great

Vol 25, No. 3, pp. 549-557, Britarn.

ENDOGENOUS

JUNE

M.

1987

C028-3932/X7 $3.CO+O.o0 Pergamon Journals Ltd.

POTENTIALS AFTER ANTERIOR LOBECTOMY*

STAPLETON?//

ERIC

HALCREN~

and KATHLEEN

TEMPORAL

A. MORENO#

TDepartment of Psychology, and Brain Research Institute, University of California, Los Angeles, CA 90024, U.S.A., and V.A. Southwest Regional Epilepsy Center, West Los Angeles V.A. Medical Center, Los Angeles, CA 90073, U.S.A.; fDepartment of Psychiatry and Biobehavioral Sciences, and Brain Research Institute, University of California, Los Angeles, CA 90024, U.S.A., and V.A. Southwest Regional Epilepsy Center, West Los Angeles V.A. Medical Center, Los Angeles, CA 90073, U.S.A.; $Brain Research Institute, University of California, Los Angeles, CA 90024, U.S.A.; and V.A. Southwest Regional Epilepsy Center, West Los Angeles V.A. Medical Center, Los Angeles, CA 90073, U.S.A. (Accepted 14 November 1986) Abstract-The scalp topography of endogenous potentials was studied in patients who had previously undergone unilateral anterior temporal lobectomy (ATL). These excisions include medial temporal lobe (MTL) structures that have been shown to generate large potentials during tasks that evoke P3 at the scalp. Following right or left ATL, patients showed no differences from unoperated control subjects in overall amplitude of P3 or any other potential measured. The topography of P3 was very similar in both ATL groups and the control subjects, with no differences in laterality. These results suggest that the MTL is not the major generator of the P3 recorded at the scalp in the tasks studied here.

INTRODUCTION

of this paper is to examine the topography of endogenous potentials, particularly N2, P3, and Slow Wave, in patients who have had medial temporal lobe (MTL) structures removed surgically as a treatment for intractable epilepsy. Endogenous potentials are long-latency responses which are sensitive to task factors which influence cognition, such as attention and stimulus probability [7]. In previous papers, we have studied endogenous potentials at the scalp in normal subjects [23], and in depth recordings from the MTL in epileptic patients with implanted electrodes [22]. At least two potentials were observed in the MTL which have task correlates similar to the P3 recorded at the scalp (for reviews on P3, see [6, 151). Other data are also consistent with the possibility that MTL electrical activity may be the source for at least part of the P3 recorded at the scalp [ 10,201. Since unilateral anterior temporal lobectomy (ATL) is performed at UCLA for relief of seizures, a pool of subjects is available who have had surgical excisions of MTL structures of known extent. If P3 or any endogenous potential reflects a substantial contribution from MTL generators, then differences in amplitude or topography might be expected in these patients. WOOD et al. [24] reported preliminary results of a study of P3 in patients following surgical THE

PURPOSE

*This research was performed as part of the doctoral dissertation of J. M. Stapleton. It was supported by NIH grant NS18741 to E. Halgren and by the Veterans Administration. 1)Present address and address for correspondence: Section of Clinical Brain Research, Laboratory of Clinical Studies, National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Building 10, Room 3C218, 9000 Rockville Pike, Bethesda, MD, 20892, U.S.A. 549

550

JUNE M. STAPLETON etal.

temporal lobe excisions. With a 41-site scalp montage, they found that P3 was not reduced on the side of excision and was actually larger on that side in some patients. JOHNSON and FEDIO [ 111 also reported preliminary results on a group of patients after ATL. They found that P3 was smaller in the left ATL group than in the normal control group, but the right ATL group was not different from controls. All three groups showed P3’s slightly larger over the right hemisphere than the left.

METHODS Event-related potentials (ERPs) were recorded using Grass 7P511 amplifiers with f amplitude roll-off at 0.1 Hz and 0.1 KHz. Potentials were digitized online using an LSI 1 l/23 computer. Scalp recordings were made at standard lo-20 system sites (Fpz, Fz, Cz, Pz, Oz, F3, C3, P3, F7, T3, T5, F4, C4, P4, F8, T4, T6) using an Electrocap system. Additional electrodes were applied bilaterally to the tip of each ethmoid process (medial and ventral to F7) and to each mastoid process. All recordings were referred to the nose tip. Post-ATL patients who underwent surgery at least 2 years ago were invited to participate at the time of a followup visit. Normal subjects were recruited via a newspaper advertisement and chosen on the basis of similarity in age to the post-.4TL subjects (mean ages kS.E.M.: right ATL group 32.2k2.8, left ATL group 33.5 52.7, no ATL group 31.6 + 2.4). Data were collected from 33 subjects, 11 in each group. All subjects were right-handed, with the exception of one left ATL patient, who was ambidextrous. Most of the patients were currently receiving anticonvulsant medication, usually carbamazepine and/or phenytoin. Three left ATL patients and one right ATL patient were drug-free. Gender distribution was as follows: left ATL group 10 males, 1 female; right ATL group 5 males, 6 females; no ATL group 7 males, 4 females. The mean time since surgery was 5.7 years for the left ATL group and 6.4 years for the right ATL group. Surgery had been performed using the en bloc resection technique of FALCONER 181. The removal includes the anterior 5 -7 cm of the temporal lobe, including most of the hippocampus, uncus, and the basolateral amygdala (Fig. 1). The removal usually extends back from the temporal pole about I cm further for right ATL than for left ATL. Access to the brain is gained by turning a bone flap which is replaced following the excision. The patients studied here would be expected to have a well-healed skull defect along the seams of the bone flap. The skull defect after left ATL was in the region of T3, P3, F3, and F7, with C3 located approximately over the center of the flap (the symmetrical electrode sites would overlie the right ATL skull defect).

FIG. 1.Sketch of the standard extent of removal in anterior temporal lobectomy by en bloc resection. The excision involves 557 cm of anterior temporal lobe, including hippocdmpus, and dorsolateral amygdala. The resection is typically about 1 cm larger if performed on the right than on the left.

Stimuli were presented under the control of an Apple microcomputer via audio speaker or videomonitor. Based on previously reported findings, tasks were developed which were expected to differentially evoke endogenous potential components or P3 subcomponents having different topographic characteristics. Tasks were then modified to increase the consistency across tasks of variables not under study (e.g., tone pitch), and to facilitate comparisons among tasks. These tasks included : (1) Auditory Oddball Attend (AA, AAR = rare tones, AAF = frequent tones): a series of 50 msec tones is presented with 20% high (1500 Hz) and 80% low (1000 Hz) at a 1323 msec inter-stimulus interval (ISI). The subject silently counts the rare high tones. Similar tasks have been used in many previous studies to evoke N2, P3, and Slow Wave. (2) Auditory Oddball Ignore (AI, AIR = rare tones, AIF = frequent tones): the same stimuli as AA are presented, but the subject is instructed to read a book and ignore all tones. SQUIRES et al. [21] reported that rare tones in this type of task may evoke an earlier latency, more frontally distributed positivity, called P3a. (3) Auditory Oddball Distract (AD, ADR=rare tones, ADF=frequent tones, ADD=distracting, “weird” tones): the tone series is 15% high, 70% low and 15% “weird”. The “weird” tones are unique, novel sounds, with the

ENDOGENOUS

POTENTIALS

AFTER

ANTERIOR

TEMPORAL

LOBECTOMY

551

same amplitude envelope as the pure tones presented in AA. The ISI is the same as AA and the task is to count only the high tones. Studies using similar tasks in either the auditory or visual modality have described a large positivity with fronto-central scalp distribution evoked to the nontarget novel stimuli [3, 41. (4) Omitted Stimulus (OM, OMR = rare tones, OMF = frequent tones): on 15% of the trials, a tone is omitted from a series of low tones with a 660 ms ISI. The subject silently counts the omissions. A similar task has been found to evoke a typical P3, but little Slow Wave [ 161. (5) Visual Oddball (VN, VNR = rare stimuli, VNF = frequent stimuli): visual stimuli are presented in the middle of a videomonitor screen, with the same ISI as AA. Twenty per cent are rare stimuli (string of x’s) and 80% frequent (string ofo’s). The subject silently counts the rare stimuli. This task evokes a P3 similar to AA, but with a somewhat longer latency [18, 191. Potentials were averaged on the LSI 1 l/23 according to task conditions using the AVERAG software package by R. Norman (Neuroscience Systems). Edited averages were made excluding trials with eyeblinks or large eye movements, and including only the last frequent stimulus before each rare so that the averages for all tasks are of approximately the same number of trials. Peak amplitude and latency measures were made by visual identification of peaks on a CRT display. Potentials were measured at the point of peak amplitude in tasks in which the potential was clearly distinguishable. In tasks where the potential was absent or obscured, an amplitude measure was taken at the same time point as for the most similar task condition, as noted below for each potential. For each subject, Nl was identified as the largest negative peak at Cz between 50 and 200 msec post-stimulus onset, chosen separately for each task condition, except OMR, which was measured at the same time point as OMF for each subject. P2 was sought as the largest positive peak at Cz between 50 and 200 msec for each frequent tone average. Measurement of P2 for the rare tones was made at the same point as the corresponding frequent. Since N2 could not be clearly identified in some tasks, it was measured at a fixed latency for each subject across all tasks (except OM). The time point was chosen on the basis of the visible N2 peak in ADD or ADR at Fpz or Fz. N2 was sought as the largest negative peak between 150 and 300 msec in OM. P3 was identified as the largest positive peak at Cz between 240 and 400 msec for each rare and measured at the same time point for the corresponding frequent. Two negative Slow Waves seemed to be visible in many of the waveforms and were measured as follows: (1) SW 1: a single point measure at 450 msec, permitting us to measure this potential in the OM task before the end of the short ISI, (2) SW2: a measure of the area from 800 to 1000 msec post-stimulus onset in all tasks except OM. All measures were made with respect to a baseline period before stimulus onset of 120 msec for AA, AI, AD and VN, and 40 msec for OM. These guidelines for peak measurement were applied with a good deal of flexibility during detailed visual inspection of the data. Statistical analyses were carried out using standard statistical packages (BMDP, SPSS, or SAS). For quantitative analyses, subjects were selected for which a large number of trials were available for all task conditions under consideration, in order to provide stable amplitude and latency measures. For appropriate tasks, rare minus frequent waveforms were computed for each subject, but these will not be presented. Visual inspection of the subtracted waveforms tended to confirm the phenomena described below, but yielded no new information.

RESULTS Peak amplitude measurements for each potential for AAR and AAF at midline sites (Fpz, Fz, Cz, Pz, Oz) were submitted to an ANOVA with 3 Groups by 2 Task conditions by 5 Sites. Measures for F3, C3, P3, F4, C4, and P4 were submitted to an ANOVA with 3 Groups by 2 Tasks by 2 Sides (left or right) by 3 Sites (frontal, central, parietal). Similarly, measures for F7, T3, T5, Ml, F8, T4, T6, and M2 were submitted to an ANOVA with 3 Groups by 2 Tasks by 2 Sides by 4 Sites. Measures for Nl, P2, N2, and P3 for AAR, AAF, AIR, AIF, ADR, ADF, ADD, OMR and OMF at midline sites were submitted to an ANOVA with 3 Groups by 9 Tasks by 5 Sites. ANOVAs were evaluated for statistical significance using a critical value of F for P-cO.05. When appropriate, they were followed by simple effects ANOVAs, evaluated using a critical value of Ffor P-co.05 and pairwise t-tests with a critical value oft for P-co.01 (since a larger number of t-tests were performed). Specific effects are reported below only if they were statistically significant using these criteria. Averaged waveforms were also submitted to topographic mapping of peaks. Nl

For N 1, in AA or across tasks, there were no significant differences involving the Grouping factor or the Laterality factor. The topography of Nl showed no clear differences across

JUNE M. STAPLETON etal

552

being usually symmetric and maximal at Cz. Thus, contrary to a previous report [ 111, there was no evidence for Nl being larger over the operated hemisphere (Table 1). P2 For P2, there were also no laterality effects or group differences across tasks, but there was a barely significant Task by Group interaction in AA. This reflects a tendency for only the left ATL group to show a larger P2 to frequent than to rare tones. There were no differences among the 3 groups in P2 topography.

Table 1. Nl amplitude

AA Rare c3 c4 AA Freq c3 c4

Left ATL

No ATL

Right ATL

-4.3 -5.1

-3.4 -3.9

-5.0 -6.3

-2.8 -3.1

-3.5 -3.9

-5.4 -6.9

P3 In AA and VN, there were no significant differences involving the Grouping factor. Across tasks, however, there was a significant Task by Group interaction. The left ATL group showed less P3 in AD than the other two groups (Fig. 2). In general, the task correlates of P3 were very similar to those seen in normal subjects in our previous study (Fig. 3). There was a tendency for the left ATL group to have slightly smaller and the right ATL group slightly larger P3 responses than the normal control group, but this was not statistically significant in most analyses. Analysis of measures for the far lateral electrodes (F7, F8, T3, T4, T5, T6, M 1, M2) yielded a significant Group by Site interaction with no significant effect of laterality. As can be seen in Table 2, the right ATL group showed a more posterior distribution with relatively larger values at posterior temporal (T5/6) and mastoid sites, whereas the left ATL group had smaller potentials at these sites. In AA, there were no differences in laterality among the groups, but overall P3 was larger on the right than on the left in this study, as in our previous study [23]. As shown in Fig. 4, the topography of P3 did not differ markedly among the three groups, but there was considerable variability in topographic details across subjects within each group. SW

Neither factor.

of the two measures

of SW showed any significant

effects involving

the Grouping

Latency

There were no significant latency differences among the three groups in any potential although there was a nonsignificant tendency for all potentials except Nl to be somewhat later in both ATL groups (Fig. 5). This effect was not quite reliable for P3 (P=O.O6) but might emerge statistically with a large number of subjects.

ENDOGENOUS

POTENTIALS

AFTER

ANTERIOR

TEMPORAL

LOBECTOMY

553

P3 15 5 if

k.---+ o--a H

N-ATL L-ATL R-ATL

1 ?I i !

-5R’ F AA

R’

R

F

R: OM

D

Al

Ai

FIG. 2. Mean peak amplitude of P3 across 9 task conditions, averaged across 5 midline sites for right anterior temporal lobectomy (R-ATL, n= 1 I), left anterior temporal lobectomy (L-ATL, n= 11) and no anterior temporal lobectomy (N-ATL, n=9) groups. The L-ATL group shows a smaller P3 response in AD than the other two groups. R=Rare tones, F= Frequent tones, D=Distractors, AA = Auditory Oddball Attend, AI = Auditory Oddball Ignore, AD = Auditory Oddball Distract, OM = Omitted Stimulus Task.

P3 --*F3

b--A

F4

Y

d alo-

7 8.

1

\ \ \ \. \\ \. \\

‘? ?

AiR

A6F

LEFT

4. \ \ \ \ \* \\ \

t

AbR

AAF

RIGHT

L-ATL

AAR

AAF

LEFT

AAR

AAF

RIGHT

N-ATL

AAR

AAF

LEFT

AAR

AAF

RIGHT

I?-ATL

FIG. 3. Peak amplitude of P3 at lateral sites for the left anterior temporal lobectomy (L-ATL, n = 8) right anterior temporal lobectomy (R-ATL, n = 7), and no anterior temporal lobectomy (N-ATL, n = 5) groups. There are no significant differences involving the grouping factor. Overall, P3 is larger on the right for all three groups of subjects. Task abbreviations same as Fig. 2.

554

JUNE M. STAPLETON et al. Table 2. P3 amplitude-AA

Fl F8 T3 T4 T5 T6 Ml M2

P3 TOPOGRAPHY LEFT ATL

rare

Left ATL

No ATL

Right ATL

3.8 3.0 5.9 4.9 3.2 3.1 1.2 -0.6

3.2 4.2 4.6 6.5 4.4 5.5 1.6 2.2

2.4 2.1 5.4 6.0 6.3 6.1 2.3 3.1

IN 6 INDIVIDUAL SUBJECTS NORMAL

RIGHT ATL

FIG. 4. Maps of the scalp topography of P3 for 6 individual subjects, 2 from each group. There are no consistent topographical differences among the groups, and there is considerable individual variability in topographic details within groups. In particular, one normal control subject displayed here is atypical in having a P3 larger on the left. On the average, P3 is larger on the right in all three groups. The scale is in absolute amplitude in tenths of a pV, with shading bands in 2 PV increments from - 10.0 to + 12.OpV. Waveforms are shown below each map in the following order: Fpz, Fz, Cz, Pz, Oz, F4, F8, C4, P4, T4, T6, E2, M2, F3, F7, C3, P3, T3, T5, El, Ml.

ENDOGENOUS

POTENTIALS

AFTER ANTERIOR

TEMPORAL

LOBECTOMY

555

LATENCY

Nl

P2

N2

P3

FIG. 5. Mean latency ( + S.E.M.) of Nl, P2, N2, P3 for the three groups of subjects. Although there is a tendency for all the potentials except Nl to be slightly later in both lobectomy groups, there are no statistically significant differences among the groups. L=left anterior temporal lobectomy, N=no anterior temporal lobectomy, R = right anterior temporal lobectomy.

DISCUSSION These data lead to the’ following general conclusions. There are no major differences among the three groups (right ATL, left ATL and no ATL) in overall amplitude, latency, laterality, or topography of Nl, P2, N2, P3, or SW. The only differences among the groups in task correlates are that for P3 the left ATL group shows a smaller response in AD than the other two groups, and there is a tendency (not always statistically significant) for the left ATL group to have Nl larger to Rares than to Frequents and P2 larger to Frequents than to Rares. The difference in the P3 in AD for the left ATL group may fail to replicate in subsequent studies. It is conceivable, however, that the response to stimuli low in “categorizability” [2] could be predominantly a left hemisphere function and hence more affected by left than right ATL. The effects of rarity on Nl and P2 in the left ATL group are similar to the effects seen in normal subjects in an earlier study [23], but this effect was not seen in the normal control group in this study. These normal subjects were older (mean age 3 1.6 f 2.4 vs 22.8 +0.8) and somewhat less educated than the earlier study which involved mainly UCLA students and staff, but it is not clear whether this is a relevant difference. In any case, the earlier results with normal subjects indicate that these effects are not specifically characteristic of patients after left ATL. The effect of rarity on SW1 seen in the earlier study was also not seen in this study. In other ways, the normal group of this study replicates the earlier results very well. Particularly for P3, the task correlates and topography seen in the two studies are very similar. In both studies, the P3 shows an effect of laterality, being larger on the right than on the left.

556

JUNE M. STAPLETON et al

Previous studies have clearly demonstrated that large potentials are generated in the MTL during tasks that elicit P3 at the scalp [9, 121. The results of this study suggest that these potentials are not major contributors to the scalp endogenous potentials recorded in these simple cognitive tasks. The strength of this conclusion, however, is limited by the fact that anterior temporal lobectomy patients not only have undergone removal of brain tissue, but also have a large skull defect. Ordinarily, the high impedance of the skull forms a major influence on the current flows from brain to scalp [13]. The skull defect in our patients undoubtedly alters these current flows and thus changes the scalp potentials. Although it is difficult to predict exactly how the scalp field would be altered, it seems highly unlikely that the effects of the skull defect would be such as to exactly counterbalance the effects of eliminating a major generator. The results of this study are consonant with previous findings that P3 is not abolished by MTL lesions, whether unilateral lesions in humans or bilateral lesions in animals [l 1, 14, 241. They also relate to the findings of ALTAFULLAH et al. [l], who found that the Slow Wave following epileptic spikes does propagate from MTL to the lateral cortical surface, but attenuates more from MTL to surface than does the large potential seen in hippocampal depth recordings during P3 tasks. This finding suggests that the MTL potentials make a partial contribution to the surface P3 but that there is also at least one other, probably more superficial, generator. It has been suggested that P3 represents a transient reduction of cortical negativity produced by an activating circuit involving midbrain reticular formation, midline thalamus, and prefrontal cortex, the same system considered to be involved in selective attention [S, 251. Although potentials correlated with P3 have been recorded in thalamus, there is as yet no evidence for locally generated potentials in any of these structures. The potentials seen in thalamus are fairly small (about 20 pV), and it is not known whether they actually volumeconduct to the scalp. Probably they are less likely to be visible at the scalp than are the MTL potentials which often exceed 100 PV in amplitude. SIMSON et al. [ 171, on the basis of their scalp topographic studies, suggested a P3 generator in the nonspecific association cortex of the inferior parietal lobule. This suggestion came from an attempt to estimate unilateral topography assuming that the midline maximum results from the overlapping fields of bilaterally symmetrical sources. They also suggested the presence of an additional contribution from frontal association cortex. The possibility that P3 is generated primarily in parietal association cortex certainly cannot be ruled out at this time, but seems unlikely to explain the P3 elicited in ADD, where it is maximal at Cz and relatively small parietally. The possibility of a diffuse cortical generator (or multiple association cortex generators) remains a plausible hypothesis. This possibility is consistent with the scalp topography, the lack of large cortical potentials, and the small attenuation ratio of P3. The conclusions of this study should not be generalized beyond the tasks studied here. It is possible that the MTL contributes to one specific subcomponent, and that effects of MTL lesions would be seen if a task specifically evoking that subcomponent were used. It is also possible that more subtle effects could be seen if a larger number of electrodes were placed over the temporal lobe, if a larger number of patients were tested, or if the same patients were tested both before and after surgery in a within-subject design. Nevertheless, these data indicate that unilateral anterior lobectomy does not produce dramatic changes in P3 amplitude or topography and hence suggests that the electrical activity of MTL does not make a major contribution to the P3 recorded at the scalp.

ENDOGENOUSPOTENTIALSAFTER ANTERIORTEMPORALLOBECTOMY

557

Acknowledgements-Our thanks to Paul Crandall, who performed the surgeries, for his cooperation in this study of his patients, and to Joan Hopgood for her very helpful work in identifying and scheduling appropriate patients.

REFERENCES I. ALTAFULLAH, I., HALGREN, E., STAPLETON, J. M. and CRANDALL, P. Interictal spike-wave complexes in the human medial temporal lobe: typical topography and comparison with cognitive potentials, Electroenceph. clin. Neurophysiol. 63, 506-516, 1986. long-term effects on scalp distribution and 2. COURCHESNE, E. Changes in P3 waves with event repetition: amplitude. Electroenceph. clin. Neurophysiol. 45, 754766, 1978. components of the event-related brain potential: changes associated with 3. COURCHESNE, E. Cognitive development. In Tutorials in Event-Related Potential Research: Endogenous Components, A. W. K. GAILLARD and W. RITTER (Editors), pp. 329-344. North Holland, Amsterdam, 1983. 4. COURCHESNE,E., HILLYARD, S. A. and GALAMBOS, R. Stimulus novelty, task relevance and the visual evoked potential in man. Electroenceph. clin. Neurophysiol. 39, 131-143, 1975. 5. DESMEDT, J. E. and DEBECKER.J. Wave form and neural mechanism of the decision P350 elecited without prestimulus CNV or readiness potential in random sequences of near-threshold auditory clicks and finger stimuli. Electroenceph. clin. Neurophysiol. 47, 648-670, 1979. surprise? Psychophysiology 18, 493-513, 1981. 6. DONCHIN, E. Surprise! the endogenous components of I. DONCHIN, E., RITTER, W. and MCCALLUM, W. C. Cognitive psychophysiology: the ERP. In Event-Related Brain Potentials in Man, E. CALLOWAY, P. TUETING and S. H. KOSLOW (Editors), pp. 349441. Academic Press, New York, 1978. 8. FALCONER, M. A. Anterior temporal lobectomy for epilepsy. In Operatiue Surgery, C. Roe and R. SMITH (Editors), Vol. 14, p. 142. Butterworths, London, 1968. 9. HALGREN, E., SQUIRES,N. K., WILSON, C. L., ROHRBAUGH, J. W., BABB, T. L. and CRANDALL, P. H. Endogenous potentials generated in the human hippocampal formation by infrequent events. Science 210, 803-805, 1980. 10. HALGREN, E., STAPLETOX, J., SMITH, M. and ALTAFULLAH, I. Generators of the human scalp P3s. In El;oked Potentials, R. Q. CRACCO and I. BODIS-WOLLNER (Editors), pp. 269-284. Alan Liss, New York, 1986. 1 I. JOHNSON, R. Jr. and FEUIO, P. ERP and P300 activity in patients following unilateral temporal lobectomy. Sot. Neurosci. Abs. 10, 847, 1984. 12. MCCARTHY, G., WOOD, C. C., ALLISON, R., GOFF, W. R., WILLIAMSON, P. D. and SPENCER, D. D. Intracranial recordings of event-related potentials in humans engaged in cognitive tasks. Sot. Neurosci. Ahs. 8, 976, 1982. 13. NUNEZ, P. L. Electric Fields ofthe Brain. New York, Oxford, 1981. 14. PALLER, K. A., ZOLA-MORGAN, S., SQUIRES, L. R. and HILLYARD, S. A. Monkeys with lesions of hippocampus and amygdala exhibit event-related brain potentials that resemble the human P300 wave. Sot. Neurosci. Abs. 10, 849, 1984. of P300. Psychol. Bull. 89, 506540, 1981. 15. PRITCHARD, W. S. Psychophysiology 16. ROHRBAUGH, J. W., SYNDULKO, K. and LINDSLEY, D. B. Cortical slow negative waves following non-paired stimuli: effects of task factors. Electroenceph. clin. Neurophysiol. 45, 551-567, 1978. of potentials in auditory and visual 17. SIMSON, R., VAUGHAN, H. G. and RITTER, W. The scalp topography discrimination tasks. Electroenceph. clin. Neurophysiol. 42, 528-535, 1977. 18. SNYDER, E., HILLYARD, S. A. and GALAMBOS, R. Similarities and differences among the P3 waves to detected signals in three modalities. Psychophysiology 17, 112-122, 1980. inferring decision19. SQUIRES, N. K., DONCHIN, E., SQUIRES, K. D. and GROSSBERG, S. Bisensory stimulation: related processes from the P300 component. J. exp. Psychol. (Hum. Percept.) 3, 299-315, 1977. 20. SQUIRES, N. K., HALGREN, E., WILSON, C. and CRANDALL, P. Human endogenous limbic potentials: crossmodality and depth/surface comparisons in epileptic subjects. In Tutorials in ELlent-Related Potential Research: Endogenous Components, A. W. K. GAILLARD and W. RITTER (Editors), pp. 217-232. North Holland, Amsterdam, 1983. 21. SQUIRES, N. K., SQUIRES, K. C. and HILLYARD, S. A. Two varieties of long-latency positive waves evoked by unpredictable auditory stimuli in man. Electroenceph. clin. Neurophysiol. 38, 387401, 1975. 22 STAPLETON,J. M. and HALC~REN,E. Depth endogenous potentials evoked in simple cognitive tasks: components and task correlates, Electroenceph. clin. Neurophysiol., in press. potentialsevoked in simple cognitive tasks, 23. STAPLETON,J. M., O’REILLY, T. and HALGREN, E. Scalpendogenous Int. J. Neurosci., in press. 24. Woou, C. C.. MCCARTHY, G., ALLISON, T., GOFF, W. R., WILLIAMSON, P. D. and SPENCER,D. D. Endogenous event-related potentials following temporal lobe excisions in humans. Sot. Neurosci. Abs. 8, 976, 1982. 25. YINC;LING,C. D. and HOSOBUIXI, Y. A subcortical correlate of P300 in man. Electroenceph. clin. Neurophysiol. 59, 72 -76, 1984.