Abstract

The thalamic medial pulvinar nucleus (PuM) is fully developed only in primates and reaches its greatest extent in humans. To assess the reciprocal functional connectivity between PuM and cortex, we studied intracerebral-evoked responses obtained after PuM and cortical electrical stimulation in 7 epileptic patients undergoing depth electroencephalographic recordings. Cortical-evoked potentials (CEPs) to PuM stimulation were recorded from all explored cortical regions, except striate cortex, anterior cingulated, and postcentral gyrus. Percentages of cortical contacts pairs responding to PuM stimulation (CEPs response rate) ranged from 80% in temporal neocortex, temporoparietal (TP) junction, insula, and frontoparietal opercular cortex to 34% in mesial temporal regions. Reciprocally, PuM-evoked potentials (PEPs) response rates were 14% after cortical stimulation in insula and frontoparietal opercular cortex, 67% in the TP junction, 76% in temporal neocortex, and 80% in mesial temporal regions. Overall, our study of functional PuM connectivity in the human brain converges with most of the data from anatomical studies in monkeys, except for a strong amygdalohippocampal functional projection to PuM and an unexpected imbalance between some of the reciprocal pathways explored. This functional quantitative approach helps to clarify the functional role of PuM as well as its implication in temporal lobe epileptic seizures.

Introduction

To date, most of our knowledge of thalamocortical connectivity derives from animal studies, raising a difficulty when extrapolation to human brain is at stake. In humans, anatomical studies rely on postmortem injections of fluorescent dyes that allow the tracking of some projections but only at distances of tens of millimeters from the injected area (Mufson et al. 1990; Lim et al. 1997; Tardif and Clarke 2001; Tardif et al. 2007). Additional information comes from gross dissections or from histological studies of remote degeneration following a focal lesion, which are based on a small number of patients (Clarke et al. 1999). Recently, diffusion tensor magnetic resonance tractography has enabled in vivo tracing of large fiber tracts in the human brain and has been applied to visualize the major thalamocortical fiber bundles (Behrens et al. 2003). However, this technique does not provide with details about the functionality of these tracts and their conduction velocity and is unable to distinguish thalamocortical from reciprocal corticothalamic connections. Some authors have attempted to use transcranial magnetic stimulation and scalp recordings to assess the functional cortical connectivity, but this method has a spatial resolution limited to the cortex and cannot be applied to the study of thalamocortical connections (Rossini and Rossi 2007).

Recently, the electrophysiological study of subcortical structures, in particular the thalamus, has known a revival in epilepsy because of growing evidence that they are involved in propagation and synchronization of epileptic discharges (Bertashius 1991; Szabo et al. 2005; Guye et al. 2006). On this line, we have shown that the thalamic medial pulvinar nucleus (PuM) is involved in most of temporal and insular lobe seizures (Rosenberg et al. 2006). Because recent attempts to control epileptic seizures by chronically stimulating thalamic nuclei have led to some encouraging results (Velasco et al. 2006; Andrade et al. 2006), we decided to investigate whether chronic PuM stimulation could represent an alternative treatment for patients with drug-resistant epileptic seizures who are not eligible for surgical resection of the epileptogenic zone. In this context, a precise knowledge of the thalamocortical and corticothalamic functional connectivity is crucial to define the cortical territories that could be functionally influenced by this therapeutic approach. Our present knowledge of PuM connectivity is mostly based on anatomical studies in monkeys, which suggested widespread reciprocal connections between PuM and temporal, parietal, frontal, and occipital cortical areas (Trojanowski and Jacobson 1977; Robinson and Cowie 1997; Romanski et al. 1997; Grieve et al. 2000; Gutierrez et al. 2000). However, PuM is in humans the thalamic structure showing the greatest phylogenetic expansion as compared with nonhuman primates and, consequently, the most susceptible to show anatomical variations as compared with other animal species. It was thus essential to reevaluate PuM connectivity in humans. For this purpose, we used electrical stimulations coupled with intracerebral recordings of evoked responses in epileptic patients chronically implanted with depth electrodes for stereoelectroencephalographic (SEEG) assessment of their epileptic focus. Furthermore, we attempted to estimate quantitatively the preferential pathways between the cortex and PuM using the same approach as that recently proposed by Matsumoto et al. (2004, 2007).

Materials and Methods

Patients

Seven patients, 4 women and 3 men, aged 34 ± 9.6 years (range = 16–47), were included in this study (Table 1). They were suffering from drug-resistant partial epilepsy suspected to originate from the temporal lobe. All patients were referred to the Epilepsy department of the Neurological Hospital in Lyon for SEEG recording of their seizures because noninvasive presurgical investigations had not permitted to delineate their epileptogenic zone precisely enough to plan a surgical cortectomy and in particular a standard temporal lobectomy. All patients were included in the phase 1 of a protocol designed to evaluate the feasibility of chronic PuM stimulation as a treatment for temporal lobe epilepsy (TLE). Noninvasive presurgical data included scalp videoelectroencephalographic monitoring of seizures, brain magnetic resonance imaging (MRI), interictal and ictal single-photon emission tomography, and 18Fluorodeoxyglucose positron emission tomography (FDG-PET).

Table 1

Individual clinical, MRI, and SEEG data

Patient number Age (years) Gender Epilepsy duration (years) MRI Seizure onset zone determined by SEEG procedure Number of electrodes Number of tested cortical contact pairs Number of tested PuM contact pairs 
16 10 Hypersignal on left and right temporal pole Right frontal cortex (first and second frontal gyrus) 11 47 
47 17 Normal Left mesial temporal structures (hippocampus, entorhinal cortex) 11 41 
30 24 Right hippocampal sclerosis Right mesial temporal structures (amygdala, hippocampus, and entorhinal cortex) 12 57 
37 27 Bilateral heterotopias Left temporal neocortex 13 59 
34 14 Normal Left hippocampus, entorhinal cortex, fusiform gyrus 41 
40 23 Left amygdala hypersignal Left temporal pole and temporal neocortex 10 46 
35 20 Normal Right occipital and temporal neocortex 10 38 
Patient number Age (years) Gender Epilepsy duration (years) MRI Seizure onset zone determined by SEEG procedure Number of electrodes Number of tested cortical contact pairs Number of tested PuM contact pairs 
16 10 Hypersignal on left and right temporal pole Right frontal cortex (first and second frontal gyrus) 11 47 
47 17 Normal Left mesial temporal structures (hippocampus, entorhinal cortex) 11 41 
30 24 Right hippocampal sclerosis Right mesial temporal structures (amygdala, hippocampus, and entorhinal cortex) 12 57 
37 27 Bilateral heterotopias Left temporal neocortex 13 59 
34 14 Normal Left hippocampus, entorhinal cortex, fusiform gyrus 41 
40 23 Left amygdala hypersignal Left temporal pole and temporal neocortex 10 46 
35 20 Normal Right occipital and temporal neocortex 10 38 

Stereotactic Implantation of Deep Brain Electrodes

Determination of SEEG Cortical Targets

The cortical regions to be explored as well as the number of electrodes were individually determined for each patient, based on individual presurgical data. Nine to 13 multiple contacts electrodes were implanted per patient. Four implantations were performed in the left hemisphere and 3 in the right one.

PuM Implantation and Recordings

PuM stimulation and/or recording were performed using the 2–4 deepest contacts of the electrode exploring the posterior part of the superior temporal gyrus (see Rosenberg et al. 2006). Therefore, no electrode was specifically implanted to record the thalamus in addition to those required by the diagnostic SEEG procedure. All patients were fully informed of the aim of this investigation and gave their written consent about the implantation and stimulation procedure, which was approved by the local ethics committee (CCPPRB Lyon—Centre Léon Bérard).

Stereotactic Implantation Technique

Intracerebral exploration was conducted according to the technique described by Talairach and Bancaud (1973), a procedure used routinely in our department (Guenot et al. 2001). Intracerebral electrodes were implanted perpendicular to the midsagittal plane (Fig. 1A) and could be left in place for ≤15 days. The electrodes had a diameter of 0.8 mm and 5–15 recordings contacts, each 2 mm long, separated by 1.5 mm one from the other.

Figure 1.

(A) Example of skull radiography performed after implantation. Each electrode is visible as a “dotted line” formed by its contacts, which are opaque to X-rays. The electrode exploring the superior temporal gyrus and the pulvinar nucleus is underlined with “black arrowheads.” Deep and lateral contacts of each electrode explore the inner and outer part of parietal, temporal, or frontal lobes, respectively, depending on their implantation coordinates. (B) Reconstruction of a thalamic electrode in the Talairach space, visualized on the appropriate axial T1 MRI and the corresponding Morel atlas stereotactic plane (3.6 mm above the AC–PC plane—white crosses: AC and PC levels). The electrode is composed of 15 adjacent contacts, each represented by a blue rectangle. Filled blue rectangles correspond to contacts used both for stimulation and PEPs and/or CEPs recordings. In the illustrated case, 3 medial contacts were located in PuM and 3 lateral ones in the temporal neocortex (first temporal gyrus). (C, D) Position of the 2 most medial contacts in the thalamus of the 7 patients explored, superimposed on a sagittal (12.6 mm lateral to interhemispheric plane, left) and axial (4.5 mm above AC–PC, right) planes from the Morel atlas. The 3 implantations in the right thalamus are flipped to be represented on the left side. Levels of sagittal and axial sections are indicated by dotted lines. Individual thalamic borders were drawn from MRI individual data. Individual diagrams are superimposed onto atlas plane taking PC as reference. Normalization was not performed in order to visualize interindividual variations in thalamic size (note that the individual AC locations illustrated by colored crosses are different from one patient to the other when PC locations are strictly superimposed at the same coordinates). Black crosses: PC and AC levels according to the atlas planes. CeM, central medial nucleus; CL, central lateral nucleus; LP, lateral posterior nucleus; MD, mediodorsal nucleus; MGN, medial geniculate nucleus; PuM, medial pulvinar; PuA, anterior pulvinar; PuL, lateral pulvinar; VA, ventral anterior nucleus; VPM, ventral posterior medial nucleus; VPL, ventral posterior lateral nucleus, a, anterior division, p, posterior division; VLp, ventral lateral posterior nucleus, d, dorsal division; v, ventral division; VLa, ventral lateral anterior nucleus; STh, subthalamic nucleus; R, reticular thalamic nucleus.

Figure 1.

(A) Example of skull radiography performed after implantation. Each electrode is visible as a “dotted line” formed by its contacts, which are opaque to X-rays. The electrode exploring the superior temporal gyrus and the pulvinar nucleus is underlined with “black arrowheads.” Deep and lateral contacts of each electrode explore the inner and outer part of parietal, temporal, or frontal lobes, respectively, depending on their implantation coordinates. (B) Reconstruction of a thalamic electrode in the Talairach space, visualized on the appropriate axial T1 MRI and the corresponding Morel atlas stereotactic plane (3.6 mm above the AC–PC plane—white crosses: AC and PC levels). The electrode is composed of 15 adjacent contacts, each represented by a blue rectangle. Filled blue rectangles correspond to contacts used both for stimulation and PEPs and/or CEPs recordings. In the illustrated case, 3 medial contacts were located in PuM and 3 lateral ones in the temporal neocortex (first temporal gyrus). (C, D) Position of the 2 most medial contacts in the thalamus of the 7 patients explored, superimposed on a sagittal (12.6 mm lateral to interhemispheric plane, left) and axial (4.5 mm above AC–PC, right) planes from the Morel atlas. The 3 implantations in the right thalamus are flipped to be represented on the left side. Levels of sagittal and axial sections are indicated by dotted lines. Individual thalamic borders were drawn from MRI individual data. Individual diagrams are superimposed onto atlas plane taking PC as reference. Normalization was not performed in order to visualize interindividual variations in thalamic size (note that the individual AC locations illustrated by colored crosses are different from one patient to the other when PC locations are strictly superimposed at the same coordinates). Black crosses: PC and AC levels according to the atlas planes. CeM, central medial nucleus; CL, central lateral nucleus; LP, lateral posterior nucleus; MD, mediodorsal nucleus; MGN, medial geniculate nucleus; PuM, medial pulvinar; PuA, anterior pulvinar; PuL, lateral pulvinar; VA, ventral anterior nucleus; VPM, ventral posterior medial nucleus; VPL, ventral posterior lateral nucleus, a, anterior division, p, posterior division; VLp, ventral lateral posterior nucleus, d, dorsal division; v, ventral division; VLa, ventral lateral anterior nucleus; STh, subthalamic nucleus; R, reticular thalamic nucleus.

Localization Checking of the Recording Contacts

To verify the final position of each implanted contact with respect to the targeted anatomical structures, frontal and sagittal X-rays at scale 1 were obtained at the end of the surgical procedure, using an X-ray source at 4.85 meters from the patient's head thus eliminating the linear enlargement due to X-ray divergence. The T1-weighted MRI midsagittal slice was superimposed on the sagittal X-ray, after enlargement at scale 1 using the limits of the cranial bone as a common reference. This allowed to define the position of the anterior commissures (AC) and posterior commissures (PC) and to calculate the anteroposterior and dorsoventral coordinates of each electrode contact. The third mediolateral coordinate of each contact was determined on the frontal X-ray with respect to the midsagittal plane. Each contact was then localized in the Talairach space (Talairach and Tournoux 1988) according to 3 coordinates in millimeters; x for the distance from the interhemispheric sagittal vertical plane, y for the distance from the coronal plane perpendicular to the horizontal AC–PC plane passing through PC, and z for the distance from horizontal AC–PC plane. In this reference system, the coordinates of the PC midline point serves as reference with x, y, and z = 0. The 3-dimensional (3-D) position of all contacts within cerebral structures in a given patient was checked by plotting their coordinates in the patient's MRI volume (T1-weighted images, 1-mm based voxel) reconstructed in the 3-D space using MRIcro Software (Rorden and Brett 2000; Fig. 1B).

This method is based on measurement of coordinates in the Talairach's stereotaxic space and not on direct MRI visualization of contacts after electrodes implantation. Therefore, even after correction for contacts locations and evaluation of PuM limits based on individual patient's MRI (Fig. 1), the contact localization in anteroposterior and dorsoventral dimensions is prone to several inaccuracies linked to MRI slice thickness (±0.5 mm), superimposition of MRI slice on X-ray (±1 mm), and coordinates measurement (±0.5 mm), which leads to an average error (square root of the sum of the squares of the different errors) of ±1.32 mm. In the mediolateral dimension, inaccuracies are limited to the determination of the midsagittal plane on X-ray (±1 mm) and coordinates measurement (±0.5 mm) leading to an average error of 1.11 mm. Combined together, these error values lead to a mean Euclidean error of 2.17 mm in the 3-D space.

The localization of the PuM contacts was checked by writing out their coordinates on corresponding planes of a stereotactic atlas of the human thalamus (Morel et al. 1997; Fig. 1B–D). Anteroposterior coordinates of thalamic electrodes ranged from 0 to −7.5 mm caudal to PC and dorsoventral ones from +2 to +6 mm above the horizontal AC–PC plane. The x coordinates of the medialmost PuM contacts of the electrodes were 7– 10 mm lateral to the interhemispheric vertical plane. Furthermore, we checked that none of PuM contacts could be located outside this structure taking into account the mean Euclidean error of 2.17 mm in contact localization. As illustrated in Figure 1C, this was achieved for all contacts except for the most medial one in patient 7, the deepest edge of which was at a distance of 1.6 mm from the medial PuM border so that the inner part of this 2-mm contact may have been out of the thalamus border. The number of contact pairs situated inside the PuM in each patient is given in Table 1.

Finally, we checked that all of the contact pairs used for cortical recording and stimulation were located in the cortical gray matter (see Supplementary Materials). Each contact was plotted in the patient 3-D MRI using MRIcro software, and the implanted structure was identified using Duvernois (1991) and Ono et al. (1990) atlases.

Stimulation Parameters

Square pulses of current were applied between 2 adjacent contacts (bipolar stimulation), delivered by a current-regulated stimulator approved for use in human subjects (IRES 600 CH Surgical, Micromed, Italy). Stimulation parameters were chosen to avoid tissue damage (Gordon et al. 1990) and were similar to those previously used in human studies (Wilson et al. 1990; Valentin et al. 2002; Catenoix et al. 2005; Zumsteg, Lozano, and Wennberg 2006; Zumsteg, Lozano, Wieser, and Wennberg 2006; Lacruz et al. 2007). They consisted of 2 series of 25 pulses of 1-ms duration and 3-mA intensity, delivered at a frequency of 0.2 Hz. To estimate the brain tissue volume in which most excitable neuronal elements, that is, axons, were activated, we referred to the Figure 1 of the Ranck's (1975) study. It defined a sphere with a 2- to 5-mm radius. However, this estimation was based on monopolar stimulation condition and it has been well established that bipolar stimulation tends to produce a more focused electrical field of stimulation (Nathan et al. 1993). Furthermore, Wetzel (1970) has pointed out that with side-by-side bipolar electrodes, the neural elements between the electrode tips are probably preferentially stimulated. It should be also mentioned that other factors might have a nonnegligible influence on stimulation efficacy including fibers electrophysiological properties (un- or myelinated axons, chronaxy) (Ranck 1981), orientations of the neuronal elements relative to the electrodes (Manola et al. 2007), and conductance differences between white and gray matters, which cannot be controlled in our experimental conditions. Taking into account all of these considerations, we estimate that the maximal tissue volume stimulated in our conditions is that of a sphere with a 4-mm radius.

The most medial PuM contacts pair was used for thalamic stimulation in all patients. Given their large number, it was not possible to stimulate all cortical contact pairs. Therefore, we selected for stimulation 1– 3 contact pairs per patient in each of the implanted cortical regions. A detailed list of studied cerebral regions is given in Table 2. Cortical stimulation was performed on the same day as, or eventually the day after, PuM stimulation. None of the patients experienced any clinical sensation resulting from cortical or PuM stimulations.

Table 2

CEPs and PEPs response rates and latencies

 CEP
 
PEPs
 
 Number of tested patients Number of responding patients Number of tested contact pairs Response rate Onset latency (mean ± SD) (ms) Number of tested patients Number of responding patients Number of tested contact pairs Response rate Onset latency (mean ± SD) (ms) 
STG (BA 22) 37 95% 11.4 ± 2.6 100% 56.6 ± 32.6 
MTG (BA 21, 37) 50 70% 13.7 ± 2.7 14 71% 56.8 ± 46.4 
ITG (BA 20, 37) 27 70% 16.9 ± 7.3 50% 25.4 ± 16.6 
FG (BA 36) 80% 37.1 ± 7.6 50% 11.72 (1 patient) 
Temporal neocortex   119 78%    29 76%  
Hippocampus 19 16% 27.3 ± 5.5 83% 27.8 ± 7.4 
Amygdala 15 20% 60.5 (1 patient) 100% 46.9 ± 7 
Temporal Pole (BA 38) 15 47% 66.4 ± 52 50% 33.2 ± 7.8 
Parahipp Gyr (BA 28, 34, 35) 12 67% 71 ± 47.3 100% 43.5 ± 28.2 
Mesial temporal structures   61 34%    20 80%  
Postcentral gyrus (BA 3, 2, 1) 0% NA NA NA NA 
TP Junction (BA 39, 40) 21 81% 15.2 ± 4.9 67% 23.4 ± 2.8 
Precuneus (BA 7) 11 46% 29.3 ± 19.8 0% NA 
Parietal cortex   34 65%    5 40%  
Insula 16 94% 17.3 ± 6.7 17% 41 (1 patient) 
FP operculum 13 85% 27.8 ± 16.1 0% NA 
Insuloopercular cortex   29 90%    7 14%  
Cing ant gyr (BA 24, 32) 0% NA 0% NA 
Cing postgyr (BA 23, 31) 67% 17.58(1 patient) 100% 82.7 ± 11.9 
Cingulate Gyrus   6 33%    3 66%  
F1, F2 gyr (BA 9, 10, 46) 14 57% 30.6 ± 8.8 33% 43 (1 patient) 
Orbital gyr (BA 11, 12, 47) 11 45% 46.9 ± 13.8 0% NA 
Frontal cortex   25 52%    6 17%  
Striate cortex (BA 17) 0% NA 0% NA 
Int occ ctx except area 17 (BA 18, 19) 13 46% 38.4 ± 13.3 25% NA 
Ext occipital ctx (BA 18, 19) 12 58% 18.9 ± 4.51 33% 37.1 (1 patient) 
Occipital cortex   30 43%    8 25%  
 CEP
 
PEPs
 
 Number of tested patients Number of responding patients Number of tested contact pairs Response rate Onset latency (mean ± SD) (ms) Number of tested patients Number of responding patients Number of tested contact pairs Response rate Onset latency (mean ± SD) (ms) 
STG (BA 22) 37 95% 11.4 ± 2.6 100% 56.6 ± 32.6 
MTG (BA 21, 37) 50 70% 13.7 ± 2.7 14 71% 56.8 ± 46.4 
ITG (BA 20, 37) 27 70% 16.9 ± 7.3 50% 25.4 ± 16.6 
FG (BA 36) 80% 37.1 ± 7.6 50% 11.72 (1 patient) 
Temporal neocortex   119 78%    29 76%  
Hippocampus 19 16% 27.3 ± 5.5 83% 27.8 ± 7.4 
Amygdala 15 20% 60.5 (1 patient) 100% 46.9 ± 7 
Temporal Pole (BA 38) 15 47% 66.4 ± 52 50% 33.2 ± 7.8 
Parahipp Gyr (BA 28, 34, 35) 12 67% 71 ± 47.3 100% 43.5 ± 28.2 
Mesial temporal structures   61 34%    20 80%  
Postcentral gyrus (BA 3, 2, 1) 0% NA NA NA NA 
TP Junction (BA 39, 40) 21 81% 15.2 ± 4.9 67% 23.4 ± 2.8 
Precuneus (BA 7) 11 46% 29.3 ± 19.8 0% NA 
Parietal cortex   34 65%    5 40%  
Insula 16 94% 17.3 ± 6.7 17% 41 (1 patient) 
FP operculum 13 85% 27.8 ± 16.1 0% NA 
Insuloopercular cortex   29 90%    7 14%  
Cing ant gyr (BA 24, 32) 0% NA 0% NA 
Cing postgyr (BA 23, 31) 67% 17.58(1 patient) 100% 82.7 ± 11.9 
Cingulate Gyrus   6 33%    3 66%  
F1, F2 gyr (BA 9, 10, 46) 14 57% 30.6 ± 8.8 33% 43 (1 patient) 
Orbital gyr (BA 11, 12, 47) 11 45% 46.9 ± 13.8 0% NA 
Frontal cortex   25 52%    6 17%  
Striate cortex (BA 17) 0% NA 0% NA 
Int occ ctx except area 17 (BA 18, 19) 13 46% 38.4 ± 13.3 25% NA 
Ext occipital ctx (BA 18, 19) 12 58% 18.9 ± 4.51 33% 37.1 (1 patient) 
Occipital cortex   30 43%    8 25%  

Note: The averaged response rates (number in bold) are calculated from the whole responding contacts pairs divide by the sum of all tested contacts pairs in the 7 patients, for the lobe or the cerebral structure considered. NA, nonavailable; STG, MTG, and ITG, superior, middle, and inferior temporal gyri; FG, fusiform gyrus; Parahipp gyr, parahippocampal gyrus; TP junction, temporoparietal junction (including angular and supramarginal gyri); FP, frontoparietal; F1 and F2, superior and middle frontal gyri; Int occ and ext occ ctx, internal occipital and external occipital cortex; BA, Brodmann area; cing ant gyr, cingulate anterior gyrus; cing post gyr, cingulate posterior gyrus.

Data Collection and Analysis

Signals were recorded through bipolar derivation between 2 adjacent contacts located either in the cortical gray matter or in PuM using a band-pass filter of 0.3–200 Hz and a sampling frequency of 512 Hz, which allowed a time resolution of 2 ms. Averaging of responses to each series of 25 stimuli was performed off-line using a time window of 4.5 s, with 500-ms pre- and 4000-ms poststimulus periods. Cortical responses to PuM stimulations were named cortical-evoked potentials (CEPs) and PuM responses to cortical stimulations were named pulvinar-evoked potentials (PEPs). They were analyzed on bipolar recordings between 2 adjacent contacts in all cases. To be considered present, CEPs and PEPs should fulfill 2 conditions (Fig. 2):

  • reproducibility between 2 consecutive series and

  • amplitude at least twice that of the background level.

Figure 2.

Example of CEPs recording from the mesial part of the superior frontal gyrus (F1) and from the middle frontal gyrus (F2) in patient number 7. Black and gray traces correspond to the 2 averaged responses of 2 series of 25 single-pulse low-frequency stimulations. Localization of the recording contact pairs (black ellipses) are plotted on the patient's MRI coronal and sagittal slices and on a normalized surface-rendering brain image (see legend of Fig. 3). Pairs 2–3, 9–10, 10–11, and 11–12 (filled ellipses) recorded significant CEPs, whereas no responses were observed from the remaining contacts pairs (pairs 1–2 and 12–13 that are represented as empty ellipses). CEPs were thus considered present in 4 out of 6 contact pairs, giving a response rate of 67%. Note that EPs morphology could vary from one pair to the other, even when contact pairs are adjacent (see 9–10 and 10–11).

Figure 2.

Example of CEPs recording from the mesial part of the superior frontal gyrus (F1) and from the middle frontal gyrus (F2) in patient number 7. Black and gray traces correspond to the 2 averaged responses of 2 series of 25 single-pulse low-frequency stimulations. Localization of the recording contact pairs (black ellipses) are plotted on the patient's MRI coronal and sagittal slices and on a normalized surface-rendering brain image (see legend of Fig. 3). Pairs 2–3, 9–10, 10–11, and 11–12 (filled ellipses) recorded significant CEPs, whereas no responses were observed from the remaining contacts pairs (pairs 1–2 and 12–13 that are represented as empty ellipses). CEPs were thus considered present in 4 out of 6 contact pairs, giving a response rate of 67%. Note that EPs morphology could vary from one pair to the other, even when contact pairs are adjacent (see 9–10 and 10–11).

Responses evoked by intracerebral stimulation can theoretically result either from orthodromic or antidromic inputs transmission along connecting pathways, so that they can reflect either feed forward transmission to the recorded area or feed back transmission from the stimulated area. It is generally accepted that the initial segment of axons is the most excitable element recruited by electrical stimulation (Rattay 1999) and that responses evoked by electrical stimulation reflect mainly postsynaptic potentials resulting from indirect mono or polysynaptic cell activation (Jankowska et al. 1975). Therefore, the responses we recorded are likely to reflect orthodromic transmission. Even though the possibility of retrograde transmission cannot be definitely ruled out, we considered CEPs as reflecting activation of thalamocortical pathways and PEPs as resulting from activation of reciprocal corticothalamic connections.

Thalamocortical functional connection for a given cortical region was estimated by calculating the percentage of cortical contact pairs showing CEPs after PuM stimulation, with respect to the total number of contact pairs exploring this region (see Fig. 2 and legend). The reciprocal pathway was quantitatively evaluated by the percentage of stimulated contact pairs inside a given cortical area, which were able to elicit PEPs. PEPs were considered present when observed on at least one bipolar derivation inside PuM. In the majority of cases, PEPs were either absent or recorded on all PuM contacts pairs.

EPs latencies were measured from the onset of their ascending slope (onset latencies). Maximal amplitudes and responses duration were also considered, and we attempted to classify responses according to their morphology.

Finally, to evaluate the potential influence of epilepsy on the EPs, we reviewed all the seizures recorded in each individual patient and determined the seizure onset zone, defined as the cerebral regions exhibiting the earliest ictal activities. All bipolar derivations involved during the first second of the epileptic discharge were considered located inside the seizure onset zone and then labeled epileptic derivations.

Results

Global Connectivity

An overview of cortical–PuM connectivity is given in Figure 3. Furthermore, localizations of CEPs and PEPs in each cerebral structure and for each exploring contact pairs are presented on individual patient MRI slices as Supplementary Material. None of the single-shock stimuli delivered to the cortex or PuM for EP recordings produced any clinically detectable response nor any subjective somatosensory, visual, or auditory sensation. PuM stimulation evoked CEPs in all explored cerebral regions except the anterior cingulate gyrus (2 patients), the striate cortex (2 patients), and the postcentral gyrus (1 patient). PEPs were also recorded after stimulation of all investigated cortical regions, except the dorsolateral frontal lobe (2 patients), the orbital cortex (2 patients), the anterior cingulate gyrus (1 patient), the frontal operculum (1 patient), and the striate cortex (1 patient). Thus, only 2 regions, the anterior cingulate gyrus and the striate cortex, did not exhibit either CEPs or PEPs. PEPs after postcentral gyrus stimulation were not tested.

Figure 3.

Overview of global PuM–cortex connectivity. All patients' data were plotted on normalized T1 MRI brain images. For each patient, all recording electrode contacts were reconstructed on their individual 3-D T1 MRI volume using MRIcro software. Then all contacts were saved as a region of interest (ROI) in an analyzed format. Right implantations were flipped on the left side. In a next step, each anatomical image was normalized to the standard T1 MNI template using SPM2 (Wellcome Department of Imaging Neuroscience, London, UK). The normalizing parameters were applied to the anatomical and ROI images, which were resampled to 2 mm of isotropic voxel size. Finally, all normalized anatomical images were averaged in order to have a brain template representative of our patient group allowing visualization of all ROIs. Left column: filled circles were plotted when at least one of the contact pairs exploring a given cerebral structure exhibited a CEPs response after PuM stimulation; empty circles: absence of CEP; yellow circle: CEP in patients with thalamic contact pair located in the posterior part of PuM (patients 1–4); blue circle: CEP in patients with thalamic contact pair located in the anterior part of PuM (patients 5–7). Right column: filled circles were drawn when stimulation through at least one of the contact pairs localized in a given cerebral structure evoked a PEP response; empty circles: absence of PEP; yellow circle: PEP in patients with thalamic contact pair located in the posterior part of PuM (patients 1–4); blue circle: PEP in patients with thalamic contact pair located in the anterior part of PuM (patients 5–7). CS, central sulcus; Sylv S, sylvian scissure; TOS, temporooccipital scissure; Ins, insula; Fusif Gyr, fusiform gyrus.

Figure 3.

Overview of global PuM–cortex connectivity. All patients' data were plotted on normalized T1 MRI brain images. For each patient, all recording electrode contacts were reconstructed on their individual 3-D T1 MRI volume using MRIcro software. Then all contacts were saved as a region of interest (ROI) in an analyzed format. Right implantations were flipped on the left side. In a next step, each anatomical image was normalized to the standard T1 MNI template using SPM2 (Wellcome Department of Imaging Neuroscience, London, UK). The normalizing parameters were applied to the anatomical and ROI images, which were resampled to 2 mm of isotropic voxel size. Finally, all normalized anatomical images were averaged in order to have a brain template representative of our patient group allowing visualization of all ROIs. Left column: filled circles were plotted when at least one of the contact pairs exploring a given cerebral structure exhibited a CEPs response after PuM stimulation; empty circles: absence of CEP; yellow circle: CEP in patients with thalamic contact pair located in the posterior part of PuM (patients 1–4); blue circle: CEP in patients with thalamic contact pair located in the anterior part of PuM (patients 5–7). Right column: filled circles were drawn when stimulation through at least one of the contact pairs localized in a given cerebral structure evoked a PEP response; empty circles: absence of PEP; yellow circle: PEP in patients with thalamic contact pair located in the posterior part of PuM (patients 1–4); blue circle: PEP in patients with thalamic contact pair located in the anterior part of PuM (patients 5–7). CS, central sulcus; Sylv S, sylvian scissure; TOS, temporooccipital scissure; Ins, insula; Fusif Gyr, fusiform gyrus.

Temporal–PuM Connections

Because the epileptogenic zone was suspected to be of temporal origin in all patients, the temporal lobe was the most extensively explored region (119 contacts pairs in the temporal neocortex, mean ± standard deviation [SD] = 17 ± 5.7 per patient, and 61 contacts pairs in the mesiotemporal structures, mean ± SD = 8.7 ± 2.9 per patient). We analyzed separately results obtained in: 1) the neocortical regions, namely the superior, middle, inferior, and fusiform temporal gyrus and 2) the mesial temporal region, that is, the hippocampus, the amygdala, the parahippocampal gyrus including entorhinal cortex and the temporal pole. Data collected for each particular structure are given in Table 2.

CEPs were recorded from 78% of the contacts pairs located in the temporal neocortex, and PEPs were elicited by stimulation of 76% of the neocortical contacts pairs stimulated. STG showed the highest percentage of reciprocal functional connection with PuM. In all, 95% of the STG contacts pairs exhibited CEPs after PuM stimulation and 100% of the STG stimulations induced PEPs. Distribution and percentages of responses were approximately similar in all patients except one, in whom CEPs were totally absent and cortical stimulations could not be performed.

The results obtained in the mesial temporal region were strikingly different. Indeed, CEPs were recorded from only 34% of the 61 tested contacts pairs, with large variations across structures. For example, the largest number of contact pairs responding to PuM stimulation (67%) was observed in the parahippocampal gyrus, whereas this rate was 47% in the temporal pole, 20% in the amygdala, and only 16% in the hippocampus. Unlike the temporal neocortex, mesial temporal structures also showed variable CEPs response rates among patients (see Table 2). Conversely, stimulation of these mesial temporal structures evoked very consistent PuM responses: parahippocampal gyrus and amygdala stimulation elicited PEPs in all patients, hippocampus stimulation in 4 of 5, and temporal pole stimulation in 3 of 6. Overall, a percentage of 80% of the stimulated contacts pairs located in mesial temporal structures were able to evoke PEPs (amygdala and parahippocampal gyrus: 100%; hippocampus: 83%; and temporal pole: 50%).

Parietal–PuM Connections

With a total of 34 contacts pairs (5 ± 4 per patient), the parietal lobe was the third more fully explored cortical area, after temporal neocortex and mesial temporal region. Functional connectivity strength varied greatly across parietal regions: The CEPs rate was 81% in the temporoparietal (TP) junction, with a small interindividual variability, whereas it dropped to 46% in the precuneus gyrus. In the postcentral gyrus, the only patient tested did not present CEPs. Thus the PuM–TP functional connectivity, including that of the angular and supramarginal gyrus, was comparable to that of the temporal neocortex regarding CEPs, whereas reciprocal connections looked slightly weaker with 67% of PEPs' responses.

Insula and Frontoparietal Operculum–PuM Connections

With the superior temporal gyrus (see above), the insula and the frontoparietal operculum were the 2 structures showing the highest CEPs rate after PuM stimulation, with 94% and 85% of CEPs' positive contacts pairs, respectively. CEPs were recorded in all patients with insular recordings. In all, 100% of the insular contacts pairs were CEPs positive in 5 patients and 67% in 1. Frontoparietal operculum CEPs rates were 100% in 3 patients and 50% in the fourth patient tested. However, unlike the temporal neocortex and the TP junction, the reciprocal corticothalamic functional connectivity was consistently much weaker. All together, only 14% of the stimulated contact pairs located in these 2 regions were able to elicit PEPs. No PEPs were recorded in response to stimulation of the frontoparietal operculum, in the only patient tested, and the PEPs' rate after stimulation of the insula was only 17% (PEPs present in 1 patient, absent for the other 4 tested).

Cingulate Gyrus–PuM Connections

No CEPs were evoked by PuM stimulation in the anterior part of the cingulate gyrus and no PEPs were recorded after stimulation of this area in 2 patients, whereas reciprocal functional connections were observed in 2 other patients between PuM and posterior cingulate gyrus (67% CEPs and 100% PEPs).

Fronto– and Occipito–PuM Connections

Twenty-five contact pairs explored the frontal lobe in 3 patients. We observed a relatively weak but constant connectivity in all these patients: 40% to 50% of CEPs but no PEPs in the orbital gyri (2 patients), 40 to 67% of CEPs, and 33% of PEPs in the dorsolateral frontal cortex (3 patients).

In 4 of the 5 patients whose occipital lobes were tested, 43% of PuM stimulations produced CEPs, and PEPs following cortical stimulations were recorded in only 1 out of 4 patients (25%). All the occipital contacts pairs recording or inducing EPs were situated in the extrastriate cortex (see Table 2).

Morphology, Latency, and Amplitude of the EPs

No distinct morphological pattern of CEPs and PEPs could be distinguished throughout the explored structures. CEPs responses were frequently polyphasic, made of one or several waves. Other morphological types were monophasic, biphasic, and repetitive responses (repetition of the same potential, separated by a significant interval with return to the baseline). For PEPs, morphological types were simpler, generally mono or biphasic. Some examples of CEPs and PEPs waves are given in Figures 2 and 4.

Figure 4.

CEPs and PEPs obtained in the same cortical structure (middle part of the middle temporal gyrus [MTG]). Middle panel: implantation site of electrodes are visualized on the surface rendering of our averaged normalized brain, each number representing one patient. Left panel: CEPs recorded from MTG when stimulating PuM. Their morphologies, frequently polyphasic, and their amplitudes are highly variable. Right panel: PEPs recorded in PuM when stimulating MTG. Like for CEPs, amplitudes vary greatly but morphology is more constant, frequently biphasic. PEPs in patient 7 were not obtained because MTG contacts pairs were not stimulated. Thick traces correspond to EPs recorded in patients 4 and 6, whose epileptogenic zone included MTG.

Figure 4.

CEPs and PEPs obtained in the same cortical structure (middle part of the middle temporal gyrus [MTG]). Middle panel: implantation site of electrodes are visualized on the surface rendering of our averaged normalized brain, each number representing one patient. Left panel: CEPs recorded from MTG when stimulating PuM. Their morphologies, frequently polyphasic, and their amplitudes are highly variable. Right panel: PEPs recorded in PuM when stimulating MTG. Like for CEPs, amplitudes vary greatly but morphology is more constant, frequently biphasic. PEPs in patient 7 were not obtained because MTG contacts pairs were not stimulated. Thick traces correspond to EPs recorded in patients 4 and 6, whose epileptogenic zone included MTG.

Mean onset latencies of CEPs and PEPs are given for each structure in Table 2. Most of CEPs' onset latencies (65%) were equal or inferior to 30 ms and strongly homogeneous across patients as indicated by relatively small SD values. The shortest CEPs latencies were observed in the temporal neocortex, the TP junction, and the insula (onset latencies <20 ms). In contrast, PEPs latencies were generally longer and more variable, except for some mesial temporal structures and TP junction. The limited number of values obtained in each structure did not allow statistical analysis except for the temporal lobe where we could compare neocortical versus mesiolimbic EPs characteristics. A Kruskal–Wallis test revealed significant differences between these 2 structures (P < 0.0001), and Dunn's multiple comparison post tests showed that CEPs onset latencies were shorter in the neocortical (13.9 ± 5 ms) than in the limbic component (58.3 ± 42.2 ms; P < 0.001), whereas onset latencies of PEPs were not significantly different (neocortex: 51.8 ± 37.5 ms, limbic: 34.9 ± 17.8 ms). Within temporal neocortex, onset latencies were shorter for CEPs (13.9 ± 5 ms) than for PEPs (51.8 ± 37.5 ms; P < 0.001), whereas in the mesiolimbic cortex, CEPs' (58.3 ± 42.2 ms) and PEPs' (34.9 ± 17.8 ms) onset latencies were not different.

Both CEPs and PEPs amplitudes showed large variations not only between patients but also between contacts pairs exploring the same structure (see Figs 2 and 4).

Impact of Epilepsy on CEPs and PEPs Responses

Cortical or PuM stimulation procedure never produced paroxysmal events like epileptic spikes or ictal subclinical discharges. We were also unable to distinguish distinct morphological patterns of response between contact pairs located in or out the seizure onset zone The influence of epilepsy on the functional efficiency of a given pathway could be quantitatively evaluated only on data obtained within the temporal lobe, where a sufficient number of epileptic derivations could be compared with “nonepileptic” ones.

The following parameters were considered: percentages of CEPs and PEPs responses, onset latencies, amplitudes, and responses duration (Fig. 5). In the nonepileptogenic temporal neocortex, 91% and 62% of contact pairs exhibited CEPs and PEPs, respectively. In contrast, only 50% of contacts pairs exploring the similar but epileptogenic neocortex give CEPs as well as PEPs. In the mesiotemporal structures, we observed a similar trend between nonepileptogenic (40% CEPs, 91% PEPs) and epileptogenic areas (27% CEPs, 67% PEPs). However, only CEPs rate in temporal neocortex was significantly lower in epileptogenic versus nonepileptogenic cortical structures (Fisher's t exact test; 2-sided P < 0.0001). Onset latencies as well as response duration tended to be longer in the epileptogenic cortex, except for CEPs duration in limbic structures, but none of these trends was statistically significant (Fig. 5). No CEPs or PEPs amplitude difference between epileptogenic and nonepileptogenic tissue was observed whatever the stimulation site (cortical or thalamic) or the cerebral structure (neocortex or mesiotemporal) considered (Fig. 5).

Figure 5.

Diagrams showing response rates, mean ± SD latencies, amplitudes and response durations in epileptic and nonepileptic temporal neocortex (left column), and mesial temporal structures (right column). For response rates, 2-sided P values were obtained after statistical analysis using the Fisher's exact test. For latency, amplitude, and duration values, the Student's t-test or the Mann–Whitney test was applied depending of the Gaussian or non-Gaussian distribution of these values. P values inferior to 0.05 are considered as significant.

Figure 5.

Diagrams showing response rates, mean ± SD latencies, amplitudes and response durations in epileptic and nonepileptic temporal neocortex (left column), and mesial temporal structures (right column). For response rates, 2-sided P values were obtained after statistical analysis using the Fisher's exact test. For latency, amplitude, and duration values, the Student's t-test or the Mann–Whitney test was applied depending of the Gaussian or non-Gaussian distribution of these values. P values inferior to 0.05 are considered as significant.

Response Rate and Distribution of Cortical Connectivity According to Contact Pairs Localization within PuM

Because the localization of contact pairs within the PuM differed between patients, we attempted to determine if there was some topical organization of cortical projections within the PuM. For this purpose, we individualized 2 groups of contacts; group 1 patients (1– 4) had contact pairs located in the medial and posterior part of PuM, whereas group 2 patients (5–7) had contact pairs in anterior part of the PuM (see Fig. 1).

In spite of large overlap (see Fig. 3), the CEPs rate was significantly higher in the temporal neocortex for stimulation of the posteromedial PuM (group 1: 88%; group 2: 61.4%; Fisher's exact test, 2-sided P value = 0.011). Conversely, in the mesial temporal region, CEPs rate was higher for stimulation of the anterior part of PuM (group 2: 47.6%; group 1: 16%; Fisher's exact test, 2-sided P value = 0.027). Finally, there was no significant difference in CEPs rates in the insula and frontoparietal operculum for stimulation of anterior or posterior PuM contact pairs. Because of small numbers, this analysis could not be performed on CEPs in other cortical regions and for PEPs data.

Discussion

This is the first study dedicated to the PuM functional connectivity and based on depth simulation and EP recordings in the human brain. We observed that this connectivity concerns large cortical territories, involving all cerebral lobes to some different degrees, and is coherent with most of the anatomical data available in monkeys. In humans, preferential and fastest pathways were found to interconnect the PuM with the temporal neocortex, the TP junction, and the insuloopercular region.

Previous animal studies have established the validity of our methodology for testing connectivity, with a demonstrated parallelism between electrophysiological recordings and anatomical findings (Tanibuchi 1992). In humans, this technique has remained rarely used because it can be applied only in epileptic patients whose presurgical evaluation necessitates intracerebral electrodes implantation. Furthermore, the spatial sampling is limited to brain areas suspected to be involved in seizures onset or propagation and thus varies from one patient to the other. However, such depth stimulations and recordings provide a unique opportunity to map in vivo the functional connections that may exist between given cerebral regions (Brazier 1964; Catenoix et al. 2005; Zumsteg, Lozano, and Wennberg 2006; Zumsteg, Lozano, Wieser, and Wennberg 2006).

However, some confounding factors or technical issues deserve attention when interpreting our data:

  • i) Even if our electrical stimulations never provoked any epileptic spike or discharge when applied in the epileptogenic zone, it cannot be ruled out that the epileptic condition of the patients could modify the evoked responses, by increasing or decreasing the neuronal excitability in the stimulated or recorded areas. To test this hypothesis, we have studied separately EPs recorded from “epileptogenic” and “nonepileptogenic” cerebral tissue in the temporal neocortex and mesial temporal structures. In both regions, our results suggested a decreased connectivity and/or excitability (decreased response rate, delayed latencies) in the epileptogenic zone, which was statistically significant for CEPs in the temporal neocortex of our patients with TLE. Thus, we may have underestimated, but not overestimated, the functional connection between PuM and its cortical targets. Another interesting point is the pathophysiological significance of this result. FDG-PET studies in temporal lobe epileptic patients have found hypometabolic cortical and subcortical areas, which could correspond to the neuronal network involved during ictal process (Chassoux et al. 2004). For the thalamus, which has been found hypometabolic in numerous PET studies (see Mauguière 2004 for review, Joo et al. 2005; Hashiguchi et al. 2007,), it has been postulated that anatomical neuronal loss in the epileptic focus is the causal factor explaining the decreased glucose metabolism (Dlugos et al. 1999). Our results support previous PET findings and indeed suggest that functional connections are reduced between PuM and its epileptogenic target areas in patients with TLE.

  • ii) Even with stimulation applied between 2 adjacent contacts separated by 1.5 mm, one issue concerns the volume of cerebral tissue effectively excited by stimulation (sphere of 4 mm radius, see Materials and Methods). This point is particularly critical to decide whether the physiological effects of our thalamic stimuli were limited to PuM or may have involved the whole pulvinar structure and eventually adjacent thalamic nuclei. Recently, Zumsteg, Lozano, and Wennberg (2006) have shown that bipolar stimulations delivered either in anterior, dorsomedian, or centromedian thalamic nuclei evoke specific pattern of distributed cortical responses coherent with the anatomical connectivity specific to each of these nuclei. Similarly, we observed differences between the functional cortical connections of anterior and posterior PuM suggesting that the effects of stimulation remains focal inside the PuM nucleus and consistent with a topical organization of PuM connections with the temporal lobe. Because of large distance, the effects of posterior PuM stimuli are unlikely to have spread to neighbor thalamic nuclei. Moreover, the higher CEPs rate in the mesial temporal regions after stimulation of the anterior PuM is unlikely to reflect a diffusion to nearby central lateral and anterior pulvinar nuclei because these nuclei are not known to be connected with mesial temporal structures. Lastly, we never observed CEPs in the striate cortex as expected in case of stimulus diffusion to the inferior pulvinar, which is connected with the striate cortex (Rezak and Benevento 1979; Lysakowski et al. 1988; Romanski et al. 1997; Adams et al. 2000).

  • iii) Another issue is that of the stimulus spread to thalamocortical fibers, which might take origin in neighbor thalamic nuclei and travel through PuM on their way to the cortex. The anatomical position of the PuM within the thalamus makes this possibility very unlikely. PuM is the most caudal thalamic nucleus and its posterior part is bordered medially by the lateral ventricle (see Duvernoy 1991); more rostrally, PuM is bounded laterally by the lateral pulvinar subdivision, ventrally by the brachium of the superior colliculus, and dorsally by the lateral posterior nucleus (see Morel's [2007] atlas). Due to this anatomical configuration, there is no clear evidence of fibers passing through the core of the PuM, especially in its most caudal extent, contrasting with the presence of fibers bundles in the inferior and lateral pulvinar subdivisions (Robinson and Cowie 1997).

The opinion that our thalamic stimulations are likely to have involved specifically the PuM nucleus are reinforced by the fact that our functional connectivity results are coherent with most of conclusions from hodologic studies in nonhuman primates (Mufson and Mesulam 1984; Baleydier and Mauguiere 1985, 1987; Aggleton et al. 1986; Insausti et al. 1987; Yeterian and Pandya 1988; Baleydier and Morel 1992; Romanski et al. 1997; Gutierrez et al. 2000). In monkeys, the cerebral cortex represents the major part of PuM-connected structures, apart from some subcortical inputs, mainly from the deep layers of the superior colliculus (Benevento and Fallon 1975), and our study shows that PuM has also a widespread cortical connectivity in the human brain. These connections are developed almost exclusively with associative cortical areas, even if some studies have reported few labeled cells in PuM resulting from injections in primary somatosensory or auditory cortex (De Vito 1978; De La Mothe et al. 2006). Among visual cortical areas, PuM connectivity concerns extrastriate cortex, particularly V4 and inferotemporal cortex (Weller et al. 2002; Adams et al. 2000; Yeterian and Pandya 1997). PuM connectivity with auditory and somatosensory associative areas like parietal BA 7 and 5, superior temporal gyrus, and superior temporal sulcus have also been extensively reported (Baleydier and Mauguiere 1977; Yeterian and Pandya 1985; Baleydier and Morel 1992; Cappe et al. 2007; Trojanowski and Jacobson 1975; Mauguiere and Baleydier 1978; Yeterian and Pandya 1989, 1991; Hackett et al. 1998; Gutierrez et al. 2000). Moreover, PuM interconnects with higher order cortex and “paralimbic” associations areas, including insular, temporopolar, prefrontal, orbital, parahippocampal and cingulate cortex, as well as the amygdala (Jones and Burton 1976; Price and Amaral 1981; Mufson and Mesulam 1984; Insausti et al. 1987; Yeterian and Pandya 1988; Barbas et al. 1991; Romanski et al. 1997). Finally, PuM has connections with frontal eye field areas (Huerta et al. 1986; Romanski et al. 1997).

In humans, we found that PuM connectivity is particularly developed with the associative cortices of temporal and parietal lobes and, to a lesser degree, with those of occipital and frontal lobes. Consistent with these data is the finding that no CEPs to PuM stimulation and no PEPs to cortical stimulation could be recorded within primary somatosensory and visual cortex, even if this result is to be considered cautiously given the few number of contact pairs tested in these areas. Furthermore, we showed that, quantitatively, functional and reciprocal connections between PuM and associative cortical areas concerned mostly the temporal neocortex and the TP junction, each one with CEPs response rate superior to 80% and PEPs response rate among the highest as well.

Some of our results, however, are slightly diverging from animal studies. This is the case for the reciprocal PuM–insula connections mentioned in some early monkey studies (Mufson and Mesulam 1984; Romanski et al. 1997) but later neglected in review articles (Augustine 1996; Shelley and Trimble 2004). Our results emphasize that connection in humans by showing an insular CEPs response rate greater than that of the TP junction and nearly equal to that of the superior temporal gyrus, 2 structures known to be tightly connected with PuM. However, we observed limited reciprocal connection from insula to PuM. PuM might thus play a substantial role as an inputs source in the multimodal functionality of the insula, extending from somatosensory, pain, and visceral domains to numerous cognitive functions (Augustine 1996; Ostrowsky et al. 2002; Shelley and Trimble 2004; Mazzola et al. 2006), but with a weak reciprocal influence of the insular cortex.

Large discrepancies between CEPs and PEPs rates were also observed in the mesiotemporal structures, where the CEPs' response rate was the lowest of all the cerebral structures that we tested (34%), whereas the PEPs' response rate was the highest, close to 80%. The asymmetrical connectivity between PuM and the above-cited structures could appear surprising. Generally, the connections between the major sensory thalamic relay nuclei and cortex are considered as “reciprocal” (for review, see Steriade et al. 1997). However, some topographical mismatches between corticothalamic and reciprocal thalamocortical pathways are known in monkey (Darian-Smith et al. 1999; Kultas-Ilinsky et al. 2003). Due to the restricted volume of our stimulations, only a limited part of cortical areas have been explored thus biasing the amount of PEPs toward low values. However, this bias does not explain the high PEPs rates observed in some cortical structures (mesial temporal structures and posterior cingulate gyrus). Our results could thus reflect a true quantitative imbalance in some reciprocal PuM-cortical pathways.

No connectivity has ever been found between pulvinar and hippocampus proper in nonhuman primates, whereas we have obtained CEPs as well as PEPs in some of our patients, resulting in 16% and 83% response rates, respectively. Aggleton et al. (1986) described pulvinar afferents from the subiculum, and it cannot be excluded that our hippocampal stimulation has involved this very neighboring structure.

The last of our results that differs from anatomical data in monkeys concerns the amygdala connectivity. Whereas a slight projection to PuM originating from the central amygdaloid nucleus has been mentioned by Price and Amaral (1981), the existence of a reciprocal pathway connecting PuM with the lateral amygdala is well established (Jones and Burton 1976; Romanski et al. 1997). Conversely, we consistently recorded PEPs after amygdala stimulation, whereas CEPs were registered in only 1 out of 5 patients tested. To summarize, our results confirm the well-known PuM connectivity with mesial temporal structures. They also suggest that, in humans, cortical mesial temporal areas are able to modulate strongly PuM activity, whereas they receive much weaker functional afferents from PuM.

Our functional connectivity data may help to better integrate and emphasize the role of PuM in previous hypothesis about the functional role of the pulvinar nucleus, which is mostly supposed to play a key role in visual attention processing and selectivity (Shipp 2003, 2004). Most of the afferent fibers to PuM conveying driving inputs to this nucleus arise from associative cortical regions. This is opposite to what is observed for first-order thalamic relay nuclei that receive only modulatory inputs from the cortex, whereas their major driving inputs come from ascending subthalamic pathways (Guillery and Sherman 2002, Sherman and Guillery 2002). Moreover, in the human thalamus, PuM is one of the most highly stained nucleus with calbindin (see Münkle et al. 2000; Table 2), a calcium-binding protein, which is a marker of the matrix thalamic network (Jones 1998, 2001, 2002). These thalamocortical matrix cells are characterized by glutamatergic projections to superficial cortical layers, showing a fuzzy topical distribution that spreads beyond cortical areas providing reciprocal cortical projections. Such an organization of connectivity between PuM and cortex would allow the modulation of corticocortical networks, some of them dominating the others when reinforced by their indirect but parallel corticothalamocortical counterpart (Shipp 2003). Because of its extensive connections with multimodal associative cortical areas, PuM as a whole may be a nodal point where numerous simultaneous perceptions converge and compete. In addition, the well-developed corticothalamic limbic pathways that modulate PuM activity may influence attention selectivity according to the emotional context.

Lastly, our results are encouraging for patients with pharmacoresistant epileptic seizures arising from insula and TP junction, in particular in the language dominant hemisphere. These regions, due to their deep anatomical situation or their functional importance, are poorly accessible to surgical resection. PuM is involved in temporoinsular seizures development (Rosenberg et al. 2006) and is tightly connected with temporal lobe and insula. Chronic electrical stimulation might thus be used to control the epileptic activity in these cortical regions.

Supplementary Material

Supplementary material can be found at: http://www.cercor.oxford journals.org/.

Funding

French Ligue Against Epilepsy; supported by the Hospices Civils de Lyon.

We wish to thanks Dr Jean Isnard and Prof. Philippe Ryvlin for their help in interpreting SEEG data. Conflict of Interest: None declared.

References

Adams
MM
Hof
PR
Gattass
R
Webster
MJ
Ungerleider
LG
Visual cortical projections and chemoarchitecture of macaque monkey pulvinar
J Comp Neurol
 , 
2000
, vol. 
419
 (pg. 
377
-
393
)
Aggleton
JP
Desimone
R
Mishkin
M
The origin, course, and termination of the hippocampothalamic projections in the macaque
J Comp Neurol
 , 
1986
, vol. 
243
 (pg. 
409
-
421
)
Andrade
DM
Zumsteg
D
Hamani
C
Hodaie
M
Sarkissian
S
Lozano
AM
Wennberg
RA
Long-term follow-up of patients with thalamic deep brain stimulation for epilepsy
Neurology
 , 
2006
, vol. 
66
 (pg. 
1
-
3
)
Augustine
JR
Circuitry and functional aspects of the insular lobe in primates including humans
Brain Res Rev
 , 
1996
, vol. 
22
 (pg. 
229
-
244
)
Baleydier
C
Mauguiere
F
Pulvinar-latero posterior afferents to cortical area 7 in monkeys demonstrated by horseradish peroxydase tracing technique
Exp Brain Res
 , 
1977
, vol. 
27
 (pg. 
501
-
507
)
Baleydier
C
Mauguiere
F
Anatomical evidence for medial pulvinar connections with the posterior cingulate cortex, the retrosplenial area, and the posterior parahippocampal gyrus in monkeys
J Comp Neurol
 , 
1985
, vol. 
232
 (pg. 
219
-
228
)
Baleydier
C
Mauguiere
F
Network organization of the connectivity between parietal area 7, posterior cingulate cortex and medial pulvinar nucleus: a double fluorescent tracer study in monkey
Exp Brain Res
 , 
1987
, vol. 
66
 (pg. 
385
-
393
)
Baleydier
C
Morel
A
Segregated thalamocortical pathways to inferior parietal and inferotemporal cortex in macaque monkey
Vis Neurosci
 , 
1992
, vol. 
8
 (pg. 
391
-
405
)
Barbas
H
Henion
TH
Dermon
CR
Diverse thalamic projections to the prefrontal cortex in the rhesus monkey
J Comp Neurol
 , 
1991
, vol. 
313
 (pg. 
65
-
94
)
Behrens
TEJ
Johansen-Berg
H
Woolrich
MW
Smith
SM
Wheeler-Kingshott
CAM
Boulby
PA
Barker
GJ
Sillery
EL
Sheehan
K
Ciccarelli
O
, et al.  . 
Non-invasive mapping of connections between human thalamus and cortex using diffusion imaging
Nat Neurosci
 , 
2003
, vol. 
6
 (pg. 
750
-
757
)
Benevento
LA
Fallon
JH
The ascending projections of the superior colliculus in the rhesus monkey (Macaca mulatta)
J Comp Neurol
 , 
1975
, vol. 
160
 (pg. 
339
-
361
)
Bertashius
K
Propagation of human complex-partial seizures: a correlation analysis
Electroencephalogr Clin Neurophysiol
 , 
1991
, vol. 
78
 (pg. 
333
-
340
)
Brazier
MA
Evoked responses recorded from the depths of the human brain
Ann N Y Acad Sci
 , 
1964
, vol. 
12
 (pg. 
33
-
59
)
Cappe
C
Morel
A
Rouiller
EM
Thalamocortical and the dual pattern of corticothalamic projections of the posterior parietal cortex in macaque monkeys
Neuroscience
 , 
2007
, vol. 
146
 (pg. 
1371
-
1387
)
Catenoix
H
Magnin
M
Guénot
M
Isnard
J
Mauguière
F
Ryvlin
P
Hippocampal-orbitofrontal connectivity in human: an electrical stimulation study
Clin Neurophysiol
 , 
2005
, vol. 
116
 (pg. 
1779
-
1784
)
Chassoux
F
Semah
F
Bouilleret
V
Landre
E
Devaux
B
Turak
B
Nataf
F
Roux
FX
Metabolic changes and electro-clinical patterns in mesio-temporal lobe epilepsy: a correlative study
Brain
 , 
2004
, vol. 
127
 (pg. 
164
-
174
)
Clarke
S
Riahi-Arya
S
Tardif
E
Cottier Eskenasy
AC
Probst
A
Thalamic projections of the fusiform gyrus in man
Eur J Neurosci
 , 
1999
, vol. 
11
 (pg. 
1835
-
1838
)
Darian-Smith
C
Tan
A
Edwards
S
Comparing thalamocortical and corticothalamic microstructure and spatial reciprocity in the macaque ventral posterolateral nucleus (VPLc) and medial pulvinar
J Comp Neurol
 , 
1999
, vol. 
410
 (pg. 
211
-
234
)
De La Mothe
LA
Blumell
S
Kajikawa
Y
Hackett
TA
Thalamic connections of the auditory cortex in marmoset monkeys: core and medial belt regions
J Comp Neurol
 , 
2006
, vol. 
496
 (pg. 
27
-
71
)
DeVito
JL
A horseradish peroxidase autoradiographic study of parietopulvinar connections in Saimiri sciureus
Exp Brain Res
 , 
1978
, vol. 
32
 (pg. 
581
-
590
)
Dlugos
DJ
Jaggi
J
O'Connor
WM
Ding
XS
Reivich
M
O'Connor
MJ
Sperling
MR
Hippocampal cell density and subcortical metabolism in temporal lobe epilepsy
Epilepsia
 , 
1999
, vol. 
40
 (pg. 
408
-
413
)
Duvernois
H
The human brain: surface, three-dimensional sectional anatomy and MRI
1991
New York
Springer-Verlag
Gordon
B
Lesser
R
Rance
N
Hart
J
Webber
R
Umatsu
S
Fisher
R
Parameters for direct cortical electrical stimulation in the human: histopathologic confirmation
Electroencephalogr Clin Neurophysiol
 , 
1990
, vol. 
75
 (pg. 
371
-
377
)
Grieve
KL
Acuna
C
Cudeiro
J
The primate pulvinar nuclei: vision and action
Trends Neurosci
 , 
2000
, vol. 
23
 (pg. 
35
-
39
)
Guenot
M
Isnard
J
Ryvlin
P
Fischer
C
Ostrowsky
K
Mauguiere
F
Sindou
M
Neurophysiological monitoring for epilepsy surgery: the Talairach SEEG method. StereoElectroEncephaloGraphy. Indications, results, complications and therapeutic applications in a series of 100 consecutive cases
Stereotact Funct Neurosurg
 , 
2001
, vol. 
77
 (pg. 
29
-
32
)
Guillery
RW
Sherman
SM
Thalamic relay functions and their role in corticocortical communication: generalizations from the visual system
Neuron
 , 
2002
, vol. 
33
 (pg. 
163
-
175
)
Gutierrez
C
Cola
MG
Seltzer
B
Cusick
C
Neurochemical and connectional organization of the dorsal pulvinar complex in monkeys
J Comp Neurol
 , 
2000
, vol. 
419
 (pg. 
61
-
86
)
Guye
M
Regis
J
Tamura
M
Wendling
F
Mc Gonigal
A
Chauvel
P
Bartolomei
F
The role of corticothalamic coupling in human temporal lobe epilepsy
Brain
 , 
2006
, vol. 
129
 (pg. 
1917
-
1928
)
Hackett
TA
Stepniewska
I
Kaas
JH
Thalamocortical connections of the parabelt auditory cortex in macaque monkeys
J Comp Neurol
 , 
1998
, vol. 
400
 (pg. 
271
-
286
)
Hashiguchi
K
Morioka
T
Yoshida
F
Kawamura
T
Miyagi
Y
Kuwabara
Y
Sasaki
M
Koga
H
Nagata
S
Sasaki
T
Thalamic hypometabolism on 18FDG-positron emission tomography in medial temporal lobe epilepsy
Neurol Res
 , 
2007
, vol. 
29
 (pg. 
215
-
222
)
Huerta
MF
Krubitzer
LA
Kaas
JH
Frontal eye field as defined by intracortical microstimulation in squirrel monkeys, owl monkeys, and macaque monkeys: I. Subcortical connections
J Comp Neurol
 , 
1986
, vol. 
253
 (pg. 
415
-
439
)
Insausti
R
Amaral
DG
Cowan
WM
The entorhinal cortex of the monkey: III. Subcortical afferents
J Comp Neurol
 , 
1987
, vol. 
264
 (pg. 
396
-
408
)
Jankowska
E
Padel
Y
Tanaka
R
The mode of activation of pyramidal tract cells by intracortical stimuli
J Physiol
 , 
1975
, vol. 
249
 (pg. 
617
-
636
)
Jones
EG
Viewpoint: the core and matrix of thalamic organization
Neuroscience
 , 
1998
, vol. 
85
 (pg. 
331
-
345
)
Jones
EG
The thalamic matrix and thalamocortical synchrony
Trends Neurosci
 , 
2001
, vol. 
24
 (pg. 
595
-
601
)
Jones
EG
Thalamic circuitry and thalamocortical synchrony
Philos Trans R Soc Lond B Biol Sci
 , 
2002
, vol. 
357
 (pg. 
1659
-
1673
)
Jones
EG
Burton
H
A projection from the medial pulvinar to the amygdala in primates
Brain Res
 , 
1976
, vol. 
104
 (pg. 
142
-
147
)
Joo
EY
Hong
SB
Han
HJ
Tae
WS
Kim
JH
Han
SJ
Seo
DW
Lee
KH
Hong
SC
Lee
M
, et al.  . 
Postoperative alteration of cerebral glucose metabolism in mesial temporal lobe epilepsy
Brain
 , 
2005
, vol. 
128
 (pg. 
1802
-
1810
)
Kultas-Ilinsky
K
Sivan-Loukianova
E
Ilinsky
IA
Reevaluation of the primary motor cortex connections with the thalamus in primates
J Comp Neurol
 , 
2003
, vol. 
457
 (pg. 
133
-
158
)
Lacruz
ME
Garcia Seoane
JJ
Valentin
A
Selway
R
Alarcon
G
Frontal and temporal functional connections of the living human brain
Eur J Neurosci
 , 
2007
, vol. 
26
 (pg. 
1357
-
1370
)
Lim
C
Mufson
EJ
Kordower
JH
Blume
HW
Madsen
JR
Saper
CB
Connections of the hippocampal formation in humans: II. The endfolial fiber pathway
J Comp Neurol
 , 
1997
, vol. 
385
 (pg. 
352
-
371
)
Lysakowski
A
Standage
GP
Benevento
LA
An investigation of collateral projections of the dorsal lateral geniculate nucleus and other subcortical structures to cortical areas V1 and V4 in the macaque monkey: a double label retrograde tracer study
Exp Brain Res
 , 
1988
, vol. 
69
 (pg. 
651
-
661
)
Manola
L
Holsheimer
J
Veltink
P
Buitenweg
JR
Anodal vs cathodal stimulation of motor cortex: a modeling study
Clin Neurophysiol
 , 
2007
, vol. 
118
 (pg. 
464
-
474
)
Matsumoto
R
Nair
DR
LaPresto
E
Bingaman
W
Shibasaki
H
Luders
HO
Functional connectivity in human cortical motor system: a cortico-cortical evoked potential study
Brain
 , 
2007
, vol. 
130
 (pg. 
181
-
197
)
Matsumoto
R
Nair
DR
LaPresto
E
Najm
I
Bingaman
W
Shibasaki
H
Luders
HO
Functional connectivity in the human language system: a cortico-cortical evoked potential study
Brain
 , 
2004
, vol. 
127
 (pg. 
2316
-
2330
)
Mauguière
F
Rosenow
F
Lüders
HO
PET Scan in the interictal period: its contribution to evaluation of the functional deficit and epileptogenic zones
Presurgical assessment of the epilepsies with clinical neurophysiology and functional imaging. Handbook of clinical neurophysiology, vol. 3
 , 
2004
Amsterdam
Elsevier
(pg. 
217
-
241
)
Mauguiere
F
Baleydier
C
Topographical organization of medial pulvinar neurons sending fibres to Brodman's areas 7, 21 and 22 in the monkey
Exp Brain Res
 , 
1978
, vol. 
31
 (pg. 
605
-
607
)
Mazzola
L
Isnard
J
Mauguière
F
Somatosensory and pain responses to stimulation of the second somatosensory area (SII) in humans. A comparison with SI and insular responses
Cereb Cortex
 , 
2006
, vol. 
16
 (pg. 
960
-
968
)
Morel
A
Stereotactic atlas of the human thalamus and basal ganglia
 , 
2007
New York
Informa Healthcare
Morel
A
Magnin
M
Jeanmonod
D
Multiarchitectonic and sterotactic atlas of the human thalamus
J Comp Neurol
 , 
1997
, vol. 
387
 (pg. 
588
-
630
)
Mufson
EJ
Brady
DR
Kordower
JH
Tracing neuronal connections in post-mortem human hippocampal complex with the carbocyanine dye DiI
Neurobiol Aging
 , 
1990
, vol. 
11
 (pg. 
649
-
653
)
Mufson
EJ
Mesulam
MM
Thalamic connections of the insula in the rhesus monkey and comments on the paralimbic connectivity of the medial pulvinar nucleus
J Comp Neurol
 , 
1984
, vol. 
227
 (pg. 
109
-
120
)
Münkle
MC
Waldvogel
HJ
Faull
RLM
The distribution of calbindin, calretinin and parvalbumin immunoreactivity in the human thalamus
J Chem Neuroanat
 , 
2000
, vol. 
19
 (pg. 
155
-
173
)
Nathan
S
Sinha
S
Gordon
B
Lesser
R
Thakor
N
Determination of current density distributions generated by electrical stimulation of the human cerebral cortex
Electroencephalogr Clin Neurophysiol
 , 
1993
, vol. 
86
 (pg. 
183
-
192
)
Ono
M
Kubik
S
Abernathey
CD
Atlas of the cerebral sulci
 , 
1990
Stuttgart (Germany)
Georg Thieme Verlag
 
218 p
Ostrowsky
K
Magnin
M
Ryvlin
P
Isnard
J
Guenot
M
Mauguière
F
Representation of pain and somatic sensation in the human insula: a study of responses to direct electrical cortical stimulation
Cereb Cortex
 , 
2002
, vol. 
12
 (pg. 
376
-
385
)
Price
JL
Amaral
DG
An autoradiographic study of the projections of the central nucleus of the monkey amygdala
J Neurosci
 , 
1981
, vol. 
1
 (pg. 
1242
-
1259
)
Ranck
JB
Jr
Which elements are excited in electrical stimulation of mammalian central nervous system: a review
Brain Res
 , 
1975
, vol. 
98
 (pg. 
417
-
440
)
Ranck
JB
Jr
Patterson
MM
Kesner
RP
Extracellular stimulation
Electrical stimulation research technique
 , 
1981
New York
Academic Press
(pg. 
1
-
36
)
Rattay
F
The basic mechanism for the electrical stimulation of the nervous system
Neuroscience
 , 
1999
, vol. 
89
 (pg. 
335
-
346
)
Rezak
M
Benevento
LA
A comparison of the organization of the projections of the dorsal lateral geniculate nucleus, the inferior pulvinar and adjacent lateral pulvinar to primary visual cortex (area 17) in the macaque monkey
Brain Res
 , 
1979
, vol. 
167
 (pg. 
19
-
40
)
Robinson
DL
Cowie
RJ
Steriade
M
Jones
EG
Mc Cormick
DA
The primate pulvinar: structural, functional, and behavioral components of visual salience
Thalamus, vol. II: experimental and clinical aspects
 , 
1997
Amsterdam
Elsevier
(pg. 
53
-
92
)
Romanski
LM
Giguere
M
Bates
JF
Goldman-Rakic
PS
Topographic organization of medial pulvinar connections with the prefrontal cortex in the rhesus monkey
J Comp Neurol
 , 
1997
, vol. 
379
 (pg. 
313
-
332
)
Rorden
C
Brett
M
Stereotaxic display of brain lesions
Behav Neurol
 , 
2000
, vol. 
12
 (pg. 
191
-
200
)
Rosenberg
DS
Mauguière
F
Demarquay
G
Ryvlin
P
Isnard
J
Fischer
C
Guénot
M
Magnin
M
Involvement of medial pulvinar thalamic nucleus in human temporal lobe seizures
Epilepsia
 , 
2006
, vol. 
47
 (pg. 
98
-
107
)
Rossini
PM
Rossi
S
Transcranial magnetic stimulation. Diagnostic, therapeutic, and research potential
Neurology
 , 
2007
, vol. 
68
 (pg. 
484
-
488
)
Shelley
BP
Trimble
MR
The insular lobe of reil—Its anatomo-functional, behavioural and neuropsychiatric attributes in humans. A review
World J Biol Psychiatry
 , 
2004
, vol. 
5
 (pg. 
176
-
200
)
Sherman
SM
Guillery
RW
The role of the thalamus in the flow of information to the cortex
Philos Trans R Soc Lond B Biol Sci
 , 
2002
, vol. 
357
 (pg. 
1695
-
1708
)
Shipp
S
The functional logic of cortico-pulvinar connections
Philos Trans R Soc Lond B Biol Sci
 , 
2003
, vol. 
358
 (pg. 
1605
-
1624
)
Shipp
S
The brain circuitry of attention
Trends Cogn Sci
 , 
2004
, vol. 
8
 (pg. 
223
-
230
)
Steriade
M
Jones
EG
McCormick
Thalamus Vol. I: organisation and function
 , 
1997
Amsterdam
Elsevier
 
959 p
Szabo
K
Poepel
A
Pohlmann-Eden
B
Hirsch
J
Back
T
Sedlaczek
O
Hennerici
M
Gass
A
Diffusion-weighted and perfusion MRI demonstrates parenchymal changes in complex partial status epilepticus
Brain
 , 
2005
, vol. 
128
 (pg. 
1369
-
1376
)
Talairach
J
Bancaud
J
Stereotactic approach to epilepsy. Methodology of anatomo-functional stereotactic investigations
Prog Neurol Surg
 , 
1973
, vol. 
5
 (pg. 
297
-
354
)
Talairach
J
Tournoux
P
Co-planar stereotaxic atlas of the human brain: 3-dimensional proportional system: an approach to cerebral imaging
 , 
1988
Stuttgart (Germany)
Thieme
Tanibuchi
I
Electrophysiological and anatomical studies on thalamic mediodorsal nucleus projections onto the prefrontal cortex in the cat
Brain Res
 , 
1992
, vol. 
580
 (pg. 
137
-
146
)
Tardif
E
Clarke
S
Intrinsic connectivity of human auditory areas. A tracing study with DiI
Eur J Neurosci
 , 
2001
, vol. 
13
 (pg. 
1045
-
1050
)
Tardif
E
Probst
A
Clarke
S
Laminar specificity of intrinsic connections in Broca's area
Cereb Cortex
 , 
2007
, vol. 
17
 (pg. 
2949
-
2960
)
Trojanowski
JQ
Jacobson
S
A combined horseradish peroxidase-autoradiographic investigation of reciprocal connections between superior temporal gyrus and pulvinar in squirrel monkey
Brain Res
 , 
1975
, vol. 
85
 (pg. 
347
-
353
)
Trojanowski
JQ
Jacobson
S
The morphology and laminar distribution of cortico-pulvinar neurons in the rhesus monkey
Exp Brain Res
 , 
1977
, vol. 
28
 (pg. 
51
-
62
)
Valentin
A
Anderson
M
Alarcon
G
Garcia Seoane
JJ
Selway
R
Binnie
CD
Polkey
CE
Responses to single pulse electrical stimulation identify epileptogenesis in the human brain in vivo
Brain
 , 
2002
, vol. 
125
 (pg. 
1709
-
1718
)
Velasco
AL
Velasco
F
Jimenez
F
Velasco
M
Castro
G
Carrillo-Ruiz
J
Fanghanel
G
Boleaga
B
Neuromodulation of the centromedian thalamic nuclei in the treatment of generalized seizures and the improvement of the quality of life in patients with lennox-gastaut syndrome
Epilepsia
 , 
2006
, vol. 
47
 (pg. 
1203
-
1212
)
Weller
RE
Steele
GE
Kaas
JH
Pulvinar and other subcortical connections of dorsolateral visual cortex in monkeys
J Comp Neurol
 , 
2002
, vol. 
450
 (pg. 
215
-
240
)
Wetzel
MC
The identification of self-stimulation elements for adjacent electrodes in the cat
Physiol Behav
 , 
1970
, vol. 
5
 (pg. 
899
-
905
)
Wilson
C
Isokawa
M
Babb
T
Crandall
P
Functional connections in the human temporal lobe. I-Analysis of limbic system pathways using neuronal responses evoked by electrical stimulation
Exp Brain Res
 , 
1990
, vol. 
82
 (pg. 
279
-
292
)
Yeterian
EH
Pandya
DN
Corticothalamic connections of the posterior parietal cortex in the rhesus monkey
J Comp Neurol
 , 
1985
, vol. 
237
 (pg. 
408
-
426
)
Yeterian
EH
Pandya
DN
Corticothalamic connections of paralimbic regions in the rhesus monkey
J Comp Neurol
 , 
1988
, vol. 
269
 (pg. 
130
-
146
)
Yeterian
EH
Pandya
DN
Thalamic connections of the cortex of the superior temporal sulcus in the rhesus monkey
J Comp Neurol
 , 
1989
, vol. 
282
 (pg. 
80
-
97
)
Yeterian
EH
Pandya
DN
Corticothalamic connections of the superior temporal sulcus in rhesus monkey
Exp Brain Res
 , 
1991
, vol. 
83
 (pg. 
268
-
284
)
Yeterian
EH
Pandya
DN
Corticothalamic connections of extrastriate visual areas in rhesus monkeys
J Comp Neurol
 , 
1997
, vol. 
378
 (pg. 
562
-
585
)
Zumsteg
D
Lozano
A
Wennberg
R
Depth electrode recorded cerebral responses with deep brain stimulation of the anterior thalamus for epilepsy
Clin Neurophysiol
 , 
2006
, vol. 
117
 (pg. 
1380
-
1387
)
Zumsteg
D
Lozano
AM
Wieser
HG
Wennberg
RA
Cortical activation with deep brain stimulation of the anterior thalamus for epilepsy
Clin Neurophysiol
 , 
2006
, vol. 
117
 (pg. 
192
-
207
)