Social cognition, which refers to the ability to interpret social information and behave accordingly in a social environment, is crucial in everyday life. But this ability has been shown to be altered in patients with epilepsy, especially in medial temporal lobe epilepsy, which leads to a deterioration in the patient's quality of life. However, the mechanisms of those deficiencies remain largely unknown. The Team objective is to achieve a structural and functional cartography of the social cognition network in 20 healthy subjects and 20 patients with drug-resistant medial temporal lobe epilepsy (before and one year after resective surgery of the epileptogenic focus). Social cognition deficiencies will be assessed using a specifically dedicated neuropsychological assessment validated in French (Batteries de Cognition Sociale BCS). Brain structural analyses will be performed on a 3-Tesla MRI (3T MRI), including an anatomical T1 sequence, a High Angular Resolution Diffusion Imaging (HARDI) to assess the morphology and macrostructural characteristics of long and short white matter tracts involved in social cognition, and quantitative MRI and Hybrid Diffusion Imaging (HYDI) to assess their microstructure. Functional connectivity will be assessed using an ultra-high-field 7-Tesla MRI (7T MRI), with acquisition in resting state and during specific social cognition tasks. Joint analysis of structural and functional connectivity will enable the team to assess the alterations of social cognition networks in patients with epilepsy and their reorganisations after epilepsy surgery.

Social cognition refers to the ability to understand others behaviours and emotions, and to adapt its own behaviour given the social context. It is a high-order cognitive function, which requires several processes: identifying some information as carrying a social meaning (such as emotions or facial expressions), decoding this information, inferring the emotional state of one's interlocutor from that information, and using that information and those inferences to adapt its own behaviour.

Social cognition deficiencies are found in many neurological (lobar frontotemporal dementia, epilepsy…) or psychiatric (schizophrenia, autistic spectrum disorders…) diseases and lead to substantial deterioration in the quality of life of the patients, and hamper their integration into society. However, brain alterations and pathophysiological mechanisms leading to those deficiencies are still poorly known.

Brain areas subserving the social cognition have been studied over the past few years, but a global view of this network is lacking. Indeed, most studies focused on a particular aspect of social cognition, such as, for example, facial emotion recognition. The main cortical areas and white matter tracts involved in the successive stages of the social cognition are resumed thereafter.

1. Neurological basis of social stimuli perception Social stimuli in everyday life mainly are in visual modality, and secondly in auditory modality. The majority of existing studies thus focused on visual social stimuli.

   The very first step of a visual social stimuli processing is its identification, and various cortical areas are devoted to the recognition of a stimuli specifically emitted by another human being. Some areas demonstrated a high specificity for human face recognition : the Occipital Face Area (OFA) in the inferior occipital gyrus, and the Fusiform Face Area (FFA) in the fusiform gyrus. Similarly, some areas are devoted to the recognition of human bodies : the Extrastriate Body area in the lateral occipito-temporal cortex, and the Fusiform Body Area in the fusiform gyrus.

   Another cortical area early involved in human interaction is the posterior part of the superior temporal sulcus (pSTS). The posterior part of the lateral and superior temporal lobe is an area specialised in movement perception, and some of its parts only react to the movement of inanimate objects (Areas MT/V5), while the pSTS specifically reacts to the perception of a movement made by a biological living being, especially another human. Electrophysiological studies confirmed that neurons in this area specifically react to the observation of movements made by other human subjects, while they are not activated during self-made movement, thus not acting as mirror neurons.

   Once a stimulus is identified as being emitted by another human, and therefore carrying a social meaning, other areas will perform the processing of social-specific information. The amygdala is a key component of this next step, and has been associated with facial emotion recognition and trustworthiness judgement . Lesions of the amygdala were associated with emotional recognition deficiency leading to social interaction impairment.

   The main white matter tracts involved in those abilities are the inferior longitudinal fasciculus and the inferior fronto-occipital fasciculus, connecting occipital cortex with temporal and frontal cortices. Those tracts' development correlated with facial recognition in healthy subjects and their alterations were found in subjects with developmental prosopagnosia.

2. Neurological basis of social interaction comprehension From the basis of those stimuli, a subject must then interpret the behaviour of another person, and infer its emotions, beliefs, and objectives. The theory of mind refers to the ability to infer the mental state of another subject, and has been recently divided in "cold" (or "cognitive") theory of mind, which designates the ability to infer another subject's thought, and "hot" (or "empathic") theory of mind, which designates the ability to infer its emotions. This last ability is close, but distinct, from empathy, which is the ability to share and experience the affective state of another subject.

   The theory of mind has been linked to a brain network called the mentalizing network, involving the temporoparietal junction, the posterior cingulum cortex, the medial precuneus and the temporal poles. On the other hand, neural bases of the empathy are based on a specific mirror network for emotions: many functional MRI studies demonstrated similar activation of cortical areas when experiencing an emotion and observing someone else experiencing it. This has been found for various emotions, such as: pain, including social pain such as the suffering experienced in a situation of social exclusion (anterior insula and anterior cingulum cortex), disgust (anterior insula), or regret (anterior cingulum cortex and orbitofrontal cortex).

   Connections between those areas mainly depend on the superior longitudinal fasciculus. Morphological measures of this tract were associated with performances in emotion recognition and empathy .Other tracts involved in empathy and theory of mind are the uncinate fasciculus ,the cingulum and the arcuate fasciculus.

3. Social cognition deficiency in patients with epilepsy Social cognition impairment associated with epilepsy has been described since the early 20th century. Indeed, psychiatric literature of this period used the term "epileptic personality" to designate a combination of cognitive symptoms in patients with epilepsy, during interictal period, including social disorders (such as egocentricity and loss of empathy). While this stereotypical depiction was later proven wrong, recent studies performed in the last decades, thanks to advances in neuropsychology and neuroimaging, have demonstrated the presence of social cognition impairment in patients with epilepsy.

   Most of those studies were done in patients with medial temporal lobe epilepsy (MTLE). Various studies demonstrated theory of mind impairment, both in its basic and more complex components, in patients with MTLE . Those patients also presented with facial emotion recognition deficiency .

   Social cognition deficiency was also found in other epilepsy syndromes. Patients with focal frontal epilepsy thus presented with theory of mind impairment, similar to patients with MTLE , as well as facial emotion recognition impairment, decreased empathy , and alteration of decision making in social context . Patients with idiopathic generalized epilepsy also demonstrated social cognition impairment, such as theory of mind impairment in patients with juvenile myoclonic epilepsy.

   However, the mechanisms of those alterations remain poorly known. As social cognition deficiencies are found in various type of epilepsy, it is supposed that these deficiencies are not due to the causal brain lesion (especially in idiopathic generalized epilepsy, where there is precisely no lesion), but may be due to an alteration of the social cognition network, secondary to the seizures. The age of onset might also be an important factor: MTLE and idiopathic generalized epilepsy classically start during adolescence, while the social cognition network, one of the last brain networks to mature, is not yet fully defined. The seizures might therefore disrupt its maturation, while having no effect on other brain networks that already achieved their final adult configuration.

4. Objectives In order to investigate the mechanisms of social cognition impairment in patients with epilepsy, we will achieve a complete structural and functional social cognition atlas in 20 patients with drug-resistant medial temporal lobe epilepsy. Those patients will be included amongst patients undergoing preoperative assessment at Pitie-Salpetriere Hospital.

As part of their preoperative assessment, patients will have a complete neuropsychological assessment, including a specifically dedicated neuropsychological assessment validated in French (Batteries de Cognition Sociale BCS).

A first neuroimaging protocol will be performed, including a structural and a functional acquisition. The structural acquisition, performed on a 3T MRI, will include an anatomical T1 sequence, a High Angular Resolution Diffusion Imaging HARDI to assess the morphology and the macrostructural characteristics of long and short white matter tracts involved in social cognition, and quantitative MRI and Hybrid Diffusion Imaging HYDI to assess their microstructure. The functional acquisition, performed on an ultra-high-field 7T MRI, will include a resting-state acquisition and activation acquisition during specific social cognition tasks.

Joint analysis of structural and functional connectivity will enable the investigator team to assess the alterations of social cognition networks in patients with epilepsy.

A second post-operative assessment will be performed at one-year post-surgery. Patients will then have a clinical evaluation, including a new neuropsychological assessment, and the same neuroimaging protocol, in order to assess the reorganization of social cognition network at one year post-surgery, and one year after epilepsy remission.

Inclusion criteria:

• Age ≥ 18 years old
• Good understanding of written and verbal French language
• Written consent signed by the subject
• Patients with focal temporal lobe epilepsy for at least one year, drug-resistant, due to a hippocampal sclerosis
• Undergoing presurgical assessment
• Affiliated to a social security system

Exclusion criteria:

• Under juridical protection
• With claustrophobia or any MRI contra-indication
• Pregnancy or lactation
• Participation in another research with an exclusion period
• Refusal to be notified in the event of anomalies discovered during the performance of an MRI sequence