Thèse de doctorat
Résumé : Epilepsy surgery is a therapeutical option in case of refractory focal epilepsy that requires to localize the epileptogenic zone. Despite a proper presurgical assessment, all patients do not reach seizure-freedom after epilepsy surgery. The reasons for epilepsy surgery failure are multiple, including inaccurate delineation of the epileptogenic zone or its intrinsic evolutive nature. The current presurgical assessment may optionally include cryogenic magnetoencephalography at the noninvasive phase, that is currently based on cryogenic cooled magnetic field sensors placed in a one-size-fits-all rigid helmet. This setup requires a thermal insulation space between the sensors and the scalp, putting sensors at a 2-4 cm distance from the brain. At the invasive phase, the presurgical assessment may include intracranial electroencephalography that allow to perform intracranial electrical stimulations. Among them, single-pulse electrical stimulations are known to elicit cortico-cortical evoked potentials, spikes and high-frequency oscillations that are more or less related to epileptogenic zone. Additionally, intracranial electroencephalography suffers from a sampling bias as the number of implanted electrodes is intrinsically limited. Its combination with an extracranial recording covering the whole scalp might allow to overcome this major limitation.The aim of this PhD thesis is to assess the added value of advanced neurophysiological techniques that share high temporal and spatial resolutions in order to better characterize the epileptogenic zone localization. For that purpose, we investigated the yield of on-scalp magnetoencephalography based on optically pumped magnetometers and cortico-cortical evoked responses at localizing the epileptogenic zone.We demonstrate that on-scalp magnetoencephalography was able to detect and localize ictal and interictal epileptiform discharges in comparison with gold-standard investigation methods (cryogenic magnetoencephalography or stereo-electroencephalography). On-scalp magnetoencephalography has a high signal-to-noise ratio, even in free-to-move patients despite spontaneous, ictal and hyperventilation-induced movements and during infancy despite the low head circumference. Two types of optically pumped magnetometers, working with rubidium or helium, demonstrated a similar detection rate of interictal epileptiform discharges. The compatibility of on-scalp magnetoencephalography with intracranial electroencephalography allow to combine the high spatial resolution and the whole-scalp covering thanks to simultaneous recordings.We demonstrate that, among single-pulse electrical stimulations responses done during stereo-electroencephalography, cortico-cortical evoked potentials are the best markers of the epileptogenic zone. An increased distance-corrected latency and variation of the cortico-cortical evoked potentials are described within the epileptogenic zone. Both are nonetheless negatively correlated with the seizure frequency at the time of the intracranial recording. The simultaneous intra- and extra-cranial recording combining stereo-electroencephalography and magnetoencephalography confirms the increased distance-corrected latency in cortico-cortical evoked fields. It also demonstrates the increased effective connectivity within the epileptogenic network compared with other networks.Overall, this PhD thesis demonstrates that both on-scalp magnetoencephalography and the investigation of cortico-cortical evoked responses may contribute to a better localization of the epileptogenic zone and they are promising tools for a future clinical use in patients with refractory focal epilepsy.