Abstracts

Rationale: In the diagnostic work-up of epilepsy surgery, noninvasive functional imaging techniques such as long-term video electroencephalography (vEEG), magneto-encephalography (MEG), positron emission tomography (PET), functional magnetic resonance imaging (fMRI) can be used to delineate the epileptogenic area. Actually intracranial EEG recording (iEEG) remains the gold standard. A comprehensive study of interictal activities is essential in presurgical investigations. The aim of the study was to assess the yield and the precise location of a source imaging method on interictal epileptic spikes. Methods: The study focused on three patients with drug resistant partial epilepsy, treated in the Neurology department at the Pitié-Salpêtrière Hospital, France. We used a distributed source model (Brainstorm), applied to the magneto-encephalography simultaneously controlled by intracranial EEG recording. iEEG was performed using intracerebral electrodes. MEG was simultaneously performed with a 151-channel whole head MEG system in a magnetically shielded room. MEG and EEG were separately analyzed offline. A realistic head model was built, based on a boundary element method, by segmentation of scalp, skull and brain surfaces. The MEG fiducial points, contacts of intracranial electrodes and patient head contour were superimposed on the three-dimensional MRI to exactly match the two surfaces. Source localization was performed with Brainstorm, an inverse method with cortically fixed distributed source developed by S. Baillet et al. (Proc IEEE Int Symp Biomed Imaging, 2004, 1, 652-655). Results: Source localization for averaged spikes allowed identification of epileptic areas recognized by iEEG, including mediotemporal sources. The proportion of detected MEG spikes increased when lateral structures or larger brain volumes, both medial and lateral, were involved. The use of MEG raws averaged according to iEEG spikes groups (determined by their location and extent), allowed us to study spatial and temporal dynamics of MEG sources during IIS. On occipital and frontal-mesial contacts, IIS matched with activation of MEG cortical sources. But, when intracranial epileptic spikes concerned hippocampic and amygdaloid areas, they were not restituted by MEG sources. MEG sources highlighted the involved focus in medio-lateral and lateral epileptic spikes. Comparison between MEG sources dynamics and iEEG ictal onset location and spreading showed close proximity except for medio-temporal lobe epilepsy. Conclusions: The final place of electromagnetic source imaging in the presurgical work-up is still to be determined. Especially in extra-temporal epilepsy, iEEG does not get the expected surgical results. We showed that study of dynamics through IIS source localization provided consistent information with simultaneous iEEG data and ictal onset localization. It can contribute to a better placement of intracerebral electrodes.