Abstracts

Rationale: Inter-ictal epileptic discharges (IEDs) are spontaneous transient events that occur between seizures with no clinical manifestations, patients being unaware of having them. They thus constitute the ideal events to study with several imaging techniques such as ElectroEncephalography (EEG), MagnetoEncephalography (MEG) or functional Magnetic Resonance Imaging (fMRI). Electric source imaging (ESI) and magnetic source imaging (MSI) of EEG and MEG signals recorded non-invasively are widely used to determine the origin of inter-ictal spikes. During the pre-surgical evaluation of patients with epilepsy, ESI and MSI can be useful for guiding the implantation of intracranial EEG (iEEG) electrodes. Findings from SEEG (StereoElectroEncephalography - EEG recorded invasively with intracranial electrodes) or anatomical lesion seen on MRI are considered the ground truth when making decision for surgery in patients with focal epilepsy. In this context, it is then crucial to provide accurate and reliable source localization results from EEG and MEG data. We recently demonstrated that a new source localization method we proposed, the Maximum Entropy on the Mean (MEM), is ideal for localizing spike generators with their spatial extent along the cortical surface (Heers et al. 2015, Brain Topogr.). Since EEG and MEG are sensitive to different aspects of the neuronal activity, combining them at the level of source localization can be even more beneficial. In our recent study, we explored and validated fusion of EEG and MEG data (MEEG) within the MEM Framework (Chowdhury et al. 2015, Brain Topogr.) suggesting very promising results. We hypothesize that combined EEG and MEG recordings will take full benefit of the complementarity between these two modalities to well-characterize the IEDs. Therefore, our present objective is to assess the clinical relevance of EEG-MEG fusion data source imaging during pre-surgical evaluation of epilepsy patients.Methods: We retrospectively selected patients with intractable focal epilepsy who underwent simultaneous EEG-MEG recordings, had subsequent SEEG, and presented visible lesion (such as FCD - Focal cortical dysplasia and PMG - Poly Microgyria) on anatomical MRI. We performed source imaging on EEG, MEG and MEEG data and qualitatively compared the results with SEEG findings and lesion localization.Results: From our preliminary results on 6 patients we noticed that fusion provided sources that were in perfect concordance with the SEEG findings for all the patients (Table 1). While ESI and MSI are also able to locate most of the sources accurately, fusion is able remove any spurious additional sources that can be misleading when considering EEG or MEG only (Figure 1). It is noteworthy that fusion source imaging was able to recover sources closest to the lesion when compared to EEG and MEG.Conclusions: Source localization results from fusion EEG-MEG data can be more useful and reliable than using EEG and MEG data separately to guide SEEG implantations or during pre-surgical evaluation.