Prognostic Value of High-resolution Electrical Source Imaging on the Success of Pediatric Focal Epilepsy Surgery

This study investigates the usefulness of high resolution electrical source imaging (HR-ESI) in the setting of presurgical evaluation of drug-resistant focal epilepsy in children. This method is based on an estimation of the intra-cerebral source that produces a signal recorded by scalp electrodes by solving the inverse problem, taking into account attenuation factors resulting from particular conductivity properties of the cerebral, peri-cerebral and cranial tissues. Electrical sources are then fused on structural magnetic resonance imaging (MRI). Scalp EEG recorded using 64 to 256 electrodes refers to as high resolution EEG (HR-EEG), leading to HR-ESI. Studies based on small population of children or on mixt population of children and adults showed that HR-ESI has accuracy values, i.e. percentage of true positives (electrical source localized in the brain area resected and success of surgery) and true negatives (electrical source localized outside the brain area resected and failure of surgery) among the total population, ranging from 50 to 80%. Discrepancies between studies could be explained by the limited number of patients included or by the mixture of pediatric and adult data. Another limitation of previously published studies is that the spatial pattern of dipole source distribution was not taken into account to determine prediction accuracy of ESI. Studies using magnetoencephalography (MEG) to perform magnetic source imaging (MSI) suggest that the spatial pattern of dipole source distribution needs to be considered, a spatially-restricted dipole distribution being associated with better post-surgical outcome when resected. To tackle these issues, the investigators aim to conduct the first large prospective multicentric study in children with focal epilepsy candidates to surgery to assess prediction accuracy of ESI based on the finding of tight clusters of dipoles. This is original as this pattern (tight versus loose cluster of dipoles) has been studied by several researchers using MEG but not using HR-EEG. The investigators make the hypothesis that HR-EEG will allow to identity good candidates for epilepsy surgery and thus to offer this underutilized treatment in more children with better post-surgical outcome. Among the secondary objectives, the investigators will address methodological issues related to the resolution of the inverse problem (methods using distributed sources models versus methods based on equivalent dipole estimation), the potential added value to model high-frequency oscillations (HFO), and the investigators will assess the cost-utility of the HR-ESI procedure.