Reconstruction of epilepsy-characteristic sources by means of a simultaneous evaluation of EEG- and MEG- data using calibrated realistic head models

点击翻译按钮获取中文摘要

The reconstruction of epilepsy-characteristic electric potentials and magnetic fields contributes already significantly to the identification of the epileptogenic zone for the therapy of pharmaco-resistant epilepsies. It is the main goal of this grant application to develop and optimize a new approach for the combined analysis of temporally and spatially highly resolving EEG and MEG as a basis for source analysis and to validate it in a group study with focal cortical dysplasia (FCD) patients who are scheduled for epilepsy surgery. Thereby, this new non-invasive method should be established as one of the standard procedures in presurgical epilepsy diagnosis. The new approach is based on hierarchical Bayesian modeling (HBM) for networks that also involve extended sources for the EEG and MEG inverse problem and on discontinuous Galerkin finite element method (DG-FEM) in individually-calibrated realistic head volume conductor models for the forward problem. For validation and evaluation we propose a multi-level strategy: First, the new methods will be validated in computer simulations. In a next step, reconstruction results of the new approach of evoked responses will be correlated with a registered anatomical atlas. Third, the combined EEG/MEG source analysis results of FCD-triggered epileptic activity will serve as a basis for the determination of regions of interest that are then further examined with novel zoomed magnetic resonance tomography (zoomed MRI) and correlated with the source analysis results. Finally, the results will be compared with other findings of presurgical evaluation and correlated to the resected brain volume and to the post-operative success with regard to seizure control in order to achieve an external validation of the reconstruction results.