F10. Co-localization of ictal and interictal activity using high density EEG data

Abstract

Despite major advances in using high density EEG (hdEEG) and electrical source imaging (ESI) for analysis of interictal activity, research demonstrating its utility in examining ictal activity is still lacking. This may be a result of technical challenges associated with capturing epileptic seizures on scalp hdEEG recordings (physical constraints, such as high number of sensors, lack of standardization in review, comfort, etc.). Furthermore, the value of using interictal discharges as a valid estimate of seizure onset zone (SOZ) still remains disputable. The aim of this study is to use electrical source imaging (ESI) to evaluate the co-localization of ictal and interictal activity through the use of high density scalp EEG. A retrospective analysis was performed on 5 patients with intractable focal temporal and extratemporal lobe epilepsy. EEG scalp recordings were obtained using high density 256 Geodesic Sensory Net (Geodesic Sensor Net, Electrical Geodesic Inc.). ESI was performed using two different commercially available ESI systems, Curry 7.0 (Compumedics) and MIPEGI (Electrical Geodesic Inc.). Sensor locations were determined using the Geodesic Photogrammetry System (GPS) and Geosource 3 (Electrical Geodesic Inc), Realistic geometric individualized head models were constructed from patient specific MRIs using the boundary element model and the finite difference model. Source reconstruction was based on distributed source models and dipole models. The concordance of irritative and ictal onset zones was estimated by the distance between maximal ictal and interictal source activity. ESI identified the seizure onset and irritative zones in all 5 patients. There was good concordance between the estimated seizure onset zone and irritative zone on the sub-lobar level. Similar results were obtained through the use of the boundary element and finite difference models. The results provide insight into whether interictal discharges provide accurate estimate of SOZ. Furthermore, these results assess similarity of source reconstruction obtained through two different types of individual head models that are commercially available.