Abstracts

Rationale: When epileptic activity is propagated, source localization analysis may be misleading. Studies have demonstrated that the propagation of seizures may share similar pathways to the course of interictal spikes. Understanding the patterns of propagation is essential and closely related to a favorable surgical outcome. In this study, we present a series of patients whose interictal spikes showed clear propagation patterns as revealed by Magnetoencephalography (MEG). Our aim was to take advantage of MEG s strengths to improve the interpretation of intracranial electroencephalography (ICEEG) recordings. Methods: We selected patients with medically intractable epilepsy who underwent MEG and ICEEG studies between Oct. 2009 and May 2010. We selected those whose interictal spikes showed clear propagation patterns. MEG was recorded by a 306-channel system (Neuromag, Helsinki, Finland). Source localization analyses were performed with Curry 6.0 (Neuroscan, Hamburg, Germany), using standard single dipole analysis, and a minimum norm method with current density reconstruction constrained by the cortical surface generated with the individual s MRI. MEG source localization, involved channels on ICEEG recordings, ictal onset zone, seizure semiology and surgical outcome were compared. Results: Patient 1: Interictal spikes started from the left parieto-temporal region and quickly propagated to the primary and supplementary motor cortex, as revealed by analysis of the MEG data. This is concordant with the seizure semiology. ICEEG recordings with subdural grids showed a misleading ictal onset zone at the propagated site. MEG-directed review of the patient s high-resolution MRI revealed a subtle cortical malformation in the posterior insular/parietal opercular region, the resection of which rendered the patient seizure free. Patient 2: MEG interictal spikes showed various propagation patterns within the insula, from the insula to the frontal operculum, and from the orbito-frontal area to the temporal pole. These findings could explain the semiologic features of her epilepsy. In this patient, evaluation of simultaneous stereo-EEG and MEG recordings confirmed the propagation pattern within the insula. Figure 1 shows the maximally involved stereo-EEG contacts coinciding with the trajectory of the moving dipole, as the latter sweeps across the insula area. Patient 3: MEG interictal activities exhibited bidirectional propagation within the left perirolandic region. ICEEG recordings with subdural grids, however, showed a much more diffuse interictal/ictal pattern over the left centro-parietal region. No resective surgery was performed due to the lack of a clear focal onset on ICEEG and the proximity of ictal activity to the sensorimotor areas. Conclusions: Using single dipole and minimum norm source localization methods, we are able to examine the propagation of MEG interictal activities in a spatio-temporal manner. This may help in the accurate localization of the true ictal onset zone especially in cases where interpretation of invasive recordings is difficult due to inadequate sampling.