Abstracts

Rationale: We sought to document the added diagnostic benefits of combined MEG and EEG source modeling, as compared to long-term EEG monitoring (LTM) findings, in patients undergoing pre-surgical evaluations.  Methods: 306 channels of MEG and 25 channels of EEG (including bilateral sub-temporal electrodes) were recorded for ~ 1 hour from 100 consecutive patients from June 2017 to July 2018. Separate dipole source models were calculated from averaged MEG/EEG (M/E) spikes that were grouped by different field topography. The number and locations of different M/E spike sources were tallied for each patient, and these were compared to the number and location of different spike types described in the patient's medical records, especially LTM reports. In no patient had source modeling been performed on LTM data.  Results: More spike foci were identified by MEG/EEG dipole modeling than by LTM interpretation - more than 2 foci (M/E 53 patients, LTM 9 patients), 2 foci (M/E 17 patients, LTM 17 patients), 1 focus (M/E 19 patients, LTM 61 patients). Localization descriptors were always sublobar for M/E spike sources, whereas those for LTM spike sources were regional in 36%, lobar in 54%, and sublobar in only 11%. With M/E dipole modeling, more spike foci were found in the same lobe in 49 patients and new spike foci were found in additional lobes in 35 patients.  Conclusions: Combined MEG/EEG recording and source modeling adds significantly to the functional localization of epileptic foci in pre-surgical evaluations.  Funding: No funding