Authors :
Presenting Author: Thandar Aung, MD, MS – University of Pittsburgh
Umit Aydin, PhD – University of Reading
Andre Pereira, BS, R. EEG T. – University of Pittsburgh Medical Center
Ajay Narayan, MD, MBA – University of Pittsbrugh
Jorge Gonzalez-martinez,, MD, PhD – University of Pittssburgh Medical Center
Rationale:
Stereoencephalography (SEEG) is the gold standard for identifying potential epileptogenic zone (EZ), but its limited spatial coverage—sampling less than 5% of the brain—can lead to inaccuracies, such as due to missing electrodes. Consequently, the seizure onset zone detected by SEEG may not accurately represent the true EZ. Magnetoencephalography (MEG), while offering broader brain coverage, has lower spatial resolution. This study aims to investigate whether simultaneous SEEG-MEG recordings can enhance EZ localization by filling coverage gaps between recorded SEEG electrodes in patients with drug resistance focal epilepsy (DRE).
Methods:
This prospective observational study was conducted with approval from the University of Pittsburgh Institutional Review Board. Informed consent was obtained from 13 consecutive patients who underwent simultaneous SEEG and MEG recording. 20 minutes of simultaneous MEG and SEEG data (sampling rate 1000 Hz) were recorded per patient using the TRIUX™ neo system with 306 channels. The MEG data were post processed with spatial-temporal signal space separation (tSSS) using Maxfilter 2.0 software and cortical surface was extracted from the patients' MRIs using FreeSurfer.
Spikes were averaged separately for each primary irritative zone, based on manual markings from SEEG. Three source localisation approaches —dipole scan, dynamical Statistical Parametric Mapping (dSPM) and coherent Maximum Entropy on the Mean (cMEM) — were used to localise epileptic spikes using Brainstorm software. Specifically, source localization was done using MEG gradiometers only, MEG magnetometers only, all MEG sensors, as well as SEEG only, and their sublobar concordance was assessed based on Desikan-Killiany atlas.
Results:
Of 13 patients (62% male), 7 had MRI-negative DRE, 1 had polymicrogyria, 2 had focal cortical dysplasia (FCD), and 3 had mesial temporal sclerosis. Twelve had focal DRE (7 neocortical, 5 temporal), while 1 had multifocal DRE. Among the 12, 10 had epilepsy surgery (6 neocortical, 4 temporal) with mean follow up 4.4 months (range 1-10). Of 10, 8 achieved complete seizure freedom (SF), including 6 with neocortical epilepsy.
For the 7 neocortical epilepsy patients (2 rolandic, 2 frontal, 1 occipital, 2 posterior temporal), MEG source localization was concordant with SEEG in 5 cases (3 with lesions, 2 MRI-negative) but showed a broader spread. In the remaining 2 MRI-negative cases, MEG differed from SEEG. The final resection plan, guided by intraoperative electrocorticogram and MEG, confirmed ILAE FCD type IIB pathology (Figure 1 shows a patient with a primary irritative zone in the parietal operculum identified by SEEG (L and S contacts). However, MEG source localization based on SEEG contacts was in the superior temporal sulcus. The final resection, guided by ECOG involved only MEG source). Both patients achieved SF, with follow-ups at 1 and 10 months.
Conclusions:
Simultaneous SEEG-MEG recordings can enhance the localization of epileptogenic zones by addressing coverage gaps inherent in SEEG alone. This approach led to successful surgical outcomes including those with MRI-negative DRE, especially in the neocortical epilepsy.
Funding: No funding