Abstracts

Interictal Epileptiform Discharge-related Distributed Source Activations in MEG Recordings Reveal Combined Local and Distant Propagation Patterns

Abstract number : 3.268
Submission category : 3. Neurophysiology / 3D. MEG
Year : 2024
Submission ID : 265
Source : www.aesnet.org
Presentation date : 12/9/2024 12:00:00 AM
Published date :

Authors :
Presenting Author: Leela Srinivasan, BS – National Institutes of Health

Elena Hayday, BS – National Institutes of Health
Kaya Scheman, BS – National Institutes of Health
Jeff Stout, PhD – National Institutes of Health
Antonio Triggiani, PhD – National Institutes of Health
William Theodore, MD – National Institutes of Health
Kareem Zaghloul, MD, PhD – National Institutes of Health
Sara Inati, MD – National Institutes of Health

Rationale: Interictal epileptiform discharges (IEDs) are sensitive but non-specific markers of the epileptogenic zone (EZ), occurring at a higher frequency than seizures, but often involving large cortical areas. In both animal models and intracranial EEG recordings, IED activity has been proposed to spread from focal sources, both contiguously over the cortical surface and distantly via white matter pathways. Here, we sought to investigate whether similar propagation patterns could be observed in IEDs recorded during magnetoencephalography (MEG) using distributed source modeling.

Methods: We studied 40 patients with epilepsy (22 male; 25±12 years) who underwent pre-surgical evaluation at the NIH Clinical Center with resting state MEG recordings. IEDs with similar waveforms were identified using an in-house algorithm and clinically validated. We performed cortical segmentation from T1-weighted 3T MR images using the standard Freesurfer pipeline and co-registered the MRI and MEG recordings. Using MNE software, for each patient we created boundary element models, filtered and smoothed the data, and applied dynamic statistical parametric mapping (dSPM) to reconstruct virtual source activity for an averaged IED epoch. The area under the curve (AUC) was summed in each dSPM activation time course around the spike peak to assess levels of activation at each virtual sensor. We used surface-based spatial clustering of sensor AUCs above the 95th percentile to identify regions with maximal IED-related activity.

Results: We found that the activation cluster with the largest area, or the primary cluster, overlapped with the resection area in 91.67% of patients who became seizure free following epilepsy surgery (N=12). In 90% of primary clusters (N=40), AUC had a significant correlation (p < 0.05) with geodesic distance from the maximal AUC, demonstrating decreasing strength of activation with distance along the cortical surface from the presumed source of activity. In 75% of patients, we identified two or more spatially distinct regions of temporally overlapping activation, consistent with discontinuous and presumably white matter propagation.
Neurophysiology