Demonstrate a patient-specific pipeline that converts ictal stereo-EEG (sEEG) into volumetric seizure-onset zone (SOZ) source models, enables 4D review within virtual reality (VR), and quantifies SOZ overlap with the post-surgical resection.

Drug-resistant focal epilepsy often requires invasive interventions. Although multimodal data guides surgery, current protocols rely on manual interpretation of 2D multiplanar views of 3D imaging alongside sEEG electrode recordings. SOZ contact resection, which has shown mixed prognostic value, could be enhanced with anatomy-aligned volumetric SOZ identification.

We implemented an end-to-end workflow: (1) select ictal segments; (2) build subject MRI/BEM forward models (FreeSurfer) and compute inverse solutions (MNE/sLORETA); (3) derive a 3D SOZ map (max |source| within a 1-s onset window), average across all ictal segments for patient, and export a single 3D NIfTI for ITK-SNAP/VR; (4) register pre-op T1 to post-op T2 FLAIR and resample the SOZ map using FLIRT; (5) segment resection and compute overlap metrics (Dice, %SOZ within resection, %resection covered by SOZ). Planned analysis will relate degree of overlap to patient post-surgical outcomes graded via the Engel scales.

Feasibility successfully demonstrated in the first patient: the pipeline produced a clean supratentorial SOZ NIfTI, aligned it to post-op imaging, and enabled overlay with the segmented resection. Onset activity was localized near the temporal resection margins. Cohort-level quantitative results are in progress; we hypothesize greater SOZ–resection overlap in better outcomes versus worse clinical outcomes.

This reproducible pipeline transforms ictal sEEG into anatomical SOZ maps suitable for visualization in sEEG4D, a 4D VR representation of spatio-temporal sEEG data, allowing for direct comparison of predicted SOZ with the resection cavity. The approach is feasible, and due to the ability to communicate spatio-temporal information through a single modality, this pipeline has the potential to enhance surgical planning and post-operative evaluation; ongoing enrollment will test whether overlap metrics predict outcome.