Directional connectivity patterns in SEEG recorded epileptiform transitions in mesial temporal lobe epilepsy

Epilepsy is characterized by dynamic disruptions in brain networks, yet the spatiotemporal organization of directional interactions underlying ictal transitions remains incompletely understood. This study investigates peri-ictal transition patterns in mesial temporal lobe epilepsy (MTLE) by characterizing directional connectivity dynamics from intracranial stereo-EEG (SEEG) seizure recordings.
Methods: We analyzed SEEG data from 10 patients with MTLE who subsequently achieved seizure freedom following surgical resection. We quantified directional connectivity using directed transfer function (DTF) measures across pre-ictal, ictal, and post-ictal phases. We then applied clustering analyses to identify reproducible dynamical connectivity patterns and evaluated their consistency across seizures and patients.
Main Results: We identified distinct out-degree connectivity patterns across seizure phases, each associated with specific anatomical regions. Regions corresponding to the epileptogenic zone exhibited elevated out-degree during pre-ictal and early ictal periods, whereas non-resected regions showed increased out-degree during ictal termination and post-ictal phases. These patterns remained consistent across patients and seizures, demonstrating a structured and phase-dependent organization of directional information flow during seizure evolution.
Significance: This study identifies stereotypical and reproducible peri-ictal directional connectivity patterns in mesial temporal lobe epilepsy. By characterizing how information flow evolves across seizure phases, our findings advance network-level understanding of ictal transitions and provide a reproducible framework for future studies of seizure network dynamics.