Mapping Seizure Behaviors to the Anterior/Mid-Cingulate–Insula Network

This project applies advanced computational signal analysis to investigate rare seizure semiologies, such as ictal smiling, laughter, neck athetotic movements, postictal face wiping, and the “chill-out” arm pose. While these behaviors are clinically recognized, their neurophysiological underpinnings remain poorly understood. Using stereo-electroencephalography (SEEG) data, we developed a MATLAB-based pipeline to preprocess signals, apply event-related desynchronization/synchronization (ERDS) analysis, and quantitatively characterize frequency dynamics from delta through high-frequency oscillations. This computational framework enables high-resolution mapping of semiology onset to the corresponding brain regions, offering an objective approach beyond traditional visual interpretation.

By integrating quantitative ERDS metrics, statistical modeling and network-level analysis, the project links clinical semiology to measurable signal features in specific brain regions. The expected outcomes include improved identification of seizure-related dynamics, novel computational biomarkers for uncommon semiologies, and tools that support presurgical evaluation. More broadly, this work demonstrates how computational neuroscience can transform semiology into a data-driven framework, enhancing epilepsy surgery planning, and paving the groundwork for AI-enabled diagnostics and therapeutics.