AI-Driven SEEG Channel Ranking for Epileptogenic Zone Localization
This work addresses a domain-specific problem for clinicians in epilepsy surgery by providing an incremental improvement in automating channel ranking to reduce time and effort in pre-surgical evaluation.
The paper tackles the problem of inefficient visual inspection of hundreds of SEEG channels for epileptogenic zone localization by proposing a machine learning approach that ranks impactful channels using XGBoost and SHAP scoring, achieving promising results in accuracy, consistency, and explainability on data from five patients.
Stereo-electroencephalography (SEEG) is an invasive technique to implant depth electrodes and collect data for pre-surgery evaluation. Visual inspection of signals recorded from hundreds of channels is time consuming and inefficient. We propose a machine learning approach to rank the impactful channels by incorporating clinician's selection and computational finding. A classification model using XGBoost is trained to learn the discriminative features of each channel during ictal periods. Then, the SHapley Additive exPlanations (SHAP) scoring is utilized to rank SEEG channels based on their contribution to seizures. A channel extension strategy is also incorporated to expand the search space and identify suspicious epileptogenic zones beyond those selected by clinicians. For validation, SEEG data for five patients were analyzed showing promising results in terms of accuracy, consistency, and explainability.