SIG | Data Science in Epilepsy: Black Boxes or Brain Models? AI Versus Mechanistic Data Science in Epilepsy

Monday, December 7, 2026 at 2:00 PM to 3:30 PM · 1 hr. 30 min. (America/Denver)
Special Interest Group

Information

In light of the rise of AI, are the days of building models of how the brain works and how seizures arise behind us? Should we still care about statistical modeling and time-series analysis of EEG if sufficiently complex AI models can make classifications without feature engineering? In this debate-style session, learners differentiate when AI/ML approaches may be most appropriate for epilepsy diagnosis and treatment, compared with supervised neurophysiological modeling approaches that can reveal mechanisms and inform the scientific study of epilepsy to develop better therapies.
Learning Objective 1
Define supervised and unsupervised learning in the context of EEG analysis for diagnosis and treatment of epilepsy
Learning Objective 2
Explore state-of-the-art statistical modeling methods and AI architectures for analysis of EEG when diagnosing and treating epilepsy
Learning Objective 3
Discriminate clinical problems that could benefit from specialized unsupervised learning tools in EEG analysis
Target Audience
CliniciansFellows/TraineesNursesScientists/Researchers
Career Stage
Early Career (typically 0-5 years from completion of training)Mid-Career (typically 6-15 years from completion of trainingSenior (typically >15 years from completion of training)
Demographics
ClinicalYoung ProfessionalsResearchFirst-time Attendee
Session ID
DATASCIENCESIG26
Presentation Id
3452229
Event ID
24056
Session Type
Session