Epilepsy is a neurological disorder characterized by seizures and, between them, brief electrical disturbances known as interictal epileptiform discharges (IEDs). These disturbances, which occur frequently during deep sleep, can disrupt memory and cognition. By training a machine learning model on expert-labeled intracranial EEG recordings from deep brain regions, we developed an automated tool that detects IEDs with accuracy comparable to human experts. This technology enhances epilepsy diagnosis and helps researchers understand how abnormal brain activity during sleep affects cognition. Additionally, by publicly sharing our dataset and AI model, we aim to advance both sleep research and epilepsy care, paving the way for AI-assisted neurology.
Annotated interictal discharges in intracranial EEG sleep data and related machine learning detection scheme
