AI-based Scalp-Subscalp EEG monitoring for epilepsy

This two-year project will further develop and validate a wearable, AI-driven seizure prediction system that combines multimodal EEG signals (scalp and optionally subscalp) to predict seizures with a horizon of 10–15 minutes. This window enables patients and caregivers to administer medication or stimulation before seizures occur.

Epilepsy affects over 50 million people worldwide, with up to 30% of patients experiencing drug-resistant seizures. For these individuals, the unpredictability of seizures poses significant risks to independence, employment, and safety, severely reducing their quality of life. Despite the availability of neurostimulation and rescue medication, timely intervention remains out of reach due to the absence of accurate seizure prediction systems for home use. 

The technology has already been demonstrated to work in the proof-of-concept study in the epilepsy monitoring unit (EMU) for short durations of few days. Here, we aim to extend such results to home environment and for ultra-longterm measurements. 

The project is led by Alpha Brain Technologies (ABT), a Dutch Deep tech SME, in collaboration with the University of Twente (UT). ABT brings the latest seizure prediction algorithms and home-use EEG hardware development expertise. UT will lead scientific validation of the algorithms and signal quality.  

AI-SEEM will progress the innovation from TRL 4 to TRL 6 by integrating hardware and AI models into a user-friendly prototype and validating it through technical, usability, and preliminary clinical pilots.