GCU Researchers Develop AI Headset to Predict Epileptic Seizures 30 Minutes Early

Researchers at Glasgow Caledonian University (GCU) have developed a wearable AI powered headset capable of predicting epileptic seizures up to 30 minutes before they occur. With a 95 percent accuracy rate, this breakthrough represents a major shift from traditional devices that only detect seizures once they have already begun. Led by Professor Hadi Larijani, the project aims to provide an essential early warning system for patients.

Predicting Seizures Before Onset

The system functions by continuously monitoring electroencephalography and electrocardiography data through a wireless, non-invasive headset. The underlying AI software was trained using extensive historical data from the William Quarrier Scottish Epilepsy Centre to recognize the subtle biological precursors of a seizure. This real-time analysis allows the device to identify patterns invisible to the human eye or standard reviews. The 30 minute window provided by the device allows individuals to move to a safe environment, sit down, or alert caregivers before losing consciousness. Beyond physical safety, the innovation addresses the anxiety and fear of the unknown many epilepsy patients face daily. By offering a reliable alert, the headset empowers users with greater independence and a significantly improved quality of life.

Transforming the Patient Experience

Originally starting as a feasibility study supported by The Data Lab, the project has since moved to hardware development via Scottish Enterprise. The headset was recognized by the Converge Challenge, a competition for academic entrepreneurs. Researchers are using surveys to refine the design based on feedback from clinicians and patients according to GCU Impact Reports.

Expanding Non-Invasive Clinical Care

This breakthrough is part of a wider effort at the university to integrate smart technology into clinical settings. Alongside the headset, GCU researchers have partnered with Safehinge Primera to develop radar-based sensors for mental health wards. The system tracks patient movement without cameras, using anonymous point clouds to ensure safety without violating privacy. Professor Craig Thomson, Associate Dean at GCU, noted that these projects demonstrate how research can meet real-world challenges to redefine the future of care. The goal is to scale these technologies for use within the NHS and global healthcare markets. By combining computer science with clinical expertise, GCU is proving that digital innovation is essential for patient safety.