Particle.news
Download on the App Store

Glasgow Caledonian University Builds AI Headset That Predicts Epileptic Seizures Minutes Ahead

UKRI funding now supports refinement toward a discreet, child‑suitable cap ahead of lengthy medical approval.

Overview

  • The prototype monitors EEG brainwaves and ECG heart signals to flag seizure risk, with researchers reporting up to 95% accuracy and confidence readouts.
  • The algorithm was trained on thousands of hours of historical recordings and validated using a custom phantom head to simulate real‑world use.
  • Early warnings are intended to give users crucial minutes to take precautions or alert others, aiming to reduce injuries and anxiety.
  • The team is moving from third‑party hardware to a lightweight, wireless design designed to be discreet and appropriate for children.
  • Regulatory accreditation is expected to take years, and there is currently no widely available device that reliably provides advanced seizure alerts in routine use.