AI system cuts false epilepsy alarms by 80 percent
Summary and headline written by AI from the source article. How we work
Researchers at the University of Pennsylvania in the United States developed an artificial intelligence system to help people with epilepsy understand their brain activity.
The system connects to electroencephalography (EEG) devices, which measure electrical signals in the brain, and allows patients to ask questions through a smartphone application. The AI analyzes the data and responds to patient inquiries about their brain activity, such as what happens during sleep or if a seizure occurred.
During a study involving 13 patients at the Hospital of the University of Pennsylvania, the system accurately answered over 92 percent of questions about brain activity. Patients and the AI exchanged more than 1,300 messages as patients sought information about their condition and responded to seizure alerts, indicating whether events were genuine seizures, false alarms, or uncertain. The researchers found that incorporating this patient feedback reduced false alarms by approximately 80 percent.
While medical expert feedback further improved accuracy, the study suggests patients’ observations can significantly improve seizure detection. The team plans to conduct larger studies with patients at home to assess the system’s performance in everyday life and explore its potential use with devices for other conditions like Parkinson’s disease.
