AI-Powered Smartphone App Helps Epilepsy Patients Track Seizures and Understand Brain Activity
Researchers at the University of Pennsylvania have developed an artificial intelligence-powered smartphone application that enables people with epilepsy to interact with their brain-monitoring devices in real time, track symptoms and receive personalised information about their health. The technology, described in a study published in Nature Medicine on October 13, 2026, could help patients better understand their seizures and improve the accuracy of automated seizure detection.
The system allows patients to communicate with their monitoring devices through a smartphone application using text-based conversations. Instead of relying solely on medical professionals to interpret monitoring data, patients can ask questions about their brain activity, sleep patterns, medication-related information and potential seizure risks.
The research team developed the platform to connect electroencephalography (EEG) monitoring systems with an artificial intelligence interface. EEG technology records electrical activity in the brain and is widely used to investigate and monitor epilepsy. The AI system processes relevant physiological information and presents it to patients in an accessible format, including personalised notifications and visual summaries.
One of the most promising findings was the system’s ability to learn from patient feedback. When users reported symptoms or corrected information about their experiences, the AI could incorporate those observations into the seizure-detection process. In a simulated evaluation, this approach reduced false alarms by approximately 80%, while maintaining the baseline sensitivity of the detection system.
False alarms are a persistent challenge in automated seizure monitoring. Frequent incorrect alerts can cause anxiety for patients and caregivers, reduce confidence in monitoring devices and make people less likely to respond consistently to notifications. Improving accuracy could therefore make seizure-monitoring systems more useful in everyday care.
The study involved 13 patients undergoing epilepsy monitoring at the Hospital of the University of Pennsylvania. Eleven participants were monitored using scalp EEG, while two received intracranial EEG monitoring. Researchers evaluated how patients interacted with the application and examined whether their feedback could improve the performance of seizure-detection algorithms.
Participants were able to ask questions about their own monitoring data, receive alerts about clinical events and report symptoms and behaviours. The application was designed to support two-way communication, allowing the system not only to provide information but also to use patient observations to refine its analysis.
Researchers also explored how sleep, medication patterns and cognitive changes might relate to seizure activity. By making these connections easier to understand, the technology could help patients recognise patterns in their daily lives and have more informed discussions with their healthcare providers.
The researchers emphasised that the technology remains at an early stage of development. The initial study was small and conducted in a controlled hospital environment, meaning the results do not yet establish how well the system will perform during long-term use at home. Larger clinical studies will be needed to assess its reliability, safety and practical benefits.
The development comes amid growing interest in AI-assisted epilepsy management. Existing wearable devices and smartphone applications can already help record seizures, monitor movement and notify caregivers when certain seizure patterns are detected. The new platform differs by allowing patients to communicate directly with a monitoring system and contribute feedback that may improve its performance.
For the millions of people worldwide living with epilepsy, better monitoring could offer greater awareness of seizure patterns and additional support between clinical appointments. More reliable alerts may also help caregivers respond appropriately when assistance is needed.
However, the researchers caution that the application is not a replacement for neurologists, established diagnostic procedures or prescribed treatment. It was tested as a research platform linked to specialised EEG equipment, and it should not be assumed to be a publicly available smartphone app that can independently detect all seizures.
The findings point towards a more interactive model of healthcare in which AI systems and patients work together to interpret medical information. If further trials confirm the early results, similar technologies could eventually help people with epilepsy manage their condition more effectively and could potentially be adapted for other medical devices and chronic illnesses.
