Detecting epileptic seizures with wearable devices

点击翻译按钮获取中文摘要

Sudden unexpected death in epilepsy (SUDEP), which occurs during or immediately after a seizure, results from the person having pauses in breathing, a convulsion placing the unconscious person in a suffocating position, developing an abnormal heart rhythm, or a combination of these. It is important for someone else, like a family member or care-giver, to assess and assist the individual having a seizure. Neutun Labs have developed an app that works with your smartphone or smartwatch to detect potential seizures and notify chosen individuals via text, phone, and/or first-responder alerts. Their current detection relies on wearables with an accelerometer. Seizures are best detected via changes in brain activity measured with electroencephalography (EEG). Recently a number of consumer grade wearable EEG headsets have entered the market. Neutun Labs would like to integrate EEG data collected with a wearable EEG headset during sleep with the kinematic data to improve seizure detection and reduce the incidence of false alarms. This project focuses on EEG signal processing and analysis combined with kinematic data analysis to develop the technology for real-time app-based event detection that produces accurate seizure alerts. The ability to mathematically calculate the likelihood that an event was a seizure is just starting to be transferred into an ability to predict the likelihood of a seizure about to occur. Epilepsy detection and prediction wearables are a nascent field that is currently focused on pulse, temperature, and skin conductance measures feeding into a smartphone app. If we can effectively and efficiently combine kinematic and EEG information, we will produce a significant leap in technology.