Wearable sensors show promise for automatic seizure detection

Researchers have reported promising findings from the use of machine learning techniques to develop wearable sensors for automatic seizure detection.

The development of effective wearable sensors could help avoid reliance on seizure diaries, in which under-reporting of seizures is a problem.

At AES 2020, the researchers from IBM Research, Harvard Medical School and Boston Children’s Hospital, Massachusetts, USA, presented findings from a study of 75 pediatric patients (mean age 11.1 years) admitted to an epilepsy monitoring unit for long-term EEG monitoring.