Ambulatory seizure forecasting with a wrist-worn device using long-short term memory deep learning

• DOI: 10.1038/s41598-021-01449-2
• Publication Date: 2021-11-09T11:02:38Z
• Authors: Mona Nasseri; Tal Pal Attia; Boney Joseph; Nicholas M. Gregg; Ewan S. Nurse; Pedro F. Viana; Gregory Worrell; Matthias Dümpelmann; Mark P. Richardson; Dean R. Freestone; Benjamin H. Brink...
• Keywords: Adult; Male; Science; Q; R; 610; Electroencephalography; Middle Aged; Wrist; Article; Cohort Studies; Wearable Electronic Devices; Young Adult; 03 medical and health sciences; Deep Learning; 0302 clinical medicine; Memory; Seizures; Medicine; Humans; Female; Forecasting

Abstract The ability to forecast seizures minutes hours in advance of an event has been verified using invasive EEG devices, but not previously demonstrated noninvasive wearable devices over long durations ambulatory setting. In this study we developed a seizure forecasting system with short-term memory (LSTM) recurrent neural network (RNN) algorithm, wrist-worn research-grade physiological sensor device, and tested the patients epilepsy field, concurrent confirmation via implanted recording device. achieved performance significantly better than random predictor for 5 6 studied, mean AUC-ROC 0.80 (range 0.72–0.92). These results provide first clear evidence that direct forecasts are possible setting many epilepsy.