Can a computer program identify seizures and other highly epileptiform brain activity in electroencephalography (EEG) signals with accuracy comparable to physicians with subspecialty training in clinical neurophysiology?
We trained a deep neural network using 6,095 scalp EEG recordings from 2,711 patients. The algorithm reached sensitivity, specificity, precision, and calibration equal to typical 61 fellowship-trained clinical neurophysiologists. By providing expert-level automatic classification of seizures and highly epileptiform brain activity on EEG, we can augment brain monitoring in patients at high risk for such events, particularly in settings without continuously available human expertise.
