EEG-based event detection using optimized echo state networks with leaky integrator neurons
Long-haul truck drivers, train drivers and commercial airline pilots routinely experience monotonous and extended driving periods in a sedentary position, which has been associated with drowsiness, microsleeps, and, consequently, serious accidents. Microsleeps are brief involuntary events of lapses in attention or responsiveness, associated with events such as prolonged eye closure which usually last from 0.5-15 s . Consequently, the detection and preferable prediction of the microsleeps in subjects, especially those working in these high-risk occupations, is very important to workplace safety. The current study aims at developing a microsleep detector using the novel recurrent neural network architecture of an echo state network (ESN) and represents progression of our research from previous methods [1, 2].