Real-time automated neural-network sleep classifier using single channel EEG recording for detection of narcolepsy episodes.

Conventional sleep staging and classification methods involve complicated settings to acquire multiple electrophysiological signals for extended recording durations, followed by specialists' analysis which is a time consuming exercise. These procedures need to be carried out in sleep clinics and are not suitable for applications based on real-time sleep monitoring and analysis. In this paper, a real-time sleep staging and classification technique is proposed using single EEG channel based on an artificial neural network classifier. This method is optimized to run on portable processing platforms with limited processing capabilities.