A neural network system for automatic classification of sleep stages

The back-propagation neural network is utilized to classify sleep stages in humans. A single-channel EEG is segmented into equally spaced intervals, each interval corresponds to one-minute in time. Measurements of the time, frequency, and energy characteristics are carried out in each interval to construct the sleep pattern vector. An adaptive training algorithm is utilized to accelerate the training process. This neural network is useful for various neurological studies and clinical diagnoses.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>