EEG entropy estimation using a Markov model of the EEG for sleep stage separation in human neonates

Entropy calculated on EEG has been shown to be a useful indicator of effects from insufficient oxygen supply. In this paper, the estimation of entropy is based on transition matrices instead of probability density functions. It is shown that the separation of sleep stages thereby can be improved. This suggests that by including time information given by the transition matrix in entropy estimates of the EEG, classification can be improved.