Single Channel Electroencephalogram Measurement with Multi-scale Entropy Analysis for Evaluating Day Time Sleep

Abstract

A general method for evaluating sleep is by identifying the stages of sleep from electroencephalogram (EEG), electrooculogram (EOG) and electromyogram (EMG). Unfortunately, visual sleep scoring is usually subjective and time consumptive. This was the reason for the development of automatic sleep staging based on multi EEG, EOG, and EMG. Recently, automatic sleep stages based on single channel EEG is also being developed to reduce the unnecessary number of wire. The problem is that this approach needs 90 min to complete the sleep stages cycle while normal short nap or day-time sleep experiment duration is usually less than 60 min. With this less than 60 min’ sleep duration issue, the purpose of this research is to develop a single channel EEG evaluation method by measuring changes in amplitude power of EEG at a specific time of day time sleep. The amplitude difference in term of a complexity of EEG will be used as an experiment index. We calculated EEG complexity with Multi-Scale Entropy (MSE) for day-time sleep experiment. The average entropy as the final result from the method could determine the EEG complexity of frequency power and have a negative correlation with the delta wave. Based on this result, average entropy was used and had demonstrated a significant difference. MSE method could be useful for single channel EEG evaluation and significantly perform better understanding of day-time sleep.