Automatic sleep level evaluation based on conditional probability for day time nap sleep evaluation

In this study, the automatic technique for the continuous light sleep level estimation and prediction is developed. The ultimate purpose is to realize the automatic sleep level evaluation for short-term day time nap sleep. The main method is based on the conditional probability obtained according to the characteristics of neuro-physiological signals including EEG, EOG and EMG. The estimated and predicted sleep level are compared with the visual inspection. The obtained results showed that the presented automatic technique is practicable to be an assistant subjective evaluation tool for sleep level analysis.