Sleep EEG Staging Based on Hilbert-Huang Transform and Sample Entropy

In this paper, a sleep EEG staging method based on Hilbert-Huang transform and sample entropy was proposed to improve the efficiency of automatic sleep staging. Firstly, the marginal spectrum of EEG was obtained by Hilbert-Huang transform. Secondly, features were extracted based on the energy distribution of the frequency band and EEG sample entropy. Thirdly, the extracted features were employed to construct an SVM classifier. Experiment results based on 10 subjects showed that the average sleep staging accuracy was 89.9%, which verified the effectiveness of the proposed method.