Sleep stage classification based on EEG Hilbert-Huang transform

Abstract

The aim of this work is to propose an automatic sleep stage classification technique of electroencephalogram's signals (EEG) using Hilbert-Huang transform. EEG signals are analyzed with the Hilbert-Huang transform, instantaneous frequency with the physical meaning is obtained; The energy-frequency distribution of EEG was used as features parameters for each sleep stage; Ultimately using nearest neighbor method for pattern classification complete classifying sleep stage. According to experimental results of 560 samples of sleep EEG, average accuracy rate of the method achieved 81.7%. In a word, The EEG Hilbert-Huang transform based method can be used as an effective sleep staging classification.