Feature Extracting Of Weak Signal In Real-Time Research Of Sleeping EEG

Sleep is the physiological function of the animal world, which is an important life-sustaining activities and physical activity. Electroencephalogram (EEG) is one of the most important neuroelectrical signals, which is often used to detect brain's neuroelectrical dysfunction. Sleep research is based on correct staging of sleep signals. However, the analysis of the EEG signal and extraction of features from it has been a challenging task due to the complexity and variability of signals. It is difficult to automatically recognize different stages of the real-time sleeping EEG signal. The previous studies of sleep stage are based on multi-channel signals, and even combine the features of electrooculogram (EOG) and electromyographic (EMG) signals with the analysis of the EEG. In our study, the features were extracted from single channel sleeping EEG signal of rats using approximate entropy (ApEn) combined with bispectrum analysis, which is more difficult.This article was focused on the feature extraction for different stage of sleep EEG in rats, using ApEn and bispectrum. The results show that ApEn and the maximal amplitudes of bispectrum can extract various features in different sleeping EEG stages. The bispectrum can detect the presence of phase coupling among different stages in rats sleeping EEG signal. The ApEn with change of bispectrum's frequency and maximal amplitudes of bispectrum can be effectively applied to automatically compartmentalize real-time sleeping EEG signal into different stages.The new method proposed in the paper is significant to automatic EEG analysis and intelligent diagnosis of brain diseases. The results also provide a new way of features extraction for other non-stationary signals in real time.