Extraction of features in EEG signals with the non-stationary signal analysis technology

The efficient detection of feature waves of EEG signals can provide more proof in clinic diagnose for doctors, then improve the veracity of diagnosis. In the course of studying the EEG, all kinds of measurement and analysis methods in time domain and frequency domain have been used in clinic, and the validity of some time-frequency testing methods to be applied to detect and analyze the feature waves of EEG waves have been discussed in theory, but which have not been found in clinic. From the view of clinic application and science investigation, all kinds of time-frequency testing and analysis methods such as Gabor transform, Wigner distribution, Choi-Williams distribution and wavelets transform have been integrated into the "virtual EEG measurement and analysis instrument" for detecting the feature waves of EEG signals. It is indicated that the satisfied results can be obtained by select different time-frequency testing and analysis methods based on the different purposes and feature waves.