Wavelet Packet Entropy in the Analysis of EEG Signals
Lisha Sun, Guoliang Chang, Hongrong Tang · 2006
Wavelet packet analysis arms with the ability to accurately resolve neuroelectric waveforms into specific time and frequency components. Wavelet packet transformation is applied to extract different kinds of dynamic EEG rhythms accurately. This paper presents a wavelet packet entropy method in the processing of EEG signal. Relative wavelet packet energy and two kinds of wavelet packet entropy are calculated as the quantitative parameter to study the complexity of the EEG signal. Experimental results show that the proposed method excels the wavelet decomposition in that it can isolate specific EEG and ERP rhythms more accurately. The proposed method is also useful for the processing of other non-stationary signals