Feature Extraction by Combining Wavelet Packet Transform and Common Spatial Pattern in Brain-Computer Interfaces

摘要

Wavelet packet transform (WPT) and common spatial pattern (CSP) are two commonly used methods for feature extraction in brain-computer interfaces. In this paper, a new feature extraction method was proposed that was based on the combination of WPT and CSP. The raw EEG signals were band pass filtered between 8 and 30 Hz, then the filtered signals were subject to WPT and reconstruction, and finally the reconstructed signals were spatially filtered by CSP algorithm. The proposed algorithm was applied to six datasets recorded during BCI experiments based on motor imagery. The results showed superior classification performance, thus verifying the feasibility and validity of the algorithm.