Improved EEG signal processing with wavelet based multiscale PCA algorithm

In this paper, wavelet based multi scale PCA algorithm is proposed and demonstrated to enhance the classification performance in identifying EEG signals. The signal decomposition is done using wavelet transform, followed by the de-correlation using PCA, to achieve maximum compression, while simultaneously preserving the dominant modes of the signal and bad data rejection. The optimum decomposition scale of wavelet transform is selected based on the energy of wavelet coefficients in each scale. The proposed algorithm employing multi scale PCA computes the principal components of the wavelet coefficients at each scale, followed by combining the results at relevant scales. The wavelet coefficients of a particular scale corresponding to the dominant eigen values are retained for signal compression. These results in effectively compresses the EEG signals while preserving the model information of the signal.