Eeg feature extraction and recognition with different mental states based on wavelet transform and accln network

The electroencephalogram (EEG) is a record of brain activity. Brain Computer Interface (BCI) technology has become one of the hotspots, especially for the identification of EEG characteristic signals. We here describe a novel method which involves the combination of discrete wavelet transformation and neural network to recognize different states of the human brain, including fatigue, consciousness and concentration from EEG signal. To eliminate the high frequency noise, raw signal was preprocessed by the wavelet denoising method and was then decomposed into multi-layer high frequency signal and low frequency signal. Thus, δ wave, θ wave, α wave, β wave were obtained by wavelet transformation. In this experiment, the frequency band energy of the different waves was regarded as the feature signal of EEG for further signal processing. The feature signal was then classified by both radial basis function (RBF) and annealed chaotic competitive learning network (ACCLN). The experimental results showed that the average accuracy of ACCLN network is 98.4%, which is much higher than the traditional method. The results together showed the effectiveness and feasibility of the proposed method. The proposed algorithm has a good practical value in the analysis of the mental states of a driver or high risk operation personnel.