结合小波包和ICA技术的脑电特征提取研究

Research on feature extraction of EEG signal based on wavelet package and ICA

为了判断流经人体电流达到感知电流阈值前后脑电特征是否发生变化,采用小波包和独立分量分析(ICA)技术相结合的方法,对脑电α波进行特征提取。利用小波包分解,重构相关频段信号来构造特征波参考信号,然后进行ICA分析,最终提取出脑电α信号特征波。实验证明,结合小波包和ICA技术能够有效地提取脑电信号特征波。

In order to judge whether the EEG （electroencephalogram） feature changes when the current flow through the hu- man body reaches human body＇s perception current threshold, the method that combines wavelet package with independent com- ponent analysis （ICA） is adopted to extract the feature of α wave from EEG. The signal in the related frequency is reconstructed for constructing the reference signal of feature wave by using wavelet package decomposition. Then ICA analysis is carried out to extract feature wave of EEG α signal. The experimental results proves that combination of wavelet package with ICA technoligy can effectively extract the feature wave of EEG signal.