基于ICA与PSD的ERD／ERS信号检测

介绍一种基于独立分量分析（ICA）空间滤波结合功率谱密度（PSD）曲线分析法用于提取大脑在想象动作时产生事件相关去同步／同步（ERD／ERS）信号的方法．其检测流程为：先对想象动作诱发的脑电（EEG）信号进行ICA分解得到独立分量与相应的解混矩阵，再按特征频段取其主要分量得到滤波后数据，然后采用短时傅里叶变换计算相关导联EEG信号在特征时段与频段的PSD曲线，引入ERD／ERS系数作为量化指标以进行想象动作的识别．计算结果表明，上述方法能够显著增强运动想象脑电信号的ERD／ERS特征信息，且通过实际分类验证，采用该方法可以获得更高的识别正确率，较传统信息检测方法平均提高8％以上．

Detection of ERD/ERS Signals Based on ICA and PSD

A new method which can extract event related desynchronization or event related synchronization(ERD/ERS) signals produced by imaginary movement in brain based on the independent component analysis(ICA)combined with the power spectral density(PSD)curves was presented.Firstly,electroencephalography(EEG)signals were decomposed into independent components to obtain the solution matrix and reconstruct the filtered data by the main components.Then PSD curves of EEG at the correlated electrodes during the feature time or frequency intervals were calculated using a short-time Fourier transform(STFT).Finally,ERD/ERS coefficient was introduced as a quantity index for the recognition of imaginary movements.The calculated results show that the proposed method can significantly enhance the feature information of ERD/ERS produced by imaginary movement,and through validation of practical classification,the proposed method can obtain higher identification rate in comparison with traditional detection methods,over 8% better on average.