The invention discloses a method for analyzing the effective connection of brain functions. The present invention is directed to the field of functional brain imaging technology in the existing connection area between the effective interest method to study the causal conditions of Granger fMRI data, EEG data sense, because of the high degree of redundancy between the multi-channel signal, resulting in effective connectivity analysis of errors and put forward. The invention uses Yonghong? Granger causality and principal component analysis conditions of Chen, calculated without considering the same spectrum estimation consistency problem in the process of conditional Granger causality, excluding negative causality has no physical meaning spectrum, eliminates the high redundancy of multi-channel signals, effectively improve the accuracy of effective connectivity analysis.
【技术实现步骤摘要】
【技术保护点】
１．一种脑功能有效连接分析方法，其特征在于，包括如下步骤：步骤１：对脑功能磁共振成像数据、脑电数据进行预处理；脑功能磁共振数据的预处理：首先对脑功能磁共振成像数据进行空间位移的校正，再将校正后的图像进行标准化处理，然后对标准化处理后的图像利用高斯核进行空间平滑处理；脑电数据的预处理：首先对脑电数据进行归一化，然后进行去漂移；步骤２：从预处理后的脑功能磁共振成像数据或者脑电数据的感兴趣区域中提取一组至少包含４道时间信号的时间信号集Ｖ，对于信号集Ｖ中的信号做广义平稳性检验，如若不平稳，采取差分或者滤波使之达到平稳性；选择任意的两道信号Ｖｘ，Ｖｙ作为目标信号，时间信号集Ｖ中剩余的信号Ｖｚ作为条件信号；步骤３：对条件信号Ｖｚ进行主成分分析，然后选择若干个主成分，并使用若干个主成分构建条件信号Ｖｚｐ来代替原来的条件信号Ｖｚ；步骤４：利用Ｙｏｎｇｈｏｎｇ Ｃｈｅｎ的矩阵分离思想的条件Ｇｒａｎｇｅｒ因果计算方法，将目标信号Ｖｘ、Ｖｙ和条件信号Ｖｚｐ带入矩阵分离思想的条件Ｇｒａｎｇｅｒ因果公式，计算频域上的Ｖｙ对Ｖｘ的条件Ｇｒａｎｇｅｒ因果值ＦＶｘ－＞Ｖｙ｜Ｖｚｐ；步骤５：利用靴带抽样，获得条件Ｇｒａ...
【技术特征摘要】
还没有人留言评论。发表了对其他浏览者有用的留言会获得科技券。