节点文献
MEA信号锋电位的主成分分类
Principle Component Sorting of MEA Signal Spike
【摘要】 分类是MEA信号锋电位研究的重要任务。卡—洛变换(KLT)是所有正交变换中方差分布最佳者。按大小排列的信号协方差矩阵的本征值,即信号的主分量,用主分量的和超过信号总量90%的最大的8个数据为分类根据,对MEA信号锋电位采用沃德法进行系统聚类,获得了较好的结果。
【Abstract】 Sorting is a basilic task of MEA signal spike research.Karhunen-Loeve transform(KLT) is optimum of variance distribution in all orthogonal transformation.Eigenvalues on the order of a covariance matrix of signal are principles components of signal.Eight principles components are used on system cluster,their sum is larger than ninety percent of totality of signal,The result of Ward method is good on the sorting of MEA signal spike.
【关键词】 锋电位;
主成分分析;
卡—洛变换;
系统聚类;
【Key words】 Spike; Principle component analysis; Karhunen-Loeve transform; System cluster;
【Key words】 Spike; Principle component analysis; Karhunen-Loeve transform; System cluster;
【基金】 国家973前期研究项目(2001CCA04100);国家自然科学基金资助项目(30170306);国家杰出青年科学基金资助项目(60025514)
- 【文献出处】 生物医学工程研究 ,Journal of Biomedical Engineering Research , 编辑部邮箱 ,2005年03期
- 【分类号】R318.04
- 【被引频次】8
- 【下载频次】91