排列复杂性度量应用于脑机接口信号分析  

Permutation Complexity Measure Applied in Brain-Computer Interface Signal Analysis

作　　者：柯大观[1] 童勤业[1] 

机构地区：[1]浙江大学生物医学工程系,杭州310027

出　　处：《传感技术学报》2007年第3期596-600,共5页Chinese Journal of Sensors and Actuators

基　　金：国家973预研项目资助(2002CCA01800);国家自然科学基金项目资助(30170267)

摘　　要：在排列分划的基础上,应用Lempel-Ziv复杂性和最新定义的格子复杂性分析脑机接口信号.由于对非线性时间序列的排列分划进行了重要改进,使这种粗粒化方法具有了普遍的适用性.与经验模式分解结合,将排列分划与常用的均值分划作了比较.实验表明,基于排列分划的复杂性度量可以取得较好的效果,甚至超过了均值分划经经验模式分解所能达到的最好结果,说明排列复杂性度量可作为脑电信号分析的新方法,尤其适用于需要快速处理的场合-比如脑机接口.With the method of permutation partition, Lempel-Ziv complexity and new-defined Lattice complexity were applied to analyze signal of brain-computer interface. Because of the important modification made on permutation partitions of nonlinear time series, this coarse graining method now can be generally used on arbitrary series. In this study, the permutation partition and the common-used average partition were compared with each other, both accompanied by empirical mode decomposition. The results showed that the complexity measure based on permutation partition could do the job even more than the best result of average partition made on empirical mode decomposition. It confirmed that permutation complexity measure could be a useful new way to analyze brain data, especially under the occasion that rapidly processing was needed-brain-computer interface for instance.

关 键 词：脑机接口 排列分划 复杂性度量 经验模式分解 

分 类 号：R318[医药卫生—生物医学工程] TN911.6[医药卫生—基础医学]