基于想象左右手运动脑电特征提取及其统计特性分析  

Statistics analysis and feature extraction of EEG for imaging left-right hands movement

作　　者：伍亚舟[1] 吴宝明[1] 何庆华[1] 张玲[2] 易东[2] 

机构地区：[1]第三军医大学大坪医院野战外科研究所五室,重庆400042 [2]第三军医大学卫生统计学教研室,重庆400038

出　　处：《北京生物医学工程》2007年第2期199-203,共5页Beijing Biomedical Engineering

基　　金：国家自然科学基金(30300418)资助

摘　　要：目的基于想象运动的脑-机接口(BCI)系统中,探讨如何获取思维脑电、提取特征并对其进行准确而有效地分类。方法利用小波变换方法构建小波系数提取不同思维脑电特征,并从统计学的角度对这些特征进行分析。结果在实际动作前0．5～1s左右,想象左右手运动时C3和C4处各自具有明显不同的脑电特征,且这些特征存在显著性差别(P＜0．05)说明小波系数能很好的反映不同思维脑电的特征。结论小波分析方法可以有效抑制或消除噪声和提取反映不同思维的脑电特征,通过对构建出的小波系数特征作统计学上的特性分析发现不同思维脑电具有显著性差别,该特征为后续分析中得以被准确地转换(识别分类)提供了更为可靠的保证。Objective To investigate the extraction and classification of EEG signals during imaging movement for designing Brain Computer Interfaces （BCI）. Methods Wavelet coefficient was constructed by using wavelet transform in order to extract feature of EEG for mental tasks, and these features were further analyzed statistially. Results Wavelet coefficient can reliably reflect feature of EEG for different mental tasks, because there are obviously different features for imaging left - right hands movement in c3 and c4 at 0.5 - 1 s befores movement, these feature have significance difference （P 〈0.05）. Conclusion Analysis method of wavelet can effectively inhibit and eliminate noise, and extract feature of EEG for mental tasks, these features provide reliable warrant for post - analysis of discrimination and classification.

关 键 词：脑-机接口 脑电 特征提取 思维作业 小波分析 

分 类 号：R318.04[医药卫生—生物医学工程]