ERP单次提取中的小波变换模极大值恢复算法  

THE RECONSTRUCTION ALGORITHM BASED ON MODULUS MAXIMA OF WAVELET TRANSFORM FOR SINGLE TRAIL ERP

作　　者：李鲁平[1] 程宏伟[1] 倪鹤鹦[1] 王兴邦[1] 马瑞山[1] 程敬之[2] 

机构地区：[1]第四军医大学空医系,西安710032 [2]西安交通大学生物医学工程与仪器研究所,西安710049

出　　处：《中国生物医学工程学报》2000年第2期181-184,共4页Chinese Journal of Biomedical Engineering

摘　　要：自发脑电的频谱不规则 ,与有效信号ERP频谱相重叠 ,传统滤波方法难以奏效。但由于随机噪声奇异性指数与有效信号奇异性指数大小不一样 ,小波变换模极大值在不同尺度下传播行为也不一样 ,据此可将有效信号从随机噪声中提取出来。我们发现利用小波变换模极大值恢复算法可以提取出单次ERP ,并将此方法用于The frequency of EEG was irregular and was the same as the eventrelated potential (ERP). The traditional filtering method was ineffective. However, the singularity index of the noise and the signal is not same. The decay of the wavelet transform modulus maxima was also different. This difference could be used to distinguish the noise and the signal. In this study, a new reconstruction method based on the modulus maxima of wavelet transform was used to estimate the single trail ERP from 11 male subjects.

关 键 词：事件关联电位 小波变换 单次提取 脑电图 

分 类 号：R741.044[医药卫生—神经病学与精神病学]