摘要
为了实现电极植入法采集到的单通道大鼠视觉诱发电位的少次提取,提出一种基于虚拟通道的独立分量分析-小波变换的大鼠视觉诱发电位少次提取方法。首先,引入虚拟通道,用快速独立分量分析方法将噪声与视觉诱发电位进行分离;然后,对分离出的视觉诱发电位进行少次叠加平均,提取出初步视觉诱发电位;最后,针对虚拟通道与实际噪声不完全吻合带来的残留噪声问题,用小波变换进一步去除初步视觉诱发电位中的残留噪声。实验结果表明,基于虚拟通道的独立分量分析8次叠加平均,就能很好地提取出初步的视觉诱发电位;小波阈值法进一步去除了残留噪声;本算法成功地实现大鼠视觉诱发电位的少次提取,同时该方法可以用于单通道其他脑电信号的提取。
In order to extract the rats’visual evoked potential（VEP） recorded by implanted the electrodes in rats’primary visual cortex through several times,a method based on the independent component analysis（ICA） of a virtual channel and wavelet transform（WT） was proposed.Firstly,the noise and the VEP were separated by ICA of the virtual channel.Then,the initial VEP was extracted by averaging the separated VEP several times.Finally,the left noise caused by inconformity of virtual channel and the real noise was decreased by wavelet transform.The results showed that the initial VEP was successfully extracted by averaging 8 times the separated VEP which was obtained by the independent component analysis of a virtual channel,and the wavelet threshold accomplished the denoising.In addition,this approach could apply to similar single channel signals’extraction.
出处
《中国生物医学工程学报》
CAS
CSCD
北大核心
2010年第3期379-383,共5页
Chinese Journal of Biomedical Engineering
基金
国家自然科学基金资助项目(60841004
60971110)
关键词
视觉诱发电位
虚拟通道的独立分量分析
小波变换
少次提取
visual evoked potential （VEP）
independent component analysis of virtual channel
wavelet transform
several trials extraction
