摘要
脑电诱发电位的单导少次提取一直是生物医学信号处理领域倍受关注的问题。独立分量分析作为解决盲源分离问题的一种有效算法已被广泛应用于诱发电位提取之中。独立分量分析处理的是多路观测信号,且要求观测信号路数大于或等于独立信号源的个数。为了能够应用独立分量分析算法实现诱发电位的单导少次提取,引入虚拟通道构建观测信号矩阵,从而得到符合实际应用条件的算法模型。4路信号仿真实验表明了虚拟通道模型可以有效提取诱发电位。对12位受试者进行模式翻转视觉诱发电位测试,仅用单导连续4次记录即可实现诱发电位的初步提取,信噪比增加约为12 dB,当采用10路虚拟通道,信噪比提高约20 dB。4路和10路虚拟通道ICA方法下得到的多导联VEP相关系数的统计结果进一步证实增加虚拟通道的数量,EP信号提取效果也会更好。
Signal estimation of single channel brain evoked potential （EP） with few-trial is of great interest. As an effective method to solve blind source separation problem, independence component analysis （ICA） algorithm has been widely applied to EP extraction. However, the ICA method requires equal or greater number of observed channels than that of independent signal sources. In order to realize single channel EP extraction, this paper built observation signal matrix for the ICA method based on the concept of virtual channels, which led to a more realistic model for EP applications. The simulation experiments with four-channel signals indicate the virtual-channel ICA method extracted EP effectively. In another experiment, EPs were recorded from 12 subjects in a pattern reversal visual evoked potential test. With the virtual-channel ICA, the VEP in a single channel was extracted with only four continuous records increasing SNR by approximately 12 dB. When using ten-channel virtual ICA, the SNR improved about 20 dB. The statistical results of the VEP correlation coefficients confirmed that the EP signal could be extracted more effectively with increased virtual channels.
出处
《中国生物医学工程学报》
CAS
CSCD
北大核心
2012年第5期704-711,共8页
Chinese Journal of Biomedical Engineering
基金
国家自然科学基金(60940023
61172108
61139001
61005088)
国家科技支撑计划项目(2012BAJ18B06)
机器人技术与系统国家重点实验室开放研究项目(SKLRS-2010-ZD-07)
关键词
诱发电位
独立分量分析
虚拟通道
单导少次提取
evoked potential （EP）
independent component analysis （ICA）
virtual channels
few-trial extraction of single channel
