小鼠视觉刺激相关的脑电波分离研究  

On Visual Stimulation Related Brain Wave Separation for Mice

作　　者：杨静[1] 项磊[1] 王冠皓 邓健康[1] 吕红[1] 刘青山[1] 

机构地区：[1]南京信息工程大学,江苏南京210044

出　　处：《数学的实践与认识》2015年第14期119-125,共7页Mathematics in Practice and Theory

基　　金：国家自然科学基金(61272223;61300162);江苏省自然科学基金(SBK201210296;SBK201204234)

摘　　要：脑电波信号是无数神经元放电的混合,而实际测量所得到的脑电波信号又是由许许多多构成不同想法和控制行为的脑信号合成的.只有将脑电波信号很好地分离,才能从中确定某种脑电波与某种行为之间的关系.由于采集到的脑电信号强度很微弱,信号的强弱差别很大,所以传统的盲源分离模型很难凑效.结合独立成分分析(Independent Component Analyse,ICA)方法,首先对原始脑电信号进行去噪处理、然后,分离出呼吸相关的脑电信号,接着,根据视觉刺激对分离后的脑电信号进行时域对齐,最终在对齐的脑电信号片段上分离出与视觉相关的脑电信号.EEG is a mixture of countless discharge of neurons,and the actual measurement brain signals is made up of many different ideas and control behavior.Unless we separate the brain signals well,can we determine the relationship between a certain brain waves with a certain behavior.Due to the weakness of brain electrical signal we obtain,and the strong difference between the signals,the traditional blind source separation model is very difficult to work.Based on Independent Component analysis（Independent Component Analyse,ICA） method,we deal with the problems as follows：first,deal with the noise of the original eeg signals,then,isolate respiratory related brain electrical signal,then,according to the visual stimulation after separation of the brain signals in time domain alignment,gain the EEG related to visual brain electrical signal on the final pieces

关 键 词：盲源分离 脑电波分离 独立成分分析 时域对齐 

分 类 号：TN911.7[电子电信—通信与信息系统]