隐秘信息的脑电检测  

EEG detection of secret information

作　　者：官金安[1,2] 段亚峰 徐世行 李东阁 印想 彭翰林 潘先攀 GUAN Jin’an;DUAN Yafeng;XU Shixing;LI Dongge;YIN Xiang;PENG Hanlin;PAN Xianpan(Key Laboratory of Cognitive Science of State Ethnic Affairs Commission,College of Biomedical Engineering,South-Central University of Nationalities,Wuhan 430074,China;Hubei Key Laboratory of Medical Information Analysis and Tumor Diagnosis & Treatment,South-Central University of Nationalities,Wuhan 430074,China)

机构地区：[1]中南民族大学生物医学工程学院,认知科学国家民委重点实验室,武汉430074 [2]中南民族大学医学信息分析及肿瘤诊疗湖北省重点实验室,武汉430074

出　　处：《中南民族大学学报（自然科学版）》2019年第2期223-226,共4页Journal of South-Central University for Nationalities：Natural Science Edition

基　　金：国家自然科学基金资助项目(91120017);中央高校基本科研业务费资助项目(CZY18047)

摘　　要：为揭示疑犯隐藏的真实信息,检测隐秘信息的脑电,设计了一个猜测受试者真实名字.结果表明:在个体对不同自我相关程度名字产生刺激,在刺激出现后的300~600ms内,本人名字诱发的正波幅值大于陌生名字刺激.通过小波变换提取特征,用支持向量机进行训练和分类.在进行5个试次叠加平均后,采用PO3通道可将自己的名字分类成功,5位被试平均正确率达98%,该方法可应用于个体隐秘信息的脑电检测.To uncover the hidden information of suspects,electroencephalogram(EEG)containing secret information was measured.An experiment to find the real names of the subjects was designed.It was found that the individuals were stimulated by self-relevant names with different degrees.The amplitude of positive wave induced by one's own name was larger than the other names during the 300-600 ms after the stimulation.The feature points were extracted by wavelet transform,then trained and classified by support vector machine.After 5 trials of superposition averaging,the average accuracy of 5 subjects could reach 98%by using PO3 channel to classify their own names.This method could be applied to EEG detection of individual hidden information.

关 键 词：自我相关程度 隐秘信息 事件相关电位 小波变换 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]