基于改进Renyi熵算法的EEG心算任务识别  

Recognition for EEG mental arithmetic tasks based on an improved Renyi entropy algorithm

作　　者：李鑫[1,2] 黄丽亚 LI Xin;HUANG Liya(School of Electronic and Optical Engineering,School of Flexible Electronics(Future Technologies),Nanjing University of Posts and Telecommunications,Nanjing 210023,China;National Joint Engineering Laboratory of RF Integration and Microassembly Technology,Nanjing 210023,China)

机构地区：[1]南京邮电大学电子与光学工程学院、柔性电子(未来技术)学院,江苏南京210023 [2]射频集成与微组装技术国家地方联合工程实验室,江苏南京210023

出　　处：《南京邮电大学学报（自然科学版）》2023年第6期44-51,共8页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：国家自然科学基金(61977039);高校教育信息化研究项目(2019JSETKT009)资助项目。

摘　　要：结构熵是度量网络复杂度的重要手段,为了弥补传统结构熵仅仅关注网络单一特性的问题,提出了一种改进Renyi熵算法来研究心算任务下的EEG脑网络,引入了两个重要网络属性——分形维数和介数中心性来提高网络复杂性的度量能力。之后,基于心算EEG数据计算两两电极间的相位锁定值(PLV),构建了复杂脑网络,并进行复杂度分析。结果表明,在α频段,心算状态下额叶与顶枕叶的脑同步性低于休息状态,心算状态的脑网络复杂性高于休息状态。利用支持向量机(SVM)实现了休息、心算状态的识别,算法识别准确率达到了88.42%。Structural entropy is an important measure of network complexity.In order to remedy the problem that traditional structural entropy focuses only on a single property of the network,this paper proposes an improved Renyi entropy algorithm to study the electroencephalogram(EEG)brain network for mental arithmetic tasks,and introduces the fractal dimension and the median centrality to improve the measurement ability of network complexity.The phase locking value(PLV)between two electrodes is calculated based on the EEG data,and a complex brain network is constructed.The complexity of different brain networks is compared.Finally,the identification of resting and mental arithmetic states is achieved by using the support vector machine(SVM).The results show that the brain synchronization in frontal and parieto⁃occipital lobe in mental calculation state is lower than that in resting state within the alpha frequency band,and that the complexity of the brain network in mental arithmetic state is higher than that in resting state.The improved Renyi entropy algorithm achieves an accuracy of 88.42%.

关 键 词：脑电 心算 复杂网络 脑网络 结构熵 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]