静息态脑电信号Lemple-Ziv复杂度与运动想象响应能力的相关性  

Correlations Between Lempel-Ziv Complexity of Resting-State EEG and Motor Imagery Performance

作　　者：王坤[1] 许敏鹏[1,2] 徐立超 陈龙 明东 Wang Kun;Xu Minpeng;Xu Lichao;Chen Long;Ming Dong(不详;Neural Engineering&Rehabilitation Lab,Tianjin University,Tianjin 300072,China)

机构地区：[1]天津大学精密仪器与光电子工程学院,天津300072 [2]天津大学医学工程与转化医学研究院,天津300072

出　　处：《航天医学与医学工程》2020年第4期283-289,共7页Space Medicine & Medical Engineering

基　　金：国家自然科学基金(81630051,61976152,81601565);天津市科技重大专项与工程(16ZXHLSY00270)。

摘　　要：目的探究与用户运动想象响应能力相关的静息态脑电特征,构建用户筛选模型。方法根据运动想象响应能力将105名受试者分成3个组,分别提取每名受试者开目静息和闭目静息态脑电不同频带(theta、alpha、beta、gamma)的Lemple-Ziv复杂度(LZC)特征,利用皮尔森相关系数与独立样本t检验分析相关性最显著的LZC特征,构建运动想象响应能力优秀受试者组与随机受试者组的筛选模型。结果开目静息态alpha频带的LZC特征与运动想象响应能力具有最显著的相关性,其中基于C4、Cp4导联的LZC特征构建的分类模型可用于不同运动想象任务响应能力的受试者筛选,最高可获得78.04%的分类正确率,优于已有文献报道的能量与谱熵特征。结论静息态脑电LZC特征是一种有效、通用的运动想象响应神经生理指标,有望为运动想象脑-机接口用户的筛选提供技术支持。Objective To explore the resting-state EEG characteristics related to the user’s motor imagery performance,which can be used to build a screening model for users.Methods The 105 test subjects were divided into three groups according to their motor imagery performances.The Lempel-Ziv complexity(LZC)values of resting-state open and closed-eye EEG of different frequency bands(theta,alpha,beta and gamma)of each subject were calculated.The LZC features with the most significant correlation were selected according to the analysis of Pearson correlation coefficient and independent sample t-test.Then,the above features were used to construct a classifier to separate the high motor imagery performance group from the low motor imagery performance group.Results The LZC of alpha-band resting open-eye EEG had the strongest significant correlations with MI performance.The LZC features on channel C4 and Cp4 were used to build the screening model of the MI performance,with the highest classification accuracy of 78.04%,which outperformed the previous power and spectral entropy features.Conclusion The LZC of resting-state EEG is an effective and general neural marker for motor imagery performance,which is helpful for user screening of motor imagery based brain-computer interface.

关 键 词：静息态 脑电 Lemple-Ziv复杂度 运动想象响应能力 

分 类 号：R318[医药卫生—生物医学工程]