【摘要】 脑电同步是脑功能区域整合的重要表现。基于时间序列的排序模式,提出一种简单易行的多通道脑电信号同步分析方法——全局排序模式同步(GMS)。仿真分析显示,该算比基于加权排序互信息的全局同步因子对弱耦合的检测更为灵敏。对26例遗忘型轻度认知障碍和20例认知功能正常的2型糖尿病患者闭眼静息态的脑电信号,采用基于小波增强的独立分量分析算法进行预处理,将32路脑电信号分为前额、中央区、顶区、枕区、左颞和右颞6个区域进行全局同步分析,并利用独立样本t检验对两组被试之间的人口学特征、神经心理学检查和同步值进行统计分析,利用皮尔森线性相关分析研究各区域同步值和认知功能之间的关系。结果显示,糖尿病轻度认知障碍患者与正常对照组相比,各脑区的全局排序模式同步值均减小,尤其是中央区(P<0.01)、顶区(P<0.05)和枕区(P<0.05)有显著性的降低,且前额(r=0.298,P=0.045)、中央区(r=0.327,P=0.026)、顶区(r=0.32,P=0.03)的全局排序模式同步值均与MOCA得分有显著的正相关性,表明GMS是与认知功能下降相关的脑电特征。更多还原

【Abstract】 EEG synchronization is considered to be the performance of brain functional area integration.A time series motif based multi-channel synchronization method——global motif-synchronization(GMS) was proposed in this study.The simulation analysis indicated that the new algorithm was more sensitive than S-estimator based normalized weighted permutation mutual information in detection weak coupling.The algorithm was used to analyze the EEG synchronization of 26 amnesic MCI and 20 normal controls of patients with diabetes in eye-closed resting state.The wavelet enhanced independent component analysis was used to eliminate artifacts.The 32-channels EEG was divided to frontal,central,parietal,occipital,left temporal and right temporal region respectively.The independent samples t-test was performed to test differences in demographic characteristics,neuropsychology and regional synchronization values between two groups.The Pearson’s linear correlation was used to study the associations between regional synchronization values and cognitive functions.The results showed that GMS values in each brain region of diabetes patients with MCI were lower than that of control group.Especially,the GMS values decreased significantly in central(P<0.01),parietal(P<0.05)and occipital(P<0.05) regions.The MOCA scores and GMS value had a significant positive correlation in frontal(r=0.298,P=0.045),central(r=0.327,P=0.026) and parietal(r=0.32,P=0.03) regions.The GMS is an important EEG characteristic that is correlated with cognitive function impairment.更多还原