目的 利用静息态BOLD-f MRI探讨正常老年人脑功能网络是否具有小世界性.方法 20名健康老年志愿者纳入研究.采集静息态BOLD-fMRI数据.应用SPM 5软件进行数据预处理,利用解剖学自动标记模板将大脑分成90个区域,提取每个区域内所有体素的时间序列平均值,计算每2个区域之间的Pea rson相关系数,构建N×N(N=90)的相关性矩阵R,并对矩阵进行Fisherr toz变换.采用直接定义阈值T和设定矩阵稀疏度两种方法确定阈值,根据小世界性的定义:当γ=Gp/Crand ＞1,且λ=Lp/Lrand≈1时(C为平均聚类系数,L为平均路径长度,p表示规则网络,rand表示随机网络),认为该网络具有小世界性.结果 在直接定义阈值T条件下,γ值为1.3157±0.3572(1.0005～2.1882),λ值为1.0149±0.0226(1.0000～1.0796);在设定矩阵稀疏度条件下,γ值为1.3299±0.2330(1.0759～1.9574),λ值为1.0152±0.0190(1.0001～1.0679).两种方法均表明该网络具有小世界性.结论 正常老年人脑功能网络具有小世界性.%Objective To explore whether the functional network in brain of healthy elderly people possess small-world property using resting-state BOLD-fMRI. Methods A total of 20 healthy volunteers were enrolled in this study. Resting-state BOLD-fMRI data were collected and preprocessed with SPM 5 software. The whole brain were divided into 90 regions using anatomical automatic labeling template. The average time course of each region was extracted. The Pearson correlation coefficient of every pair regions was calculated, and the correlation matrix R (NXN, N=90) was generated by Fisher r to z transformation. Threshold was set by two methods: Direct defining the correlation value T and setting the sparsity of matrix. According to the definition of small-world property, when the formulas y=Cp/Crand>1, and λ= LP/Lrand≈ (C representing average clustering coefficient, L representing average shortest path, p representing regular networks and rand representing random networks) were met, it was considered that there was small-world property. Results When the threshold T was set directly, the value of y was 1. 3157±0. 3572 (1. 0005-2. 1882) and that of was 1. 0149±O. 0226 (1.0000-1.0796). While under the condition of setting the sparsity of matrix, the value of y was 1. 3299 ± 0. 2330 (1. 0759-1. 9574) and that of X was 1. 0152±0. 0190 (1. 0001-1. 0679). The small-world property was satisfied by both of the two methods. Conclusion The functional network in brain of healthy old people have the property of small-world-ness.
展开▼
