2型糖尿病脑病患者大脑功能磁共振网络小世界特性  

Small-world features of brain functional MRI network in patients with type 2 diabetic encephalopathy

作　　者：王晓艳[1] 王鹏程[1] 侯立霞[1] 刘珊珊[1] 宋莉[1] 徐龙春[2] 赵雷[2] 

机构地区：[1]泰山医学院放射学院,山东泰安271016 [2]泰山医学院附属医院放射科,山东泰安271016

出　　处：《中国医学物理学杂志》2016年第5期501-504,共4页Chinese Journal of Medical Physics

基　　金：山东省优秀中青年科学家科研奖励基金(BS2011DX038)

摘　　要：目的：以伴有脑病的2型糖尿病患者为研究对象,检验其大脑功能网络是否具有小世界特性,并比较与正常人的差异是否有统计学意义。方法：采集17例糖尿病脑病患者与15例健康志愿者被试的功能磁共振数据,采用gretna软件包和spm8在matlab2010b平台上对数据进行预处理。利用解剖学自动标记模板将全脑划分为90个脑区,每个脑区代表1个节点,计算两两节点之间的Pearson相关系数,通过Fisher-z变换将90×90的矩阵变成Z矩阵。以矩阵稀疏度为阈值,将Z值转换为二值矩阵,矩阵稀疏度阈值取0.05~0.40,以0.01为步长,计算共16个阈值点上的小世界参数。结果：在0.05~0.40的网络稀疏度阈值范围内,糖尿病脑病组与正常对照组的γ、λ、σ、最短路径长度均随阈值增大而减少,两组同时满足小世界网络特性。糖尿病脑病组与正常对照组相比,γ、σ值变大,λ变小,并且糖尿病脑病组最短路径长度变大,聚类系数变小,两组在整个阈值范围内的差异有统计学意义。结论：与正常对照组相比,糖尿病脑病患者脑网络虽然具有小世界特性,但与正常对照组相比发生了变化。Objective To observe whether the brain function network of the patient with type 2 diabetic encephalopathy has small-world features, and to observe the statistically significant differences by comparing the patients with normal persons.Methods The functional magnetic resonance imaging（fMRI） data of 17 patients with diabetic encephalopathy and 15 healthy volunteers were collected. The data were preprocessed with the gretna package and spm8 on matlab2010 b platform.The automatic labeling template was used to divide the whole brain into 90 brain regions, and each brain region represented a node. The Pearson correlation coefficients between two nodes were calculated and the 90 × 90 matrix was changed into Z matrix by using Fisher-z transformation. Matrix sparsity was taken as threshold to transform Z value into binary matrix. The small-world parameters of 16 threshold points were calculated with the sparsity of 0.05-0.40 and step of 0.01. Results For the network sparsity of 0.05-0.40, all the γ, λ, σ and the shortest path length（L_P） of diabetic encephalopathy group and control group decreased with the increasing of sparsity. Both the two groups were satisfied with small-world features.Compared with the controls, the diabetic encephalopathy patients had larger γ, σ, L_p and smaller λ, clustering coefficient.The differences in the whole range of threshold value were statistically significant. Conclusion The brain fMRI network of patients with diabetic encephalopathy has small-world features, with some differences with the controls.

关 键 词：糖尿病脑病 功能磁共振 大脑功能网络 小世界特性 

分 类 号：R445.2[医药卫生—影像医学与核医学] R749.1[医药卫生—诊断学]