基于最小生成树的MCI脑网络分类  

Classification of MCI brain network based on minimum spanning tree

作　　者：苗丽雯 田程 李婷 李佩珍 王彬[1] 曹锐[1] MIAO Li-wen;TIAN Cheng;LI Ting;LI Pei-zhen;WANG Bin;CAO Rui(Department of Computer Science and Technology,Taiyuan University of Technology,Taiyuan 030024,China)

机构地区：[1]太原理工大学计算机科学与技术学院,山西太原030024

出　　处：《计算机工程与设计》2018年第2期585-589,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(61373101;61503272);虚拟现实技术与系统国家重点实验室开放课题基金项目(BUAA-VR-15KF-16);山西省工业攻关基金项目(20140321002-01);山西省青年科学基金项目(2015021090;201601D202042);山西省回国留学人员科研基金项目(2016-037)

摘　　要：为更好地将脑网络的拓扑属性应用于轻度认知障碍的分类研究中,提出利用最小生成树构造无偏差脑网络,通过其拓扑属性准确刻画网络之间的差异,避免传统网络中连接强度带来的影响。分别提取早期轻度认知障碍、晚期轻度认知障碍和正常老年人这3组被试的无权网络和最小生成树的拓扑属性作为分类特征,使用支持向量机进行分类研究。实验结果表明,基于最小生成树的分类方法比传统无权网络具有更好地分类效果,表明最小生成树能更准确度量脑网络的结构变化,可以用于阿尔兹海默病的早期辅助诊断。To improve the classification performance of mild cognitive impairment(MCI)based on brain network topological,the minimum spanning tree(MST)method was proposed to construct unbiased brain network,which accurately described the diffe-rence between networks through its topological property,and avoided the influence of connection strength in the traditional network.The classification characteristics of early-MCI,late-MCI and normal control in the un-weighted brain network and MST were extracted and classified using support vector machine(SVM).Experimental results show that the classification method based on the MST has better effects than the traditional un-weighted network,which indicated that the MST can measure the structural changes of the brain network more accurately,which can be used in the early diagnosis of Alzheimer’s disease.

关 键 词：最小生成树 轻度认知障碍 阿尔兹海默病 支持向量机 脑网络 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]