基于网络拓扑特性的MCI分类

发布时间：2019-02-14 08:14

【摘要】：阿尔茨海默氏病(Alzheimer′s disease,AD)和轻度认知障碍(Mild cognitive impairment,MCI)是经常发生在老年人中的脑疾病,其主要表现为认知和智力的障碍。机器学习和模式识别方法已经被应用到对AD和MCI的诊断和分类中。最近,研究人员提出利用大脑连接网络实现对疾病的诊断和分类。大部分的研究主要集中在网络中提取一些局部的特性(如聚类系数),并利用机器学习的方法(如支持向量机)来实现对疾病的分类。然而,存在的研究表明AD以及MCI是和一个大规模的脑连接网络相关,而不仅是大脑的若干区域。因此,本文提出一种新的基于网络整体拓扑结构信息的分类方法,并将其用于对MCI疾病的分类。实验结果表明,本文的方法能够对分类结果有重要的改进。
[Abstract]:Alzheimer's disease (Alzheimer's disease,AD) and mild cognitive impairment (Mild cognitive impairment,MCI) are common brain disorders in the elderly, which are mainly characterized by cognitive and intellectual impairment. Machine learning and pattern recognition methods have been applied to the diagnosis and classification of AD and MCI. Recently, researchers proposed the use of brain connectivity networks for the diagnosis and classification of diseases. Most of the researches focus on extracting some local characteristics (such as clustering coefficients) from the network and using machine learning (such as support vector machine) to classify diseases. However, existing studies suggest that AD and MCI are associated with a large network of brain connections, not just several regions of the brain. Therefore, this paper proposes a new classification method based on the network topology information, and applies it to the classification of MCI diseases. The experimental results show that the proposed method can improve the classification results.
【作者单位】：南京航空航天大学计算机科学与技术学院;
【基金】：江苏省杰出青年基金(SBK201310351)资助项目高等教育博士学科点专项科研基金(20123218110009)资助项目南京航空航天大学基本科研业务费专项资金(NE2013105)资助项目安徽省高校省级自然科学研究(KJ2013Z095)资助项目
【分类号】：TP393.0;R318.0

【参考文献】