电子学报2025，Vol.53Issue(3)：1000-1013,14.DOI:10.12263/DZXB.20240770
使用最近邻域聚合图神经网络的阿尔茨海默病分类方法
Method on Alzheimer's Disease Classification Utilizing Graph Neural Network with Nearest Neighborhood Aggregation
摘要
Abstract
Alzheimer's disease(AD)is a chronic neurodegenerative disease,and its accurate classification is advan-tage to achieve early diagnosis of AD so as to take timely treatment and intervention.In this paper,a novel method on AD Classification utilizing graph neural network with nearest neighborhood aggregation(GraphNAGE)is proposed.Firstly,the graph data modeling is performed to represent AD samples as graph data.By feature selection method based on mutual in-formation(MI),the high-importance volume features are selected from the 114 dimensional volume features of cerebral cor-tex and subcortical regions of interest(CCS-ROI)in the sample,and used for node modeling.Meanwhile,a relationship modeling method based on similarity measurement,modeling the relationships between samples using high importance vol-ume features,genetic genes,demographic information,and cognitive scores,is presented.Subsequently,the graph neural network with nearest neighborhood aggregation is constructed.For each node in the graph data,the nearest neighbor sam-pling is performed based on the weights of edge related to it.Then,the sampled data of neighboring nodes and central node are aggregated using the mean aggregation method.At last,a full-connected layer and a softmax layer are used to imple-ment AD classification.The proposed AD classification method is evaluated on the Alzheimer's disease prediction of longi-tudinal evolution(TADPOLE)dataset.The accuracy(ACC),F1 score and area under curve(AUC)of proposed AD classifi-cation method are 98.20%,97.34%and 97.80%,respectively.The experimental results show that the proposed AD classifi-cation method fully exploits the correlation between AD samples.Its performance is superior to conventional AD classifica-tion methods based on machine learning,deep learning and graph neural network.关键词
阿尔茨海默病(AD)/图神经网络(GNN)/节点建模/关系建模/相似性度量/最近邻域聚合Key words
Alzheimer's disease(AD)/graph neural network/node modeling/relationship modeling/similarity mea-surement/nearest neighborhood aggregation分类
电子信息工程引用本文复制引用
韩亮,刘媛,蒲秀娟,谈云帆,任青..使用最近邻域聚合图神经网络的阿尔茨海默病分类方法[J].电子学报,2025,53(3):1000-1013,14.基金项目
国家自然科学基金(No.62171066) National Natural Science Foundation of China(No.62171066) （No.62171066）
