Potential biomarkers for distinguishing people with Alzheimer's disease from cognitively intact elderly based on the rich-club hierarchical structure of white matter networks.

The aim of this study is to identify potential biomarkers that may distinguish people with Alzheimer's disease (AD) from cognitively intact elderly. We analyzed the features of rich-club hierarchical network between the AD and a control group by diffusion tensor imaging. We detected that the changes between the two groups were located mainly in the feeder and local connections. Then, we calculated the betweenness centrality of the rich nodes and the strength values of all feeder connections, and we chose the nodes and connections that showed the most significant differences as features. We found that 1) Feeder and local connections were changed in the AD group; 2) Rich nodes of the left putamen and precuneus had significant differences in betweenness centrality between the AD and control groups; 3) Three connections showed significant differences. The obtained features were fed into a linear discriminant analysis for classifying AD from cognitively intact elderly. The classification accuracy is superior to that of traditional biomarkers (hippocampal volume and clinical scores). Our results suggested that rich-club hierarchical network analysis is a viable tool for finding potential biomarkers. The obtained features can be applied as potential biomarkers for distinguishing AD patients from cognitively intact elderly.