You should be logged in to see the presentation
The dynamic functional connectivity (FC) has been used successfully to investigate the dysfunction of Alzheimer's disease (AD) patients. However, the reconfiguration intensity of nodal FC at different time scales received little attention. In this paper, we performed a feature, time-distance nodal connectivity diversity (tdNCD), to analyze the network reconfiguration intensity in every specific region of interest (ROI) using a large multicenter dataset (N=809). Meta-analysis technique was used to reveal the difference of the tdNCD between ADs and normal controls (NCs) from 7 sites. Besides, to explore the distinguishability of the tdNCD in the NC and AD groups, a machine learning model based on a fully connected neural network (FC-net) was created. The tdNCD value of AD is significantly higher in some ROIs of default mode network and lower in some regions of subcortical and cerebellum network than that in NC group (p < 0.05, Bonferroni corrected). The classification results showed that the F1-score=81.8% when using tdNCD and FC to classify AD from NC with a leave-one-site-out cross-validation, which is significantly higher than that only using FC as features. Collectively, we demonstrated that dynamic reconfiguration of nodal FC exists abnormity in AD. And tdNCD highlights the potential for further understanding core mechanisms of AD's dysfunction.
Chinese Academy of Sciences, China
There are still no questions. Be the first to post one!
You should be logged in order to use this feature.