Impairment of anatomical brain network in Alzheimer's disease: A study based on diffusion tensor tractography

Abstract

Recent studies on Alzheimer's disease (AD) have shown disrupted topological organizations of both functional and anatomical brain networks in AD patients, suggesting that AD might be a disease related to disconnection between brain regions. White matter (WM)-based anatomical network can be weighted by fractal anisotropy (FA), which reflects the integrity of axons and myelin sheaths. Thus, study on the WM-based brain network from the perspectives of topological organization and pair-wise edges might provide information about global organization alteration and WM integrity impairment in AD. In the current work, in addition to the topological characteristics of brain networks, we further investigated the disconnection of edges in anatomical brain networks of AD patients via diffusion tensor tractography. The results showed that both the brain networks of AD patients and normal controls (NCs) exhibited small-world property and such small-world property tended to decline with cognitive impairment. Some network metrics (e.g., clustering coefficient, and characteristic path length and mean FA of all edges) were found to be correlated with the cognitive performance. More importantly, compared to NCs, AD patients had decreased FA weight of edges between certain cortical regions, indicating network disconnection and damage of WM integrity. In summary, our results provided further support for the existence of small-world property in anatomical brain networks of AD patients, and revealed reduced edge weights in AD patients' anatomical brain networks.