Abstract:Alzheimer's disease is a common progressive neurodegenerative disorder in the elderly. Diffusion tensor imaging (DTI) has been widely used to explore structural integrity and to describe white matter degeneration in Alzheimer's disease. Previous research has indicated that the change of connections between white matter fiber tracts is very important for investigating the brain function of Alzheimer's disease patients. However, whether white matter features can be used as potential biomarkers for predicting Alzheimer's disease tendency requires more in-depth research. In this study, we investigated the relationship between the damage in white matter tracts and the decline of cognitive function in Alzheimer's disease. DTI data were collected from 38 Alzheimer's disease patients and 30 normal controls. Fiber assignment by continuous tracking approach was used to establish connections between different brain regions of the whole brain, network-based statistical analysis and support vector machine classification analysis were used to explore the connection of whole-brain fiber bundles between the two groups. Most importantly, our results showed that the connections between brain regions of Alzheimer's disease patients were damaged, and the damage were mainly located in the right hemisphere, there was a certain degree of lateralization effect. Using whole-brain fiber bundle connection network as a feature for classification, we found it helped to improve the classification accuracy in Alzheimer's disease patients, which is useful for early clinical diagnosis of Alzheimer's disease. These findings further suggested that we can use the whole-brain fiber bundle connection network of Alzheimer's disease patients as a potential diagnostic indicator of Alzheimer's disease in the future.

Changes in the connection network of whole-brain fiber tracts in patients with Alzheimer's disease have a tendency of lateralization