Application of Diffusion Tensor Imaging Based on Automatic Fiber Quantification in Alzheimer's Disease

Area of Science:

• Neuroimaging
• Neuroscience
• Biomarkers

Background:

• Alzheimer's disease (AD) significantly impacts white matter microstructure.
• The specific regional alterations and their potential as biomarkers for distinguishing AD from normal controls (NC) remain unclear.

Purpose of the Study:

• To investigate regional white matter microstructural differences between AD patients and NC.
• To evaluate the diagnostic value of these alterations as potential biomarkers for AD detection.

Main Methods:

• Utilized automatic fiber quantification (AFQ) on 33 AD patients and 25 NC.
• Assessed fractional anisotropy (FA), mean diffusivity (MD), volume, and curvature across 20 segmented fiber bundles.
• Employed mRMR and LASSO algorithms for feature selection and logistic regression with ROC curves to assess predictive power.

Main Results:

• Significantly increased MD values were observed in AD patients compared to NC.
• These differences were predominantly located in the left cingulum hippocampus (HCC), left uncinate fasciculus (UF), and superior longitudinal fasciculus (SLF).
• Mean diffusivity (MD) index demonstrated the highest performance in distinguishing between NC and AD.

Conclusions:

• Abnormal white matter microstructure, identified through DTI-based AFQ analysis, aids in understanding AD pathogenesis.
• These findings highlight the potential of DTI-based white matter analysis as a tool for exploring AD pathogenesis and diagnosis.