Diffusion weighted imaging-based maximum density path analysis and classification of Alzheimer's disease.

Area of Science:

• Neuroimaging
• Neurology
• Medical Physics

Background:

• Characterizing brain changes in Alzheimer's disease (AD) is crucial for prognosis and clinical trial assessment.
• Diffusion weighted imaging (DWI) is a key tool for investigating white matter integrity.

Purpose of the Study:

• To investigate white matter integrity in elderly controls (CTL), mild cognitive impairment (MCI), and AD patients using a new fiber-tract modeling method.
• To assess the utility of diffusion tensor imaging (DTI) measures along maximum density paths (MDPs) for classifying AD.

Main Methods:

• Employed a novel fiber-tract modeling method involving clustering tractography and shortest path graph search to derive MDPs.
• Calculated fractional anisotropy (FA) and mean diffusivity (MD) along MDPs in 50 CTL, 113 MCI, and 37 AD patients.
• Utilized support vector machine (SVM) classification with DTI measures along MDPs as features.

Main Results:

• Found significant differences in MD and FA between AD patients and CTL subjects.
• Observed significant MD differences between CTL and late MCI subjects.
• Demonstrated associations between MD and FA with clinical scores.

Conclusions:

• The novel MDP method effectively captures white matter organization changes in AD.
• DTI measures along MDPs show potential as biomarkers for AD classification and monitoring disease progression.