Investigating the Diagnostic Utility of Multi-Shell Diffusion MRI and CSF Biomarkers for Mild Cognitive Impairment Classification

Motivation: While recent advances in quantitative MRI have resulted in a myriad of methods capable of examining cerebral tissue microstructure, the pre-dementia classification performances of these models have not been extensively studied. Goal(s): To evaluate diffusion MRI biomarkers as a noninvasive diagnostic tool for early detection of Alzheimer's disease. Approach: We utilize multi-shell diffusion MRI from ADNI to investigate the comparative effectiveness of diffusion MRI models (MAP, SMI, NODDI, and C-NODDI) versus established CSF biomarkers in classifying mild cognitive impairment. Results: Multi-shell diffusion biomarkers match CSF biomarker performance in MCI classification. Greater performance can be achieved when using both MRI and CSF biomarkers simultaneously. Impact: This research highlights diffusion MRI's potential for classifying prodromal Alzheimer's disease, paving the way for non-invasive pre-dementia diagnostics. It additionally provides MRI researchers insights into the comparative classification performance of various diffusion MRI models.