Motivation: With the recent approval of Alzheimer's drugs, predicting which patients will convert to AD has become more important. The AD drugs may be administered early to the converters to delay the onset of Alzheimer's disease. Goal(s): To predict which patients with MCI will convert to AD using diffusion imaging with a multi-compartment diffusion model. Approach: We used a longitudinal image database of MCI patients and derived diffusion features from anatomical regions segmented in the native space of patients. A Cox regression model was used to determine the hazard ratio. Results: A multi-shell RSI diffusion model can predict MCI-to-AD conversion. Impact: Diffusion imaging derived biomarkers that can predict MCI to AD conversion will be useful to stratify patients to start recently approved Alzheimer's drugs early in their disease progression. This may improve better patient outcome.
Uluğ et al. (Tue,) studied this question.