Cerebrovascular brain-age

Abstract

The structural brain-age makes predictions based on changes in tissue integrity. Adding cerebrovascular MRI biomarkers may add sensitivity to physiological and metabolic changes, hence complementing structural brain-age, and possibly improving its early pathology sensitivity. Baseline and follow-up T1w, FLAIR, and ASL data of 233 healthy participants and combinations of features and algorithms were used to predict ‘Cerebrovascular brain-age’. The ExtraTrees algorithm utilising T1w, ASL, and FLAIR features performed best and showed good longitudinal reproducibility.