Cerebral Small Vessel Disease (cSVD) is the main cause of Vascular Cognitive Impairment and Dementia (VCID). Existing MRI modalities primarily detect parenchymal lesions of cSVD, which reflect downstream changes rather than directly visualizing small cerebral blood vessels. Here, we developed and applied a novel deep learning (DL) based pipeline to segment and quantify small cerebral blood vessels in a cohort of elderly persons at risk of cSVD, based on high-resolution black blood (BB) MRI. By correlating vessel density with cognitive scores, we demonstrate this method’s potential as a noninvasive
biomarker for VCID.