Quantifying Cerebral Small Vessel Geometry with Self-Supervised Shape-Aware 3D U-Net on 7T TOF MRI: Associations with Age and Cognitive Function

Motivation: Aging and cerebral vascular disease can both contribute to structural alterations of intracranial vessels. Goal(s): Our study aims to evaluate the role of vessel geometry of cerebral small vessels in an aged cohort with aging, cognitive function, and vascular risk factors Approach: Vessel density, mean diameter and length were automatically extracted from high resolution TOF using shape-aware 3D segmentation with local contrast guided attention Results: All derived vascular matrics are associated with cognitive function and highly correlated with body mass index (BMI). Vessel mean diameter is associated with pulse pressure and hyperlipidemia. Vessel density and mean length are related with reduced GM perfusion. Impact: Quantitative morphological measures of cerebral small vessels can be effectively segmented using deep learning, which can be used as sensitive markers of aging, cognitive dysfunction, and brain vascular function