A comprehensive normative atlas of cerebrovascular morphology and its associations with stroke prognosis

Brain vasculature is an essential structural component of the brain and plays an important role in maintaining neural function. Its morphological alterations are increasingly implicated in physiological aging and neurological diseases. However, due to the lack of automated, reproducible tools for morphological analysis, quantitative characterization of brain vasculature and its pathological changes at a population level remains a major challenge. Here, we present MARVAL (Morphological Analysis of ceRebroVAscuLature), an integrated platform for comprehensive cerebrovascular morphometry comprising a high-quality annotated time-of-flight magnetic resonance angiography (TOF-MRA) dataset, a generalizable automated segmentation pipeline with enhanced recall in small vessels, and a comprehensive toolkit for deep phenotyping and group analysis. Leveraging MARVAL in a large community-based healthy cohort, we established a high-resolution standard template of the brain vasculature in East Asians, revealing key normative characteristics of vascular aging. We also systematically demonstrated, for the first time, the modulating effects of non-imaging phenotypes on brain vascular morphology. Furthermore, in a stroke cohort, we identified novel morphological biomarkers associated with stroke prognosis. Together, our findings provide a foundational reference for cerebrovascular health and establish vascular morphometry as a framework to advance precision neurology.