An Automated Framework for Large-Scale Graph-Based Cerebrovascular Analysis

TL;DR

The paper presents CaravelMetrics, an automated framework that constructs skeleton-based vascular graphs from binary masks to enable multiscale cerebrovascular analysis using atlas-based regional parcellation. It extracts 15 morphometric, topological, fractal, and geometric features, applied to 570 IXI subjects with TOF-MRA to reveal aging-related declines in vessel length, volume, and bifurcation density, along with increased tortuosity and education-related vessel length increases. The approach demonstrates population-scale applicability, modality-agnostic input, and the ability to perform whole-brain and regional analyses, contributing to normative modeling and vascular aging studies. While robust to moderate segmentation noise, the work highlights site effects and suggests future work on pathological cohorts and robustness assessments to enhance clinical applicability.

Abstract

We present CaravelMetrics, a computational framework for automated cerebrovascular analysis that models vessel morphology through skeletonization-derived graph representations. The framework integrates atlas-based regional parcellation, centerline extraction, and graph construction to compute fifteen morphometric, topological, fractal, and geometric features. The features can be estimated globally from the complete vascular network or regionally within arterial territories, enabling multiscale characterization of cerebrovascular organization. Applied to 570 3D TOF-MRA scans from the IXI dataset (ages 20-86), CaravelMetrics yields reproducible vessel graphs capturing age- and sex-related variations and education-associated increases in vascular complexity, consistent with findings reported in the literature. The framework provides a scalable and fully automated approach for quantitative cerebrovascular feature extraction, supporting normative modeling and population-level studies of vascular health and aging.

Figures (6)

• Figure 1: Overview of the vascular feature extraction framework.
• Figure 2: Scanner-induced systematic bias. Consistent IOP shift across all features demonstrates acquisition artifacts.
• Figure 3: Age effects on vessel morphometry. Significant age correlations (red bars) after FDR correction.
• Figure 4: Age-related vessel changes. Group differences across multiple morphometric and topological features.
• Figure 5: Region-specific age effects. Vessel length (left) and curvature (right) correlations across 30 brain arterial regions.