Leptomeningeal Collateral Detection on DSA via Vessel-Graph Neural Networks

2026-06-12

AI Summary

This study addresses the limitations of current digital subtraction angiography (DSA)–based assessment of leptomeningeal collaterals (LMCs), which relies on subjective manual scoring and lacks objective, vessel-level granularity. To overcome this, the work proposes modeling LMC detection as a single-vessel classification task on DSA angiograms and introduces a novel graph–pixel hybrid neural network architecture. This framework integrates a topology-aware graph branch with a dense pixel branch within a shared node probability space, enabling automatic identification and quantification of individual LMCs. Evaluated via five-fold cross-validation, the method achieves a PR-AUC of 0.434, significantly outperforming pure graph-based (0.403) and pure pixel-based (0.362) models, thereby advancing collateral assessment from subjective scoring toward objective, precise vessel-level analysis.

Abstract

Leptomeningeal collaterals (LMCs) are an important prognostic factor in acute ischemic stroke. Existing automated methods rely on CT angiography (CTA), but individual LMCs are often too small to be resolved on CTA, limiting these methods to coarse collateral scoring. Digital subtraction angiography (DSA) visualizes individual collaterals at superior resolution, yet current assessment remains subjective, relying on manual grading scales that suffer from poor inter-rater agreement. We present a framework that formulates collateral detection as the classification of individual vessel segments on a graph derived from DSA. A hybrid graph-pixel architecture combines a topology-aware graph branch with a dense pixel branch, fused in a shared node-probability space. In a five-fold cross-validation setting, the fused model achieves a PR-AUC of 0.434, outperforming the graph-only (0.403) and pixel-only (0.362) baselines. To our knowledge, this is the first method to enable the individualization of LMCs in DSA, allowing for precise per-vessel quantitative assessment. This integration shifts DSA assessment toward objective evaluation, supporting future biomarker and pattern discovery for individual LMCs.

Problem

• Leptomeningeal collaterals
• Digital subtraction angiography
• Acute ischemic stroke
• Collateral assessment
• Vessel segmentation

Innovation

• vessel-graph neural networks
• leptomeningeal collaterals
• digital subtraction angiography
• graph-pixel fusion
• individual vessel classification