TFFM: Topology-Aware Feature Fusion Module via Latent Graph Reasoning for Retinal Vessel Segmentation

2026-01-27

AI Summary

This work proposes a topology-aware segmentation framework to address the frequent topological discontinuities—such as gaps and fragmented segments—in existing retinal artery-vein segmentation methods, which hinder reliable graph-based clinical analysis despite high pixel-level accuracy. The approach integrates local features with global vascular structural dependencies through an implicit graph reasoning mechanism, leveraging a Graph Attention Network (GAT) and a Topology Feature Fusion Module (TFFM) to explicitly model vessel connectivity. A hybrid loss function combining soft clDice and Tversky loss is introduced to penalize topological breaks during training. Evaluated on the Fundus-AVSeg dataset, the method achieves a Dice score of 90.97% and a 95% Hausdorff distance of 3.50 pixels, reducing vessel fragmentation by approximately 38% compared to baseline models, thereby yielding anatomically coherent vascular trees suitable for automated biomarker quantification.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Graph-based Machine Learning
• Knowledge Representation and Reasoning: Qualitative Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Precise segmentation of retinal arteries and veins carries the diagnosis of systemic cardiovascular conditions. However, standard convolutional architectures often yield topologically disjointed segmentations, characterized by gaps and discontinuities that render reliable graph-based clinical analysis impossible despite high pixel-level accuracy. To address this, we introduce a topology-aware framework engineered to maintain vascular connectivity. Our architecture fuses a Topological Feature Fusion Module (TFFM) that maps local feature representations into a latent graph space, deploying Graph Attention Networks to capture global structural dependencies often missed by fixed receptive fields. Furthermore, we drive the learning process with a hybrid objective function, coupling Tversky loss for class imbalance with soft clDice loss to explicitly penalize topological disconnects. Evaluation on the Fundus-AVSeg dataset reveals state-of-the-art performance, achieving a combined Dice score of 90.97% and a 95% Hausdorff Distance of 3.50 pixels. Notably, our method decreases vessel fragmentation by approximately 38% relative to baselines, yielding topologically coherent vascular trees viable for automated biomarker quantification. We open-source our code at https://tffm-module.github.io/.

Problem

• retinal vessel segmentation
• topological discontinuity
• vascular connectivity
• graph-based analysis
• artery-vein segmentation

Innovation

• Topology-aware segmentation
• Graph Attention Network
• Feature fusion
• clDice loss
• Retinal vessel segmentation