VAOT: Vessel-Aware Optimal Transport for Retinal Fundus Enhancement

TL;DR

VAOT is a novel vessel-aware optimal transport framework that enhances retinal fundus images by preserving vascular structures, outperforming existing methods in image quality and clinical relevance.

Contribution

This work introduces VAOT, a new unpaired image enhancement method that maintains vascular integrity using optimal transport and structure-preserving regularizers.

Findings

Outperforms state-of-the-art baselines in synthetic degradation benchmarks.

Improves vessel and lesion segmentation accuracy.

Reduces noise while preserving vessel topology.

Abstract

Color fundus photography (CFP) is central to diagnosing and monitoring retinal disease, yet its acquisition variability (e.g., illumination changes) often degrades image quality, which motivates robust enhancement methods. Unpaired enhancement pipelines are typically GAN-based, however, they can distort clinically critical vasculature, altering vessel topology and endpoint integrity. Motivated by these structural alterations, we propose Vessel-Aware Optimal Transport (\textbf{VAOT}), a framework that combines an optimal-transport objective with two structure-preserving regularizers: (i) a skeleton-based loss to maintain global vascular connectivity and (ii) an endpoint-aware loss to stabilize local termini. These constraints guide learning in the unpaired setting, reducing noise while preserving vessel structure. Experimental results on synthetic degradation benchmark and downstream evaluations in vessel and lesion segmentation demonstrate the superiority of the proposed methods against several state-of-the art baselines. The code is available at https://github.com/Retinal-Research/VAOT