Cross-Domain Vessel Segmentation via Latent Similarity Mining and Iterative Co-Optimization

arXiv cs.CV / 4/3/2026

💬 OpinionIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper targets retinal vessel segmentation, where performance drops sharply under domain shift between training and test data.
It introduces a domain-transfer framework that uses latent vascular similarity mining across domains, based on pretrained conditional diffusion models.
A deterministic inversion step creates intermediate, domain-agnostic latent prototypes that guide target-domain image synthesis.
The method iteratively co-optimizes a generation network and a segmentation network via cyclic parameter updates to jointly improve synthesis quality and segmentation accuracy.
Experiments on cross-domain clinical setups with modality discrepancies show state-of-the-art results, especially in difficult scenarios.

Abstract

Retinal vessel segmentation serves as a critical prerequisite for automated diagnosis of retinal pathologies. While recent advances in Convolutional Neural Networks (CNNs) have demonstrated promising performance in this task, significant performance degradation occurs when domain shifts exist between training and testing data. To address these limitations, we propose a novel domain transfer framework that leverages latent vascular similarity across domains and iterative co-optimization of generation and segmentation networks. Specifically, we first pre-train generation networks for source and target domains. Subsequently, the pretrained source-domain conditional diffusion model performs deterministic inversion to establish intermediate latent representations of vascular images, creating domain-agnostic prototypes for target synthesis. Finally, we develop an iterative refinement strategy where segmentation network and generative model undergo mutual optimization through cyclic parameter updating. This co-evolution process enables simultaneous enhancement of cross-domain image synthesis quality and segmentation accuracy. Experiments demonstrate that our framework achieves state-of-the-art performance in cross-domain retinal vessel segmentation, particularly in challenging clinical scenarios with significant modality discrepancies.