Improve Retinal Artery/Vein Classification via Channel Couplin

TL;DR

This paper introduces a novel deep learning loss function that enforces consistency between vessel, artery, and vein predictions, improving retinal artery/vein classification accuracy.

Contribution

It proposes the Channel-Coupled Vessel Consistency Loss and intra-image pixel-level contrastive loss to enhance A/V classification by leveraging anatomical relationships.

Findings

Achieved state-of-the-art results on RITE, LES-AV, and HRF datasets.

Demonstrated improved prediction coherence and accuracy.

Validated effectiveness of proposed losses in retinal vessel classification.

Abstract

Retinal vessel segmentation plays a vital role in analyzing fundus images for the diagnosis of systemic and ocular diseases. Building on this, classifying segmented vessels into arteries and veins (A/V) further enables the extraction of clinically relevant features such as vessel width, diameter and tortuosity, which are essential for detecting conditions like diabetic and hypertensive retinopathy. However, manual segmentation and classification are time-consuming, costly and inconsistent. With the advancement of Convolutional Neural Networks, several automated methods have been proposed to address this challenge, but there are still some issues. For example, the existing methods all treat artery, vein and overall vessel segmentation as three separate binary tasks, neglecting the intrinsic coupling relationships between these anatomical structures. Considering artery and vein structures are subsets of the overall retinal vessel map and should naturally exhibit prediction consistency with it, we design a novel loss named Channel-Coupled Vessel Consistency Loss to enforce the coherence and consistency between vessel, artery and vein predictions, avoiding biasing the network toward three simple binary segmentation tasks. Moreover, we also introduce a regularization term named intra-image pixel-level contrastive loss to extract more discriminative feature-level fine-grained representations for accurate retinal A/V classification. SOTA results have been achieved across three public A/V classification datasets including RITE, LES-AV and HRF. Our code will be available upon acceptance.