Learning to Fuse and Reconstruct Multi-View Graphs for Diabetic Retinopathy Grading

2026-02-25

AI Summary

This work addresses the limitation of existing methods in multi-view fundus image fusion, which often neglect inter-view correlations and fail to effectively exploit cross-view consistency. To overcome this, we propose MVGFDR, an end-to-end multi-view graph fusion framework that explicitly decouples shared and view-specific features by constructing a multi-view graph. The framework enhances viewpoint-invariant representation learning through three key components: frequency-domain anchors derived from DCT coefficients, selective node fusion, and a cross-view mask reconstruction mechanism. Extensive experiments on MFIDDR—the largest multi-view fundus dataset to date—demonstrate that MVGFDR significantly outperforms current state-of-the-art approaches, achieving notable improvements in the accuracy of diabetic retinopathy grading.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Diabetic retinopathy (DR) is one of the leading causes of vision loss worldwide, making early and accurate DR grading critical for timely intervention. Recent clinical practices leverage multi-view fundus images for DR detection with a wide coverage of the field of view (FOV), motivating deep learning methods to explore the potential of multi-view learning for DR grading. However, existing methods often overlook the inter-view correlations when fusing multi-view fundus images, failing to fully exploit the inherent consistency across views originating from the same patient. In this work, we present MVGFDR, an end-to-end Multi-View Graph Fusion framework for DR grading. Different from existing methods that directly fuse visual features from multiple views, MVGFDR is equipped with a novel Multi-View Graph Fusion (MVGF) module to explicitly disentangle the shared and view-specific visual features. Specifically, MVGF comprises three key components: (1) Multi-view Graph Initialization, which constructs visual graphs via residual-guided connections and employs Discrete Cosine Transform (DCT) coefficients as frequency-domain anchors; (2) Multi-view Graph Fusion, which integrates selective nodes across multi-view graphs based on frequency-domain relevance to capture complementary view-specific information; and (3) Masked Cross-view Reconstruction, which leverages masked reconstruction of shared information across views to facilitate view-invariant representation learning. Extensive experimental results on MFIDDR, by far the largest multi-view fundus image dataset, demonstrate the superiority of our proposed approach over existing state-of-the-art approaches in diabetic retinopathy grading.

Problem

• Diabetic Retinopathy Grading
• Multi-View Fusion
• Inter-View Correlation
• Fundus Images
• View Consistency

Innovation

• Multi-View Graph Fusion
• Diabetic Retinopathy Grading
• View-Invariant Representation
• Masked Cross-view Reconstruction
• Frequency-Domain Anchors