Neovascularization Segmentation via a Multilateral Interaction-Enhanced Graph Convolutional Network

2025-08-05

AI Summary

Choroidal neovascularization (CNV) segmentation in optical coherence tomography angiography (OCTA) images faces challenges including irregular morphology, projection artifacts, ambiguous boundaries, and limited annotated data. To address these, this work introduces CNVSeg—the first publicly available OCTA dataset with expert-annotated CNV masks—and proposes MTG-Net, a multi-task graph convolutional network. MTG-Net incorporates two novel modules: Multi-Edge Interactive Graph Reasoning (MIGR) and Multi-Edge Reinforced Graph Reasoning (MRGR), which jointly model geometric structures and semantic relationships between CNV lesions and vasculature. Additionally, an uncertainty-weighted loss function is designed to mitigate imaging noise. Extensive experiments demonstrate that MTG-Net achieves Dice scores of 87.21% for CNV segmentation and 88.12% for vessel segmentation—outperforming state-of-the-art methods. This framework provides a robust technical foundation for quantitative assessment of wet age-related macular degeneration (AMD).

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Web data integration and cleaning
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Choroidal neovascularization (CNV), a primary characteristic of wet age-related macular degeneration (wet AMD), represents a leading cause of blindness worldwide. In clinical practice, optical coherence tomography angiography (OCTA) is commonly used for studying CNV-related pathological changes, due to its micron-level resolution and non-invasive nature. Thus, accurate segmentation of CNV regions and vessels in OCTA images is crucial for clinical assessment of wet AMD. However, challenges existed due to irregular CNV shapes and imaging limitations like projection artifacts, noises and boundary blurring. Moreover, the lack of publicly available datasets constraints the CNV analysis. To address these challenges, this paper constructs the first publicly accessible CNV dataset (CNVSeg), and proposes a novel multilateral graph convolutional interaction-enhanced CNV segmentation network (MTG-Net). This network integrates both region and vessel morphological information, exploring semantic and geometric duality constraints within the graph domain. Specifically, MTG-Net consists of a multi-task framework and two graph-based cross-task modules: Multilateral Interaction Graph Reasoning (MIGR) and Multilateral Reinforcement Graph Reasoning (MRGR). The multi-task framework encodes rich geometric features of lesion shapes and surfaces, decoupling the image into three task-specific feature maps. MIGR and MRGR iteratively reason about higher-order relationships across tasks through a graph mechanism, enabling complementary optimization for task-specific objectives. Additionally, an uncertainty-weighted loss is proposed to mitigate the impact of artifacts and noise on segmentation accuracy. Experimental results demonstrate that MTG-Net outperforms existing methods, achieving a Dice socre of 87.21% for region segmentation and 88.12% for vessel segmentation.

Problem

• Accurate segmentation of CNV regions and vessels in OCTA images
• Address irregular CNV shapes and imaging limitations like artifacts
• Lack of publicly available datasets for CNV analysis

Innovation

• Multilateral graph convolutional interaction-enhanced network
• Multi-task framework with graph-based modules
• Uncertainty-weighted loss for noise mitigation