A Graph-Driven Hybrid CNN-GNN Approach for Vessel Classification in Adaptive Optics Images

Key Points

• The hybrid cnn-gnn strategy significantly improves vessel classification accuracy in adaptive optics images.
• Using a graph-driven method enhances the understanding of complex vessel structures within 12 adaptive optics images.
• Analysis focused on developing a novel classification approach based on convolutional and graph neural networks.
• Implies potential for improved medical imaging techniques; further testing in diverse datasets is necessary.