MICCAI 2026 Tutorial

4th Graph Learning in Medical Image Analysis (GraphMedIA)

Conference: MICCAI 2026
Venue: Abu Dhabi, United Arab Emirates
Tutorial Page: GraphMedIA 2026
Conference Website: MICCAI 2026

We are pleased to announce the 4th GraphMedIA tutorial, Graph Learning in Medical Image Analysis, which will be held in conjunction with MICCAI 2026. This half-day tutorial provides a structured overview of graph learning for medical image analysis, covering classical semi-supervised graph methods, graph neural networks, graph transformers, hypergraph learning, graph foundation models, and graph agentic systems.

The 2026 edition introduces a dedicated Mini Interactive Hands-On Session and Live Demonstration. I will lead this session, focusing on practical graph and hypergraph learning pipelines for medical imaging. The session includes constructing graph and hypergraph representations from medical imaging data, implementing hypergraph neural networks, training and evaluating models on real-world medical tasks, and exploring structural modelling, message passing, higher-order interactions, scalability, and practical implementation considerations.

The hands-on materials will use state-of-the-art open-source hypergraph learning frameworks, including THU-HyperG and DeepHypergraph. Pre-prepared Jupyter/Colab notebooks will be provided to support accessibility and reproducibility.