8th Workshop on GRaphs in biomedicAl Image anaLysis

• 20.03.2026: Excellent news - our proposal for GRAIL 2026 was accepted as an in-person workshop at MICCAI 2026. Stay tuned for further announcements.
• 10.04.2026: The GRAIL 2026 website is now live. Paper submission, program details, keynote announcements, and exact workshop logistics will be updated here as they become available.

GRAIL 2026 is the eighth international Workshop on GRaphs in biomedicAl Image anaLysis, organized as an in-person satellite event of MICCAI 2026. Building on the successful joint format of last year, GRAIL continues as a unified forum for graph-based, higher-order, and topology-informed methods in biomedical image analysis and related domains.

Graphs provide a flexible and scalable mathematical framework for modeling complex, unstructured data, objects, and their interactions in an interpretable and mathematically grounded way. They underpin a wide range of methods such as spectral analysis, dimensionality reduction, and network analysis. Since 2017, geometric deep learning has tightly integrated graph signal processing with deep neural architectures, leading to rapid progress across medical imaging, shape analysis, brain connectomics, population modeling, patient multi-omics, and drug discovery. In parallel, graph-based learning has become a major research direction at leading machine learning and computer vision venues such as CVPR, ICLR, NeurIPS, and the Learning on Graphs (LoG) conference.

Since its inception in 2017, GRAIL has served as a focused venue within MICCAI for connecting methodological advances in graph learning with clinically and biologically relevant applications. In 2025, GRAIL broadened its scope through a collaborative format with Topology-Guided Imaging (TGI) and Hypergraphs in MedIA (HGMIA), reflecting strong community interest in a shared forum covering graphs, higher-order structures, and topology-inspired methods. For 2026, GRAIL continues this inclusive direction, with TGI co-organizers from last year joining as co-chairs.

GRAIL 2026 is designed as a compact, half-day workshop featuring invited keynote talks and peer-reviewed paper presentations. The workshop emphasizes scientific depth, methodological rigor, and strong application grounding, while fostering exchange between complementary perspectives in graph learning and topological methods. We also plan to video-record the sessions and make them publicly available after the conference.

Scientific Tracks

Track 1: Graphs and Applications

This track focuses on graph representations and learning methods based on pairwise relationships, as well as their applications to biomedical image analysis and related data modalities.

• Graph analytics and machine/deep learning on graphs
• Graph neural networks (GNNs) for biomedical image and data analysis
• Signal processing on graphs, including non-learning-based approaches
• Probabilistic graphical models for biomedical data
• Graph generative models
• Graph foundation models and integration with non-graph foundation models
• Graph datasets, benchmarks, and evaluation methodologies
• Learning on small or limited biomedical datasets
• Statistical testing and group-level analysis on graph structures
• Explainable AI (XAI) for graph-based and geometric deep learning
• Inductive biases, symmetry, and equivariance in graph-based models
• Combinations with other paradigms such as self-supervised or federated learning

Track 2: Higher-Order Topologies and Applications

This track highlights methods that go beyond pairwise graph structures by incorporating higher-order relationships and topological priors.

• Hypergraphs, multiview graphs, multiplex graphs, and PolyConnect structures
• Topological deep learning (TDL) and topological signal processing
• Persistent homology and topology-aware learning methods
• Higher-order representations for biomedical images and multimodal data
• Integration of topology-based methods with deep learning architectures
• Theoretical foundations and practical applications of higher-order relational learning in medicine

Applications covered include but are not limited to:

• Image segmentation, registration, classification
• Graph representations in pathology imaging and whole-slide image analysis
• Graph-based approaches for intra-operative surgical support
• Graph-based shape modeling and dimensionality reduction
• Graphs for large-scale patient population analyses
• Combining multimodal/multi-omics data through graph structures
• Graph analysis of brain networks and connectomics

Through this structure, GRAIL 2026 aims to continue serving as a unifying forum for graph-based and topology-informed research in biomedical image analysis, reflecting both the evolution of the field and the interests of the broader MICCAI community.

As GRAIL 2026 will be an in-person workshop, registration will happen via the main conference website. Please visit the MICCAI 2026 Registration page for details.

The detailed GRAIL 2026 workshop agenda is tbd.

GRAIL 2026 will be organized as a half-day in-person workshop at MICCAI 2026 and will feature:

• Invited keynote talks
• Peer-reviewed paper presentations
• Discussion and networking opportunities

Workshop date: tbd
Conference dates: October 4-8, 2026
Workshop event days at MICCAI: October 4 and October 8, 2026
Venue: ADNEC Centre, Abu Dhabi, United Arab Emirates
Room: tbd

The keynote speaker lineup for GRAIL 2026 is tbd.

We will announce invited speakers here as soon as confirmations are finalized.

GRAIL 2026 paper submission will be handled through the OpenReview platform. Submission details, links, and author instructions are currently tbd. Interested authors are encouraged to stay tuned for further announcements.

We expect to accept papers in the following format:

• Complete papers: papers describing original research with MICCAI workshop length (max 8 pages text + max 2 pages references, supplemental material allowed). Submissions should be anonymous and formatted following the LNCS Style. All complete papers will be peer-reviewed through a double-blind review process.

Further details on tracks, submission instructions, review process, and proceedings will be announced soon.