Full papers
Complete papers describing original research, with a maximum of 8 pages of text plus 2 pages of references; supplemental material allowed.
Computer Vision/CORE Unranked
GRAIL 2026
8th Workshop on GRaphs in biomedicAl Image anaLysis
GRAIL 2026 is the eighth international workshop on Graphs in Biomedical Image Analysis, held as an in-person satellite event of MICCAI 2026 in Strasbourg, France. It serves as a unified forum for graph-based, higher-order, and topology-informed methods in biomedical image analysis, integrating research from graph neural networks, topological deep learning, and their applications in medical imaging, connectomics, pathology, and multi-omics data.
Key deadlines
Verified September 9, 2026Abstract registration
June 29, 2026
AoE
Full paper
June 29, 2026
AoE
Workshop timeline
Submission and decisions
Abstract registrationKey deadline
June 29, 2026 · AoE
Full paperKey deadline
June 29, 2026 · AoE
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Research areas in scope
01
Graphs and Applications
Graph analytics and machine/deep learning on graphsGraph neural networks (GNNs) for biomedical image and data analysisSignal processing on graphs, including non-learning-based approachesProbabilistic graphical models for biomedical dataGraph generative modelsGraph foundation models and integration with non-graph foundation modelsGraph datasets, benchmarks, and evaluation methodologiesLearning on small or limited biomedical datasetsStatistical testing and group-level analysis on graph structuresExplainable AI (XAI) for graph-based and geometric deep learningInductive biases, symmetry, and equivariance in graph-based modelsCombinations with other paradigms such as self-supervised or federated learning
02
Higher-Order Topologies and Applications
Hypergraphs, multiview graphs, multiplex graphs, and PolyConnect structuresTopological deep learning (TDL) and topological signal processingPersistent homology and topology-aware learning methodsHigher-order representations for biomedical images and multimodal dataIntegration of topology-based methods with deep learning architecturesTheoretical foundations and practical applications of higher-order relational learning in medicine
Policies worth checking twice
- Submissions were handled via OpenReview with mandatory author accounts and conflict-of-interest disclosure
- All submissions underwent double-blind peer review
- Papers must be formatted following the LNCS Style
- Submissions were anonymous
- Paper submission is closed
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026