Paper submissions
Submissions are invited for peer-reviewed papers addressing the three core themes of the workshop: stable adaptation, faithful reasoning, and evaluation/trustworthiness of medical AI models.
AI/CORE Unranked
SAFER 2026
First Workshop on Stable Adaptation and Faithful Evaluation of Reasoning in Medical Foundation Models
The SAFER 2026 workshop, held in conjunction with MICCAI in Strasbourg, France, brings together researchers from academia, industry, and clinical practice to advance trustworthy medical AI. It focuses on developing foundation models that enable stable adaptation, faithful reasoning, and robust evaluation across multimodal clinical data, with an emphasis on transparency, safety, and reliability in imaging and vision-language systems.
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Research areas in scope
01
Stable and Efficient Adaptation of Medical VLMs
Test-time adaptation and continual learning for medical imagingEfficient parameter/prompt-tuning for new clinical domainsUnsupervised, semi-supervised, and weakly-supervised adaptationRobustness to domain shifts in imaging and other clinical dataHuman-in-the-loop reinforcement learning for clinical feedback
02
Faithful Reasoning and Explainability in Medical VLMs
Chain-of-thought and step-by-step reasoning for medical decisionsVisual grounding of reasoning on medical images (e.g., CT, MRI, Pathology)Causality-informed reasoning for image-based diagnosis and prognosisRetrieval-augmented generation (RAG) for multimodal medical dataGenerative models for synthetic medical data and report generation
03
Evaluation, Safety, and Trustworthiness
Hallucination detection, mitigation, and quantificationFaithfulness metrics and benchmarks for multimodal modelsUncertainty quantification in LLMs and VLMsInterpretability and explainability of model reasoningPrivacy-preservation during adaptation and reasoning
Policies worth checking twice
- Papers must be original and not under review elsewhere.
- Dual submission to other conferences or journals is not permitted.
- Submissions must adhere to the specified page limits (implied by standard MICCAI workshop practices, though not explicitly stated).
- Camera-ready papers are required after acceptance.
- The workshop follows a blind review process (inferred from standard MICCAI practices and absence of non-blind instructions).
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026