High-quality, openly accessible data are a cornerstone of progress in medical imaging and artificial intelligence. Yet the datasets available to our community still represent only a fraction of the diversity of patients, diseases, clinical environments, and regions in which these technologies are ultimately expected to operate. The MICCAI Open Data initiative was established to help address this gap by providing a dedicated platform for sharing high-quality medical imaging datasets, with particular attention to underrepresented populations and diseases. Building on the first two editions in 2024 and 2025, the 2026 initiative marks the third year of this community effort.

This year, the initiative took an important additional step: from promoting open data sharing to actively supporting the creation, curation, and release of open datasets. With support from a MICCAI Society Award for the Advancement of Health Equity, the Open Data initiative introduced micro-grants to help researchers overcome some of the practical barriers involved in preparing and releasing open datasets. Eight projects were awarded either a small grant of USD 650 or a large grant of USD 1,200 to support activities including data collection and curation, annotation and quality control, documentation, and public release. Five of these funded projects, three supported by a large grant and two by a small grant, were able to complete and submit both their paper and dataset to the 2026 Open Data Track. Through this programme, we particularly aimed to support datasets representing populations, diseases, and regions that remain underrepresented in publicly available medical imaging data.

The MICCAI 2026 Open Data Track received 29 submissions, of which 16 were accepted. The papers collected in this special issue document not only the datasets themselves, but also the considerable effort required to make medical data accessible, well described, reusable, and useful to the wider research community. Together, they illustrate the breadth of data needed to develop and evaluate medical AI across different populations, diseases, imaging modalities, and clinical settings.

We are delighted to present this Special Issue on MICCAI Open Data 2026. We hope that these contributions will serve not simply as datasets for the next generation of algorithms, but as resources that enable new questions to be asked, findings to be reproduced and validated, and medical AI to be developed for a broader and more representative range of patients. By combining a venue for dataset dissemination with concrete support for dataset creation and sharing, the Open Data initiative aims to help make open, diverse, and carefully curated data an increasingly integral part of research in our community.

Guest editors

• Martijn Starmans, PhD, Erasmus MC, the Netherlands
• Apostolia Tsirikoglou, PhD, Karolinska Institutet, Sweden
• Lidia Garrucho Moras, PhD, University of Barcelona, Barcelona, Spain
• Kaouther Mouheb, Erasmus MC, the Netherlands
• Laura Arbelaez Ossa, PhD, University of Barcelona, Barcelona, Spain
• Nadieh Khalili, PhD, Radboud University Medical Center, the Netherlands
• Alessa Hering, PhD, Radboud University Medical Center, the Netherlands

Reviewers

We sincerely thank all reviewers for their time, expertise, and thoughtful feedback. Their contributions have been invaluable in ensuring the quality of this special issue. We gratefully acknowledge:

Anabik Pal, Andrea Loddo, Choyeon Hong, Deepa Krishnaswamy, Fengyi Jiang, Feriel Ramdhane, Giorgio Cazzaniga, Gonzalo Esteban Mosquera Rojas, Huixin Zhan, Jan Tagscherer, Kazi Rifat Ahmed, Ketki Kinkar, Kumar Abhishek, Luc Builtjes, Matteo Maspero, Raffaele Mineo, Satrajit Chakrabarty, Shuang Hao, Swathi Parvathaneni, Tanja Lossau, Xuan Loc Pham