Special Issue on MICCAI Open Data 2026

Background. High-quality, openly accessible datasets are a cornerstone of medical imaging and artificial intelligence research. However, many diseases, populations, and healthcare environments remain underrepresented in public datasets, limiting the inclusiveness and generalizability of medical AI.

Innovation. Building on the success of the 2024 and 2025 editions, the MICCAI Open Data initiative reached several important milestones in 2026. Shortly before MICCAI 2026, the initiative was officially recognized as a MICCAI Special Interest Group (SIG), establishing Open Data as a permanent community activity within MICCAI. Supported through a MICCAI Society Award for the Advancement of Health Equity, we further expanded the initiative by introducing micro-grants to support the creation, annotation, curation, documentation, and public release of open datasets, particularly for underrepresented populations, diseases, and regions.

Impact. The MICCAI 2026 Open Data Track received 29 submissions, of which 16 were accepted and published in this MELBA Special Issue. Eight projects received micro-grant funding, and several funded projects successfully completed and released their datasets as part of the initiative, demonstrating a tangible pathway from community support to open data dissemination.

Vision. The initiative promotes FAIR data principles, data-centric AI, reproducibility, and global collaboration. By supporting not only dataset dissemination but also dataset creation, we aim to lower barriers to open science and improve the diversity and representativeness of medical AI research worldwide.

Role. I served as Guest Editor together with Apostolia Tsirikoglou, Lidia Garrucho Moras, Kaouther Mouheb, Laura Arbelaez Ossa, Nadieh Khalili, and Alessa Hering. This Special Issue represents the maturation of the MICCAI Open Data initiative from a conference event into a permanent MICCAI Special Interest Group and an ecosystem that actively enables the development, dissemination, and reuse of impactful open datasets.