Empowering cancer research in Europe: the EUCAIM cancer imaging infrastructure

Artificial intelligence (AI) is a powerful technology with the potential to disrupt cancer detection, diagnosis and treatment. However, the development of new AI algorithms requires access to large and complex real-world datasets. Although such datasets are constantly being generated, access to them is limited by data fragmentation across numerous repositories and sites, heterogeneity, lack of annotations, and potential privacy issues. The European Cancer Imaging Initiative is a flagship of Europe's Beating Cancer Plan, aiming to unlock the power of AI for cancer patients, clinicians, and researchers by establishing a federated European infrastructure for cancer images through the EU-funded EUropean Federation for CAncer IMages (EUCAIM) project. This infrastructure, called Cancer Image Europe, builds on the AI for Health Imaging network (AI4HI), established European Research Infrastructures (Euro-BioImaging, BBMRI-ERIC, EATRIS, ECRIN, and ELIXIR), and numerous related partners providing access to research tools, images, and related clinical, pathology and molecular data. The infrastructure targets clinicians, researchers, and innovators by providing the means to develop and validate data-intensive AI-based and other IT-enabled clinical decision-making systems supporting precision medicine. Common data models, including a linking hyperontology, quality standards, compliance with the FAIR (Findability, Accessibility, Interoperability and Reusability) principles, data annotation, curation and anonymization services are provided to ensure data quality and interoperability, consistency and privacy. In summer 2024, the EUCAIM project released the first prototype of an EU-wide infrastructure, with a comprehensive dashboard integrating applications for dataset discovery, federated search, data access request, metadata harvesting, annotation, secure processing environments and federated processing. CRITICAL RELEVANCE STATEMENT: EUCAIM's federated infrastructure for cancer image data advances medical research and related AI development in Europe. It addresses the current fragmentation and heterogeneity of data repositories is legally compliant, and facilitates collaboration among clinicians, researchers, and innovators. KEY POINTS: AI solutions to advance cancer care rely on large, high-quality real-world datasets. EUCAIM's federated infrastructure for cancer image data empowers cancer research in Europe. It provides access to research tools, images, and related clinical, pathology and molecular data.