CANCER-AGI: Towards Trustworthy Artificial General Intelligence for Cancer Care and Research

We are developing a Horizon Europe project concept under HORIZON-HLTH-2027-03-TOOL-08, “Towards Artificial General Intelligence (AGI) for healthcare”, focusing on the potential of advanced AI to transform cancer registries and cancer data ecosystems into intelligent, interoperable and trustworthy infrastructures.

The project will explore how next-generation AI and multimodal foundation models could reason across heterogeneous cancer data, including cancer registry information, electronic health records, pathology, medical imaging, molecular/genomic data, treatment history and patient outcomes.

A key concept is to move beyond the traditional role of cancer registries as data collection and reporting systems towards “learning cancer registries” that can support advanced research, personalised medicine and evidence generation.

Potential applications include:
• AI-supported analysis of complete cancer patient journeys;
• prediction of treatment response, recurrence, toxicity and outcomes;
• identification of similar patient populations and real-world treatment pathways;
• detection of gaps and variations in cancer care;
• AI-assisted generation of research hypotheses and evidence;
• integration of registry data with multimodal clinical and molecular information;
• trustworthy and explainable AI with explicit uncertainty assessment;
• privacy-preserving and federated approaches for learning across European cancer registries.

The project will not aim to replace clinical decision-making. Instead, it will investigate how advanced AI capabilities can provide trustworthy, evidence-based insights to clinicians, researchers, cancer registries and policymakers.

We are interested in establishing a multidisciplinary European consortium bringing together cancer registries, hospitals and cancer centres, universities and research organisations, AI and machine-learning experts, health-data and interoperability specialists, patient organisations, ethics and governance experts, and relevant European health-data infrastructures.

Partners with expertise in cancer registry data, multimodal health data, AI/foundation models, federated learning, personalised oncology, clinical research, health-data governance, ethics, interoperability and evaluation are particularly welcome.

The ultimate ambition is to contribute to a European roadmap for the responsible development and evaluation of advanced AI capabilities for cancer care and research, while identifying the data, infrastructure, methodological, ethical and governance requirements needed to move towards trustworthy AGI-enabled healthcare.