A pre-competitive consortium to train a shared multimodal cancer foundation model
Companies, hospitals and funders pool effort to train one very large AI on scans, slides, genomes and outcomes from millions of patients, kept at their hospitals, and share the resulting model.
The existing idea of patient-level multimodal foundation models for treatment selection depends on data no single organisation holds. The proposal is the governance and infrastructure to build one as shared infrastructure: a consortium (like the Structural Genomics Consortium or IMI) with federated training across dozens of health systems, pre-agreed data-use terms, open or consortium-licensed weights, a neutral host, and evaluation on sequestered prospective data. Members compete on applications built on top, not on the base model.
- AI that is built but not validated or deployed · Thousands of cancer AI models are published; a handful are in clinical use, and fewer have shown they help patients.
- Data silos · Records, scans, genomes and outcomes sit in separate systems that cannot talk. Every patient's experience is lost to the next.
- Secrecy and intellectual property block collaboration · Companies with complementary drugs rarely test them together, and data that could answer questions stays locked up.
Pages like this
not linked directly; found by shared links- IdeaA federated learning consortium of cancer centres that jointly own the models
Shares Owkin, Patient-level multimodal foundation models for treatment selection, Secrecy and intellectual property block collaboration, Pathology & radiology foundation models.
- IdeaAI second reads to stop borderline lesions being upgraded to cancer
Shares PathAI, Paige AI, Pathology & radiology foundation models, Digital pathology & AI.
- IdeaAI-first reading for high-volume common cancer diagnoses, pathologist for the exceptions
Shares PathAI, Owkin, Paige AI, Pathology & radiology foundation models.
- IdeaA standard evaluation pathway for AI-assisted pathology, from reader study to deployment
Shares PathAI, Paige AI, Digital pathology & AI, AI that is built but not validated or deployed.
- IdeaVersion control and locked reference sets for AI algorithms used as companion diagnostics
Shares PathAI, Paige AI, Pathology & radiology foundation models, Digital pathology & AI.
- IdeaWhole-patient digital twins validated in prospective randomised trials
Shares Patient-level multimodal foundation models for treatment selection, Pathology & radiology foundation models, Digital pathology & AI, AI that is built but not validated or deployed.
- IdeaA registry of external validation datasets for cancer AI models, with mandatory reporting
Shares Patient-level multimodal foundation models for treatment selection, Pathology & radiology foundation models, Digital pathology & AI, AI that is built but not validated or deployed.
- ProductArteraAI Breast
Shares Patient-level multimodal foundation models for treatment selection, Pathology & radiology foundation models, Digital pathology & AI, AI that is built but not validated or deployed.