MSK is a founding member of the Cancer AI Alliance (CAIA), an exciting collaborative research initiative designed to accelerate development and deployment of AI models to improve cancer care.
This unique alliance unites four leading NCI-designated cancer centers — MSK, Dana-Farber Cancer Institute, Fred Hutch Cancer Center, and The Sidney Kimmel Comprehensive Cancer Center and Whiting School of Engineering at Johns Hopkins — with technology powerhouses that include Ai2, Amazon, Deloitte, Google, Microsoft, Nvidia, and Slalom.
By harnessing the power of cutting-edge technologies with wide-ranging oncology-focused datasets from MSK and the other participating cancer centers, the CAIA partners are working together to accelerate the pace of cancer treatments, cures, and prevention in a way that has never been previously possible.
“CAIA represents a strategic shift leveraging collective strength,” says Anaeze Offodile II, MD, MPH, EVP and Chief Strategy Officer. “By combining MSK’s clinical expertise with the Alliance’s capital, network of technology partners, data, and federated framework, we can accelerate meaningful advances in cancer care while upholding the highest standards of security and integrity. Together, we move closer to the shared goal of reducing — and ultimately eliminating — the burden of cancer.”
A Unique Collaborative Partnership
More than a year in the planning and since its conceptual founding, CAIA now includes deidentified oncology datasets from all four of the alliance’s healthcare partners, along with $40 million in funding, state-of-the-art AI technology, and expertise from its seven technology partners — all leaders in the field.
The CAIA aims to develop the platform, collaborations, and resources needed to create a comprehensive, shared, federated learning laboratory for cancer AI — in a manner that will ensure that sensitive patient data stays within each cancer center, prioritizing privacy and security.
There are numerous benefits to training AI models on diverse datasets from multiple cancer centers, including reducing biases and enhancing the robustness of results. It is anticipated that the CAIA will create new opportunities for MSK’s researchers, without precluding the possibility of other collaborations and partnerships.
“With scale and diversity of data, and the capacity of the CAIA platform, MSK’s scientists will be empowered to accelerate research advances at a pace not possible by siloed approaches,” says Sohrab Shah, PhD, Chief of Computational Oncology. “The scale of the CAIA is important to overcoming bias limitations of small datasets inherent in many AI models. The CAIA will therefore promote AI models that will benefit cancer patients more broadly, and models that will augment research, including in rare cancers.”
Next Steps
MSK is committed to playing a key role in the CAIA’s development and expansion. Dr. Offodile serves as executive sponsor for this initiative, with Dr. Shah and Stephen Cleaver, PhD, Chief Research Informatics Officer, representing MSK on the CAIA Steering Committee. Additional support is provided by Nikolaus Schultz, PhD, Director of the Cancer Data Science Initiative and Attending Computational Oncologist and Dave Arnold, PhD, Chief Data & Analytics Officer, DigITs.
Currently, the CAIA is facilitating eight pilot projects — two from each of the participating healthcare institutions — with the goal of achieving proof of concept. It is anticipated that MSK will develop a full-scale application process in 2026 for those wishing to leverage CAIA resources to further their research.
For recent news about CAIA, see here: https://www.canceralliance.ai/blog/caia-federated-learning-cancer-ai.