A group of leading healthcare organizations specializing in cancer treatment has entered into a partnership to better leverage the potential of AI to advance the field. With $40 million in cash and resources from leading technology backers, the Cancer AI Alliance (CAIA) could be a major step forward in precision medicine.
Members of the alliance include Fred Hutchinson, Johns Hopkins University, Dana-Farber, and Sloan Kettering, specifically the cancer research arms of these organizations, who are coordinating the new effort.
Tom Lynch, president and president of Fred Hutch, made the remarks on stage at the Intelligent Applications Summit in Seattle, where the institute is based. This represents an unprecedented capacity, and we can agree that by working together, progress is possible. ”
He said patients with rare childhood cancers are admitted to one center, but the scientific knowledge to better treat them is siled in another center, wrapped in its own methods and handling protocols. I gave an example of this. Perhaps in 10 years, that knowledge will have filtered through the scientific literature, but as he pointed out, that's not that long for children with nonresponsive leukemia.
Of course, AI is not a miracle worker. Also, tugging at the heartstrings does not mean that the problem will be quickly and easily solved by a hypothetical treatment discovery model. But if treatments and research aren't found across these organizations to help move things forward, the entire field will slow down.
The problem is that sharing data between healthcare organizations is not easy due to regulations, safety considerations, and inconsistencies between formats and databases. Even if research exists at Johns Hopkins to help children with leukemia at Sloan Kettering, there is no guarantee that it will exist in a form that can be shared in a legally and technically viable way.
A new organization aims to solve this problem through federated learning. Federated learning is a type of secure data collaboration that can be used for training purposes in AI and other computational systems, although the raw data remains private.
If research organizations can contribute to a common goal, such as training the cancer drug discovery and diagnostic models we all know exist, while complying with HIPAA and other data controls, we will be happy to do so. . CAIA's goal is to build a collaborative system based on this model, but it's still a long way off, said Jeff Leek, vice president and chief data officer at Fred Hutch.
While it's certainly possible, he explained, it's a difficult technical problem that can only be tackled with key participants. Bringing these cancer research centers together and bringing together funding and expertise from Microsoft, AWS, Nvidia, and Deloitte was a necessary first step, and it was not an easy step. Now the actual shared infrastructure, standards, and specific goals (such as pursuing a specific cancer or treatment model) can begin to take shape.
The $40 million is a combination of operating cash, services and intangible assets from the four companies mentioned above, and is expected to be deployed on an unspecified schedule, except that CAIA expects it to be operational by the end of this year. . The effort is expected to “generate first insights” by the end of 2025.