In 2019, Mayo Clinic began a 10-year strategic partnership with Google focused on innovation and cloud computing. Here, Cris Ross, the Rochester, Minn.-based system's CIO, provides updates on what the partnership has accomplished and where it's going.
Note: Responses have been edited lightly for clarity and style.
Question: How did Mayo Clinic choose Google as its innovation and cloud computing partner?
Cris Ross: We wanted to join together with an organization that believed in our vision for how to cure, connect and transform healthcare so people get the care they expect and deserve. After a detailed process, the team chose Google based on its talent and technology as well as our shared vision for the future of healthcare. Google's focus on innovation and commitment to excellence and security made it the right partner to help us transform healthcare.
Q: If you could distill Mayo Clinic’s partnership with Google down to one main goal, what would that be?
CR: This partnership will change how care is delivered and will help us grow as a healthcare organization. Patients will benefit as we improve and discover new ways of providing their care.
Q: How has Mayo Clinic’s partnership with Google already advanced the system's innovation and virtual care initiatives?
CR: The first stage in our partnership is to build Mayo Clinic Cloud on the Google Cloud platform. In partnership with them, we have begun to populate Mayo Clinic Cloud with data and have constructed an AI factory which is being used by nearly 200 teams to advance scientific discoveries with AI tools. We expect to see many new algorithms to improve care coming from the AI factory. We will complete construction of Mayo Clinic Cloud in early 2023.
A second, equally important part of our partnership is our joint research and innovation program. For example, Mayo and Google teams are working together to develop a new way to expedite and improve quality for radiation therapy planning, a critical component of cancer care. Together, we have found ways to more accurately and safely deliver radiation therapies through better planning and targeting. Our initial work has been in head and neck cancers, a complex area for radiation therapy, but we are expanding it into other cancers. When available, this technology will benefit far more than Mayo Clinic patients through Mayo Clinic Platform and could improve quality and access to complex radiation therapy — anywhere.
Q: In which ways will Mayo Clinic’s partnership with Google advance the system’s innovation and virtual care initiatives in the years to come?
CR: Our main goal is to benefit patients, and thanks to technology and the Mayo Clinic Platform business model, we will be able to do this at scale and truly transform healthcare. One stream of work will be for Mayo Clinic scientists and our partners to use data to improve care of our patients. A second stream of work will be to develop algorithms which could be used at Mayo Clinic and elsewhere. A third stream of work will be collaborating with Google on projects beyond the capabilities of our scientists alone.
Q: How has Mayo Clinic leveraged Google’s technology to improve its ability to conduct medical research?
CR: Mayo Clinic Cloud includes a repository of de-identified longitudinal patient records which we have constructed with Google and our partner Nference. Mayo Clinic Platform Discover product line provides access to those high-quality, comprehensive, longitudinal de-identified patient data that few in the industry can offer. Mayo Clinic Platform’s principal partner, Nference, is already using this data with life sciences companies for drug discovery. In addition to applying sophisticated de-identification methods, we have also applied advanced techniques to secure this data, and data never leaves the Mayo Clinic Cloud. This repository of de-identified data is also used by internal researchers.
Q: Why should all hospitals understand the importance of machine learning in improving treatment precision and clinical outcomes?
CR: Not every hospital will develop algorithms based on machine learning, but all of them will use them. Most already are today — voice transcription replacing dictation, computer-aided claims coding and sepsis detectors are all based on algorithms or continuous machine learning. They have applied consumer products like search, wayfinding maps on their websites and other things. Algorithms for care management will begin to become more common.
Mayo Clinic Platform program Validate will ensure these AI-based algorithms are fit for purpose, that they serve well the population they are intended for. And the Deliver product is working out how they can be available for clinicians or people who want more control over their health. These algorithms are already powerful diagnostic tools. For most health systems, these algorithms will be embedded in EHR workflow, products like radiation oncology or EKG devices.