Health AI Partnership: an innovation and learning network for health AI software

Duke Health, Mayo Clinic, UC Berkeley, DLA Piper, and the Gordon and Betty Moore Foundation have established the Health AI Partnership to standardize industry best practices on leveraging artificial intelligence (AI) software (“medical AI”) and distribute key insights and guidance in the form of an open-source curriculum.

While health systems have been attracted to the possible benefits of medical AI, industry best practices for the procurement, integration, and lifecycle management of such tools have yet to be standardized. Health care delivery settings have largely been left on their own to establish practices and policies for medical AI procurement, integration, and lifecycle management.

Medical AI has the potential to substantially improve patient outcomes and make care delivery system operations more efficient. However, if delivery systems continue to procure and implement medical AI without clear strategies or guidance, then the software may result in more harm than good, such as by perpetuating systemic biases or inaccuracies.

Without building consensus around best practices and building capabilities across delivery settings, the quality of medical AI innovations integrated into clinical practice will remain questionable, with real risk of harm to clinicians and patients.

In response, the Health AI Partnership plans to engage various delivery systems and stakeholders to collect and synthesize the industry’s use of medical AI during a 12-month project. The group will build upon capabilities that currently exist across delivery systems to procure and manage therapeutics, laboratory diagnostics, and other medical devices. The group will also actively seek and incorporate feedback from users, regulators, policy experts, and health payers.

Key insights and best practices will then be aggregated to develop an online professional education curriculum that will empower delivery systems to bolster medical AI procurement, integration, and lifecycle management activities. The guidance and curriculum will be made open source and publicly available online. The program is being funded by the Gordon and Betty Moore Foundation. Project leaders include:

  • Duke Health: Suresh Balu, Michael Gao, Manesh Patel, Mark Sendak
  • DLA Piper: Danny Tobey
  • Mayo Clinic: Mark Lifson, Ajai Sehgal, David Vidal
  • UC Berkeley: Deirdre Mulligan, Inioluwa Deborah Raji
  • Community Representative: Alexandra Valladares

Notably, the Health AI Partnership is the first to bring together stakeholders from across care delivery systems to systematically gather information from industry leaders and develop curriculum on best practices for managing medical AI products. The group aims to rapidly strengthen the health care sector’s adoption of medical AI capabilities and capture the industry’s potential to improve quality of care for all.

For more updates on this project, please fill out this form and our team will be in touch soon: https://bit.ly/healthaipartnership 

To read more about the launch of the Health AI Partnership at the 2021 HIMSS Machine Learning & Artificial Intelligence symposium, please visit: https://www.healthcareitnews.com/news/duke-mayo-clinic-others-launch-innovative-ai-collaboration