CLIIR and the Moore Foundation Announce White Paper on Opportunities for AI-enabled Diagnosis

The availability of digitized health data has expanded extensively in the past decade alongside significant growth in associated technologies - spanning broad enterprise EHR platforms to targeted point-of-care prediction tools. These data and technologies are being applied in myriad ways to assess patient health status and inform care. However, applications of data and technology to support clinicians during the diagnostic process are underdeveloped.

In this white paper, we illustrate opportunities to develop AI “wayfinding” tools that better support the diagnostic process. Next, we identify and discuss the types of data assets that need to be available to develop such tools. These include two broad categories of data - traditional patient-centric (clinical) data and new clinician-centric data that reflects their actions during the diagnostic process and the contextual factors surrounding the clinician and patient during this diagnostic process. Lastly, we address broader healthcare system drivers that could speed development of AI tools that support the diagnostic process. These include strengthening incentives specifically for improved diagnostic performance and better characterizing the dynamic diagnostic refinement processes for different common presenting symptoms.

View the full White Paper here.

View the associated JAMA Viewpoint here.