2023 Hearst Health Prize - AI for Incidental CAC

Congratulations to Stanford Medicine on winning the 2023 Hearst Health Prize for its artificial intelligence solution that helps identify patients at risk for heart attack by scoring incidental coronary artery calcium (CAC) on non-gated chest CTs.

Bunkerhill is proud to be a collaborator on the testing, validation, FDA clearance, and clinical distribution of this life saving AI population health screening tool.

https://www.youtube.com/watch?v=5sa80IvlG3Q

This algorithm runs automatically in the background of all non-gated chest CTs (e.g. lung cancer screens) and provides an exact coronary artery calcium (Agatston) score. When a patient with CAC who is unaware of cardiovascular disease is identified, a notification including the patient’s CAC score and MESA risk percentile will be sent to a PCP and/or Cardiology Clinic for follow up.

After building and validating the accuracy of the algorithm (Nature Digital Medicine validation study below), the Stanford team conducted an RCT to determine the patient follow-up impacts associated with the use of the algorithm, which includes an automated patient follow-up workflow. 

The RCT compared the algorithm (experiment arm) to the current standard of care of qualitative CAC assessments on non-gated CTs (control arm). The study found that patients with elevated CAC received treatment 51% of the time in the experiment arm, versus 7% in the control arm.  A 7x improvement in patient follow-up rates is attributable to use of the algorithm and the associated AI-enabled follow-up workflow.

The algorithm was FDA cleared in March 2023 and is being used as a cardiovascular screening tool in clinical practice at 100+ hospitals in the U.S. via the Bunkerhill Health platform: https://www.bunkerhillhealth.com/cac

Validation Study in Nature Digital Medicine: https://www.nature.com/articles/s41746-021-00460-1

Clinical Follow up RCT in AHA Circulation: https://www.ahajournals.org/doi/full/10.1161/CIRCULATIONAHA.122.062746