Artificial Intelligence Revolutionizes Cardiac Disease Detection Using Routine Scans

Researchers at Mass General Brigham, in collaboration with the U.S. Department of Veterans Affairs (VA), have developed an innovative AI tool capable of identifying hidden coronary artery calcium (CAC) in previously acquired CT scans. This breakthrough allows for the detection of individuals at heightened risk of cardiovascular events without the need for specialized imaging. The newly developed system, named AI-CAC, leverages deep learning algorithms to analyze routine, nongated chest CT scans—scans that are often performed for screening purposes such as lung cancer detection but do not synchronize with the heartbeat. Traditionally, only gated scans are used to accurately assess CAC, but the AI-CAC can reliably identify calcium deposits and estimate their severity even in standard scans.

Published in NEJM AI, the study demonstrated that AI-CAC possesses an accuracy of 89.4% in detecting CAC presence and correctly classifies the severity of calcium scores with 87.3% precision. Notably, the AI model’s predictions correlated strongly with long-term outcomes, with individuals possessing CAC scores over 400 exhibiting a 3.49-fold increase in 10-year mortality risk. Importantly, nearly all patients flagged by the AI as having high CAC scores would benefit from preventive treatments like lipid-lowering therapy, as confirmed by cardiologist evaluations.

The initiative holds significant promise for opportunistic screening because thousands of nongated chest scans are already in digital repositories, especially within the VA healthcare system, which contains millions of such images. Harnessing AI to analyze these scans proactively could shift medical practice from reactive to preventative, catching cardiovascular risk factors early and improving patient outcomes.

However, the study's limitations include its focus on a veteran population, and future research aims to validate these findings in broader, more diverse groups. The team also plans to evaluate how lipid-lowering medications influence CAC scores over time. Overall, AI integration into routine imaging could transform how clinicians assess cardiovascular risk, making prevention more accessible and effective.