AI-CAC: Revolutionizing Heart Disease Detection Through CT Scans

In an innovative breakthrough, researchers from Mass General Brigham and the United States Department of Veterans Affairs (VA) have unveiled a cutting-edge artificial intelligence tool, known as AI-CAC, which can analyze routine chest CT scans to reveal hidden risks of heart disease. This development holds significant promise for early cardiovascular risk assessment, potentially transforming preventive healthcare practices.

The study, published on June 23, 2025, in the New England Journal of Medicine AI, demonstrates that AI-CAC can effectively detect coronary artery calcium (CAC) levels—an established indicator of impending cardiac events. Senior author Dr. Hugo Aerts, director of the Artificial Intelligence in Medicine (AIM) Program at Mass General Brigham, emphasized the importance of this tool in utilizing previously conducted scans for proactive health management. "Millions of chest CT scans are taken each year for various reasons, often overlooking crucial cardiovascular risk information," Dr. Aerts stated.

The AI-CAC tool was trained on data from approximately 8,052 chest CT scans collected from veterans at 98 VA medical centers. It achieved an impressive accuracy of 89.4% in identifying the presence of CAC and 87.3% in assessing moderate cardiovascular risk based on CAC scores. These findings suggest that the implementation of AI-CAC could significantly enhance the early detection of cardiovascular issues, allowing clinicians to engage with patients before heart disease progresses to more severe stages.

Raffi Hagopian, MD, a cardiologist and lead author of the study, noted the potential of AI-CAC to convert routine imaging into a valuable tool for cardiovascular risk evaluation. "This approach could shift medical practice from reactive to proactive, thus reducing the long-term morbidity and mortality associated with heart disease," Hagopian remarked.

The implications of AI-CAC are far-reaching. According to the researchers, individuals with a CAC score exceeding 400 face a 3.49-fold increase in the risk of death over a decade compared to those with a score of zero. Furthermore, cardiologists verified that nearly all patients identified with high CAC scores would benefit from lipid-lowering therapies, underscoring the tool's clinical relevance.

Despite its promise, the study's authors acknowledge limitations, particularly that the algorithm was developed using a singular veteran population. Future research will aim to validate AI-CAC's efficacy across broader demographics and evaluate its impact on treatment outcomes.

This advancement in artificial intelligence could herald a new era in cardiovascular healthcare, where insights gleaned from existing medical imaging can lead to preemptive interventions and improved patient outcomes. As healthcare systems increasingly integrate AI technologies, the potential for enhanced diagnostic capabilities and personalized patient care continues to expand.

In conclusion, the development of AI-CAC reflects a significant stride towards harnessing artificial intelligence for the proactive management of heart disease, potentially saving lives and improving the quality of care for millions of patients. The researchers' ongoing efforts to refine and validate this tool will be critical in determining its role in the future landscape of cardiovascular health management.