Mass General Brigham researchers developed and tested an AI tool called AI-CAC, in collaboration with the US Department of Veterans Affairs, that automatically detects hidden signs of heart disease by analyzing routine chest CT scans that were originally taken for unrelated purposes, such as lung cancer screening.

AI-CAC is a deep learning algorithm that quantifies coronary artery calcium (CAC) levels — a key predictor of heart attack risk — from non-gated chest CT scans, which are the standard type taken for non-cardiac purposes. CAC is not normally measured from such scans. The tool was developed using 446 expert segmentations and tested across data from 98 VA medical centers. A study published in NEJM AI in 2025 showed high accuracy and predictive value for future cardiac events and 10-year mortality. As of mid-2025, the tool remained in a research/validation phase; sources indicate it could be implemented widely but have not confirmed routine clinical deployment.