Every CT Scan Is Now a Heart Test
McLaren’s new AI program reviews thousands of past and future CT scans to flag early heart disease risk
Grand Blanc–based McLaren Health Care has begun using artificial intelligence to screen up to 40,000 chest CT scans each year for hidden signs of heart disease. The health system announced that the tool will review CT scans performed over the past year and continue prospectively across its hospitals in Michigan.
The scans were not ordered to check the heart. Most were performed for pneumonia, trauma, lung cancer screening or other unrelated conditions. The system analyzes those existing images for coronary artery calcium and aortic valve calcification — two markers associated with cardiovascular disease, as detailed in McLaren’s announcement.
McLaren estimates that 8% to 10% of patients undergoing chest CT scans will be identified as having previously unknown cardiovascular risk, translating to roughly 3,000 to 4,000 patients annually across its 12 hospitals.
The program begins with a one-year retrospective review of prior scans before transitioning to real-time screening. Patients identified as high risk will receive letters in the mail, and their primary care physicians will also be notified, the Detroit Free Press reported.
This is not a new imaging test. It is the use of existing imaging volume as a preventive cardiology channel.
When Every CT Scan Becomes a Heart Check
McLaren partnered with Bunkerhill Health to deploy the system. The AI model, developed at Stanford University and cleared by the U.S. Food and Drug Administration, is part of Bunkerhill’s Carebricks platform.
The software automatically quantifies coronary artery calcium and aortic valve calcification on chest CT scans performed for non-cardiac reasons. Coronary artery calcium scoring has traditionally required a dedicated cardiac CT scan. In this case, no additional imaging is ordered. The algorithm analyzes scans that were already performed.
Dr. Justin Klamerus, McLaren’s chief clinical officer, said during a Feb. 18 news conference that heart disease remains the leading cause of death in the United States and that earlier identification can prevent complications. The system will “look back one year in time to see if we can identify any of these early signals, any… calcifications that would suggest there’s underlying risk,” he said, according to the Detroit Free Press.
The health system described the initiative as the first of its kind in Michigan to use AI in this way.
Chest CT scans are common across hospital systems. By applying automated analysis to that existing volume, McLaren is effectively layering a screening function onto routine imaging without additional appointments or radiation exposure.
Activating a Year of Stored Scans
The retrospective review converts archived scans into an active risk assessment program. Once a patient is flagged, McLaren initiates a notification workflow. Patients receive letters in the mail, and primary care physicians receive electronic alerts.
Dr. Samer Kazziha, chief medical director of the McLaren Heart & Vascular Institute, said the ability to automatically calculate calcium scores “would be very, very valuable” and would allow physicians to identify patients at intermediate to high risk before symptoms appear.
Nishith Khandwala, co-founder and CEO of Bunkerhill Health, cited an example during the public rollout. A patient visited the emergency department after a turkey bone became lodged in the throat. A CT scan performed for that issue was later analyzed by the AI system, which identified a high risk for heart disease. The patient subsequently saw a cardiologist and underwent bypass surgery.
The example shows how an imaging study ordered for one problem can lead to preventive intervention in another.
McLaren projects that thousands of patients each year could move into cardiovascular evaluation because of the program. The system is scheduled to begin with retrospective screening and then move fully into real-time analysis of all chest CT scans performed within the health system.
Other imaging providers are also embedding AI into routine workflows. RadNet has described integrating AI tools directly into imaging operations as scan volumes rise and workforce shortages persist.
Also read: “We Want to Utilize AI to Bridge That Gap,” Says RadNet Eastern Operations CEO
The scans were already being done. The change is that they are now being used to identify cardiovascular risk before it becomes an emergency.