Artificial Intelligence Evaluates Cardiovascular Risk from CT Scans

Chest computed tomography (CT) is a common diagnostic tool, with approximately 15 million scans conducted each year in the United States, though many are underutilized or not fully explored. While clinicians traditionally assess cardiovascular risk using CT scans that require contrast, a new study has shown that artificial intelligence (AI) can effectively determine cardiovascular risk from routine non-contrast chest CT scans. This method, which assesses coronary calcium and the dimensions of heart chambers and heart muscle, offers a more affordable and less invasive option for identifying cardiovascular risk.

In the study, investigators at Cedars-Sinai (Los Angeles, CA, USA utilized two AI models to analyze data related to coronary calcium and the sizes of heart muscle chambers from the imaging records of nearly 30,000 patients. Findings revealed that these measurements provide more reliable indicators of cardiac risk than the traditional method of a radiologist identifying visual abnormalities. The research showed that non-contrast chest CT can estimate cardiac chambers and left ventricular myocardium from non-contrast chest CT to improve risk classification. Specifically, patients exhibiting higher atrial or ventricular volumes and an abnormal left ventricular mass index were found to be at a higher risk of all-cause or cardiovascular mortality. Furthermore, both left atrial volume and left ventricular mass index were associated with a heightened risk of both all-cause and cardiovascular mortality, even after adjusting for relevant confounding factors and imaging variables.

Image: AI can identify comprehensive cardiovascular risk from CT scans obtained without contrast dye (Photo courtesy of Cedars-Sinai)

“These results are likely practice-changing for many patients because this technology can accurately identify cardiovascular risk without the use of invasive tests or contrast dye that some patients cannot receive,” said Piotr J. Slomka, PhD, director of Innovation in Imaging at Cedars-Sinai, professor of Medicine in the Division of Artificial Intelligence in Medicine and senior author of the study.

Related Links:
Cedars-Sinai