CT Scan Heart Risk: AI Detection
- A new VA and Mass General Brigham collaboration has yielded an artificial intelligence (AI) tool designed to detect individuals at high risk for cardiovascular events.
- The AI-CAC model demonstrated 89.4% accuracy in determining the presence of CAC in scans.
- The model was trained using chest CT scans from veterans across 98 VA medical centers.
Uncover how an innovative AI tool is revolutionizing heart health assessment. Leveraging existing CT scans, this cutting-edge system, developed in collaboration with the VA and Mass General brigham, accurately identifies individuals at high risk of cardiovascular events by detecting coronary artery calcium (CAC).This primary_keyword, a key indicator, is assessed with remarkable precision. The research highlights how the AI model demonstrates remarkable accuracy, predicting 10-year mortality risks with exceptional results. This advancement signals a shift towards proactive disease prevention, possibly reducing healthcare costs and improving patient outcomes.News Directory 3 brings you this urgent report on a major breakthrough in medical technology. Discover what’s next.
AI Tool Predicts Heart Attack risk from CT Scans
Updated June 24, 2025
A new VA and Mass General Brigham collaboration has yielded an artificial intelligence (AI) tool designed to detect individuals at high risk for cardiovascular events. The AI-CAC tool analyzes existing chest CT scans to identify high levels of coronary artery calcium (CAC), a key indicator of potential heart issues.
The AI-CAC model demonstrated 89.4% accuracy in determining the presence of CAC in scans. For scans with CAC, the model accurately assessed whether the score exceeded 100, indicating moderate cardiovascular risk, with 87.3% accuracy. the study, which appeared in Nejm who, suggests widespread implementation of this AI tool could considerably improve cardiovascular risk assessment.
The model was trained using chest CT scans from veterans across 98 VA medical centers. Researchers then tested the AI-CAC’s performance on 8,052 CT scans to simulate routine CAC screening.
The AI-CAC tool also proved predictive of 10-year mortality.Patients with CAC scores over 400 faced a 3.49 times greater risk of death within a decade compared to those with a score of zero. Cardiologists confirmed that nearly all (99.2%) patients identified by the model as having very high CAC scores (above 400) would benefit from lipid-lowering therapy.
”Millions of chest CT scans are taken each year,often in healthy people,such as to screen for lung cancer. Our study shows that significant facts about cardiovascular risk is going unnoticed in these scans,” said Hugo Aerts, director of the Artificial Intelligence in Medicine (AIM) program at Mass General Brigham.
Raffi Hagopian, a cardiologist and researcher at the VA Long Beach Healthcare System, emphasized the potential for AI in shifting medicine toward proactive disease prevention. “Using AI for tasks like CAC detection can help shift medicine from a reactive approach to the proactive prevention of disease, reducing long-term morbidity, mortality and healthcare costs,” Hagopian said.
What’s next
Future studies will assess the tool’s effectiveness in the general population and its ability to monitor the impact of lipid-lowering medications on CAC scores. The initial algorithm was developed using data exclusively from a veteran population.