AI Enhances Heart Disease Risk Prediction, Mayo Clinic Study Finds

Mayo Clinic research has identified a new method to improve the prediction of long-term cardiovascular disease risk. This innovative method utilizes artificial intelligence (AI) to enhance a standard imaging test, offering a promising avenue for earlier identification of heart disease risk and prevention of serious outcomes such as heart attack and stroke.

Study Methodology and Findings

The comprehensive study involved nearly 12,000 adults and spanned approximately 16 years. Researchers applied AI to participants' standard coronary artery calcium scans to precisely quantify the fat surrounding the heart. The predictive value of this AI-driven measurement was then rigorously compared against and combined with two established risk assessment methods: the American Heart Association PREVENT equation and the traditional coronary artery calcium score.

The research clearly indicated that the volume of heart fat independently predicts cardiovascular events. Significantly, when combined with the coronary artery calcium score and the PREVENT equation, this measurement substantially enhanced the accuracy of long-term risk prediction, particularly in patients initially categorized as low-risk.

"The study demonstrates how automatic measurement of pericardial fat can improve risk prediction, especially for patients with borderline or intermediate risk profiles. This approach could facilitate more personalized prevention strategies."
— Zahra Esmaeili, Researcher, Mayo Clinic's Department of Cardiovascular Medicine and first author of the study

Key Findings

The study yielded several critical insights:

• Approximately 10% of participants developed cardiovascular disease during the extensive study period.
• A higher volume of fat around the heart independently correlated with an elevated risk of cardiovascular events, even after accounting for traditional risk factors and coronary calcium scores.
• Participants exhibiting the highest coronary fat volume showed increased risk across all coronary calcium levels, indicating its predictive power even in seemingly healthier individuals.
• Incorporating coronary fat measurements improved the accuracy of predicting cardiovascular events beyond established models, offering a more refined risk assessment.
• Crucially, the study concluded that additional diagnostic information could be extracted from existing coronary artery calcium scans without requiring further tests or incurring additional costs, making the method highly practical.

Francisco Lopez-Jimenez, M.D., a preventive cardiologist and co-director of the AI in Cardiology program at Mayo Clinic, emphasized the practicality of this breakthrough. He noted that "this measurement method is practical and scalable since it uses imaging already widely performed on patients." This widespread applicability could enable earlier and more effective clinical interventions.

Further studies are currently being planned to determine the optimal integration of coronary fat measurement into routine clinical practice and its potential role in guiding treatment decisions for patients worldwide.