Researchers from the University of Western Australia, in collaboration with medical technology partner Artrya, have successfully developed a new fully automated artificial intelligence algorithm that outperforms existing methods in predicting the risk of heart attacks. The findings were published in *Circulation: Cardiovascular Imaging*.
The first author, Dr. Gavin Huangfu, who is affiliated with the University of Western Australia's Medical School and other institutions, noted that while coronary artery calcium (CAC) scoring has transformed preventive cardiology for asymptomatic individuals, the measurement approach has inherent limitations. Dr. Huangfu explained: "Although it is known that disease near the origin of the arteries carries greater risk, standard scoring does not account for the location of coronary calcium plaques. Highly calcified plaques are generally considered higher risk, yet stable plaques actually carry a lower risk of cardiac events. Addressing these shortcomings requires analysis of each individual plaque — something human researchers struggle to do, but which AI can easily achieve." The research team found that an algorithm called CAC-DAD can measure both the overall coronary calcium burden and the distance of lesions from the coronary ostia, while also reclassifying high-density plaques as low risk when appropriate, all in a user-friendly way. Dr. Huangfu said: "The CAC-DAD score is more accurate and effective than the standard Agatston score at predicting cardiac events in vulnerable populations, particularly before and after procedures. Combining the two further improves risk prediction and opens numerous possibilities for clinical use."
Senior author Professor Girish Dwivedi, also from the University of Western Australia's Medical School, emphasized that heart disease remains the leading cause of death in developed countries. Professor Dwivedi stated: "The best treatment is prevention, which requires accurate risk stratification to identify high-risk individuals for targeted intervention. Calcium scoring is a key predictor of first heart attack risk, so improving its accuracy is of major importance for risk management. A shift toward personalized care is essential, and CAC-DAD offers clear advantages in this regard. It is simple yet powerful, and once validated, its output can be used to guide patient management across the medical community. We look forward to validating the predictive ability of CAC-DAD in larger international cohorts."
More information:
Gavin Huangfu et al., Novel CAC Dispersion and Density Score Predicts Myocardial Infarction and Cardiovascular Mortality, *Circulation: Cardiovascular Imaging* (2025). Journal information: *Circulation: Cardiovascular Imaging*
