AUTHORS (FIRST NAME, LAST NAME):
Morteza Naghavi, Anthony Reeves, Matthew Budoff, Dong Li, Kyle Atlas, Chenyu Zhang, Thomas Atlas, Sion K. Roy, Nathan D. Wong, Claudia Henschke, and David Yankelevitz.
Abstract:
Background:
Traditionally derived coronary artery calcium (CAC) scoring offers valuable information beyond traditional risk factors that significantly improve early detection of patients at risk for cardiovascular events. We examined whether artificial intelligence-enabled methods utilizing the CAC scan may provide further improvement in overall cardiovascular disease (CVD) risk prediction.
Methods:
We applied artificial intelligence-enabled automated cardiac chambers volumetry to CAC scans (AI-CAC) of 5830 asymptomatic individuals (52.2% women, age 61.7±10.2 years) that were previously obtained for CAC scoring in the baseline examination (2000-2002) of the Multi-Ethnic Study of Atherosclerosis (MESA). The primary outcome was a composite of all cardiovascular events comprised of stroke, myocardial infarction, angina, resuscitated cardiac arrest, all cardiovascular disease related deaths, heart failure, and atrial fibrillation. We used the 15-year outcomes data and assessed discrimination using the time-dependent area under the curve (AUC) and Uno’s C-statistic between AI-CAC with Agatston CAC Score.
Results:
Over 15-years follow-up, 1773 cardiovascular events accrued. AI-CAC automated cardiac chambers volumetry took on average 21 seconds per CAC scan. The C-statistic for cardiovascular events between AI-CAC and Agatston CAC Score was 0.746 (CI: 0.724-0.768) versus 0.707 (CI: 0.683-0.723) for females, respectively (p<.0001) and 0.680 (CI: 0.653-0.707) versus 0.657 (CI: 0.632-0.682) for males, respectively (P=0.0012). The category-free Net Reclassification Index of AI-CAC over CAC at 15-years follow-up for cardiovascular events was 0.29 (p<.0001).
Conclusion:
In this multi-ethnic longitudinal population study followed for 15 years, the addition of AI-CAC measurements significantly improved on Agatston CAC score for all cardiovascular event prediction.
Comments