Distribution of coronary artery calcium volume and density by age, sex, and race using AI-based quantification algorithm.
Authors
Affiliations (6)
Affiliations (6)
- Department of Radiology and Imaging Sciences, Translational Laboratory for Cardiothoracic Imaging and Artificial Intelligence, Emory University School of Medicine, Atlanta, GA, USA.
- Division of Cardiology, Department of Medicine, Emory University School of Medicine, Atlanta, GA, USA.
- Siemens Medical Solutions, Princeton NJ, USA.
- Johns Hopkins Ciccarone Center for the Prevention of Cardiovascular Disease, Baltimore, MD, USA.
- Lipid Clinic and Cardiovascular Risk Reduction Program, Department of Medicine, Emory University School of Medicine, USA.
- Department of Radiology and Imaging Sciences, Translational Laboratory for Cardiothoracic Imaging and Artificial Intelligence, Emory University School of Medicine, Atlanta, GA, USA. Electronic address: [email protected].
Abstract
The Agatston score is commonly used for coronary artery calcium (CAC) quantification which integrates calcium volume and density into a single measure. Recent studies suggest CAC volume and mean density offer independent prognostic value. However, normative distributions of these parameters across demographic subgroups are not well characterized. A total of 23,844 patients were analyzed who underwent non-contrast ECG-gated cardiac CT at 120 kVp from 2010 to 2023 at our institution. Patients included were older than 35, asymptomatic, and without prior atherosclerotic cardiovascular disease. CAC volume and mean density were directly quantified using a validated deep learning-based software. Participants were stratified by sex, self-reported race (White, Black), and age (in 10-year strata). Percentile distributions were constructed for CAC total volume and average mean density using a LOESS-based approach to account for the zero-inflated nature of the data. Between-group comparisons were conducted with Mann-Whitney U. The cohort (mean age 58 ± 9 years) included 41 % women and 74 % White and 10 % Black participants. Men had significantly higher CAC volume than women across all race/age strata (59 mm<sup>3</sup> (IQR 14-223) vs 28 mm<sup>3</sup> (IQR 8-100), p < 0.001). Black men and women generally had lower CAC volumes than their White counterparts, with the exception of Black women, who demonstrated higher CAC volumes than White women in several age strata. Average mean CAC density increased with age and was consistently higher in White patients compared to Black patients (194 HU (IQR 163-226) vs 171 HU (IQR 152-205), p < 0.001), independent of sex. Sex-based differences in CAC volume persisted after stratification by race. Age-related increases in both volume and density were observed in all groups. This analysis provides percentile distributions of directly measured CAC volume and average mean density across age, sex, and race. The data may better contextualize CAC interpretation and risk stratification.