Fully automated quantification method (FQM) of coronary calcium in an anthropomorphic phantom

Objective Coronary artery calcium (CAC) score is a strong predictor for future adverse cardiovascular events. Anthropomorphic phantoms are often used CAC studies on computed tomography (CT) to allow evaluation or variation of scanning reconstruction parameters within across scanners against reference standard. This results in large number datasets. Manual assessment these datasets time consuming and cumbersome. Therefore, this study aimed develop validate fully automated, open-source quantification method (FQM) coronary standardized phantom. Materials methods A standard, commercially available anthropomorphic thorax phantom was with an insert containing nine calcifications different sizes densities. To simulate two patient sizes, extension ring used. Image data were acquired four state-of-the-art CT systems using routine scoring acquisition protocols. For interscan variability, each repeated five times small translations and/or rotations. Vendor-specific scores (Agatston, volume, mass) calculated as vendor-specific software. Both the international standard well adjustments implemented FQM. Reference FQM compared Bland-Altman analysis, intraclass correlation coefficients, risk reclassifications, Cohen's kappa. Also, robustness assessed varied acquisitions settings validation dynamic Further, image quality metrics implemented: noise power spectrum, task transfer function, contrast- signal-to-noise ratio among others. Results validated imQuest Three software packages: Hounsfield unit (HU) threshold, minimum area designate group voxels calcium, usage isotropic volume score. The high agreement ICC's (median [95% CI]) excellent (1.000 [0.999-1.000] 1.000 [1.000-1.000]). An interplatform reliability ? = 0.969 0.973 found. TTF gave maximum deviation 3.8% NPS comparable imQuest. Conclusions We developed open-source, robust quantify scans automated algorithm contains fast comparison differences parameters.