Consistency of automated coronary calcium score and extent of emphysema with different CT scanners and radiation dose protocols in lung cancer screening

ObjectiveTo assess the consistency of automated measurements of coronary artery calcification (CAC) burden and emphysema extent on computed tomography (CT) images acquired with different scanners and radiation dose protocols in a lung cancer screening (LCS) population. Materials and MethodsThe analysis included two LCS cohorts, named inter-scanner cohort, whose subjects underwent two consecutive screening rounds with two different dual-source CT scanners, and inter-dose cohort, whose subjects underwent a low-dose CT scan and an ultra-low dose CT scan. Exclusion criteria for CAC measurements were software failure, previous history of CVD and/or of coronary stenting, whereas for emphysema assessment software failure only. CT images were retrospectively analyzed by a fully automated AI software for CAC scoring, using three predefined Agatston score categories (0-99, 100-399, and [≥] 400), and emphysema quantification, using the percentage of low attenuation areas (%LAA). Demographic and clinical data were obtained from the written questionnaire completed by each participant at the first visit. Agreement for CAC and %LAA categories was measured by the k-Cohen Index with Fleiss-Cohen weights (Kw) and 95% Confidence Interval (CI). ResultsIn the inter-scanner cohort, an overlap of CAC strata was observed in 218/245 (90%) volunteers with an almost perfect agreement (Kw= 0.91, 95%CI 0.88-0.95), while an overlap of %LAA strata in 182/256 (71%) volunteers, with a substantial agreement (Kw= 0.70, 95%CI 0.63-0.76). In the inter-dose cohort, an overlap of CAC strata was observed in 275/327 (84%) volunteers, with an almost perfect agreement (Kw= 0.86, 95%CI 0.82-0.90), while an overlap of %LAA strata was found in 204/356 (57%) volunteers, with a moderate agreement (Kw= 0.57, 95%CI 0.51-0.63). ConclusionAutomated CAC and emphysema quantification showed consistent results when applied on CT images acquired with different scanners and different radiation dose CT protocols in two LCS cohorts.