Diagnostic agreement of non-ECG-gated chest computed tomography for coronary artery calcium scoring: comparison with ECG-gated cardiac computed tomography using Agatston and volume methods.

Objectives: To evaluate the agreement and comparative performance of non-ECG-gated computed tomography (CT) for coronary artery calcification (CAC) quantification compared with ECG-gated CT and to assess measurement variability between Agatston and volume scoring methods.

Methods: In this retrospective diagnostic accuracy study, 130 patients undergoing pretranscatheter aortic valve implantation received both ECG-gated cardiac CT (3-mm slice) and non-ECG-gated chest CT (5-mm slice) in the same session. Agatston and volume scores were calculated for total CAC and individual vessels. Analyses included receiver operating characteristic curves, Pearson correlation, Fisher's z -transformation, Bland-Altman plots, and Deming regression.

Results: Both methods showed excellent agreement for detecting high-risk CAC [Agatston: area under the curve (AUC), 0.994; volume: AUC, 0.996; P  = 0.24]. Correlation between ECG-gated and non-ECG-gated CT was very strong (Agatston: r  = 0.978; volume: r  = 0.981), with volume scoring significantly outperforming Agatston in the left anterior descending artery ( z = -2.02, P  = 0.044). Bland-Altman analysis revealed greater bias and wider limits of agreement for Agatston scores. Deming regression showed Agatston scores were consistently underestimated (slope = 0.61), whereas volume scores had slopes closer to unity (1.07), reflecting stronger agreement.

Conclusion: Calcium quantification on non-ECG-gated CT demonstrated strong agreement with ECG-gated reference values. Volume scoring showed superior reproducibility and reduced susceptibility to variability inherent in non-ECG-gated imaging.