Key result
Automated CAC scoring shows excellent agreement with manual reads on standard CT scans.
Why the study?
The influence of CT slice thickness on the accuracy of deep learning-based automatic coronary artery calcium scoring software had not yet been explored.
Does CT slice thickness influence the accuracy of deep learning-based, automatic coronary artery calcium scoring software compared to manual scoring?
Cross-Sectional (n=844)
Observers for manual scoring were blinded to the CAC scores on other slice thickness images and to scores by another reader
No
Does CT slice thickness influence the accuracy of deep learning-based, automatic coronary artery calcium scoring software compared to manual scoring?
Effect estimate: weighted κ 0.851 (95% CI 0.823-0.879)
Deep learning-based automatic CAC scoring is highly reliable on standard 3 mm CT slices but exhibits lower categorical agreement and more false positives on 1.5 mm slices due to increased image noise.
Supports automatic CAC scoring on standard 3 mm CT; leaves open validation and noise mitigation for 1.5 mm scans.
Background: The influence of computed tomography (CT) slice thickness on the accuracy of deep learning (DL)-based, automatic coronary artery calcium (CAC) scoring software has not been explored yet. Methods: This retrospective study included 844 subjects (477 men, mean age of 58.9±10.7 years) who underwent electrocardiogram (ECG)-gated CAC scoring CT scans with 1.5 and 3 mm slice thickness values between September 2013 and October 2020. Automatic CAC scoring was performed using DL-based software (3D patch-based U-Net architectures). Manual CAC scoring was set as the reference standard. The reliability of automatic CAC scoring was evaluated using intraclass correlation coefficients (ICCs) for both the 1.5 and 3 mm datasets. The agreement of CAC severity categories [Agatston score (AS) 0, 1-100, 101-400, >400] between automatic CAC scoring and the reference standard was analyzed using weighted kappa (κ) statistics for both 1.5 and 3 mm datasets. Results: The CAC scoring agreement between the automatic CAC scoring and reference standard was excellent (ICC 0.982 for 1.5 mm, 0.969 for 3 mm, respectively). The categorical agreement of CAC severity between two methods was excellent for both 1.5 and 3 mm scans, with better agreement for 3 mm scans (weighted κ: 0.851 and 0.961, 95% confidence intervals: 0.823-0.879 and 0.945-0.974, respectively). Conclusions: Automatic CAC scoring shows excellent agreement with the reference standard for both 1.5 and 3 mm scans but results in lower agreement in the CAC severity category for 1.5 mm scans.
No takes yet. Share an insight, caveat, or question.
Kim et al. (2023) conducted a cross-sectional in Asymptomatic individuals undergoing cardiac CT for health evaluation (n=844). Deep learning-based automatic coronary artery calcium (CAC) scoring software vs. Manual CAC scoring (reference standard) was evaluated on Categorical agreement of CAC severity (Agatston score 0, 1-100, 101-400, >400) between automatic CAC scoring and manual reference standard on 1.5 mm scans (weighted κ 0.851, 95% CI 0.823-0.879). Automatic CAC scoring showed excellent agreement with manual scoring for categorical CAC severity on standard 3 mm CT scans (weighted κ 0.961) but demonstrated lower agreement on 1.5 mm scans (weighted κ 0.851) largely due to image noise.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: