Fully Automated Agatston score calculation from ECG gated Cardiac CT using Deep learning and Multiorgan Segmentation: A Validation study

Purpose: To evaluate deep learning-based calcium segmentation and quantification on ECG gated Cardiac CT scans compared with manual evaluation. Methods: : Automated calcium quantification was performed using a combination of deep-learning convolution neural networks based on Mask R CNN for multiorgan segmentation. Calcifications were identified automatically, after which the algorithm automatically excluded all non-coronary calcifications using 2D erosion, dilation, volume, and maximum intensity threshold and by applying cardiac, aortic, and epicardial fat segmentation. This study used 40 patients to train and test the segmentation model. Results: : 110 patients were tested for the validation of the algorithm. The Pearson correlation coefficient between the reference actual and the computed predictive scores on the test set show high level of correlation (0.84; p < 0.001) and high limits of agreement in Bland-Altman plot. The proposed method correctly classifies the risk group in 75.2% of the cases and classifies the subjects in the same group. 81% of the predictive scores lie in the same categories and only seven patients out of 110 were more than one category off. For the presence/absence of coronary artery calcifications, the deep learning model achieved a sensitivity of 90 % and a specificity of 94 %. Conclusion: Fully automated deep learning-based calcium quantification on cardiac-CTs shows good correlation compared to reference standards. Automating this process may reduce evaluation time and potentially optimize clinical calcium scoring without additional resources.