Deep learning-based reconstruction on cardiac CT yields distinct radiomic features compared to iterative and filtered back projection reconstructions

Abstract

We aimed to determine the effects of deep learning-based reconstruction (DLR) on radiomic features obtained from cardiac computed tomography (CT) by comparing with iterative reconstruction (IR), and filtered back projection (FBP). A total of 284 consecutive patients with 285 cardiac CT scans that were reconstructed with DLR, IR, and FBP, were retrospectively enrolled. Radiomic features were extracted from the left ventricular (LV) myocardium, and from the periprosthetic mass if patients had cardiac valve replacement. Radiomic features of LV myocardium from each reconstruction were compared using a fitting linear mixed model. Radiomics models were developed to diagnose periprosthetic abnormality, and the performance was evaluated using the area under the receiver characteristics curve (AUC). Most radiomic features of LV myocardium (73 of 88) were significantly different in pairwise comparisons between all three reconstruction methods (P < 0.05). The radiomics model on IR exhibited the best diagnostic performance (AUC 0.948, 95% CI 0.880-1), relative to DLR (AUC 0.873, 95% CI 0.735-1) and FBP (AUC 0.875, 95% CI 0.731-1), but these differences did not reach significance (P > 0.05). In conclusion, applying DLR to cardiac CT scans yields radiomic features distinct from those obtained with IR and FBP, implying that feature robustness is not guaranteed when applying DLR.

Figure 1

Flow chart for patient enrollment. LVAD, left ventricular assist device; DLR, learning-based reconstruction; IR, iterative reconstruction; FBP, filtered back projection.

Figure 2

Distribution plot for intraclass correlation of radiomic features of (a) myocardium and (b) periprosthetic mass. Radiomic features are lined on the x axis and color coded based-on which family of radiomics features they belong to, while their corresponding intraclass correlation coefficient is plotted on the y axis.

Figure 3

Receiver operating characteristic curves of radiomics models from three different reconstruction methods. DLR, learning-based reconstruction; IR, iterative reconstruction; FBP, filtered back projection.

Figure 4

Axial cardiac CT image of a 77-year-old female patient. The myocardium (purple color) is segmented by excluding the LV blood pool and trabeculae to improve reproducibility for delineating the endocardial border. CT, computed tomography; LV, left ventricular.

Figure 5

Cardiac CT images of an 83-year-old female exhibiting leaflet thrombosis of the bioprosthetic aortic valve. An ROI is drawn along the hypoattenuated leaflet thickening of the bioprosthetic aortic valve (green color). CT, computed tomography.