Key result
A deep learning-based workflow using 3D U-Nets achieved highly accurate automatic segmentation of the left atrium, right atrium, and pericardium on cardiac CT images, with Dice coefficients of 0.960, 0.945, and 0.967, respectively.
Why the study?
Preoperative left atrium and epicardial adipose tissue volumes on CT are associated with AF recurrence risk, prompting the design of a deep learning-based workflow for reliable automatic segmentation.
Does a deep learning-based 3D U-Net workflow provide reliable automatic segmentation of atria, pericardium, and epicardial adipose tissue on cardiac CT in patients with atrial fibrillation?
Observational (n=157)
No
Does a deep learning-based 3D U-Net workflow provide reliable automatic segmentation of atria, pericardium, and epicardial adipose tissue on cardiac CT in patients with atrial fibrillation?
A 3D U-Net deep learning workflow can reliably and rapidly automate the segmentation of the atria, pericardium, and epicardial adipose tissue on contrast-enhanced cardiac CT in patients with atrial fibrillation.
Supports feasible automated CT segmentation in AF; leaves open prospective clinical validation and workflow integration.
Background Preoperative measurements of left atrium (LA) and epicardial adipose tissue (EAT) volumes in computed tomography (CT) images have been reported to be associated with an increased risk of atrial fibrillation (AF) recurrence. We aimed to design a deep learning-based workflow to provide a reliable automatic segmentation of atria, pericardium and EAT, which can facilitate future applications in AF. Methods A total of 157 patients with AF who underwent radiofrequency catheter ablation were enrolled in this study. The 3D U-Net models of LA, right atrium (RA) and pericardium were used to develop the pipeline of total, LA-and RA-EAT automatic segmentation. We defined the attenuation range between -190 to -30 HU as fat within the pericardium to obtain total EAT. Regions between the dilated endocardial boundaries and endocardial walls of LA or RA within the pericardium were used to detect the voxels attributed to fat, resulting in LA-EAT and RA-EAT. Results The LA, RA and pericardium segmentation models achieved Dice coefficients of 0.960 ± 0.010, 0.945 ± 0.013 and 0.967 ± 0.006, respectively. The 3D segmentation models correlated well with ground truth for LA, RA and pericardium (r=0.99 and p < 0.001 for all). For the results of EAT, LA-EAT and RA-EAT segmentation, Dice coefficients of our proposed method were 0.870 ± 0.027, 0.846 ± 0.057 and 0.841 ± 0.071, respectively. Conclusions Our proposed workflow for automatic LA/RA and EAT segmentation applying 3D U-Nets on CT images was reliable in patients with AF.
No takes yet. Share an insight, caveat, or question.
Wang et al. (2023) conducted an observational in Atrial Fibrillation (n=157). 3D U-Net deep learning automatic segmentation vs. Manual segmentation was evaluated on Dice coefficient for left atrium (LA) segmentation. A deep learning-based workflow using 3D U-Nets achieved highly accurate automatic segmentation of the left atrium, right atrium, and pericardium on cardiac CT images, with Dice coefficients of 0.960, 0.945, and 0.967, respectively.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: