Key result
YOLOv2-based object detection outperforms segmentation for aortic valve extraction, achieving a ~0.81 Dice score.
Why the study?
Does object detection improve the accuracy of aortic valve annulus extraction compared to segmentation in contrast-enhanced cardiac CT images?
Does object detection improve the accuracy of aortic valve annulus extraction compared to segmentation in contrast-enhanced cardiac CT images?
Object detection using YOLOv2 outperforms segmentation for automatic extraction of the aortic valve annulus from cardiac CT images, showing promise for preoperative planning.
Answers are grounded in this paper and cite their sources.
May support automated aortic valve extraction in CT; leaves open clinical validation for preoperative planning.
PURPOSE: This study evaluates the use of deep learning techniques to automatically extract and delineate the aortic valve annulus region from contrast-enhanced cardiac CT images. Two approaches, namely, segmentation and object detection, were compared to determine their accuracy. MATERIALS AND METHODS: A dataset of 32 contrast-enhanced cardiac CT scans was analyzed. The segmentation approach utilized the DeepLabv3+ model, while the object detection approach employed YOLOv2. The dataset was augmented through rotation and scaling, and five-fold cross-validation was applied. The accuracy of both methods was evaluated using the Dice similarity coefficient (DSC), and their performance in estimating the aortic valve annulus area was compared. RESULTS: The object detection approach achieved a mean DSC of 0.809, significantly outperforming the segmentation approach, which had a mean DSC of 0.711. Object detection also demonstrated higher precision and recall, with fewer false positives and negatives. The aortic valve annulus area estimation had a mean error of 2.55 mm. CONCLUSIONS: Object detection showed superior performance in identifying the aortic valve annulus region, suggesting its potential for clinical application in cardiac imaging. The results highlight the promise of deep learning in improving the accuracy and efficiency of preoperative planning for cardiovascular interventions.
No takes yet. Share an insight, caveat, or question.
Inomata et al. (2024) studied Aortic valve annulus evaluation (n=32). Object detection (YOLOv2) vs. Segmentation (DeepLabv3+) was evaluated on Dice similarity coefficient (DSC) for aortic valve annulus region extraction. The object detection approach using YOLOv2 achieved a mean Dice similarity coefficient of 0.809, outperforming the segmentation approach for automatic aortic valve extraction.
Synapse has enriched 10 related papers. Consider them for comparative context: