The proposed ResDUnet model for left ventricle segmentation from echocardiographic images showed a dice similarity increase of 8.4% compared to deeplabv3 and 1.2% compared to basic U-net.
Does the ResDUnet deep learning model improve left ventricle segmentation from echocardiographic images compared to deeplabv3 and basic U-net?
The ResDUnet model improves automated left ventricle segmentation from echocardiograms, potentially enhancing clinical workflow for assessing LV function.
Effect estimate: 8.4% increase vs deeplabv3, 1.2% increase vs basic U-net
Echocardiography is the modality of choice for the assessment of left ventricle function. Left ventricle is responsible for pumping blood rich in oxygen to all body parts. Segmentation of this chamber from echocardiographic images is a challenging task, due to the ambiguous boundary and inhomogeneous intensity distribution. In this paper we propose a novel deep learning model named ResDUnet. The model is based on U-net incorporated with dilated convolution, where residual blocks are employed instead of the basic U-net units to ease the training process. Each block is enriched with squeeze and excitation unit for channel-wise attention and adaptive feature re-calibration. To tackle the problem of left ventricle shape and size variability, we chose to enrich the process of feature concatenation in U-net by integrating feature maps generated by cascaded dilation. Cascaded dilation broadens the receptive field size in comparison with traditional convolution, which allows the generation of multi-scale information which in turn results in a more robust segmentation. Performance measures were evaluated on a publicly available dataset of 500 patients with large variability in terms of quality and patients pathology. The proposed model shows a dice similarity increase of 8.4% when compared to deeplabv3 and 1.2% when compared to the basic U-net architecture. Experimental results demonstrate the potential use in clinical domain.
Amer et al. (Wed,) conducted a other in Left ventricle segmentation from echocardiographic images (n=500). ResDUnet (Residual Dilated UNet) vs. deeplabv3 and basic U-net architecture was evaluated on Dice similarity (8.4% increase vs deeplabv3, 1.2% increase vs basic U-net). The proposed ResDUnet model for left ventricle segmentation from echocardiographic images showed a dice similarity increase of 8.4% compared to deeplabv3 and 1.2% compared to basic U-net.