Key result
The proposed deep learning model combining a U-Net encoder, Atrous Spatial Pyramid Pooling, and a DeeplabV3+ decoder achieved a Dice Similarity Coefficient of 0.9287 for left ventricle segmentation.
Why the study?
Manual left ventricle delineation on echocardiography is time-consuming, error-prone, and prone to interobserver variability, creating a need for effective automatic segmentation.
Does an encoder-decoder deep learning model with Atrous Spatial Pyramid Pooling improve left ventricle segmentation accuracy in echocardiography compared to standard models?
Does an encoder-decoder deep learning model with Atrous Spatial Pyramid Pooling improve left ventricle segmentation accuracy in echocardiography compared to standard models?
Absolute Event Rate: 0.9287% vs 0.9262%
A novel deep learning model combining a U-Net encoder, DeeplabV3+ decoder, and Atrous Spatial Pyramid Pooling achieves high accuracy for automated left ventricle segmentation in echocardiography.
Automated LV segmentation may reduce echo assessment time and variability; leaves open prospective clinical validation.
Assessment of cardiac function using echocardiography is an essential and widely used method. Assessment by manually labeling the left ventricle area can generally be time-consuming, error-prone, and has interobserver variability. Thus, automatic delineation of the left ventricle area is necessary so that the assessment can be carried out effectively and efficiently. In this study, encoder-decoder based deep learning model for left ventricle segmentation in echocardiography was developed using the effective CNN U-Net encoder and combined with the deeplabv3+ decoder which has efficient performance and is able to produce sharper and more accurate segmentation results. Furthermore, the Atrous Spatial Pyramid Pooling module were added to the encoder to improve feature extraction. Tested on the Echonet-Dynamic dataset, the proposed model gives better results than the U-Net, DeeplabV3+, and DeeplabV3 models by producing a dice similarity coefficient of 92.87%. The experimental results show that combining the U-Net encoder and DeeplabV3+ decoder is able to provide increased performance compared to previous studies.
No takes yet. Share an insight, caveat, or question.
Azizi et al. (2023) studied Left Ventricle Segmentation (n=10,030). Proposed deep learning model (U-Net encoder + ASPP + DeeplabV3+ decoder) vs. U-Net, DeeplabV3+, and DeeplabV3 models was evaluated on Overall Dice Similarity Coefficient (DSC) on Echonet-Dynamic dataset. The proposed deep learning model combining a U-Net encoder, Atrous Spatial Pyramid Pooling, and a DeeplabV3+ decoder achieved a Dice Similarity Coefficient of 0.9287 for left ventricle segmentation.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: