Abstract
Ultrasound images are one of the most widely used medical images in clinical medicine. However, ultrasound images generally have the characteristics of strong noise, weak edges, and complex organizational structures, so the segmentation of ultrasound images is very difficult. Aiming at the problems of low efficiency and poor recognition accuracy in the existing ultrasound image segmentation algorithm, a deep layer aggregation for a residual dense network (DLA-RDNet) is proposed to segment the left ventricle ultrasound images. First, to locate the left ventricular region, perform morphological operations on the left ventricular ultrasound image to obtain the target image. Then, a deep aggregation residual for dense networks is proposed for left ventricular ultrasound image segmentation, that is, a residual dense network (RDNet) is designed to extract image features, and the feature information is closely integrated through the deep aggregation (DLA) method. Finally, the network is pruned through the deep supervision (DS) method, which simplifies the network structure and improves the network operation speed. We segment the left ventricular ultrasound images by the proposed algorithm. The experimental results show that the average precision rate(AP) is 95.68%, the average intersection over union (IoU) is 97.13%, the Dice coefficient is 97.15%, the average perpendicular distance(APD) is 0.31 mm, and the good contours rate (GC) is 99.32%. Compared with the other six segmentation algorithms, the proposed algorithm achieves a more effective segmentation of the left ventricle ultrasound images. Therefore, the proposed algorithm can accurately segment the left ventricle ultrasound images, which meets the needs of the segmentation of the left ventricle ultrasound images in clinical medicine.
Similar content being viewed by others
References
D’Andrea A et al (2020) New ultrasound technologies for ischemic heart disease assessment and monitoring in cardiac rehabilitation. J Clin Med 9(10):3131
Seo J, and Kim Y-S (2017) "Ultrasound imaging and beyond: recent advances in medical ultrasound. "Biomed Eng Lett 7.2: 57–58
Jang J et al (2015) A reconstruction method of blood flow velocity in left ventricle using color flow ultrasound. Comput Math Methods Med 2015:1–15
Wu S et al. (2017) Automatic segmentation of ultrasound tomography image. BioMed Res Int 2017
Zyuzin Vas et al. (2018) "Identification of the left ventricle endocardial border on two-dimensional ultrasound images using the convolutional neural network Unet." Ural Symposium Biomed Eng, RadioelectronicsInform Technol (USBEREIT). IEEE, 2018
Yang C et al (2017) Left ventricle segmentation via two-layer level sets with circular shape constraint. Magn Reson Imaging 38:202–213
Ma Y, Wang L, Ma Y, Dong M, du Shiqiang, Sun X (2016) An SPCNN-GVF-based approach for the automatic segmentation of left ventricle in cardiac cine MR images. Int J Comput Assist Radiol Surg 11(11):1951–1964
Avendi MR, Kheradvar A, Jafarkhani H (2016) A combined deep-learning and deformable-model approach to fully automatic segmentation of the left ventricle in cardiac MRI. Med Image Anal 30:108–119
Koo HJ et al (2020) Automated Segmentation of Left Ventricular Myocardium on Cardiac Computed Tomography Using Deep Learning. Korean J Radiol 21.6:660
Bouzidi S, Emilien A, Benois-Pineau J, Quesson B, Ben Amar C, Desbarats P (2017) Segmentation of left ventricle on dynamic MRI sequences for blood flow cancellation in thermotherapy. Signal Process Image Commun 59:37–49
Arrieta C, Uribe S, Sing-Long C, Hurtado D, Andia M, Irarrazaval P, Tejos C (2017) Simultaneous left and right ventricle segmentation using topology preserving level sets. Biomed Signal Process Control 33:88–95
Kass M, Witkin A, Terzopoulos D (1988) Snakes: active contour models. Int J Comput Vis 1(4):321–331
Shin H-C, Roth HR, Gao M, Lu L, Xu Z, Nogues I, Yao J, Mollura D, Summers RM (2016) Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning. IEEE Trans Med Imaging 35(5):1285–1298
Du X et al (2019) Cardiac-DeepIED: automatic pixel-level deep segmentation for cardiac bi-ventricle using improved end-to-end encoder-decoder network. IEEE J Transl Eng Health Med 7:1–10
Abdelhafiz D, Bi J, Ammar R, Yang C, Nabavi S (2020) Convolutional neural network for automated mass segmentation in mammography. BMC Bioinform 21(1):1–19
Yang Z et al (2017) Improved variational autoencoders for text modeling using dilated convolutions. Int Conf Mach Learn. PMLR
Chen L-C, Papandreou G, Kokkinos I, Murphy K, Yuille AL (2017) Deeplab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs. IEEE Trans Pattern Anal Mach Intell 40(4):834–848
Chen L-C et al (2018) Encoder-decoder with atrous separable convolution for semantic image segmentation. Proceed Eur Conf Comp Vision (ECCV)
Zhao H et al. (2017) Pyramid scene parsing network. Proceed IEEE Conf Comp Vision Patt Recog
Sun W, Wang R (2018) Fully convolutional networks for semantic segmentation of very high resolution remotely sensed images combined with DSM. IEEE Geosci Remote Sens Lett 15(3):474–478
Ronneberger O, Fischer P and Brox T (2015) "U-net: Convolutional networks for biomedical image segmentation. "International Conference on Medical image computing and computer-assisted intervention. Springer, Cham
Badrinarayanan V, Kendall A, Cipolla R (2017) Segnet: a deep convolutional encoder-decoder architecture for image segmentation. IEEE Trans Pattern Anal Mach Intell 39(12):2481–2495
Lin G et al. (2017) Refinenet: Multi-path refinement networks for high-resolution semantic segmentation. Proceed IEEE Conf Comput Vision Patt Recogn
Peng C et al. (2017) Large kernel matters--improve semantic segmentation by global convolutional network. Proceed IEEE Conf Comput Vision Patt Recogn
Jégou S et al. (2017) The one hundred layers tiramisu: Fully convolutional densenets for semantic segmentation. Proceed IEEE Conf Comput Vision Patt Recogn Workshops
Zhou Z, et al. (2018) Unet++: a nested u-net architecture for medical image segmentation. Deep learning in medical image analysis and multimodal learning for clinical decision support. Springer, Cham. 3–11
Li C, Fan Y, Cai X (2021) PyConvU-net: a lightweight and multiscale network for biomedical image segmentation. BMC Bioinform 22(1):1–11
He K et al. (2017) "Mask r-cnn." Proceed IEEE Int Conf Comp Vision
Xiong J et al. (2021) "Edge-Sensitive Left Ventricle Segmentation Using Deep Reinforcement Learning." Sensors 21.7: 2375
Yu F et al. (2018) Deep layer aggregation. Proceed IEEE Conf Comp Vision Patt Recog
He K et al (2016) Deep residual learning for image recognition. Proceed IEEE confer Comput Vision Patt Recogn
Huang G et al. (2017) Densely connected convolutional networks." Proceed IEEE Conf Comp Vision Patt Recogn
Lee C-Y, et al. (2015) "Deeply-supervised nets." Artif Intell Statist. PMLR
Lin T-Y et al. (2017)" Feature pyramid networks for object detection. Proceed IEEE Conf Comput Vision Patt Recogn
Zhang Y et al. (2018) Residual dense network for image super-resolution." Proceed IEEE Conf Comput Vision Patt Recogn
Hu K, Zhang Z, Niu X, Zhang Y, Cao C, Xiao F, Gao X (2018) Retinal vessel segmentation of color fundus images using multiscale convolutional neural network with an improved cross-entropy loss function. Neurocomputing 309:179–191
Yilmaz P, Wallecan K, Kristanto W, Aben J-P, Moelker A (2018) Evaluation of a semi-automatic right ventricle segmentation method on short-axis MR images. J Digit Imaging 31(5):670–679
Acknowledgments
This work was supported by the Research Project on Science and Technology of Chongqing Education Committee in 2020 (KJQN202005803, KJQN202005801), the Seventh Batch of Pilot Projects of Comprehensive Education Reform in Chongqing in 2021 (21JGS63), the Demonstration Course of Ideological and Political of Chongqing Vocational College of Public Transportation in 2021 (KCSZ-21-01), Research Project of Vocational College of Public Transportation in 2020 (YSKY2020-04). The authors thank the Editor-in-Chief, the Associate Editor, and the anonymous reviewers for their constructive comments and suggestions.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Wu, X., Li, X., Mou, G. et al. Identification of the left ventricle endocardial border on two-dimensional ultrasound images using deep layer aggregation for residual dense networks. Appl Intell 52, 16089–16103 (2022). https://doi.org/10.1007/s10489-022-03392-9
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10489-022-03392-9