Abstract
Breast cancer is the most common cancer in women, and ultrasound imaging is one of the most widely used approaches for diagnosis due to its non-radioactive process, ease of operation and low cost. Moreover, image segmentation plays a great role in medical image analysis, since it affects the accuracy of computer aided diagnosis (CAD) results. However, the malignant mass of breast in ultrasound images often appears irregular boundary and indistinct margin which is difficult to distinguish from other surrounding tissues. Therefore, breast ultrasound images segmentation is significant for diagnosis, and it has attracted the attention of researchers for many years. In this paper, we propose a weighted feature aggregation network with fusing domain knowledge for two-dimensional breast ultrasound images segmentation. (I) We modify the U-Net by adding a classification branch, in which BI-RADS category information is applied as the classification label. (II) In order to deal with the artifacts in ultrasound, such as posterior shadowing, we conduct Squeeze-and-Excitation (SE) block and aggregation mechanism to compose the up-sampling part in U-Net. (III) We employ the conditional random field (CRF) to optimize segmentation to make the boundaries more continuous and integral after getting the output of U-Net. The experiment conducted on a challenging and representative dataset includes more than three thousand two-dimensional breast ultrasound images. Our method achieves Jaccard Index of 84.9%, Matthew correlation coefficient of 90.9%, and Dice Coefficient of 90.8% in testing which demonstrates the potential clinical value of our work.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
D’Orsi, C., et al.: ACR BI-RADS® Atlas, Breast Imaging Reporting and Data System (2013)
Liu, S., Qi, L., Qin, H., Shi, J., Jia, J.: Path aggregation network for instance segmentation. In: IEEE CVPR (2018)
Su, H., et al.: Region segmentation in histopathological breast cancer images using deep convolutional neural network. In: IEEE 12th International Symposium on Biomedical Imaging (ISBI), pp. 55–58 (2015)
Yap, M.H., Goyal, M., Osman, F., et al.: Automated breast ultrasound lesions detection using convolutional neural networks. IEEE J. Biomed. Health Inform., 1 (2017). https://doi.org/10.1109/jbhi.2017.2731873
Moon, W.K., et al.: Tumor detection in automated breast ultrasound images using quantitative tissue clustering. Med Phys. 41(4), 042901 (2014)
Moraru, L., Moldovanu, S., Biswas, A.: Optimization of breast lesion segmentation in texture feature space approach. Med. Eng. Phys. 36(1), 129–135 (2014)
Cao, Z., et al.: Breast tumor detection in ultrasound images using deep learning. In: Wu, G., Munsell, B.C., Zhan, Y., Bai, W., Sanroma, G., Coupé, P. (eds.) Patch-MI 2017. LNCS, vol. 10530, pp. 121–128. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-67434-6_14
Ronneberger, O., Fischer, P., Brox, T.: U-Net: convolutional networks for biomedical image segmentation. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 234–241. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-24574-4_28
Caruana, R.: Multitask learning. Mach. Learn. 28(1), 41–75 (1997)
Lin, T.Y., Dollár, P., Girshick, R., He, K., Hariharan, B., Belongie, S.: Feature pyramid networks for object detection. In: IEEE CVPR, pp. 2117–2125 (2017)
Hu, J., Shen, L., Albanie, S.: Squeeze-and-excitation networks, pp. 7132–7141 (2018). https://doi.org/10.1109/cvpr.2018.00745
Zheng, S., Jayasumana, S., Romera-Paredes, B., et al.: Conditional random fields as recurrent neural networks. In: ICCV, pp. 1529–1537 (2015)
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770–778 (2016)
Simonyan, K., Zisserman, A.: Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556 (2014)
He, K., Gkioxari, G., Dollár, P., Girshick, R.: Mask R-CNN. In: 2017 ICCV, pp. 2980–2988 (2017)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Liu, Y., An, X., Cong, L., Dong, G., Zhu, L. (2020). Embedding Weighted Feature Aggregation Network with Domain Knowledge Integration for Breast Ultrasound Image Segmentation. In: Hu, Y., et al. Medical Ultrasound, and Preterm, Perinatal and Paediatric Image Analysis. ASMUS PIPPI 2020 2020. Lecture Notes in Computer Science(), vol 12437. Springer, Cham. https://doi.org/10.1007/978-3-030-60334-2_7
Download citation
DOI: https://doi.org/10.1007/978-3-030-60334-2_7
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-60333-5
Online ISBN: 978-3-030-60334-2
eBook Packages: Computer ScienceComputer Science (R0)
