基于深度学习的乳腺超声图像分割网络
A Breast Ultrasound Image Segmentation Network Based on Deep Learning
DOI: 10.12677/jisp.2026.151003, PDF,    科研立项经费支持
作者: 张玉宇*, 杨国亮, 杨 浩：江西理工大学电气工程与自动化学院，江西 赣州
关键词: 医学图像分割；乳腺超声图像；Mamba；Transformer；特征交互；Medical Image Segmentation； Breast Ultrasound Image； Mamba； Transformer； Feature Interaction
摘要: 乳腺肿瘤对女性健康的威胁越来越大，因此使用计算机辅助对乳腺超声图像进行分割成为不可或缺的一部分。由于乳腺超声图像像素低、噪声大使得分割变得极具挑战性。本文提出的基于深度学习的乳腺超声图像分割模型，通过Mamba与Transformer结合对图像进行特征提取，获得更加准确的语义信息；设计特征交互门控(FIG)以及多注意力聚合(MAA)模块对深层和浅层特征进行精细化控制，大幅度提升了乳腺超声图像分割的效率以及精度。模型在公开的数据集BUSI进行实验，其中Dice系数以及HD95分别达到75.91%、21.18 mm，优于现有经典模型。本研究为复杂环境下的医学图像分割提供了新思路，具有一定的应用价值。
Abstract: Breast tumors pose a growing threat to women’s health, so computer-aided segmentation of breast ultrasound images becomes an indispensable part. Segmentation of breast ultrasound images is challenging due to the low pixel size and noise. The breast ultrasound image segmentation model based on deep learning proposed in this paper combines Mamba with Transformer to extract image features and obtain more accurate semantic information. Feature Interactive Gating (FIG) and Multi-Attention Aggregation (MAA) modules were designed to fine-control the deep and shallow features, which greatly improved the efficiency and accuracy of breast ultrasound image segmentation. The model is tested on the public dataset BUSI, where the Dice coefficient and HD95 reach 75.91%and 21.18 mm respectively, which are better than the existing classical models. This study provides a new idea for medical image segmentation in complex environment and has certain application value.
文章引用：张玉宇, 杨国亮, 杨浩. 基于深度学习的乳腺超声图像分割网络[J]. 图像与信号处理, 2026, 15(1): 25-37. https://doi.org/10.12677/jisp.2026.151003

参考文献

[1] Wang, Z.Y., et al. (2023) Advances in Fundamental and Translational Breast Cancer Research in 2022. China Oncology, 33, 95-102.
[2] Sung, H., Ferlay, J., Siegel, R.L., Laversanne, M., Soerjomataram, I., Jemal, A., et al. (2021) Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: A Cancer Journal for Clinicians, 71, 209-249. [Google Scholar] [CrossRef] [PubMed]
[3] Dong, F. (2020) Clinical Applications of Contrast-Enhanced Ultrasound in Diagnosis of Breast Diseases: Present Situation and Prospect. Chinese Journal of Medical Ultrasound (Electronic Edition), 17, 1151-1154.
[4] Almajalid, R., Shan, J., Du, Y. and Zhang, M. (2018) Development of a Deep-Learning-Based Method for Breast Ultrasound Image Segmentation. 2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA), Orlando, 17-20 December 2018, 1103-1108. [Google Scholar] [CrossRef] 
[5] Kayalibay, B., Jensen, G. and van der Smagt, P. (2017) CNN-Based Segmentation of Medical Imaging Data.
[6] Ronneberger, O., Fischer, P. and Brox, T. (2015) U-net: Convolutional Networks for Biomedical Image Segmentation. Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015: 18th International Conference, Munich, 5-9 October 2015, Part III 18.
[7] Shareef, B., Vakanski, A., Freer, P.E. and Xian, M. (2022) ESTAN: Enhanced Small Tumor-Aware Network for Breast Ultrasound Image Segmentation. Healthcare, 10, Article No. 2262. [Google Scholar] [CrossRef] [PubMed]
[8] Yap, M.H., Pons, G., Martí, J., Ganau, S., Sentís, M., Zwiggelaar, R., et al. (2018) Automated Breast Ultrasound Lesions Detection Using Convolutional Neural Networks. IEEE Journal of Biomedical and Health Informatics, 22, 1218-1226. [Google Scholar] [CrossRef] [PubMed]
[9] Hu, Y., Guo, Y., Wang, Y., Yu, J., Li, J., Zhou, S., et al. (2018) Automatic Tumor Segmentation in Breast Ultrasound Images Using a Dilated Fully Convolutional Network Combined with an Active Contour Model. Medical Physics, 46, 215-228. [Google Scholar] [CrossRef] [PubMed]
[10] Chen, J., Lu, Y., Yu, Q., Luo, X., Adeli, E., Wang, Y., Lu, L., Yuille, A. L. and Zhou, Y. (2021) Transunet: Transformers Make Strong Encoders for Medical Image Segmentation.
[11] Cao, H., Wang, Y., Chen, J., Jiang, D., Zhang, X., Tian, Q., et al. (2023) Swin-Unet: Unet-Like Pure Transformer for Medical Image Segmentation. In: Karlinsky, L., et al., Eds., Computer Vision—ECCV 2022 Workshops, Springer, 205-218. [Google Scholar] [CrossRef] 
[12] Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., et al. (2021) Swin Transformer: Hierarchical Vision Transformer Using Shifted Windows. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, 10-17 October 2021, 9992-10002. [Google Scholar] [CrossRef] 
[13] Mo, Y., Han, C., Liu, Y., Liu, M., Shi, Z., Lin, J., et al. (2023) Hover-Trans: Anatomy-Aware Hover-Transformer for Roi-Free Breast Cancer Diagnosis in Ultrasound Images. IEEE Transactions on Medical Imaging, 42, 1696-1706. [Google Scholar] [CrossRef] [PubMed]
[14] Gu, A., Goel, K. and Ré, C. (2021) Efficiently Modeling Long Sequences with Structured State Spaces.
[15] Gu, A. and Dao, T. (2023) Mamba: Linear-Time Sequence Modeling with Selective State Spaces.
[16] Jiao, J., Liu, Y., Liu, Y., Tian, Y., Wang, Y., Xie, L., et al. (2024) VMamba: Visual State Space Model. Advances in Neural Information Processing Systems 37, Vancouver, 10-15 December 2024, 103031-103063. [Google Scholar] [CrossRef] 
[17] Wang, Z., Zheng, J.-Q., Zhang, Y., Cui, G. and Li, L. (2024) Mamba-Unet: Unet-Like Pure Visual Mamba for Medical Image Segmentation.
[18] Yuan, L., Chen, Y., Wang, T., Yu, W., Shi, Y., Jiang, Z., et al. (2021) Tokens-to-Token Vit: Training Vision Transformers from Scratch on Imagenet. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, 10-17 October 2021, 538-547. [Google Scholar] [CrossRef] 
[19] Zhu, X., Cheng, D., Zhang, Z., Lin, S. and Dai, J. (2019) An Empirical Study of Spatial Attention Mechanisms in Deep Networks. 2019 IEEE/CVF International Conference on Computer Vision (ICCV), Seoul, 27 October-2 November 2019, 6687-6696. [Google Scholar] [CrossRef] 
[20] Qin, Z., Zhang, P., Wu, F. and Li, X. (2021) FcaNet: Frequency Channel Attention Networks. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, 10-17 October 2021, 763-772. [Google Scholar] [CrossRef] 
[21] Li, C., Xue, X., Zeng, C., Xu, Y., Xu, X. and Zhao, S. (2025) RegCDNet: A Regnet-Based Framework for Remote Sensing Image Change Detection Combining Feature Enhancement and Gating Mechanism. IEEE Access, 13, 91466-91479. [Google Scholar] [CrossRef] 
[22] Al-Dhabyani, W., Gomaa, M., Khaled, H. and Fahmy, A. (2020) Dataset of Breast Ultrasound Images. Data in Brief, 28, Article ID: 104863. [Google Scholar] [CrossRef] [PubMed]
[23] He, Q., Yang, Q. and Xie, M. (2023) HCTNet: A Hybrid CNN-Transformer Network for Breast Ultrasound Image Segmentation. Computers in Biology and Medicine, 155, Article ID: 106629. [Google Scholar] [CrossRef] [PubMed]