基于改进DenseNet的乳腺钼靶肿块分类方法

doi:10.3778/j.issn.1002-8331.2101-0403

摘要/Abstract

摘要： 乳腺X线摄影技术是目前乳腺癌早期发现和诊断的重要手段。然而乳腺X线图像中肿块边缘模糊，分类相对困难，因此提升乳腺肿块的诊断精度从而及早预防和治疗仍是医学领域的一大挑战。针对乳腺肿块的特点，提出了一种结合密集卷积神经网络（DenseNet）和压缩激励（SE）模块的新网络（DSAMNet），该网络融合了二者优势，既加强特征重用，又实现特征提取过程中的特征重标定。根据SE模块嵌入DenseNet的不同位置，提出了模型SE-DenseNet-A、SE-DenseNet-B和SE-DenseNet-C。对SE-DenseNet的池化函数进行改进，提出了模型DSAMNet-A、DSAMNet-B和DSAMNet-C。综合不同结构和不同深度的网络模型在公开数据集CBIS-DDSM上进行训练和测试。实验结果表明，DSAMNet-B有更加优异的性能，其准确率比DenseNet模型的准确率提高了10.8%，AUC达到了0.929。

关键词: 乳腺钼靶图像, 计算机辅助诊断, 卷积神经网络, 图像分类, DenseNet

Abstract: Mammography technology is currently an important method for early detection and diagnosis of breast cancer. However, it is relatively difficult to classify tumor masses due to the blurred edges of mammograms. Therefore, it is still a major challenge in the medical field to improve the diagnostic accuracy of breast masses so as to prevent and treat them as early as possible. According to the characteristics of breast masses, a new network（DSAMNet） combining dense convolutional neural network（DenseNet） and Squeeze-and-Excitation（SE） modules is proposed. This network combines the advantages of the two, which not only strengthens feature reuse, but also realizes feature recalibration in the feature extraction process. In this paper, the models SE-DenseNet-A, SE-DenseNet-B, and SE-DenseNet-C are proposed in consideration of the different positions where the SE module is embedded in DenseNet. The pooling function of SE-DenseNet is improved, and models DSAMNet-A, DSAMNet-B and DSAMNet-C are proposed. Network models with different structures and different depths are trained and tested on the public dataset CBIS-DDSM. The experimental results show that DSAMNet-B has more excellent performance, whose accuracy rate is 10.8% higher than that of the DenseNet model, and the area under the curve（AUC） reaches 0.929.

Key words: mammogram, computer aided diagnosis（CAD）, convolutional neural network（CNN）, image classification, DenseNet

白茹, 余慧, 安建成, 曹锐. 基于改进DenseNet的乳腺钼靶肿块分类方法[J]. 计算机工程与应用, 2022, 58(15): 270-277.

BAI Ru, YU Hui, AN Jiancheng, CAO Rui. Mass Classification of Breast Mammogram Based on Improved DenseNet[J]. Computer Engineering and Applications, 2022, 58(15): 270-277.

参考文献

[1] BRAY F，FERLAY J，SOERJOMATARAM I，et al.Global cancer statistics 2018：globocan estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J].Ca-a Cancer Journal for Clinicians，2018，68（6）：394-424.
[2] 李贺，郑荣寿，张思维，等.2014年中国女性乳腺癌发病与死亡分析[J].中华肿瘤杂志，2018，40（3）：166-171.
LI H，ZHENG R S，ZHANG S W，et al.Incidence and mortality of female breast cancer in China，2014[J].Chinese Journal of Oncology，2018，40（3）：166-171.
[3] 陈万青，郑荣寿.中国女性乳腺癌发病死亡和生存状况[J].中国肿瘤临床，2015，42（13）：668-674.
CHEN W Q，ZHENG R S.Incidence，mortality and survival analysis of breast cancer in China[J].Chinese Journal of Clinical Oncology，2015，42（13）：668-674.
[4] 张巧丽，赵地，迟学斌.基于深度学习的医学影像诊断综述[J].计算机科学，2017，44（z2）：1-7.
ZHANG Q L，ZHAO D，CHI X B.Review for deep learning based on medical imaging diagnosis[J].Computer Science，2017，44（z2）：1-7.
[5] MURTAZA G，SHUIB L，WAHAB A W A，et al.Deep learning-based breast cancer classification through medical imaging modalities：state of the art and research challenges[J].Artificial Intelligence Review，2020，53（3）：1655-1720.
[6] 海金金.基于CNN的乳腺癌钼靶影像病理学分级算法研究[D].郑州：战略支援部队信息工程大学，2018.
HAI J J.Research on automated breast cancer histopathological grading through digital mammograms based on CNN[D].Zhenzhou：PLA Strategic Support Force Information Engineering University，2018.
[7] VAN ZELST J C，TAN T，MANN R M，et al.Validation of radiologists’ findings by computer-aided detection（CAD） software in breast cancer detection with automated 3D breast ultrasound：a concept study in implementation of artificial intelligence software[J].Acta Radiologica，2020，61（3）：312-320.
[8] TIAN C，XU Y，FEI L，et al.Deep learning for image denoising：a survey[C]//International Conference on Genetic & Evolutionary Computing，2018：563-572.
[9] 雒培磊，李国庆，曾怡.一种改进的基于深度学习的遥感影像拼接方法[J].计算机工程与应用，2017，53（20）：180-186.
LUO P L，LI G Q，ZENG Y.Modified approach to remote sensing image mosaic based on deep learning[J].Computer Engineering and Applications，2017，53（20）：180-186.
[10] 吴仁彪，赵婷，屈景怡.基于深度SE-DenseNet的航班延误预测模型[J].电子与信息学报，2019，41（6）：1510-1517.
WU R B，ZHAO T，QU J Y.Flight delay prediction model based on deep SE-DenseNet[J].Journal of Electronics & Information Technology，2019，41（6）：1510-1517.
[11] LEVY D，JAIN A.Breast mass classification from mammograms using deep convolutional neural networks[J].arXiv：1612.00542，2016.
[12] 孙利雷，徐勇.基于深度学习的乳腺X射线影像分类方法研究[J].计算机工程与应用，2018，54（21）：13-19.
SUN L L，XU Y.Research on classification method of mammography based on deep learning[J].Computer Engineering and Applications，2018，54（21）：13-19.
[13] 侯霄雄，许新征，朱炯，等.基于AlexNet和集成分类器的乳腺癌计算机辅助诊断方法[J].山东大学学报（工学版），2019，49（2）：74-79.
HOU X，XU X，ZHU J，et al.Computer aided diagnosis method for breast cancer based on AlexNet and ensemble classifiers[J].Journal of Shandong University（Engineering Science），2019，49（2）：74-79.
[14] VANG Y S，CHEN Z，XIE X，et al.Deep learning framework for multi-class breast cancer histology image classification[C]//International Conference on Image Analysis & Recognition，2018：914-922.
[15] ABDEL RAHMAN A S，BELHAOUARI S B，BOUZERDOUM A，et al.Breast mass tumor classification using deep learning[C]//2020 IEEE International Conference on Informatics，IoT，& Enabling Technologies（ICIoT），2020：271-276.
[16] ERTOSUN M G，RUBIN D L.Probabilistic visual search for masses within mammography images using deep learning[C]//2015 IEEE International Conference on Bioinformatics and Biomedicine，2015：1310-1315.
[17] HUANG G，LIU Z，VAN DER MAATEN L，et al.Densely connected convolutional networks[C]//2017 IEEE Comference on Computer Vision and Pattern Recognition，2017：2261-2269.
[18] LI H，ZHUANG S，LI D，et al.Benign and malignant classification of mammogram images based on deep learning[J].Biomedical Signal Processing & Control，2019，51：347-354.
[19] DAS K，CONJETI S，ROY A G，et al.Multiple instance learning of deep convolutional neural networks for breast histopathology whole slide classification[C]//International Symposium on Biomedical Imaging，2018：578-581.
[20] HU J，SHEN L，ALBANIE S，et al.Squeeze-and-excitation networks[J].IEEE Transactions on Pattern Analysis & Machine Intelligence，2020，42（8）：2011-2023.
[21] IOFFE S，SZEGEDY C.Batch normalization：accelerating deep network training by reducing internal covariate shift[J].arXiv：1502.03167，2015.
[22] GLOROT X，BORDES A，BENGIO Y.Deep sparse rectifier neural networks[C]//International Conference on Artificial Intelligence and Statistics，2011：315-323.
[23] HEATH M，BOWYER K，KOPANS D，et al.The digital database for screening mammography[C]//5th International Workshop on Digital Mammography，2001.
[24] LEE R S，GIMENEZ F，HOOGI A，et al.A curated mammography data set for use in computer-aided detection and diagnosis research[J].Scientific Data，2017，4（1）：170177.