一种层次化的乳腺肿瘤分割方法

南京大学学报(自然科学版) ›› 2018, Vol. 54 ›› Issue (1): 64–.

一种层次化的乳腺肿瘤分割方法

袭肖明1，2，杜亨方1，2，孟宪静1，2，张春云1，2，张　光3，于　振4，尹义龙5*

出版日期:2018-01-31 发布日期:2018-01-31
作者简介:1.山东财经大学计算机科学与技术学院，济南，250014；
2.山东省数字媒体重点实验室，山东财经大学，济南，250014；
3.山东省千佛山医院，济南，250014；
4.山东省农村信用社联合社，济南，250014；
5．山东大学计算机科学与技术学院，济南，250101
基金资助:
基金项目：国家自然科学基金(61701280，61573219，61671274)，山东省自然科学基金(ZR2016FQ18)，山东省高等学校优势学科人才团队培育计划，山东省重点研发计划(2017CXGC1504)，山东财经大学优势学科人才团队培育计划
收稿日期：2017－12－08
*通讯联系人，E－mail：ylyin@sdu.edu.cn

Hierarchical segmentation of breast tumor in ultrasound image

Xi Xiaoming1，2，Du Hengfang1，2，Meng Xianjing1，2，Zhang Chunyun1，2，Zhang Guang3，Yu Zhen4，Yin Yilong5*

Online:2018-01-31 Published:2018-01-31
About author:1.School of Computer Science and Technology，Shandong University of Finance and Economics，Ji’nan，250014，China；
2．The Shandong Province Key Laboratoty of Digital Media Technology，Shandong University of
Finance and Economics，Ji’nan，250014，China；
3．Qianfoshan Hospital of Shandong Province，Ji’nan，250014，China；
4．Rural Credit Cooperative of Shandong Province，Ji’nan，250014，China；
5．School of Computer Science and Technology，Shandong University，Ji’nan，250101，China

摘要/Abstract

摘要： 超声图像是乳腺癌辅助诊断常用的工具之一．肿瘤分割是乳腺超声图像分析的基础．乳腺超声图像中的灰度不同质性、纹理及形状的多变性等复杂特点使得肿瘤的精确分割较为困难．提出了一种层次化的分割框架．首先将局部灰度聚类假设引入活动轮廓模型作为底层分割模型，对图像进行初始分割；然后提出基于超像素和支持向量机(Support Vector Machine，SVM)的高层分割模型，对初始结果再进行高层分割．在高层分割过程中，首先使用简单线性迭代聚类(Simple Linear Interactive Cluster，SLIC)提取超像素，然后提取超像素的灰度、纹理和局部特征，最后使用SVM进行分类．高层分割模型是基于底层模型的分割结果学习获取的，能够检测到底层模型可能分割错误的区域，与底层模型具有较好的互补性．因此，提出的层次化分割框架具有较好的鲁棒性．在自建乳腺超声数据库上的实验结果证明了提出方法的有效性和鲁棒性．

Abstract: Breast ultrasound image analysis is commonly used to help breast cancer diagnosis.Breast tumor segmentation is the first step of image analysis and plays an important role for the automatic breast ultrasound image analysis.However，complex characteristics such as intensity inhomogeneity，variance of texture and shape in breast ultrasound images makes accurate tumor segmentation difficult.In order to deal with the complex characteristics appeared in ultrasound images，this paper proposes a hierarchical breast tumor segmentation method.The method is developed with active contour model framework.In the proposed framework，active contour model is first to be used as low－level segmentation model for initial segmentation.Based on the initial segmentation result，a high－level segmentation model based on superpixel and Support Vector Machine(SVM) is proposed.Simple Linear Interactive Cluster(SLIC) is firstly used to divide the image into superpixels.And then，intensity，texture and local features are extracted for each superpixel.At last，SVM is trained based on these superpixels.Because the high－model is trained by introducing the segmentation result information，the high－model can be used for recognizing tumor regions misclassified by low－level segmentation model.In the process of image segmentation by using low－level segmentation model，complex characteristics of the ultrasound images may result in segmentation error of some local tumor regions which appear large intensity difference or texture variation.However，high－level segmentation model has ability to learn certain characteristics of the low－level segmentation model，which can be complementary to low－level segmentation model.Therefore，the high－level segmentation model can be used to correct the segmentation error caused by using low－level segmentation model，which can be used to improve the segmentation performance.We conduct the experiments on our self－constructed database and the experimental results demonstrate the effectiveness and robustness of the proposed method.

袭肖明1，2，杜亨方1，2，孟宪静1，2，张春云1，2，张　光3，于　振4，尹义龙5*. 一种层次化的乳腺肿瘤分割方法[J]. 南京大学学报(自然科学版), 2018, 54(1): 64–.

Xi Xiaoming1，2，Du Hengfang1，2，Meng Xianjing1，2，Zhang Chunyun1，2，Zhang Guang3，Yu Zhen4，Yin Yilong5*. Hierarchical segmentation of breast tumor in ultrasound image[J]. Journal of Nanjing University(Natural Sciences), 2018, 54(1): 64–.

参考文献

[1]　Shi J，Zhou S C，Liu X，et al.Stacked deep polynomial network based representation learning for tumor classification with small ultrasound image dataset.Neurocomputing，2016，194：87－94.
[2] Cheng J Z，Ni D，Chou Y H，et al.Computer－aided diagnosis with deep learning architecture：Applications to breast lesions in US images and pulmonary nodules in CT scans.Scientific Reports，2016，6：24454.
[3] Han S，Kang H K，Jeong J Y，et al.A deep learning framework for supporting the classification of breast lesions in ultrasound images.Physics in Medicine & Biology，2017，62(19)：7714－7728.
[4] Althuis M D，Dozier J M，Anderson W F，et al.Global trends in breast cancer incidence and mortality 1973－1997.International Journal of Epidemiology，2005，34(2)：405－412.
[5] Xian M，Zhang Y T，Cheng H D．Fully automatic segmentation of breast ultrasound images based on breast characteristics in space and frequency domains.Pattern Recognition，2015，48(2)：485－497.
[6] Cheng H D，Shan J，Ju W，et al.Automated breast cancer detection and classification using ultrasound images：A survey.Pattern Recognition，2010，43(1)：299－317.
[7] Zhang L，Ren Y P，Huang C C，et al.A novel automatic tumor detection for breast cancer ultrasound Images ∥ 2011 8th International Conference on Fuzzy Systems and Knowledge Discovery.Shanghai，China：IEEE，2011：401－404.　
[8] Madabhushi A，Metaxas D N．Combining low－，high－level and empirical domain knowledge for automated segmentation of ultrasonic breast lesions.IEEE Transactions on Medical Imaging，2003，22(2)：155－169.
[9] Zhang Q，Xiao Y，Chen S，et al.Quantification of elastic heterogeneity using contourlet－based texture analysis in shear－wave elastography for breast tumor classification.Ultrasound in Medicine and Biology，2015，41(2)：588－600.
[10] Tian J W，Sun L T，Guo Y H，et al.Computerized－aid diagnosis of breast mass using ultrasound image.Medical Physics，2007，34(8)：3158－3164.
[11] Gómez W，Pereira W C A，Infantosi A F C．Analysis of co－occurrence texture statistics as a function of gray－level quantization for classifying breast ultrasound.IEEE Transactions on Medical Imaging，2012，31(10)：1889－1899.
[12] Chabi M L，Borget I，Ardiles R，et al.Evaluation of the accuracy of a computer－aided diagnosis(CAD)system in breast ultrasound according to the radiologist’s experience.Academic Radiology，2012，19(3)：311－319.
[13] Tan T，Platel B，Twellmann T，et al.Evaluation of the effect of computer－aided classification of benign and malignant lesions on reader performance in automated three－dimensional breast ultrasound.Academic Radiology，2013，20(11)：1381－1388.
[14] Joo S，Yang Y S，Moon W K，et al.Computer－aided diagnosis of solid breast nodules：Use of an artificial neural network based on multiple sonographic features.IEEE Transactions on Medical Imaging，2004，23(10)：1292－1300.
[15] Cheng H D，Jiang X H，Sun Y，et al.Color image segmentation：Advances and prospects.Pattern Recognition，2001，34(12)：2259－2281.
[16] Horsch K，Giger M L，Venta L A，et al.Computerized diagnosis of breast lesions on ultrasound.Medical Physics，2002，29(2)：157－164.　
[17] Hojjatoleslami S A，Kittler J．Region growing：A new approach.IEEE Transactions on Image Processing，1998，7(7)：1079－1084.
[18] Poonguzhali S，Ravindran G．A complete automatic region growing method for segmentation of masses on ultrasound images ∥ Proceedings of the International Conference on Biomedical and Pharmaceutical Engineering.Singapore，The Public of Singapore：IEEE，2006：88－92.
[19] Jung I S，Thapa D，Wang G N．Automatic segmentation and diagnosis of breast lesions using morphology method based on ultrasound ∥ Wang L P，Jin Y C．Fuzzy Systems and Knowledge Discovery.Springer Berlin Heidelberg，2005：1079－1088.
[20] Shan J，Cheng H D，Wang Y X．A completely automatic segmentation method for breast ultrasound images using region growing ∥ The 11th Joint International Conference on Information Sciences.Taiwan，China：Atlantis Press，2008：1－6.
[21] Chang R F，Wu W J，Moon W K，et al.Segmentation of breast tumor in three－dimensional ultrasound images using three－dimensional discrete active contour model.Ultrasound in Medicine and Biology，2003，29(11)：1571－1581.
[22] Chen D R，Chang R F，Wu W J，et al.3－D breast ultrasound segmentation using active contour model.Ultrasound in Medicine and Biology，2003，29(7)：1017－1026.
[23] Sarti A，Corsi C，Mazzini E，et al.Maximum likelihood segmentation of ultrasound images with Rayleigh distribution.IEEE Transactions on Ultrasonics，Ferroelectrics，and Frequency Control，2005，52(6)：947－960.
[24] Chen C M，Lu H H S，Lin Y C．An early vision－based snake model for ultrasoundimage segmentation.Ultrasound in Medicine and Biology，2000，26(2)：273－285.
[25] Chang R F，Wu W J，Tseng C C，et al.3－D snake for US in margin evaluation for malignant breast tumor excision using mammotome.IEEE Transactions on Information Technology in Biomedicine，2003，7(3)：197－201.
[26] Jumaat A K，Rahman W E Z W A，Ibrahim A，et al.Segmentation and characterization of masses in breast ultrasound images using active contour ∥ Proceedings of the IEEE International Conference on Signal and Image Processing Applications.Kuala Lumpur，Malaysia：IEEE，2011：404－409.
[27] Li C M，Xu C Y，Gui C F，et al.Level set evolution without re－initialization：A new variational formulation ∥ Proceedings ofthe IEEE Computer Society Conference on Computer Vision and Pattern Recognition.San Diego，CA，USA：IEEE，2005：430－436.
[28] Li C M，Huang R，Ding Z H，et al.A level set method for image segmentation in the presence of intensity inhomogeneities with application to MRI.IEEE Transactions on Image Processing，2011，20(7)：2007－2016.
[29] Rodrigues R，Pinheiro A，Braz R，et al.Towards breast ultrasound image segmentation using multi－resolution pixel descriptors ∥ The 21st International Conference on Pattern Recognition.Tsukuba，Japan：IEEE，2012：2833－2836.
[30] Lee S Y，Huang Q H，Jin LW，et al.A graph－based segmentation method for breast tumors in ultrasound images ∥ The 4th International Conference on Bioinformatics and Biomedical Engineering.Chengdu，China：IEEE，2010：1－4.
[31] Braz R，Pinheiro A M G，Moutinho J，et al.Breast ultrasound images gland segmentation ∥ Proceedings of the IEEE International Workshop on Machine Learning for Signal Processing.Santander，Spain：IEEE，2012：1－6.
[32] Yang H Y，Christopher L A，Duric N，et al.Performance analysis of EM－MPM and K－means clustering in 3D ultrasound image segmen－tation ∥ Proceedings of the IEEE International Conference on Electro/Information Technology.Indianapolis，IN，USA：IEEE，2012：1－4.
[33] Liu B，Cheng H D，Huang J H，et al.Fully automatic and segmentation－robust classification of breast tumors based on local texture analysis of ultrasound images.Pattern Recognition，2010，43(1)：280－298.
[34] Moon W K，Shen Y W，Bae M S，et al.Computer－aided tumor detection based on multi－scale blob detection algorithm in automated breast ultrasound images.IEEE Transactions on Medical Imaging，2013，32(7)：1191－1200.
[35] Achanta R，Shaji A，Smith K，et al.SLIC superpixels compared to state－of－the－art superpixel methods.IEEE Transactions on Pattern Analysis and Machine Intelligence，2012，34(11)：2274－2282.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed