臺灣博碩士論文加值系統

根據統計，乳腺癌不僅是常見的癌症，而且也是女性癌症死亡的第二大原因。因此，早期的檢查和治療是降低死亡率的最有效的方法。近年來，超音波技術中的彈性成像已成為乳腺癌的診斷的一種有效途徑。本文提出了基於三維卷積神經網路(CNN)结合乳房弹性超音波影像的電腦輔助診斷(CAD)系統，對腫瘤進行良惡性分類。首先，我們使用U-Net分割方法從三維B-mode影像中提取腫瘤區域。完成分割後，我們得到了只含有腫瘤區域的B-mode影像和彈性影像，其中彈性影像包含了腫瘤組織彈性的資訊信息。我們也從U-Net中獲得了腫瘤的遮罩影像，遮罩影像包含了腫瘤輪廓的形狀特徵。最後，用我們的CAD系统結合五類影像，分別是B-mode影像，彈性影像，遮罩影像，只含有腫瘤的B-mode影像和只含有腫瘤的彈性影像來診斷腫瘤的良惡性。此外我們使用集成學習的方法降低預測的變異數進而提高診斷性能。在此實驗中，利用63個良性腫瘤和20個惡恶性腫瘤組成的83個經過病理驗證的腫瘤來測試我們所提出的方法。從實驗結果可知，我們提出的基於3-D CNN的集成模型結合了B-mode影像，遮罩影像和只含有腫瘤的彈性影像有最好的性能，腫瘤診斷系統的準確率、靈敏度、特異性分别為90.36% (75/83)，90.00% (18/20)和90.48% (57/63)，ROC曲線下的面積為0.9374。

Breast cancer is the common and second leading cause of cancer death in women. Therefore, early examination and treatment are the most effective way to reduce the mortality rate. Recently, the breast elastography techniques had become a helpful examination for cancer diagnosis. As a result, the computer-aided diagnosis (CAD) system based on 3-D convolutional neural networks (CNN) using breast elastography is proposed to distinguish the tumors as benign or malignant. At first, the U-Net segmentation method is used to extract the tumor region from 3-D B-mode images. After generating the tumor region, the mask images contain complete shape features. Thus, the tumors are diagnosed in the proposed CAD systems using B-mode images, elastographic images, mask images, masked elastographic images, and masked B-mode images. In addition, the ensemble method is adopted to reduce prediction variance and enhance diagnosis performance. In this experiment, totally 83 biopsy-proved lesions composed of 63 benign tumors and 20 malignant tumors are used to evaluate our proposed method. According to results, the best diagnosis performance is using the ensemble method combined with B-mode images, mask images, and masked elastographic images and the accuracy, sensitivity, specificity, and AUC are 90.36% (75/83), 90.00% (18/20), 90.48% (57/63), and 0.9374, respectively.

口試委員會審定書-----------------------------i
致謝-----------------------------------------ii
摘要-----------------------------------------iii
Abstract-------------------------------------iv
Table of Contents----------------------------vi
List of Figures------------------------------vii
List of Tables-------------------------------ix
Chapter 1 Introduction-----------------------1
Chapter 2 Material---------------------------5
2.1 Patients and Lesion Characters-----5
2.2 Data Acquisition-------------------6
Chapter 3 Methods----------------------------8
3.1 Tumor Segmentation---------------9
3.2 Tumor Diagnosis------------------14
3.2.1 3-D DenseNet-------------------15
3.2.2 Ensemble Method----------------18
Chapter 4 Eperiment Result-------------------21
4.1 Statistical Analysis---------------21
4.2 Result-----------------------------22
Chapter 5 Discussion and Conclusion----------42
References-----------------------------------46

[1]L. F. Xie et al., "Blocking the Glycolytic Pathway Sensitizes Breast Cancer to Sonodynamic Therapy," Ultrasound in Medicine and Biology, vol. 44, no. 6, pp. 1233-1243, Jun 2018.
[2]R. L. Siegel, K. D. Miller, and A. Jemal, "Cancer Statistics, 2018," Ca-a Cancer Journal for Clinicians, vol. 68, no. 1, pp. 7-30, Jan-Feb 2018.
[3]A. Takemura, A. Shimizu, and K. Hamamoto, "Discrimination of breast tumors in ultrasonic images using an ensemble classifier based on the AdaBoost algorithm with feature selection," IEEE Transactions on Medical Imaging, vol. 29, no. 3, pp. 598-609, 2010.
[4]M. Heijblom, W. Steenbergen, and S. Manohar, "Clinical photoacoustic breast imaging: the twente experience," IEEE pulse, vol. 6, no. 3, pp. 42-46, 2015.
[5]T. M. Kolb, J. Lichy, and J. H. Newhouse, "Occult cancer in women with dense breasts: Detection with screening US - Diagnostic yield and tumor characteristics," Radiology, vol. 207, no. 1, pp. 191-199, Apr 1998.
[6]K. M. Kelly, J. Dean, W. S. Comulada, and S. J. Lee, "Breast cancer detection using automated whole breast ultrasound and mammography in radiographically dense breasts," European Radiology, vol. 20, no. 3, pp. 734-742, Mar 2010.
[7]T. P. Matthews, K. Wang, C. P. Li, N. Duric, and M. A. Anastasio, "Regularized Dual Averaging Image Reconstruction for Full-Wave Ultrasound Computed Tomography," Ieee Transactions on Ultrasonics Ferroelectrics and Frequency Control, vol. 64, no. 5, pp. 811-825, May 2017.
[8]L. Dong et al., "Quantitative Compression Optical Coherence Elastography as an Inverse Elasticity Problem," Ieee Journal of Selected Topics in Quantum Electronics, vol. 22, no. 3, May-Jun 2016.
[9]T. Nelson et al., Three-dimensional ultrasound. Lippincott Williams & Wilkins Philadelphia, PA:, 1999.
[10]J. E. Lindop, M. T. Graham, A. H. Gee, and R. W. Prager, "3D elastography using freehand ultrasound," Ultrasound in Medicine and Biology, vol. 32, no. 4, pp. 529-545, Apr 2006.
[11]J. M. Chang, W. K. Moon, N. Cho, J. S. Park, and S. J. Kim, "Radiologists’ performance in the detection of benign and malignant masses with 3D automated breast ultrasound (ABUS)," European journal of radiology, vol. 78, no. 1, pp. 99-103, 2011.
[12]A. F. de Winter, M. P. Jans, R. J. P. M. Scholten, W. Deville, D. van Schaardenburg, and L. M. Bouter, "Diagnostic classification of shoulder disorders: interobserver agreement and determinants of disagreement," Annals of the Rheumatic Diseases, vol. 58, no. 5, pp. 272-277, May 1999.
[13]Y. C. Lai, Y. S. Huang, D. W. Wang, C. M. Tiu, Y. H. Chou, and R. F. Chang, "Computer-Aided Diagnosis for 3-D Power Doppler Breast Ultrasound," Ultrasound in Medicine and Biology, vol. 39, no. 4, pp. 555-567, Apr 2013.
[14]W. K. Moon et al., "Classification of Breast Tumors Using Elastographic and B-Mode Features: Comparison of Automatic Selection of Representative Slice and Physician-Selected Slice of Images," Ultrasound in Medicine and Biology, vol. 39, no. 7, pp. 1147-1157, Jul 2013.
[15]B. D. de Vos, J. M. Wolterink, P. A. de Jong, T. Leiner, M. A. Viergever, and I. Išgum, "ConvNet-Based Localization of Anatomical Structures in 3-D Medical Images," IEEE transactions on medical imaging, vol. 36, no. 7, pp. 1470-1481, 2017.
[16]J.-Z. Cheng et al., "Computer-aided diagnosis with deep learning architecture: applications to breast lesions in US images and pulmonary nodules in CT scans," Scientific reports, vol. 6, p. 24454, 2016.
[17]D. Yi et al., "Optimizing and visualizing deep learning for benign/malignant classification in breast tumors," arXiv preprint arXiv:1705.06362, 2017.
[18]G. Huang, Z. Liu, K. Q. Weinberger, and L. van der Maaten, "Densely connected convolutional networks," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, vol. 1, no. 2, p. 3.
[19]A. C. o. R. B.-R. Committee, Breast imaging reporting and data system. American College of Radiology, 1998.
[20]E. I. Zacharaki et al., "Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme," Magnetic Resonance in Medicine: An Official Journal of the International Society for Magnetic Resonance in Medicine, vol. 62, no. 6, pp. 1609-1618, 2009.
[21]A. Rodtook, K. Kirimasthong, W. Lohitvisate, and S. S. Makhanov, "Automatic initialization of active contours and level set method in ultrasound images of breast abnormalities," Pattern Recognition, vol. 79, pp. 172-182, 2018.
[22]K. Zhang, L. Zhang, H. Song, and D. Zhang, "Reinitialization-free level set evolution via reaction diffusion," IEEE Transactions on Image Processing, vol. 22, no. 1, pp. 258-271, 2013.
[23]C. Li, C. Xu, C. Gui, and M. D. Fox, "Distance regularized level set evolution and its application to image segmentation," IEEE transactions on image processing, vol. 19, no. 12, p. 3243, 2010.
[24]H.-K. Zhao, T. Chan, B. Merriman, and S. Osher, "A variational level set approach to multiphase motion," Journal of computational physics, vol. 127, no. 1, pp. 179-195, 1996.
[25]S. Osher and J. A. Sethian, "Fronts propagating with curvature-dependent speed: algorithms based on Hamilton-Jacobi formulations," Journal of computational physics, vol. 79, no. 1, pp. 12-49, 1988.
[26]O. Ronneberger, P. Fischer, and T. Brox, "U-net: Convolutional networks for biomedical image segmentation," in International Conference on Medical image computing and computer-assisted intervention, 2015, pp. 234-241: Springer.
[27]J. Long, E. Shelhamer, and T. Darrell, "Fully convolutional networks for semantic segmentation," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2015, pp. 3431-3440.
[28]Y.-S. Huang, E. Takada, S. Konno, C.-S. Huang, M.-H. Kuo, and R.-F. Chang, "Computer-Aided tumor diagnosis in 3-D breast elastography," Computer methods and programs in biomedicine, vol. 153, pp. 201-209, 2018.
[29]C. Ju, A. Bibaut, and M. van der Laan, "The relative performance of ensemble methods with deep convolutional neural networks for image classification," Journal of Applied Statistics, pp. 1-19, 2018.
[30]G. Valentini, "An experimental bias-variance analysis of SVM ensembles based on resampling techniques," IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), vol. 35, no. 6, pp. 1252-1271, 2005.
[31]F. Chollet, "Keras," ed, 2015.
[32]R. Kohavi, "A study of cross-validation and bootstrap for accuracy estimation and model selection," in Ijcai, 1995, vol. 14, no. 2, pp. 1137-1145: Montreal, Canada.
[33]J. A. Hanley and B. J. Mcneil, "The Meaning and Use of the Area under a Receiver Operating Characteristic (Roc) Curve," Radiology, vol. 143, no. 1, pp. 29-36, 1982.
[34]J. F. Box, "Guinness, Gosset, Fisher, and small samples," Statistical science, vol. 2, no. 1, pp. 45-52, 1987.
[35]R. L. Wasserstein and N. A. Lazar, "The ASA''s Statement on p-Values: Context, Process, and Purpose," American Statistician, vol. 70, no. 2, pp. 129-131, May 2016.
[36]C. J. Murray, A. D. Lopez, and W. H. Organization, "The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020: summary," 1996.
[37]P. M. Shankar, J. M. Reid, H. Ortega, C. W. Piccoli, and B. B. Goldberg, "Use of Non-Rayleigh Statistics for the Identification of Tumors in Ultrasonic B-Scans of the Breast," Ieee Transactions on Medical Imaging, vol. 12, no. 4, pp. 687-692, Dec 1993.
[38]E. Mendelson, J.-F. Chen, and P. Karstaedt, "Assessing tissue stiffness may boost breast imaging specificity," Diagnostic Imaging, vol. 31, no. 12, pp. 15-17, 2009.
[39]K. Liu, G. X. Kang, N. B. Zhang, and B. B. Hou, "Breast Cancer Classification Based on Fully-Connected Layer First Convolutional Neural Networks," Ieee Access, vol. 6, pp. 23722-23732, 2018.
[40]I. J. A. Simpson, M. W. Woolrich, J. L. R. Andersson, A. R. Groves, J. A. Schnabel, and A. s. D. Neuroimaging, "Ensemble Learning Incorporating Uncertain Registration," Ieee Transactions on Medical Imaging, vol. 32, no. 4, pp. 748-756, Apr 2013.
[41]N. Thai-Nghe, Z. Gantner, and L. Schmidt-Thieme, "Cost-sensitive learning methods for imbalanced data," in Neural Networks (IJCNN), The 2010 International Joint Conference on, 2010, pp. 1-8: IEEE.
[42]Y. M. Sohn, M. J. Kim, E. K. Kim, J. Y. Kwak, H. J. Moon, and S. J. Kim, "Sonographic Elastography Combined With Conventional Sonography How Much Is It Helpful for Diagnostic Performance?," Journal of Ultrasound in Medicine, vol. 28, no. 4, pp. 413-420, Apr 2009.
[43]H. Zhi, B. Ou, B. M. Luo, X. Feng, Y. L. Wen, and H. Y. Yang, "Comparison of ultrasound elastography, mammography, and sonography in the diagnosis of solid breast lesions," Journal of Ultrasound in Medicine, vol. 26, no. 6, pp. 807-815, Jun 2007.
[44]P. F. Song et al., "Two-Dimensional Shear-Wave Elastography on Conventional Ultrasound Scanners With Time-Aligned Sequential Tracking (TAST) and Comb-Push Ultrasound Shear Elastography (CUSE)," Ieee Transactions on Ultrasonics Ferroelectrics and Frequency Control, vol. 62, no. 2, pp. 290-302, Feb 2015.