臺灣博碩士論文加值系統

在女性癌症人口中以罹患女性乳癌為主要病因。在多數的以開發國家中，乳癌已經成為女性最主要的死亡原因。近幾年來不少學者投入如何將超音波診斷結合電腦輔助診斷的議題，此論文提出一個結合影像特徵及腫瘤圈選的研究，進一步將取得的影像特徵用以增加腫瘤範圍圈選的準確度，並探討由超音波影像擷取出的特徵是否能與病理特徵或一些常見之癌症類型有所關聯。
在腫瘤圈選的研究當中，我們試圖去開發一套結合腫瘤的影像特徵和強健的影像弱邊界切割演算法擷取腫瘤，稱為結合式距離規則化水平集演化(combined distance regularized level set evolution, cDRLSE)。首先，我們將超音波影像以手動的方式劃分出腫瘤區域及非腫瘤區域，並擷取出影像特徵，再藉由支持向量機器(support vector machine)訓練並分類，用以得到初始位置。第二步驟，應用改良式DRLSE取得第二次的初始位置。第三步驟，使用高斯濾波器去平滑雜訊，接著使用DRLSE找出大略的腫瘤範圍。第四步驟，將影像做後處理，以便找尋最適當的腫瘤範圍。並將cDRLSE與另外兩種圈選方式做比較(方法描述在附錄內)，可以發現cDRLSE的切割結果比另外兩者來的好。再透過腫瘤影像的灰階強度(gray scale intensity)及紋理(texture)，我們可以獲得大量影像紋理特徵(image texture feature)用以建立影像特徵資料庫。其中紋理分析包括統計分析(mean and variance, skewness and kurtosis)、熵(entropy)及碎形維度(fractal dimension)還有灰階共生矩陣(gray-level co-occurrence matrix)都包含在使用範圍。在計算灰階共生矩陣時，我們包含不同距離參數，以達到取得大量紋理資料之目的。最後，綜合以上各種紋理特徵，以及病理特徵，我們去分析兩者之間是否有所關聯。根據SPSS分析結果，腫瘤病名僅與病理等級和分類有關連性，而大部分的紋理特徵與病理特徵並沒有強烈關聯，但在病理特徵中的尺寸(size), 淋巴結數量(lymph nodes), 腫瘤壞死(tumor necrosis)與一些影像特徵有關連性。

Breast cancer is the most common type of cancer in women worldwide, which inspires researchers worldwide to develop computer-aided-diagnosis with ultrasound images. In this thesis, we propose the research of combining texture features with tumor segmentation based on level set evolution in breast ultrasound image. Furthermore we try to find if the texture features coming from ultrasound images have correlation with histological features or some specific types of breast cancer.
First a segmentation method, which combines the texture features and one rubout weak-edge segmentation algorithm, called combined distance regularized level set evolution (cDRLSE), is developed for capturing the contour of tumors. The proposed cDRLASE consists of the following steps. First, apply the texture features for support vector machine (SVM) to decide the initialization area. Second, re-initialize by improved DRLSE, where the edge indicator of external energy was improved so that the contour can easily and precisely reach to the weak boundary. Third, Gaussian filtering is used to smooth noise, and apply DRLSE for roughly capturing the tumor area. Last, with the post-processing to find a proper tumor area. Meanwhile, manual segmentation on BUS images under the supervision of Dr. Chou, an experienced doctor, are performed to define the contour separating the tumor and non-tumor region and used as ground truth contour for evaluation. Comparing the cDRLSE method with two other segmentation methods in Appendix, cDRLSE outperforms the other two in the average.
Then, BUS tumor image texture features are extracted by gray scale intensity analysis and texture analysis. The image features includes mean, variance, skewness, kurtosis, entropy, fractal dimension and gray-level co-occurrence matrix. During the procedure of computing gray level co-occurrence matrix, various distance parameters are included, which lead us to build a large number of texture features.
Finally, we investigate if there are any correlations between texture features extracted from breast ultrasound images, pathological features and the histological type by SPSS. According to the analysis, most of the features have no relation among histological type, except only histological grade and category show relations. Also, few histological features were found correlate with texture features; size, lymph nodes and tumor necrosis are the ones associate with some texture features.

摘要 I
Abstract II
致謝 III
Chapter 1 Introduction 1
1.1 Background 1
1.2 Research Motivation and Purpose 4
1.3 Literature Review 6
Chapter 2 Segmentation Method 8
2.1 Traditional Level Set Method 8
2.2 Distance Regularized Level Set Evolution (DRLSE) 9
2.3 Improved DRLSE (iDRLSE) 12
2.4 Combined DRLSE – Initialization with Classifying Texture Features for DRLSE 15
2.5 Automatic setting for iteration number 20
2.6 Post Processing for cDRLSE 21
Chapter 3 Feature Extraction Methods 22
3.1. Introduction 22
3.2. Statistical Texture Parameters 22
3.2.1. Basic Statistical Texture Parameters 22
3.2.2. Entropy 23
3.3. Fractal dimension 23
3.4. Gray Level Co-occurrence Matrix (GLCM) 24
3.4.1. Co-occurrence Matrix 24
3.4.2. Gray Level Co-occurrence Matrix (GLCM) 25
Chapter 4 Experiment Process and Results 30
4.1. Data Collecting 30
4.1.1. Pathological Dataset 30
4.1.2. Image Dataset 33
4.2. Segmentation Process 34
4.2.1. Building Dataset 34
4.2.2. Training and Classifying Dataset 36
4.2.3. Post Processing Stage 38
4.2.4. Evaluation Metrics 40
4.2.4.1. Misclassification Error (ME) 40
4.2.4.2. Relative foreground area error (RFAE) 40
4.2.4.3. Region Non-Uniformity (NU) 41
4.3. Segmentation Results 42
4.3.1. Initialization by Classifying Features’ Lattices 42
4.3.2. Demonstration of Segmentation Results 45
4.4. ROI Features Extraction 53
4.5. Features Extraction Result 54
4.6. Relationship Between Features and Pathological Type 57
5.1. Conclusion 65
5.2. Future Work 66
Reference 67
Appendix A Segmentation by Classifying Features 73
Appendix B Segmentation with DRLSE and Initial area in center region. 76

[1] 行政院衛生福利部, "民國102年死因統計年報," 2013
[2] H.J. Gdynia, H.P. Muller, A.C. Ludolph, H.Koninger, R. Huber, "Quantitative muscle ultrasound in neuromuscular disorders using the parameters 'intensity', 'entropy', and 'fractal dimension'," Eur J Neurol, vol. 16, pp. 1151-1158, 2009.
[3] D.R. Chen, R.F. Chang, C.J. Chen, M.F. Ho, S.J. Kuo, S.T. Chen, S.J. Hung, W.K. Moon, "Classification of breast ultrasound images using fractal feature," Clinical Imaging, vol. 29, pp. 235-245, 2005.
[4] B. Liu, H.D. Cheng, J.Huang, J. Tian, X. Tang, J. Liu, "Fully automatic and segmentation-robust classification of breast tumors based on local texture analysis of ultrasound images," Pattern Recognition, vol. 43, pp. 280-298, 2010.
[5] J. Canny, "A computational approach to edge detection," IEEE Trans. Pattern Anal. Machine Intell., vol. PAMI-8, pp. 679-698, 1986.
[6] R. Adams, L. Bischof, "Seeded region growing," IEEE Trans. Pattern Anal. Machine Intell., vol. 16, pp. 641-647, 1994.
[7] M. Kass, A. Witkin, D. Terzopoulos, "Snakes: Active contour models," International Journal of Computer Vision, vol. 1, pp. 321-331, 1988.
[8] C. Xu, J.L. Prince, "Snakes, shapes, and gradient vector flow," IEEE Trans. Image Process, vol. 7, pp. 359-369, 1998.
[9] S.C. Zhu, T.S. Lee, A.L. Yuille, "Region competition: unifying snakes, region growing, energy/Bayes/MDL for multi-band image segmentation," in Proc. IEEE 5th Int. Conf. Computer Vision, pp. 416-423, 1995.
[10] V. Caselles, R. Kimmel, G. Sapiro, "Geodesic active contours," International Journal of Computer Vision, vol. 22, pp. 61-79, 1997.
[11] R. Malladi, J. A. Sethian, B. C. Vemuri, "Shape modeling with front propagation: a level set approach," IEEE Trans. Pattern Anal. Machine Intell., vol. 17, pp. 158-175, 1995.
[12] S. Kichenassamy, A. Kumar, P. Olver, A. Tannenbaum, A. Yezzi, "Gradient flows and geometric active contour models," in Proc. IEEE 5th Int. Conf. Computer Vision, pp. 810-815, 1995.
[13] S. Osher, J.A. Sethian, "Fronts propagating with curvature-dependent speed: Algorithms based on Hamilton-Jacobi formulations," Journal of Computational Physics, vol. 79, pp. 12-49, 1988.
[14] X. Chenyang, A. Yezzi, J.L. Prince, "On the relationship between parametric and geometric active contours," in Proceedings of Thirty-Fourth Asilomar Conference on Signals, Systems and Computers, vol. 1, pp. 483-489, 2000.
[15] T.F. Chan, L.A. Vese, "Active contours without edges," IEEE Trans. Image Process, vol. 10, pp. 266-277, 2001.
[16] X. Yang, H.C. Yu, Y. Choi, W. Lee, B. Wang, J. Yang, H. Hwang, J.H. Kim, J. Song, B.H. Cho, H. You, "A hybrid semi-automatic method for liver segmentation based on level-set methods using multiple seed points," Comput Methods and Programs in Biomedicine, vol. 113, pp. 69-79, 2014.
[17] P.R. Bai, Q.Y. Liu, L. Li, S.H. Teng, J. Li, and M.Y. Cao, "A novel region-based level set method initialized with mean shift clustering for automated medical image segmentation," Computers in Biology and Medicine, vol. 43, pp. 1827-1832, 2013.
[18] M. Ammar, S. Mahmoudi, M.A. Chikh, A. Abbou, "Endocardial border detection in cardiac magnetic resonance images using level set method," Journal of Digital Imaging, vol. 25, pp. 294-306, 2012.
[19] A. Belaid, D. Boukerroui, Y. Maingourd, J.F. Lerallut, "Phase-based level set segmentation of ultrasound images," IEEE Trans. Information Technology in Biomedicine, vol. 15, pp. 138-147, 2011.
[20] J.E. Ball, L.M. Bruce, "Digital mammographic computer aided diagnosis (CAD) using adaptive level set segmentation," in Proc. 29th Annual Int. Conf. IEEE EMBS, pp. 4973-8, 2007.
[21] J.A. Sethian, "Level Set Methods and Fast Marching Methods: Evolving Interfaces in Computational Geometry, Fluid Mechanics, Computer Vision, and Materials Science," Second ed., Cambridge University Press, 1999.
[22] S. Osher, R. Fedkiw, "Level Set Methods and Dynamic Implicit Surfaces," Springer-Verlag New york, 2003.
[23] L. Chunming, X. Chenyang, G. Changfeng, M.D. Fox, "Level set evolution without re- initialization: a new variational formulation," in Proc. IEEE Conf. Computer Vision and Pattern Recognition, vol. 1, pp. 430-436, 2005.
[24] R. Kimmel, A. Amir, A. M. Bruckstein, "Finding shortest paths on surfaces using level sets propagation," IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 17, pp. 635-640, 1995.
[25] L. Chunming, X. Chenyang, G. Changfeng, M. D. Fox, "Distance Regularized Level Set Evolution and Its Application to Image Segmentation," IEEE Trans. Image Processing, vol. 19, pp. 3243-3254, 2010.
[26] G. Castellano, L. Bonilha, L.M. Li, F. Cendes, "Texture analysis of medical images," Clinical Radiology, vol. 59, pp. 1061-1069, 2004.
[27] R. Sivaramakrishna, K.A. Powell, M.L. Lieber, W.A. Chilcote, R. Shekhar, "Texture analysis of lesions in breast ultrasound images," Computerized Medical Imaging and Graphics, vol. 26, pp. 303-307, 2002.
[28] R.M. Haralick, K. Shanmugam, I.H. Dinstein, " Textural features for image classification," IEEE Trans. System, Man Cybernetics, vol. SMC-3, pp. 610-621, 1973.
[29] H.D. Cheng, J. Shan, W. Ju, Y. Guo, L. Zhang, "Automated breast cancer detection and classification using ultrasound images: A survey," Pattern Recognition, vol. 43, pp. 299-317, 2010.
[30] B.B. Mandelbrot, "The Fractal Geometry of Nature. Revised edition," W.H. Freeman and Company, 1983.
[31] X. Qin, Z. Cong, B. Fei, "Automatic segmentation of right ventricular ultrasound images using sparse matrix transform and a level set," Phys. Med. Biol., vol. 58, pp. 7609-7624, 2013.
[32] Z. Tao, H.D. Tagare, "Tunneling descent level set segmentation of ultrasound images," Inf. Process Med. Imaging, vol. 19, pp. 750-761, 2005.
[33] Q. Lin, S. Liu, S.S. Parajuly, Y. Deng, L. Boroczky, S. Fu, Y. Wu, Y. Pen, "Ultrasound lesion segmentation using clinical knowledge-driven constrained level set," in Proc. 35th Annual Int. Conf. IEEE EMBC, pp. 6067-6070, 2013.
[34] R.F. Chang1, W.J. Wu1, W.K. Moon, D.R. Chen, "Automatic ultrasound segmentation and morphology based diagnosis of solid breast tumors," Breast Cancer Research and Treatment, vol. 89, pp. 179-185, 2005
[35] C.M. ZHU, G.C. GU, H.B. LIU, J. SHEN, H. YU, "Segmentation of ultrasound images based on texture feature and graph cut," in Proc. IEEE Conf. Computer Science and Software Engineering, pp. 795-798, 2008.
[36] J. Gomes, O. Faugeras, “Reconciling distance functions and Level Sets,” Journal of Visual Communication and Image Representation, vol. 11, pp. 209-223, 2000.
[37] 王鈞奕, “改良式結合弱邊界資訊的距離正規化水平及演化演算法之研究, “ 國立清華大學電機工程學系, 2013.
[38] C. Tomasi, R. Manduchi, "Bilateral filtering for gray and color images," in Proc. 6th IEEE Conf. Computer Vision, pp. 839-846, 1998.
[39] 吳俊緯, “結合圖形邊界地圖資訊的距離正規化水平集演算法-用於全自動的指骨輪廓圈選研究, “ 國立清華大學電機工程學系, 2014.
[40] K. Wu, Y. Wang, Y. Pan, J. Zhang, B. Qian, C. Chang, "Classifying uterine myoma and adenomyosis based on ultrasound image fractal and texture features," in Proc. 27th Annual Int. Conf. IEEE EMBS, vol. 2, pp. 1790-1793, 2005.
[41] F.J. McEvoy, J. Tomkiewicz, J.G. Stottrup, J.L. Overton, C. McEvoy, E. Svalastoga, "Determination of fish gender using fractal analysis of ultrasound images," Vet Radiol. Ultrasound, vol. 50, pp. 519-524, 2009.
[42] W. Gómez, W.C.A. Pereira, A.F.C. Infantosi, "Analysis of Co-Occurrence Texture Statistics as a Function of Gray-Level Quantization for Classifying Breast Ultrasound," IEEE Trans. Medical Imaging, vol. 31, pp. 1889-1899, 2012.
[43] D.A. Clausi, "An analysis of co-occurrence texture statistics as a function of grey level quantization," Canadian Journal of Remote Sensing, vol. 28, pp. 45-62, 2002.
[44] L.K. Soh, C. Tsatsoulis, "Texture analysis of SAR sea ice imagery using gray level co-occurrence matrices," IEEE Trans. Geoscience and Remote Sensing, vol. 37, pp. 780-795, 1999.
[45] Matlab R2014.a, Natick, The MathWorks Inc., 2014.
[46] M. Bevk, I. Kononenko, "A statistical approach to texture description of medical images:Apreliminary study," in Proc. 15th IEEE Conf. Symposium CBMS, pp. 239-244, 2002.
[47] American Cancer Society, "How is breast cancer classified?," 2014, http://www.cancer.org/cancer/breastcancer/detailedguide/breast-cancer-classifying.
[48] American Cancer Society, "Understanding Your Pathology Report: Breast Cancer," 2014, http://www.cancer.org/treatment/understandingyourdiagnosis/understandingyourpathologyreport/breastpathology/breast-cancer-pathology.
[49] 行政院衛生署國民健康局, "台灣癌症登記短表摘錄手冊," 2008, http://tcr.cph.ntu.edu.tw/uploadimages/Coding_Manual_33_200805.pdf.
[50] American Cancer Society, "Understanding your mammogram report – BI-RADS categories," 2014, http://www.cancer.org/treatment/understandingyourdiagnosis/examsandtestdescriptions/mammogramsandotherbreastimagingprocedures/mammograms-and-other-breast-imaging-procedures-mammo-report
[51] Breastcancer.org, "Your Diagnosis," 2015, http://www.breastcancer.org/symptoms/diagnosis.
[52] C.C. Liu, P.C. Chung, "Objects extraction algorithm of color image using adaptive forecasting filters created automatically," International Journal of Innovative Computing, Information and Control, vol. 7, pp. 5771-5787, 2010.
[53] M. Sezgin, "Survey over image thresholding techniques and quantitative performance evaluation," Journal of Electronic imaging, vol. 13, pp. 146-168, 2004.