臺灣博碩士論文加值系統

本論文的研究目的在於對乳房超音波影像進行自動化的腫瘤切割，以一組接近腫瘤邊緣的參數做為起始輪廓，改進傳統的人工圈選，並加強整體輪廓的收斂效果。在本論文中，我們使用主動式輪廓模型(active contour model)的能量原理，運用一組封閉曲線作為起始輪廓，利用主動式輪廓模型演算法中的內部能量以及外部能量對超音波影像進行輪廓偵測，再結合梯度向量流(Gradient Vector Flow, GVF)中梯度擴散特性，改善了傳統主動式輪廓模型能量中所沒有包含的梯度向量資訊，以強化邊界能量的引導。在影像邊緣資訊方面，本研究在影像邊緣資訊(edge map)中加入了紋理的特徵，將一幅影像分為不同區域進行紋理的量化及直方圖比較，以區域性二位元樣型(Local binary pattern, LBP)將圖片依照紋理的特徵進行區域合併及分割，以達到輪廓切割的目的。在實驗結果方面，本篇論文實驗了包含良性與惡性的超音波腫瘤影像，對於各種形狀的腫瘤區域進行改良性的主動式輪廓切割演算法，切割結果可被接受比例可達95%，有效達到自動化切割準確性提升的目標。

In this study, we propose a new active contour model based segmentation method for breast tumor in ultrasound images. An initial contour is marked automatically and the corresponding parameters are used to enhance the convergence effect of overall contour. To improve the poor convergence in concave boundaries of traditional active contour model, the gradient vector flow is considered as the guidance of edged energy. The additional information makes the contour converge to the correct boundary closely. Moreover, local binary pattern (LBP) is used as texture features at the stage of evaluating edge information. Contour of the tumor is then estimated by internal energy and external energy of the active contour model according to the proposed features. Experimental results show that the overall accepted rate is up to 95% both on benign and malignant tumors of the ultrasound images. The proposed method can achieve the purpose of enhancing the accuracy of automatic segmentation effectively.

書名頁 i
論文口試委員審定書 ii
授權書 iii
中文摘要 iv
英文摘要 v
誌謝 vi
目錄 vii
表目錄 viii
圖目錄 ix

第一章 緒論 1
1.1研究動機與目的 3
1.2研究方法概述 5
1.3研究流程架構 6
第二章 文獻探討 9
2.1乳房超音波影像切割 10
2.2主動式輪廓模型(Active Contour Models) 11
2.3梯度向量流模式(Gradient Vector Flow,GVF) 19
2.4區域性二位元樣型(Local Binary Pattern, LBP) 23
第三章 研究方法 26
3.1影像前處理 27
3.1.1去除雜訊 27
3.1.2鏈碼 29
3.1.3初始輪廓指定 30
3.2外部能量定義 31
3.3以區域紋理為基礎的邊緣檢測 32
3.4改良式的主動式輪廓切割演算法 35
第四章 實驗結果評估 38
4.1評估指標定義 38
4.2實驗切割結果 41
4.3實驗數據 44
4.4實驗結果比較 49
第五章 結論與未來展望 60
參考文獻 62

[1]財團法人乳癌防治基金會，http://www.breastcf.org.tw/breast.htm
[2]M. Kass, A. Witkin, and D. Terzopoulos, “Snake:Active contour models,”International Journal of Computer, vol.1, no.4, pp. 321-331, 1988.
[3]A. Amini, T. Weymouth, and R. Jain, “Using dynamic programming for solving variational problems in vision,” IEEE Trans. on Pattern Anal. Machine Intell., vol. 12, pp. 855-867, 1990.
[4]C. Y. Xu, and J. L. Prince, “Snake Shapes and Gradient Vector Flow,” IEEE Trans. on Image Processing., vol.7, no.3, 1988.
[5]T. F. Chan, and L.A. Vese, “Active Contours Without Edges,” IEEE Trans. on Image processing, vol. 7, pp. 266-277, 2001.
[6]M. Rousson, T. Brox, and R. Deriche, “Active Unsupervised Texture Segmentation on a Diffusion Based Feature Space,” Computer Vision and Pattern Recognition, CVPR 2003. IEEE Computer Society Conference, 18-20 June, 2003, vol. 2, pp. 699-704.
[7]H. W. Park, T. Schoepflin, and Y. Kim, “Active Contour Model with Gradient Directional Information: Directional Snake,” IEEE Trans. on Circuits and System for Video Technology, vol. 11, no. 2, pp. 252-256, 2001.
[8]M. A. Savelonas, D. K. Iakovidis, and D. E. Maroulis, “An LBP-Based Active Contour Algorithm for Unsupervised Texture Segmentation,” Pattern Recognition, ICPR 2006. IEEE International Conference, 20-24 Aug., 2006, Hong Kong, China, vol. 2, pp. 279-282.
[9]C. Sagiv, N. A. Sochen, and Y. Y. Zeevi, “Integrated active contours for texture segmentation,” IEEE Trans. on Image Processing., vol.15, no. 6, pp. 1633-1645, 2006.
[10]T. Ojala, M. Pietikainen, and D. Harwood, “A Comparative Study of Texture Measures with Classification based on Feature Distributions,” Pattern Recognition, vol. 29, pp. 51-59, 1996.
[11]T. Ojala, M. Pietikainen, and D. Harwood, “Unsupervised texture segmentation using feature distributions,” Pattern Recognition, vol. 32, pp. 477-486, 1999.
[12]ACS, “Breast Cancer Facts and Figures 2007-2008,” American Cancer Society 2008.
[13]A. Madabhushi, D. N. Metaxas, “Combining Low-,High-level and Empirical Domain Knowledge for Automated Segmentation of Ultrasonic Breast Lesions,” IEEE Trans. Med Imaging, vol. 22, pp. 155-169, 2003.
[14]S.-F. Huang, R.-F. Chang, “Characterization of Spiculation on Ultrasound Lesions,” IEEE Trans. Med Imaging, vol. 23, 2004.
[15]J. Shen, Y. Wang, “Boundary Extraction of Breast Ultrasonic Images, ” Proceedings of the 28th IEEE EMBS Annual International Conference, Aug 30-Sept 3,2006,New York City, USA.
[16]Y. Wang, J. Shen, Y. Guo, and W. Wang,.“Computerized Classification of Breast Tumors with Morphologic and Texture Features of Ultrasonic Images,” CBMS, pp. 23-28, 2008.
[17]謬紹綱，數位影像處理，第二版，688-689頁，九十五年。
[18]J. F. Canny, “A Computational Approach to Edge Detection, ” IEEE Trans. on Pattern Anal. Machine Intell., vol. 8, no. 6, pp. 679-698, 1986.
[19]陳幹旻，「以特徵分佈為基礎之彩色紋理影像切割」，元智大學電機暨資訊工程研究所碩士論文，民國88年。