乳房超音波影像(BUS)腫瘤切割一直以來都是很多人探討的議題，但它並不像其他影像有鮮明的輪廓，而且具有大量的雜訊干擾，使用傳統的邊緣偵測並不能有效的達到腫瘤切割的目的，因此本論文結合影像形態學運算和DRLSE提出了一套全自動的BUS影像腫瘤切割演算法，針對徐詠璿學姐提出之cDRLSE BUS腫瘤切割法進行改善。cDRLSE最大問題在於耗費大量的時間尋找initial contour，主要因為該方法採用GLCM之影像特徵抽取和SVM訓練之故，因此本論文提出一個新的初始輪廓找法，利用影像本身的結構，配合形態學操作與Ranking，可得到低時間成本與有效的初始輪廓。接著利用此輪廓進行DRLSE的演化，將超音波乳房影像的腫瘤位置以及輪廓圈選出來。為了評估切割結果的好壞，我們搭配ME、RFAE以及MHD三種錯誤測量指標，與外擴式DRLSE法、內縮式DRLSE法和cDRLSE進行比較，發現本論文提出方法的結果基本上是優於外擴法與內縮法，證明了初始輪廓對於DRLSE演化的重要性，而跟cDRLSE相比之下三種錯誤測量指標則好壞互見，推測可能是由於Ranking失敗或cDRLSE有額外做後處理來改善腫瘤切割的準確度所造成，但在時間的花費上本論文提出之方法比cDRLSE明顯減少了許多，也省去了做後處理的步驟。
本論文提出之切割方法有下述幾個優點：一為全自動的乳房腫瘤切割，不需要額外再手動設定初始位置；二為尋找初始輪廓之方法相當快速，若腫瘤內部平滑均勻，其初始輪廓則會很貼近腫瘤真實邊界；三為經過DRLSE之結果，若初始輪廓大部分落於ground truth內，與其他方法相比，確實能更靠近手繪腫瘤真實邊界。

Lesion segmentation for breast ultrasound images has been studied by many people, but it is hard to achieve using traditional edge detection because ultrasound images don’t have sharp contours and exist a lot of noise. In this thesis, we propose an automatic segmentation method for breast ultrasound images which combines morphological image processing and distance-regularized level-set evolution method (DRLSE), and improve combined DRSLE (cDRLSE) proposed by Yung-Hsuan Hsu. The most serious problem of cDRLSE is that it spend a lot of time on obtaining initial contour through the operation which applied the texture features (e.g., gray level co-occurrence matrix (GLCM)) for support vector machine (SVM), therefore we propose a new method based on morphological operations to search for initial contour which is effective and efficient. Using the new initial contours, we can capture the tumor area more precisely after applying DRLSE. To evaluate the result of segmentation, we compare it with expansion DRLSE method, contraction DRLSE method and cDRLSE method using three evaluation metrics, including misclassification error (ME), relative foreground area error (RFAE) and modified Hausdorff distance (MHD). We find that the proposed method is basically better than expansion DRLSE method and contraction DRLSE method which confirms the importance of initial contour to DRLSE. However, it is better than cDRLSE method in RFAE but worse than cDRLSE in ME and MHD, probably due to ranking failure in the proposed method or improved segmentation accuracy of post processing in cDRLSE method; even so, the proposed method not only spends less time obviously but also has no need to apply post processing.
The proposed method has the following properties:
1.A fully automatic segmentation method for breast ultrasound images which has no need to set initial contour manually.
2.The way to obtain initial contour is efficient, and moreover if the tumor is smooth, the initial contour will be close to the tumor’s real boundaries.
3.Compared with other methods, the segmentation result of the proposed method is truly closer to ground truth image if the proposed initial contour mostly lies inside the ground truth.

摘要 i
Abstract ii
致謝 iv
目錄 v
圖目錄 vii
表目錄 x
第一章 簡介 1
1.1 前言 1
1.2文獻回顧 1
1.2.1 Maximum-likelihood Method 2
1.2.2 cDRLSE Method 3
1.3研究目的 6
1.4論文架構 6
第二章 理論與方法 7
2.1系統流程圖 7
2.2 BUS Preprocessing 8
2.2.1 Image Cropped 8
2.2.2 Speckle Reduction 9
2.3 Initial Contour Searching Based on Morphology 11
2.3.1 Morphological Image Processing 11
2.3.1.1侵蝕Erosion 11
2.3.1.2膨脹Dilation 13
2.3.1.3斷開Opening 15
2.3.1.4閉合Closing 15
2.3.1.5補集Complement 16
2.3.1.6形態學重建Morphological Reconstruction 18
2.3.1.7區域最小Regional Minima 21
2.3.2 Search for initial contour 22
2.4 DRLSE Segmentation 28
2.4.1傳統Level Set 28
2.4.2距離正規化水平集演化(DRLSE) 31
2.4.3 Edge indicator g 33
2.4.4 DRLSE疊代次數之自動化 35
第三章 實驗結果 37
3.1 Data set 37
3.2錯誤測量指標(MOEs) 42
3.2.1 ME 42
3.2.2 RFAE 42
3.2.3 MHD 43
3.3實驗結果與討論 43
3.3.1初始輪廓 44
3.3.2切割結果 52
3.3.2.1 Proposed method result 52
3.3.2.2 Compare with DRLSE 54
3.3.2.3 Compare with cDRLSE 58
第四章 結論 61
4.1 結論 61
4.2 未來展望 62
參考文獻 63

[1]行政院衛生福利部, "民國103年死因統計年報, " 2014.
[2]Huang, Yu-Len, and Dar-Ren Chen. "Watershed segmentation for breast tumor in 2-D sonography." Ultrasound in medicine & biology 30.5 (2004): 625-632.
[3]Gomez, W., et al. "Computerized lesion segmentation of breast ultrasound based on marker-controlled watershed transformation." Medical physics 37.1 (2010): 82-95.
[4]Huang, C‐S., et al. "Microcalcifications of non‐palpable breast lesions detected by ultrasonography: correlation with mammography and histopathology." Ultrasound in obstetrics & gynecology 13.6 (1999): 431-436.
[5]Madabhushi, Anant, and Dimitris N. Metaxas. "Combining low-, high-level and empirical domain knowledge for automated segmentation of ultrasonic breast lesions." IEEE transactions on medical imaging 22.2 (2003): 155-169.
[6]Hao, Xiaohui, et al. "Segmenting high-frequency intracardiac ultrasound images of myocardium into infarcted, ischemic, and normal regions." IEEE transactions on medical imaging 20.12 (2001): 1373-1383.
[7]Sarti, A., et al. "Maximum likelihood segmentation with Rayleigh distribution of ultrasound images." Computers in Cardiology, 2004. IEEE, 2004.
[8]徐詠璿, "結合影像特徵分析的距離正規化水平集演算法-用於乳房超音波影像的腫瘤範圍圈選, " 國立清華大學電機工程學系, 2016.
[9]Li, Chunming, et al. "Distance regularized level set evolution and its application to image segmentation." IEEE transactions on image processing 19.12 (2010): 3243-3254.
[10]Castellano, G., et al. "Texture analysis of medical images." Clinical radiology 59.12 (2004): 1061-1069.
[11]Gómez, Walter, W. C. A. Pereira, and Antonio Fernando C. Infantosi. "Analysis of co-occurrence texture statistics as a function of gray-level quantization for classifying breast ultrasound." IEEE transactions on medical imaging 31.10 (2012): 1889-1899.
[12]Clausi, David A. "An analysis of co-occurrence texture statistics as a function of grey level quantization." Canadian Journal of remote sensing 28.1 (2002): 45-62.
[13]Soh, L-K., and Costas Tsatsoulis. "Texture analysis of SAR sea ice imagery using gray level co-occurrence matrices." IEEE Transactions on geoscience and remote sensing 37.2 (1999): 780-795.
[14]Vapnik, Vladimir. The nature of statistical learning theory. Springer Science & Business Media, 2013.
[15]Yu, Yongjian, and Scott T. Acton. "Speckle reducing anisotropic diffusion." IEEE Transactions on image processing 11.11 (2002): 1260-1270.
[16]林宏桀, et al. "乳房超音波腫瘤切割之研究. " (2008).
[17]Perona, Pietro, and Jitendra Malik. "Scale-space and edge detection using anisotropic diffusion." IEEE Transactions on pattern analysis and machine intelligence 12.7 (1990): 629-639.
[18]Gonzalez, Rafael C., and Richard E. Woods. "Digital image processing." Prentice Hall (2008).
[19]Haralick, Robert M., Stanley R. Sternberg, and Xinhua Zhuang. "Image analysis using mathematical morphology." IEEE transactions on pattern analysis and machine intelligence 4 (1987): 532-550.
[20]Vincent, Luc. "Morphological grayscale reconstruction in image analysis: applications and efficient algorithms." IEEE transactions on image processing2.2 (1993): 176-201.
[21]Shan, Juan. A fully automatic segmentation method for breast ultrasound images. Diss. Utah State University, 2011.
[22]Osher, Stanley, and James A. Sethian. "Fronts propagating with curvature-dependent speed: algorithms based on Hamilton-Jacobi formulations." Journal of computational physics 79.1 (1988): 12-49.
[23]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).
[24]Osher, Stanley, and Ronald Fedkiw. Level set methods and dynamic implicit surfaces. Vol. 153. Springer Science & Business Media, 2006.
[25]Sussman, Mark, et al. "An improved level set method for incompressible two-phase flows." Computers & Fluids 27.5 (1998): 663-680.
[26]Li, Chunming, et al. "Level set evolution without re-initialization: a new variational formulation." 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05). Vol. 1. IEEE, 2005.
[27]Li, Chunming, et al. "Distance regularized level set evolution and its application to image segmentation." IEEE transactions on image processing 19.12 (2010): 3243-3254.
[28]Sezgin, Mehmet. "Survey over image thresholding techniques and quantitative performance evaluation." Journal of Electronic imaging 13.1 (2004): 146-168.
[29]Liu, Chen-Chung, and Pei-Chung Chung. "Objects extraction algorithm of color image using adaptive forecasting filters created automatically." International Journal of Innovative Computing, Information and Control 7.10 (2011): 5771-5787.