(3.236.222.124) 您好!臺灣時間:2021/05/13 21:27
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:鄭勝文
研究生(外文):Sheng-Wen Zheng
論文名稱:根植於動態輪廓之乳房腫瘤影像擷取
論文名稱(外文):A Dynamic contours-Based Breast Tumors Segmentation Algorithm for Digital Mammograms
指導教授:劉正忠劉正忠引用關係
指導教授(外文):Chen-Chung Liu
學位類別:碩士
校院名稱:國立勤益科技大學
系所名稱:電子工程系
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2013
畢業學年度:101
語文別:中文
論文頁數:72
中文關鍵詞:乳房X光攝影檢查術腫瘤側面乳房X光片乳房X光片影像分析協會隨機漫步
外文關鍵詞:mammographycancermedio-lateral obliqueMammogram Image Analysis Societyrandom walk
相關次數:
  • 被引用被引用:0
  • 點閱點閱:79
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
乳腺癌是由乳腺導管或腺泡細胞的異常分裂和繁殖所增生的惡性腫瘤,而乳腺癌早期發現早期治療的治癒率會越高。檢測初期的乳腺癌,乳房X光片檢測是目前效率最高、成本最低且經常被拿來檢測乳房腫瘤的技術。本文的主要目的是提出一個準確和有效的演算法,擷取側面乳房X光片(Mediolateral-oblique, MLO)的乳腺腫瘤。
對於已移除胸大肌之數位乳房X光片,先使用4次Otsu閾值法進行二值化,再使用三層特徵做過濾,第一層先針對過於細小的乳腺區域及較大的肌肉部分,分別做刪除過大及過小的面積,第二層的過濾是根據乳腺管的細長特性,故將主副軸長比過於懸殊的面積移除。
第三層利用像素與像素之間灰階值的數理差異,計算兩相對位置的像素,在不同的距離及不同角度情況下,統計所發生次數而得到一個共發生矩陣,再使用共發生矩陣為基礎去計算一些數量的特徵值,藉由腫瘤和肌肉組織在特徵值上的不同,使用常態分佈的離群值協助判斷,可過濾出只剩腫瘤的二值化影像。然後將腫瘤及其附近的矩形區域擷取獨立出來,取出邊緣像素以作為初始輪廓。
接下來的第三個部份,利用梯度向量流蛇型 (Gradient vector field snake, GVF-Snake) 輪廓及隨機漫步(Random walk)找出腫瘤輪廓。GVF-Snake會對初始輪廓做多次形變,利用一個適當的外部力計算方式,使得GVF-Snake在每次的形變中,可以接近腫瘤的真實輪廓,最終輪廓所圍起的區域就是最接近真實腫瘤部份的區域。Random walk計算像素點和種子點間的機率問題,由最大狄利克雷邊界(Dirichlet boundary)機率找出腫瘤的二元遮罩影像,再使用索貝爾(Sobel)邊緣檢測找出腫瘤邊緣。
比對GVF-Snake和Random walk兩種動態輪廓邊緣擷取方法的實驗結果並與專家所劃分出的腫瘤輪廓範圍做確認,可以發現本文所提出的方法能擷取出正確的腫瘤區域。

Breast cancer is one of the most common cancers in women. Screening and early detection is essential for better outcomes in breast cancer treatment. For screening of breast cancer, mammography has been utilized widely in various clinical settings for many years. With regards to the mammogram processing, extracting the region of interest (the breast tumour) accurately from a mammogram is one kernel stage of the mammogram processing. It significantly influences the overall analysis accuracy and processing speed of the mammogram.
The main goal of this paper is to propose an accurate and efficient algorithm of breast tumor extraction on the medio-lateral oblique (MLO) mammograms. The proposed algorithm consists of three stages. At the first stage, we adapt interactive Otsu thresholding scheme to binarize the input breast region, then to delete those too large objects (muscle parts) and too small objects (noises) from the binarized breast region to obtain the first stage object map. At the second stage, we filter out slim objects (ductal) by evaluating the ratio of major axis to minor axis of each object in the first stage object map to obtain the second stage object map. At the third stage, the co-occurrence matrix of each object of the second stage object map is constructed to evaluate the corresponding co-occurrence matrix feature values. Moreover, the third moment, compactness, entropy, extent and solidity of each object are calculated at the same time. Above items are scored according each item’s statistic distribution. The object with having highest total score is considered as a tumor. At the fourth stage, gradient vector flow snake contour (GVF-Snake) and random walk (Random walk) are employed on the tumor segmented in the third stage to accurately segment breast tumor.
The presented algorithm is tested on the digital mammograms from the Mammogram Image Analysis Society (MIAS) database. The experimental results show that the tumor extracted by the presented algorithm approximately follows that extracted by the biopsy of MIAS. In the future we will develop a high performance breast mass analysis based on accurate breast tumor segmentation to power the computer aided detection of breast cancer.

摘要 i
Abstract iii
致謝 v
目錄 vi
表目錄 viii
圖目錄 ix
第一章 緒論 1
1.1 研究動機 1
1.2 歷史文獻 1
1.3 論文架構 3
第二章 相關原理 5
2.1 Otsu's門檻值法(Otsu's method ) 5
2.2 統計矩(Statistics moment) 7
2.3 共發生矩陣(Co-occurrence Matrix) 9
2.4 梯度向量流蛇型輪廓(GVF-snake) 13
2.4.1 蛇型輪廓 13
2.4.2 梯度向量流 16
2.5 隨機漫步(Random walk) 17
第三章 X光片腫瘤擷取演算法 19
3.1 初始二元影像標記 20
3.2 初階過濾 25
3.3 第二階過濾 28
3.3.1 緊密度排序 30
3.3.2 共發生矩陣特徵值極值排序 30
3.3.3 熵、第三矩排序 32
3.3.4 特徵值離群值排序 34
3.3.5 標記區塊對邊界盒面積比(Extent,拉伸比) 37
3.3.6 凸形封包面積比(Solidity,稠度) 38
3.3.7 總積分計算 39
3.4 擷取腫瘤邊緣 40
第四章 實驗結果 44
4.1 乳房腫瘤擷取準確率評估數據 47
4.2 腫瘤輪廓擷取圖 56
第五章 結論 68
參考文獻 69

[1] R. D. Yapa, K. Harada, “Breast skin-line estimation and breast segmentation in mammograms using fast-marching method,” International Journal of Biomedical Sciences, Vol. 3, No. 1, pp 54-62, 2008.
[2] S. M. Kwok, R. Chandrasekhar, Y, Attikiouzel, M. T. Rickard, “Automatic pectoral muscle segmentation on mediolateral oblique view mammograms,” IEEE Trans. on Medical Imaging, Vol. 23, No. 9, pp.1129-1140, 2004.
[3] T.S. Subashini, V. Ramalingam, S. Palanivel, “Automated assessment of breast tissue density in digital mammograms,” Computer Vision and Image Understanding, Vol. 114, No. 1, pp. 33-43, 2010.
[4] J. Nagi, S. A. Kareem, F. Nagi, S. K. Ahmed, “Automated breast profile segmentation for ROI detection using digital mammograms,” 2010 IEEE EMBS Conference on Biomedical Engineering &; Sciences, pp.87- 92, 2010.
[5] G. Kom, A. Tiedeu, M. Kom, “Automated detection of masses in mammograms by local adaptive thresholding,” Computers in Biology and Medicine, Vol. 37, pp. 37-48, 2007.
[6] A. R. Domiınguez, A. K. Nandi, “Detection of masses in mammograms via statistically based enhancement, multilevel thresholding segmentation, and region selection,” Computerized Medical Imaging and Graphics, Vol. 32, pp. 304-315, 2008.
[7] Zhang, Y., Tomuro, N., Furst, J. and Raicu, D. “Image Enhancement and Edge-based Mass Segmentation in Mammogram,” In Proceedings of the SPIE Symposium on Medical Imaging (SPIE-2010), pp.13-18, 2010.
[8] Y. Yuan, M. L. Giger, H. Li, K. Suzuki, and C. Sennett, “A Dual-stage Method for Lesion Segmentation on Digital Mammograms “, Med. Phys. Vol. 34, No. 11, pp. 4180-4193, 2007.
[9] E. Song, L. Jiang, R. Jin, L. Zhang, Y. Yuan, and Q. Li, “Breast Mass Segmentation in Mammography Using Plane Fitting and Dynamic Programming”, Academic radiology, Vol. 16, No. 7, pp. 826-35, 2009.
[10] A. R. Domínguez1, and A. K. Nandi, “Toward Breast Cancer Diagnosis Based on Automated Segmentation of Masses in Mammograms,” Pattern Recognition, Vol. 42, pp. 1138-1148, 2009.
[11] L. Jiang, E. Song, X. Xu, G. Ma, and B. Zhang, “Automated Detection of Breast Mass Spiculation Levels and Evaluation of Scheme Performance,” Acad Radiol, pp. 15:1534-1544, 2008.
[12] W. Xu, S. Xia, M. Xiao, and H. Duan, “A Model-based Algorithm for Mass Segmentation in Mammograms,” Engineering in Medicine and Biology 27th Annual Conference, pp. 2543- 2546, 2005.
[13]蔡忠諺,基於GVF-Snake模型的乳房區域擷取方法,碩士論文,國立中興大學,2010。
[14]莊凱文,多元回歸分析之色彩轉換系統,碩士論文,國立勤益科技大學,2012。
[15]謬紹綱,數位影像處理−應用MATLAB,東華書局,2005。
[16]劉源俊教授,釋說新語之十八-什麼是moment,東吳大學/中央大學,http://sci.ncu.edu.tw/wp-content/uploads/s18.pdf。
[17]林宸生教授,數位信號與光電量測應用於機電整合系統之工程教育網,逢甲大學,http://140.134.32.129/scteach/mech/text/yin04_5.html。
[18] 廖友千,紋路特徵值分析應用於乳房X光攝影之腫瘤偵測,碩士論文,國立成功大學,2002。
[19]于南書,最佳特徵選擇:乳房X光片腫瘤偵測,碩士論文,國立成功大學,2004。
[20]黃俊忠,利用X光乳房攝影產生之紋理特徵影像在腫瘤偵測上之研究,碩士論文,國立中央大學,2006。
[21]林冠良,應用灰階共生矩陣於彩色濾光片瑕疵檢測之研究,碩士論文,義守大學,2009。
[22]李綺芳,結合主成份分析與模組化半徑基底函數類神經網路於影像語意內容分析問題,碩士論文,國立雲林科技大學,2007。
[23]李德治、童惠玲,應用統計學,博碩文化,2008。
[24]邱彥儒,主動輪廓模組影像分割技術之自動化體積定量研究,碩士論文,中臺科技大學,2006。
[25]林松柏,傳統、小波理論與動態輪廓模式之波浪影像邊緣偵測處理,碩士論文,國立成功大學,2005。
[26] M. Kass, A. Witkin, and D. Terzopoulos , “Snakes: Active contour models,”International Journal of Computer Vision, 1, no. 4, 321-331, 1987.
[27] C. Xu and J.L. Prince , “Snakes, shapes, and gradient vector flow,” IEEETrans. on Image Processing, 7, no. 3, 359-369 ,1998
[28] Leo Grady, “Random Walks for Image Segmentation,” IEEE transactions on pattern analysis and machine intelligence, Vol. 28, No. 11, November 2006.
[29] R. Modzelewski, E. Janvresse, T. de la Rue, P. Vera , “Brain perfusion heterogeneity measurement based on Random Walk algorithm: Choice and influence of inner parameters,” Computerized Medical Imaging and Graphics, Vol. 34, pp. 289–297, 2010.
[30] H. K. Hsiao, C. C. Liu, C. Y. Yu, S. W. Kuo, S. S. Yu, “A Novel Optic Disc Detection Scheme on Retinal Images,” Expert System with Applications, Vol. 39, pp. 10600–10606, 2012.
[31] H. Abed, A. Roya, S. Jamshid, “ReliefF-Based Feature Selection for Automatic Tumor Classification of Mammogram Images,” Machine Vision and Image Processing (MVIP), pp. 1 – 5, 2011.
[32] M. Karnan, K. Thangavel , “Automatic detection of the breast border andnipple position on digital mammograms using genetic algorithm for asymmetryapproach to detection of microcalcifications.” Computer Methods and Programs inBiomedicine, 87, 12-20 ,2007.
[33] U. Bick, M.L. Giger, R.A. Schmidt, R.M. Nishikawa, D.E. Wolverton, K. Doi , “Automated segmentation of digitized mammograms.” Acad. Radiol, 2, 1-9 ,1995.

連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔