臺灣博碩士論文加值系統

隨著生物醫學應用影像技術來協助判讀的需要增加，以往需要大量顯微鏡操作及人力觀察辨識的工作也日漸需要一套標準化的系統來協助處理，以降低在時間及人力上的大量耗費。以微細胞核（Micronuclei ,MN）為例，在解讀的過程中除了得耗用大量人力判讀外，其辨識的結果也還存在不少變異性，因為這與專業訓練成熟度存在密切關係，研究人員的主觀看法皆可能會影響到最後的診斷結果，因此極需客觀的資訊工具來數位化這些過程，讓第三者也可以在事後以最迅速的方式來再次確認及給予建議。所以本研究主要運用影像技術及類神經網路對細胞影像進行自動化細胞計數，過程中會先將玻片上的影像自動數位化到電腦中，針對取得的玻片影像進行處理，將原影像轉成灰階圖再運用Otsu演算法自動尋找最佳閥值（Optimal Threshold）進行影像分割，並視情況結合形態學（Morphology）的技巧來消除雜訊，保留特徵型態，接著根據取得的輪廓資訊去計算影像形狀特徵值及將此輪廓套回原影像取得內部結構的特徵值，最後再將這些特徵值交由類神經網路的貝氏網路（Bayesian Network）進行型態辨識，找出辨識微細胞核的最佳特徵值組合。簡言之，步驟分別為：（1）玻片數位化，（2）色彩轉換，（3）影像分割，（4）初步細胞個數估算，（5）特徵值擷取，（6）找出最佳的特徵值組合，（7）自動化型態辨識。另外，本研究也提供了玻片影像還原程式，以方便研究人員可以隨時調閱細胞影像檔出來驗證診斷結果，並瀏覽其鄰近的細胞影像。

In recent decades, demand for applying image-processing techniques to help biomedical diagnosis has increased rapidly. The traditional system processes, take Micronuclei (MN) scoring for example, use human eyes with microscopic which not only take too much time and human resources, but with unstable results for recognition. In addition, the results are highly dependant on the expertise of the researcher who conducted the experiment. As a result, an objective tool to digitize the process and automatic recognition will eliminate the subjective judgment from the researcher. Besides, the digitized slides and results can also be retrieved and verified anytime in the future conveniently.
In this study, we propose to integrate image processing techniques and Neural Networks to the automatic counting of cells. First, we digitize images on the slide automatically and store them to the computer. Then we convert the original images into gray scales and use Otsu algorithm to find the optimal threshold for image segmentation. In some cases, it is also required to apply morphological techniques to eliminate noises and preserve the important features. We then derive the features of shape from the information of the object contours and the internal structure obtained by referencing to the original image. Finally, we take these features as input to the Bayesian Network for the recognition of the Micronuclei. To sum up, the steps are as follows: 1) slide digitization, 2) color space transformation, 3) image segmentation, 4) initially counting of the cells, 5) feature retrieval, 6) optimal features finding, and 7) automatic Micronuclei recognition. In addition, an image restoration program is also developed which makes it more convenient for researchers to find the image files afterwards for further confirming and facilitates the browsing of nearby images.

書名頁…………………………………………………………………………………I
論文口試委員審定書………………………………………………………………II
教育部授權書 III
國科會授權書 IV
國家圖書館授權書 V
中文摘要.... VI
英文摘要… VII
誌謝 ……………………………………………………………………………….VIII
目錄………………………………………………………………………………….IX
表目錄 …………………………………………………………….………………XII
圖目錄…........... XIII
第一章、 緒論 1
1.1研究動機與背景 1
1.2研究目的 3
1.3研究範圍與限制 4
1.4論文架構 6
第二章、 文獻探討 7
2.1微細胞核（MICRONUCLEI, MN） 7
2.1.1微核頻率（Frequency of micronucleus, MNF, MN‰） 9
2.2 近代細胞影像辨識 10
2.2.1細胞影像處理技術 10
2.2.2應用類神經網路於細胞辨識 12
2.3影像切割 12
2.3.1影像二值化 13
2.3.2目標物影像擷取 14
2.4細胞特徵值 14
2.4.1結構特徵值 15
2.4.2統計特徵值 16
2.5類神經網路 17
2.5.1倒傳遞類神經網路（Backpropagation, BP） 17
2.5.2貝氏網路（Bayesian Network） 18
第三章、 研究架構與方法 21
3.1 研究架構 21
3.2玻片數位化 23
3.3影像前處理 24
3.4影像切割 24
3.4.1影像二值化 24
3.4.2影像定位 25
3.5初步細胞計數 28
3.6影像特徵值 31
3.6.1從資訊角度解讀人工判斷微細胞核的標準 31
3.6.2特徵值的擷取 32
3.6.2.1根據特徵值來建立MN資料庫 39
3.6.3特徵值的選取 40
3.7細胞辨識 43
第四章、 實驗結果 45
4.1開發環境的介紹 45
4.1.1顯微鏡的控制模組與通訊模組 45
4.1.2拍照的設備 47
4.1.3 玻片的實驗範圍 48
4.1.4 儲存環境與程式開發環境 51
4.2 程式功能與介面介紹 52
4.2.1數位化玻片程式 52
4.2.2 影像分析程式 52
4.2.3 玻片影像還原程式 56
4.3實驗設計 57
4.4 WEKA軟體的實驗結果 57
4.4.1 獨立玻片的實驗結果 58
4.4.2多張玻片的實驗結果 60
4.5 實驗分析 64
第五章、 結論 67
5.1 研究成果 67
5.2 研究限制 68
5.2.1玻片標準化製作 68
5.2.2數位化的環境因素 68
5.3 未來展望 69
參考文獻………..........................................................................................................71
附錄A - 細胞分類表 74
附錄B - WEKA的實驗結果 75
附錄C - 主核與小核分開的微細胞核實驗 83

1.張武修, “HUMN International Collaborative Project on Micronucleus Frequencies in Human Populations 人群微細胞核頻率國際合作計劃－[ 臺灣計劃 ] “, http://www.ym.edu.tw/humn/.
2.E. Bengtsson, “Computerized Cell Image Analysis：Past, Present, and Future”, Scandinavian conference on Image Analysis (SCIA 2003), Springer-Verlag, Lecture Notes in Computer Science 2749, pp. 395-407.
3.K.A. Marghani, S.S. Dlay, and B.S. Sharif Univ. of Newcastle upon Tyne (United Kingdom) Andrew J. Sims Freeman Hospital (United Kingdom), “Automated morphological analysis approach for classifying colorectal microscopic images.” 2003 SPIE.
4. A. Lozano, M.E. Gonsebatt, A.E. Buenfil, and J. Márquez, “Image Analysis of Cell Micronuclei Micrographs to Evaluate Their Use as Indicators of Cell Damage”, AIP Conference Proceedings, September 15, 2003, Vol 682(1) pp. 211-216.
5.M. Fenech, W.P. Chang, M. Kirsch-Volders, N. Holland, S. Bonassi, and E. Zeiger, “HUMN project： detailed description of the scoring criteria for the cytokinesis-block micronucleus assay using isolated human lymphocyte cultures”, Genetic Toxicology and Environment Mutagenesis, Elsevier Science, 2002, pp.65-75.
6.S.G. Luis and M.M. Regina, “Micronuclei and chromatid buds are the result of related genotoxic events”, Departamento de Toxicologia Genetica y Ambiental, Instituto de Investigaciones Biomedicas, UNAM, Distrito Federal, Mexico.
7.T. Markiewicz, L. Moszczynski, and S. Osowski, “Myelogenous leukemia cell image preprocessing for feature generation”, V Int. Workshop - Computational Methods in Electrical Engineering, Jaznowiec, 2003, pp.70-73.
8.J. Iivarinen and A. Visa, “An Adaptive Texture and Shape Based Defect Classification”, Proceedings of the 14th International Conference on Pattern Recognition-Volume 1 - Volume 1, August 16 - 20, 1998, Brisbane, Australia .
9.T. Markiewicz and L. Moszczynski, “Analysis of Features for Blood Cell Recognition”, VI International Workshop - Computational Problems of Electrical Engineering, Zakopane 2004.
10.F. Albregtsen, H. Schulerud, and L. Yang, “Texture Classification of Mouse Liver Cell Nuclei Using Invariant Moments of Consistent Regions.” CAIP 1995: pp.496-502.
11.P. Berger, J. Lavallee, R. Rouiller, F. Laurent, R. Marthan, and JM. Tunon-De-Lara, “Assessment of bronchial inflammation using an automated cell recognition system based on colour analysis.” Eur Resp J 1999;14:1394–1402.
12.F. Ortiz, F. Torres, E. De Juan, and N. Cuenca, “Colour Mathematical Morphology For Neural Image Analysis”, Real-Time Imaging, 8 (2002) pp.455-465.
13.Q. Zheng, B.K. Milthorpe, and A. S. Jones, “Direct Neural Network Application for Automated Cell Recognition”, Cytometry Part A 57A:1-9, 2004. © 2003 Wiley-Liss, Inc.
14.N. Otsu, “A Threshold Selection Method From Gray Level Histograms”, IEEE Transactions on Systems, Man, and Cybernetics, SMC-9 (1979) 62-66.
15.J.R. Weaver and J.L. Au, “Application of automatic thresholding in image analysis scoring of cells in human solid tumors labeled for proliferation markers.” Cytometry; 29(2):128-35, 1997.
16.張祐育，”子宮內膜異位症之組織影像特徵參數分析”，南台科技大學，電機工程所碩士論文，2004年。
17.M.K. Hu, “Visual pattern recognition by moment invariants”, Information Theory, IEEE Transactions on, Volume: 8, Issue: 2, Feb 1962.
18.J. Iivarinen and A. Visa, “ Shape recognition of irregular objects” , In D. P. Casasent (Ed.), Intelligent Robots and Computer Vision XV: Algorithms, Techniques, Active Vision, and Materials Handling, Proc. SPIE 2904, 1996, pp. 25-32.
19.C. Sun, “Symmetry detection using gradient information”, Pattern Recognition Letters, ELSEVIER,3 April 1995
20.繆紹綱編著，數位影像處理活用-Matlab，全華科技圖書股份有限公司，1999年。
21.J. Kilian, “Simple Image Analysis by Moments, version 0.2”, OpenCV library documentation, 2001.
22.D.F. Specht, “Probabilistic Neural Networks”, Neural Networks, vol. 3, no.1 Jan 1990, pp. 109-118.
23.羅華強編著，類神經網路－MATLAB的應用，清蔚科技出版，2001年。
24.Weka, http://www.cs.waikato.ac.nz/~ml/weka/
25.OpenCV News Group, http://groups.yahoo.com/group/OpenCV/