臺灣博碩士論文加值系統

在本論文中，主要探討自動化切割與分類應用於醫學影像之研究。一般在臨床上，醫學影像造影與判讀皆需要人工介入進行處理，需耗費許多醫療人力資源同時也增加醫師的疲勞度。因此藉由電腦快速的運算能力以便自動的切割出血球影像中各種類型血球的位置和大小，以期能由電腦自動判讀出正確的血球影像而不需人工的介入。而自動判讀的依據往往是根據觀察血球的面積、形狀的變化、或是有任何病變的特徵，因此需擷取出最具影響力之特徵，使其達到有效的切割與分類之效果。
首先，本論文在第三章提出一個自動化的切割方法找出溶血性貧血中紅血球病變的部份進行。其切割方法為先去除背景和雜訊，之後將單一顆的紅血球進行邊緣變化的分析找出病變。而重疊的紅血球利用鏈碼技術找出高凹點的位置，其凹點的資訊再運用circle-fit演算法將重疊的血球分別分割出來。本技術能有效的找出病變的血球數量。
在第四章中，本論文又進一步的提出可將不同的溶血性貧血形態進行分類的方法。利用八方向的鏈碼演算法(chain code)找出邊緣的變化與方向，而後再做特徵的擷取將鏈碼做差值運算及鏈碼依序方向的變化。則再將擷取出來的特徵利用分類器進行訓練測試，從實驗結果顯示，本技術能有效的達到溶血性貧血類型之辨識。
第五章提出一個在巨紅血球貧血中計算出細胞核葉片數的自動化切割。首先將細胞核利用色彩空間切割出來，再運用距離轉化技術(distance transform)找出物件中心位置，在此同時利用圓擴張的方式將細胞核的葉片分別分離出來。本技術能將不規則的細胞核切割並計算出來，並分辨出其病徵與嗜中性白血球的不同之處。

In this thesis we mainly discuss the application of automatic image segmentation and classification in the research of medical image. Clinically, medical imaging and interpretation requires human works, costing abundant of medical human resource and increasing the loading of staffs. Therefore, the thesis seeks for an automatic way to identify the images correctly via the vast computing ability of computers to automatically cut and categorize the location and size of various types of the blood cells from the smear image without human efforts. Automatic identification is based on observing the area, shape varieties and other features of disease of the blood cells. Thus, it is essential to extract the most significant feature to effectively categorize the cells.
First, an automatic image segmentation scheme to filter the Hemolytic Anemia blood cells from the healthy ones is proposed in Chapter 3. The proposed method first removes the background and noises in the blood images and then each single erythrocyte is analyzed by using the changes in the edges. For segmenting the overlapping erythrocytes, the edge directions of the cells are acquired by using chain code technique to find out the points of high concavity that can be adapted to circle-fit algorithm to separate each erythrocyte. The proposed scheme could effectively analyze the numbers of normal erythrocytes and poikilocytes according to the experimental results.
In Chapter 4, this thesis proposes an automatic method for segmentation and classification of erythrocytes from the patients with Hemolytic Anemia. By applying 8-connection chain code, the direction of edges change can be analyzed. Then extract the variation of eight directions from the individual erythrocyte as feature for classifying four main Hemolytic Anemia types. From the experimental results, the proposed method can calculate the quantity of erythrocytes and recognize the type of Hemolytic Anemia effectively.
Chapter 5 presents an automatic segmentation method for calculating blood cell of nucleus with Megaloblastic Anemia. First, use RGB color space to segment the nuclei from the blood image. Then utilize distance transform to figure out the center points information which is then adapted to expansion of circle to separate each nucleus from the object in the image. The proposed method will eventually analyze the numbers of nucleus in the blood smears. The proposed scheme can segment and calculate irregular clustered nuclei. Therefore, the number of nucleus can determine whether the multisegmented neutrophils are Megaloblastosic Anemia or not.

Abstract in Chinese I
Abstract in English III
Contents V
Lists of Tables VIII
List of Figures IX
Chapter 1 Introduction 1
1.1 Background 1
1.2 Image Segmentation and Classification 3
1.3 Thesis Organization 6
Chapter 2 Preliminaries 7
2.1 Image Binarization 7
2.2 Image Restoration 9
2.3 Sobel Operator for Edge Detection 11
2.4 Circle-fit Algorithm 13
2.5 Chain Codes Technique 14
2.6 Distance Transform 16
Chapter 3 Automatic Image Segmentation for Hemolytic Anemia in Thin Blood Smears 19
3.1 The Proposed Scheme 19
3.1.1 Image Preprocessing 20
3.1.2 Recognition of Abnormal Erythrocyte 21
3.1.3 Segmentation of Overlapping Erythrocytes 23
3.2 Experimental Results 27
3.2.1 Comparisons of the Proposed Method with Manual Recognition 30
Chapter 4 Automatic Image Segmentation and Classification based on Direction Texton Technique for Hemolytic Anemia in Thin Blood Smears 34
4.1 The Proposed Scheme 35
4.1.1 Image Preprocessing 36
4.1.2 Recognition of Isolated Erythrocytes 37
4.1.3 Segmentation of Overlapping Erythrocytes 38
4.1.4 Feature Extraction 38
4.2 Experimental Results 43
4.2.1 Comparisons of the Proposed Method with Manual Recognition 45
4.2.2 Analysis of the Proposed Features and Results of Various Classifiers 48
Chapter 5 Segmentation of Complex Nucleus for Megaloblastic Anemia in Thin Blood Smears 50
5.1 The Proposed Scheme 50
5.1.1 Image Preprocessing 52
5.1.2 Segmentation of Nucleus 52
5.1.3 Calculate the Number of Nucleus 54
5.2 Experimental Results 57
5.2.1 Comparisons of the Proposed Method with Manual Segmentation 59
Chapter 6 Conclusions and Future Works 62
6.1 Conclusions 62
6.2 Future Works 63
Bibliography 65

[1]M. S. Atkins and B. T. Mackiewich, “Fully Automatic Segmentation of the Brain in MRI,” IEEE Transactions in Medical Imaging, Vol. 17, No. 1, 1998, pp. 98–107.
[2]B. J. Bain, “Diagnosis from the Blood Smear,” The New England Journal of Medicine, Vol. 353, No. 5, 2005, pp. 498-507.
[3]B. Ballaro, A. M. Florena, V. Franco, D. Tegolo, C. Tripodo, and C. Valenti, “An automated Image Analysis Methodology for Classifying Megakaryocytes in Chronic Myeloproliferative Disorders,” Medical Image Analysis, Vol. 12, No. 6, 2008, pp. 703-712.
[4]S. D. Cataldo, E. Ficarra, A. Acquaviva, and E. Macii, “Automated Segmentation of Tissue Images for Computerized IHC Analysis,” Computer Methods and Programs in Biomedicine, Vol.100, No. 1, 2010, pp. 1-15.
[5]L. Chen and H. Y. H. Chuang, “An Efficient Algorithm for Complete Euclidean Distance Transform on Mesh-connected SIMD,” Parallel Computing, Vol. 21, No. 5, 1995, pp. 84-852.
[6]F. Cloppet and A. Boucher, “Segmentation of Complex Nucleus Configurations in Biological Images,” Pattern Recognition Letters, Vol. 31, No. 8, 2010, pp. 755-761.
[7]A. Datta and S. Soundaralakshmi, “Constant-Time Algorithm for the Euclidean Distance Transform on Reconfigurable Meshes,” Journal of Parallel and Distributed Computing, Vol. 61, No. 10, 2001, pp. 1439-1455.
[8]A. Datta and S. Soundaralakshmi, “Fast and Scalable Algorithms for the Euclidean Distance Transform on A Linear Array with A Reconfigurable Pipelined Bus System,” Journal of Parallel and Distributed Computing, Vol. 64, No. 3, 2004, pp. 360-369.
[9]D. F. Dunn and W. E. Higgins, “Optimal Gabor Filters for Texture Segmentation,” IEEE Transactions on Image Processing, Vol. 4, No. 7, 1995, pp. 947-964.
[10]S. K. S. Fan and Y. Lin, “A Multi-level Thresholding Approach Using A Hybrid Optimal Estimation Algorithm,” Pattern Recognition Letter, Vol.28, No. 5, 2007, pp. 662–669.
[11]H. Farid and E. P. Simoncelli, “Differentiation of Discrete Multidimensional Signals,” IEEE Transactions on Image Processing, Vol. 13, No. 4, 2004, pp. 1057-7149.
[12]L. Grady and E. L. Schwartz, “Isoperimetric Graph Partitioning for Image Segmentation,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 28, No. 3, 2006, pp. 469-475.
[13]C. Gunduz-Demir, M. Kandemir, A. B. Tosun, and C. Sokmensuer, “Automatic Segmentation of Colon Glands Using Object-graphs,” Medical Image Analysis, Vol. 14, No. 1, 2010, pp.1–12.
[14]S. Gustavson and R. Strand, “Anti-aliased Euclidean Distance Transform,” Pattern Recognition Letters, Vol. 32, No. 2, 2011, pp.252-257.
[15]K. S. Hahn, S. Jung, Y. Han, and H. Hahn, “A New Algorithm for Ellipse Detection by Curve Segments,” Pattern Recognition Letters, Vol. 29, No. 13, 2008, pp. 1836-1841.
[16]D. Y. Huang and C. H. Wang, “Optimal Multi-level Thresholding Using A Two-stage Otsu Optimization Approach,” Patter Recognition Letters, Vol. 30, No. 3, 2009, pp. 275-284.
[17]A. Jimeno-Morenilla, R. Molina-Carmona, and J. L. “Sanchez-Romero, Mathematical Morphology for Design and Manufacturing,” Mathematical and Computer Modelling, Vol. 54, No. 7-8, 2011, pp. 1753–1759.
[18]C. Jung, C. Kim, S. W. Chae, and S. Oh, “Unsupervised Segmentation of Overlapped Nuclei Using Bayesian Classification,” IEEE Transactions on Biomedical Engineering, Vol. 57, No. 12, 2010, pp. 2825-2832.
[19]S. J. Kim and M. Y. Yang, “Triangular Mesh Offset for Generalized Cutter,” Computer Aided Design, Vol. 37, No. 10, 2005, pp. 999-1014.
[20]M. N. Kolountzakis and K. N. Kutulakos, “Fast Computation of the Euclidian Distance Maps for Binary Images,” Information Processing Letters, Vol. 42, No. 4, 1992, pp. 181-184.
[21]K. Li, Y. Pan, and M. Hamdi, “Solving Graph Theory Problems Using Reconfigurable Pipelined Optical Buses,” Parallel Computing, Vol. 26, NO. 6, 2000, pp. 723-735.
[22]L. M. Lifshitz and S. M. Pizer, “A Multiresolution Hierarchical Approach to Image Segmentation based on Intensity Extrema,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 12, No. 6, 1990, pp. 529-540.
[23]X. Liu and D. L. Wang, “Texture Classification Using Spectral Histograms,” IEEE Transactions on Image Processing, Vol. 12, No. 6, 2003, pp. 662-670.
[24]Y. Lucet, “New Sequential Exact Euclidean Distance Transform Algorithms based on Convex Analysis,” Image and Vision Computing, Vol. 24, No. 12, 2009, pp. 37-44.
[25]G. R. Martin and A. C. Yu, “Compact Representation of Contours Using Directional Grid Chain Code,” Signal Processing: Image Communication, Vol. 23, No. 2, 2008, pp. 87-100.
[26]Medical Image Open Source of Website: PathologyOutlines.com
[27]Medical Image Open Source of Website: http://www.medicalhistology.us/twiki/bin/view/Main/BloodAndBoneMarrowAtlas10#TopOfPage
[28]Medical Image Open Source of Website: http://www.ihcworld.com/imagegallery/displayimage.php?album=4&pos=24
[29]F. Meyer, “An Overview of Morphological Segmentation,” International Journal of Pattern Recognition and Artificial Intelligence, Vol. 15, No. 7, 2001, pp. 1089–1118.
[30]R. Molina-Carmona, A. Jimeno, and R. Rizo-Aldeguer, “Morphological Offset Computing for Contour Pocketing,” Journal of Manufacturing Science and Engineering, Vol. 129, No. 2, 2007, pp. 400-406.
[31]W. H. Organization, “The World Health Report 2002: Reducing Risks Promoting Healthy Life,” 2002.
[32]F. A. Pujol, J. M. G. Chamizo, A. Fuster, M. Pujol, and R. Rizo, “Use of Mathematical Morphology in Real-time Path Planning,” The International Journal of Systems and Cybernetics, Vol.31, No.1, 2002, pp. 115–123.
[33]Y. D. Qu, C. S. Cui, S. B. Chen and J. Q. Li, “A Fast Subpixel Edge Detection Method Using Sobel–Zernike Moments Operator,” Image and Vision Computing, Vol. 23, No. 1, 2005, pp. 11-17.
[34]R. Rodríguez, “A Strategy for Blood Vessels Segmentation based on the Threshold which Combines Statistical and Scale Space Filter: Application to the Study of Angiogenesis,” Computer Methods and Programs in Biomedicine, Vol. 82, No. 1, 2006, pp. 1-9.
[35]A. Rosenfeld and J. L. Pfalz, “Distance Function on Digital Pictures,” Pattern Recognition, Vol. 1, No.1, 1968, pp. 33–61.
[36]P. K. Saha and J. K. Udupa, “Optimum Image Thresholding via Class Uncertainty and Region Homogeneity,” IEEE Transaction Pattern Analyze, Vol.23, No. 7, 2001, pp. 689–706.
[37]H. Sanchez-Cruz, E. Bribiesca, and R. M. Rodriguez-Dagnino, “Efficiency of Chain Codes to Represent Binary Objects,” Pattern Recognition, Vol. 40, No. 6, 2007, pp. 1660-1674.
[38]A. B. Tosuna, M. Kandemir, C. Sokmensuerb, and C. Gunduz-Demira, “ Object-oriented Texture Analysis for the Unsupervised Segmentation of Biopsy Images for Cancer Detection,” Pattern Recognition, Vol. 42, No. 6, 2009, pp. 1104 -1112.
[39]N. Theera-Umpon and S. Dhompongsa, “Morphological Granulometric Features of Nucleus in Automatic Bone Marrow White Blood Cell Classification,” IEEE Transactions on Information Technology in Biomedicine, Vol. 11, No. 3, 2007, pp. 353-359.
[40]Weka: Data Mining Software in Java Class, Available on http://www.cs.waikato.ac.nz/ml/weka/.
[41]D. Yu, T. D. Pham, X. Zhou, and S. T. C. Wong, “Recognition and Analysis of Cell Nuclear Phases for High-content Screening based on Morphological Features,” Pattern Recognition, Vol. 42, No. 4, 2009, pp. 498-508.
[42]Y. H. Yu and C. C. Chang, “A New Edge Detection Approach based on Image Context Analysis,” Image and Vision Computing, Vol. 24, No. 10, 2006, pp. 1090-1102.
[43]E. Zahara, S.K.S Fan, and D.M. Tsai, “Optimal Multi-thresholding Using A Hybrid Optimization Approach,” Pattern Recognition Letter, Vol. 26, No. 8, 2005, pp. 1082–1095.
[44]C. T. Zahn, “Graph-theoretical Methods for Detecting and Describing Gestalt Clusters,” IEEE Transactions on Computers, Vol. 20, No. 1, 1971, pp. 68-86.