生物醫學影像往往潛藏許多醫學資訊，藉由影像解讀以及特徵分析，進而幫助醫學上對生理現象與疾病能有更多的了解。對於分析生物影像特徵時，我們能透過四種不同的特徵描述包含:大小、形狀、顏色、紋理，清楚地描述所需分析之醫學特徵，進而選擇所需之影像分析工具，以獲得重要的醫學資訊。本論文提出一個電腦輔助方法，應用於剖析人類皮膚膠原纖維特徵之生物醫學影像，重要的膠原纖維特徵包含其密度、方向多樣性、粗細，利用賈伯濾波器能分析影像上紋理與大小之特徵，以定量分析膠原纖維之方向與粗細，另外透過賈伯濾波器與Frangi濾波器依據膠原纖維特性萃取膠原纖維形狀特徵，並給予支持向量機訓練出一個準確分類器，以分割膠原纖維區域，進而分析膠原纖維之密度。演算法能克服利用儀器進行特徵分析之不便，且演算法與其相關文獻進行比較，本論文能提供一個全面性膠原纖維特徵之分析， 不僅在醫學影像分析上具有相當的發展潛力，對於醫學研究也具極大的醫學價值。

The medical images normally contain abundant medical information. With image interpretations and feature analyses of medical images, physiological processes or diseases can be further understood, and this could be crucial for medical advances. Regarding feature analyses, four feature descriptions (including size, shape, color, and texture) will be analyzed to clarify desired medical features. These feature descriptions help people understand each medical feature clearly, and then the image processing tool is applied to analyze every medical feature. This thesis presents an algorithm of a computer-assisted method to dissect and quantify collagen fiber features of human skin including collagen fiber density, orientation diversity and thickness in the medical image. The Gabor filter is able to extract image texture and size information, which used for quantifying collagen fiber orientation and thickness. Moreover, the Gabor filter and the Frangi filter are utilized for extracting shape information of collagen fiber according to the properties of collagen fiber, and then the support vector machine method use shape information to obtain an accurate classification to segment collagen fiber region and further analyze collagen fiber density. The proposed algorithm is able to overcome inconveniently using the instrument for feature analysis. Comparing with other related works, the proposed algorithm provides full analyses of collagen fiber features, which has not only potential in biomedical image analyzing, but also significant value to medical research.

摘 要 i
Abstract iii
誌 謝 v
List of Tables xi
List of Figures xiii
Chapter 1 Introduction 1
1.1 Introduction 1
1.2 Motivation 2
1.3 Structure of this Thesis 5
Chapter 2 Background Information 7
2.1 Image Information 7
2.2 Physical Background of the Acquired Images 9
Chapter 3 Surveys of Related Works in the Literatures 13
3.1 Feature Extraction 13
3.1.1 Fourier Transform 13
3.1.2 Wavelet Transform 15
3.1.3 Gabor Filter 17
3.1.4 Frangi Filter 22
3.1.5 Steerable Filter 24
3.1.6 Gray-level Co-occurrence Matrix 25
3.1.7 Fractal Feature Extraction 26
3.2 Clustering and Classification 27
3.2.1 K-means Clustering 28
3.2.2 Support Vector Machine (SVM) 29
3.2.3 Artificial Neural Networks (ANN) 33
3.2.4 Bayesian Networks 34
3.3 Image Segmentation 35
3.3.1 Image Thresholding 35
3.3.2 Otsu’s Method 36
3.3.3 Region Growing 38
3.3.4 Edge Detection 39
3.3.5 Graph-theoretical Method 40
Chapter 4 Proposed Algorithms 41
4.1 Block Diagram 41
4.2 Image Preprocessing 43
4.2.1 Wiener Filter 43
4.2.2 Contrast Limit Adaptive Histogram Equalization 48
4.3 Feature Extraction of Collagen Fiber 53
4.3.1 Convolution with a Gabor Filter Bank 55
4.3.2 The Extraction of Directionality Feature and Scale Feature 70
4.3.3 Structure Analysis and Eigen Decomposition 72
4.3.4 Vessel Enhancement 75
4.4 Collagen Fiber Segmentation 79
4.4.1 Analyses of Different Features for SVM method 80
4.4.2 The Different Combinations of Feature Vector 85
4.5 Density Evaluation 87
4.6 Orientation Diversity Evaluation 90
4.7 Thickness Evaluation 92
4.8 Experimental Results 93
4.8.1 Density Evaluation 101
4.8.2 Orientation Diversity Evaluation 104
4.8.3 Thickness Evaluation 107
4.9 Comparison with Previous Works 109
Chapter 5 Conclusions and Future Works 123
5.1 Conclusions 123
5.2 Future Works 124
Acknowledgments 125
References 127

[1]K. Wolff, L. Goldsmith, B. Gilchrest, S. Katz, A. Paller, and D. Leffell, Fitzpatrick's Dermatology In General Medicine, Seventh Edition: Two Volumes: Mcgraw-hill, 2007.
[2]C.-K. Sun, S.-W. Chu, S.-Y. Chen, T.-H. Tsai, T.-M. Liu, C.-Y. Lin, et al., Higher harmonic generation microscopy for developmental biology, Journal of structural biology, vol. 147, pp. 19-30, 2004.
[3]S.-Y. Chen, H.-Y. Wu, and C.-K. Sun, In vivo harmonic generation biopsy of human skin, Journal of biomedical optics, vol. 14, pp. 060505-060505-3, 2009.
[4]M.-R. Tsai, S.-Y. Chen, D.-B. Shieh, P.-J. Lou, and C.-K. Sun, In vivo optical virtual biopsy of human oral mucosa with harmonic generation microscopy, Biomedical optics express, vol. 2, p. 2317, 2011.
[5]Y.-H. Cheng, C.-F. Lin, T.-F. Shih, and C.-K. Sun, A novel intravital multi-harmonic generation microscope for early diagnosis of oral cancer, SPIE BiOS, pp. 85770R-85770R-6, 2013.
[6]M.-R. Tsai, Y.-W. Chiu, M. T. Lo, and C.-K. Sun, Second-harmonic generation imaging of collagen fibers in myocardium for atrial fibrillation diagnosis, Journal of biomedical optics, vol. 15, pp. 026002-026002-6, 2010.
[7]S.-Y. Chen, S.-U. Chen, H.-Y. Wu, W.-J. Lee, Y.-H. Liao, and C.-K. Sun, In vivo virtual biopsy of human skin by using noninvasive higher harmonic generation microscopy, IEEE J. Sel. Top. Quantum Electron, vol. 16, pp. 478-492, 2010.
[8]T. Yasui, M. Yonetsu, R. Tanaka, Y. Tanaka, S.-i. Fukushima, T. Yamashita, et al., In vivo observation of age-related structural changes of dermal collagen in human facial skin using collagen-sensitive second harmonic generation microscope equipped with 1250-nm mode-locked Cr: Forsterite laser, Journal of biomedical optics, vol. 18, pp. 031108-031108, 2013.
[9]J. Varani, M. K. Dame, L. Rittie, S. E. Fligiel, S. Kang, G. J. Fisher, et al., Decreased collagen production in chronologically aged skin: roles of age-dependent alteration in fibroblast function and defective mechanical stimulation, The American journal of pathology, vol. 168, pp. 1861-1868, 2006.
[10]E. Sivridis, A. Giatromanolaki, and M. I. Koukourakis, “Stromatogenesis and tumor progression, International Journal of Surgical Pathology, vol. 12, pp. 1-9, 2004.
[11]E. Brown and T. McKee, Dynamic imaging of collagen and its modulation in tumors in vivo using second-harmonic generation, Nature medicine, vol. 9, pp. 796-800, 2003.
[12]G. Falzon, S. Pearson, and R. Murison, Analysis of collagen fibre shape changes in breast cancer, Physics in medicine and biology, vol. 53, p. 6641, 2008.
[13]J. F. Ribeiro, E. H. M. dos Anjos, M. L. S. Mello, and B. de Campos Vidal, Skin Collagen Fiber Molecular Order: A Pattern of Distributional Fiber Orientation as Assessed by Optical Anisotropy and Image Analysis, PloS one, vol. 8, p. e54724, 2013.
[14]W. Lo, W.-L. Chen, C.-M. Hsueh, A. A. Ghazaryan, S.-J. Chen, D. H.-K. Ma, et al., Fast Fourier Transform–Based Analysis of Second-Harmonic Generation Image in Keratoconic Cornea, Investigative ophthalmology ＆ visual science, vol. 53, pp. 3501-3507, 2012.
[15]R. Cicchi, C. Matthäus, T. Meyer, A. Lattermann, B. Dietzek, B. R. Brehm, et al., Characterization of collagen and cholesterol deposition in atherosclerotic arterial tissue using non‐linear microscopy, Journal of biophotonics, vol. 7, pp. 135-143, 2014.
[16]R. Cicchi, N. Vogler, D. Kapsokalyvas, B. Dietzek, J. Popp, and F. S. Pavone, From molecular structure to tissue architecture: collagen organization probed by SHG microscopy, Journal of biophotonics, vol. 6, pp. 129-142, 2013.
[17]W. Hu, H. Li, C. Wang, S. Gou, and L. Fu, Characterization of collagen fibers by means of texture analysis of second harmonic generation images using orientation-dependent gray level co-occurrence matrix method, Journal of Biomedical Optics, vol. 17, pp. 0260071-0260079, 2012.
[18]J. S. Bredfeldt, Y. Liu, C. A. Pehlke, M. W. Conklin, J. M. Szulczewski, D. R. Inman, et al., Computational segmentation of collagen fibers from second-harmonic generation images of breast cancer, Journal of Biomedical Optics, vol. 19, pp. 016007-016007, 2014.
[19]T. Igarashi, K. Nishino, and S. K. Nayar, The appearance of human skin, Technical Report: CUCS-024-05, 2005.
[20]M.-R. Tsai, C.-H. Chen, and C.-K. Sun, Third and second harmonic generation imaging of human articular cartilage, Proceeding of SPIE, San Iose, CA, vol. 7183, pp. 71831V-1, 2009.
[21]R.C. Gonzalez and R.E. Woods, Digital Image Processing: Prentice-Hall, Englewood Cliffs, NJ, 2002.
[22]A. Zoumi, A. Yeh, and B. J. Tromberg, Imaging cells and extracellular matrix in vivo by using second-harmonic generation and two-photon excited fluorescence, Proceedings of the National Academy of Sciences, vol. 99, pp. 11014-11019, 2002.
[23]N. Otsu, A threshold selection method from gray-level histograms, Automatica, vol. 11, pp. 23-27, 1975.
[24]R. Adams and L. Bischof, Seeded region growing, Pattern Analysis and Machine Intelligence, IEEE Transactions, vol. 16, pp. 641-647, 1994.
[25]R. Maini and H. Aggarwal, Study and comparison of various image edge detection techniques, International Journal of Image Processing (IJIP), vol. 3, pp. 1-11, 2009.
[26]J. Shi and J. Malik, Normalized cuts and image segmentation, Pattern Analysis and Machine Intelligence, IEEE Transactions, vol. 22, pp. 888-905, 2000.
[27]P. F. Felzenszwalb and D. P. Huttenlocher, Efficient graph-based image segmentation, International Journal of Computer Vision, vol. 59, pp. 167-181, 2004.
[28]J. A. Hartigan and M. A. Wong, Algorithm AS 136: A k-means clustering algorithm, Applied statistics, pp. 100-108, 1979.
[29]C. Cortes and V. Vapnik, Support-vector networks, Machine learning, vol. 20, pp. 273-297, 1995.
[30]C.-W. Hsu, C.-C. Chang, and C.-J. Lin, A practical guide to support vector classification, 2003.
[31]T. Kohonen, An introduction to neural computing, Neural networks, vol. 1, pp. 3-16, 1988.
[32]N. Friedman, D. Geiger, and M. Goldszmidt, Bayesian network classifiers, Machine learning, vol. 29, pp. 131-163, 1997.
[33]M. Adel, M. Rasigni, T. Gaidon, C. Fossati, and S. Bourennane, Statistical-based linear vessel structure detection in medical images, Image Processing (ICIP) 16th IEEE International Conference, pp. 649-652, 2009.
[34]Y. Tamada, S. Imoto, and S. Miyano, Parallel algorithm for learning optimal bayesian network structure, Journal of Machine Learning Research, vol. 12, pp. 2437-2459, 2011.
[35]A. V. Oppenheim, A. S. Willsky, and S. H. Nawab, Signals and systems: Prentice Hall, 1997.
[36]S. Theodoridis and K. Koutroumbas, Pattern Pecognition: Academic Press, 2003.
[37]S. G. Mallat, “A theory for multiresolution signal decomposition: The wavelet representation, IEEE Trans. Pattern Anal. and Machine Intell., vol. 11, no. 7, pp. 674 - 693, 1989.
[38]D. Gabor. “Theory of communications, Journal of International Electrical Engineers, vol. 93, part III, No. 26, pp. 427 - 457, 1946.
[39]Q. Li, J. You, L. Zhang, and P. Bhattacharya, A multiscale approach to retinal vessel segmentation using Gabor filters and scale multiplication, Systems, Man and Cybernetics, pp. 3521-3527, 2006.
[40]J. V. Soares, J. J. Leandro, R. M. Cesar, H. F. Jelinek, and M. J. Cree, Retinal vessel segmentation using the 2-D Gabor wavelet and supervised classification, Medical Imaging, IEEE Transactions, vol. 25, pp. 1214-1222, 2006.
[41]J. G. Daugman, Uncertainty relation for resolution in space, spatial frequency, and orientation optimized by two-dimensional visual cortical filters, JOSA A, vol. 2, pp. 1160-1169, 1985.
[42]A. F. Frangi, W. J. Niessen, K. L. Vincken, and M. A. Viergever, Multiscale vessel enhancement filtering, Medical Image Computing and Computer-Assisted Interventation—MICCAI’98, ed: Springer, pp. 130-137, 1998.
[43]S. You, E. Bas, D. Erdogmus, and J. Kalpathy-Cramer, Principal Curved Based Retinal Vessel Segmentation towards Diagnosis of Retinal Diseases, Healthcare Informatics, Imaging and Systems Biology (HISB), 2011 First IEEE International Conference, pp. 331-337, 2011.
[44]T. Lindeberg, Scale-space theory in computer vision: Springer, 1993.
[45]Z. Hongqing, S. Huazhong, and L. Limin, Blood vessels segmentation in retina via wavelet transforms using steerable filters, Computer-Based Medical Systems, 2004. CBMS 2004. Proceedings. 17th IEEE Symposium, pp. 316-321, 2004.
[46]C.-C. Chen, J. S. DaPonte, and M. D. Fox, Fractal feature analysis and classification in medical imaging, Medical Imaging, IEEE Transactions on, vol. 8, pp. 133-142, 1989.
[47]J. S. Lim, Two-dimensional signal and image processing: Prentice-Hall, Englewood Cliffs, NJ, 1990.
[48]E. D. Pisano, S. Zong, B. M. Hemminger, M. DeLuca, R. E. Johnston, K. Muller, et al., Contrast limited adaptive histogram equalization image processing to improve the detection of simulated spiculations in dense mammograms, Journal of Digital Imaging, vol. 11, pp. 193-200, 1998.
[49]C. E. Shannon, A mathematical theory of communication, ACM SIGMOBILE Mobile Computing and Communications Review, vol. 5, pp. 3-55, 2001.
[50]S. M. Kay, Fundamentals of statistical signal processing: estimation theory: Prentice-Hall, 1993.
[51]Javier R. Movellan: “Tutorial on Gabor Filters, Tutorial paper [Online] http://mplab.ucsd.edu/tutorials/pdfs/gabor.pdf.
[52]A. M. MacEachren, “Compactness of geographic shape: Comparison and evaluation of measures, Geografiska Annaler. Series B. Human Geography, pp. 53 - 67, 1985.
[53]S. Heuke, N. Vogler, T. Meyer, D. Akimov, F. Kluschke, H.-J. Röwert-Huber, et al., Detection and Discrimination of Non-Melanoma Skin Cancer by Multimodal Imaging, in Healthcare, 2013, pp. 64-83.
[54]M. Unser, Texture classification and segmentation using wavelet frames, Image Processing, IEEE Transactions, vol. 4, pp. 1549-1560, 1995.
[55]T. Kasparis, D. Charalampidis, M. Georgiopoulos, and J. Rolland, Segmentation of textured images based on fractals and image filtering, Pattern Recognition, vol. 34, pp. 1963-1973, 2001.