臺灣博碩士論文加值系統

醫學影像是醫生在臨床診斷上非常重要的診斷依據之一，因此，影像的品質不僅會影響醫生診斷的正確性，更會影響自動化的電腦輔助診斷系統的判讀。影像物件的對比強化是改善影像品質很關鍵的步驟，這不僅能讓醫生更易正確的判讀物件，也能幫助自動化的電腦輔助診斷系統的物件切割。而物件切割的好壞，當然更直接地影響到自動化的電腦輔助診斷系統提供的輔助診斷品質。
本研究關注在兩個主要的議題：如何改善醫學影像中物件與背景的顏色對比度，以及醫學影像中物件的切割與分類。其中，對比強化是針對如何增強眼底影像中血管與背景的顏色對比度，而物件偵測則是針對肝組織切片影像中傷疤的切割與分類。主要內容如下：
(1)提出一個利用色彩轉換增強對比度的方法—現有的影像對比度增強方法，皆是基於原始影像的色彩紋理在作處理。如果原始影像的品質不佳或者醫生要診斷不同物件時，現有的影像對比度增強方法常常無法提供適當的色彩對比。本研究所提出來的方法是利用一張參考影像的色彩紋理來改善原始眼底影像的色彩紋理。這個方法能打破現有影像對比度增強方法的限制。其次，尋找彩色眼底影像中，能提供最佳血管與背景對比度的色彩頻道—現有的研究指出原始眼底影像中，綠色頻道能提供最佳血管與背景對比度。但實驗發現，在經由所提出的色彩轉換方法轉換後，紅色頻道能提供最佳血管與背景對比度。
(2) 首先使用區域十字門檻方法，由肝組織切片影像中將傷疤肝組織與正常肝組織分離開來，再利用雙層識別演算法來確認肝組織中的傷疤所屬狀態。最後，用參數設定基因演算法來決定所用到參數的最佳值。實驗結果證明所提出的方法對肝組織傷疤的判定有接近90%的正確率。

Medical images are some of the very important referenced data for clinical diagnosis. The quality of a medical image not only affects the result of clinical diagnosis but also affects the performance of an automatic computer diagnosis assistant system. The object contrast enhancement is an essential step to improve the quality of medical images. This not only lets a doctor correctly identify objects in a medical image easier but also improves the object segmentation of an automatic computer diagnosis assistant system. Of course, the quality of object segmentation will affect the performance of the diagnosis assistant of an automatic computer diagnosis assistant system.
This research is focused on two major issues - improving the color contrast between vessels and the background in a retinal image and object segmentation of medical images. Contrast enhancement is applied on retinal images to enhance the color contrast between vessels and backgrounds. Object detection is employed to extract and classify the scar tissues from a liver tissue section image. The topics are summarized as follows.
(1) Propose a color transformation method to enhance the contrast. The existing color contrast enhancement methods are based on the color textures of the original retinal images. If the image quality is poor, or diagnosis needs to be done for different parts of a retinal image, the existing methods might not obtain well color contrast. The proposed method employs the color textures of another image to improve the color texture of the original retinal image. This can break the restriction of the existing contrast enhancement methods. Furthermore, the color channel of retinal images with the highest contrast between vessels and backgrounds is studied. The existing researches show that the green channel of a color retinal image has the highest contrast. But in experiments, it is found that R channel will provide the highest color contrast after the proposed color transforming.
(2) First, a local cross thresholding method is provided to separate the scar liver tissues from normal liver tissues on a liver tissue section image. Moreover, a two-layer recognition algorithm is presented to identify the scar stage of liver tissue. Furthermore, a parameter decider genetic algorithm (PDGA) is proposed to decide the most suitable values of the parameters used in the presented diagnosis system. The experimental results show that, in segmenting scar tissues and normal tissues on liver cirrhosis images, the average accuracy is close to 90%.

Contents
誌謝辭 i
中文摘要 ii
Abstract iii
Contents v
List of Figures vii
List of Tables viii
Chapter 1 Introduction 1
1.1 Motivation 1
1.2 Challenges 4
1.3 Materials 5
1.4 Thesis Structures 6
Chapter 2 Related Works 8
2.1 Related Works of Color Enhancement 8
2.2 Related Works of Object Segmentation 10
Chapter 3 MRA-Based Color Transformation Scheme 12
3.1 Color Space Transformation 13
3.2 Normalization 14
3.3 Sampling and Sorting 14
3.4 Multiple Regression Analysis 15
3.5 Inverse Normalization and Color Space Transformation 16
3.6 Color Transformation in Different Color Space 17
Chapter 4 Performance Evaluations 20
Chapter 5 Liver Scar Stage Diagnosis System 26
5.1 Image Pre-processing 26
5.2 Contrast Enhancement 28
5.3 Local Cross Thresholding 28
5.4 Feature Extraction 30
5.5 Liver Tissue Scar Stage Identification 31
5.6 Parameter Decider Genetic Algorithm (PDGA) 33
The performance of the TSILSS diagnosis system is deeply affected by the parameters rG, mc, rb, rs, r1, w1, r2, w2, …, r5, and w5. In this study, a parameter decider genetic algorithm (PDGA) is presented to give the fittest values of rG, mc, rb, rs, r1, w1, r2, w2, …, r5, and w5. 33
Chapter 6 Results Analysis 36
Chapter 7 Conclusions 39
7.1 Color Contrast Enhancement 39
7.2 Object Segmentation and Detection 39
Bibliography 41

Bibliography

[1] G. Michelson, et al., “Eye imaging identifies cardiovascular risk factors,” World Ophthalmology Congress. (2012)；IS-TEL-FR65.
[2] S. Wild, G. Roglic, A. Green, R. Sicree, H. King, “Global prevalence of diabetes: Estimates for the year 2000 and projections for 2030,” Diabetes Care, vol. 27, no. 5, pp. 1047-1053, 2004.
[3] B.S. Courtney, Z. Wei, A.C. Douglas, G. Jennifer, “Utilization and accuracy of biopsy in patients with hepatocellular carcinoma in a community based setting,” AASLD Abstracts, no.744, pp.S-927-S-928, 2012.
[4] J.H. Chow, C. Chow, The Encyclopedia of Hepatitis and Other Liver Diseases. Infobase Publishing, 2006.
[5] M. Al‐Sarraf, et al., “Primary liver cancer. A review of the clinical features, blood groups, serum enzymes, therapy, and survival of 65 cases,” Cancer, vol.33, pp.574-582, 1974.
[6] K.W. Jung, et al., “Cancer survival in Korea 1993-2002: a population-based study,” Journal of Korean medical science, vol.22, pp.S5-S10, 2007.
[7] A. Andreou, et al., “Improved long-term survival after major resection for hepatocellular carcinoma: a multicenter analysis based on a new definition of major hepatectomy,” Journal of gastrointestinal surgery, vol.17, pp.66-77, 2013.
[8] T. Lin, W. Tsu, C. Chen, “Mortality of hepatoma and cirrhosis of liver in Taiwan,” British journal of cancer, vol.54, 969, 1986.
[9] Y. Qian, J.G. Fan, “Obesity, fatty liver and liver cancer,” Hepatobiliary & pancreatic diseases international : HBPD INT, vol.4, pp.173-177, 2005.
[10] E.A. Tsochatzis, J. Bosch, A.K. Burroughs, “Liver cirrhosis,” The Lancet, vol.383, pp.1749-1761, 2014.
[11] D. Schuppan, N.H. Afdhal, “Liver cirrhosis,” The Lancet, vol.371, pp.838-851, 2008.
[12] J.P. Iredale, “Models of liver fibrosis: exploring the dynamic nature of inflammation and repair in a solid organ,” The Journal of clinical investigation, vol.117, pp.539-548, 2007.
[13] J. Bruix, et al., “Surgical resection of hepatocellular carcinoma in cirrhotic patients: prognostic value of preoperative portal pressure,” Gastroenterology, vol.111, pp.1018-1022, 1996.
[14] L. Dai, et al., “Antifibrotic effects of ZK14, a novel nitric oxide-donating biphenyldicarboxylate derivative, on rat HSC-T6 cells and CCl4-induced hepatic fibrosis,” Acta Pharmacologica Sinica, vol.31, pp.27-34, 2010.
[15] T. Fujii, et al., “Mouse model of carbon tetrachloride induced liver fibrosis: Histopathological changes and expression of CD133 and epidermal growth factor,” BMC gastroenterology, vol.10, 79, 2010.
[16] J. Bruix, et al., “Clinical management of hepatocellular carcinoma. Conclusions of the Barcelona-2000 EASL conference,” Journal of hepatology, vol.35, pp.421-430, 2001.
[17] Research Section, Digital Retinal Image for Vessel Extraction (DRIVE) Database, Utrecht, The Netherlands, Univ. Med. Center Utrecht, Image Sci. Inst., 2009. nhttp://www.isi.uu.nl/Research/Databases/DRIVE
[18] AA-HA-R. Youssif, AZ. Ghalwash, AASA-R. Ghoneim, “Optic disc detection from normalized digital fundus images by means of a vessels direction matched filter,” IEEE Transactions on Medical Imaging, vol. 27, pp.11–18, 2008.
[19] Structured Analysis of the Retina (STARE) Project Website, Clemson, SC, Clemson Univ., 2009. http://www.clemson.edu/ces/
[20] M.J. Ruwart, et al., “The integrated value of serum procollagen III peptide over time predicts hepatic hydroxyproline content and stainable collagen in a model of dietary cirrhosis in the rat,” Hepatology, vol.10, pp.801-806, 1989.
[21] K. Huang, Q. Wang, Z. Wu, “Natural color image enhancement and evaluation algorithm based on human visual system,” Computer Vision and Image Understanding, vol.103, issue 1, pp. 52-63, 2006.
[22] A.M. Rera, “Realization of the contrast limited adaptive histogram equalization (CLAHE) for real-time image enhancement,” Journal of VLSI Signal Processing, vol.38, issue 1, pp. 35-44, 2004.
[23] R. Jegatha, K. Lakshmi, “Retinal blood vessel segmentation using gray-level and moment invariants-based features,” J. Comput. Appl., vol.5, issue 2,pp. 271-280, 2012.
[24] X.Z. Bai, F.G. Zhou, B.D. Xue, “Infrared image enhancement through contrast enhancement by using multiscale new top-hat transform,” Infrared Physics & Technology, vol.54, issue 2, pp. 61-69, 2011.
[25] M.M. Fraz, P. Remagnino, A. Hoppe, B. Uyyanonvara, A.R. Rudnicka, C.G. Owen, S.A. Barman, “An ensemble classification-based approach applied to retinal blood vessel segmentation,” IEEE Transactions on Biomedical Engineering, vol.59, issue 9, pp. 2538-2548, 2012.
[26] G. Azzopardi, N. Petkov, “Automatic detection of vascular bifurcations in segmented retinal images using trainable COSFIRE filters,” Pattern Recognition Letters, vol.34, issue 8, pp. 922-933, 2013.
[27] P. Feng, Y.J. Pan, B. Wei, W. Jin, D.L. Mi,” Enhancing retinal image by the Contourlet transform”, Pattern Recognition Letters, vol.28, issue 4, pp.516-552, 2007.
[28] A. Osareh, B. Shadgar, R. Markham, “A computational-intelligence-based approach for detection of exudates in diabetic retinopathy images”, IEEE Transactions on Information Technology in Biomedicine, vol.13, issue 4, pp.535-545, 2009.
[29] G.D. Joshi and J. Sivaswamy, “Colour retinal image enhancement based on domain knowledge”, Proc. of the IEEE Sixth Indian Conference on Computer Vision, Graphics and Image Processing, pp.591-598, 2008.
[30] A. Gandhamal, S. Talbar, S. Gajre, A. Fadzil, M. Hani, D. Kumar, “Local gray level S-curve transformation – A generalized contrast enhancement technique for medical images,” Computers in Biology and Medicine, vol.83, pp.120-133, 2017.
[31] E. Reinhard, M. Ashikhmin, B. Gooch, and P. Shirley, “Color transfer between images,” IEEE Computer Graphics and its Applications, pp. 34-41, 2001.
[32] Y. Chang, S. Saito, and M. Nakajima, “A framework for transfer colors based on the basic color categories”, Proceedings of the Computer Graphics International (CGI’03), pp.176-183, 2003.
[33] X. Xiao and L. Ma, “Color transfer in correlated color space,” VRCIA ’06: Proceedings of the 2006 ACM International Conference on Virtual Reality Continuum and its Applications, New York, NY, USA: ACM; pp. 305-309, 2006.
[34] A. Maslennikova and V. Vezhnevets, “Interactive local color transfer between images,” Proceedings of the International Conference on Computer Graphics & Vision (GraphiCon '07), pp. 75-78, 2007.
[35] C. L. Wen, C. H. Hsieh, B. Y. Chen, O. Ming, “Example-based multiple local color transfer by strokes,” Pacific Graphics, vol.27, no.7, pp.1765–1762, 2008.
[36] P.C. Chung, K.W. Chuang, C.C. Liu, C.Y. Yu, C.C. Huang, “An object color transformation scheme using regression analysis,” International Journal of Computer, Consumer and Control (IJ3C), vol. 1, no.2, pp.40-49, 2012.
[37] K.W. Chuang, C.C. Liu, G.N. Hu, S.S. Yu, S.W. Zheng, “A MRA-based color transformation scheme between objects,” 2012 International Symposium on Computer, Consumer and Control (IS3C), pp. 878 - 881, 2012.
[38] M. Emre Celebi, Hassan Kingravi, Fatih Celiker, “Fast color space transformations using minimax approximations,” IET Image Processing, vol. 4, no. 2, pp. 70-80, 2010.
[39] Wikipedia Foundation, Inc., Lab color space, Aug. 2014, http://en.wikipedia.org/wiki/Lab_color_space
[40] E Ricci, R Perfetti, “Retinal blood vessel segmentation using line operators and support vector classification,” IEEE Transactions on Medical Imaging, vol.26, no.10, pp. 1357-1365, 2007.
[41] Y. Yin, M. Adel, S. Bourennane, “Retinal vessel segmentation using a probabilistic tracking method,” Pattern Recognition, vol.45, issue 4, pp. 1235-1244, 2012.
[42] S. Shahbeig, “Automatic and quick blood vessels extraction algorithm in retinal images,” IET Image Processing, vol.7, issue 4, pp. 392-400, 2013.
[43] U.T. Nguyen, A. Bhuiyan, L.A. Park, R. Kawasaki, T.Y. Wong, J.J. Wang, P. Mitchell, K. Ramamohanarao, “An automated method for retinal arteriovenous nicking quantification from color fundus images,” IEEE Transactions On Biomedical Engineering, vol. 60, no. 11, pp. 3194-3203, 2013.
[44] A. Hoover, V. Kouznetsova and M. Goldbaum, “Locating blood vessels in retinal Images by piece-wise threshold probing of a matched filter response,” IEEE Transactions on Medical Imaging, vol. 19, no. 3, pp. 203-210, March 2000.
[45] A. Hoover and M. Goldbaum, “Locating the optic nerve in a retinal image using the fuzzy convergence of the blood vessels,” IEEE Transactions on Medical Imaging, vol. 22, no. 8, pp. 951-958, August 2003.
[46] J.J. Staal, M.D. Abramoff, M. Niemeijer, M.A. Viergever, B. van Ginneken, “Ridge based vessel segmentation in color images of the retina,” IEEE Transactions on Medical Imaging, vol. 23, issue 4, pp. 501-509, 2004.
[47] N. Otsu, “A threshold selection method from gray-level histograms,” IEEE Trans. Syst. Man. Cybern., vol.SMC-9, no.1, pp.62-66, 1979.
[48] M.A. Ebrahimi, M.H. Khoshtaghaza, S. Minaei, B. Jamshidi, “Vision-based pest detection based on SVM classification method,” Computers and Electronics in Agriculture, vol.137, pp.52-58, 2017.
[49] C.C. Liu, C.Y. Tsai, J. Liu, C.Y. Yu, S.S. Yu, “A pectoral muscle segmentation algorithm for digital mammograms using Otsu thresholding and multiple regression analysis,” Computers & Mathematics with Applications, vol.64, pp.1100-1107, 2012.
[50] S.G. Shu, H.H. Lin, S.W. Kuo, S.S. Yu, “Excluding background initial segmentation for radiographic image segmentation,” IJICIC, vol.5, no.11(A), pp.3849-3860, 2009.
[51] R. Geetha Ramani, L. Balasubramanian, “Retinal blood vessel segmentation employing image processing and data mining techniques for computerized retinal image analysis,” Biocybernetics and Biomedical Engineering, vol.36, pp.102-118, 2016.
[52] M. Wdowiak, T. Markiewicz, S. Osowski, J. Patera, W. Kozlowski, “Novel segmentation algorithm for identification of cell membrane staining in HER2 images,” Pattern Recognition Letters, vol.84, pp.225-231, 2016.
[53] C.C. Liu, C.Y. Tsai, T.S. Tsui, S.S. Yu, “An improved GVF snake based breast region extrapolation scheme for digital mammograms,” ESWA, vol.39, pp.4505-4510, 2012.
[54] E. Imani, M. Javidi, H.R. Pourreza, “Improvement of retinal blood vessel detection using morphological component analysis,” Computer Methods and Programs in Biomedicine, vol.118, pp.263-279, 2015.
[55] Y. Chen, P.W. Hao, “Optimal transform in perceptually uniform color space and its application in image retrieval,” The 7th International Conference on Signal Processing, vol. 2, pp. 1107-1110, 2004.
[56] X.Z. Bai, F.G. Zhou, B.D. Xue, “Image enhancement using multi scale image features extracted by top-hat transform,” Optics & Laser Technology, vol.44, no. 2, pp. 328-36, 2012.
[57] M. Liao, Y.Q. Zhao, X.H. Wang, P.S. Dai, “Retinal vessel enhancement based on multi-scale top-hat transformation and histogram fitting stretching,” Optics & Laser Technology, vol. 58, pp. 56-62, 2014.
[58] Y. Yun, “Hybrid genetic algorithm with adaptive local search scheme,” Computers & Industrial Engineering, vol.51, no.1, pp.128-141, 2006.
[59] V. Raghavan, P. Bollmann, G.S. Jung, “A critical investigation of recall and precision as measures of retrieval system performance,” ACM Transactions on Information Systems (TOIS), vol.7, no.3, pp.205-229, 1989.
[60] J.S. Syu, Segmentation Blood Vessels in Retinal Image Based on Modified Multi-Scale Line Detection, Master Thesis, National Chung-Hsing University, Taichung, 2017.
[61] V. Bhateja, M. Misra, S. Urooj, “Non-linear polynomial filters for edge enhancement of mammogram lesions,” Computer Methods and Programs in Biomedicine, vol.129, pp.125-134, 2016.
[62] J.M. Chaves-González, M.A. Vega-Rodríguez, J.A. Gómez-Pulido, J.M. Sánchez-Pérez, “Detecting skin in face recognition systems: A colour spaces study,” Digital Signal Processing, vol.20, no.3, pp.806-823, 2010.
[63] A. Hanbury, “Constructing cylindrical coordinate colour spaces,” Pattern Recognition Letters, vol.29, no.4, pp.494-500, 2008.
[64] C.W. Park, J.Y. Ryu, “Development of a new automatic gamma control system for mobile LCD applications,” Displays, vol.29, no.4, pp.393-400, 2008.
[65] N. Otsu, “A threshold selection method from gray-level histograms,” Automatica, vol.11(285-296), pp.23-27, 1975.
[66] H. Samet, M. Tamminen, “Efficient component labeling of images of arbitrary dimension represented by linear bintrees,” Pattern Analysis and Machine Intelligence, IEEE Transactions on, vol.10, no.4, pp.579-586, 1988.
[67] S. Booth, D.A. Clausi, “Image segmentation using MRI vertebral cross-sections,” IEEE Proceedings of the IEEE Canadian Conference on Electrical and Computer Engineering, vol.2, pp.1303-1307, 2001.
[68] R.C. Gonzalez, R.E. Woods, Digital Image Processing, (4th Edition) 4th Edition, Pearson-Prentice-Hall, 2008.
[69] J. MacQueen, “Some methods for classification and analysis of multivariate observations,” In Proceedings of the fifth Berkeley symposium on mathematical statistics and probability, vol. 1, pp. 281-297: Oakland, CA, USA, 1967.
[70] K.F. Man, K.S. TANG, S. Kwong, Genetic algorithms: concepts and designs: Springer Science & Business Media, 2012.
[71] J. Sanchis, M.A. Martínez, X. Blasco, “Integrated multiobjective optimization and a priori preferences using genetic algorithms,” Information Sciences, vol.178, no.4, pp.931-951, 2008.
[72] Y. Yun, “Hybrid genetic algorithm with adaptive local search scheme,” Computers & Industrial Engineering, vol.51, no.1, pp.128-141, 2006.