臺灣博碩士論文加值系統

肝臟是人體最重要的器官，在台灣肝臟疾病趨於普遍，主要是肝細胞發炎或是肝細表的損壞，由全世界調查顯示每年新的肝癌相關疾病患者數超過100萬人，原因為肝臟疾病早期很難被觀察診斷出來，都是已經出現嚴重的病症，患者才感覺出來，故早期診斷概念受到廣泛的重視。
所以透過醫學影像處理技術，來防護相關疾病發生的可能性，邊際檢測是主要前處理階段，針對目標區域進行切割，而影像透過共生矩陣的輔助，能讓醫學圖像數值化，然而在辨識這些紋理特徵上，百家爭鳴各種方法皆有，但很多方法經常出現找不到最重要幾個特徵集合或是刪除相關或相關性大的特徵，讓圖像的正確辨識率降低。本研究呈現出對於是否有血色肝癌病患的影像，較為關鍵性的特徵集合作分類；首先經由GVF Snake方法來獲取肝臟部位區域影像，加上紋理分析和分類來描述這些影像相關的特徵值；應用不同的特徵選擇方法，找出相關的特徵集來建構出一個分類的知識庫；利用交叉比對來評估分類器的正確性；最終提昇診斷之效益，可作為未來電腦輔助診斷之參考。

The liver is the most complex human organ, and the liver disease patients tend to common in Taiwan, Inflammation of the liver cells which are mainly small table, or liver damage, a new worldwide survey shows that the number of new liver cancer patients more than 100 million people, because the liver disease symptoms is difficult to observe diagnosed, it always are already serious illness, the patient was feeling it, so the concept of early diagnosis of widespread attention.
Therefore, the texture features using the typical way for the value of medical images, the marginal detection is the main pre-treatment stage for the target area to be cut, but in recognition of these texture methods, several features have some shortcomings that several features are not the most important or relevant, or delete the relevant characteristics of large, so that the correct image recognition rate are affected by it.
This study shows whether classification of liver-related diseases, first way to get through the active contours regions, coupled with texture analysis and classification to describe the characteristic values associated with these images, using a variety of feature selection methods to identify associated set to construct a classification knowledgebase; using cross validation to assess the correctness of classification , so we hope enhance improve the diagnosis to doctor, and be used as a reference for computer-aided diagnosis.

壹、 緒論 1
一、 研究背景 1
二、 研究動機 2
三、 研究目的 4
四、 研究流程 4
貳、 文獻探討 5
一、 磁振造影圖像(Magnetic Resonance Image; MRI) 5
二、 邊界檢測(Border detection) 6
三、 目標區的分割(Segmentation) 8
四、 紋理特徵(Texture Feature) 10
五、 特徵選擇(Feature Selection) 14
參、 研究方法 17
一、 資料蒐集(Data Set Collection;MRI) 18
二、 邊際檢測(Border Detection) 18
三、 灰階共生矩陣(Gray Level co-occurrence Matrix;GLCM) 20
四、 主成分分析(Principal Analysis) 26
五、 傳統特徵選擇(Feature Selection) 30
六、 分類模式(Classification Model) 34
七、 評估方法(Cross-Validation Method) 35
肆、 實驗與結果分析 37
一、實驗環境 37
二、實驗資料 37
三、實驗結果與討論 37
伍、 結論與未來研究方向 51
一、結論 51
二、未來研究方向 52
參考文獻 53
附錄 57

[1] 王晉中(2010)，「Matlab 7在工程上的應用」，台北：高立圖書有限公司。
[2] 莊克士(2006)，「醫學影像物理學」，台北：合記圖書出版社。
[3] 鄭慶明(2006)，「醫學影像診斷學」，台北：合記圖書出版社。
[4] 張簡復(2010)，「數位影像處理使用Matlab 7.X」，台北：松崗管理股份有限公司。
[5] Lin, R.H(2009), An intelligent Model for Liver Disease Diagnosis, Artificial Intelligence in Medicine, 47, pp. 53-62.
[6] Esneault, S., Lafon, C., and Dillenseger, J.L(2010), Liver Vessels Segmentation Using a Hybrid Geometrical Moments / Graph Cuts Method, IEEE Transactions on Biomedical Engineering, 57(2), pp.276-283.
[7] Campadelli, P., Casiraghia, E., and Espositob, A(2009), Liver Segmentation from Computed Tomography Scans：a Survey and a New Algorithm, Artificial Intelligence in Medicine, 45, pp.185-196.
[8] Poyanli, A(1999), Computed Tomography Scan of the Liver, European Journal of Radiology, 32, pp.15-20.
[9] Schwenzer, N.F., Springer, F., Schraml, C., Stefan, N., Machann, J., and Chick, F(2009), Non-invasive Assessment and Quantification of Liver Steatosis by Ultrasound, Computed Tomography and Magnetic Resonance, Journal of Hepatology, 51(3), pp.433-445.
[10] Chen, E.L., Tsai, H.M., and Chang, C.I(1998), An Automatic Diagnostic System for CT Liver Image Classification, IEEE Transactions on Biomedical Engineering, 45(6), pp.783-794.
[11]Mougiakakou, S., and Alexandra, K.I(2007), Differential Diagnosis of CT Focal Liver Lesions Using Texture Features, Feature Selection and Ensemble Driven Classifiers, Artificial Intelligence in Medicine, 41, pp. 25-37.
[12]Haralick, R.M., and Shanmugam, K(1973), Textural Features for Image Classification, IEEE Transactions on Medical Imaging, 3, pp.610-621.
[13]Haralick, R.M(1979), Statistical and Structural Approaches to Texture, IEEE Transactions on Medical Imaging, 67, pp. 786-806.
[14]Daniel, J. “Texture Analysis of Human Liver(2002), Journal of Magnetic Resonance Imaging, 68, pp.68-74.
[15]Svoles, A.E., and Todd, P.A(1998), Time and Space Results of Dynamic Texture Feature Extraction in MR and CT Image Analysis, IEEE Transactions on Information Technology in Biomedicine, 21(2), pp.48-54.
[16]Chen, K.L(1995), Quantitative Characterization of Electron Micrograph Image Using Fractal Feature, IEEE Transaction on Information in Biomedicine, 42, pp.1033-1037.
[17]Lee, C.C., and Shih, C.Y. Classification of Liver Disease from CT Images Using Sigmoid Radial Basis – Function Neural Network(2009), World Congress on Computer Science and Information Engineering, 24, pp. 656-660.
[18]Wu, C.M., Chen, Y.C., and Hsieh, K.S(1992), Texture Features for Classification in Ultrasonic Liver Images, IEEE Transaction on Information Technology in Biomedicine, 11, pp.141-152.
[19]Pauls, S., Andreas, G., Wolfgang, H., Eckhart, F., Sandra, P., and Stefan, A.S(2009), Liver Perfusion with Dynamic Multidetector-Row Computed Tomography as an Objective Method to Evaluate the Efficacy of Chemotherapy in Patients with Colorectal Cancer, 33, pp.289-294.
[20]Celebi, M.E., Kingravia, H.A., Uddina, B., and Lyatomid, H(2000), A Methodological Approach to the classification of Dermoscopy Images, Computerized Medical Imaging and Medical Imaging and Graphics, 31, pp.362-373.
[21]Lixu, G., Chen, G., Qian, L., and Xu, J(2009), An Improved Level Set for Liver Segmentation and Perfusion Analysis in MRIs, IEEE Transactions on Information Technology in Biomedicine, 13(1), pp.94-103.
[22]Xu, C., and Prince, J.L(2004) Snakes, Shapes, and Gradient Vector Flow, IEEE Transactions on Image Processing, 7(3), pp.94-106.
[23]Van, B.G., Heimann, T., Styner, M.A., Arzhaeva, Y., Aurich, V., Bauer, C., Beck, A., Becker, C., Beichel, R., Bekes, G., Bello, F., Binnig, G., Bischof, H., Bornik, A., Cashman, P., Ying, C., Cordova, A., Dawant, B.M., Fidrich, M., and Furst, J.D(2009), Comparison and Evaluation of Methods for Liver Segmentation From CT Datasets, IEEE Transactions on Medical Imaging, 28(8), pp.1251-1265.
[24]Wen, L.W., and Hsiehb, K.S(2010), A Robust Algorithm for the Fractal Dimension of Images and its Applications to the Classification of Natural Images and Ultrasonic Liver Images, Signal Processing, 90, pp. 1897-1904.
[25]Mougiakakou, M.G., Gletsos, M., Matsopoulos, G.K., Nikita, K.S., Nikita, A.S., and Kelekis, D(2003), A Computer – Aided Diagnostic System to Characterize CT Focal Liver Lesions: Design and Optimization of a Neural Network Classifier, IEEE Transaction on Information Technology in Biomedical, 7(3), pp.153-162.
[26]Yu, L., and Liu, H(2005), Toward Integrating Feature Selection Algorithm for Classification and Clustering, IEEE Transaction on Information Technology in Biomedicine, 17(4), pp.491-502.
[27]Liu, J., and Iba, H(2001), Selecting Informative Genes with Parallel Genetic Algorithms in Tissue Classification, Genorm Information, 12, pp.14-23.
[28]Chung, S.H., and Ho, S.B(2008), Classification of Liver Diseases MRI Images Using First-Order Statistics – Grid Computing Approach, IEEE Transaction of information Technology in Biomedicine, 5, pp.25-27.
[29]Ludwig, O., and Nunes, U(2008), Novel Maximum - Margin Training Algorithms for Supervised Neural Networks, IEEE Transactions on Neural Networks, 6, pp.972-984.
[30]Silveira, M., and Marques, J.(2006), Automatic Segmentaion of the lungs using Multiple Active Contours and Outlier Model, IEEE Transaction on Information Technology in Biomedicine, 28, pp.63-69.
[31]Celebi, M.E., Kingravi, H., Uddin, B., Iyatomi, H., Aslandogan, Y.A., Stoecker, W.V., and Moss, R.H(2007), A Methodological Approach to the Classification of Dermoscopy Images, Computer Medical Imaging and Graphics, 31, pp.362-373.
[32]Fung, G.M., and Mangasarian, O.L(2004), A Feature Selection Newton Method for Support Vector Machine Classification, Computational and Applications, 28, pp.185-202.
[33]Rief, M., Wagner, M., Franiel, T., Bresan, V., Taupitz, M., Klessen, C., Hamm, B., and Asbach, P(2009) Detection of Focal Liver Lesions in Unenhanced and Ferucarbotran-Enhanced Magnetic Resonance Imaging：A comparison of T2-Weighted Breath-Hold and Respiratory-Triggered Sequences, Science Direct, 27, pp.1223-1229.