簡易檢索 / 詳目顯示

回結果列表
研究生： 黃冠維
Huang, Kuan-Wei
論文名稱： 基於卷積神經網路於三倍頻顯微術影像中之幹細胞偵測
Stem Cell Detection based on Convolutional Neural Network via Third Harmonic Generation Microscopy Images
指導教授： 李國君
Lee, Gwo-Giun
學位類別： 碩士
Master
系所名稱： 電機資訊學院 - 電機工程學系
Department of Electrical Engineering
論文出版年： 2017
畢業學年度： 106
語文別： 英文
論文頁數： 74
中文關鍵詞： 電腦輔助診斷 、三倍頻顯微術 、細胞分割演算法 、三維重建技術 、卷積神經網路 、手調式初始化
外文關鍵詞： Computer-Aided Diagnosis (CAD), Third Harmonic Generation (THG), cell segmentation algorithm, three-dimensional reconstruction, Convolutional Neural Network (CNNs), hand-crafted initialization
相關次數： 點閱：67  下載：0 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 幹細胞在於修復受損的組織以及維持人類每一天的健康中扮演一個重要的角色，因此在醫師們分析幹細胞之前，幹細胞的觀察與偵測是最首要的步驟。在這篇論文中，基於細胞分割演算法以及細胞本質的特徵，我們提出一個電腦副助診斷系統的標準來偵測位於基底層的幹細胞，此方法提供一個一致性與精準的結果來協助診斷的評估。
    除此之外，我們使用卷積神經網路來識別基底細胞與幹細胞，因為卷積神經網路在處理非常大量的數據時，有非常傑出的成果。實際上，在取得生物醫學影像的過程是非常複雜的，以至於很難收集到足夠的影像。因此，根據預先知道的知識或醫生的建議，我們採用手調式初始化來克服訓練資料不足的問題。實驗結果顯示出手調式初始化的精準度比隨機分佈的核較高、在收斂時間上也比較短，因為在最佳化理論中，較好的初始值可能會得到較好的結果。

    Stem cell plays an important role in repairing destroyed tissues and keeping human healthy every day; and thus stem cell observation and detection are principle procedures before analyzed by physicians. In this thesis, we proposed a criterion of Computer-Aided Diagnosis (CAD) system to detect stem cells in the stratum basale based on cell segmentation algorithm and intrinsic characteristics of cells, which can provide consistent and accurate results for assisting the assessment of diagnosis.
    In addition, we utilize Convolutional Neural Networks (CNNs) to recognize basal cells and stem cells since CNN has excellent performance on processing abundant data. Actually, the procedure of acquiring biomedical images is too complicated to collect, hence hand-crafted initialization is adopted to overcome the issue of the lack of training data according to prior knowledge or the suggestion from medical doctors. The experimental results show that the accuracy of hand-crafted initialization is higher than random distribution kernels and the convergence time is shorter also since a better condition may lead better results in optimization theory.

    摘 要 i Abstract iii 誌 謝 v Table of Contents vii List of Tables x List of Figures xi Chapter 1 Introduction 1 1.1 Introduction 1 1.2 Background Information 2 1.2.1 Traditional Biopsy 2 1.2.2 Optical in vivo Virtual Biopsy 3 1.2.2.1 Background of the Acquired Images 3 1.2.2.2 Setup of the Optical Image 4 1.3 Human Skin 4 1.4 Background of Stem Cell 5 1.5 Motivation 7 1.6 Organization of this Thesis 8 Chapter 2 Surveys of Related Works for the Literatures 9 2.1 Related work of Classification 9 2.1.1 Neural Network 9 2.1.2 Support Vector Machine 11 2.1.3 Convolutional Neural Network 13 2.2 Related Work of Clustering 14 2.2.1 K-means Clustering 14 2.2.2 Restricted Boltzmann Machine 16 2.2.3 Convolutional AutoEencoder 18 2.3 Data Pre-processing for Neural Networks 18 2.3.1 Batch Normalization 19 2.3.2 K-Fold Cross Validation 20 2.3.3 Dropout 21 Chapter 3 Overview of Cell Segmentation Algorithm 23 3.1 Nuclei Segmentation Algorithm and Block Diagram 23 3.1.1 Nuclei Initialization 25 3.1.1.1 Blob Detection 26 3.1.1.2 Outlier Removal 28 3.1.1.3 Efficient Ellipse Hough Transform 30 3.1.1.4 Distance Transform and Cell Boundary Initialization 31 3.1.2 Gradient Map Generation 32 3.1.3 Nuclei Detection 33 3.1.4 Nuclei Validation 35 3.2 Cytoplasm Boundary Initialization and Block Diagram 35 3.2.1 Introduction of Basic Active Contour Model and Snake Force 36 3.2.2 Cytoplasm Initialization 38 3.2.3 Cytoplasm Detection 39 Chapter 4 Proposed Methodology 41 4.1 Stem Cell Detection 43 4.2 Patch Generation 46 4.3 Hand-crafted Initialization for Convolutional Neural Networks 46 4.3.1 Framework Design of Convolutional Neural Network 47 4.3.1.1 Xavier Initialization 48 4.3.1.2 Gaussian Filter Initialization 50 4.3.2 Activation Function 52 4.3.3 Stochastic Gradient Descent 53 Chapter 5 Experimental Results and Discussion 55 5.1 Epidermal Stem Cells Detection 55 5.2 Epidermal Stem Cells Classification based on Convolutional Neural Network 60 5.2.1 Hand-crafted Initialization 61 5.3 Comparison and Discussion 65 Chapter 6 Conclusion and Future Work 68 6.1 Conclusion 68 6.2 Future Work 68 Acknowledgments 70 References 71

    [1]. C.-K. Sun, “Higher harmonice generation microscopy,” in Microscopy Techniques, ed: Springer, 2005, pp. 17-56

    [2]. 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.

    [3]. C.-S. Hsieh, S.-U. Chen, Y.-W. Lee, Y.-S. Yang, and C.-K. Sun, "Higher harmonic generation microscopy of in vitro cultured mammal oocytes and embryos," Optics express, vol. 16, pp. 11574-11588, 2008.

    [4]. 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.

    [5]. Alberts B., Johnson A., Lewis J., Raff M., Roberts K., and Walter P., “Epidermis and Its Renewal by Stem Cells,” Molecular Biology of the Cell, 4th edition, 2002.

    [6]. J.-X. Zhou, S.-Y. Chen, W.-M. Liu, Y.-J. Cao and E.-K Duan, “Enrichment and identification of human ‘fetal’ epidermal stem cells.” Human Reproduction, vol. 19, no. 4, pp.968-974, 2004.

    [7]. T. Kohonen, "An introduction to neural computing," Neural networks, vol. 1, pp. 3-16, 1988.

    [8]. C. Cortes, and V. Vapnik, “Support-vector networks,” Machine learning, vol. 20, no. 3, pp. 273-297, 1995.

    [9]. N. Friedman, D. Geiger, and Goldszmidt M. “Bayesian network classifiers,” Machine Learning, 29:131–163, 1997.

    [10]. S. Haykin and N. Network, "A comprehensive foundation," Neural Networks, vol. 2, 2004.

    [11]. Y. LeCun, L. Bottou, G. B. Orr, and K.-R. Müller, "Efficient BackProp," in Neural Networks: Tricks of the Trade, G. B. Orr and K.-R. Müller, Eds., ed Berlin, Heidelberg: Springer Berlin Heidelberg, pp. 9-50, 1998.

    [12]. Lecun, Y. (1989). Generalization and network design strategies. In R. Pfeifer, Z. Schreter, F. Fogelman, & L. Steels (Eds.), Connectionism in perspective. Zurich, Switzerland: Elsevier.

    [13]. Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner, "Gradient-based learning applied to document recognition," Proceedings of the IEEE, vol. 86, pp. 2278-2324, 1998.

    [14]. J. A. Hartigan and M. A. Wong, "Algorithm AS 136: A k-means clustering algorithm," Applied statistics, pp. 100-108, 1979.

    [15]. C. M. Bishop, Pattern recognition and machine learning vol. 1: springer New York, 2006.

    [16]. A. Fischer, and C. Igel, “An Introduction to Restricted Boltzmann Machines.” Lecture Notes in Computer Science, Berlin: Springer, vol. 7441, pp. 14–36, 2012

    [17]. G. E. Hinton, "Training products of experts by minimizing contrastive divergence," Neural computation, vol. 14, pp. 1771-1800, 2002.

    [18]. W. R. Gilks and P. Wild, "Adaptive rejection sampling for Gibbs sampling," Applied Statistics, pp. 337-348, 1992.

    [19]. A. Ng, "Sparse autoencoder," CS294A Lecture notes, vol. 72, pp. 1-19, 2011.

    [20]. J. Masci, U. Meier, D. Cireşan, and J. Schmidhuber, "Stacked convolutional auto-encoders for hierarchical feature extraction," in International Conference on Artificial Neural Networks, 2011, pp. 52-59.

    [21]. S. Ioffe and C. Szegedy, "Batch normalization: Accelerating deep network training by reducing internal covariate shift," arXiv preprint arXiv:1502.03167, 2015.

    [22]. G. G. Lee, H.-H. Lin, M.-R. Tsai, S.-Y. Chou, W.-J. Lee, Y.-H. Liao, et al., "Automatic cell segmentation and nuclear-to-cytoplasmic ratio analysis for third harmonic generated microscopy medical images," Biomedical Circuits and Systems, IEEE Transactions on, vol. 7, pp. 158-168, 2013.

    [23]. Y.-H. Chou, S.-Y. Hung, G.-G. (Chris) Lee, W.-H. Weng, Y.-H. Liao, C.-K. Sun, H.-T. Shih, “Nuclei location enhancement based on improved efficient ellipse Hough transform for third harmonic generation microscopy imaging,” 2015 IEEE China Summit and International Conference on Signal and Information Processing, Chengdu, pp. 807 - 811, Jul. 2015.

    [24]. T. Lindeberg, "Feature detection with automatic scale selection," International journal of computer vision, vol. 30, pp. 79-116, 1998.

    [25]. J. Illingworth and J. Kittler, "A survey of the Hough transform," Computer vision, graphics, and image processing, vol. 44, pp. 87-116, 1988.

    [26]. Y. Xie and Q. Ji, "A new efficient ellipse detection method," in Pattern Recognition, 2002. Proceedings. 16th International Conference on, 2002, pp. 957-960.

    [27]. Rafael C. Gonzalez, Richard E. Woods, Digital Image Processing, 3rd Edition, Pearson Education, 2008.

    [28]. D. W. Paglieroni, "Distance transforms: Properties and machine vision applications," CVGIP: Graphical models and image processing, vol. 54, pp. 56-74, 1992.

    [29]. L. Vincent and P. Soille, "Watersheds in digital spaces: an efficient algorithm based on immersion simulations," IEEE Transactions on Pattern Analysis & Machine Intelligence, pp. 583-598, 1991.

    [30]. J. M. Gauch, "Image segmentation and analysis via multiscale gradient watershed hierarchies," Image Processing, IEEE Transactions on, vol. 8, pp. 69-79, 1999.

    [31]. A. M. MacEachren, "Compactness of geographic shape: Comparison and evaluation of measures," Geografiska Annaler. Series B. Human Geography, pp. 53-67, 1985.

    [32]. M. Kass, A. Witkin, and D. Terzopoulos, "Snakes: Active contour models," International journal of computer vision, vol. 1, pp. 321-331, 1988.

    [33]. L. D. Cohen, "On active contour models and balloons," CVGIP: Image understanding, vol. 53, pp. 211-218, 1991.

    [34]. C. Xu and J. L. Prince, "Snakes, shapes, and gradient vector flow," Image Processing, IEEE Transactions on, vol. 7, pp. 359-369, 1998.

    [35]. 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.

    [36]. A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet classification with deep convolutional neural networks,” Advances in Neural Information Processing Systems 25, vol. 60, pp. 1097-1105, 2012.

    [37]. X. Glorot, Y. Bengio, “Understanding the difficulty of training deep feedforward neural networks,” Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, vol. 9: 249-256, 2010.

    [38]. X. Glorot, A. Bordes and Y. Bengio. “Deep Sparse Rectifier Neural Networks,” Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics (AISTATS-11), vol. 15, pp. 315-323, 2011.

    [39]. A. L. Maas, A. Y. Hannun, A. Y. Ng, “Rectifier Nonlinearities Improve Neural Network Acoustic Models,” ICML Workshop on Deep Learning for Audio, Speech, and Language Processing, 2013.

    [40]. L. Bottou, "Stochastic gradient descent tricks." Neural Networks: Tricks of the Trade. Springer Berlin Heidelberg, vol. 7700, pp. 421-436, 2012.

    無法下載圖示 校內：2022-12-29公開
    校外：不公開
    電子論文尚未授權公開，紙本請查館藏目錄
    QR CODE