跳到主要內容

簡易檢索 / 詳目顯示

回結果列表
研究生： 唐嘉梅
Chia-Mei Tang
論文名稱： 深度學習於醫學影像處理之應用
Applications of Deep Learning in Medical Image Processing
指導教授： 蘇木春
Mu-Chun Su
口試委員:
學位類別： 碩士
Master
系所名稱： 資訊電機學院 - 資訊工程學系
Department of Computer Science & Information Engineering
論文出版年： 2018
畢業學年度： 106
語文別： 中文
論文頁數： 102
中文關鍵詞： 醫學影像 、影像處理 、深度學習 、鈣化偵測 、壓迫性骨折偵測
外文關鍵詞： medical image, image processing, deep learning, calcified plaques detection, vertebral compression fractures detection
相關次數： 點閱：14  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

    • 近年來深度學習是個非常熱門的議題，許多事情都可以用深度學習來解決，本論文希望能將深度學習應用在醫學上，希望能幫助醫師節省判斷時間並降低誤判率，也可以讓就醫的人可以及早發現問題，及早治療。
    本論文用CNN建立一套可以自動偵測心臟血管鈣化系統，系統主要有三個功能。一、結合影像處理的方法，擷取出心臟血管影像；二、利用CNN架構來訓練偵測鈣化的模型；三、使用影像處理的方法來定位鈣化斑塊的所在位置。
    另外本論文也用Mask R-CNN建立另一套偵測脊椎壓迫性骨折的系統，系統主要有兩大功能。一、利用Mask R-CNN訓練分割脊椎骨影像的模型；二、利用Mask R-CNN訓練偵測四個點頂的模型，並用結果來評估是否為壓迫性骨折。
    本系統的心臟血管擷取演算法有成功擷取出87%的血管，鈣化偵測準確率也有接近8成，誤判的部分每1張影像中僅會出現0.21個鈣化斑塊誤判，鈣化位置定位效果也有7成以上，這些數據說明本論文所使用的方法能有效的找出心臟血管鈣化斑塊及鈣化位置。
    分割脊椎骨影像演算法能夠將97%以上的脊椎骨影像分割出來，找脊椎骨影像上的六個頂點正確率達到96%，偵測脊椎壓迫性骨折形狀及程度正確率也都有達到7成以上，應能夠輔助醫師有效的偵測出脊椎壓迫性骨折。

    In recent years, deep learning is a very hot topic. People can solve a lot of problem with deep learning. This paper hopes to apply deep learning in medical area, hoping to help doctors save their judgment time and reduce misjudgment rate. Thus doctors can help patient to find diseases earlier and treat it earlier.
    This paper uses Convolutional Neural Network to build a system that can automatically detect calcified plaque in cardiac blood vessels. The system has three main features. First, extract cardiac blood vessels images by using image processing methods. Second, use Convolutional Neural Network to train a calcified plaque detection model. Third, locate calcified plaque by using image processing methods.
    Moreover, this paper also builds another system for detecting vertebral compression fractures by Mask R-CNN. The system has two major features. First, train a model which can separate spines images from vertebral medical images by using Mask R-CNN. Second, train another model to detect the four vertex by using Mask R-CNN. Finally using these results to do the vertebral compression fractures detection.
    The cardiac blood vessels extraction algorithm of this system has succeeded extract 87% of cardiac blood vessels completely, and the accuracy of calcified plaque detection almost reach 80%. The misjudgment is about 0.21 calcified plaques per image. And also location accuracy of calcified plaque is more than 70%. For these tests proves that the method we used in this paper can efficiently find the calcified plaque on cardiac blood vessels.
    The segmentation of vertebral images algorithm can separate 97% or more of vertebral bones images, and also find 6 vertices of each vertebral bones images correctly. 70% or more of vertebral compression fractures has been detected by the system. These results indicate that the system can assist doctors to detect vertebral compression fractures efficiently.

    摘 要 i ABSTRACT iii 誌 謝 v 目 錄 vii 圖目錄 ix 表目錄 xiii 第一章、 緒論 1 1-1 研究動機 1 1-2 研究目的 3 1-3 論文架構 4 第二章、 相關研究 5 2-1 心臟醫學影像研究 5 2-1-1 心臟血管介紹 5 2-1-2 病變種類 7 2-1-3 心臟血管影像資料集與論文探討 9 2-2 脊椎醫學影像研究 11 2-2-1 脊椎骨介紹 11 2-2-2 病變種類 12 2-2-3 脊椎影像資料集與論文探討 12 2-3 深度學習相關研究 15 2-3-1 CNN 15 2-3-2 Mask R-CNN 19 2-3-3 深度學習網路開發套件 20 第三章、 研究方法 21 3-1 心臟血管鈣化偵測 21 3-1-1 軟體流程架構 21 3-1-2 擷取心臟血管影像演算法 21 3-1-3 鈣化偵測演算法 27 3-1-4 鈣化位置定位演算法 30 3-2 脊椎壓迫性骨折偵測 39 3-2-1 軟體流程架構 39 3-2-2 分割脊椎骨影像演算法 39 3-2-3 脊椎壓迫性骨折偵測演算法 44 第四章、 實驗設計與結果分析 51 4-1 心臟血管鈣化偵測實驗 51 4-1-1 實驗設計與評估指標 51 4-1-2 心臟血管擷取實驗結果與分析 53 4-1-3 判定鈣化區域的閥值實驗 56 4-1-4 CNN層數實驗結果與分析 57 4-1-5 鈣化位置定位實驗結果與分析 59 4-2 脊椎壓迫性骨折偵測實驗 62 4-2-1 找脊椎CT影像最適二值化閥值實驗 62 4-2-2 偵測脊椎骨影像及脊椎骨影像六個頂點實驗結果與分析 69 4-2-3 脊椎壓迫性骨折偵測實驗結果與分析 73 第五章、 結論與展望 80 5-1 結論 80 5-2 未來展望 82 參考文獻 83

    1. 衛生福利部統計室, "國人各年度十大死因統計結果," [Online]. Available: https://dep.mohw.gov.tw/DOS/np-1776-113.html. [Accessed 20 - June - 2018].
    2. 中華民國骨質疏鬆症學會, "2017 台灣成人骨質疏鬆症防治之共識及指引," [Online]. Available: http://www.toa1997.org.tw/download/2017_TOA_guideline-0615.pdf. [Accessed 12 - June - 2018].
    3. 萬芳醫院, 心臟血管圖, 2018.
    4. 衛生福利部國民健康署, "認識冠心病," [Online]. Available: https://www.hpa.gov.tw/Pages/Detail.aspx?nodeid=632&pid=1188. [Accessed 21 - June - 2018].
    5. I. Išgum , A. Rutten, M. Prokop, and B. van Ginneken, "Detection of coronary calcifications from computed tomography scans for automated risk assessment of coronary artery disease," Medical Physics, vol. 34, no. 4, pp. 1450-1461, April 2017.
    6. A. Jain and D. Zongker, “Feature selection: evaluation, application, and small sample performance,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 19, no. 2, pp. 153-158, 1997.
    7. "Wikipedia - Vertebral column," [Online]. Available: https://en.wikipedia.org/wiki/Vertebral_column. [Accessed 21 - June - 2018].
    8. H. K. Genant, M. Jergas, L. Palermo, M. Nevitt, R. S. Valentin, D. Black, and S. R. Cummings, "Comparison of semiquantitative visual and quantitative morphometric assessment of prevalent and incident vertebral fractures in osteoporosis," Journal of Bone and Mineral Research, vol. 11, no. 7, pp. 984-996, July 1996.
    9. M. Grigoryan, A. Guermazi, F. W. Roemer, P. D. Delmas, and H. K. Genant, “Recognizing and reporting osteoporotic vertebral fractures,” European Spine Journal, vol. 12, Suppl. 2, pp. S104-S112, October 2003.
    10. H. K. Genant, C. Y. Wu, C. V. Kuijk, and M. C. Nevitt, "Vertebral Fracture Assessment Using a Semiquantitative Technique," Journal of Bone and Mineral Research, vol. 8, no. 9, pp. 1137-1148, September 1993.
    11. J. E. Burns, J. Yao, and R. M. Summers, "Vertebral Body Compression Fractures and Bone Density: Automated Detection and Classification on CT Images," Radiology, vol. 284, no. 3, pp. 788-797, September 2017.
    12. T. Baum, J. S. Bauer, T. Klinder, M. Dobritz, E. J. Rummeny, P. B. Noël, and C. Lorenz, "Automatic detection of osteoporotic vertebral fractures in routine thoracic and abdominal MDCT," European Radiology, vol. 24, no. 4, pp. 872-880, April 2014.
    13. J. Yao, S. D. O’Connor, and R. M. Summers, “Automated spinal column extraction and partitioning,” 3rd IEEE International Symposium on Biomedical Imaging: Nano to Macro, 2006., pp. 390-393, 2006.
    14. "Wikipedia - Convolutional neural network," [Online]. Available: https://en.wikipedia.org/wiki/Convolutional_neural_network. [Accessed 21 - June - 2018].
    15. "Convolutional Neural Networks," [Online]. Available: http://cs231n.github.io/convolutional-networks/. [Accessed 22 - June - 2018].
    16. H. Lee, R. Grosse, R. Ranganath, and A. Y. Ng, "Convolutional Deep Belief Networks for Scalable Unsupervised Learning of Hierarchical Representations," in Proceedings of the 26th Annual International Conference on Machine Learning, pp. 609-616, 2009.
    17. K. He, G. Gkioxari, P. Dollár, and R. Girshick, "Mask R-CNN," in IEEE International Conference on Computer Vision, pp. 2980-2988, 2017.
    18. R. Girshick, J. Donahue , T. Darrell, and J. Malik, "Rich feature hierarchies for accurate object detection and semantic segmentation," in IEEE Conference on Computer Vision and Pattern Recognition, pp. 580-587, 2014.
    19. "Tensorflow," [Online]. Available: https://www.tensorflow.org. [Accessed 21 - June - 2018].
    20. "Wikipedia - Keras," [Online]. Available: https://en.wikipedia.org/wiki/Keras. [Accessed 21 - June - 2018].
    21. Z. Guo and R. W. Hall, "Parallel thinning with two-subiteration algorithms," Communications of the ACM, vol. 32, no. 3, pp. 359-373, March 1989.
    22. "Wiki: Depth-first_search," [Online]. Available: https://en.wikipedia.org/wiki/Depth-first_search. [Accessed 24 - June - 2018].
    23. "Labelme: Image Polygonal Annotation with Python," [Online]. Available: https://github.com/wkentaro/labelme. [Accessed 11 - June - 2018].
    24. "Wikipedia - F1 score," [Online]. Available: https://en.wikipedia.org/wiki/F1_score. [Accessed 16 - June - 2018].
    25. "Otsu's method," [Online]. Available: https://en.wikipedia.org/wiki/Otsu%27s_method. [Accessed 22 - June - 2018].
    26. "OpenCV," [Online]. Available: https://opencv.org. [Accessed 22 - June - 2018].

    QR CODE
    :::