基於深度學習之心臟血管鈣化偵測系統

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：16

、訪客IP：3.145.206.169

姓名

詹鈞婷(Chun-Ting Chan) 查詢紙本館藏

畢業系所

資訊工程學系

論文名稱

基於深度學習之心臟血管鈣化偵測系統
(A Coronary Artery Calcification Detection System based on Deep Learning)

相關論文

★ 以Q-學習法為基礎之群體智慧演算法及其應用	★ 發展遲緩兒童之復健系統研製
★ 從認知風格角度比較教師評量與同儕互評之差異：從英語寫作到遊戲製作	★ 基於檢驗數值的糖尿病腎病變預測模型
★ 模糊類神經網路為架構之遙測影像分類器設計	★ 複合式群聚演算法
★ 身心障礙者輔具之研製	★ 指紋分類器之研究
★ 背光影像補償及色彩減量之研究	★ 類神經網路於營利事業所得稅選案之應用
★ 一個新的線上學習系統及其於稅務選案上之應用	★ 人眼追蹤系統及其於人機介面之應用
★ 結合群體智慧與自我組織映射圖的資料視覺化研究	★ 追瞳系統之研發於身障者之人機介面應用
★ 以類免疫系統為基礎之線上學習類神經模糊系統及其應用	★ 基因演算法於語音聲紋解攪拌之應用

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

[檢視]

[下載]

本電子論文使用權限為同意立即開放。
已達開放權限電子全文僅授權使用者為學術研究之目的，進行個人非營利性質之檢索、閱讀、列印。
請遵守中華民國著作權法之相關規定，切勿任意重製、散佈、改作、轉貼、播送，以免觸法。

摘要(中)

近年來，心血管疾病一直都是台灣前三大死亡原因，在全球更是最主要的死亡原因。如果病人的心臟鈣化指數高的話，就有極高度的心血管疾病的風險。對醫生來說，要從大量的心臟血管的CT醫學影像中標註出鈣化斑塊是一項頗耗費時間與精力的工作。因此，本論文希望能夠將醫學影像處理技術和深度學習做結合，來開發出一套心臟血管鈣化偵測系統，以便減輕醫師的工作量以及減少因疲勞而升高的誤判率。
本論文所開發出心臟血管鈣化偵測系統包含四個模組：（1）心臟血管擷取模組：利用影像處理技術，從心臟血管的CT醫學影像中擷取出冠狀動脈血管的部分、（2）鈣化斑塊偵測模組：使用YOLOv3模型並搭配影像處理技術，偵測並且標記出鈣化斑塊的發生位置、（3）鈣化斑塊血管分段定位模組：使用CNN模型判斷鈣化斑塊的位置是在血管的上中下的哪個位置、（4）編輯工具模組：可以搭配上述的三個模組，在讀取心臟血管的CT醫學影像後，自動生成一張心臟鈣化評估報告圖，透過此編輯工具，可讓醫師直接增修圖上的鈣化斑塊的位置與數量，藉此減輕醫師的工作量。
實驗結果顯示，本系統的心臟血管擷取模組可成功擷取90.32%的血管影像，鈣化斑塊偵測的結果，F1 score達到0.8，平均偵測到的鈣化斑塊正確率達到90%，鈣化斑塊血管位置準確率則有81%，整體準確率大於八成。因此，本系統具備一定程度之可用性。

摘要(英)

In recent years, cardiovascular disease has been the top three causes of death in Taiwan, and it is the leading cause of death in the world. If the patient′s heart calcification index is high, there is a high risk of cardiovascular disease. For doctors, labeling calcified plaques from a large number of CT medical images of coronary arteries is a time-consuming and energetic task. Therefore, this thesis hopes to combine medical image processing technology with deep learning to develop a coronary artery calcification detection system to reduce the workload of physicians and reduce the false positive rate due to fatigue.
This thesis developed a coronary artery calcification detection system, which consists of four modules: (1) the coronary artery capture module: using image processing technology to extract the vascular part of the coronary artery from a CT medical image of the coronary artery, (2) the calcified plaque detection module: using the YOLOv3 model with image processing technology to detect and mark the location of calcified plaques, (3) the calcified plaque vascular segmentation positioning module: using a CNN model to decide which position of the calcified plaque is in the upper, middle, and lower parts of the coronary artery, (4) the editing tool module: via the use of above-mentioned three modules, it can automatically generate a coronary artery calcification evaluation report after reading the CT medical image of the heart. The editing tool allows the physician to directly modify the position and the number of the calcified plaques on the report in order to reduce the workload of the physician.
The experimental results are as follows. The coronary artery capture module can successfully segment the coronary artery from a CT medical image of the heart with 90.32% correct rate. The F1 score of the calcified plaque detection module reaches 0.8. The average classification accuracy and positioning accuracy rate of the calcified plaque classification and coronary artery positioning module is 90% and 81%, respectively. Overall accuracy of the coronary artery calcification detection system is greater than 80%. Therefore, the system has a certain degree of usability.

關鍵字(中)

★ 醫學影像
★ 深度學習
★ 心臟血管鈣化
★ 偵測系統

關鍵字(英)

★ medical images
★ deep learning
★ coronary artery calcifications
★ detection system

論文目次

摘要 i
ABSTRACT ii
致謝 iv
目錄 vi
圖目錄 viii
表目錄 xii
第一章、緒論 1
1-1 研究動機 1
1-2 研究目的 2
1-3 論文架構 3
第二章、相關研究 4
2-1 醫學影像研究 4
2-1-1 醫學影像介紹 4
2-1-2 心臟血管 6
2-1-3 鈣化斑塊種類 8
2-1-4 相關文獻探討 9
2-2 神經網路模型CNN相關研究 16
2-3 物件偵測模型YOLO相關研究 19
2-3-1 YOLO 19
2-3-2 YOLOv2 24
2-3-3 YOLOv3 26
第三章、研究方法 28
3-1 心臟血管影像擷取模組 29
3-2 鈣化斑塊偵測模組 39
3-2-1 偵測模型建立與訓練 39
3-2-2 精準偵測範圍 40
3-3 鈣化斑塊血管分段定位模組 42
3-4 心臟血管鈣化偵測系統 44
3-4-1 系統流程 45
3-4-2 系統介面操作流程與介紹 46
3-4-3 編輯功能使用說明 49
第四章、實驗設計與結果 53
4-1 評估方式 53
4-1-1 混淆矩陣 53
4-1-2 IOU 55
4-2 心臟血管影像擷取實驗 56
4-2-1 閥值設定之結果比較與分析 57
4-2-2 使用MaskRCNN偵測血管區域 62
4-3 鈣化斑塊偵測實驗 65
4-3-1 交叉驗證fold數量之比較 66
4-3-2 不同神經網路之結果比較與分析 68
4-3-3 IOU閥值實驗結果與分析 70
4-3-4 影像二值化閥值實驗結果比較與分析 73
4-4 鈣化斑塊位置定位實驗 78
4-4-1 CNN不同架構之比較 81
4-4-2 實驗結果與分析 82
第五章、結論與未來展望 85
5-1 結論 85
5-2 未來展望 86
參考文獻 87

參考文獻

[1] 衛生福利處統計部. [Online]. Available: https://dep.mohw.gov.tw/DOS/np-1776-113.html. [Accessed: 10 - Jun - 2019].
[2] The top 10 causes of death. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death. [Accessed: 13 - Jun - 2019].
[3] Developments in medical imaging – timeline. [Online]. Available: https://www.sciencelearn.org.nz/resources/1906-developments-in-medical-imaging-timeline. [Accessed: 14 - Jun - 2019].
[4] The First X-ray, 1895. [Online]. Available: https://www.the-scientist.com/foundations/the-first-x-ray-1895-42279. [Accessed: 18 - Jun - 2019].
[5] D. S. Kermany et al., "Identifying medical diagnoses and treatable diseases by image-based deep learning," Cell, vol. 172, no. 5, pp. 1122-1131. e9, 2018.
[6] Coronary circulation. [Online]. Available: https://en.wikipedia.org/wiki/Coronary_circulation. [Accessed: 20 - Jun - 2019].
[7] S. Voros et al., "Coronary atherosclerosis imaging by coronary CT angiography: current status, correlation with intravascular interrogation and meta-analysis," JACC: Cardiovascular Imaging, vol. 4, no. 5, pp. 537-548, 2011.
[8] J. Yeboah et al., "Comparison of novel risk markers for improvement in cardiovascular risk assessment in intermediate-risk individuals," Jama, vol. 308, no. 8, pp. 788-795, 2012.
[9] C. W. Pavitt et al., "Deriving coronary artery calcium scores from CT coronary angiography: a proposed algorithm for evaluating stable chest pain," The international journal of cardiovascular imaging, vol. 30, no. 6, pp. 1135-1143, 2014.
[10] I. Mylonas et al., "Quantifying coronary artery calcification from a contrast-enhanced cardiac computed tomography angiography study," European Heart Journal–Cardiovascular Imaging, vol. 15, no. 2, pp. 210-215, 2013.
[11] S. Voros and Z. Qian, "Agatston score tried and true: by contrast, can we quantify calcium on CTA?," Journal of cardiovascular computed tomography, vol. 6, no. 1, pp. 45-47, 2012.
[12] F. De Dombal et al., "Computer-aided diagnosis of acute abdominal pain," Br Med J, vol. 2, no. 5804, pp. 9-13, 1972.
[13] K. Doi, "Computer-aided diagnosis in medical imaging: historical review, current status and future potential," Computerized medical imaging and graphics, vol. 31, no. 4-5, pp. 198-211, 2007.
[14] S. Katsuragawa and K. Doi, "Computer-aided diagnosis in chest radiography," Computerized Medical Imaging and Graphics, vol. 31, no. 4-5, pp. 212-223, 2007.
[15] P. Schmid-Saugeona et al., "Towards a computer-aided diagnosis system for pigmented skin lesions," Computerized Medical Imaging and Graphics, vol. 27, no. 1, pp. 65-78, 2003.
[16] R. M. Rangayyan et al., "A review of computer-aided diagnosis of breast cancer: Toward the detection of subtle signs," Journal of the Franklin Institute, vol. 344, no. 3-4, pp. 312-348, 2007.
[17] J. M. Wolterink et al., "Automatic coronary artery calcium scoring in cardiac CT angiography using paired convolutional neural networks," Medical image analysis, vol. 34, pp. 123-136, 2016.
[18] Image Reconstruction Planes. [Online]. Available: https://www.ipfradiologyrounds.com/hrct-primer/image-reconstruction/. [Accessed: 21 - Jun - 2019].
[19] I. Išgum et al., "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, 2007.
[20] J. M. Keller et al., "A fuzzy k-nearest neighbor algorithm," IEEE transactions on systems, man, and cybernetics, no. 4, pp. 580-585, 1985.
[21] B. Scholkopf and A. J. Smola, Learning with kernels: support vector machines, regularization, optimization, and beyond. MIT press, 2001.
[22] 唐嘉梅, "深度學習於醫學影像處理之應用," 碩士, 資訊工程學系, 國立中央大學, 桃園縣, 2018.
[23] Y. LeCun et al., "Deep learning," nature, vol. 521, no. 7553, p. 436, 2015.
[24] A Beginner′s Guide To Understanding Convolutional Neural Networks. [Online]. Available: https://adeshpande3.github.io/A-Beginner%27s-Guide-To-Understanding-Convolutional-Neural-Networks/. [Accessed: 20 - Jun - 2019].
[25] Applied Deep Learning - Part 4: Convolutional Neural Networks. [Online]. Available: https://towardsdatascience.com/applied-deep-learning-part-4-convolutional-neural-networks-584bc134c1e2. [Accessed: 18 - Jun - 2019].
[26] B. Christian and T. Griffiths, "Chapter 7: Overfitting," in Algorithms to live by: The computer science of human decisions: Macmillan, pp. 149–168, 2016.
[27] Multilayer perceptron. [Online]. Available: https://en.wikipedia.org/wiki/Multilayer_perceptron. [Accessed: 18 - Jun - 2019].
[28] C. Szegedy et al., "Going deeper with convolutions," in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1-9, 2015.
[29] J. Redmon et al., "You only look once: Unified, real-time object detection," in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 779-788, 2016.
[30] R. Girshick et al., "Rich feature hierarchies for accurate object detection and semantic segmentation," in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 580-587, 2014.
[31] R. Girshick, "Fast r-cnn," in Proceedings of the IEEE international conference on computer vision, pp. 1440-1448, 2015.
[32] S. Ren et al., "Faster r-cnn: Towards real-time object detection with region proposal networks," in Advances in neural information processing systems, pp. 91-99, 2015.
[33] A. Neubeck and L. V. Gool, "Efficient Non-Maximum Suppression," in 18th International Conference on Pattern Recognition (ICPR′06), vol. 3, pp. 850-855, doi: 10.1109/ICPR.2006.479, 2006.
[34] J. Redmon and A. Farhadi, "YOLO9000: better, faster, stronger. ," CoRR abs/1612.08242, 2016.
[35] J. Redmon and A. Farhadi, "Yolov3: An incremental improvement," arXiv preprint arXiv:1804.02767, 2018.
[36] A Closer Look at YOLOv3. [Online]. Available: https://www.cyberailab.com/home/a-closer-look-at-yolov3. [Accessed: 12 - Jun - 2019].
[37] Confusion matrix. [Online]. Available: https://en.wikipedia.org/wiki/Confusion_matrix. [Accessed: 21 - Jun - 2019].
[38] S. Nowozin, "Optimal Decisions from Probabilistic Models: The Intersection-over-Union Case," in 2014 IEEE Conference on Computer Vision and Pattern Recognition, pp. 548-555, doi: 10.1109/CVPR.2014.77, 2014.
[39] Illustration of IOU calculation. [Online]. Available: https://devblogs.nvidia.com/exploring-spacenet-dataset-using-digits/. [Accessed: 22 - Jun - 2019].
[40] N. Otsu, "A threshold selection method from gray-level histograms," IEEE transactions on systems, man, and cybernetics, vol. 9, no. 1, pp. 62-66, 1979.
[41] K. He et al., "Mask r-cnn," in Proceedings of the IEEE international conference on computer vision, pp. 2961-2969, 2017.

指導教授

蘇木春(Mu-Chun Su)

審核日期

2019-8-16

推文