基於醫學影像目標強化與深度學習演算法在心血管斑塊偵測的應用

Jun-Ting Chen; 陳鈞廷

請用此 Handle URI 來引用此文件： http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/80624

完整後設資料紀錄

DC 欄位	值	語言
dc.contributor.advisor	陳定立(Ting-Li Chen)
dc.contributor.author	Jun-Ting Chen	en
dc.contributor.author	陳鈞廷	zh_TW
dc.date.accessioned	2022-11-24T03:11:04Z	-
dc.date.available	2021-11-03
dc.date.available	2022-11-24T03:11:04Z	-
dc.date.copyright	2021-11-03
dc.date.issued	2021
dc.date.submitted	2021-10-25
dc.identifier.citation	[1] G. A. Roth, G. A. Mensah, C. O. Johnson, G. Addolorato, E. Ammirati, L. M. Bad dour et al., “Global Burden of Cardiovascular Diseases and Risk Factors, 1990-2019: Update From the GBD 2019 Study,” J Am Coll Cardiol, vol. 76, no. 25, pp. 2982–3021, 12 2020, [PubMed Central:PMC7755038] [DOI:10.1016/j.jacc.2020.11.010] [PubMed:33069326]. [2] F. Pugliese, M. G. Hunink, K. Gruszczynska, F. Alberghina, R. Malagó, N. van Pelt et al., “Learning curve for coronary CT angiography: what constitutes sufficient training?” Radiology, vol. 251, no. 2, pp. 359–368, May 2009, [DOI:10.1148/ra diol.2512080384] [PubMed:19401570]. [3] G. L. Raff, A. Abidov, S. Achenbach, D. S. Berman, L. M. Boxt, M. J. Budoff, V. Cheng, T. DeFrance, J. C. Hellinger, and R. P. Karlsberg, “SCCT guidelines for the interpretation and reporting of coronary computed tomographic angiography,” J Cardiovasc Comput Tomogr, vol. 3, no. 2, pp. 122–136, 2009. [4] S. C. Saur, H. Alkadhi, L. Desbiolles, G. Székely, and P. C. Cattin, “Automatic detection of calcified coronary plaques in computed tomography data sets,” in International Conference on Medical Image Computing and ComputerAssisted Intervention. Springer, 2008, pp. 170–177. [5] S. Lankton, A. Stillman, P. Raggi, and A. R. Tannenbaum, “Soft plaque detection and automatic vessel segmentation.” Georgia Institute of Technology, 2009. [6] M. A. Zuluaga, D. Hush, E. J. D. Leyton, M. H. Hoyos, and M. Orkisz, “Learning from only positive and unlabeled data to detect lesions in vascular ct images,” in International conference on medical image computing and computer-assisted intervention. Springer, 2011, pp. 9–16. [7] F. Zhao, B. Wu, F. Chen, X. Cao, H. Yi, Y. Hou et al., “An automatic multi-class coronary atherosclerosis plaque detection and classification framework,” Medical biological engineering computing, vol. 57, no. 1, pp. 245–257, 2019. [8] M. Zreik, R. W. Van Hamersvelt, J. M. Wolterink, T. Leiner, M. A. Viergever, and I. Išgum, “A recurrent cnn for automatic detection and classification of coronary artery plaque and stenosis in coronary ct angiography,” IEEE transactions on medical imaging, vol. 38, no. 7, pp. 1588–1598, 2018. [9] S. Candemir, R. D. White, M. Demirer, V. Gupta, M. T. Bigelow, L. M. Prevedello et al., “Automated coronary artery atherosclerosis detection and weakly supervised localization on coronary ct angiography with a deep 3-dimensional convolutional neural network,” Computerized Medical Imaging and Graphics, vol. 83, p. 101721, 2020. [10] A. Tejero-de Pablos, K. Huang, H. Yamane, Y. Kurose, Y. Mukuta, J. Iho et al., “Texture-based classification of significant stenosis in ccta multi-view images of coronary arteries,” in International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, 2019, pp. 732–740. [11] K.He,X.Zhang,S.Ren,andJ.Sun,“Deepresiduallearningforimagerecognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770–778. [12] D. Tran, H. Wang, L. Torresani, J. Ray, Y. LeCun, and M. Paluri, “A closer look at spatiotemporal convolutions for action recognition,” in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 6450–6459. [13] A. F. Agarap, “Deep learning using rectified linear units (relu),” arXiv preprint arXiv:1803.08375, 2018. [14] B. Xu, N. Wang, T. Chen, and M. Li, “Empirical evaluation of rectified activations in convolutional network,” arXiv preprint arXiv:1505.00853, 2015. [15] S. Woo, J. Park, J.-Y. Lee, and I. S. Kweon, “Cbam: Convolutional block attention module,” in Proceedings of the European conference on computer vision (ECCV), 2018, pp. 3–19. [16] P. Branco, L. Torgo, and R. P. Ribeiro, “A survey of predictive modeling on imbal anced domains,” ACM Computing Surveys (CSUR), vol. 49, no. 2, pp. 1–50, 2016. [17] J. M. Johnson and T. M. Khoshgoftaar, “The effects of data sampling with deep learning and highly imbalanced big data,” Information Systems Frontiers, vol. 22, no. 5, pp. 1113–1131, 2020. [18] I. Loshchilov and F. Hutter, “Decoupled weight decay regularization,” in International Conference on Learning Representations, 2018. [19] L. N. Smith, “Cyclical learning rates for training neural networks,” in 2017 IEEE winter conference on applications of computer vision (WACV). IEEE, 2017, pp. 464–472. [20] P. Goyal, P. Dollár, R. Girshick, P. Noordhuis, L. Wesolowski, A. Kyrola et al.,“Accurate, large minibatch sgd: Training imagenet in 1 hour,” 2018.
dc.identifier.uri	http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/80624	-
dc.description.abstract	偵測冠狀動脈斑塊和狹窄的存在是評估心血管疾病(Cardiovascular diseases)風險的常見方法。冠狀動脈電腦斷層血管攝影(CCTA)是心血管疾病最常見的評估之一。在每位患者的冠狀動脈樹上手動檢測斑塊與狹窄通常需要花費專家大約 10 分鐘。過去關於全自動或半自動檢測斑塊的演算法需要嚴謹控制區域分割(region based segmentation)算法的參數或機器學習(Machine Learning)算法中特徵提取的設計。最近的相關研究展示了深度學習算法用於斑塊和狹窄檢測的可行性。在這篇研究中目標為使用 3D 卷積神經網絡 (CNN) 開發檢測斑塊和分類的快速演算法流程。基於完全的卷積層架構，我們的模型在訓練階段比遞迴神經層(Recurrent Layer)具有更好的穩定性。此外，因為模型僅需要使用橫截面 (CSP) 影像作為輸入，而不包括包含重複信息的其他視圖，因此這個流程是非常有效率的。對於每位患者，執行整個演算法流程所需的時間平均不到 25 秒。演算法流程流程的第一步是在給定的冠狀動脈中心線的每個控制點上擷取橫截面影像。接著將幾個連續數個中心線控制點擷取的橫截面影像堆疊形成一個 3D 影像，並作為 CNN 模型的輸入。CNN 模型將會預測每個中心線控制點的斑塊類型(非鈣化、混合型或鈣化)。有了這些預測資訊，我們不僅可以檢測斑塊的精確位置，或是總結出任何冠狀動脈段中是否有存在斑塊。此外，我們考慮將血管內徑估計作為斑塊狹窄程度偵測的初步流程。	zh_TW
dc.description.provenance	Made available in DSpace on 2022-11-24T03:11:04Z (GMT). No. of bitstreams: 1 U0001-2110202120303100.pdf: 6018241 bytes, checksum: 318af8092ba4ce134c272feed3b3e9d3 (MD5) Previous issue date: 2021	en
dc.description.tableofcontents	Acknowledgements ii 摘要 iii Abstract v Contents vii List of Figures x List of Tables xii Denotation xiii Chapter 1 Introduction 1 1.1 Medical Background.......................... 1 1.2 Problem Statement and Significance.................. 1 1.3 Literature Review............................ 3 1.4 Challenges ............................... 4 1.5 Proposed Method............................ 4 1.6 Results ................................. 5 Chapter 2 Data 7 2.1 Data Collection............................. 7 2.2 Data Properties............................. 8 2.3 Reference Standard........................... 9 Chapter 3 Methodology 10 3.1 Cross Sectional Plane Extraction.................... 10 3.2 Piecewise Linear Image Normalization ................ 12 3.3 Model Architecture........................... 13 3.3.1 3D ResNet............................... 13 3.3.2 R(2+1)D Architecture......................... 14 3.4 Training Strategy............................ 19 3.4.1 Weighted Loss............................. 19 3.4.2 Epoch Data Sampling......................... 19 3.4.3 Optimizer and Learning Rate Decay Schedule . . . . . . . . . . . . 20 Chapter 4 Experiments and Evaluation 22 4.1 Resolution of CSP slices........................ 23 4.2 Length of Input Sequence ....................... 24 4.3 Model Architecture........................... 24 4.3.1 Model Scale.............................. 25 4.3.2 Leaky ReLU and CBAM ....................... 26 4.4 Result.................................. 26 Chapter 5 Summary 29 5.1 Discussion ............................... 29 5.2 Conclusion and Future Work ............................... 29 References 31
dc.language.iso	en
dc.subject	心血管疾病	zh_TW
dc.subject	深度學習	zh_TW
dc.subject	卷積神經網絡	zh_TW
dc.subject	冠狀動脈硬化斑塊	zh_TW
dc.subject	CoronaryPlaque	en
dc.subject	CNN	en
dc.subject	Cardiovascular Disease	en
dc.subject	Deep Learning	en
dc.title	基於醫學影像目標強化與深度學習演算法在心血管斑塊偵測的應用	zh_TW
dc.title	Medical Image Object Enhancements and Deep Learning for Cardiovascular Plaque Detection	en
dc.date.schoolyear	109-2
dc.description.degree	碩士
dc.contributor.coadvisor	陳逸昆(I-Kun Chen)
dc.contributor.oralexamcommittee	黃裕城(Hsin-Tsai Liu),王偉仲(Chih-Yang Tseng)
dc.subject.keyword	深度學習,卷積神經網絡,心血管疾病,冠狀動脈硬化斑塊,	zh_TW
dc.subject.keyword	Deep Learning,CNN,Cardiovascular Disease,CoronaryPlaque,	en
dc.relation.page	34
dc.identifier.doi	10.6342/NTU202103993
dc.rights.note	同意授權(限校園內公開)
dc.date.accepted	2021-10-26
dc.contributor.author-college	電機資訊學院	zh_TW
dc.contributor.author-dept	資料科學學位學程	zh_TW
顯示於系所單位：	資料科學學位學程

文件中的檔案：

檔案	大小	格式
U0001-2110202120303100.pdf 授權僅限NTU校內IP使用（校園外請利用VPN校外連線服務）	5.88 MB	Adobe PDF

顯示文件簡單紀錄

系統中的文件，除了特別指名其著作權條款之外，均受到著作權保護，並且保留所有的權利。