請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/80104完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 陳君厚(Chun-houh Chen) | |
| dc.contributor.author | Ming-Hsiu Lu | en |
| dc.contributor.author | 呂明修 | zh_TW |
| dc.date.accessioned | 2022-11-23T09:26:18Z | - |
| dc.date.available | 2022-02-16 | |
| dc.date.available | 2022-11-23T09:26:18Z | - |
| dc.date.copyright | 2022-02-16 | |
| dc.date.issued | 2022 | |
| dc.date.submitted | 2022-02-09 | |
| dc.identifier.citation | [1] 台灣衛生福利部統計處. 歷年死因統計結果2013-2020. https://dep.mohw.gov.tw/DOS/lp-5069-113.html. Accessed: 2021-12-25. [2] NVIDIA NGC. HeAortaNet. https://catalog.ngc.nvidia.com/orgs/nvidia/teams/med/models/heaortanet. Accessed: 2021-12-25. [3] Priscilla Ornellas Neves, Joalbo Andrade, and Henry Monção. Coronary artery calcium score: current status. Radiologia brasileira, 50:182–189, 2017. [4] Gianmarco Santini, Daniele Della Latta, Nicola Martini, Gabriele Valvano, Andrea Gori, Andrea Ripoli, Carla L. Susini, Luigi Landini, and Dante Chiappino. An automatic deep learning approach for coronary artery calcium segmentation. In Hannu Eskola, Outi Väisänen, Jari Viik, and Jari Hyttinen, editors, EMBEC NBC 2017, pages 374–377, Singapore, 2018. Springer Singapore. [5] W. Wang, H. Wang, Q. Chen, Z. Zhou, R. Wang, H. Wang, N. Zhang, Y. Chen, Z. Sun, and L. Xu. Coronary artery calcium score quantification using a deep learning algorithm. Clinical Radiology, 75(3):237.e11–237.e16, 2020. [6] Ran Shadmi, Victoria Mazo, Orna BregmanAmitai, and Eldad Elnekave. Fully convolutional deeplearning based system for coronary calcium score prediction from noncontrast chest ct. In 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018), pages 24–28, 2018. [7] Slicer official website. https://www.slicer.org/. Accessed: 2021-12-25. [8] Andriy Fedorov, Reinhard Beichel, Jayashree Kalpathy-Cramer, Julien Finet, Jean-Christophe Fillion-Robin, Sonia Pujol, Christian Bauer, Dominique Jennings, Fiona Fennessy, Milan Sonka, et al. 3d slicer as an image computing platform for the quantitative imaging network. Magnetic resonance imaging, 30(9):1323–1341, 2012. [9] Monai official website. https://monai.io/. Accessed: 2021-12-26. [10] Nvidia clara official website. https://developer.nvidia.com/clara. Accessed: 2021-12-26. [11] Flask official website. https://flask.palletsprojects.com/en/2.0.x/. Accessed: 2021-12-26. [12] Docker official website. https://www.docker.com/. Accessed: 2021-12-26. [13] Bodypart 醫學影像智慧檢索系統. https://www.medalab.ai/research/bodypart. Accessed: 2021-12-26. [14] Matthias Peter Hilty, Philippe Guerci, Yasin Ince, Fevzi Toraman, and Can Ince. Microtools enables automated quantification of capillary density and red blood cell velocity in handheld vital microscopy. Communications biology, 2(1):1–15, 2019. [15] Microvision medical official website. https://www.microvisionmedical.com. Accessed: 2021-1-4. [16] Sam Hutchings, Sarah Watts, and Emrys Kirkman. The cytocam video microscope. A new method for visualising the microcirculation using incident dark field technology. Clinical hemorheology and microcirculation, 62(3):261–271, 2016. [17] Sumeyra U Demir, Roya Hakimzadeh, Rosalyn Hobson Hargraves, Kevin R Ward, Eric V Myer, and Kayvan Najarian. An automated method for analysis of microcirculation videos for accurate assessment of tissue perfusion. BMC medical imaging, 12(1):1–13, 2012. [18] Chao Liu, Hernando Gomez, Srinivasa Narasimhan, Artur Dubrawski, Michael R Pinsky, and Brian Zuckerbraun. Real-time visual analysis of microvascular blood flow for critical care. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 2217–2225, 2015. [19] Ossama Mahmoud, Mahmoud El-Sakka, and Barry GH Janssen. Two-step machine learning method for the rapid analysis of microvascular flow in intravital video microscopy. Scientific Reports, 11(1):1–12, 2021. [20] Perikumar Javia, Aman Rana, Nathan Shapiro, and Pratik Shah. Machine learning algorithms for classification of microcirculation images from septic and non-septic patients. In 2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA), pages 607–611. IEEE, 2018. [21] Yuli Sun Hariyani, Heesang Eom, and Cheolsoo Park. Da-capnet: Dual attention deep learning based on u-net for nailfold capillary segmentation. IEEE Access, 8:10543–10553, 2020. [22] Drive: Digital retinal images for vessel extraction. https://drive.grand-challenge.org. Accessed: 2021-1-4. [23] Muthu Rama Krishnan Mookiah, Stephen Hogg, Tom J MacGillivray, Vijayaraghavan Prathiba, Rajendra Pradeepa, Viswanathan Mohan, Ranjit Mohan Anjana, Alexander S Doney, Colin NA Palmer, and Emanuele Trucco. A review of machine learning methods for retinal blood vessel segmentation and artery/vein classification. Medical Image Analysis, 68:101905, 2021. [24] Brian B Avants, Nick Tustison, Gang Song, et al. Advanced normalization tools (ants). Insight j, 2(365):1–35, 2009. [25] Yoshinobu Sato, Shin Nakajima, Nobuyuki Shiraga, Hideki Atsumi, Shigeyuki Yoshida, Thomas Koller, Guido Gerig, and Ron Kikinis. Three-dimensional multi-scale line filter for segmentation and visualization of curvilinear structures in medical images. Medical Image Analysis, 2(2):143–168, 1998. [26] Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention, pages 234–241. Springer, 2015. [27] 葉育彰. 右美托咪定對腎捐贈者的微循環與剩餘腎功能及腎受贈者移植腎功能的影響：機器學習輔助分析微循環型態和參數. IRB/REC 案號: 202003094RINA. | |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/80104 | - |
| dc.description.abstract | 隨著深度學習在影像處理領域的發展，有越來越多研究者開始以深度學習技術應用於醫學影像分析，在此領域中影像分割是一個常見的議題，如從圖像中找到精確的器官、腫瘤或血管等等，這些分割結果可能會直接應用於最後的結果 (eg. 評估大小)，或是作為後續分類、計算分數的前置資料。 在影像分析演算法開發與部署的過程中，會隨著案例不同而有各自的問題需要處理，在演算法開發上，我們以微循環影片分析做為案例，因為微循環影像的複雜度導致血管標註工作需要耗費大量人力，我們嘗試使用傳統電腦視覺方法生成的標註輔以深度學習模型強大的泛化能力來完成血管分割的任務；而在演算法部署上，我們以心血管鈣化分數做為案例，因為演算法的處理流程中會有耗時的後處理，導致使用 PyTorch For-Loop 推論架構會有大量時間的資源閒置，我們嘗試設計一個事件驅動的架構來處理。 在最後成果上，在微循環影片分析上，我們發現以 SATO 血管分割演算法生成的標註結合醫學影像常使用的 UNet 可以捕捉到比原先生成的標註更多的血管，展示了以電腦視覺方法生成的標註可以訓練出更優秀的深度學習模型的潛力；而在心血管鈣化分數計算上，事件驅動的架構可以顯著提升整體推論速度，同時也成功將基於 HeAortaNet 的心血管鈣化分數演算法應用於健保醫學影像資料庫。 | zh_TW |
| dc.description.provenance | Made available in DSpace on 2022-11-23T09:26:18Z (GMT). No. of bitstreams: 1 U0001-2401202213324600.pdf: 8529271 bytes, checksum: d0f319610fab77d359d755fb03625173 (MD5) Previous issue date: 2022 | en |
| dc.description.tableofcontents | 中文摘要 i Abstract ii 目錄 iv 圖目錄 vii 表目錄 ix 第一章論文架構 1 第二章心血管鈣化分數演算法序論 2 2.1 研究背景與動機 2 2.2 研究目標 4 第三章研究主題與相關文獻 6 3.1 DICOM 資料格式 7 3.2 鈣化分數與 HeAortaNet 7 3.2.1 鈣化分數 7 3.2.2 HeAortaNet 8 3.3 醫學影像 AI 開發框架 10 3.4 Python 平行處理與GIL 11 3.5 事件驅動程式設計 12 3.6 Web API 與 Flask 14 3.7 Container 虛擬化技術 15 3.8 健保署醫學影像資料庫 16 第四章鈣化分數計算處理流程 17 4.1 Web Server 與API 19 4.2 讀取 DICOM TAG 20 4.3 前過濾與資料特性 21 4.4 前處理-模型推論-後處理 22 4.5 後過濾與資料特性 23 第五章系統分析與設計 25 5.1 系統需求分析 25 5.1.1 PyTorch 中後處理平行化的問題 26 5.2 事件驅動架構 29 5.2.1 事件驅動 Service 30 5.2.2 任務階段的設計 33 5.2.3 Service 間的串接 36 5.3 框架無關的 GPU 平行 Inference 設計 39 5.3.1 Model 外包裝 40 5.3.2 GPU 程序間的隔離與溝通 42 5.3.3 GPU 程序的管理 45 5.4 API 編寫 48 第六章實驗結果 52 第七章結論與未來方向 55 第八章微循環影片分析演算法序論 56 8.1 研究背景與動機 56 8.2 研究目標 57 第九章相關文獻與研究問題 58 9.1 舌下微循環影像 58 9.2 微循環影像分析 59 9.3 影像對位 59 9.4 SATO 血管分割演算法 60 9.5 UNet 60 9.6 研究問題 61 第十章研究方法 62 10.1 影像校正與挑選 62 10.2 SATO 血管分割與訓練資料生成 64 10.3 模型訓練 65 第十一章實驗結果 66 11.1 參數設定 66 11.2 驗證集結果 68 第十二章結論與未來方向 69 參考文獻 70 | |
| dc.language.iso | zh-TW | |
| dc.subject | 深度學習 | zh_TW |
| dc.subject | 鈣化分數 | zh_TW |
| dc.subject | 微循環 | zh_TW |
| dc.subject | 血管分割 | zh_TW |
| dc.subject | 醫學影像 | zh_TW |
| dc.subject | 事件驅動 | zh_TW |
| dc.subject | Deep Learning | en |
| dc.subject | Vessel Segmentation | en |
| dc.subject | Medical Image | en |
| dc.subject | Calcium Score | en |
| dc.subject | Event-Driven | en |
| dc.subject | Microcirculation | en |
| dc.title | 心血管鈣化分數演算法佈署與微循環影片分析演算法開發 | zh_TW |
| dc.title | Cardiovascular Calcium Score Algorithm Deployment and Microcirculation Video Analysis Algorithm Development | en |
| dc.date.schoolyear | 110-1 | |
| dc.description.degree | 碩士 | |
| dc.contributor.coadvisor | 王偉仲(Wei-Chung Wang) | |
| dc.contributor.oralexamcommittee | 黃裕城(Chun-Chieh Wu),葉育彰(Pay-Liam Lin),(Min-Hui Lo),(Chuan-Yao Lin) | |
| dc.subject.keyword | 鈣化分數,事件驅動,微循環,血管分割,醫學影像,深度學習, | zh_TW |
| dc.subject.keyword | Calcium Score,Event-Driven,Microcirculation,Vessel Segmentation,Medical Image,Deep Learning, | en |
| dc.relation.page | 73 | |
| dc.identifier.doi | 10.6342/NTU202200171 | |
| dc.rights.note | 同意授權(全球公開) | |
| dc.date.accepted | 2022-02-11 | |
| 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 | |
|---|---|---|---|
| U0001-2401202213324600.pdf | 8.33 MB | Adobe PDF | 檢視/開啟 |
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。