請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/73728
完整後設資料紀錄
DC 欄位 | 值 | 語言 |
---|---|---|
dc.contributor.advisor | 李財坤(Tsai-Kun Li) | |
dc.contributor.author | Jheng-Siang Ye | en |
dc.contributor.author | 葉政祥 | zh_TW |
dc.date.accessioned | 2021-06-17T08:08:53Z | - |
dc.date.available | 2019-08-28 | |
dc.date.copyright | 2019-08-28 | |
dc.date.issued | 2019 | |
dc.date.submitted | 2019-08-16 | |
dc.identifier.citation | [1]https://www.kaggle.com/paultimothymooney/detect-retina-damage-from-oct-images,Detect Retina Damage From OCT Images, Paul mooney, 2018.
[2]http://retinatw.org/a03.php, 老年性黃斑部病變, 中華民國視網膜病變協會, 2016. [3]http://www.sinphar.com.tw/magazine?cls=12&cls2=2&id=87, 淺談老年性黃斑部病變, 杏國新藥-AMD藥品部 許競爵經理. [4]http://www.icare99.com.tw/expert_view.html?nid=38839#.XOQK4cgzaUk, 3C不離手黃斑部病變報到, 陳世真醫師, 2014 Feb. [5]https://www.benq.com/zh-tw/knowledge-center/eyecare/tv-eyecare-education-age-related-macular-degeneration.html, 造成年齡相關性黃斑部變性的病因及疾病分類, 國家網路醫藥編輯部, 2018 Apr. [6]http://blog.udn.com/dogsongs/25051041, 視神經玻璃膜疣{Optic Nerve Drusen}, watwih, 2015 Jun. [7] https://www.ovs.cuhk.edu.hk/眼病小字典/老年黃斑病變, 老年黃斑病變, 香港中文大學眼科中心. [8]https://www.chimei.org.tw/main/cmh_department/59012/magazin/vol115/vol_115_28.html, 「黃斑水腫得恢復 延誤治療效果無」, 奇美醫院眼科部 林主國主治醫師, 2016 Dec. [9]https://www.careonline.com.tw/2018/04/diabetes-eye.html, 「血糖太高很傷眼,保護眼睛這樣做!」, 白映俞醫師, 2018 Apr. [10]https://sub.chgh.org.tw/news_in.aspx?siteid=50&ver=&usid=&mnuid=1643&modid=246&mode=&nid=3594&noframe=, 什麼是糖尿病引起之黃斑部水腫 (DME)?, 振興醫院眼科部, 2018 Feb. [11]http://www.staywelleye.center/2018/03/brvo.html, 視網膜分枝靜脈阻塞BRVO治療案例分享, 和欣視光眼科診所 林醫師, 2018 Mar. [12] https://blog.csdn.net/qq_40994943/article/details/85126310, 經典cnn模型彙總(alexnet,vggnet,inception), 路上的病人, 2018 Dec. [13]https://blog.csdn.net/guyuealian/article/details/80150510, CNN經典網絡模型, pan_jinquan, 2018 Apr. [14]https://blog.csdn.net/huangfei711/article/details/80484488, VGGNet總結, HF飛哥, 2018 May. [15]http://ramok.tech/tag/vgg-16/, Java Image Cat&Dog Recognition with Deep Neural Networks, KLEVIS RAMO, 2018 Jan. [16]https://my.oschina.net/u/876354/blog/1637819, 大話CNN經典模型:GoogLeNet(從Inception v1到v4的演進), 雪餅, 2018 Mar. [17]https://medium.com/@sh.tsang/review-inception-v3-1st-runner-up-image-classification-in-ilsvrc-2015-17915421f77c,Review:Inception-v3 — 1st Runner Up (Image Classi -fication) in ILSVRC 2015, Sik-Ho Tsang, 2018 Sep. [18]https://blog.csdn.net/yimingsilence/article/details/79208092, 注意力機制, 默一鸣, 2018 Jan. [19]https://blog.csdn.net/Evan123mg/article/details/80058077, SENet(Squeeze-and-Excitation Networks)網路詳解, 2k-Gamer, 2018 Apr. [20]https://www.twblogs.net/a/5c4d5c18bd9eee6e7d8229b5, CBAM論文解讀和算法原理解析筆記, 碼元, 2018 Jan. [21]https://www.cell.com/cell/fulltext/S0092-8674(18)30154-5, Identifying Medical Diagnoses and Treatable Diseases by Image-Based Deep Learning, Daniel S. Kermany, 2018 Feb. [22]https://www.researchgate.net/publication/283433254_Driving_posture_recognition_by_convolutional_neural_networks, Driving posture recognition by convolutional neural networks, Chao Yan, 2015 Sep. [23]https://www.cnblogs.com/pinard/p/6483207.html, 卷積神經網路(CNN)模型結構, 劉建平, 2017 Mar. [24]https://blog.csdn.net/v_JULY_v/article/details/51812459, 通俗理解卷積神經網路, v_JULY_v, 2016 Jul. [25]https://medium.com/@chih.sheng.huang821/機器學習-統計方法-模型評估-驗證指標-b03825ff0814, 機器學習統計方法: 模型評估-驗證指標(validation index), Tommy Huang, 2018 Jul. [26]https://zh.wikipedia.org/wiki/視網膜色素上皮,視網膜色素上皮, 維基百科, 2017 Jul. [27]https://blog.csdn.net/Quincuntial/article/details/78605507, Squeeze-and-Excitation Networks論文翻譯——中英文對照, Tyan, 2017 Nov. [28]https://www.kaggle.com/kmader/inceptionv3-for-retinopathy-gpu-hr, InceptionV3 for Retinopathy (GPU-HR), Kevin Mader, 2018 May. [29]https://buzzorange.com/techorange/2019/04/03/deepmind-ophthalmology/, DeepMind推出首款商業化AI產品:眼科AI準度94%超越人類醫師!, 郭家宏, 2019 Apr. [30]Richard H.R. Hahnloser, Rahul Sarpeshkar, Misha A. Mahowald, Rodney J. Douglas, and H. Sebastian Seung. Digital selection and analogue amplification coexist in a cortex-inspired silicon circuit. Nature, Vol. 405, No. 6789, p. 947, 2000. [31]Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton. ImageNet classification with deep convolutional neural networks. In Advances in Neural Information Processing Systems. pp. 1097-1105, 2012. [32]Dan C. Cireşan, Ueli Meier, Jonathan Masci, Luca Maria Gambardella, and Jürgen Schmidhuber. Flexible, High Performance Convolutional Neural Networks for Image Classication. In IJCAI Proceedings-International Joint Conference on Artificial Intelligence, Vol. 22, p. 1237, Barcelona, Spain, 2011. [33]Dan C. Cireşan, Ueli Meier, and Jürgen Schmidhuber. Multi-column Deep Neural Networks for Image Classification. In Computer Vision and Pattern Recognition(CVPR), 2012 IEEE conference on, pp. 3642-3649, IEEE, 2012. [34]David H. Hubel and Torsten N. wiesel. Receptive fields, Binocular Interaction And Functional Architecture In The Cat's Visual Cortex. The Journal of Physiology, Vol. 160, No. 1, pp. 106-154, 1962. [35]Strategic Management by M. J. Ryall, Philip Sadler - Business & Economics - 2003 [36]Anne Westhues, Jean Lafrance and Glen Schmidt. “A SWOT Analysis of Social Work Education in Canada”. Social Work Education: The International Journal, 20(1): pp. 35-56, 2001. [37]www.mindtools.com, 'SWOT Analysis: Discover New Opportunities, Manage and Eliminate Threats', 2018 Feb. [38]Ansoff, H. Igor. “Corporate Strategy”. McGraw Hill, New York, 1965. [39]Hull, David L. “Deconstructing Darwin: Evolutionary Theory in Context” in Journal of the History of Biology, Vol. 38, No. 1, pp. 137-152, 2005. [40]Porter, Michael E. Competitive Strategy. Free Press, ISBN 0-684-84148-7, 1980. [41]https://kknews.cc/health/89px48l.html, 近視人口增加已成全球公共衛生難題, 手機人民網, 2016 Mar. [42] https://tw.news.yahoo.com/糖尿病治療拉警報-病患已破4億人-2030年恐有4000萬患者難以取得胰島素-024001191.html, 糖尿病治療拉警報》病患已破4億人,未來將有4000萬患者難以取得胰島素!, 蔡亦寧, 2018 Nov. [43]https://www.inside.com.tw/article/17163-alphabet-s-deepmind-takes-on-billion-dollar-debt-as-loss-spirals, DeepMind這世界最知名AI公司,去年虧了5 億美元, iFanr 責任編輯:Chris, 2019 Aug. | |
dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/73728 | - |
dc.description.abstract | 本論文研究目的是建立AI眼疾辨識系統來輔助眼科醫生診斷,達到早期發現與早期治療,從而改善臨床結果。本論文採用來自於Kaggle之OCT影像數據集,此影像數據集由知名醫院之專業人員嚴格審核與標註,數據共有4種類別。AI辦識眼疾辨識系統之建立,首先利用轉移學習(Transfer Learning)以及注意力機制(Attention Mechanism),其中轉移學習使用VGG16模型,注意力機制分成1path二個模組、2path二個模組,共四種模組。其次使用OCT影像之數據集來訓練神經網絡。最後對四種模組個別進行參數調整,並找出各模型最佳參數與比較四個模型的優略。最後,篩選結果是1pathb(Conv:1path_1x3 , 3x1)與2patha(Conv:2path_1x1,1x1_1x3,3x1_add)模型之Dropout_Rate設在20%,Dense_cells數目設定512、1024為佳。實驗結果Test_acc目前達92%-93%左右,稍低於優秀專業眼科醫生93.3%判斷水準[29],但OCT醫學影像判讀速度遠遠超過眼科醫生,表示AI眼疾辨識系統存在臨床應用之價值。AI眼疾辨識輔助診斷系統之商業化要成功,應具備:(1)AI眼疾辨識輔助診斷之市場具吸引力(2)極高品質的OCT醫學影像之標註模式,(3)能夠研發優越AI辨識模型之研發精兵組織,(4)以雲端服務為主力產品之商業模式,(5)其他層面配合。前五項紮實做好就可創造公司商業價值與客戶使用價值。再透過一系列商業競爭策略之分析:(1)價值網分析,(2)GE矩陣分析,(3)SWOT分析,(4)BCG矩陣分析,(5)公司演化與競爭策略矩陣分析,來檢視Ω公司之商業模式的可行性與獲取策略方針。 | zh_TW |
dc.description.abstract | The research purpose of this paper is to establish an Artificial Intelligence(AI) of eye-related diseases identification system to assist ophthalmologists in diagnosis, so as to achieve early detection and treatment, thus improving clinical results.In this paper, Optical coherence tomography (OCT) image data set from Kaggle was used. This image data set was strictly checked and marked by professionals from well-known hospitals and there are 4 types of data. The establishment of this AI eye disease identification system was first employed Transfer Learning and Attention Mechanism, in which the VGG16 model was used for transfer learning, and the Attention Mechanism was divided into two modules of 1path and two modules of 2path, totaling four modules. Moreover, the data set of OCT images was used to train the neural network. Finally, the parameters of the four modules are adjusted individually, and the best parameters of each model was found. The advantages and disadvantages of the four models were also compared. In conclusion, 1pathb (Conv:1path_1x3, 3x1) and 2patha (Conv:2path_1x1, 1x1_1x3, 3x1_add) models were selected, and the accuracy of Dropout_Rate setting 20% and the number of Dense_cells setting 512 and 1024 is better. The experimental results of Test accuracy(Test_acc) are currently about 92%-93%, which is slightly lower than the diagnostic accuracy of 94.5% of professional ophthalmologists. However, OCT medical image diagnosis is much faster than ophthalmologists, indicating that the AI of eye-related diseases identification system has clinical application value. To succeed in the business model of the AI eye-related diseases recognition assistant diagnosis system, the necessary conditions are: (1) market appeal of AI eye disease recognition and auxiliary diagnosis (2) very high quality OCT medical image mark mode, (3) to organize An excellent R&D team to develop AI models, (4) cloud services are the main products, and (5) other support. The top five can be achieved to create company value and customer value.Through a series of business competition strategy analysis: (1) network value analysis, (2) GE matrix analysis, (3) SWOT analysis, (4) BCG matrix analysis, (5) company evolution and competition strategy matrix analysis, Verify the feasibility of the business model and provide a clear strategic direction. | en |
dc.description.provenance | Made available in DSpace on 2021-06-17T08:08:53Z (GMT). No. of bitstreams: 1 ntu-108-P06e43007-1.pdf: 4345822 bytes, checksum: 431dc896a36dcd0f352518e604866d03 (MD5) Previous issue date: 2019 | en |
dc.description.tableofcontents | 目錄
口試委員會審定書 i 誌謝 ii 中文摘要 iii ABSTRACT iv 目錄 vi 圖目錄 viii 表目錄 xi 第一章 研究背景與目的 1 第一節. 研究動機 1 第二節. 研究目的 2 第三節. 數據集介紹 3 第二章 文獻探討 5 第一節. 眼科醫學 5 第二節. 神經網路模型 14 第三章 研究方法 31 第一節. 研究流程 31 第二節. 實驗設計 32 第三節. 研究限制 36 第四節. 影像處理基本原理介紹 37 第五節. 模型評估 43 第四章 研究結果與分析 47 第一節. 小樣本初級篩選 47 第二節. 小樣本進階篩選 52 第三節. 大樣本進階篩選 54 第四節. 商品化後AI眼疾辨識報告呈現方式 57 第五節. AI眼疾辨識輔助診斷之市場潛力 59 第六節. AI眼疾辨識輔助診斷系統之商業化模式 61 第七節. 商業模式與策略之分析 66 第八節. 商業計劃書 73 第五章 研究結論與建議 77 第一節. 研究結論 77 第二節. 研究後建議 78 附表 79 參考文獻 82 | |
dc.language.iso | zh-TW | |
dc.title | AI眼疾辨識導入attention機制下實作出優越模型並分析商業模式 | zh_TW |
dc.title | Attention mechanism-Assisted Artificial Intelligence for Diagnosis of Eye Diseases: Implementation of Superior Models and Analysis of Business Models | en |
dc.type | Thesis | |
dc.date.schoolyear | 107-2 | |
dc.description.degree | 碩士 | |
dc.contributor.oralexamcommittee | 何佳安(Ja-an Annie Ho),孔令傑(Ling-Chieh Kung) | |
dc.subject.keyword | 人工智慧,機器學習,眼疾辨識系統,輔助診斷,影像數據集,標註,轉移學習,注意力機制,參數調整,神經網路,卷積,池化,醫學影像,雲端服務,商業模式,價值網分析,產品市場策略分析,競爭策略分析,商業計劃書, | zh_TW |
dc.subject.keyword | Artificial Intelligence(AI),machine learning,identification,assist,diagnosis,ophthalmologists,OCT,eye disease,neural network,convolution,pooling,parameters,transfer learning,Attention Mechanism,models,accuracy,dropout rate,auxiliary,recognition,cloud services,competition strategy,network value analysis,GE matrix analysis,SWOT analysis,BCG matrix analysis,Ansoff Matrix analysis,Business Plan, | en |
dc.relation.page | 86 | |
dc.identifier.doi | 10.6342/NTU201903847 | |
dc.rights.note | 有償授權 | |
dc.date.accepted | 2019-08-17 | |
dc.contributor.author-college | 進修推廣學院 | zh_TW |
dc.contributor.author-dept | 生物科技管理碩士在職學位學程 | zh_TW |
顯示於系所單位: | 生物科技管理碩士在職學位學程 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-108-1.pdf 目前未授權公開取用 | 4.24 MB | Adobe PDF |
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。