惡性腫瘤(癌症)為台灣十大死因之首，其中肺癌又是癌症之首。肺癌是肺部惡性腫瘤顧名思義是產生於肺部的異常腫塊，異常腫塊的產生是因位於肺部細胞產生異常變性，
變性的肺部細胞不斷的異常增生，分裂形成特殊的腫塊，稱為「腫瘤」。
　　
一天需看將近500張胸部X-RAY影像，其中判斷有無肺部腫瘤，須依照其醫生各自的經驗進行判讀，且過程中需耗費大量的人工時間與精神力。

本研究的合作之醫院為位於台中的童綜合醫院。使用院方提供之胸部X-RAY 資料集與童綜合醫院院方資料Label，並利用人工智慧方法中影像辨識偵測的邏輯，選用InceptionResNetV2、InceptionV3、VGG、ResNet(18.50)、MobileNetV2，透過遷移學習修改ANN層，比較各個大中小規模的模型績效，與驗證是否績效與模型規模成正比，最後將最佳的模型，經過實驗設計找出最佳的參數組合，提出最適合胸部X-RAY影像分類有無肺部腫瘤之最佳方法，協助醫生藉以發現是否胸腔X-RAY 影像中有無肺部腫瘤。


本研究明確的檢測成果:(1)協助醫生偵測肺部腫瘤，可減少醫生辨識腫瘤的人工時間與大幅降低人為判讀失誤之風險，造福社會大眾。(2)判斷肺部腫瘤的分類模型。

Malignant tumors (cancer) are the top ten causes of death in Taiwan, and lung cancer is also the top cause of cancer. Lung cancer is a malignant tumor of the lung, as the name implies, an abnormal mass that occurs in the lungs. The abnormal mass is caused by abnormal degeneration of cells located in the lung.
The degenerated lung cells continue to proliferate abnormally and divide to form special masses called "tumors."
　　

Nearly 500 chest X-RAY images need to be seen in one day. To determine whether there are lung tumors, the interpretation must be based on the doctor's own experience, and the process requires a lot of manpower and mental energy.


The cooperative hospital in this study is from Tungs' Taichung MetroHarbor Hospital in Taichung. Use the chest X-RAY dataset provided by the hospital and Label from Tungs' Taichung MetroHarbor Hospital , and use the logic of image recognition and detection in the artificial intelligence method, this study select InceptionResNetV2, InceptionV3, VGG, ResNet(18.50), MobileNetV2, and modify the ANN layer through transfer learning , Compare the performance of various large, medium and small scale models, and verify whether the performance is proportional to the model scale, and finally find the best parameter combination through experimental design for the best model, and propose the most suitable chest X-RAY image classification with or without lung tumor. The best method can help doctors find out whether there are lung tumors in the X-RAY image of the thoracic cavity.


The clear test results of this study: (1) Assist doctors in detecting lung tumors, which can reduce the manual time for doctors to identify tumors and greatly reduce the risk of human errors in interpretation, which will benefit the public. (2) classification model for judging lung tumors .

第1章緒論1
第2章文獻探討5
第3章 研究步驟與過程14
第4章實驗結果23
第5章結論與未來展望27
參考文獻 29

[1] 衛生福利部統計處 (2021). 109 年死因統計結果分析.
[2] WHO.(2021).Cancer.Retrieved from https://www.who.int/en/news- room/fact-sheets/detail/cancer (Jul,6,2021).
[3] 劉劍英. 認識肺癌. 林口長庚肺腫瘤及內視鏡科
[4] UWMC 造影服務部.(2018).CT 掃描劑量
[5] Hubel, D. H., & Wiesel, T. N. (1962). Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. The Journal of physiology, 160(1), 106-154.
[6] LeCun, Y., Boser, B., Denker, J., Henderson, D., Howard, R., Hubbard, W., & Jackel, L. (1989). Backpropagation Applied to Handwritten Zip Code Recog- nition. Neural Computation, 1(4), 541-551. doi: 10.1162/neco.1989.1.4.541
[7] Sheetal,Rajpal., Navin,Lakhyani., Ayush, Kumar.Singh., Rishav Kohli.,Naveen,Kumar.(2021).Using handpicked features in conjunction with ResNet-50 for improved detection of COVID-19 from chest X-ray images.Chaos, Solitons and Fractals 145,Elsevier
[8] Zhiyong Liu, Chuan Yang, Jun Huang, Shaopeng Liu, Yumin Zhuo, Xu Lu, Deep learning framework based on integration of S-Mask R-CNN and Inception- v3 for ultrasound image-aided diagnosis of prostate cancer,Future Generation Computer Systems,Volume 114,2021,Pages 358-367,ISSN 0167-739X
[9] Wen-Chi Kuo, Wen-Tien Hsiao, Chii-Jen Chen, "Computer-aided analysis sys- tem for bone age in x-ray image using deep neural network," Proc. SPIE 11792, International Forum on Medical Imaging in Asia 2021, 117920D (20 April 2021); https://doi.org/10.1117/12.2590718
[10] S. A. Sanjaya and S. Adi Rakhmawan, "Face Mask Detection Using Mo- bileNetV2 in The Era of COVID-19 Pandemic," 2020 International Conference on Data Analytics for Business and Industry: Way Towards a Sustainable Econ- omy (ICDABI), 2020, pp. 1-5, doi: 10.1109/ICDABI51230.2020.9325631.
[11] N. Tajbakhsh et al., "Convolutional Neural Networks for Medical Image Anal- ysis: Full Training or Fine Tuning?," in IEEE Transactions on Medical Imaging, vol. 35, no. 5, pp. 1299-1312, May 2016, doi: 10.1109/TMI.2016.2535302.
[12] P.Jagadeesha.,V.S Jayanthib.,Radhika Bhaskerc.(2021).A frame work for the detection and Diagnosis of Lung Tumors using Deep learning Methods.TurkishJournalofComputerandMathematicsEducation.
[13] Chaunzwa, T. L., Hosny, A., Xu, Y., Shafer, A., Diao, N., Lanuti, M., Chris- tiani, D. C., Mak, R. H., & Aerts, H. (2021). Deep learning classification of lung cancer histology using CT images. Scientific reports, 11(1), 5471.
[14] 桑慧敏. (2019). 一生受用的統計學 大數據分析之鑰 (三). 桑慧敏.
[15] He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 770-778).
[16] Lin, T. Y., Dollár, P., Girshick, R., He, K., Hariharan, B., & Belongie, S. (2017). Feature pyramid networks for object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 2117-2125).
[17] Lin, T. Y., Goyal, P., Girshick, R., He, K., & Dollár, P. (2017). Focal loss for dense object detection. In Proceedings of the IEEE international conference on computer vision (pp. 2980-2988).
[18] Lin, Z. (2011, May). The cell image segmentation based on the KL transform and OTSU method. In 2011 International Conference on Multimedia and Signal Processing (Vol. 1, pp. 25-28). IEEE.
[19] Liong, S. T., Gan, Y. S., Huang, Y. C., Liu, K. H., & Yau, W. C. (2019). Integrated Neural Network and Machine Vision Approach For Leather Defect Classification. arXiv preprint arXiv:1905.11731.
[20] Masci, J., Meier, U., Fricout, G., & Schmidhuber, J. (2013, August). Multi- scale pyramidal pooling network for generic steel defect classification. In The 2013 International Joint Conference on Neural Networks (IJCNN) (pp. 1-8). IEEE.
[21] Ng, H. F. (2006). Automatic thresholding for defect detection. Pattern recog- nition letters, 27(14), 1644-1649.
[22] Ren, S., He, K., Girshick, R., & Sun, J. (2015). Faster r-cnn: Towards real- time object detection with region proposal networks. In Advances in neural information processing systems (pp. 91-99).
[23] Saeed, N., King, N., Said, Z., & Omar, M. A. (2019). Automatic defects detec- tion in CFRP thermograms, using convolutional neural networks and transfer learning. Infrared Physics & Technology, 102, 103048.
[24] Saravanan, T., Bagavathiappan, S., Philip, J., Jayakumar, T., & Raj, B. (2007). Segmentation of defects from radiography images by the histogram con- cavity threshold method. Insight-Non-Destructive Testing and Condition Mon- itoring, 49(10), 578-584.
[25] Siche, D., Klimm, D., Hölzel, T., & Wohlfart, A. (2004). Reproducible defect etching of SiC single crystals. Journal of crystal growth, 270(1-2), 1-6.
[26] Simonyan, K., & Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556.
[27] Sun, W., Song, Y., Liu, C., Peng, T., Wang, W., & Chen, X. (2015). Basal plane dislocation-threading edge dislocation complex dislocations in 6H-SiC sin- gle crystals. Materials Express, 5(1), 63-67.
[28] SyvA ̈jvA ̈rvi, M., Yakimova, R., & JanzE ́n, E. (2000). Anisotropic etching of SiC. Journal of the Electrochemical Society, 147(9), 3519-3522.
[29] Tastimur, C., Yetis, H., Karaköse, M., & Akin, E. (2016). Rail defect detec- tion and classification with real time image processing technique. International Journal of Computer Science and Software Engineering, 5(12), 283.
[30] Tsanakas, J. A., Chrysostomou, D., Botsaris, P. N., & Gasteratos, A. (2015). Fault diagnosis of photovoltaic modules through image processing and Canny edge detection on field thermographic measurements. International Journal of Sustainable Energy, 34(6), 351-372.
[31] Van de Sande, K. E., Uijlings, J. R., Gevers, T., & Smeulders, A. W. (2011, November). Segmentation as selective search for object recognition. In 2011 International Conference on Computer Vision (pp. 1879-1886). IEEE.
[32] Hardik Deshmukh(2020).Medical X-ray Image Classifica- tion using Convolutional Neural Network.Retrieved from https : //towardsdatascience.com/medical − x − ray − %EF%B8%8F − image − classification − using − convolutional − neural − network − 9a6d33b1c2a (Jul,13,2021)。
[33] Abiyev, R. H., & Ma’aitah, M. (2018). Deep Convolutional Neural Networks for Chest Diseases Detection. Journal of healthcare engineering, 2018, 4168538. https://doi.org/10.1155/2018/4168538