目的 : 診斷是否遭到幽門桿菌感染的準確性會受到醫師經驗、胃腸道複雜環境因素與檢驗方法的限制，目前為止還未有快速又高準確度的檢測方式，因此本研究提出一個基於機器學習之影像辨識技術分析胃的體部(Body)及竇部(Antrum)影像以輔助醫師做幽門桿菌感染之診斷。
材料與方法 : 本研究使用的資料集之機構審查委員會(Institutional Review Board, IRB)編號為ANHRF109-07，研究資料來源自安南醫院獲得的302位病患的Body及Antrum部位，本研究使用不使用任何前處理、自適應對比增強加上限制對比度自適應直方圖均衡方法來對深度學習結合分類模型進行幽門桿菌感染為陰性或陽性的分類。
結果 : 本研究使用深度學習結合分類模型方法，在兩種部位的實驗綜合結果顯示使用scSE-CatBoost模型的平均評估可達到準確率為0.90，靈敏度為1.00，特異度為0.81，陽性預測值為0.82，陰性預測值為1.00，曲線下面積為0.88。
結論 : 實驗結果顯示使用scSE的深度學習結合分類模型能夠提高預測的準確度，輔助醫師評估患者胃部是否遭到幽門桿菌感染，以此規劃並調整治療計畫。

Purpose: The accuracy of diagnosis of Helicobacter infection is limited by the experience of physicians, complicated gastrointestinal environment factors and testing methods. So far, there is no fast and high-accuracy detection method. Therefore, this study proposes an image based on machine learning. The identification technology analyzes the images of the body and antrum of the stomach to assist the doctor in the diagnosis of Helicobacter infection.
Materials and methods: The Institutional Review Board (IRB) number of the data set used in this study is ANHRF109-07. The research data is sourced from the Body and Antrum parts of 302 patients obtained from Annan Hospital. This study uses no pre-processing, The adaptive contrast enhancement and the adaptive histogram equalization method of restricted contrast are used to classify the negative or positive Helicobacter infection of the deep learning combined with the classification model.
Results: This study uses deep learning combined with a classification model method. The comprehensive results of the experiments in the two parts show that the average evaluation using the scSE-CatBoost model can achieve an accuracy of 0.90, a sensitivity of 1.00, a specificity of 0.81, a positive predictive value of 0.82, and a negative The predicted value is 1.00 and the area under the curve is 0.88.
Conclusions: Experimental results show that using scSE's deep learning combined with a classification model can improve the accuracy of prediction and assist physicians in assessing whether the patient’s stomach is infected by Helicobacter pylori, so as to plan and adjust treatment plans.

摘要 i
Abstract iii
致謝 v
目錄 vi
表目錄 viii
圖目錄 ix
符號縮寫 xi
第一章 緒論 1
1.1 動機 1
1.2 目的 2
1.3 相關文獻探討 3
第二章 材料與方法 5
2.1 前言 5
2.2 Body與Antrum資料集 7
2.3標記座標擷取影像 9
2.4影像前處理(Image Pretreatment) 11
2.4.1無前處理(No Preprocessing) 11
2.4.2自適應對比度增強(Adaptive Contrast Enhancement, ACE) 12
2.4.3限制對比度自適應直方圖均衡化(Contrast Limited Adaptive Histogram
Equalization, CLAHE) 14
2.4.4 ACE+CLAHE 15
2.5 深度學習模型(Deep Learning Model) 16
2.5.1卷積神經網路(Convolutional Neural Networks, CNN) 16
2.5.2卷積塊注意模型(Convolutional Block Attention Module, CBAM) 17
2.5.3 Spatial and Channel Squeeze & Excitation Block(scSE) 20
2.6 分類模型(Classification Module) 22
2.6.1最近鄰居法(K-Nearest Neighbor, KNN) 23
2.6.2支持向量機(Support Vector Machine, SVM) 24
2.6.3自適應提升(Adaptive Boosting, AdaBoost) 24
2.6.4隨機森林(Random Forest, RF) 25
2.6.5梯度提升決策樹(Gradient Boosting Decision Tree, GBDT) 26
2.6.6極限梯度提升(eXtreme Gradient Boosting, XGBoost) 27
2.6.7光梯度提升(Light Gradient Boosting, LGBoost) 27
2.6.8分類梯度提升(Categorical Boosting, CatBoost) 28
2.7 分類模型的績效評估 29
2.7.1準確率(Accuracy) 30
2.7.2陽性預測值(Positive Predictive Value, PPV) 30
2.7.3陰性預測值(Negative Predictive Value, NPV) 30
2.7.4靈敏度(Sensitivity) 31
2.7.5特異度(Specificity) 31
2.7.6 Area Under Curve (AUC) 32
第三章 結果 33
3.1 前言 33
3.2 分類結果 33
3.3 Body與Antrum部位綜合評估 52
第四章 討論 54
4.1 前言 54
4.2 平均分類結果評估探討 54
4.3 Body及Antrum樣本影像之預測綜合評估探討 55
第五章 結論 56
參考文獻 57

[1]P. Correa, "Helicobacter pylori and gastric carcinogenesis," The American journal of surgical pathology, vol. 19, pp. S37-43, 1995.
[2]S. Suerbaum and P. Michetti, "Helicobacter pylori infection," New England Journal of Medicine, vol. 347, pp. 1175-1186, 2002.
[3]M. J. Blaser and J. C. Atherton, "Helicobacter pylori persistence: biology and disease," The Journal of clinical investigation, vol. 113, pp. 321-333, 2004.
[4]J. G. Kusters, A. H. Van Vliet, and E. J. Kuipers, "Pathogenesis of Helicobacter pylori infection," Clinical microbiology reviews, vol. 19, pp. 449-490, 2006.
[5]F. E. Rihane, D. Erguibi, O. Elyamine, B. Abumsimir, M. M. Ennaji, and F. Chehab, "Helicobacter pylori co-infection with Epstein-Barr virus and the risk of developing gastric adenocarcinoma at an early age: Observational study infectious agents and cancer," Annals of Medicine and Surgery, vol. 68, p. 102651, 2021.
[6]S. Shichijo, S. Nomura, K. Aoyama, Y. Nishikawa, M. Miura, T. Shinagawa, et al., "Application of convolutional neural networks in the diagnosis of Helicobacter pylori infection based on endoscopic images," EBioMedicine, vol. 25, pp. 106-111, 2017.
[7]S. Zhou, H. Marklund, O. Blaha, M. Desai, B. Martin, D. Bingham, et al., "Deep learning assistance for the histopathologic diagnosis of Helicobacter pylori," Intelligence-Based Medicine, vol. 1, p. 100004, 2020.
[8]J. Parsonnet, G. D. Friedman, D. P. Vandersteen, Y. Chang, J. H. Vogelman, N. Orentreich, et al., "Helicobacter pylori infection and the risk of gastric carcinoma," New England Journal of Medicine, vol. 325, pp. 1127-1131, 1991.
[9]K. Togashi, "Applications of artificial intelligence to endoscopy practice: The view from Japan Digestive Disease Week 2018," ed: Wiley Online Library, 2019.
[10]G. Urban, P. Tripathi, T. Alkayali, M. Mittal, F. Jalali, W. Karnes, et al., "Deep learning localizes and identifies polyps in real time with 96% accuracy in screening colonoscopy," Gastroenterology, vol. 155, pp. 1069-1078. e8, 2018.
[11]X. Jia and M. Q.-H. Meng, "A deep convolutional neural network for bleeding detection in wireless capsule endoscopy images," in 2016 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2016, pp. 639-642.
[12]Y. Zhu, Q.-C. Wang, M.-D. Xu, Z. Zhang, J. Cheng, Y.-S. Zhong, et al., "Application of convolutional neural network in the diagnosis of the invasion depth of gastric cancer based on conventional endoscopy," Gastrointestinal endoscopy, vol. 89, pp. 806-815. e1, 2019.
[13]G. Liu, J. Hua, Z. Wu, T. Meng, M. Sun, P. Huang, et al., "Automatic classification of esophageal lesions in endoscopic images using a convolutional neural network," Annals of translational medicine, vol. 8, 2020.
[14]Z. Sobri and H. A. M. Sakim, "Texture color fusion based features extraction for endoscopic gastritis images classification," International Journal of Computer and Electrical Engineering, vol. 4, pp. 674-678, 2012.
[15]S. Jain, A. Seal, A. Ojha, A. Yazidi, J. Bures, I. Tacheci, et al., "A deep CNN model for anomaly detection and localization in wireless capsule endoscopy images," Computers in Biology and Medicine, p. 104789, 2021.
[16]C. Zhang, J. Wu, Z. Chen, W. Liu, M. Li, and S. Jiang, "Dense-CNN: Dense convolutional neural network for stereo matching using multiscale feature connection," Signal Processing: Image Communication, vol. 95, p. 116285, 2021.
[17]P. M. Narendra and R. C. Fitch, "Real-time adaptive contrast enhancement," IEEE transactions on pattern analysis and machine intelligence, pp. 655-661, 1981.
[18]S. M. Pizer, E. P. Amburn, J. D. Austin, R. Cromartie, A. Geselowitz, T. Greer, et al., "Adaptive histogram equalization and its variations," Computer vision, graphics, and image processing, vol. 39, pp. 355-368, 1987.
[19]I. Goodfellow, Y. Bengio, and A. Courville, Deep learning: MIT press, 2016.
[20]J. Wu, "Introduction to convolutional neural networks," National Key Lab for Novel Software Technology. Nanjing University. China, vol. 5, p. 495, 2017.
[21]S. Albawi, T. A. Mohammed, and S. Al-Zawi, "Understanding of a convolutional neural network," in 2017 International Conference on Engineering and Technology (ICET), 2017, pp. 1-6.
[22]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.
[23]A. G. Roy, N. Navab, and C. Wachinger, "Concurrent spatial and channel ‘squeeze & excitation’in fully convolutional networks," in International conference on medical image computing and computer-assisted intervention, 2018, pp. 421-429.
[24]R. F. Murray, "Classification images: A review," Journal of vision, vol. 11, pp. 2-2, 2011.
[25]E. Fix, Discriminatory analysis: nonparametric discrimination, consistency properties vol. 1: USAF school of Aviation Medicine, 1985.
[26]G. Guo, H. Wang, D. Bell, Y. Bi, and K. Greer, "KNN model-based approach in classification," in OTM Confederated International Conferences" On the Move to Meaningful Internet Systems", 2003, pp. 986-996.
[27]C. Cortes and V. Vapnik, "Support-vector networks," Machine learning, vol. 20, pp. 273-297, 1995.
[28]W. S. Noble, "What is a support vector machine?," Nature biotechnology, vol. 24, pp. 1565-1567, 2006.
[29]Y. Freund and R. E. Schapire, "A decision-theoretic generalization of on-line learning and an application to boosting," Journal of computer and system sciences, vol. 55, pp. 119-139, 1997.
[30]R. E. Schapire, "Explaining adaboost," in Empirical inference, ed: Springer, 2013, pp. 37-52.
[31]T. K. Ho, "Random decision forests," in Proceedings of 3rd international conference on document analysis and recognition, 1995, pp. 278-282.
[32]L. Breiman, "Random forests," Machine learning, vol. 45, pp. 5-32, 2001.
[33]J. H. Friedman, "Greedy function approximation: a gradient boosting machine," Annals of statistics, pp. 1189-1232, 2001.
[34]Y. Xi, X. Zhuang, X. Wang, R. Nie, and G. Zhao, "A research and application based on gradient boosting decision tree," in International Conference on Web Information Systems and Applications, 2018, pp. 15-26.
[35]T. Chen and C. Guestrin, "Xgboost: A scalable tree boosting system," in Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, 2016, pp. 785-794.
[36]G. Ke, Q. Meng, T. Finley, T. Wang, W. Chen, W. Ma, et al., "Lightgbm: A highly efficient gradient boosting decision tree," Advances in neural information processing systems, vol. 30, pp. 3146-3154, 2017.
[37]J. Cai, X. Li, Z. Tan, and S. Peng, "An assembly-level neutronic calculation method based on LightGBM algorithm," Annals of Nuclear Energy, vol. 150, p. 107871, 2021.
[38]L. Prokhorenkova, G. Gusev, A. Vorobev, A. V. Dorogush, and A. Gulin, "CatBoost: unbiased boosting with categorical features," arXiv preprint arXiv:1706.09516, 2017.
[39]B. Dhananjay and J. Sivaraman, "Analysis and classification of heart rate using CatBoost feature ranking model," Biomedical Signal Processing and Control, vol. 68, p. 102610, 2021.
[40]N. Chen and D. Blostein, "A survey of document image classification: problem statement, classifier architecture and performance evaluation," International Journal of Document Analysis and Recognition (IJDAR), vol. 10, pp. 1-16, 2007.
[41]J. Lever, M. Krzywinski, and N. Altman, "Classification evaluation," ed: Nature Publishing Group, 2016.
[42]J. M. Lobo, A. Jiménez‐Valverde, and R. Real, "AUC: a misleading measure of the performance of predictive distribution models," Global ecology and Biogeography, vol. 17, pp. 145-151, 2008.