臺灣博碩士論文加值系統

牙根斷裂是導致根管治療後牙齒拔除的常見原因。過去研究指出，使用假牙冠保護經過根管治療的牙齒可以降低牙根斷裂的發生機率。然而，在臨床實踐中，即使接受了根管治療和牙冠裝戴的牙齒，仍有可能出現牙根斷裂，給患者和醫生帶來困擾。牙根斷裂的成因尚未明確，且涉及多種因素，因此在治療前很難預估牙齒未來是否會出現裂縫。近年來，人工智能技術取得了飛速發展。因此，本研究收集了嘉義基督教醫院近五年內發生牙根斷裂的臨床病例，篩選出在該院接受根管治療和假牙製作的牙齒。此外，還收集了在臨床上接受相同治療並存活超過五年的牙齒作為對照組。為了更加深入地了解可能導致牙裂的因素，本研究使用基於深度神經網路的 Network-based Network 架構進行分析。這種架構能夠針對治療前表格資料中的數值型與類別型特徵，預測後續病人是否會發生牙裂，從而為臨床治療計劃的制定提供重要幫助。根據實驗結果顯示，研究提出的 Network-based Network 架構在預測後續病人是否會牙裂的準確率上達到了約 82.1%，這一成績相對於傳統機器學習方法如支援向量機的預測準確率提高了約 10%，並且相較於傳統的深度神經網路，也提高約 8% 的預測準確率。這一結果表明，本研究提出的 Network-based Network 架構在預測牙根是否會斷裂方面具有優越性，相較於傳統的機器學習方法和深度神經網路，表現出更高的預測準確性。

Root fracture is a common cause for tooth extraction following root canal treatment. Past studies have indicated that the use of crowns to protect teeth treated with root canal therapy can reduce the likelihood of root fracture. However, in clinical practice, even teeth that have undergone root canal treatment and crown placement may still suffer from root fracture, posing challenges for both patients and doctors. The causes of root fractures remain unclear and involve multiple factors, making it difficult to predict the occurrence of cracks before treatment. In recent years, artificial intelligence technology has seen rapid development. Therefore, this study collected clinical cases of root fractures that occurred in the last five years at Chiayi Christian Hospital, selecting teeth that underwent root canal treatment and crown fabrication. In addition, teeth that received the same treatment and survived for more than five years were collected for the control group. To gain a deeper understanding of the factors that may lead to tooth cracking, this study employed a Network-based Network structure based on deep neural networks for analysis. This architecture is capable of predicting the subsequent occurrence of tooth cracks by analyzing numerical and categorical features in the pre-treatment tabular data, providing significant assistance in the formulation of clinical treatment plans. According to experimental results, the proposed Network-based Network architecture achieved approximately 82.1% accuracy in predicting the occurrence of tooth cracks, an improvement of about 10% compared to traditional machine learning methods such as support vector machines and about 8% higher than traditional deep neural networks. This result demonstrates the superiority of the proposed Network-based Network structure in predicting root fractures, exhibiting higher predictive accuracy compared to traditional machine learning methods and deep neural networks.

摘要...i
Abstract...ii
誌謝...iv
目錄...v
表目錄...vii
圖目錄...viii
第一章 文獻探討...1
1.2 人工智慧...1
1.2.1 機器學習的訓練方法...2
1.2.2 機器學習的方法...4
1.2.3 深度學習應用於牙科研究...8
1.3 目標...10
第二章 研究內容與方法...12
2.1 蒐集病人資料...12
2.2 病人資料的前處理...13
2.2.1 對資料特徵進行數值型特徵與類別型特徵的分類...13
2.2.2 對類別型特徵進行的前處理...14
2.2.3 對數值型特徵進行的前處理...16
2.3 利用深度學習分析牙根斷裂...17
2.3.1 深度神經網路...17
2.3.2 Embedding...18
2.3.3 Batch normalization...18
2.3.4 Network-based Network (NbN)...18
第三章 實驗...20
3.1 資料集...20
3.2 效能指標...20
3.2.1 K 折交叉驗證...20
3.2.2 準確率與 f1-score...21
3.3 實驗設置...22
3.3.1 ReLU:...22
3.3.2 Adam:...22
3.4 Baselines...23
3.5 實驗結果...24
3.5.1 訓練過程...25
3.5.2 測試速度...32
3.5.3 測試結果...33
3.5.4 節點與隱藏層的調整...35
3.5.5 降維測試結果...48
第四章 結論...50
文獻探討...51
Extended Abstract...54
Abstract...54
1. Introduction...55
2. Methods...55
2.2 Preprocessing of Patient Data...55
2.2.1 Preprocessing of categorical features...56
2.2.2 Preprocessing of numerical features...56
2.3 Deep Learning Analysis of Root Fractures...56
3. Results...56
3.1 Baselines...56
3.5 Experimental Results...57
4. Conclusion...57

[1]　Touré, B., Faye, B., Kane, A. W., Lo, C. M., Niang, B., & Boucher, Y. 2011, “Analysis of reasons for extraction of endodontically treated teeth: a prospective study”, Journal of endodontics, 37(11), 1512-1515.
[2]　RIVERA, E. M., & WALTON, R. E. 2007, “Longitudinal tooth fractures: findings that contribute to complex endodontic diagnoses”, Endodontic Topics, 16(1), 82-111.
[3]　Sedgley, C. M., & Messer, H. H. 1992, “Are endodontically treated teeth more brittle?”, Journal of endodontics, 18(7), 332-335.
[4]　Reeh, E. S., Messer, H. H., & Douglas, W. H. 1989, “Reduction in tooth stiffness as a result of endodontic and restorative procedures”, Journal of endodontics, 15(11), 512-516.
[5]　Dai, L. X., Tong, W. L., & Guo, J. L. 2013, “The role of occlusal factors in the occurrence of vertical root fracture”, Shanghai kou Qiang yi xue= Shanghai Journal of Stomatology, 22(1), 68-71.
[6]　Barreto, M. S., do Amaral Moraes, R., da Rosa, R. A., Moreira, C. H. C., Só, M. V. R., & Bier, C. A. S. 2012, “Vertical root fractures and dentin defects: effects of root canal preparation, filling, and mechanical cycling”, Journal of endodontics, 38(8), 1135-1139.
[7]　Wilcox, L. R., Roskelley, C., & Sutton, T. 1997, “The relationship of root canal enlargement to finger-spreader induced vertical root fracture”, Journal of Endodontics, 23(8), 533-534.
[8]　Saw, L. H., & Messer, H. H. 1995, “Root strains associated with different obturation techniques”, Journal of Endodontics, 21(6), 314-320.
[9]　Vertucci FJ. Management of vertical root fracture. J Endod 1985;11(3):126-31.
[10]　Jenhani, I., Amor, N. B., & Elouedi, Z. 2008, “Decision trees as possibilistic classifiers”, International journal of approximate reasoning, 48(3), 784-807.
[11]　Charbuty, B., & Abdulazeez, A. 2021, “Classification based on decision tree algorithm for machine learning”, Journal of Applied Science and Technology Trends, 2(01), 20-28.
[12]　Granik, M., & Mesyura, V. 2017, May, “Fake news detection using naive Bayes classifier”, In 2017 IEEE first Ukraine conference on electrical and computer engineering (UKRCON) (pp. 900-903). IEEE.
[13]　Zhang, Y. 2012, “Support vector machine classification algorithm and its application”, In Information Computing and Applications: Third International Conference, ICICA 2012, Chengde, China, September 14-16, 2012. Proceedings, Part II 3 (pp. 179-186). Springer Berlin Heidelberg.
[14]　Altman, N. S. 1992, “An introduction to kernel and nearest-neighbor nonparametric regression”, The American Statistician, 46(3), 175-185.
[15]　Wang, C., Long, Y., Li, W., Dai, W., Xie, S., Liu, Y., ... & Duan, Y. 2020, “Exploratory study on classification of lung cancer subtypes through a combined K-nearest neighbor classifier in breathomics”, Scientific reports, 10(1), 5880.
[16]　Ho, T. K. 1995, August, “Random decision forests”, In Proceedings of 3rd international conference on document analysis and recognition (Vol. 1, pp. 278-282). IEEE.
[17]　Schonlau, M., & Zou, R. Y. 2020, “The random forest algorithm for statistical learning”, The Stata Journal, 20(1), 3-29.
[18]　Li, Z., Liu, F., Yang, W., Peng, S., & Zhou, J. 2021, “A survey of convolutional neural networks: analysis, applications, and prospects”, IEEE transactions on neural networks and learning systems.
[19]　Schmidt, R. M. 2019, “Recurrent neural networks (rnns): A gentle introduction and overview”, arXiv preprint arXiv:1912.05911.
[20]　Singh, A., Dargar, S. K., Gupta, A., Kumar, A., Srivastava, A. K., Srivastava, M., ... & Ullah, M. A. 2022, “Evolving long short-term memory network-based text classification”, Computational Intelligence and Neuroscience, 2022.
[21]　Gui, J., Sun, Z., Wen, Y., Tao, D., & Ye, J. 2021, “A review on generative adversarial networks: Algorithms, theory, and applications”, IEEE transactions on knowledge and data engineering, 35(4), 3313-3332.
[22]　Khanagar, S. B., Al-Ehaideb, A., Maganur, P. C., Vishwanathaiah, S., Patil, S., Baeshen, H. A., ... & Bhandi, S. 2021, “Developments, application, and performance of artificial intelligence in dentistry–A systematic review”, Journal of dental sciences, 16(1), 508-522.
[23]　Boreak, N. 2020, “Effectiveness of artificial intelligence applications designed for endodontic diagnosis, decision-making, and prediction of prognosis: a systematic review”, J Contemp Dent Pract, 21(8), 926-934.
[24]　Saghiri, M. A., Asgar, K., Boukani, K. K., Lotfi, M., Aghili, H., Delvarani, A., ... & Garcia‐Godoy, F. 2012, “A new approach for locating the minor apical foramen using an artificial neural network”, International endodontic journal, 45(3), 257-265.
[25]　Campo, L., Aliaga, I. J., De Paz, J. F., García, A. E., Bajo, J., Villarubia, G., & Corchado, J. M. 2016, “Retreatment predictions in odontology by means of CBR systems”, Computational Intelligence and Neuroscience, 2016, 39-39.
[26]　Mahmoud, Y. E., Labib, S. S., & Mokhtar, H. M. 2015, December, “Clinical prediction of teeth periapical lesion based on machine learning techniques”, In The Second International Conference on Digital Information Processing, Data Mining, and Wireless Communications (DIPDMWC2015) (pp. 9-15).
[27]　Ekert, T., Krois, J., Meinhold, L., Elhennawy, K., Emara, R., Golla, T., & Schwendicke, F. 2019, “Deep learning for the radiographic detection of apical lesions”, Journal of endodontics, 45(7), 917-922.
[28]　Orhan, K., Bayrakdar, I. S., Ezhov, M., Kravtsov, A., & Özyürek, T. A. H. A. 2020, “Evaluation of artificial intelligence for detecting periapical pathosis on cone‐beam computed tomography scans”, International endodontic journal, 53(5), 680-689.
[29]　Johari, M., Esmaeili, F., Andalib, A., Garjani, S., & Saberkari, H. 2017, “Detection of vertical root fractures in intact and endodontically treated premolar teeth by designing a probabilistic neural network: an ex vivo study”, Dentomaxillofacial Radiology, 46(2), 20160107.
[30]　Fukuda, M., Inamoto, K., Shibata, N., Ariji, Y., Yanashita, Y., Kutsuna, S., ... & Ariji, E. 2020, “Evaluation of an artificial intelligence system for detecting vertical root fracture on panoramic radiography”, Oral Radiology, 36, 337-343.
[31]　Hatvani, J., Horváth, A., Michetti, J., Basarab, A., Kouamé, D., & Gyöngy, M. 2018, “Deep learning-based super-resolution applied to dental computed tomography”, IEEE Transactions on Radiation and Plasma Medical Sciences, 3(2), 120-128.
[32]　Hiraiwa, T., Ariji, Y., Fukuda, M., Kise, Y., Nakata, K., Katsumata, A., ... & Ariji, E. 2019, “A deep-learning artificial intelligence system for assessment of root morphology of the mandibular first molar on panoramic radiography”, Dentomaxillofacial Radiology, 48(3), 20180218.
[33]　Kuwada, C., Ariji, Y., Fukuda, M., Kise, Y., Fujita, H., Katsumata, A., & Ariji, E. 2020, “Deep learning systems for detecting and classifying the presence of impacted supernumerary teeth in the maxillary incisor region on panoramic radiographs”, Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology, 130(4), 464-469.
[34]　Chen, H., Zhang, K., Lyu, P., Li, H., Zhang, L., Wu, J., & Lee, C. H. 2019, “A deep learning approach to automatic teeth detection and numbering based on object detection in dental periapical films”, Scientific reports, 9(1), 3840.
[35]　Luo, Y., Zhou, H., Tu, W. W., Chen, Y., Dai, W., & Yang, Q. 2020, July, “Network on network for tabular data classification in real-world applications”, In Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval (pp. 2317-2326).
[36]　Panda, S. K., & Jana, P. K. 2015, “Efficient task scheduling algorithms for heterogeneous multi-cloud environment”, The Journal of Supercomputing, 71, 1505-1533.
[37]　Kingma, D. P., & Ba, J. 2014, “Adam: A method for stochastic optimization”, arXiv preprint arXiv:1412.6980.
[38]　Glorot, X., Bordes, A., & Bengio, Y. 2011, June, “Deep sparse rectifier neural networks”, In Proceedings of the fourteenth international conference on artificial intelligence and statistics (pp. 315-323). JMLR Workshop and Conference Proceedings.