植入式醫療器材主要是用於治療重大創傷，以及輔助治療疾病的主要手段之一，為避免植入物在治療後於患者體內損壞，因此探討植入物的機械性能是必要的過程。在過去的探討方式多為機械實驗測試以及有限元素分析進行骨板植入物的相關研究，但是在現今對於客製化的要求逐漸提升，而對骨板植入物的設計優化的探討日益加重的情況下，機械實驗以及有限元素分析的方式需要耗費大量的時間以及開發的成本，因此本研究的目的是結合有限元素分析以及深度學習技術，開發能夠對骨板設計進行探討的快速計算模式。
在本研究中，使用ANSYS Workbench 2022 R1進行骨板模型的建立，並用於準備後續深度學習所需資料，骨板的負載是參考四點彎曲測試的實驗環境，在邊界條件方面，承重滾輪設定為完全拘束，負載滾輪施以向下500N的負載。深度學習的方面是在Python 3.8版本環境進行開發，並使用卷積神經網路(CNN)作為預測之模型，並對模型架構進行準確度的探討。使用的訓練資料中，會將骨板模型的設計參數提取出來並將其進行矩陣化，以符合卷積神經網路之輸入端的資料型態，而輸出端則考慮骨板的六個螺絲孔洞周圍最大馮米塞斯(von Mises)應力。
有限元素分析的研究表明可以評估骨板的植入應力，而本研究基於 CNN 的深度學習模型可以快速預測骨板設計的機械性能，這種深度學習模型有望減少骨板設計優化的成本和時間，因此有限元分析與深度學習技術的集成是預測四點彎曲載荷下金屬骨板設計的彎曲強度的可行且有效的工具。

Implantable medical devices are mainly used to treat trauma and assist in the treatment of diseases. In order to avoid damage to the implant in the patient's body after treatment, it is necessary to explore the mechanical properties of the implant. In the past, the research methods of bone plate implants were mainly based on mechanical experiments and finite element analysis. However, the requirements for customization are gradually increasing, and the discussion on the design optimization of bone plate implants is becoming more and more serious. In some cases, the method of mechanical experiment and finite element analysis needs a lot of time and development cost. Therefore, the purpose of this study was to combine finite element analysis and deep learning technology to develop a fast calculation method that can discuss the bone plate design model.
In this study, ANSYS Workbench 2022 R1 was used to establish the bone plate model, and a total of 24,143 finite element models were solved for deep learning. The load of the bone plate was based on the experimental environment of the four-point bending test. In order to reduce the solution time, a half-bone plate model was developed. In the boundary conditions, the load-bearing cylinder was set to be fully restrained, and a downward load of 500N was applied to the load cylinder. The deep learning model was developed in the Python 3.8 version environment, and the convolutional neural network (CNN) was used as the prediction model. For the preparation of the training and testing data, the input variables were transformed into a matrix, and the maximum von Mises stress on the screw holes was used as the output performance.
The finite element study showed that the implant stress of the bone plates could be evaluated. The CNN-based deep learning model could quickly predict the mechanical performances of the bone plate designs. This deep learning model was expected to reduce the cost and time for design optimization of bone plates. Integration of finite element analysis with deep learning technique was a feasible and effective tool for predicting the flexural strength of the metal bone plate designs under a four-point bending load.

[1]Avery, C. M. E., Bujtár, P., Simonovics, J., Dézsi, T., Váradi, K., Sándor, G. K. B., & Pan, J. (2013). A finite element analysis of bone plates available for prophylactic internal fixation of the radial osteocutaneous donor site using the sheep tibia model. Medical Engineering & Physics, 35(10), 1421-1430.
[2]Le, C., Kolasangiani, K., Nayyeri, P., & Bougherara, H. (2023). Experimental and numerical investigation of 3D-Printed bone plates under four-point bending load utilizing machine learning techniques. Journal of the Mechanical Behavior of Biomedical Materials, 143, 105885.
[3]Yamada, Y., Yoshimura, N., & Sakurai, T. (1968). Plastic stress-strain matrix and its application for the solution of elastic-plastic problems by the finite element method. International Journal of Mechanical Sciences, 10(5), 343-354.
[4]Geng, J.-P., Tan, K. B. C., & Liu, G.-R. (2001). Application of finite element analysis in implant dentistry: A review of the literature. The Journal of Prosthetic Dentistry, 85(6), 585-598.
[5]De Jager, N., Pallav, P., & Feilzer, A. J. (2005). The influence of design parameters on the FEA-determined stress distribution in CAD–CAM produced all-ceramic dental crowns. Dental Materials, 21(3), 242-251.
[6]De Jager, N., de Kler, M., & van der Zel, J. M. (2006). The influence of different core material on the FEA-determined stress distribution in dental crowns. Dental Materials, 22(3), 234-242.
[7]Schutte, S., van den Bedem, S. P. W., van Keulen, F., van der Helm, F. C. T., & Simonsz, H. J. (2006). A finite-element analysis model of orbital biomechanics. Vision Research, 46(11), 1724-1731.
[8]Elangovan, S., Semeer, S., & Prakasan, K. (2009). Temperature and stress distribution in ultrasonic metal welding—An FEA-based study. Journal of Materials Processing Technology, 209(3), 1143-1150.
[9]Panagiotopoulou, O. (2009). Finite element analysis (FEA): Applying an engineering method to functional morphology in anthropology and human biology. Annals of Human Biology, 36(5), 609-623.
[10]Kamaya, M., & Kawakubo, M. (2011). A procedure for determining the true stress–strain curve over a large range of strains using digital image correlation and finite element analysis. Mechanics of Materials, 43(5), 243-253.
[11]Gervais, B., Vadean, A., Raison, M., & Brochu, M. (2016). Failure analysis of a 316L stainless steel femoral orthopedic implant. Case Studies in Engineering Failure Analysis, 5-6, 30-38.
[12]Szabó, B., & Babuška, I. (2021). Finite Element Analysis: Method, Verification and Validation.
[13]Lloyd, S., Mohseni, M., & Rebentrost, P. (2013). Quantum algorithms for supervised and unsupervised machine learning. arXiv preprint arXiv:1307.0411.
[14]Scheurer, M. S., & Slager, R.-J. (2020). Unsupervised Machine Learning and Band Topology. Physical Review Letters, 124(22), 226401.
[15]Vurtur Badarinath, P., Chierichetti, M., & Davoudi Kakhki, F. (2021). A Machine Learning Approach as a Surrogate for a Finite Element Analysis: Status of Research and Application to One Dimensional Systems. Sensors, 21(5), 1654.
[16]Najjar, I. M. R., Sadoun, A. M., Alsoruji, G. S., Elaziz, M. A., & Wagih, A. (2022). Predicting the mechanical properties of Cu–Al2O3 nanocomposites using machine learning and finite element simulation of indentation experiments. Ceramics International, 48(6), 7748-7758.
[17]Zhang, C., Li, Y., Jiang, B., Wang, R., Liu, Y., & Jia, L. (2022). Mechanical properties prediction of composite laminate with FEA and machine learning coupled method. Composite Structures, 299, 116086.
[18]Deng, L., & Yu, D. (2014). Deep Learning: Methods and Applications. Foundations and Trends® in Signal Processing, 7(3–4), 197-387.
[19]LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436-444.
[20]Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. MIT press.
[21]Fakoor, R., Ladhak, F., Nazi, A., & Huber, M. (2013). Using deep learning to enhance cancer diagnosis and classification. Proceedings of the international conference on machine learning, Association for computing machinery, 3937-3949.
[22]Ching, T., Himmelstein, D. S., Beaulieu-Jones, B. K., Kalinin, A. A., Do, B. T., Way, G. P., Ferrero, E., Agapow, P.-M., Zietz, M., Hoffman, M. M., Xie, W., Rosen, G. L., Lengerich, B. J., Israeli, J., Lanchantin, J., Woloszynek, S., Carpenter, A. E., Shrikumar, A., Xu, J., . . . Greene, C. S. (2018). Opportunities and obstacles for deep learning in biology and medicine. Journal of The Royal Society Interface, 15(141), 20170387.
[23]Koeppe, A., Bamer, F., & Markert, B. (2020). An intelligent nonlinear meta element for elastoplastic continua: deep learning using a new Time-distributed Residual U-Net architecture. Computer Methods in Applied Mechanics and Engineering, 366, 113088.
[24]Dalvi, C., Rathod, M., Patil, S., Gite, S., & Kotecha, K. (2021). A Survey of AI-Based Facial Emotion Recognition: Features, ML & DL Techniques, Age-Wise Datasets and Future Directions. IEEE Access, 9, 165806-165840.
[25]Fang, Z., Roy, K., Mares, J., Sham, C.-W., Chen, B., & Lim, J. B. P. (2021). Deep learning-based axial capacity prediction for cold-formed steel channel sections using Deep Belief Network. Structures, 33, 2792-2802.
[26]Patel, R., Thong, E. H. E., Batta, V., Bharath, A. A., Francis, D., & Howard, J. (2021). Automated Identification of Orthopedic Implants on Radiographs Using Deep Learning. Radiology: Artificial Intelligence, 3(4), e200183.
[27]Baruah, S., Sarkar, S., Singh, I. V., & Mishra, B. K. (2022). A computational framework based on FEA, ML and GA for estimation of welding residual stresses. Finite Elements in Analysis and Design, 205, 103753.
[28]Madani, A., Bakhaty, A., Kim, J., Mubarak, Y., & Mofrad, M. R. (2019). Bridging finite element and machine learning modeling: stress prediction of arterial walls in atherosclerosis. Journal of biomechanical engineering, 141(8), 084502.
[29]Zhang, Z., & Gu, G. X. (2020). Finite‐Element‐Based Deep‐Learning Model for Deformation Behavior of Digital Materials. Advanced Theory and Simulations, 3(7), 2000031.
[30]Ford, E., Maneparambil, K., Rajan, S., & Neithalath, N. (2021). Machine learning-based accelerated property prediction of two-phase materials using microstructural descriptors and finite element analysis. Computational Materials Science, 191, 110328.
[31]Silva, G. C., Beber, V. C., & Pitz, D. B. (2021). Machine learning and finite element analysis: An integrated approach for fatigue lifetime prediction of adhesively bonded joints. Fatigue & Fracture of Engineering Materials & Structures, 44(12), 3334-3348.
[32]Vurtur Badarinath, P., Chierichetti, M., & Davoudi Kakhki, F. (2021). A machine learning approach as a surrogate for a finite element analysis: Status of research and application to one dimensional systems. Sensors, 21(5), 1654.
[33]黃泗堯, (2019), 基於深度學習之端到端語音辨識與物件偵測於雙臂機器人之整合應用, 碩士論文, 國立臺灣科技大學.
[34]徐子仁, (2018), 基於深度學習的人臉辨識系統, 碩士論文, 國立臺灣科技大學.
[35]Nie, Z., Jiang, H., & Kara, L. B. (2020). Stress field prediction in cantilevered structures using convolutional neural networks. Journal of Computing and Information Science in Engineering, 20(1), 011002.
[36]Abueidda, D. W., Koric, S., & Sobh, N. A. (2020). Topology optimization of 2D structures with nonlinearities using deep learning. Computers & Structures, 237, 106283.
[37]Hsu, C. C., Lin, J., & Chao, C. K. (2011). Comparison of multiple linear regression and artificial neural network in developing the objective functions of the orthopaedic screws. Computer methods and programs in biomedicine, 104(3), 341-348.
[38]Hajializadeh, F., & Ince, A. (2021). Integration of artificial neural network with finite element analysis for residual stress prediction of direct metal deposition process. Materials Today Communications, 27, 102197.
[39]Seventekidis, P., & Giagopoulos, D. (2021). A combined finite element and hierarchical Deep learning approach for structural health monitoring: Test on a pin-joint composite truss structure. Mechanical Systems and Signal Processing, 157, 107735.
[40]Fernandez-Navamuel, A., Zamora-Sánchez, D., Omella, Á. J., Pardo, D., Garcia-Sanchez, D., & Magalhães, F. (2022). Supervised Deep Learning with Finite Element simulations for damage identification in bridges. Engineering Structures, 257, 114016.
[41]Zargar, S. A., & Yuan, F. G. (2021). Impact diagnosis in stiffened structural panels using a deep learning approach. Structural Health Monitoring, 20(2), 681-691.
[42]Li, X., Liu, Z., Cui, S., Luo, C., Li, C., & Zhuang, Z. (2019). Predicting the effective mechanical property of heterogeneous materials by image based modeling and deep learning. Computer Methods in Applied Mechanics and Engineering, 347, 735-753.
[43]Wang, T., Shao, M., Guo, R., Tao, F., Zhang, G., Snoussi, H., & Tang, X. (2021). Surrogate model via artificial intelligence method for accelerating screening materials and performance prediction. Advanced Functional Materials, 31(8), 2006245.
[44]Liang, L., Liu, M., Martin, C., & Sun, W. (2018). A deep learning approach to estimate stress distribution: a fast and accurate surrogate of finite-element analysis. Journal of The Royal Society Interface, 15(138), 20170844.
[45]Lee, C. H., Hsu, C. C., & Huy, D. C. (2014). An optimization study of the screw orientation on the interfacial strength of the anterior lumbar plate system using neurogenetic algorithms and experimental validation. Journal of biomechanical engineering, 136(11), 111007.
[46]Yeh, J. (2017, December 25). [資料分析&機器學習] 第5.1講: 卷積神經網絡介紹(Convolutional Neural Network). Medium. https://medium.com/jameslearningnote/資料分析-機器學習-第5-1講-卷積神經網絡介紹-convolutional-neural-network-4f8249d65d4f.
[47]Pellissetti, M. F., & Schuëller, G. I. (2006). On general purpose software in structural reliability – An overview. Structural Safety, 28(1), 3-16.
[48]Shabliy, L., & Cherniaev, A. (2014). Optimization of gas turbine compressor blade parameters for gas-dynamic efficiency under strength constraints. 2014 4th International Conference On Simulation And Modeling Methodologies, Technologies And Applications.
[49]Xiangjie, J., Cai, Z., & Chong, W. (2020). RESPONSE SURFACE OPTIMIZATION OF MACHINE TOOL COLUMN BASED ON ANSYS WORKBENCH. Academic Journal of Manufacturing Engineering, 18(2).
[50]ASTM, A. (2014). E290-14 Standard Test Methods for Bend Testing of Material for Ductility.