人工智能在髋关节置换术中的研究进展
Progress of Artificial Intelligence in Total Hip Arthroplasty
DOI: 10.12677/ACM.2024.142493, PDF, HTML, XML, 下载: 76  浏览: 168 
作者: 刘宇钊:山东大学附属省立医院关节外科,山东 济南;孙 水*:山东大学附属省立医院关节外科,山东 济南;山东第一医科大学附属省立医院关节外科,山东 济南
关键词: 人工智能全髋关节置换术机器学习深度学习综述Artificial Intelligence Total Hip Arthroplasty Machine Learning Deep Learning Review
摘要: 全髋关节置换术(total hip arthroplasty, THA)可以通过减轻患者疼痛、恢复下肢功能以及矫正步态畸形,从而改善患者日常生活质量。随着医疗技术的不断进步,THA的数量也逐年增加。人工智能(artificial intelligence, AI)近年来在医疗领域应用十分广泛,本文通过对AI在THA中的研究进展进行总结,以期为患者和临床医生提供个性化治疗新思路。
Abstract: Total Hip Arthroplasty (THA) can improve the quality of daily life of patients by relieving pain, re-storing lower limb function, and correcting gait deformities. As medical technology continues to advance, the number of THA is increasing year by year. Artificial intelligence (AI) has been widely used in the medical field in recent years, and this paper summarizes the research progress of AI in THA in order to provide patients and clinicians with new ideas for personalized treatment.
文章引用:刘宇钊, 孙水. 人工智能在髋关节置换术中的研究进展[J]. 临床医学进展, 2024, 14(2): 3521-3527. https://doi.org/10.12677/ACM.2024.142493

1. 引言

人工智能(artificial intelligence, AI)这一术语由约翰·麦卡锡在1965年提出,当时,“有生命的计算机”这一概念占据了人们的思维 [1] 。AI是一种机器获取信息、将其转化为理论知识而产生改变环境的反应的迭代过程。这是一个广泛的概念,其中涉及到虚拟(计算)和物理(机器)元素 [2] 。AI算法包括多个分支,如机器学习(machine learning, ML)和深度学习(deep learning, DL)等。ML在医疗保健领域和康复领域,以及DL在医学影像领域越来越受欢迎 [2] [3] 。AI在改善医疗保健方面有巨大潜力,几年内,人工智能可能会改变日常临床实践方式 [4] 。髋关节通常被描述为由髋臼和股骨头组成的球窝关节,作为连接躯干和下肢的主要关节,在肢体活动中发挥重要作用 [5] 。据估计,在跨步过程中,髋关节需要承受2.5倍于身体重量的力来平衡髋臼–股骨支点处的压力 [6] 。全髋关节置换术是最成功的骨科手术之一,为患有终末期髋关节骨关节炎等髋关节疾病的患者提供了可靠的治疗效果,特别是缓解疼痛、恢复髋关节功能和改善整体生活质量 [7] 。据估计,到2030年,每年的THA数量将达到57.2~138.5万例 [8] 。但髋关节在标准的人体轴位、冠状位和矢状位上定向不佳,所以很难成像 [9] 。AI在自动识别THA植入物等许多方面显示出优势,本文对AI在THA的应用进行综述。

2. DL预测骨盆矢状倾角(predicting pelvic sagittal, PSI)和髋关节中心(hip joint center, HJC)

最近研究表明,PSI影响髋臼假体的功能位置,是THA术后发生撞击和脱位的关键因素 [10] [11] ,HJC也具有重要的生物力学意义,是髋关节假体放置的重要参考 [12] 。之前提出的一种监督学习方法需要患者的CT图像,这对患者的辐射较大 [13] 。而在许多医院,CT并非常规检查项目。Ata Jodeiri等人提出的DL框架方法仅使用x线图像便可对PSI进行自动、稳定的评估,这对改善髋臼假体放置位置非常重要 [14] 。Seong等人基于DL开发的特异性模型能在骨盆x线上识别HJC,对THA术前评估假体位置和降低HJC评估差异性具有参考价值 [15] 。这些方法仅使用x线便可完成评估,能够很大程度减少患者的辐射暴露。但这些方法需要在临床中进行充分的验证,以确保其在实践中的功能性和安全性。

3. AI HIP在THA术前规划的应用

成功的THA需要选择合适的假体尺寸,因此准确的术前规划有助于提前预知术后结果,这对外科医生在术中做出准确判断十分重要 [16] 。然而,精准规划对时间、人力、和复杂的工作流程的要求限制了其应用 [17] 。随着AI的发展,一种新的术前规划软件——AI HIP应运而生,这是一种基于CT数据的三维图像处理系统,可以将AI与医疗大数据结合进行THA术前精准规划。一项前瞻性研究使用AI HIP、3D模型、2D数字模型进行THA术前规划,结果显示AI HIP在预测假体尺寸方面表现出优越性 [18] 。另一项回顾性研究比较了三维AI HIP软件和传统二维手工模型在预测THA假体尺寸和位置的准确性,结果显示AI HIP具有更高的可靠性 [19] 。总之,AI HIP在THA术前规划中具有更高的准确性,但其临床意义需进一步探究。

4. ML预测THA住院时间、支付模式和手术时间

临床上患者比较关心的就是住院时间、费用和手术时间等问题。下肢关节置换术中,综合护理模式的应用降低了患者的住院时间、30天内再住院率和费用,但未解决潜在风险问题 [20] 。Prem N. Ramkumar等人基于ML将患者术前大数据导入算法模型,在预测住院时间和费用方面表现出良好的有效性和可靠性 [21] 。2022年,Igor Lazic等人通过调整输入和输出数据,能够建立一个ML模型用于预测THA的不规则手术时间,这一ML模型有望在临床实践中得到广泛应用 [22] 。

5. ML预测骨矿物质密度和改进假体选择

THA可以明显改善髋关节病变引起的疼痛,但仍有许多患者对术后自理能力和下肢无力感到失望。THA可以分为使用骨水泥和不使用骨水泥,研究表明,非骨水泥THA的总体生存率较低 [23] ,但目前仍没有通用的标准来决定使用哪种手术方式,临床医生必须根据实际情况来进行选择。骨骼和肌肉的质量可以为临床医生的决策提供有效的参考 [24] 。如何选择合适的假体是术前决策的关键,这可以减少骨折等术后并发症的发生。骨矿物质密度也是另一个重要的影响因素,可以用来评估患者的长期预后。Carlo Ricciardi等人利用ML技术和生物识别特异性分析患者假体选择,ML还可以预测股骨近端和远端的骨矿物质密度来评估患者长期预后,有效的预测这两个因素可以帮助临床医生更好的决策 [25] 。

6. ML预测THA术后患者满意度

THA可以为患者减轻疼痛并改善生活质量。通常使用患者报告的结果测量和患者满意度评估手术效果。后者可以独特且全面的反应患者术后主观生活质量和术后改善情况 [26] [27] 。然而,患者满意度受很多因素影响,如年龄、性别、心理状况和术前期望等 [28] [29] [30] 。最近研究表明,大约10%~20%的患者在接受THA术后不满意 [31] [32] 。因此,预测术后患者满意度可以为患者提供个性化的术前咨询、减少患者不切实际的手术期望。先前一项研究开发了一种有ML算法,利用患者人口统计学、患者并发症和牛津髋关节评分等因素预测患者术后2年的满意度。这一算法在预测患者THA术后满意度上表现良好并确定了具体的影响因素,可以用于帮助临床医生改善术前咨询和THA术前的健康优化 [33] 。

7. ML预测THA术后输血率

当前,大多数外科医生在进行THA时选择使用氨甲环酸、促红细胞生成素或自体血再输注来改善术中失血 [34] 。然而,THA术后输血率的发生率仍高达9% [35] ,因此预测术后输血率对患者术后满意度十分重要。Wayne Brian Cohen-Levy等人开发了4种ML算法并分别评估其性能,均在预测THA患者术后特异性输血率方面表现优异,并可能改善术前计划和手术结果 [36] 。

8. ML预测THA术后服用阿片类药物风险

阿片类药物能够有效缓解各种因素引起的疼痛,其使用范围极其广泛,术前服用阿片类药物会增加术后服用慢性阿片类药物的风险,抑郁史、基线疼痛评分较高、年龄小等因素同样会增加使用风险 [37] [38] [39] [40] 。由于资源有限,减少阿片类药物使用的周围神经调节等方法不能为患者普遍使用 [41] 。因此需要进一步开发预测模型来更好地分配资源。Rodney A Gabriel等人比较了各种ML预测模型,结果显示集成学习可以很好的改善阿片类药物持续使用的预测模型,准确识别高危患者,为术前优化提供个性化干预措施 [42] 。

9. DL基于x线自动识别THA假体植入物

由于年龄等因素,髋关节翻修术的比率不断增加。翻修术前最重要的步骤之一便是识别假体型号,这对术前计划有很大意义。有研究表明,每个翻修病例大约需要20分钟来确定假体型号,仍有10%的假体术前无法确定,2%的假体术中无法识别。这会导致手术时间的和医疗费用的增加 [43] 。有研究证明了一种新的DL方法可以在髋关节x线中识别4种不同的髋关节假体,具有在翻修术前对假体进行分类的潜力 [44] 。Jaret M Karnuta等人训练、验证并外部测试了一个DL系统软件,通过x线图像对假体进行分类并表现出出色性能,有助于改善髋关节翻修术前计划 [45] 。

10. DL、ML预测THA术后假体松动

由于人口老龄化等原因,THA的数量不断增加,这导致髋关节翻修术的数量也显著增加 [46] [47] 。THA失败的常见原因包括假体松动、骨溶解和术后感染等,其中最主要的是假体松动。但其检测仍是一个棘手的难题,通常需要在翻修手术过程中才能确定 [48] [49] [50] 。X线、关节造影、核磁共振成像等方法可用于帮助假体松动诊断,但这些方法通常都有辐射、不敏感且缺乏有效性等缺点 [51] 。因此,需要更加便捷有效的方法来识别假体松动的发生。Mattia Loppini等人基于DL开发了一套自动放射影像错误识别系统,通过卷积神经网络进行分析,可以准确地检测到髋关节假体的松动 [52] 。Romil F Shah等人通过训练一系列卷积神经网络模型来评估ML算法识别假体松动的能力。结果表明虽然目前发展下ML算法无法独立完成假体松动检测,但仍是一种临床决策的有用工具 [53] 。

11. DL预测THA术后脱位风险

脱位是THA术后最常见的早期并发症,也是导致翻修术的主要原因之一 [46] [54] 。脱位可以导致患者出现严重疼痛和肢体功能丧失等问题,因此准确预测初次THA术后脱位风险对制定个性化手术方案和术后康复计划至关重要。Alireza Borjali等人开发了一种基于DL的自然语言处理模型以检测初次THA术后假体位置。该模型可以准确预测髋关节脱位风险并改善患者预后 [55] 。Pouria Rouzrokh等人开发了一种基于DL的影像分类模型,能够结合临床危险因素对THA术后脱位进行快速评估,这一研究展示了自动成像模型在骨科应用的潜力 [56] 。

12. 小结

目前,越来越多患有髋关节炎、股骨头坏死或先天性髋关节发育不良等疾病的患者选择接受THA来改善下肢功能和日常生活质量。虽然THA是一种十分成熟的手术方式,但仍有许多因素需要进一步改善。随着新兴科技和医疗技术的不断进步,AI已经在THA的许多方面展现出优势。在THA术前,AI可用于预测PSI和HJC、住院时间、患者满意度和骨矿物质密度等。此外,AI-HIP可以为患者优化术前计划。在THA术后,AI可用于预测输血率、识别髋关节假体类型、预测术后假体松动和脱位风险。这表明AI在THA中的应用具有很大的发展潜力,但随着患者需求的不断提高也将面临更大挑战。随着国内外的不断深入研究,相信在将来AI在THA中的应用会有进一步的飞跃。

NOTES

*通讯作者。

参考文献

[1] Bini, S.A. (2018) Artificial Intelligence, Machine Learning, Deep Learning, and Cognitive Computing: What Do These Terms Mean and How Will They Impact Health Care? The Journal of Arthroplasty, 33, 2358-2361.
https://doi.org/10.1016/j.arth.2018.02.067
[2] Anderson, D. (2019) Artificial Intelligence and Applications in PM&R. American Journal of Physical Medicine & Rehabilitation, 98, e128-e129.
https://doi.org/10.1097/PHM.0000000000001171
[3] Do, S., Song, K.D. and Chung, J.W. (2020) Basics of Deep Learning: A Radiologist’s Guide to Understanding Published Radiology Articles on Deep Learning. Korean Journal of Radiology, 21, 33-41.
https://doi.org/10.3348/kjr.2019.0312
[4] Laur, O. and Wang, B. (2022) Musculoskeletal Trauma and Artificial Intelligence: Current Trends and Projections. Skeletal Radiology, 51, 257-269.
https://doi.org/10.1007/s00256-021-03824-6
[5] Polkowski, G.G. and Clohisy, J.C. (2010) Hip Biomechanics. Sports Medicine and Arthroscopy Review, 18, 56-62.
https://doi.org/10.1097/JSA.0b013e3181dc5774
[6] Dimon, JH. (1974) Surgical Anatomy of the Hip. Surgical Clinics of North America, 54, 1327-1335.
https://doi.org/10.1016/S0039-6109(16)40488-3
[7] Varacallo, M., Luo, T.D. and Johanson, N.A. (2022) Total Hip Arthroplasty Techniques. StatPearls Publishing, Treasure Island, FL.
[8] Singh, J.A., Yu, S., Chen, L. and Cleveland, J.D. (2019) Rates of Total Joint Replacement in the United States: Future Projections to 2020-2040 Using the National Inpatient Sample. The Journal of Rheumatology, 46, 1134-1140.
https://doi.org/10.3899/jrheum.170990
[9] Chang, C.Y. and Huang, A.J. (2013) MR Imaging of Normal Hip Anatomy. Magnetic Resonance Imaging Clinics of North America, 21, 1-19.
https://doi.org/10.1016/j.mric.2012.08.006
[10] Shon, W.Y., Sharma, V., Keon, O.J., Moon, J.G. and Suh, D.H. (2014) Can Pelvic Tilting Be Ignored in Total Hip Arthroplasty? International Journal of Surgery Case Reports, 5, 633-636.
https://doi.org/10.1016/j.ijscr.2014.07.015
[11] Bhaskar, D., Rajpura, A. and Board, T. (2017) Current Concepts in Acetabular Positioning in Total Hip Arthroplasty. Indian Journal of Orthopaedics, 51, 386-396.
https://doi.org/10.4103/ortho.IJOrtho_144_17
[12] Durand-Hill, M., Henckel, J., Satchithananda, K., et al. (2016) Calculating the Hip Center of Rotation Using Contralateral Pelvic Anatomy. Journal of Orthopaedic Research, 34, 1077-1083.
https://doi.org/10.1002/jor.23118
[13] Shun, M., Wang, Z.J. and Rui, L. (2016) A CNN Regression Approach for Real-Time 2D/3D Registration. IEEE Transactions on Medical Imaging, 35, 1352-1363.
https://doi.org/10.1109/TMI.2016.2521800
[14] Jodeiri, A., Zoroofi, R.A., Hiasa, Y., et al. (2020) Fully Automatic Estimation of Pelvic Sagittal Inclination from Anterior-Posterior Radiography Image Using Deep Learning Framework. Computer Methods and Programs in Biomedicine, 184, Article 105282.
https://doi.org/10.1016/j.cmpb.2019.105282
[15] Jang, S.J., Kunze, K.N., Vigdorchik, J.M., Jerabek, S.A., Mayman, D.J. and Sculco, P.K. (2022) John Charnley Award: Deep Learning Prediction of Hip Joint Center on Standard Pelvis Radiographs. The Journal of Arthroplasty, 37, S400-S407.e1.
https://doi.org/10.1016/j.arth.2022.03.033
[16] Thirion, T., Georis, P. and Gillet, P. (2019) [Preoperative Planning Interest of a Total Hip Prosthesis]. Revue Médicale de Liège, 74, 593-597.
[17] Chen, X., Liu, X., Wang, Y., et al. (2022) Development and Validation of an Artificial Intelligence Preoperative Planning System for Total Hip Arthroplasty. Frontiers in Medicine, 9, Article 841202.
https://doi.org/10.3389/fmed.2022.841202
[18] Huo, J., Huang, G., Han, D., et al. (2021) Value of 3D Preopera-tive Planning for Primary Total Hip Arthroplasty Based on Artificial Intelligence Technology. Journal of Orthopaedic Surgery and Research, 16, Article No. 156.
https://doi.org/10.1186/s13018-021-02294-9
[19] Ding, X., Zhang, B., Li, W., et al. (2021) Value of Preoperative Three-Dimensional Planning Software (AI-HIP) in Primary Total Hip Arthroplasty: A Retrospective Study. The Journal of International Medical Research, 49.
https://doi.org/10.1177/03000605211058874
[20] Mouille, B., Higuera, C., Woicehovich, L., et al. (2016) How to Succeed in Bundled Payments for Total Joint Replacement. Catalyst Carryover, 2, No. 5.
[21] Ramkumar, P.N., Navarro, S.M., Haeberle, H.S., et al. (2019) Development and Validation of a Machine Learning Algorithm after Primary Total Hip Arthroplasty: Applications to Length of Stay and Payment Models. The Journal of Arthroplasty, 34, 632-637.
https://doi.org/10.1016/j.arth.2018.12.030
[22] Lazic, I., Hinterwimmer, F., Langer, S., et al. (2022) Prediction of Complications and Surgery Duration in Primary Total Hip Arthroplasty Using Machine Learning: The Necessity of Modified Algorithms and Specific Data. Journal of Clinical Medicine, 11, Article 2147.
https://doi.org/10.3390/jcm11082147
[23] Radl, R., Aigner, C., Hungerford, M., Pascher, A. and Windhager, R. (2000) Proximal Femoral Bone Loss and Increased Rate of Fracture with a Proximally Hydroxyapatite-Coated Femoral Component. The Journal of Bone and Joint Surgery, 82, 1151-1155.
https://doi.org/10.1302/0301-620X.82B8.0821151
[24] Gargiulo, P., Edmunds, K.J., Gíslason, M.K., et al. (2018) Patient-Specific Mobility Assessment to Monitor Recovery after Total Hip Arthroplasty. Proceedings of the Institution of Mechanical Engineers Part H, Journal of Engineering in Medicine, 232, 1048-1059.
https://doi.org/10.1177/0954411918797971
[25] Ricciardi, C., Jónsson, H., Jacob, D., et al. (2020) Improving Prosthetic Selection and Predicting BMD from Biometric Measurements in Patients Receiving Total Hip Arthroplasty. Diagnostics (Basel, Switzerland), 10, Article 815.
https://doi.org/10.3390/diagnostics10100815
[26] Siljander, M.P., McQuivey, K.S., Fahs, A.M., Galasso, L.A., Serdahely, K.J. and Karadsheh, M.S. (2018) Current Trends in Patient-Reported Outcome Measures in Total Joint Arthroplasty: A Study of 4 Major Orthopaedic Journals. The Journal of Arthroplasty, 33, 3416-3421.
https://doi.org/10.1016/j.arth.2018.06.034
[27] Halawi, M.J., Jongbloed, W., Baron, S., Savoy, L., Cote, M.P. and Lieberman, J.R. (2020) Patient-Reported Outcome Measures Are Not a Valid Proxy for Patient Satisfaction in Total Joint Arthroplasty. The Journal of Arthroplasty, 35, 335-339.
https://doi.org/10.1016/j.arth.2019.09.033
[28] Mancuso, C.A., Salvati, E.A., Johanson, N.A., Peterson, M.G. and Charlson, M.E. (1997) Patients’ Expectations and Satisfaction with Total Hip Arthroplasty. The Journal of Arthroplasty, 12, 387-396.
https://doi.org/10.1016/S0883-5403(97)90194-7
[29] Anakwe, R.E., Jenkins, P.J. and Moran, M. (2011) Predict-ing Dissatisfaction after Total Hip Arthroplasty: A Study of 850 Patients. The Journal of Arthroplasty, 26, 209-213.
https://doi.org/10.1016/j.arth.2010.03.013
[30] Palazzo, C., Jourdan, C., Descamps, S., et al. (2014) Determinants of Satisfaction 1 Year after Total Hip Arthroplasty: The Role of Expectations Fulfilment. BMC Musculoskeletal Dis-orders, 15, Article No. 53.
https://doi.org/10.1186/1471-2474-15-53
[31] Fontana, M.A., Lyman, S., Sarker, G.K., Padgett, D.E. and Mac-Lean, C.H. (2019) Can Machine Learning Algorithms Predict Which Patients Will Achieve Minimally Clinically Im-portant Differences from Total Joint Arthroplasty? Clinical Orthopaedics and Related Research, 477, 1267-1279.
https://doi.org/10.1097/CORR.0000000000000687
[32] Huber, M., Kurz, C. and Leidl, R. (2019) Predicting Pa-tient-Reported Outcomes Following Hip and Knee Replacement Surgery Using Supervised Machine Learning. BMC Medical Informatics and Decision Making, 19, Article No. 3.
https://doi.org/10.1186/s12911-018-0731-6
[33] Zhang, S., Chen, J.Y., Pang, H.N., Lo, N.N., Yeo, S.J. and Liow, M.H.L. (2021) Development and Internal Validation of Machine Learning Algorithms to Predict Patient Satisfaction after Total Hip Arthroplasty. Arthroplasty (London, England), 3, Article No. 33.
https://doi.org/10.1186/s42836-021-00087-3
[34] Fillingham, Y.A., Ramkumar, D.B., Jevsevar, D.S., et al. (2018) The Safety of Tranexamic Acid in Total Joint Arthroplasty: A Direct Meta-Analysis. The Journal of Arthroplasty, 33, 3070-3082.e1.
https://doi.org/10.1016/j.arth.2018.03.031
[35] Bedard, N.A., Pugely, A.J., Lux, N.R., Liu, S.S., Gao, Y. and Callaghan, J.J. (2017) Recent Trends in Blood Utilization after Primary Hip and Knee Arthroplasty. The Journal of Arthroplasty, 32, 724-727.
https://doi.org/10.1016/j.arth.2016.09.026
[36] Cohen-Levy, W.B., Klemt, C., Tirumala, V., et al. (2023) Artificial Neural Networks for the Prediction of Transfusion Rates in Primary Total Hip Arthroplasty. Archives of Orthopaedic and Trauma Surgery, 143, 1643-1650.
https://doi.org/10.1007/s00402-022-04391-8
[37] Goesling, J., Moser, S.E., Zaidi, B., et al. (2016) Trends and Predictors of Opioid Use after Total Knee and Total Hip Arthroplasty. Pain, 157, 1259-1265.
https://doi.org/10.1097/j.pain.0000000000000516
[38] Inacio, M.C.S., Hansen, C., Pratt, N.L., Graves, S.E. and Roughead, E.E. (2016) Risk Factors for Persistent and New Chronic Opioid Use in Patients Undergoing Total Hip Arthroplasty: A Retrospective Cohort Study. BMJ Open, 6, e010664.
https://doi.org/10.1136/bmjopen-2015-010664
[39] Bedard, N.A., Pugely, A.J., Dowdle, S.B., Duchman, K.R., Glass, N.A. and Callaghan, J.J. (2017) Opioid Use Following Total Hip Arthroplasty: Trends and Risk Factors for Prolonged Use. The Journal of Arthroplasty, 32, 3675-3679.
https://doi.org/10.1016/j.arth.2017.08.010
[40] Bedard, N.A., Pugely, A.J., Westermann, R.W., Duchman, K.R., Glass, N.A. and Callaghan, J.J. (2017) Opioid Use after Total Knee Arthroplasty: Trends and Risk Factors for Prolonged Use. The Journal of Arthroplasty, 32, 2390-2394.
https://doi.org/10.1016/j.arth.2017.03.014
[41] Swami, A. and Jain, R. (2013) Scikit-Learn: Machine Learning in Python. Journal of Machine Learning Research, 12, 2825-2830.
[42] Gabriel, R.A., Harjai, B., Prasad, R.S., et al. (2022) Machine Learning Approach to Predicting Persistent Opioid Use Following Lower Extremity Joint Arthroplasty. Regional Anesthesia and Pain Medicine, 47, 313-319.
https://doi.org/10.1136/rapm-2021-103299
[43] Wilson, N.A., Jehn, M., York, S. and Davis, C.M. (2014) Revi-sion Total Hip and Knee Arthroplasty Implant Identification: Implications for Use of Unique Device Identification 2012 AAHKS Member Survey Results. The Journal of Arthroplasty, 29, 251-255.
https://doi.org/10.1016/j.arth.2013.06.027
[44] Gong, Z., Fu, Y., He, M. and Fu, X. (2022) Automated Identifica-tion of Hip Arthroplasty Implants Using Artificial Intelligence. Scientific Reports, 12, Article No. 12179.
https://doi.org/10.1038/s41598-022-16534-3
[45] Karnuta, J.M., Murphy, M.P., Luu, B.C., et al. (2023) Artificial Intelligence for Automated Implant Identification in Total Hip Arthroplasty: A Multicenter External Validation Study Exceeding Two Million Plain Radiographs. The Journal of Arthroplasty, 38, 1998-2003.e1.
https://doi.org/10.1016/j.arth.2022.03.002
[46] Bozic, K.J., Kurtz, S.M., Lau, E., Ong, K., Vail, T.P. and Berry, D.J. (2009) The Epidemiology of Revision Total Hip Arthroplasty in the United States. The Journal of Bone and Joint Surgery, 91, 128-133.
https://doi.org/10.2106/JBJS.H.00155
[47] Kurtz, S., Ong, K., Lau, E., Mowat, F. and Halpern, M. (2007) Pro-jections of Primary and Revision Hip and Knee Arthroplasty in the United States from 2005 to 2030. The Journal of Bone and Joint Surgery, 89, 780-785.
https://doi.org/10.2106/00004623-200704000-00012
[48] Trampuz, A. and Zimmerli, W. (2008) Diagnosis and Treatment of Implant-Associated Septic Arthritis and Osteomyelitis. Current Infectious Disease Reports, 10, 394-403.
https://doi.org/10.1007/s11908-008-0064-1
[49] Parvizi, J., Ghanem, E., Sharkey, P., Aggarwal, A., Burnett, R.S.J. and Barrack, R.L. (2008) Diagnosis of Infected Total Knee: Findings of a Multicenter Database. Clinical Orthopaedics and Related Research, 466, 2628-2633.
https://doi.org/10.1007/s11999-008-0471-5
[50] Zoccali, C., Teori, G. and Salducca, N. (2009) The Role of FDG-PET in Distinguishing between Septic and Aseptic Loosening in Hip Prosthesis: A Review of Literature. Interna-tional Orthopaedics, 33, 1-5.
https://doi.org/10.1007/s00264-008-0575-2
[51] Signore, A., Sconfienza, L.M., Borens, O., et al. (2019) Con-sensus Document for the Diagnosis of Prosthetic Joint Infections: A Joint Paper by the EANM, EBJIS, and ESR (with ESCMID Endorsement). European Journal of Nuclear Medicine and Molecular Imaging, 46, 971-988.
https://doi.org/10.1007/s00259-019-4263-9
[52] Loppini, M., Gambaro, F.M., Chiappetta, K., Grappiolo, G., Bianchi, A.M. and Corino, V.D.A. (2022) Automatic Identification of Failure in Hip Replacement: An Artificial Intel-ligence Approach. Bioengineering (Basel, Switzerland), 9, Article 288.
https://doi.org/10.3390/bioengineering9070288
[53] Shah, R.F., Bini, S.A., Martinez, A.M., Pedoia, V. and Vail, T.P. (2020) Incremental Inputs Improve the Automated Detection of Implant Loosening Using Machine-Learning Al-gorithms. The Bone & Joint Journal, 102-B, 101-106.
https://doi.org/10.1302/0301-620X.102B6.BJJ-2019-1577.R1
[54] Salassa, T., Hoeffel, D., Mehle, S., Tatman, P. and Gioe, T.J. (2014) Efficacy of Revision Surgery for the Dislocating Total Hip Arthroplasty: Report from a Large Community Registry. Clinical Orthopaedics and Related Research, 472, 962-967.
https://doi.org/10.1007/s11999-013-3344-5
[55] Borjali, A., Magnéli, M., Shin, D., Malchau, H., Muratoglu, O.K. and Varadarajan, K.M. (2021) Natural Language Processing with Deep Learning for Medical Adverse Event Detection from Free-Text Medical Narratives: A Case Study of Detecting Total Hip Replacement Dislocation. Computers in Biol-ogy and Medicine, 129, Article 104140.
https://doi.org/10.1016/j.compbiomed.2020.104140
[56] Rouzrokh, P., Ramazanian, T., Wyles, C.C., et al. (2021) Deep Learning Artificial Intelligence Model for Assessment of Hip Dislocation Risk Following Primary Total Hip Arthroplasty from Postoperative Radiographs. The Journal of Arthroplasty, 36, 2197-2203.e3.
https://doi.org/10.1016/j.arth.2021.02.028