Developing a Machine-Learning Predictive Model for Retention of Posterior Cruciate Ligament in Patients Undergoing Total Knee Arthroplasty

Abstract

Objective: Predicting whether the posterior cruciate ligament (PCL) should be preserved during total knee arthroplasty (TKA) procedures is a complex task in the preoperative phase. The choice to either retain or excise the PCL has a substantial effect on the surgical outcomes and biomechanical integrity of the knee joint after the operation. To enhance surgeons' ability to predict the removal and retention of the PCL in patients before TKA, we developed machine learning models. We also identified significant feature factors that contribute to accurate predictions during this process.

Methods: Patients' data on TKA continuously performed by a single surgeon who had intended initially to undergo implantation of cruciate-retaining (CR) prostheses was collected. During the sacrifice of PCL, we utilized anterior-stabilized (AS) tibial bearings. The dataset was split into CR and AS categories to form distinct groups. Relevant information regarding age, gender, body mass index (BMI), the affected side, and preoperative diagnosis was extracted by reviewing the medical records of the patients. To ensure the authenticity of the research, an initial step involved capturing X-ray images before the surgery. These images were then analyzed to determine the height of the medial condyle (MMH) and lateral condyle (LMH), as well as the ratios between MLW and MMH and MLW and LMH. Additionally, the insall-salvati index (ISI) was calculated, and the severity of any varus or valgus deformities was assessed. Eight machine-learning methods were developed to predict the retention of PCL in TKA. Risk factor analysis was performed using the SHApley Additive exPlanations method.

Results: A total of 307 knee joints from 266 patients were included, among which there were 254 females and 53 males. A stratified random sampling technique was used to split patients in a 70:30 ratio into a training dataset and a testing dataset. Eight machine-learning models were trained using data feeding. Except for the AUC of the LGBM Classifier, which is 0.70, the AUCs of other machine learning models are all lower than 0.70. In importance-based analysis, ISI, MMH, LMH, deformity, and age were confirmed as important predictive factors for PCL retention in operations.

Conclusion: The LGBM Classifier model achieved the best performance in predicting PCL retention in TKA. Among the potential risk factors, ISI, MMH, LMH, and deformity played essential roles in the prediction of PCL retention.

Keywords

Knee joint / Machine learning / Posterior cruciate ligament / Total knee arthroplasty

References

[1] NguyenL-CL, LehilMS, BozicKJ. Trends in total knee arthroplasty implant utilization. J Arthroplasty. 2015;30(5):739–742.

[2] MeiF, LiJ, ZhangL, Gao J, LiH, ZhouD, et al. Posterior-stabilized versus cruciate-retaining prostheses for total knee arthroplasty: an overview of systematic reviews and risk of bias considerations. Indian J Orthop. 2022;56(11):1858–1870.

[3] JiangC, LiuZ, WangY, Bian Y, FengB, WengX. Posterior cruciate ligament retention versus posterior stabilization for total knee arthroplasty: a meta-analysis. PLoS One. 2016;11(1):e0147865.

[4] Shoifi AbubakarM, Nakamura S, KuriyamaS, ItoH, Ishikawa M, FuruM, et al. Influence of posterior cruciate ligament tension on knee kinematics and kinetics. J Knee Surg. 2016;29(8):684–689.

[5] SongSJ, ParkCH, BaeDK. What to know for selecting cruciate-retaining or posterior-stabilized total knee arthroplasty. Clin Orthop Surg. 2019;11(2):142–150.

[6] ChurchillDL, IncavoSJ, JohnsonCC, Beynnon BD. The influence of femoral rollback on patellofemoral contact loads in total knee arthroplasty. J Arthroplasty. 2001;16(7):909–918.

[7] LongoUG, Ciuffreda M, ManneringN, D'AndreaV, LocherJ, SalvatoreG, et al. Outcomes of posterior-stabilized compared with cruciate-retaining total knee arthroplasty. J Knee Surg. 2018;31(4):321–340.

[8] LombardiAV, Mallory TH, FadaRA, HartmanJF, CappsSG, KefauverCA, et al. An algorithm for the posterior cruciate ligament in total knee arthroplasty. Clin Orthop Relat Res. 2001;392:75–87.

[9] RossiR, Cottino U, BruzzoneM, DettoniF, Bonasia DE, RossoF. Total knee arthroplasty in the varus knee: tips and tricks. Int Orthop. 2019;43(1):151–158.

[10] WangY, ZhangL, LinJ, XingD, LiuQ, ZhouD. Preoperative factors predicting the preservation of the posterior cruciate ligament in total knee arthroplasty. Orthop Surg. 2022;14(9):2203–2209.

[11] HuberNR, Missert AD, GongH, HsiehSS, LengS, YuL, et al. Random search as a neural network optimization strategy for convolutional-neural-network (CNN)-based noise reduction in CT. Proc SPIE - Int Soc Opt Eng. 2021;11596.

[12] LundbergSM, ErionG, ChenH, DeGrave A, PrutkinJM, NairB, et al. From local explanations to global understanding with explainable AI for trees. Nat Mach Intell. 2020;2(1):56–67.

[13] ChristodoulouE, MaJ, CollinsGS, Steyerberg EW, VerbakelJY, Van CalsterB. A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models. J Clin Epidemiol. 2019;110:12–22.

[14] AlzamzamiF, HodaM, El SaddikA. Light gradient boosting machine for general sentiment classification on short texts: a comparative evaluation. IEEE Access. 2020;8:101840–101858.

[15] ChristenB, Heesterbeek P, WymengaA, WehrliU. Posterior cruciate ligament balancing in total knee replacement: the quantitative relationship between tightness of the flexion gap and tibial translation. J Bone Joint Surg Br Vol. 2007;89(8):1046–1050.

[16] JacobsWCH, Clement DJ, WymengaAB. Retention versus removal of the posterior cruciate ligament in total knee replacement: a systematic literature review within the Cochrane framework. Acta Orthop. 2005;76(6):757–768.

[17] HatayamaK, Terauchi M, HashimotoS, SaitoK, Higuchi H. Factors associated with posterior cruciate ligament tightness during cruciate-retaining total knee arthroplasty. J Arthroplasty. 2018;33(5):1389–1393.

[18] Martinez-CanoJP, GobbiRG, GiglioPN, Arendt E, CostaGB, HinckelBB. Magnetic resonance imaging overestimates patellar height compared with radiographs. Knee Surg Sports Traumatol Arthrosc. 2022;30(10):3461–3469.

[19] VastaS, Andrade R, PereiraR, BastosR, Battaglia AG, PapaliaR, et al. Bone morphology and morphometry of the lateral femoral condyle is a risk factor for ACL injury. Knee Surg Sports Traumatol Arthrosc. 2018;26(9):2817–2825.

[20] LiG, ZhouC, ZhangZ, Foster T, BedairH. Articulation of the femoral condyle during knee flexion. J Biomech. 2022;131:110906.

[21] FengY, TsaiT-Y, LiJ-S, Wang S, HuH, ZhangC, et al. Motion of the femoral condyles in flexion and extension during a continuous lunge. J Orthop Res. 2015;33(4):591–597.

[22] MatziolisG, LoosM, BöhleS, Schwerdt C, RoehnerE, HeineckeM. Effect of additional distal femoral resection on flexion deformity in posterior-stabilized total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc. 2020;28(9):2924–2929.

RIGHTS & PERMISSIONS

2024 2024 The Authors. Orthopaedic Surgery published by Tianjin Hospital and John Wiley & Sons Australia, Ltd.