Can TKA outcomes be predicted with computational simulation? Generation of a patient specific planning tool

Abstract

Background: Computer simulations of knee movement allow Total Knee Arthroplasty (TKA) dynamic outcomes to be studied. This study aims to build a model predicting patient reported outcome from a simulation of post-operative TKA joint dynamics. Methods: Landmark localisation was performed on 239 segmented pre-operative computerized tomography (CT) scans to capture patient specific soft tissue attachments. The preoperative bones and 3D implant files were registered to post-operative CT scans following TKA. Each post-operative knee was simulated undergoing a deep knee bend with assumed ligament balancing of the extension space. The kinematic results from this simulation were used in a Multivariate Adaptive Regression Spline algorithm, predicting attainment of a Patient Acceptable Symptom State (PASS) score in captured 12 month post-operative Knee Injury and Osteoarthritis Outcome Scores (KOOS). An independent series of 250 patients was evaluated by the predictive model to assess how the predictive model behaved in a pre-operative planning context. Results: The generated predictive algorithm, called the Dynamic Knee Score (DKS) contained features, in order of significance, related to tibio-femoral force, patello-femoral motion and tibio-femoral motion. Area Under the Curve for predicting attainment of the PASS KOOS Score was 0.64. The predictive model produced a bimodal spread of predictions, reflecting a tendency to either strongly prefer one alignment plan over another or be ambivalent. Conclusion:A predictive algorithm relating patient reported outcome to the outputs of a computational simulation of a deep knee bend has been derived (the DKS). Simulation outcomes related to tibio-femoral balance had the highest correlation with patient reported outcome.

Can minimal clinically important differences in patient reported outcome measures be predicted by machine learning in patients with total knee or hip arthroplasty? A systematic review

“…One study [ 48 ] included knee arthroscopies and one [ 49 ] patients with osteochondral allograft for cartilage defects. Next, five studies [ 50 – 54 ] did not calculate MCIDs, even though they researched PROMs in TJA patients. Finally, six studies [ 26 , 55 – 58 ] aimed to make comparable predictions as searched for this review except failing to apply machine learning.…”

Patient-reported impairment following TKA is reduced when a computationally simulated predicted ideal alignment is achieved

“…The simulation replicates a deep knee bend performed in an Oxford Knee Rig (OKR), and includes modelled collateral ligaments, a quadriceps tendon and other passive soft-tissue restraints. All ligaments were modelled as one or two bundles of nonlinear springs as described by Abdul-Rahman et al [ 1 ] with fixed parameters further adapted using a process previously described by Theodore et al [ 22 , 23 , 26 ]. In this way, the model captures the combination of patient-specific elements, component geometry and component position and orientation that contribute to the dynamic joint motion.…”

“…Then, a set of alternate surgical plans the surgeon might have followed were simulated. From each simulation, the DKS score is generated, giving a score from 0 to 100 that predicts the probability of a Patient Acceptable Symptom State (PASS) being reached, using cut off values [ 12 ] and via a process described in a previous publication [ 23 ]. Patients were grouped by whether (a) the surgically achieved position scored the highest DKS of the variants or (b) whether an alternative simulation scored higher, which would indicate an alternate approach to the surgery was predicted to be more likely to reach a PASS outcome (Fig.…”

“…In previous work, the authors have described a computational simulation of patient-specific knee motion undertaking a deep knee bend [ 35 ] that has been previously mechanically validated [ 22 ]. This simulation has further had a Machine Learning (ML) model of patient outcome generated [ 23 , 26 ], called the dynamic knee score (DKS). ML models in general are a relatively underexplored [ 14 ] tool in TKA, but one with many potentially valuable applications being explored [ 13 ].…”

Active Neurodynamics at Home in Patients with Knee Osteoarthritis: A Feasibility Study

“…For this reason, it could be interesting in the future to include computational simulation studies, which reduce costs and provide results more quickly [53][54][55]. Computational prediction models have been designed for total knee arthroplasty to evaluate biomechanical effects [56] and patient-reported outcomes, including KOOS [57]. This simulation decreases the likelihood of impairments after total knee arthroplasty and may be useful in planning [58].…”