Dataset Information

Machine Learning Predicts Femoral and Tibial Implant Size Mismatch for Total Knee Arthroplasty.

ABSTRACT:

Background

Despite reasonable accuracy with preoperative templating, the search for an optimal planning tool remains an unsolved dilemma. The purpose of the present study was to apply machine learning (ML) using preoperative demographic variables to predict mismatch between templating and final component size in primary total knee arthroplasty (TKA) cases.

Methods

This was a retrospective case-control study of primary TKA patients between September 2012 and April 2018. The primary outcome was mismatch between the templated and final implanted component sizes extracted from the operative database. The secondary outcome was mismatch categorized as undersized and oversized. Five supervised ML algorithms were trained using 6 demographic features. Prediction accuracies were obtained as a metric of performance for binary mismatch (yes/no) and multilevel (undersized/correct/oversized) classifications.

Results

A total of 1801 patients were included. For binary classification, the best-performing algorithm for predicting femoral and tibial mismatch was the stochastic gradient boosting model (area under the curve: 0.76/0.72, calibration intercepts: 0.05/0.05, calibration slopes: 0.55/0.7, and Brier scores: 0.20/0.21). For multiclass classification, the best-performing algorithms had accuracies of 83.9% and 82.9% for predicting the concordance/mismatch of the femoral and tibial implant, respectively. Model predictions of greater than 51.0% and 47.9% represented high-risk thresholds for femoral and tibial sizing mismatch, respectively.

Conclusions

ML algorithms predicted templating mismatch with good accuracy. External validation is necessary to confirm the performance and reliability of these algorithms. Predicting sizing mismatch is the first step in using ML to aid in the prediction of final TKA component sizes. Further studies to optimize parameters and predictions for the algorithms are ongoing.

SUBMITTER: Polce EM

PROVIDER: S-EPMC8167319 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundAseptic loosening is one of the major reasons for late revision in total knee arthroplasty (TKA). The risk of aseptic loosening can be detected using radiostereometric analysis (RSA), whereby micromovements (migration) can be measured, and thus RSA is recommended in the phased introduction of orthopedic implants. Decrease in bone mineral density (BMD), as measured by dual-energy x ray absorptiometry (DXA), is related to the breaking strength of the bone, which is measured concurrently by RSA. The aim of the study was to evaluate bone remodeling and implant migration with cemented asymmetrical tibial and uncemented femoral components after TKA with a follow up period of 2 years.MethodsThis was a prospective longitudinal cohort study of 29 patients (number of female/male patients 17/12, mean age 65.2 years), received a hybrid Persona® TKA (Zimmer Biomet, Warsaw, IN, USA) consisting of a cemented tibial, an all-polyethylene patella, and uncemented trabecular metal femoral components. Follow up: preoperative, 1 week, and 3, 6, 12 and 24 months after surgery, and double examinations for RSA and DXA were performed at 12 months. RSA results were presented as maximal total point of motion (MTPM) and segmental motion (translation and rotation), and DXA results were presented as changes in BMD in different regions of interest (ROI).ResultsMTPM at 3, 6, 12, and 24 months was 0.65 mm, 0.84 mm, 0.92 mm, and 0.96 mm for the femoral component and 0.54 mm, 0.60 mm, 0.64 mm, and 0.68 mm, respectively, for the tibial component. The highest MTPM occurred within the first 3 months. Afterwards most of the curves flattened and stabilized. Between 12 and 24 months after surgery, 16% of femoral components had migrated by more than 0.10 mm and 15% of tibial components had migrated by more than 0.2 mm. Percentage change in BMD in each ROI for distal femur was as follows: ROI I 26.7%, ROI II 9.2% and ROI III 3.3%. BMD and at the proximal tibia: ROI I 8.2%, ROI II 8.6% and ROI III 7.0% after 2 years compared with 1 week postoperative results. There was no significant correlation between maximal percentwise change in BMD and MTPM after 2 years.ConclusionMigration patterns and changes in BMD related to femoral components after TKA in our study correspond well with previous studies; we observed marginally greater migration with the tibial component.

Project description:BackgroundIt is unclear whether there is a clinical benefit to adding hydroxyapatite (HA) coatings to total knee implants, especially with the tibial component, where failure of the implant more often occurs. A systematic review of the literature was undertaken to identify all prospective randomized trials for determining whether the overall clinical results (as a function of durability, function, and adverse events) favored HA-coated tibial components.MethodsA comprehensive literature search was performed for the years 1990 to September 16, 2010. We restricted our search to randomized controlled trials involving participants receiving either an HA-coated tibia or other forms of tibial fixation. The primary outcome measures evaluated were durability, function, and acute adverse events.ResultsData from 926 evaluable primary total knee implants in 14 studies were analyzed. Using an RSA definition for durability, HA-coated tibial components (porous or press-fit) without screw fixation were less likely to be unstable at 2 years than porous and cemented metal-backed tibial components (RR = 0.58, 95% CI: 0.34-0.98; p = 0.04, I(2) = 39%, M-H random effects model). There was no significant difference in durability, as measured from revision and evaluated at 2 and 8-10 years, between groups. Also, functional status using different validated measures showed no significant difference at 2 and 5 years, no matter what measure was used. Lastly, there was no significant difference in adverse events. Limitations included small numbers of evaluable patients (≤ 50) in 7 of the 14 trials identified, and a lack of "hard" evidence of durability with need for replacement (i.e. frank failure, pain, or loss of functionality).InterpretationIn patients > 65 years of age, an HA-coated tibial implant may provide better durability than other forms of tibial fixation. Larger trials should be undertaken comparing the long-term durability, function, and adverse events of HA-coated implants with those of other porous-coated tibial implants in younger, more active OA patients.

Project description:PurposeThe anterior flange height of the current femoral component increases with an increasing distal femoral anteroposterior dimension. During total knee arthroplasty (TKA), we have observed that a large femur may have a thinner anterior condyle, whereas a small femur may have a thicker anterior condyle. The first purpose of this study was to examine whether the femoral anterior condyle height decreases as the distal femoral anteroposterior size increases and whether gender differences exist in anterior condyle height.MethodsA total of 1218 knees undergoing TKA intraoperative and computed tomography scans from 303 healthy knees were used to measure the anterior lateral condylar height (ALCH), anterior medial condylar height (AMCH), and the lateral anteroposterior (LAP) and medial anteroposterior (MAP) dimensions of distal femurs. The LAP and MAP measurements were used for adjustments to determine whether gender differences exist in anterior condyle heights. Linear regression analysis was performed to determine correlations between ALCH and LAP or between AMCH and MAP.ResultsThere were significant differences between males and females in ALCH in both the CT and TKA groups and AMCH in the CT group (all P<0.01). After adjusting for LAP and MAP, there were significant gender differences in the lateral and medial condylar heights in both groups (P<0.01). There were significant negative correlations between ALCH and LAP values and between AMCH and MAP values in both CT and TKA measurements, with the LAP and MAP values increasing as ALCH and AMCH decreased.ConclusionsThe results demonstrate that femoral anterior condylar height decreased with increasing anteroposterior dimension in both the medial and lateral condyle. In addition, this study also showed that anterior condylar heights are highly variable, with gender differences. The data may provide an important reference for designing femoral anterior flange thickness to precisely match the natural anterior condylar anatomy.