Effect of the underlying cadaver data and patient-specific adaptation of the femur and pelvis on the prediction of the hip joint force estimated using static models

The prediction of the hip joint force (HJF) is a fundamental factor for the prevention of edge loading in total hip arthroplasty. Naturally, the loading of the liner of the acetabular component depends on the HJF acting on the artificial joint. In contrast to dynamic musculoskeletal models, static models for HJF prediction do not require motion analysis of the patient. However, patient-specific adaptability and validity of static models have to be scrutinized. In this study, a modular framework for HJF prediction using static models is introduced to compare the results of different cadaver templates that are the basis of most static and dynamic models, and different scaling laws for the patient-specific adaptation with in vivo HJF of ten patients for one-leg stance and level walking. The results revealed the significant effect of the underlying cadaver template used for the prediction of the HJF (p < 0.01). A higher degree of patient-specific scaling of the cadaver template often did not significantly reduce the prediction error. Three static models with the lowest prediction errors were compared to results of dynamic models from literature. The prediction error of the peak HJF of the static models (median absolute errors below 15% body weight in magnitude and below 5 degrees in direction) was similar in magnitude and even smaller in direction compared to dynamic models. The necessary reduction of a load-based target zone for the prevention of edge loading due to the uncertainty of the HJF prediction has to be considered in the preoperative planning. The framework for HJF prediction is openly accessible at https://github.com/RWTHmediTEC/HipJointForceModel. (c) 2021 Elsevier Ltd. All rights reserved.

Automated summary

The prediction of hip joint force (HJF) is crucial for preventing edge loading in total hip arthroplasty. This study introduces a modular framework for predicting HJF using static models and compares it with different cadaver templates and scaling laws for patient-specific adaptation. The results show that the prediction error of the static models is similar to or even smaller than that of dynamic models.