Probabilistic finite element predictions of the human lower limb model in total knee replacement

Abstract

The purpose of this paper is to explore both an extended and a reduced set of input parameters of the Finite Element (FE) model of the human lower limb with a Total Knee Replacement (TKR) implant. The most influential parameters in determining the size and the shape of the performance envelopes of eight kinematics and peak contact pressure output variables of the tibio-femoral joint and the patello-femoral joint are sought. The lower limb FE model, which includes bones, TKR implant, soft tissues and applied forces of realistic size, is used in the context of the stair ascent simulation. Two probabilistic methods are used together with the FE model to generate the performance envelopes and to explore the sensitivities of the input parameters of the FE model: the Monte Carlo simulation and the Response Surface Method (RSM). A total of four probabilistic FE analyses assess how the uncertainties in an extended set of 77 input variables and a reduced set of 22 input variables obtained from the RSM/sensitivity analyses affect the performance envelopes. It is shown that the FE model with the reduced set of variables is able to replicate the full FE model. The differences between the Monte Carlo envelopes of performance obtained with the FE model with the full set of variables and the FE model with the reduced set of variables were on average over all output measures under 1.67mm for translations, 1.75° for rotations and under 2MPa for peak contact pressures. The differences between the RSM and the Monte Carlo envelopes of performances obtained with the reduced set of input variables were, on average, over all output measures under 0.75mm for translations, 1.26° for rotations and 2.39MPa for peak contact pressures. While saving computational time with the reduced set of variables, the findings are especially of high importance to the orthopedic surgeons who would like to know the most important parameters that can influence the performance of the TKR for a given human activity.