MUSCULOSKELETAL MODELLING FOR KNEE CONTACT FORCE DURING GAIT AFTER KNEE REPLACEMENT

The purpose of this study was to evaluate different musculoskeletal models, mainly focus on the number of muscles and DOF of knee joint, for estimated knee contact force (KCF) during gait. Four different OpenSim models: Gait2354, Gait2392, Gait2354 and Gait2392 with 3 DOFs of knee joint, were evaluated using gait data from four patients with unilateral knee instrumented implants. The predicted results of muscle forces and peak KCFs from OpenSim simulation were compared with EMG data and in vivo KCF data separately. The results indicated that the correlations between predicted muscle forces and EMG data were relatively high for Gait2354 and Gait2392. The musculoskeletal models generally overestimated the peak KCFs compared to the in vivo measurement with the minimum differences in Gait2392. Therefore, this study suggests increasing the number of muscles in musculoskeletal model could improve the simulation results in some extent, but adding additional DOFs of knee joint without ligament structures worse the predicted KCF results. Gait2392 is a preferred model among the four tested models for studying muscle forces and KCF after total knee replacement. Additional subject-specific attributes are need to improve the ability of musculoskeletal model to predict KCF.