Evaluation of the accuracy of musculoskeletal simulation during squats by means of instrumented knee prostheses.

Standard musculoskeletal simulation tools now offer widespread access to internal loading conditions for use in improving rehabilitation concepts or training programmes. However, despite broad reliance on their outcome, the accuracy of such loading estimations, specifically in deep knee flexion, remains generally unknown. The aim of this study was to evaluate the error of tibio-femoral joint contact force (JCF) calculations using musculoskeletal simulation compared to in vivo measured JCFs in subjects with instrumented total knee endoprostheses during squat exercises. Using the early but common "Gait2392_simbody" (OpenSim) scaled musculoskeletal models, tibio-femoral JCFs were calculated in 6 subjects for 5 repetitions of squats. Tibio-femoral JCFs of 0.8-3.2 times bodyweight (BW) were measured. While the musculoskeletal simulations underestimated the measured knee JCFs at low flexion angles, an average error of less than 20% was achieved between approximately 25°-60° knee flexion. With an average error that behaved almost linearly with knee flexion angle, an overestimation of approximately 60% was observed at deep flexion (ca. 80°), with an absolute maximum error of ca. 1.9BW. Our data indicate that loading estimations from early musculoskeletal gait models at both high and low knee joint flexion angles should be interpreted carefully.