Influence of musculoskeletal model implemented in the AnyBody Modeling System on calculated loadings acting in the lower extremity during gait.

PURPOSE: An evaluation of the influence of different musculoskeletal models implemented in the AnyBody Modeling System on muscle forces and joint reaction forces in the lower limb, calculated as the result of numerical simulation of gait, was the main aim of the paper, using identical scaling and optimisation procedures. The study investigated whether the models predict significant differences in muscle forces and joint reaction forces.

METHODS

Kinematic parameters and ground reaction forces were recorded for 20 healthy adults of both sexes (21-24 years), walking at a self-selected speed and used as inputs for musculoskeletal modelling and simulation in AnyBody. Two musculoskeletal models (Twente Lower Extremity Model [TLEM]) available in the AnyBody repository, TLEM 1.1 and TLEM 2.0, were used. Time waveforms of force generated by 20 muscles and hip, knee and ankle joint reaction forces were obtained. Statistical parametric mapping and root-mean-square differences were used to evaluate differences between the two models.

RESULTS

Significant differences between models were observed in both muscle force and joint reaction force waveforms. The greatest differences were visible in pelvic stabilisers, hip flexors and muscles responsible for a push-off and lower limb stabilisation, as well as in the medio-lateral components of joint reaction forces.

CONCLUSIONS

The results confirm that the geometrical model of the musculoskeletal system has a significant influence on the calculated muscle and joint reaction forces. Recognising these model-dependent differences is essential for the correct interpretation of simulation results and the informed use of musculoskeletal modelling in future studies.