The effect of musculoskeletal model scaling methods on ankle joint kinematics and muscle force prediction during gait for children with cerebral palsy and equinus gait

• Published in: Computers in biology and medicine 2021-07, Vol.134, p.104436-104436, Article 104436
• Main Authors: Ma, Yunru, Jiang, Shuyun, Mithraratne, Kumar, Wilson, Nichola, Yu, Yan, Zhang, Yanxin
• Format: Article
• Language: English
• Subjects: Ankle; Cerebral palsy; Children; Electromyography; EMG-Driven; Force; Gait; Gastrocnemius muscle; Joints (anatomy); Kinematics; Mathematical models; Methods; Motion capture; Muscles; Optimization; Paralysis; Patients; Predictions; Scaling; Skeletal muscle; Skin; Software; Static optimisation; Tibialis anterior muscle
• ISSN: 0010-4825; 1879-0534
• DOI: 10.1016/j.compbiomed.2021.104436

Clinical gait analysis incorporated with neuromusculoskeletal modelling could provide valuable information about joint movements and muscle functions during ambulation for children with cerebral palsy (CP). This study investigated how imposing pre-calculated joint angles during musculoskeletal model scaling influence the ankle joint angle and muscle force computation. Ten children with CP and equinus gait underwent clinical gait analysis. For each participant, a “default” (scaled without pre-calculated joint angles) and a “PJA” (scaled with pre-calculated ankle joint angles) model were generated to simulate their gait. Ankle joint angles were calculated with an inverse kinematic (IK) and direct kinematic (DK) approach. Triceps surae and tibialis anterior muscle forces were predicted by static optimisation and EMG-assisted modelling. We found that PJA-derived ankle angles showed a better agreement with what derived from the DK approach. The tibialis anterior muscle prediction was more likely to be affected by the scaling methods for the static optimisation approach and the gastrocnemius muscle force prediction was more likely to be influenced for the EMG-assisted modelling. This study recommends using the PJA model since the good consistency between IK and DK-derived joint angles facilitates communication among different research disciplines. •Musculoskeletal model scaling methods affect joint angle estimation.•Both musculoskeletal model scaling and muscle force modelling methods affect muscle force prediction.•Scaling methods influence tibialis anterior muscle force prediction for the static optimisation approach.•Scaling methods influence gastrocnemius muscle force prediction for the EMG-assisted modelling.•Scaling with pre-calculated joint angles is recommended.