Giorgio Davico1, Claudio Pizzolato2, David G Lloyd2, Steven J Obst3, Henry P J Walsh4, Christopher P Carty5. 1. School of Allied Health Sciences, Griffith Centre of Biomedical and Rehabilitation Engineering (GCORE), Menzies Health Institute Queensland, Griffith University, Gold Coast, Australia. Electronic address: giorgio.davico@unibo.it. 2. School of Allied Health Sciences, Griffith Centre of Biomedical and Rehabilitation Engineering (GCORE), Menzies Health Institute Queensland, Griffith University, Gold Coast, Australia. 3. School of Health, Medical and Applied Sciences, Central Queensland University, Bundaberg, Australia. 4. Department of Orthopaedics, Children's Health Queensland Hospital and Health Services, Brisbane, Australia; Queensland Children's Motion Analysis Service, Queensland Paediatric Rehabilitation Service, Children's Health Queensland Hospital and Health Services, Brisbane, Australia. 5. School of Allied Health Sciences, Griffith Centre of Biomedical and Rehabilitation Engineering (GCORE), Menzies Health Institute Queensland, Griffith University, Gold Coast, Australia; Department of Orthopaedics, Children's Health Queensland Hospital and Health Services, Brisbane, Australia; Queensland Children's Motion Analysis Service, Queensland Paediatric Rehabilitation Service, Children's Health Queensland Hospital and Health Services, Brisbane, Australia.
Abstract
BACKGROUND: Cerebral palsy is a complex neuromuscular disorder that affects the sufferers in multiple different ways. Neuromusculoskeletal models are promising tools that can be used to plan patient-specific treatments for cerebral palsy. However, current neuromusculoskeletal models are typically scaled from generic adult templates that poorly represent paediatric populations. Furthermore, muscle activations are commonly computed via optimisation methods, which may not reproduce co-contraction observed in cerebral palsy. Alternatively, calibrated EMG-informed approaches within OpenSim can capture pathology-related muscle activation abnormalities, possibly enabling more feasible estimations of muscle and joint contact forces. METHODS: Two identical twin brothers, aged 13, one with unilateral cerebral palsy and the other typically developing, were enrolled in the study. Four neuromusculoskeletal models with increasing subject-specificity were built in OpenSim and CEINMS combining literature findings, experimental motion capture, EMG and MR data for both participants. The physiological and biomechanical validity of each model was assessed by quantifying its ability to track experimental joint moments and muscle excitations. FINDINGS: All developed models accurately tracked external joint moments; however EMG-informed models better tracked muscle excitations compared to neural solutions generated by static optimisation. Calibrating muscle-tendon unit parameters with EMG data allowed for more physiologically plausible joint contact forces estimates. Further scaling the maximal isometric force of muscles with MR-derived muscle volumes did not affect model predictions. INTERPRETATION: Given their ability to identify atypical joint contact forces profiles and accurately reproduce experimental data, calibrated EMG-informed models should be preferred over generic models using optimisation methods in informing the management of cerebral palsy.
BACKGROUND:Cerebral palsy is a complex neuromuscular disorder that affects the sufferers in multiple different ways. Neuromusculoskeletal models are promising tools that can be used to plan patient-specific treatments for cerebral palsy. However, current neuromusculoskeletal models are typically scaled from generic adult templates that poorly represent paediatric populations. Furthermore, muscle activations are commonly computed via optimisation methods, which may not reproduce co-contraction observed in cerebral palsy. Alternatively, calibrated EMG-informed approaches within OpenSim can capture pathology-related muscle activation abnormalities, possibly enabling more feasible estimations of muscle and joint contact forces. METHODS: Two identical twin brothers, aged 13, one with unilateral cerebral palsy and the other typically developing, were enrolled in the study. Four neuromusculoskeletal models with increasing subject-specificity were built in OpenSim and CEINMS combining literature findings, experimental motion capture, EMG and MR data for both participants. The physiological and biomechanical validity of each model was assessed by quantifying its ability to track experimental joint moments and muscle excitations. FINDINGS: All developed models accurately tracked external joint moments; however EMG-informed models better tracked muscle excitations compared to neural solutions generated by static optimisation. Calibrating muscle-tendon unit parameters with EMG data allowed for more physiologically plausible joint contact forces estimates. Further scaling the maximal isometric force of muscles with MR-derived muscle volumes did not affect model predictions. INTERPRETATION: Given their ability to identify atypical joint contact forces profiles and accurately reproduce experimental data, calibrated EMG-informed models should be preferred over generic models using optimisation methods in informing the management of cerebral palsy.
Authors: Mohammad Fazle Rabbi; Laura E Diamond; Chris P Carty; David G Lloyd; Giorgio Davico; Claudio Pizzolato Journal: Sci Rep Date: 2022-03-04 Impact factor: 4.379