Neuro-musculoskeletal simulation of instrumented contracture and spasticity assessment in children with cerebral palsy

• Published in: Journal of neuroengineering and rehabilitation Vol. 13; no. 1; p. 64
• Main Authors: van der Krogt, Marjolein Margaretha, Bar-On, Lynn, Kindt, Thalia, Desloovere, Kaat, Harlaar, Jaap
• Format: Journal Article
• Language: English
• Published: London BioMed Central 16.07.2016; BioMed Central Ltd
• Subjects: Adolescent; Biomechanics; Biomedical and Life Sciences; Biomedical Engineering and Bioengineering; Biomedicine; Cerebral palsy; Cerebral Palsy - complications; Cerebral Palsy - physiopathology; Child; Complications and side effects; Computer Simulation; Contracture - physiopathology; Electromyography; Hamstring Muscles - physiopathology; Humans; Male; Muscle Spasticity - etiology; Muscle Spasticity - physiopathology; Muscle, Skeletal - physiopathology; Neurology; Neurosciences; Reflex, Stretch - physiology; Rehabilitation Medicine; Risk factors; Spasticity; Belgium; Biomechanics; Cerebral palsy; Muscle stiffness; Neuro-musculoskeletal modeling; Electromyography; Rehabilitation; Muscle spasticity
• ISSN: 1743-0003; 1743-0003
• DOI: 10.1186/s12984-016-0170-5

Background Increased resistance in muscles and joints is an important phenomenon in patients with cerebral palsy (CP), and is caused by a combination of neural (e.g. spasticity) and non-neural (e.g. contracture) components. The aim of this study was to simulate instrumented, clinical assessment of the hamstring muscles in CP using a conceptual model of contracture and spasticity, and to determine to what extent contracture can be explained by altered passive muscle stiffness, and spasticity by (purely) velocity-dependent stretch reflex. Methods Instrumented hamstrings spasticity assessment was performed on 11 children with CP and 9 typically developing children. In this test, the knee was passively stretched at slow and fast speed, and knee angle, applied forces and EMG were measured. A dedicated OpenSim model was created with motion and muscles around the knee only. Contracture was modeled by optimizing the passive muscle stiffness parameters of vasti and hamstrings, based on slow stretch data. Spasticity was modeled using a velocity-dependent feedback controller, with threshold values derived from experimental data and gain values optimized for individual subjects. Forward dynamic simulations were performed to predict muscle behavior during slow and fast passive stretches. Results Both slow and fast stretch data could be successfully simulated by including subject-specific levels of contracture and, for CP fast stretches, spasticity. The RMS errors of predicted knee motion in CP were 1.1 ± 0.9° for slow and 5.9 ± 2.1° for fast stretches. CP hamstrings were found to be stiffer compared with TD, and both hamstrings and vasti were more compliant than the original generic model, except for the CP hamstrings. The purely velocity-dependent spasticity model could predict response during fast passive stretch in terms of predicted knee angle, muscle activity, and fiber length and velocity. Only sustained muscle activity, independent of velocity, was not predicted by our model. Conclusion The presented individually tunable, conceptual model for contracture and spasticity could explain most of the hamstring muscle behavior during slow and fast passive stretch. Future research should attempt to apply the model to study the effects of spasticity and contracture during dynamic tasks such as gait.