SpasticMyoElbow: Physical Human-Robot Interaction Simulation Framework for Modelling Elbow Spasticity.

Area of Science:

• Rehabilitation Engineering
• Neuroscience
• Robotics

Background:

• Robotic devices offer potential for spasticity assessment in post-stroke patients.
• Limited patient data hinders quantitative modeling and validation of robotic spasticity assessment.
• Current clinical spasticity assessment remains manual.

Purpose of the Study:

• To present a simulation framework for designing and validating elbow spasticity models.
• To address data limitations by enabling simulation with synthetic data.
• To support the development of robotic assessment techniques for spasticity.

Main Methods:

• Developed a simulation framework including a robot-assisted assessment environment, motion controllers, and a stretch reflex controller.
• Utilized synthetic data for simulations, bypassing the need for experimental patient data.
• Replicated constant-velocity stretch experiments and evaluated four spasticity models.

Main Results:

• A spasticity reflex model incorporating muscle fibre velocity and length feedback demonstrated superior accuracy in capturing joint resistance during passive elbow stretching.
• This model outperformed a force-dependent model in simulating spastic patient resistance characteristics.
• The simulation framework successfully generated synthetic data for evaluating spasticity models.

Conclusions:

• The developed simulation framework can generate extensive virtual patient datasets for spasticity research.
• An advanced spasticity reflex model provides a more accurate representation of joint resistance in spastic patients.
• This approach mitigates data scarcity issues, paving the way for robust robotic spasticity assessment validation.