Model-Based Estimation of Ankle Joint Stiffness.

Area of Science:

• Biomechanics
• Wearable Technology
• Biomedical Engineering

Background:

• Estimating biomechanical parameters like joint stiffness is crucial for understanding human movement.
• Wearable sensors offer a promising, non-invasive approach for in-situ biomechanical assessments.
• Existing models often struggle with the nonlinear dynamics of biological systems.

Purpose of the Study:

• To develop and validate a novel method for estimating sagittal plane ankle joint stiffness using wearable sensors.
• To create a nonlinear biomechanical model of the lower leg driven by electromyographic (EMG) signals.
• To implement a robust filtering algorithm capable of handling model uncertainties and system nonlinearities.

Main Methods:

• Formulated a novel nonlinear biomechanical model of the lower leg incorporating 2D kinematics for muscle lever arm and torque calculations.
• Developed a square-root cubature Kalman filter to address the model's nonlinearities and nonsmooth dynamics.
• Validated the estimation approach through in silico simulations and an experimental study using body-worn sensors and a dedicated single-joint test-bench.

Main Results:

• The developed filter successfully reconstructed joint angle positions, velocities, and torques.
• Accurate estimation of ankle joint stiffness was achieved during experimental movements.
• The system demonstrated effectiveness in reducing estimation errors caused by model uncertainties.

Conclusions:

• The novel biomechanical model and square-root cubature Kalman filter provide an effective solution for estimating ankle joint stiffness.
• Wearable measurement technologies, combined with advanced filtering algorithms, show significant potential for in-situ biomechanical analysis.
• This approach can advance the understanding and assessment of musculoskeletal health and performance.