Methods for integrating postural control into biomechanical human simulations: a systematic review.

Area of Science:

• Biomechanics
• Neuroscience
• Robotics

Background:

• Human postural control relies on integrating sensory information for balance maintenance.
• Forward dynamic simulations of biomechanical models are crucial for predicting motion and understanding muscle forces.
• Predictive postural control simulations hold potential for developing future therapy recommendations for patients with postural disorders.

Approach:

• Conducted a systematic literature research on Scopus, Web of Science, and PubMed.
• Analyzed 102 eligible studies focusing on biomechanical models, sensory representation, integration, and control methods.
• Discussed various modeling approaches for simulating human postural control in standing and gait.

Key Points:

• Diverse methods exist for sensory representation and integration in postural control models, though neural underpinnings remain unclear.
• Increasing model complexity and nonlinearities reduce the use of optimal control methods like linear quadratic regulators and model predictive control.
• Reflex-based and proportional-derivative (PD) controllers are prevalent in musculoskeletal models, yielding promising human-like motion.

Conclusions:

• A comprehensive understanding of existing postural control models is essential for advancing predictive simulation capabilities.
• Current models show promise in replicating human-like motion, particularly when incorporating physiological processes.
• Further research into the neural basis of sensory integration is needed to refine postural control simulations.