Trunk acceleration for neuroprosthetic control of standing: a pilot study

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Postural instability is a significant challenge for individuals with paralysis.
• Existing neuroprosthetic control strategies often lack adaptability to dynamic disturbances.

Purpose of the Study:

• To investigate the feasibility of using trunk acceleration feedback for center of pressure (COP) control in a standing neuroprosthesis.
• To assess the efficacy of artificial neural networks (ANNs) in predicting COP changes based on trunk acceleration.
• To evaluate the impact of ANN-driven feedback control on postural stability.

Main Methods:

• Able-bodied subjects performed bipedal stance with induced postural perturbations.
• Three-dimensional trunk acceleration data were used to train ANNs to predict COP.
• An ANN model was implemented to control ankle muscle excitation in a computer simulation of a neuroprosthesis user.

Main Results:

• ANN predictions of COP showed moderate to strong correlations with actual COP (0.67–0.77).
• Feedback control significantly reduced average upper-body loading (42%) during perturbation onset and recovery.
• Peak upper-body loading was reduced by 29% compared to constant excitation control.

Conclusions:

• Trunk acceleration feedback control shows promise for enhancing postural stability in neuroprosthesis users.
• ANNs can effectively predict COP changes, enabling adaptive control strategies.
• This approach has the potential to improve functional recovery and reduce fall risk in individuals with paralysis.