Data-Driven Dynamic Motion Planning for Practical FES-Controlled Reaching Motions in Spinal Cord Injury

IEEE Trans Neural Syst Rehabil Eng, 2023 · DOI: 10.1109/TNSRE.2023.3272929 · Published: June 8, 2023

Simple Explanation

Functional electrical stimulation (FES) can help people with spinal cord injuries (SCI) regain arm movement. However, SCI can weaken muscles, making it hard to control FES-driven reaching motions. This study developed a new method that uses data about a person's specific muscle capabilities to plan feasible reaching trajectories, improving target reach and accuracy compared to simple reaching paths. The trajectory optimization method should be practically implemented to improve the FES-driven reaching performance.

Study Duration

Not specified

Participants

Single human participant with high tetraplegia

Evidence Level

Simulation study

Key Findings

1
Trajectory optimization improved the ability to reach targets compared to naive direct-to-target paths.
2
Trajectory optimization improved the accuracy for the feedforward-feedback and model predictive controllers (p < 0.001).
3
Even with an ideal model of the participant’s muscle capabilities and the dynamics of the system, trajectory optimization is necessary to avoid paths which include uncontrollable configurations and produce accurate reaches.

Research Summary

This study developed a data-driven trajectory optimization method for reaching motions that accounts for person-specific muscle weakness and loss of function in individuals with SCI. The method involves identifying a person-specific mathematical model of the arm, creating a dynamic simulation with modeled muscle capabilities, and developing a trajectory optimization routine. The performance of controlling the arm along optimized planned trajectories was compared to naive direct-to-target paths using feedback, feedforward-feedback, and model predictive control structures.

Practical Implications

Improved Reaching Performance

The trajectory optimization method can be implemented to improve FES-driven reaching performance in individuals with SCI.

Personalized Neuroprosthetics

The subject-specific, data-driven approach allows for personalized neuroprosthetics that account for individual muscle capabilities.

Advanced Control Strategies

More advanced controllers such as feedforward-feedback or MPC can be used to further improve performance by incorporating knowledge of the arm’s dynamics.

Study Limitations

1
The dynamic simulation does not include many nonlinearities and sources of uncertainty which exist in individuals with SCI.
2
Some of the found trajectories were on the edge of controllability, and even small deviations would lead to large errors.
3
Implementation of this controller for use in daily life would require improvements in both modeling accuracy and optimization time

Your Feedback

Was this summary helpful?

Back to Assistive Technology