Review of control strategies for robotic movement training after neurologic injury

Journal of NeuroEngineering and Rehabilitation, 2009 · DOI: 10.1186/1743-0003-6-20 · Published: June 16, 2009

Assistive Technology Neurology Neurorehabilitation

Simple Explanation

Robotic devices are increasingly used to help people recover movement after neurologic injuries like stroke and spinal cord injury. These devices physically interact with the participant's limbs during movement training, or 'coach' the participant without physical contact. Control strategies for these devices fall into categories like assisting, challenge-based, haptic simulation (practicing daily tasks in a virtual environment), and non-contact coaching. The goal is to provoke motor plasticity and improve motor recovery. While much work has focused on developing assistive strategies, it's important to consider that too much assistance could be detrimental. The review emphasizes the need for 'assistance-as-needed' and comparison of different control algorithms in clinical trials.

Study Duration

Not specified

Participants

Not specified

Evidence Level

Review

Key Findings

1
The review identifies four categories of active assistance control strategies: impedance-based, counterbalance-based, EMG-based, and performance-based adaptive assistance.
2
Challenge-based controllers, resistive strategies and error-amplification strategies can provide insights that might be missed by focusing solely on assistive-type algorithms.
3
Adaptive controllers can automatically tune assistance to individual changing needs, but including a 'forgetting' factor can prevent participants from slacking.

Research Summary

This review examines various control strategies for robotic therapy devices used in neurorehabilitation, focusing on assistive, challenge-based, haptic simulation, and coaching approaches. Active assist exercise is the primary control paradigm that has been explored so far in robotic therapy development. Because providing too much assistance may have negative consequences for learning, a commonly stated goal in active assist exercise is to provide 'assistance-as-needed' The paper highlights the need for rigorous clinical trials comparing different control algorithms and for improved models of human motor recovery to guide the design of robotic therapy control strategies.

Practical Implications

Personalized Therapy

Tailoring robotic therapy to individual patient needs, injury type, and recovery stage may improve outcomes.

Algorithm Comparison

Head-to-head comparisons of control algorithms in clinical trials are crucial for determining the most effective strategies.

Model-Based Design

Developing computational models of motor learning and recovery can inform the design of more effective robot therapy control algorithms.

Study Limitations

1
Clinical evidence is limited regarding the relative effectiveness of different robotic therapy controllers.
2
The question of the most effective control algorithms is still wide open, in part because the randomized controlled trials necessary to identify these algorithms are expensive and time-consuming.
3
It is still unclear whether robotic control approaches have the potential to produce greater benefits than is possible with simpler techniques, such as rote, unassisted practice

Your Feedback

Was this summary helpful?

Back to Assistive Technology