Assessing walking ability using a robotic gait trainer: opportunities and limitations of assist‑as‑needed control in spinal cord injury

Journal of NeuroEngineering and Rehabilitation, 2023 · DOI: https://doi.org/10.1186/s12984-023-01226-4 · Published: January 1, 2023

Spinal Cord Injury Assistive Technology Biomechanics

Simple Explanation

Walking impairments following neurological disorders are commonly assessed using clinical scores, which have limitations. Robot-assisted locomotor training is an established practice, offering controlled environment assessment of walking ability. The study proposes an adaptive assist-as-needed (AAN) control for the Lokomat exoskeleton. It reduces support based on the patient's ability to follow a gait pattern displayed on a screen. The hypothesis is that the robotic support values, determined by the AAN algorithm, correlate with an individual's walking ability, providing useful quantitative information.

Study Duration

May to November 2016

Participants

15 participants with spinal cord injury and 12 unimpaired controls

Evidence Level

Not specified

Key Findings

1
The AAN controller is usable across different injury severity levels in patients with spinal cord injury, showing feasibility.
2
Robotic knee stiffness at terminal swing is a key variable, explaining a substantial portion of variance in 10MWT and TUG scores.
3
Adding maximum hip flexor torque to the model improves explained variance above 85%, showing muscle strength's contribution.

Research Summary

The study investigates using a robotic gait trainer (Lokomat) with an assist-as-needed (AAN) control to assess walking ability in individuals with spinal cord injury (SCI). Key findings include the feasibility of the AAN controller across different injury severity levels and the identification of robotic knee stiffness at terminal swing as a significant predictor of overground walking speed. While the current implementation is not ready for clinical assessment, the approach is safe and demonstrates the potential for robotic gait trainers to provide objective measures of walking ability.

Practical Implications

Quantitative Walking Assessment

The AAN software can be used to quantify the support required by a patient during robotic gait training.

Adaptive Training

The adaptive software may enable more challenging training conditions tuned to the ability of individuals.

Assist-as-Needed Training

The approach could be integrated as assist-as-needed training, enhancing patient challenge and potential outcomes.

Study Limitations

1
The intra-rater reliability between two sessions needs to be improved before the test can be used in therapy.
2
We could not develop a model to predict the walking-related functions of the patients continuously from the non-ambulatory to the ambulatory phase.
3
The model used to predict the walking tests is a simple linear regression: it may be that a generalized linear model with another link function would lead to better predictions.

Your Feedback

Was this summary helpful?

Back to Spinal Cord Injury