Enhancing the Prediction of Locomotion Transition With High-Density Surface Electromyography.
IEEE Journal of Biomedical and Health Informatics
|March 3, 2025
Summary
High-density surface electromyography (HDsEMG) significantly improves locomotion transition prediction for assistive devices. HDsEMG captures subtle muscle changes, enhancing accuracy compared to traditional methods, even with electrode detachment.
Area of Science:
- Biomedical Engineering
- Rehabilitation Technology
- Human Movement Science
Background:
- Accurate prediction of locomotion transitions is crucial for seamless integration with lower limb assistive technologies like exoskeletons and prostheses.
- Current methods using inertial and bipolar electromyography (EMG) sensors lack the precision to detect subtle muscle activation changes, especially during early transition phases.
- This limitation hinders clinical utility and optimal performance of assistive devices.
Purpose of the Study:
- To evaluate the effectiveness of high-density surface electromyography (HDsEMG) in detecting muscle activation changes during stair-related locomotion transitions.
- To compare the prediction accuracy of HDsEMG with traditional bipolar EMG methods.
- To assess the robustness of HDsEMG against signal loss and the impact of reduced electrode count.
Main Methods:
- Utilized two high-density surface electromyography (HDsEMG) sensors to capture muscle activation patterns during stair transitions.
- Compared HDsEMG data with bipolar EMG recordings from the same muscles.
- Implemented image-inpainting signal processing to simulate and evaluate the effects of electrode signal loss.
- Analyzed transition prediction accuracy at different time points relative to toe-off and with varying electrode configurations.
Main Results:
- HDsEMG significantly increased transition prediction accuracy from 70.2% to 91.1% at toe-off and from 89.8% to 99.2% with a 400-ms delay.
- HDsEMG demonstrated superior ability in capturing subtle muscle activation changes during early transition stages.
- Reducing electrode count to 21 per muscle minimally impacted performance (88.3% accuracy at toe-off).
- HDsEMG showed robustness against signal loss, with only a 3% decrease in accuracy even with 30% electrode detachment.
Conclusions:
- HDsEMG offers a significant advancement in locomotion transition prediction accuracy for assistive technology interfaces.
- The ability of HDsEMG to detect subtle muscle activations is key to improving prediction, especially in early transition phases.
- HDsEMG's robustness to signal loss and potential for optimized electrode distribution present a promising solution for clinical applications.
