A comparative analysis of three non-invasive human-machine interfaces for the disabled
Vikram Ravindra1, Claudio Castellini1
1Robotics and Mechatronics Center, German Aerospace Center (DLR) , Weßling , Germany.
Frontiers in Neurorobotics
|November 12, 2014
Summary
Pressure sensors offer a stable and cost-effective alternative to surface electromyography (sEMG) for controlling robotic prosthetics. This human-machine interface (HMI) shows comparable accuracy to sEMG in detecting finger forces.
Area of Science:
- Rehabilitation Robotics
- Human-Machine Interface (HMI) Design
- Biomedical Engineering
Background:
- Surface electromyography (sEMG) is a traditional human-machine interface (HMI) for rehabilitation robotics, but its signal variability leads to unreliable patient intent detection.
- Novel HMIs are being explored to improve control signals for robotic prostheses, aiming to overcome sEMG limitations.
Purpose of the Study:
- To compare the performance of three HMIs—sEMG, ultrasound imaging, and pressure sensing—for detecting finger forces.
- To evaluate HMIs based on prediction accuracy, temporal stability, wearability, and cost.
Main Methods:
- A psychophysical experiment was conducted with ten subjects performing a finger-flexion task.
- Comparative analysis of sEMG, ultrasound imaging, and pressure sensing for finger force detection.
Main Results:
- Pressure sensing and sEMG demonstrated comparable prediction accuracies, outperforming ultrasound imaging in this experiment.
- Pressure sensing exhibited significantly better stability over time compared to sEMG.
- Pressure sensors are as wearable as sEMG electrodes and considerably more cost-effective.
Conclusions:
- Pressure sensing presents a viable alternative or augmentation to sEMG for controlling multi-fingered hand prostheses.
- The cost-effectiveness and stability of pressure sensors make them a promising HMI for rehabilitation robotics.
