Towards Hybrid EEG-EMG-Based Control Approaches to be Used in Bio-robotics Applications: Current Status, Challenges and Future Directions

Abstract

In the last few decades, bio-robotics applications such as exoskeletons, prosthetics and robotic wheelchairs have progressed from machines in science fiction to nearly commercialized products. Though there are still several challenges associated with electromyography (EMG) signals, the advances in use of EMG signals for controlling such bio-robotics applications have been enormous. Similarly, recent trends and attempts in developing electroencephalography- (EEG) based control methods have shown the potential of this area in the modern bio-robotics field. However, the EEG-based control methods are also yet to be perfected. A new approach of combining both these control methods, which take the advantages, and diminish the disadvantages, of each system might therefore be a promising approach. In this paper, we review hybrid fusion of EMG- and EEG-based control approaches in the bio-robotics field which have been attempted or developed to date. We provide a design overview of the method and consider the main features and merits/disadvantagages for the approaches that have been analyzed. We also discuss the current challenges regarding these hybrid EEG-EMG control approaches and propose some potential future directions.

Control interfaces to actively support the arm function of men with Duchenn Muscular Dystrophy

“…2) are generally a suitable solution to derive the movement intention from highly paralyzed patients [112,118,113,121,125]. Interfaces based on parallel systems have also been implemented in hybrid sensor modalities [198] to expand the sources of information.…”

“…However, researchers are improving their performance by implementing artificial learning algorithms [56,219,220], shared-control strategies [131,221] and combining brain-computer interfaces with other sensor modalities [198].…”

“…The concept behind hybrid control interfaces is to exploit the advantages and diminish the limitations of each individual system by combining them. Hybrid approaches can improve accuracy, reliability, and robustness compared to individual use of control interfaces [198]. Moreover, hybrid control interfaces have the potential advantage of adapting better to the needs and capabilities of the user since they can rely on several information sources.…”

“…For instance, if the user gets fatigued and EMG signals present poor quality, the hybrid system could adapt and give priority to other information sources. EEG-EMG-based control interfaces are an example of a hybrid system that has been extensively used to operate active prostheses, orthoses, and robotic wheelchairs [198]. In active lower-extremity prostheses, hybrid interfaces combine EMG and mechanical inputs to identify the phase of the gait cycle and actions like sitting or standing [199,200,201].…”

“…For instance, an EMG-force hybrid interface could use EMG signals to distinguish voluntary from un-voluntary forces, and force signals for the control of the direction and amplitude of the movements. Hybrid approaches have been shown to improve accuracy, reliability or robustness in comparison to individual use of control interfaces [198].…”