Hybrid EEG-fEMG based Human-Machine Interface for Communication and Control Applications

This paper demonstrated a human-machine interface (HMI) system to communication and control applications based on facial-electromyography (fEMG) and electroencephalography (EEG) signals. For the proposed fEMG-based HMI system, four commands with an average accuracy of 76.7% can be generated by the facial muscles pattern of four Thai syllables speech as left, right, front, and back positions. Moreover, we also designed the hybrid EEG-fEMG-based HMI system to reduce an error from involuntary command by using only single-channel EEG, and two-channel fEMG with the proposed algorithms achieved 84.5% average accuracy. The proposed hybrid HMI systems can be further developed to enhance communication and control abilities of the patient.