The effects evaluation of a long-term neurofeedback training using coupling EEG-EMG features

Brain-computer interfaces (BCIs) have been widely used to improve or restore neural functions. For stroke patients, BCIs based on motor training show a promising potential in motor rehabilitation. However, the neural mechanism and the effects of different time course in motor rehabilitation remain unclear. To this end, our study focused on the BCI based neurofeedback training (NFT) design and its evaluation method. During motor imagery and execution (MI/ME) tasks, electroencephalogram (EEG) and electromyogram (EMG) were synchronously recorded and probed. We found the multi-band changes of coupling EEG-EMG features. Additionally, the longterm motor NFT significantly improved the cortical-muscle activation, while non-feedback training improved less. These relevant results give a theoretical basis to the development and application of new neural rehabilitation technology.