A neurofeedback training paradigm for motor imagery based Brain-Computer Interface

The performance of motor imagery based brain-computer interface (BCI) is mainly depending on subject´s ability of self-modulation EEG signals. A proper training would help naïve subjects to modulate brain activity proficiently. A few works of neurofeedback training showed that the performance was similar by using different feedback type because they did not provide the distinguishing characteristic to train subjects. To improve the performance of neurofeedback training, we presented a training paradigm which provided dissimilar information of imagination strength in visual feedback. The strength based feedback showed the difference in the inter-trials and it would help subjects to follow the right way to modulate brain signals. The experiment results verified the effectiveness of the proposed training paradigm.