Decoding human right and left hand motor imagery from EEG single trials using Sample Entropy

This study aims to researching whether human intentions to move left and right hands can be decoded from Sample Entropy (SampEn) features in non-invasive EEG. Eleven healthy subjects participated in the experiment. We found the a waves SampEn value of imagined left hand movement is larger than imagined right hand movement in most of male electrodes, especially F3, C3 and T3, and the a waves SampEn value of imagined right hand movement is larger than imagined left hand movement in most of female electrodes, especially Fp1 and P3. Experiment results show that SampEn can express the EEG dynamic features of imagined left and right hand movement. Besides it has clear physiological significance. Fisher discriminator analysis is utilized to dynamically classify imagined left and right hand movement according to SampEn features. As a result, we gained a male average maximum classification accuracy of 76.97%. Lastly, we discussed the influence of different time length and t value on the classification accuracy.