Study Of The Implementation Of Motor Imagery Based Feedback System And Its Application

In recent years,brain-computer interface(BCI)based on motor imagery(MI)has attracted wide attentions because of its potential application prospects in such fields as motor control,neurological rehabilitation and intelligent control under the special environment.MI BCI is expected to play an important role in improving the motion control ability,the information processing efficiency in brain and the rehabilitation of cerebral palsy due to the similar neural mechanisms with the actual movement.MI feedback converts the brain electrophysiological signals into vividly visual and sound signals,such as music,image and so on.In the feedback system,the subjects receive the feedback information for self-regulation and control,and thus to enhance the corresponding abilities.The main contributions of current dissertation is as below:1.Base on the existing BCI system,the multi-process and multi-thread design are used to realize the EEG acquisition,design and development of different feedback modes,which is finally resulted in a feedback system platform.The developed system includes the direct display of MI related information including power spectrum,ERD and the classifier outputs,etc.Moreover,based on those MI related information,we realized the feedback controls for different scenes,aiming for the flexible application purposes.2.Based on the developed feedback system,we design and carry out the experiments based on the normal subjects to evaluate the performance of the MI feedback system.In the experiment,we grouped the subjects into e feedback group and the control group.For the feedback group,the three feedback modes,including the direct feedback of ERD,power spectrum and classifier ouput,the feedback control by the frequency of interest band,and the feedback control by ERD and classifier output.While the control group only carries out traditional MI training experiment.Each group includes 8 subjects,and each subject performed 5 training sessions,with two consecutive sessions being 6 days interval.8 minutes resting EEG data will be collected before every training,and then collect the corresponding experimental task data.By analyzing the MI control accuracy,and brain networks,we found that the both MI ability and the efficiency of the brain network processing information of the feedback group were significantly improved compared with the group.3.Moreover,we applied the developed MI feedback system to assist the rehabilitation of cerebral palsy patients.Currently,10 children with cerebral palsy are involved in the MI feedback training.During the feedback training,we used the developed system to instruct the CP subjects to perform the MI training.After every 10 days MI feedback training,the corresponding EEG dataset including 8 minutes resting EEG data and MI task EEG data is recorded.Each subject includes 5 sessions training.The analysis of MI ability and brain networks reveals that the MI feedback system could effectively improve the functional brain network topology,which is more consistent with the normal MI physiological basis.The preliminary application to CP demonstrates the developed MI feedback system is helpful for the rehabilitation of CP.