The Analysis Of EEG During Imagined Movements Based On Brain Computer Interface

Brain-Computer Interface(BCI)is a direct information communication and control channel established between human and computer or other electrical devices and it is a new communication system that does not depend on the brain's normal output pathways of peripheral nerves and muscles.EEG-based BCI may provide an effective communication and control channels with world for the paralyzer,especially those "locked-in" but with intact ideation.BCI is winning more and more attentions.On the base of foregoing accomplishment,this paper focuses on some topics described as following:(1)Feature extraction of the EEG signalsFor the data that provided by Graze BCI,the EEG features during imagined right and left hand movement were extracted by using power spectral entropy(PSE), approximate entropy(ApEn),sample entropy(SampEn)and Embedding-space decomposition(ESD).Both methods are compared and the advantages and disadvantages of both are discussed.It acquires good results with these four methods, and the maximal accuracy achieves 90%by using PSE.(2)Design of the Mental Tasks ClassifierThis paper presents a design approach of time-variable linear classifier based on time accumulation.We used this method to distinguish the left and right hand motor imaginary tasks and gained the satisfying result.It is obvious that the classification accuracy by using time-variable linear classifier is higher than using Fisher linear classifier.(3)Estimate the performance of BCI systemThe performance of BCI system with different features is estimated by some indexes,which includes classification accuracy,signal-to-noise ratio(SNR)and the mutual information(MI).It was found that the performance of BCI system with PSE and time variable linear classifier is the best.The experimental results show that while the useable information reflecting different conditions of the mental tasks has been properly extracted and the proper classifier has been applied,there is highly accurate discrimination for the states of different mental tasks.From the experimental results,the PSE is a sensitive parameter for EEG of imaginary hand movements.It acquires a good classification result by the time-variable linear classifier.These methods are both simple and quick and provide promising methods for on-line BCI system.