Application Research Of Synchronized EEG-fMRI Brain-Computer Interface System

Brain science research is the frontier and key work in the field of biomedicine,which can promote us to further understand the brain and develop the brain.With the development of science and technology,EEG and functional magnetic resonance imaging(fMRI)have their own advantages in temporal and spatial resolution.Synchronized EEG-fMRI fusion research can fully combine the advantages of both,observe the brain noninvasively,and provide multimodal analysis for brain source localization and observing the mechanism of brain activity.However,because of the huge magnetic field intensity in the magnetic environment,EEG signals will be completely "submerged" in the magnetic artifacts.How to remove these two artifacts is a problem that must be solved in the study of synchronous EEG-fMRI fusion.Firstly,this paper analyzes the causes and respective characteristics of gradient artifacts and BCG artifacts.Gradient artifacts is introduced with magnetic field reversal of magnetic resonance scanner,roughly in the period of one slice.The frequency amplitude of BCG artifacts is close to that of ECG.By observing the waveform of ECG signal and detecting the R peaks of ECG,it is found that the ECG has a certain periodicity,but the waveform of each period is quite different.Through peak detection of ECG,the interval time between R peaks in ECG can be calculated.Secondly,this paper proposes an online artifacts removal algorithm.For gradient artifacts,we propose to construct the optimal basis by combining the correlation coefficient and eigenvalue size,and remove gradient artifacts real time by improving the optimal basis group method.The algorithm is compared with the traditional gradient artifacts removal algorithm to verify the effectiveness of the algorithm.For BCG artifacts,we propose a clustering-OBS algorithm to remove BCG artifacts.Based on the characteristics of each BCG artifacts period,the BCG artifacts template was obtained for each cluster,and then the BCG artifacts was removed by using the optimal basis group method.Finally,the EEG signal is reconstructed.By comparing the algorithm in this paper with other commonly used algorithm,the effectiveness of the algorithm in this paper is verified by evaluating several indicators.Finally,the algorithm proposed in this paper was applied to the real time experiment of synchronous EEG-fMRI brain-computer interface system.Through the establishment of the online experiment platform of synchronous EEG-fMRI braincomputer interface,the experiment was conducted using the character input experimental paradigm based on P300 signal.The validity of this algorithm is verified by comparing the accuracy of experimental results of this algorithm with those of other algorithms.Through this study,we can lay a solid foundation for future research on synchronous EEG-fMRI fusion.