Research On The Improvement Of Whole Brain FNIRS-EEG Bimodal Detection Performance

Functional neuroimaging methods provide a new method for learning how the brain works.As a noninvasive optical imaging method,functional near-infrared spectroscopy(fNIRS)has been widely employed in studies of brain function.By the channels' location on the scalp,fNIRS can detect neuro signals' source from the cortex.However,fNIRS has problems such as insufficient time resolution and can't detect neuron activity directly,which affect the effectiveness of fNIRS used in brain functional detection.As a mature brain functional detection technology,Electroencephalography(EEG)has sufficient time resolution,while it can not detect the source of neuroactivity.Both fNIRS and EEG place few limitations on the subjects and environment,therefore can detect subjects' brain activity in a relaxed state.By combining fNIRS and EEG,bimodal acquisition technology can have complementary advantages of the two methods,help people to study neural mechanism in cognitive psychology experiments better.The current fNIRS-EEG bimodal technology is still in its infancy,has problems such as low signal-to-noise ratio,poor system integration,and few detection channels.Improving fNIRS-EEG bimodal detection performance is very important for employing this technology in brain function research.This thesis focused on the improvement of the fNIRS-EEG bimodal detection performance,and the main contents are as follows:(1)Studying key issues affecting whole brain fNIRS detection performance based on our homemade optical fiber fNIRS system.First,study superficial signals which affect the fNIRS signal quality.By theoretical analysis and model experiments,this study established a method based on short separation channels with adaptive filtering to remove superficial interference and verified the effectiveness of the method in vivo experiment.Then this study developed on-line superficial interference removing function embedded in fNIRS software.Second,this study analyzed key issues affecting fNIRS system stability by theoretical analysis and phantom tests.To solve the issues,this study designed light sources heat dissipation system and fiber-optics probe.Then,focused on the visualization of fNIRS acquisition software,this study analyzed software requirements and developed fNIRS acquisition software with high visualization by Matlab-Lab VIEW hybrid programming.At last,through an in vivo N-back experiment verified that the fNIRS system can detect brain functional activity during working memory reliably.(2)Combining fNIRS and EEG system in both physical structure and digital signal level,improving bimodal detection performance.In physical structure level,this study developed multi-channels optic-electronic integrated probe cap using the special material basement and probe fixtures to achieve the joint arrangement of fNIRS probes and EEG electrodes.In signal synchronization level,this study developed photoelectric marking method to achieve simultaneous bimodal detection with high precision.Focused on bimodal detection utilized in brain-computer-interface,this study did research on online neural feedback based on fNIRS and EEG signals,and proved the efficiency of neural feedback by a motor imagery experiment.Then this study developed closed-loop bimodal acquisition software based on Lab VIEW by remote data access method.At last,through an in vivo Stroop experiment verified that the bimodal system can detect hemodynamic response and neuroelectrical activity simultaneously and effectively.(3)In order to promote the application of the whole brain bimodal detection,studying gender-related brain function during working memory.Focused on few bimodal data analysis methods,this study established brain networks analysis methods based on the bimodal signal.Multiple data analysis methods including functional connectivity,effective connectivity,and small world network property were employed to analyze fNIRS and EEG data,as well as the correlation between two types of signal.Results of brain activity and connectivity indicated that women surpassed men in verbal working memory.Men achieve similar performance to women by enhancing brain activation,optimizing brain networks and employing preferred visuospatial strategies to encode memory in the memory tasks.This study further proved that brain networks analysis based on bimodal detection can help people to extract the information of brain function more comprehensively and effectively.This study devoted to improving the performance of fNIRS-EEG bimodal detection in various aspects including signal quality,bimodal system integration,software visualization and data analysis methods.This study can help researchers to employ bimodal detection technology study brain function in more details.