A Research On Neural Mechanism Of Other-race Effect Based On Independent Components Analysis

Other-race effect(ORE)is a kind of deviation phenomenon that is much easier to identify own-race faces than to recognize other-race faces.The study of the neural mechanisms of ORE has become one of the hottest spots in the field of cognitive neuroscience.With the rapid development of the functional neuroimaging technology,using functional magnetic resonance imaging(fMRI)and electroencephalogram(EEG)to study the neural mechanism of interracial effect has become a reality.It has been reported that there are some neural response difference in some traditional face-sensitive cortical area such as the right fusiform face area(FFA).With the deepening study of the neural mechanism of the effect of ORE,some recent studies also found that there are significant differences in the effective network between the cortical area involved in face processing.As a data-driven method,independent component analysis(ICA)plays an increasingly important role in functional neuroimaging research in recent years.Both fMRI and functional near infrared spectroscopy(fNIRS)are the non-invasive neuroimaging methodology.However,compared to fMRI,fNIRS have some advantages such as the relative high temporal resolution,insensitivity to move,and low experimental cost.In order to reveal the brain activation patterns of ORE,this paper investigates the neural mechanism of ORE using ICA and fNIRS imaging technology.The process is as follows:First,the fNIRS data of 95 subjects were collected by fNIRS imaging technique,and the primitive spatial data and optical data were preprocessed to obtain the registration spatial data and oxygenated hemoglobin concentration data.Second,Group ICA was performed on the data of oxygenated hemoglobin concentration,and the time series algorithm was developed according to the data characteristics and experimental design,so as to analyze the differences of brain mechanism caused by different racial faces under the same conditions.Third,the correlation analysis between each of independent components of own-race faces and the ideal hemodynamics responses elicited by own-race faces was performed.This correlation analysis was also performed between the independent components and their ideal hemodynamics responses related to other-race faces.Then these two correlation coefficients were compared to determine whether there was a significant difference.After the significant difference was obtained,the corresponding brain region distribution and spatial coordinates were found.Finally,this paper concluded that the brain regions showing greater activation for other-race faces than for own-race faces activation degree are:Precentral_R,Calcarine_L,Frontal_Mid_Orb_L,occipital lobe,gyri occipitales superiores.The brain regions showing greater activation for own-race faces than for other-race faces are:Frontal_Sup_L,Frontal_Inf_Tri_L,Frontal_Sup_L.This paper presents a complete fNIRS data processing process by using Group ICA,and reveals the brain activation patterns of ORE through the analysis of near-infrared brain function data and lay the foundation for further research.The findings of the present study are consistent with those of the previous studies using fMRI,EEG.Further,the present study reveals the race-related difference in the spatial-temporal pattern of face processing.The present study can not only helps us understand the neural mechanism of face processing,but also provide a bionics basis for improving the machine learning algorithm for face recognition.