The Neural Mechanism Of Facial Processing In Rapid Serial Visual Presentation

Face is very important in social exchanges and biological evolution. It has been studied since Darwin. Recently it has increasingly attracts the attention of many psychologists and cognitive neuroscientists. Face can express some information such as sex, age, charm, expression, identity, race, among which expression and identity information were the focuses of most scientists. Many studies from behavior, neuropsychology, neurophysiology, and brain imaging have indicated some important characteristics about facial expression and identity processing. But there are conflicting evidences about the relationship between identity and expression. According to orthodox face-recognition models, the processing of information regarding the identity of faces takes place within a dedicated cognitive route that is independent of the cognitive route which processes other types of face-related information. It is suggested that there is a parallel pathway between identity processing and expression processing, and some studies also support it; whereas other studies have found support for the notion that there are two separate systems for the processing of identity and that of expression which are not completely independent but are rather interconnected. Colligating the results of previous studies, the present study hypothesized that the relationship between processing of identity and processing of expression is not completely independent and is interconnected. They are independent because there are different brain areas in processing identity and expression information, while they are interconnected because there is a same brain area in processing these two kinds of information. So at present it is of both practical and academic significance to further investigate the relationship between identity and expression in order to solve the controversy about this relationship.Moreover, only a limited number of previous studies have touched on the relationship between the processing of facial and attention; and these studies only use visual search task to examine the effect of processing identity and expression from the angle of space-based attention instead of time-based attention. Rapid serial visual presentation paradigm can examine the character of attention from time dimension, and using this paradigm in the research of facial processing will help us further understand the relationship facial processing and attention.Applying high density ERP technique and dipole source analysis, we studied the neural mechanism of facial processing. In this study, we use the subtraction technique to eliminate the overlapped ERP component caused by rapid-presented the visual sitimuli. We mainly analyzed the ERP component elicited by T2. The first study indicated the neural mechanism of facial expression processing. We based on the ERP effect that happy, fearful , and neutral facial expressions in the RSVP and dipole source analysis technique infer the neural model of facial expression processing. The second study included two experiments and indicated the neural mechanism of facial identity processing. The first experiment compared the ERP effect between face and non-face in the RSVP; the second experiment further compared the ERP effect between upright face and invert face in the RSVP. Through connect Garner interference paradigm and RSVP paradigm, the third study further indicated the relationship between processing of identity and processing of expression. The main findings are as follows:ERP components are modulated by facial emotion valence in the early phase of perception and cognitive evaluation. Early anterior N1 and posterior P1 reflect the bias for processing the fearful expression, the fearful expression elicited larger amplitude than happy and neutral expressions. The amplitude of VPP and the amplitude of N170 in short interval condition can distinguish the emotional expression from neutral face stimuli. The amplitude of N3 and P3 can distinguish among emotional facial expression such as fearful expression, happy expression, and neutral faces. The results of dipole source analysis show that the source which fearful expression elicited larger N1 is the prefrontal cortex; the sources that fearful expression elicited larger P1 are the parahippocampal cortex and the inferior occipital cortex; the source which happy expression elicited larger VPP and N170 is the right fusiform gyrus, and the sources which happy and fearful expressions elicited larger N3 and P3 are the superior temporal cortex and the anterior cingulate cortex.Face elicited larger P1, VPP, and P3 components and shorter latency of N1 than house. For the N170, attention has greater influence on the procession of house than the processing of face. The result of dipole source analysis for the difference wave between face and house show that the source that face elicited larger P1 is the inferior occipital cortex, the source that face elicited larger VPP is the right fusiform gyrus, and source that face elicited larger P3 is anterior cingulate cortex. Both upright and inversion faces show the effect attentional blink; inversion face show lower accuracy than upright face. P1, VPP; and N170 show the effect of inversion, upright face elicited larger the amplitude of P1 than inversion face in parietal-occipital sites, while inversion face elicited larger the amplitude and longer the latency of VPP and N170 than upright faces, and there is the right hemisphere dominance for N170. The result of dipole source analysis for the difference wave between inversion faces and upright faces show that the source that upright faces elicited larger P1 is the inferior occipital cortex, while the source that inversion face elicited larger VPP and N170 is the right fusiform gyrus.Emotion task elicited larger VPP amplitude than identity task, and the VPP amplitude show the interference from processing expression information to processing identity information. There is the right hemisphere dominance for N170, the right hemisphere elicited larger amplitude and longer latency. The result of dipole source analysis for the ERP difference wave evoked by the identity task was subtracted from the ERP evoked by the expression task showed that the source which the expression task elicited larger VPP is the bilateral fusiform gyrus, and the source which the expression task elicited larger N3 is the superior temporal cortex.The N1 component in expression processing is influenced by attentional resource that decided by interval in expression processing, while the one in identity processing is not influenced by it. Attentional resource has a little influence on P1, while the attentional resource in relatively abundant elicited larger VPP, N170, N3, and P3 amplitudes than in relatively deficient. P3 amplitude is the index of attentional resource and difficult of task: the more difficult the task, the smaller the P3 amplitude.The present study proposes hypothetical model involving three stages of processing of emotion facial expression. In this model, the first stage allows automatic processing for fearful facial expressions, while emotionally charged facial expressions are distinguished from neutral faces in the second stage and different emotion facial expressions are differentiated in the third stage. Processing in the second and third stages is sensitive to attentional resource availability while processing in the first stage remains relatively independent.In light of the present results, we propose neural model involving face processing. In this model, the relationship between processing of identity and processing of expression is not completely independent but is interconnected. They are interconnected because fusiform gyrus is the same brain area in processing of identity and processing. This model can explain the result of previous studies well. Because the processing of identity must be through fusiform gyrus, while processing of expression may not, there is relative independence for processing of expression and processing of identity, thus the prosopagnosia can still recognize facial expression of emotion; however, amygdale and superior temporal cortex (especially for the amygdale) are very important in processing of expression, so the patients with damages in these areas will have deficits in the processing of expression while the ability of recognition is not impaired. However, the processing of identity and processing of expression will also involve fusiform gyrus, these processing will show interconnection and interaction in some conditions.