| The study of the neural mechanism of facial processing has become an important aspect in the cognitive neuroscience. Bruce-Young model and Haxby model have successively been proposed based on results from behavioral research, neuropsychology, and functional neuroimaging respectively. Adolphs model, furthermore, has set up a time framework for the above models in light of the findings of the electromagnetic neurophysiology. However, many details of face processing remain unclear. For example, how does the facial beauty and sexually dimorphic features be processed? Are they different from facial structure coding and facial expressions in processing pathway? Does the unconscious emotional stimulus has any effect on processing facial beauty and how? ...We'll try to answer those interesting questions in this dissertation. Spatiotemporal analysis based on both multi-channel event-related potentials (ERPs) recordings and a statistical parametric mapping (SPM) method has been applied to investigate the brain mechanism of the processing of facial beauty and facial sexual features.A database of the facial pictures has in advance been constructed with those collected from websites and processed for uniform with the software of PhotoShop. 600 pictures (male 300; female 300) of young faces have been graded on the beauty level by 23 observers (male 12, female 11; average age of 29.96). Among them, 360 pictures are selected with the beauty values as beautiful, plain and ugly for each gender, each group thus consists of 60 pictures and all of them will be used in the following three experiments.In experiment 1, thirty-four (17 males and 17 females) healthy subjects joined in. A 2 X 2 X 3 three-way factorial design was used to examine the gender of subjects (male, female), the gender of faces of stimuli (male, female) and facial beauty (beautiful, plain and ugly). The SPM of F-value suggested that: (1) significant interaction of the above factors appeared at the small region of right occipital in 160-170 ms after stimuli, at the small region of left occipital in 210 ms and in more regions after 270 ms; (2) the significant interaction between the gender of subjects and the gender of faces appeared at many regions in 30 ms, 80 ms, 130 ms and 150 ms after stimuli, at a larger region on the right hemisphere during 160-210 ms and involving more locations later; (3) the significant main effects of facial beauty appeared in the left posterior scalp in 90-100 ms, in a larger region of anterior scalp in 120 and 170 ms as well as in more locations after 300 ms. The results suggest that the processing of facial beauty and sexual information goes far earlier than the stage of facial structure coding represented by N170. Such findings can be well interpreted by the hypothesis of reward value proposed in the study of facial attractiveness together with the hypothesis of fast early perceptual processing to highly salient stimuli of Adolphs model. Such important information to individual adaptation as facial beauty and sexual features undergoes a fast early processing independently and then gets into the reward system or alert system for coarse classification rather than appears after the stage of facial construction coding acknowledged before. Simultaneously, it is first integrated to facial processing and then to attention-related subtle processing.In experiment 2, we'll further examine the impact of facial orientation to the facial beauty processing. A total of 20 healthy subjects (10 males) were taken part in. A 2 X 3 factorial design of repeated measures was used to investigate the orientation (inverted and upright) and the facial beauty (beautiful, plain and ugly). The SPMs of F-value recorded suggested: (1) the significant interaction between facial orientation and facial beauty appeared at the region of the occipital part in 150-170 ms after stimuli, and at the anterior scalp in 310 ms and 480 ms later; (2) the significant main effects of facial orientation appeared early at the left frontal area in 110-160 ms, extended to a larger region in 190-230 ms. In contrast to the previous studies which concluded that inverted faces other than non-face stimuli have effects on N170 resulted in specialization of facial processing, our results suggest that facial orientation itself may be a salient clue or character, which can enter pathway of fast early processing. Facial orientation, thus, should be examined in the same important way as facial beauty and facial sexual information.In experiment 3, a masked priming paradigm was applied to examine the impact of emotional stimuli to the facial beauty processing. A total of 24 healthy (12 males)subjects joined in. Two 3X3 factorial designs (under subliminal condition and under supraliminal condition) of repeated measures was used to examine emotional prime (positive, neutral, and negative) and facial beauty (beautiful, plain, and ugly). The SPMs of F-value suggested: (1) under subliminal condition: the significant interaction of emotional priming and facial beauty appeared at the left frontal area in 210-220 ms, the left joint of temporal, parietal and occipital areas in 360-380 ms, the small left frontal area. The significant main effects of emotional factor appeared at the left posterior scalp as early as in 80-100 ms. (2) under supraliminal condition: the significant interaction of emotional priming and facial beauty appeared at the left occipital area in 210-230 ms, the right joint of temporal, parietal and occipital areas in 380 ms. The significant main effects of emotional factor appeared at the posterior scalp as early as in 60-80 ms and 140-180ms. The results suggest without awareness consideration, the effects of emotion to facial beauty processing appear at the early stage. On the one hand, emotion processing has been brought forward in line with the processing of sexual features, facial beauty and facial orientation and so on. On the other hand, due to the specialization of this experimental design, of which in fact the emotional prime induced different emotional states that have impact on facial beauty processing. Moreover, the findings strongly suggest that priming effect exists under subliminal condition. It needs further investigation.Except for the Adolphs model, which mentioned the fast early processing for salient stimuli around 120 ms, the processing of facial beauty, sexual features and facial expressions has been derived from facial structure coding in the three existing models of facial recognition. Taking advantage with the reward hypothesis of facial beauty, we are quite sure that, processing the important information to individuals comes earlier than facial constructional coding and enters the reward or alert system, for the early processing of those salient stimuli has been found in all of our three experiments. Therefore, we propose a particular model, in which the important facial information including expression, beauty, gender, orientation, etc could be processed in early stage independent from facial integrated processing. |
PDF Full Text Request