| BACKGROUND: According to the findings from neuropsychological studies, prosopagnosia patients could not recognize familiar faces but they could recognize facial expression. The fact established dissociation between face identity and facial expression processing. Yin reported the inversion effect of face, which means recognition accuracy for inverted faces, was down and the reaction time was longer. Brace and Young proposed the cognition model of face recognition, which emphasize parallel processing of facial information such as identity and facial expression. The evidences from single-cell recordings in the temporal cortex in macaque found that populations of cells in inferior temporal gyms and in the fundus of the superior temporal sulcus were shown to be selectively responsive to human and monkey faces. Using fMRI, facial identity processing was located in the right fusiform gyms. Haxby proposed the neural network model of face processing, which was made of core system and extended system. They involved separately in visual analysis, spatial attention, emotion, identity and name. The limbic lobe play an important role in facialexpression processing. The neuroanatomical structure in emotion processing was consisted of orbit-frontal cortex, hippocampus, cingulated gyrus and amygdala. The definition of emotion is people's attitude, experience and reaction to objective identity. Because ERP could provide a lot of information, many scholars use it to record the waveform evoked by different facial expression (positive, negative and neutral) picture. It ia gong to be the most common experimental model. During face processing studies, Bentin first report a negative waveform whose latency was about 170 ms in temporo-occipital area selectively responsive to face identity (structure analysis). It reflected the early structure encoding stage and was not related to face familiarity, sexuality, age and race. The deep electrode record N200, which was related to face recognition, in ventral temporo-occipital cortex before an operation with epilepsy patients. The inverted face could delay the latency of N170 and enlarge the amplitude. Most studies in face processing concentrated in the latency and amplitude of N170 or N200. They did not involve in the ERP components the other area and stage. During the depression cognition studies, the depression patients had negative deviation in attributional styles, negative self schema, automatic thoughts and negative ways of coping. There was significant evidence about the difference between depression patients and normal in Neuropsychological, biological, neurological image and functional image studies. The evidence from ERP research showed that amplitude of P3 related to auditory sense in depression patients group was less than control group. The latency and amplitude change of inner resource component such as N2 and P3 reveal that there was higher stage cognition abnormal. The latency and amplitude change of outer resource component such as Nl and P2 reveal that there was higher stage cognition abnormality and alert, select and attention disorder. Post imperative negative variation (PINV), the chief component of CNV, was delayed in depression patients. The amplitude of CNV and the area of anticipatewave, the indicate signal of CNV, was less than control group. The facts revealed that there was disorder in arousal, attention and motivation depression patients. During the relationship between depression types and ERP, the latency of P3 was related to psychosocial movement inhibition and the amplitude of P3 was significant down in psychosocial depression group. Compared with the depression patients without suicide attempt, the ones with suicide attempt demonstrated that amplitude of P2, P3 and CNV decreased significantly and PINV enlarged specially. This evidence revealed that amplitude of P3 was negative related to suicide risk and helplessness. The self concentration and preoccupy of the depression patients with suicide attempt dispersed the indicative signal stimuli attention and induced the abnormality of amplitude of CNV. The latency of P3 in typical group of depression patients was significantly delayed, but there was not significant difference between the atypical group and control group. Moreover, there was not significant difference between single and double phase depression group. The abnormality of ERP was related to the specific depression symptoms other than the depression severity. Other evidence showed that there was abnormality in the predominant hemisphere and the right cerebral hemisphere (modulation of complicated tone processing) of the depression patients with somatic anhedonia was abnormal. Previous research was limited in several ERP components. For example, most scholar proposed the neurophysical mechanism of depression only with the observation of P300, which was related to auditory stimuli. We investigate the effects of both facial expression and orientation with spatiotemporal pattern of event-related potentials (ERP) of patients with major depression and normal people. Compared of the difference of "emotional waveform" between major depression and normal people, we study the influence of depression in emotional facial recognition and the mechanism involved. METHODS: 32 healthy subjects from the Southern Medical University and 12 majordepression patients, who were chosen (1) according to CCMD-3 diagnosis standard,(2) without organic origin brain diseases, electroshock, physical diseases, (3) righthandedness, (4) voluntary to attend this trail, were chosen from July to September,2004. They had normal eyesight (or corrected-to-normal), without mental disease andmental family history. Three categories of stimuli were used: positive, neutral andnegative emotional schematic face pictures. The corresponding inverted stimuli andthree Chinese words (happy, think, fear) were also presented. Thus, these ninepictures were viewed fifty times in the total of four hundred and fifty picturepresentations. The pictures were displayed separated for 1.5 second each andpresented in random. Inter-trial intervals was 1500 ms. Subject's task was to press akey if the successive stimulus was the same. Subjects were instructed to respond asaccurately and as quickly as possible. All responses were made with the right hand.Participants were seated in a dimly lit sound attenuated cabin, with response buttonsunder their left and right hands, at 100 cm from a computer screen (resolution, 800 x600 pixels, vertical refresh rate, 85 Hz). The size of pictures presented in the center ofthe screen is 10.6 x 12.3 cm2 with visual angle 6.0 x 7.0°. The total trial was made offive blocks each of which would take 135 seconds. Subjects could have a rest forthirty seconds between the successive block. Before Event-related EEG was recorded,subjects had a short training block of 20 stimuli. Event-related EEG was recordedfrom 19 channels of international 10-20 system (see Fig. 2) with linked earlobes asreference. EEG was amplified by means of a ERP system developed in our lab,digitized with sampling frequency of 1000 Hz and bandwidth of (0.10, 30) Hz and 50Hz notch. Electrode impedance during the experiments did not exceed 5 k￡l. EEGepochs from 100 ms before to 1000 ms after target stimulus (see Fig. 1) wereaveraged. Trials contaminated with ocular, muscular or any other type of artifactswere inspected visually and rejected. Sweeps exceeding ± 50 microvolt in anychannels and those with incorrect performance were excluded from the ERP averaging offline. ERP waveform elicited by the different (facial expression, orientation) pictures and accuracy would be statistic analyzed. In control group, the 3(expression: positive, neutral and negative)x2(orientation: upright, inverted) factorial design of repeated-measures was applied with statistical parametric mapping (SPM) of F-value. In the contrast between depression and control, the group t-test was applied to their "emotion-wave" (EP), which was subtracted from emotional (positive, negative) wave to neutral wave, SDS scales, response time and accuracy. The t-value of "emotional wave" was presented with statistical parametric mapping. In the early slope of ERP, six waveforms had been elicited in control. According to latency, their name was P50, N150, P150, P200, N210 and P33O separately. Six waveforms had been evoked in depression. According to latency, their name was P80, N150, P150, P200, N210 and P33O separately. There was N210 specificity for human faces in T6, 01 and 02 electrode in both of groups, but N210 was not found in depression group. The feature of ERP waveform in anterior and central region was different from posterior region.RESULTS: In control group, the significant interaction effect between facial expression and orientation factors was found in right temporal-parietal-occipital areas (340-390 ms). The Significant facial expression effects occurred separately in bilateral parietal-occipital (120-180 ms) and right frontal-partial (160-180 ms) regions, whole scalp (200-280 ms), and prefrontal areas (320-380 ms). The significant facial orientation effects occurred in frontal-parietal region (85-105 ms), whole scalp (125-185 ms), bilateral frontal-partial region (225-265 ms), left temporal, frontal-parietal and prefrontal ones (305-345 ms) correspondingly. In the contrast between depression and control, SDS scale was more, response time was longer and accuracy was down in major depression. There was significant difference betweenthem. The ERP demonstrated similar waveform (N210) between depression patients and normal group, but its components were different. N210 was record on T6> OK 02 in both groups, however, not record on T5 in major depression. As revealed by group t-Test, there was significant difference between their "emotion-wave". In "a" group, EP occurred separately in left temporal region (260-275 ms), whole scalp (305-335 ms), and left temporal-partial (395-410 ms) region and spread from left occipital region to anterior (485-590 ms) region. In "b" group, EP occurred separately in left fronto-temporal region (140-245 ms) and spread to posterior and right region, in left, whole scalp (290-385 ms) and spread to left fronto-temporal region, temporal region (410-440 ms), in posterior occipital region (470-485 ms). In "c" group, EP occurred separately in right temporo-occipital region (410-450 ms) and spread to anterior and left region. In "d" group, EP occurred separately in anterior and right temporo-occipital region (190-235 ms), left fronto- partial (265-295 ms) region, left temporal-partial (445-505 ms) region.CONCLUSION: The spatiotemporal patterns of ERP suggested the information processing of both facial expression and orientation involved extensive brain regions dynamically. The findings provided electrophysiological evidence that dysfunction ofleft cerebral cortex existed in patients with major depression.
|
PDF Full Text Request