محرک های پویا بازنمایی قطعی از حالت چهره در آمیگدال را نشان می دهد
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|37802||2014||6 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Neuropsychologia, Volume 56, April 2014, Pages 47–52
Abstract Face-selective regions in the amygdala and posterior superior temporal sulcus (pSTS) are strongly implicated in the processing of transient facial signals, such as expression. Here, we measured neural responses in participants while they viewed dynamic changes in facial expression. Our aim was to explore how facial expression is represented in different face-selective regions. Short movies were generated by morphing between faces posing a neutral expression and a prototypical expression of a basic emotion (either anger, disgust, fear, happiness or sadness). These dynamic stimuli were presented in block design in the following four stimulus conditions: (1) same-expression change, same-identity, (2) same-expression change, different-identity, (3) different-expression change, same-identity, and (4) different-expression change, different-identity. So, within a same-expression change condition the movies would show the same change in expression whereas in the different-expression change conditions each movie would have a different change in expression. Facial identity remained constant during each movie but in the different identity conditions the facial identity varied between each movie in a block. The amygdala, but not the posterior STS, demonstrated a greater response to blocks in which each movie morphed from neutral to a different emotion category compared to blocks in which each movie morphed to the same emotion category. Neural adaptation in the amygdala was not affected by changes in facial identity. These results are consistent with a role of the amygdala in category-based representation of facial expressions of emotion.
. Introduction Transient changes in facial musculature that signal current emotional state are critical for effective social interactions. A prominent model of face perception has proposed that a neural pathway from the occipital face area (OFA) to the posterior superior temporal sulcus (pSTS) is involved in processing transient facial signals such as facial expression and eye gaze. In this model the STS is thought to have reciprocal connections with the amygdala which is recruited for further analysis of facial expression (Haxby, Hoffman, & Gobbini, 2000). The sensitivity of the STS and amygdala to a range of facial expressions has been demonstrated across a variety of experiments (Adolphs et al., 1994, Adolphs et al., 1999, Andrews and Ewbank, 2004, Baseler et al., 2013, Engell and Haxby, 2007, Harris et al., 2012 and Narumoto et al., 2001). >However, relatively little is known regarding how facial expression is encoded in these regions. Models of facial expression perception have debated whether facial expressions are represented as belonging to discrete categories of emotion or as gradations along continuous dimensions (see Bruce & Young, 2012). Although usually treated as incompatible opposites there is evidence for both accounts. Evidence for categorical perception of expression is shown by the consistency with which basic emotions are recognized (Ekman, 1972) and by the increased sensitivity to changes in facial expression that alter the perceived emotion (Calder et al., 1996 and Etcoff and Magee, 1992). In contrast, continuous or dimensional models are better able to explain the systematic confusions that occur when labeling facial expressions (Woodworth & Schlosberg, 1954). Continuous models can also account for the fact that we are readily able to perceive differences in intensity of a given emotional expression (Calder et al., 1997 and Young et al., 1997) and for variation in the way that basic emotions are expressed (Rozin, Lowery, & Ebert, 1994). Previously, we offered evidence supporting a synthesis of the above accounts at the neural level by demonstrating that expression is represented in the brain in both a categorical and a continuous manner (Harris et al., 2012). Specifically, by morphing static images of faces to create equal physical changes between images that either fell within the same emotion category or crossed the boundary between different emotion categories, we showed that the amygdala is more sensitive to between than within-category changes (showing a more categorical representation of facial expression) whereas the pSTS is equally sensitive to within and between-category change (indicating a more continuous representation). In this current study, we aimed to further explore the categorical representation of expression in the amygdala using dynamic stimuli. Dynamic changes in facial expression can provide a stringent test of categorical representations, as dynamic movies necessarily incorporate continuous transient changes in expression, and many of these changes need to be disregarded in order to assign dynamic expressions into discrete categories. We used short movies that always showed a change from a neutral resting expression to an intense emotional expression. These movies were created by animating morphed images of facial expressions of basic emotions from the Ekman and Friesen (1976) series. We then used a block fMR-adaptation design to compare neural responses to blocks involving a series of these short movies in which the final expressions were either the same (e.g. all fear) or different (mixed emotions). So within a block participants we saw a series of movies which displayed a dynamic change from a neutral expression to the apex of an emotion. In the same-expression change conditions the same change in expression was displayed across all movies (e.g. all neutral to fear). In the different expression conditions each movie had a different facial expression change (neutral to fear, neutral to disgust, neutral to happy etc). These same and different expression blocks could be presented with either the same or different facial identity. In the same identity conditions each movie would show the same person across the block, whilst in the different identity conditions each movie would show a different person. This design incorporates contrasts that provide substantial criteria for a category-based response to moving expressions. A neural region using a predominately categorical representation of expression should show a greater response to the different compared to the same change in expression conditions, as these conditions involve a change in the emotion category. Moreover, a region showing a response based primarily on emotional categories should also be relatively insensitive to changes in facial identity. However, if a region does not represent expression into emotion categories it should respond equally to the same and different expression blocks, because all of the movies are based on morphed sequences of images that undergo continuous changes. From Harris et al.’s (2012) results with static expressions, we predicted that the amygdala, but not the pSTS, would demonstrate a categorical representation of expression.
نتیجه گیری انگلیسی
. Results The location of all face-selective regions is shown in Fig. 1 and Table 1. The localiser was able to identify a posterior part of right STS and a region in the right amygdala which responded more to faces than to non-face stimuli. A further two regions, the OFA and FFA, also showed a preferential response to faces and were identified in both the left and right hemispheres. To determine whether there was any difference in the neural response across the hemispheres in the adaptation experiment, we conducted a 2×2×4 ANOVA with Hemisphere (left, right), Region (OFA, FFA) and Condition (same-expression change, same-identity, same-expression change, different-identity, different-expression change, same-identity, different-expression change, different-identity) as the main factors (participants in which the OFA and FFA could only be identified unilaterally (see Table 1), were not included in this ANOVA). There was no main effect of Hemisphere (F(1,12)=2.73, p=0.13). There was also no significant Hemisphere⁎Condition (F(3,36)=0.45), Hemisphere⁎Region (F(1,12)=0.25) or Hemisphere⁎Region⁎Condition (F(3,36)=1.17, p=0.34) interactions. Accordingly, for participants that demonstrated bilate ral OFA and FFA, the neural responses were combined across hemispheres. Location of regions that were more responsive to faces compared to non-face ... Fig. 1. Location of regions that were more responsive to faces compared to non-face stimuli in the Localiser scan. MNI coordinates (mm) of slices: x=46, y=−52, z=−24. FFA: fusiform face area, OFA: occipital face area, pSTS: posterior superior temporal sulcus, and AMG: amygdala. Figure options Table 1. MNI coordinates (mm) of face-selective regions. Coordinates for the center of gravity were averaged across all participants. Standard error is reported in parenthesis. Region n x y z FFA 19 L 18 −41 (1.0) −54 (1.5) −21 (1.0) R 19 43 (1.1) −55 (3.2) −22 (1.6) OFA 19 L 15 −39 (2.1) −84 (1.5) −16 (0.9) R 19 43 (1.6) −80 (2.0) −14 (1.2) STS 18 R 53 (1.7) −51 (2.6) 4.7 (1.0) Amygdala 16 R 17 −9 −18 Table options Next, we determined whether there was any difference between the response in the face-selective regions to dynamic changes in facial expression and to changes in identity. A 4×2×2 ANOVA with Region (pSTS, amygdala, FFA, OFA) Expression (same, different) and Identity (same, different) as the main factors, revealed significant main effects of Expression (F(1,14)=7.30, p=0.02) and Region (F(3,42)=63.71, p<0.0001) and a marginal, but not statistically significant effect of Identity (F(1,14)=4.35, p=0.06). There was also a significant interaction between Region×Expression (F(3,42)=3.06, p=0.04).The main focus of the analysis is the pSTS and the amygdala as these regions have been previously implicated in the processing of facial expression ( Harris et al., 2012 and Haxby et al., 2000). To further investigate whether these two regions demonstrated a different pattern of response, we conducted a 2×2×2 ANOVA with the factors Region (amygdala, pSTS), Expression (same, different) and Identity (same, different). This revealed a significant main effect of Region (F(1,14)=48.42, p<0.001) and Expression (F(1,14)=5.67, p=0.03) but not identity (F(1,14)=0.01) There was also a significant Region⁎Expression interaction (F(1,14)=6.99, p=0.02), suggesting a dissociable representation of expression in these regions. Therefore, to investigate how the response to dynamic changes in facial expression differed between the face-selective regions, the patterns of response in the face-selective regions of interest were considered individually. Fig. 2c shows the peak responses in the posterior part of the right STS. A 2×2 ANOVA with the factors Expression (same, different) and Identity (same, different) revealed no significant effect of Expression (F(1,17)=0.66), or Identity (F(1,17)=0.20). There was also no significant Expression⁎Identity interaction (F(1,17)=1.97, p=0.18). In contrast, the amygdala was sensitive to blocks of faces in which the dynamic change in expression varied across the block. A 2×2 repeated measures ANOVA found a significant main effect of Expression (F(1,15)=5.10, p=0.04) but not Identity (F(1,15)=0.23). There was no significant interaction Expression⁎identity (F(1,15)=0.08). The main effect of Expression was due to the bigger response to the different-expression conditions compared to the same-expression conditions (different expression: 0.19%, same expression: 0.05%). This pattern held regardless of whether the blocks showed the same or different identities. Example stimuli and neural responses to dynamic changes in facial expression: ... Fig. 2. Example stimuli and neural responses to dynamic changes in facial expression: (a) alternate frames from an example neutral to happiness movie. (b) Example of the sequence of movies within a block from the four experimental conditions: (upper, left) same-expression, same-identity; (upper, right) same-expression, different-identity; (lower, left) different-expression, same-identity; (lower, right) different-expression, different-expression. (c) Peak responses to the different conditions in the pSTS and amygdala. The results show that the pSTS was responsive to all conditions, consistent with a continuous representation of facial expression. However, the amygdala was more sensitive to a series of expression movies which displayed different changes in the emotion category, demonstrating a more categorical representation of emotion. Figure options The responses to the different conditions in the FFA are shown in Fig. 3. A 22×2 ANOVA revealed no significant main effect of Expression (F(1,18)=0.44) but there was a main effect of Identity (F(1,18)=6.37, p=0.02). There was a borderline but not significant Expression⁎Identity interaction (F(1,18)=3.48, p=0.08). The main effect of Identity was due to a bigger response to the different-identity conditions compared to the same-identity conditions. The OFA showed a similar pattern of response to that found in FFA. There was no significant effect of Expression (F(1,18)=0.73), but there was a significant effect of Identity (F(1,18)=10.15, p=0.01). There was also a significant Expression⁎Identity interaction (F(1,18)=4.47, p=0.05). The interaction was due to a significantly bigger response to different-identity condition compared to same-identity condition for the same-expression change (t(18)=3.31, p=0.004) but not for the different-expression change conditions (t(18)=1.59, p=0.39). Peak responses to the different stimulus conditions in the OFA and FFA. Both ... Fig. 3. Peak responses to the different stimulus conditions in the OFA and FFA. Both regions showed a significant increase in response to changes in expression and to changes in identity. Figure options In summary, the results from this experiment reveal that the amygdala was sensitive to the emotion category, with a greater response to blocks of movies that varied in the category of emotion compared to blocks of movies displaying the same change in emotion. This is consistent with a more categorical representation of expression. This is dissociable form the response in the pSTS which did not discriminate between blocks with same change and different changes in expression.