Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

On the neural networks of empathy: A principal component analysis of an fMRI study

On the neural networks of empathy: A principal component analysis of an fMRI study Background: Human emotional expressions serve an important communicatory role allowing the rapid transmission of valence information among individuals. We aimed at exploring the neural networks mediating the recognition of and empathy with human facial expressions of emotion. Methods: A principal component analysis was applied to event-related functional magnetic imaging (fMRI) data of 14 right-handed healthy volunteers (29 +/- 6 years). During scanning, subjects viewed happy, sad and neutral face expressions in the following conditions: emotion recognition, empathizing with emotion, and a control condition of simple object detection. Functionally relevant principal components (PCs) were identified by planned comparisons at an alpha level of p < 0.001. Results: Four PCs revealed significant differences in variance patterns of the conditions, thereby revealing distinct neural networks: mediating facial identification (PC 1), identification of an expressed emotion (PC 2), attention to an expressed emotion (PC 12), and sense of an emotional state (PC 27). Conclusion: Our findings further the notion that the appraisal of human facial expressions involves multiple neural circuits that process highly differentiated cognitive aspects of emotion. Introduction on correctly recognizing and reacting to rapid fluctuations Human emotional facial expressions contain information in the emotional states of others [4,5]. This capability may which is essential for social interaction and communica- have played a key role in our ability to survive and evolve tion [1-3]. Social interaction and communication depend [6]. Page 1 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 In the research field of psychological childhood develop- Each psychological state may be associated with distinct ment the competence to detect, share, and utilise cogni- neural networks containing cortical areas which interact tive patterns and emotional states of the other was within and across neutral networks. The interactions fos- conceptualized as mentalizing [7]. The important aspect is ter an exponentially complex environment within which that the emotion expressed in someone else's face is not the processes driving emphatic responses occur. While the only detected but also valuated from the subjective per- processes have been explored to some extent, much work spective of the observer. In fact, the appraisal of and the is needed to reveal the precise nature of the complex inter- resonance with an emotion observed in somebody else is action among the constituents of empathy. central to the concept of empathy [8-11]. Accordingly, empathy has been suggested to involve dis- We can acquire two important sources of information tinct, distributed neural networks. Such distributed net- upon perceiving a face: the identification and emotional works demand the application of a network analysis, expression of the individual [12-15]. Support for distinct which subjects voxels of the image matrix to a multivariate networks of facial identification and emotional recogni- rather than a univariate analysis. This feature allows a net- tion have been provided by studies of those with trau- work analysis to overcome two significant limitations of matic brain injuries [16], electroencephalograph (EEG) univariate analyses that are based on categorical compari- studies [17,18], magnetoencephalography (MEG) studies sons: they are unable to distinguish regional networks [19], and also by studies examining the influence of emo- because the constituent voxels may not all change at the tional expression on familiar faces [20]. EEG studies have defined level of significance and may also show areas of also hypothesized that these two networks operate in a activation which are unrelated to the phenomenon being parallel system of recognition in which emotional effects studied [27]. and structural facial features are simultaneously identified through distinct neural networks [8,9]. The network analysis applied to the blood oxygen level dependent (BOLD) signals in this functional magnetic When inferring the emotional expression of another, we resonance imaging (fMRI) study is a principal component are automatically compelled to compare our assessment analysis (PCA), which decomposes the image matrix into of their emotional state with our own. A key component statistically uncorrelated components. Each component of perspective taking and emphatic experiences is the abil- represents a distinct neural network, and the extreme ity to compare the emotional state of oneself with another voxel values of a component image, its nodes. Thus, the [8-11,21]. However, a distinct separation between first component images map the functional connectivity of and third-person experiences is necessary [8,22]. People constituent regions activated during neural stimulation. are normally able to correctly attribute emotional states Previous studies have proposed and verified the hypothe- and actions to the proper individual, whether they are our sis of functional connectivity in regions-of-interest [28] own or someone else's [23]. Confusing our emotional and voxel-based analyses [29-31]. In contrast to regions state with another would vitiate the function of empathy that show enhanced metabolic or blood flow levels corre- and cause unnecessary emotional distress and anxiety lated with mental states [31], the networks deduced by [24]. PCA incorporate no a priori assumptions regarding the neural stimulation. Therefore, a number of distinct psychological states func- tioning in concert may be necessary in order for a proper Virtually unexplored until now, the neural networks asso- emphatic experience to occur [25]. They may include the ciated with recognizing and empathizing with human evaluation of emotion of the self, evaluation of another's facial expressions of emotions are here described using emotion, comparing those emotions, and anticipating PCA. and reacting to our own or another's emotion, among others. In accord with this line of thought, Decety and We attempt to elucidate the functional neural networks Jackson [26] have proposed three major functional com- central to the processes of recognizing and empathizing ponents of emphatic experiences. First, the actions of with emotional facial stimuli recorded in fMRI acquisi- another person automatically enact a psychological state tions of healthy subjects. in oneself which mirror those actions and create a repre- sentative state which incorporates the perceptions of both Statistical testing identified four principal components the self and another. Secondly, there must be a distinct (PC) as relevant neural networks. The correlation of the separation between the perception of the self and the subjects' scores of emotional experience with the PC's fur- other person. Finally, the ability to cognitively assume the ther supported their functional relevance. Our analysis perception of another person while being aware of self/ shows the coordinated action of brain areas involved in other separation is important. the processing of visual perception and emotional Page 2 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 appraisal underlying the facial expressions of human Visual stimuli emotions. These regions, including pre-frontal control From photographs of facial affect [38], we selected for areas and occipital visual processing areas presumably presentation those with happy, sad and neutral facial constitute nodes of the networks implicated in appraising expressions found to be correctly identified in more than other people's emotion in their facial expressions. 90 percent of raters; 14 happy, 14 sad, and 14 neutral faces were used. Only faces which had been found to be cor- Methods rectly identified in more than 90 percent of raters [38] Subjects were used to ensure that the subjects internally generated Fourteen healthy, right-handed subjects (28.6 +/- 5.5 the corresponding emotion. This corresponded to a simi- years; 7 men, 7 women) participated in this study. The lar approach of a recent study [39]. As control stimuli we subjects had normal or corrected to normal vision. Right- produced a number of scrambled images from these pho- handedness was assessed using the Oldfield's question- tographs equal to the number of intact faces. All images naire [32]. In addition, the subjects' emotional compe- were digitized and controlled for luminescence. tence was tested with the German, 20-item version of the Experimental task design Toronto Alexithymia Scale (TAS-20) [33-35]. None of the participating subjects were classified as alexithymic (mean Faces were presented for durations ranging between 300 TAS-20 sum score 34.14 +/- 6.26, range 23). Subjects were and 500 ms, since this short presentation time is sufficient also evaluated with the Beck Depression Inventory [36] for conscious visual perception avoiding habituation [40]. and the scales of emotional experience (SEE) [37]. The presentation time of the faces was jittered to enhance the detection of the stimulation-related BOLD activity changes in this event-related fMRI study. Thereafter, Sc Figure 1 hematic illustration of the experimental design Schematic illustration of the experimental design. Page 3 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 scrambled faces were presented for durations between 11 subject was co-registered to the mean image of the func- and 12 s (Figure 1). This long second stimulation period tional images and also normalized to the MNI-space. was chosen to provide the subjects sufficient time to engage in appraising the emotional facial expression seen For each of the three instruction sets, the happy, sad and in the first stimulation interval and to allow changs of skin neutral face presentations were modelled as well as the conductance to occur [40]. The faces were presented in corresponding scrambled faces. The models employed the random order on a laptop connected to a projector (LCD haemodynamic response function provided by SPM2. data projector, VPL-S500E, Sony, Toyko, Japan) on a Data were temporally filtered using a Gaussian low-pass screen which was placed approximately 50 cm from the filter of 4 s and a high-pass filter of 100 s. All data were mirror in the head coil. Immediately before each fMRI scaled to the grand mean. Realignment parameters as scan the subjects were instructed to view the faces in a determined in the realignment step were used as con- mind set according to one of the following cognitive founding covariates. The duration of all events was mod- instructions: a) identify the emotion expressed in the faces elled with 4 s for face presentations and 8 s for the delay (RECOGNIZE), b) empathize with the emotion expressed period. The repeated condition images of the 18 experi- in the faces (SHARE EMOTION), c) count the earrings in mental conditions were averaged for each of the 14 sub- the faces shown (control condition: DETECT EARRINGS). jects, yielding a total of 252 averaged condition images. Each instruction was given twice in random order across Data were modelled using the canonical haemodynamic the subjects, yielding six separate scans per subject. Thus, response function provided by SPM2. The duration of all the visual stimuli were identical in the different experi- events was modelled explicitly with 4 s for face presenta- mental conditions, but the visual information to be proc- tions and 8 s for the delay period. To identify the brain essed differed according to the instructions. Since each areas related to viewing the faces a comparison with view- condition was repeated twice, 28 happy, 28 sad and 28 ing the scrambled faces was calculated. Only areas with a neutral faces were presented. The presentation was done p < 0.05 corrected at cluster level with a cluster threshold in random order across the subjects to counteract possible > 20 voxels were accepted (Table 1). sequence or habituation effects. After each condition, the subjects were debriefed about how well they could per- The PCA employed in house software of which some form the task. modules were adapted from SPM2. Extracerebral voxels were excluded from the analysis using a mask derived Functional magnetic resonance imaging from the gray matter component yielded by segmentation The subjects lay supine in the MRI scanner and viewed the of the high resolution anatomical 3D image volume into faces on a screen via a mirror which was fixed to the head gray matter, white matter and cerebrospinal fluid using coil. Image presentation was controlled by a TTL-stimulus the segmentation module of SPM2. Voxel values of the coming from the MRI-scanner as described in detail else- segmented image ranged between 0 and 1; the mask where [41]. Scanning was performed on a Siemens Vision included only those exceeding 0.35, excluding most of 1.5 T scanner (Erlangen, Germany) using an EPI-GE white matter. Calculation of the residual matrix is the first sequence: TR = 5 s, TE = 66 ms, flip angle = 90°. The whole step. From a matrix whose rows corresponded to the 252 brain was covered by 30 transaxial slices oriented parallel conditons and columns to the 180 thousands voxels in a to the bi-commissural plane with in-plane resolution of single image volume the mean voxel value of each row 3.125 × 3.125 mm, slice thickness of 4 mm, and interslice was subtracted from each element as was the mean voxel gap of 0.4 mm. Each acquisition consisted of 255 vol- value of each column subtracted from each element. umes. The first 3 volumes of each session did not enter the Thereafter, the grand mean of all voxel values in the orig- analysis. A high-resolution 3D T1-weighted image (TR = inal matrix was added to each element. The result of this 40 ms, TE = 5 ms, flip angle = 40°) consisting of 180 sag- normalization procedure is the residual matrix for which ittal slices with in plane resolution of 1.0 × 1.0 mm was the row, column and grand means vanish. Using the sin- also acquired for each subject. gular value decomposition implemented in Matlab, the residual matrix was then decomposed into 252 compo- Data analysis nents, consisting of an image, an expression coefficient, Image data were analysed with SPM2 (Wellcome Depart- and an eigenvalue for each component. The eigenvalue ment of Cognitive Neurology, London, UK; http:// was proportional to the square root of the fraction of var- www.fil.ion.ucl.ac.uk/spm). Images were slice-time cor- iance described by each component, while the expression rected, realigned, normalized to the template created in coefficients described the amount that each subject and the Montreal Neurology Institute (MNI), and spatially condition contributed to the component. The principal smoothed with a 10 × 10 × 10 mm Gaussian filter. The components (PCs) were ranked according to the propor- normalization step resampled the images to a voxel size of tion of variance that each component explains, i.e, PC 1 2 × 2 × 2 mm. The anatomical T1-weighted image of each explains the greatest amount of variance. The expression Page 4 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 Table 1: Cerebral activations related to viewing emotional face expressions as compared with viewing scrambled faces Anatomical location Coordinates Brodmann area Cognitive instruction (Task) x y z Recognize Empathize Object detection Lingual gyrus R 28 -72 -10 BA 18 + Fusiform gyrus R 36 -69 -10 BA 19 + + Cuneus R 28 -78 16 BA 31 + + Superior frontal gyrus L -3 22 52 BA 6 + + + Inferior frontal gyrus L -49 2 20 BA 44 + + + Inferior frontal gyrus R 55 8 16 BA 44 + + + Middle frontal gyrus R 43 29 30 BA 9 + + Supramarginal gyrus L -46 -49 28 BA 42 + Parietal operculum R 50 -24 10 BA 40 + Sup. temporal gyrus R 34 16 -6 BA 38 + Hypothalamus L -3 -6 -11 + Cerebellum L -14 -60 -22 + Activations in categorical analysis with SPM2 at p < 0.05 FDR with clusters > 20 voxels. The coordinates (x, y, z; mm) of the peak activations (+) are listed in stereotactic space [43]. L refers to left hemisphere; R refers to right hemisphere. coefficients and voxel values of a PC were orthonormal Results and their orthogonality reflected the statistical independ- Before the fMRI experiment the subjects' capacity to expe- ence of the PCs. The PC image displays the degree to rience emotions was tested with the German, 20-item ver- which the voxels covaried in each PC, their voxel values sion of the Toronto Alexithymia Scale (TAS-20 [35]). It (loadings) ranged between -1 and 1. was found that the 14 participating subjects had a normal mean TAS-20 sum score (34.1 +/- 6.3, range 23). This In order to provide a neurophysiological interpretation of indicated that each of the subjects had a high capacity of the components, statistical tests, e.g. unpaired t-tests and introspection and emotional awareness. In the fMRI ses- tests of correlation (Pearson) were applied to the expres- sion the subjects stated that they could readily identify the sion coefficients. The formal criteria for relevant PCs were: seen emotion and generate the corresponding emotion (1) the statistical tests identified the condition differenti- internally as instructed. They detected 92 +/- 0.2 percent ating PCs at a significance level of p < 0.001, and (2) the of the faces wearing earrings. PCs fulfilled the Guttman-Kaiser criterion, the most com- mon retention criterion in PCA [42] in which PCs associ- The categorical analysis showed that viewing the emo- ated with eigenvalues of the covariance matrix larger in tional face expressions as compared with viewing the magnitude than the average of all eigenvalues are scrambled faces resulted in activations of right visual cor- retained, implying in this analysis that they ranked among tical areas and bilaterally the inferior frontal and superior the first 61 PCs. The voxels describing the nodes of a neu- frontal gyrus (Table 1). Recognizing emotional facial ral network associated with a relevant PC image volume expressions showed the most extensive activation pattern st fulfilled the conditions that the voxel values lie in the 1 involving also the hypothalamus, the left supramarginal th percentile or the 99 percentile of the volume's voxel gyrus, cortical areas at the right temporal parietal junction, value distribution, and that the voxels belong to clusters and the left hypothalamus (Table 1). Empathizing with of greater than 50 voxels. the seen emotion as compared with object detection (con- trol condition) resulted in one activation area which The anatomical locations of the peak activations and of occurred in the left inferior frontal gyrus. Note, that no the coordinates of the maximal PC loadings of the signif- activation occurred in the anterior prefrontal or orbitof- icant PCs are reported in Talairach space [43]. A freely dis- rontal cortex. tributed Matlab script [44] effected the transformation from MNI space. The network analysis revealed that out of the total of 61 retained PCs four differentiated the experimental condi- Correlation of the expression coefficients of the signifi- tions as found by formal statistical testing (Table 2); for cant PCs with the TAS-20, the Beck Depression Inventory, the 11 statistical tests described below, the probability and the SEE scales was conducted using a Pearson two- threshold corrected for multiple comparisons is p < 0.001. tailed correlation (SPSS for Windows, Version 12.0.1.). PC1 explained 16.8 percent of the variance and distin- The significance level of the correlations was set to p < guished between viewing the happy, sad and neutral faces 0.01. from viewing the scrambled faces during RECOGNIZE, Page 5 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 Table 2: Cerebral circuits in processing of emotional face expressions Coordinates Anatomical location Broadmann area Cluster size XY Z PC1: Face identification (16.8%) Positive loadings 4 53 4 Superior frontal gyrus R BA 10 248 -4 48 -2 Anterior cingulate L BA 32 81 1 31 -7 Anterior cingulate R BA 32 76 38 26 45 Middle frontal gyrus L BA 8 66 -50 -63 31 Angular gyrus L BA 39 53 44 -61 28 Supramarginal gyrus R BA 39 396 Negative loadings -21 -77 -10 Lingual gyrus L BA 18 306 28 -72 -10 Lingual gyrus R BA 18 314 34 -70 15 Cuneus R BA 31 428 -6 -75 44 Precuneus L BA 7 54 PC2: Identification of expressed emotion (4.7%) Positive loadings -30 -74 -10 Fusiform gyrus L BA 19 158 36 -69 -13 Fusiform gyrus R BA 19 252 40 -87 8 Middle occipital gyrus R BA 19 60 6 3 59 Superior frontal gyrus L BA 6 87 -49 2 20 Inferior frontal gyrus L BA 44 65 -3 -6 -11 Hypothalumus L 50 4 -6 -11 Hypothalumus R 63 Negative loadings -5 -73 46 Precuneus L BA 7 51 2 -63 52 Precuneus R BA 7 71 -63 -10 0 Superior temporal gyrus L BA 22 66 -2 4 6 Thalamus L 259 7 -3 6 Thalamus 300 -14 -30 -22 Pons L 80 50 -60 -27 Neocerebellum R 71 PC 12: Attention to expressed emotion (1.6%) Positive loadings 4 -86 17 Cuneus R BA 18 230 54 8 36 Middle frontal gyrus R BA 9 115 55 17 16 Inferior frontal gyrus R BA 44 56 -51 -35 -2 Middle temporal gyrus L BA 22 64 59 -56 3 Middle temporal gyrus R BA 22 119 8-8 12 Thalamus R 377 Negative loadings -12 -77 48 Precuneus L BA 7 117 8 -69 48 Precuneus R BA 7 68 -42 -48 56 Inferior parietal lobule L BA 40 128 -32 60 6 Middle frontal gyrus L BA 10 246 43 29 30 Middle frontal gyrus R BA 9 132 PC 27: Sense of emotional state (0.9%) Positive loadings -48 -57 -6 Middle occipital gyrus L BA 37 68 -28 52 21 Middle frontal gyrus L BA 10 153 -550 Caudate L 66 -13 -18 -25 Pons L 131 -1 -47 -36 Medulla oblongata 192 Negative loadings -42 -59 -17 Fusiform gyrus L BA 37 58 -46 -49 28 Supramarginal gyrus L BA 42 86 51 -34 50 Inferior parietal lobule R BA 40 54 55 -19 16 Parietal operculum R BA 40 53 -55 -40 8 Superior temporal gyrus L BA 22 162 38 8 -26 Superior temporal gyrus R BA 38 59 Shown are the relevant principal components (PC) and their designations. Indicated in the parentheses beside each PC entry is the amount of th variance accounted for by that PC. The coordinates (x, y, z; mm) are listed in stereotactic space [43]. Only those voxels with values lying in the 99 st percentile (positive loadings) or the 1 percentile (negative loadings) and in cluster with sizes greater than 50 voxels were included. L refers to left hemisphere; R refers to right hemisphere. Page 6 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 SHARE EMOTION and DETECT EARRINGS. Accordingly, subjects exhibited score values suggestive of depression, PC1 represented a neural network associated with face this correlation was obtained in the normal range of the identification. The areas with the positive loadings Beck's depression inventory. Nevertheless, a more nega- included the right dorsolateral and superior frontal cortex, tive emotional experience was related to recognizing neu- the left anterior cingulate, and bilaterally the inferior pari- tral faces. Further, the TAS-20 scores correlated negatively etal region. The areas with the negative loading included with PC2 expression coefficients computed for viewing bilaterally the lingual gyrus, the precuneus, and the happy and neutral faces in the control condition DETECT cuneus, which are areas involved in higher order process- EARRINGS. Note, that the TAS-20 classified all subjects as ing of visual information (Table 2, Figures 2 and 3). highly emotionally sensitive. Thus, this correlation sug- gests that the more sensitive to processing of emotion our PC2 explained 4.7 percent of the variance and differenti- subjects were, the more they were so during implicit ated viewing the happy, sad, and neutral faces from view- processing of faces. The SEE-scale score values related to ing scrambled faces during the RECOGNIZE and SHARE experience of emotional control correlated negatively EMOTION conditions (Table 2, Figures 2 and 3). There- with PC 1 expression coefficients computed for viewing of fore, PC2 was expected to depict a neural network associ- sad faces in the RECOGNIZE condition (Figure 5). A sim- ated with the identification of an expressed emotion. In ilar correlation was found for PC27 expression coeffi- fact, the areas with positive loadings included bilaterally cients computed for viewing of neutral faces in the the fusiform gyrus, the right middle occipital gyrus, the RECOGNIZE condition. This suggested that recognizing right superior frontal gyrus, and the left inferior frontal sad and neutral faces was most pronounced in the subjects gyrus. The areas with negative loadings included bilater- whose scores indicated relatively impaired emotional ally the precuneus, the left superior temporal gyrus, as control. Finally, the SEE-scale score values of experience of well as the thalamus, pons, and neocerebellum (Table 2). self-control correlated with PC 12 expression coefficients The relative localization of the cortical areas in the occip- computed for the SHARE EMOTION condition after view- ital cortex and the superior frontal gyrus involved in PC1 ing happy and neutral faces (Figure 6). This suggested that and PC2 is illustrated in Figure 3. subjects with a high level of self-control most strongly empathized with happy and neutral faces. Thus, viewing PC12 explained 1.6 percent of the variance and reflected sad faces may have impaired the subject's perception of the contrast of the conditions SHARE EMOTION and emotional control, while processing the happy face DETECT EARRINGS as well as the contrast of the condi- expression appears to have improved it (Table 3). In con- tions RECOGNIZE and DETECT EARRINGS during and trast, the processing of neutral faces was related to a rela- after viewing happy, sad, or neutral faces. The compari- tively negative emotion, possibly due to the ambiguous sons showed that PC12 represented a neural network character of the neutral faces (Table 3). associated with the subjects' attention to an expressed emotion. Areas involved in this network included the Discussion right cuneus, bilaterally the precuneus, middle frontal and The novel finding of this event-related fMRI study is that temporal gyrus, and left inferior parietal lobule (Table 2). recognizing and empathizing with emotional face expres- sions engaged a widespread cortical network involving PC27 explained 0.9 percent of the variance and repre- visual areas in the temporal and occipital cortex which are sented the contrast of the conditions SHARE EMOTION known to be involved in face processing [12,45-62]. Pre- with RECOGNIZE during and after viewing happy, sad, vious fMRI studies have identified several brain regions and neutral faces. We hypothesized, therefore, that this PC that are consistently activated when perceiving facial stim- characterizes the neural network subserving the sensation uli. These regions include the fusiform gyrus, which is of the emotional state associated with empathy. The often referred to as the "fusiform face area" [45,49-55,61- regions involved were the left fusiform and middle occip- 63], a face-selective region in the occipito-temporal cortex ital gyrus, the left middle frontal gyrus, bilaterally the infe- [45-48,56-60], and the superior temporal sulcus rior parietal lobule and the superior temporal gyrus (Table [12,13,48,52]. 2, Figure 4). Subcortical structures, such as the left caudate and brain stem were also involved. Because the involved cortical areas are active within and across multiple neural networks, the difficulty of assign- Correlation of the PC expression coefficients with the ing a single function to a cortical area becomes apparent behavioral scales of emotional processing yielded the fol- [2]. In fact, disagreement over assigning functions to cor- lowing observations (Table 3). Beck's depression inven- tical areas does exist, with some arguing that function may tory subject scores correlated significantly with the PC2 be tied to elements of experiment design [64]. However, expression coefficients computed for the RECOGNIZE inability to agree over the specific function of cortical condition after viewing neutral faces. Since none of the areas would further suggest that a description of "face net- Page 7 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 B pose SPM2 in a sa in tion ing) negative loading; pre-SM positiv Figure 2 r P ain areas involved in C of s an d on the 1 ( e d in P e u cu p xn e pr e rio C g u canonical sin ession is 2 ( ttal plane sh r: fr re pon re d loadin cune tal – ne t gh gati y A, hy us, thalamus: blu e PC1 and PC2 rus g) g owing the areas involved le-sub ve loading; anterior por- : gre pothalamus: yellow – e ject M n – po Rs image of superim- ie tiv – e load- Brain areas involved in the PC1 and PC2 superim- posed on the canonical single-subject MR image of SPM2 in a sagittal plane showing the areas involved in PC1 (cuneus: red – negative loading; anterior por- tion of superior frontal gyrus: green – positive load- ing) and in PC2 (precuneus, thalamus: blue – negative loading; pre-SMA, hypothalamus: yellow – positive expression loading). Brain on the canonical single-subje an axia in the fusi ing) relative Figure 3 ar l plane showing the la eas form face area in to volved PC1 (red, negative in P C (PC2, yellow, positive load- 1 and P teral position of ct MR im C loa 2 sup d age of SPM2 in ing) erimposed activity Brain areas involved in PC1 and PC2 superimposed work" or "face pathway" may be more accurate than "face on the canonical single-subject MR image of SPM2 in region" [65]. an axial plane showing the lateral position of activity in the fusiform face area (PC2, yellow, positive load- Notably, cortical areas in the anterior lateral and medial ing) relative to PC1 (red, negative loading). Note the prefrontal cortex known to participate in manipulating medial prefrontal involvement in PC1 (green, positive load- and monitoring information and controlling behavior ing). were also involved [26,66]. Multivariate image analysis using PCA permits the characterization of different net- works that include brain areas changing brain activity in ously to be involved in attention [71], and the anterior related to a given task and brain areas contributing to the portion of the superior frontal gyrus has been implicated function without necessarily changing their activity [28- in theory of mind paradigms [26,71]. 31,67]. Specifically, we applied inferential statistical tests to identify the PCs that effectively differentiated between PC2 represented the functional neural network associated the experimental conditions [31,68]. The four thereby with detecting emotional facial expressions, since the net- identified PCs showed more cerebral areas involved than work nodes included bilaterally the fusiform gyrus, corre- in the simple task comparions. These PCs revealed corre- sponding to the so-called fusiform face area. Some studies lations with behavioral data obtained prior to the fMRI have suggested that the fusiform face area mediates the acquisitions, which highlighted their functional rele- lower order processing of simple face recognition [72], vance. while others have implicated it in higher order processing of faces at a specific level [45] including emotional detec- Functional neural networks tion [73] and identity discrimination [58,64]. Also Of the four differentiating PCs, PC1 distingished the neu- included among the nodes were the posterior portion of ral network involved in face indentification, since the the superior frontal and the inferior frontal gyrus, which involved lingual gyrus, cuneus and precuneus have been have been implicated in higher order processing of faces known to be involved in face recognition and visual involving the perception of gaze direction [74], attention processing [68,69]. The lingual gyrus has been linked to to faces [75], eye and mouth movements [56], and empa- an early stage of facial processing which occurs before spe- thy [25,76]. Thus, our analysis, relating the concerted cific identification occurs [70]. Conversely, the angular action of the fusiform and superior frontal gyrus to the gyrus and the anterior cingulate have been shown previ- detection of emotional face expressions, substantiates pre- Page 8 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 and are thought to regulate the emotions, drives, and motivated behavior [2,79,80]. The third statistically relevant PC, PC12, was related to the attentive processing of expressed emotions. Accordingly, the involved cortical areas, e.g. the precuneus, cuneus, middle frontal gyrus, and thalamus have all been linked to tasks related to controlling attention [81-83]. The infe- rior frontal gyrus has been shown to be involved in response inhibition during attention [84], while the mid- dle temporal gyrus has been implicated in attention to facial stimuli [47] and to visual attention in general [81,82]. PC27 appeared to represent a network mediating a 'feel- ing' or sense of an observed emotional state. In fact, the superior temporal gyrus has been shown to play an important role in perceiving self/other distinctions, and, importantly, in experiencing a sense of agency [85,86]. Also, the superior temporal gyrus seems to be important for "theory of mind" capabilities [87-89]. Similarly, the right inferior parietal lobule has been related to an experi- enced sense of agency [85,86], while the temporal parietal junction has been thought to be crucial to larger networks mediating spatial unity of the self and body [89,90], and attention [81,91,92]. In PC27 MR ima loading) Figure 4 volvem supe ge of SPM2 in an axial pla en rimpose t of thed left s on th uperior temporal gyrus in e canonical sin ne (blue, negative gle-subject Involvement of the left superior temporal gyrus in Activity within and across functional networks PC27 superimposed on the canonical single-subject MR image of SPM2 in an axial plane (blue, negative Empathic processes allow individuals to quickly asses the loading). emotional states and needs of other individuals while rap- idly transmitting our own experiences and needs [93]. Thus, emphatic experiences are crucial to rapid and suc- vious work showing that the conjoint activity of the fusi- cessful social interaction [94,95]. The recognition of emo- form gyrus and cortex lining the superior temporal sulcus tional states and cognitive processing of another's are related to higher order processing of facial features emotional expression are critical skillsets utilized by all [45,53,56,58,64,72-74]. We propose that these areas con- humans upon perceiving another person's face. In order stitute a system involved in facial affect processing. to properly behave in a social situation, one must under- stand the context of the situation before an appropriate PC2 also implicated subcortical structures such as the tha- action can be taken. lamus and the hypothalamus bilaterally, and the cerebel- lum. These structures belong to the anterior cingulate – The participation of the prefrontal cortex in the four rele- ventral striatum – thalamus – hypothalamus loop [76-78] vant PCs in our study not only support the view that dif- Table 3: Correlation of behavioral data vs. expression coefficients Toronto alexthymia scale-20 Beck depression inventory Experience of control of emotion Experience of self control R value P value R value P value R value P value R value P value Condition/PC Condition/PC Condition/PC Condition/PC CH 2 -0.670 0.009 RaN 2 0.667 0.009 RS 1 -0.687 0.007 EaH 12 0.757 0.002 CN 2 -0.683 0.007 RN 27 -0.675 0.008 EaN 12 0.778 0.001 C = DETECT EARRINGS (Control), E = SHARE EMOTION, R = RECOGNIZE, H = Happy, S = Sad, N = Neutral, aH = after Happy, aN = after Neutral Page 9 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 icant carrier of emotional information is an important source for these evaluative and appraisal functions. The four neural networks involved in the recognition of faces and the cognitive control of emotions capture the functions that this study aimed to explore. Our results support numerous studies that hypothesize distinct neu- ral networks for recognizing facial features and emotional expressions [11,13-17,96,97]. In particular, although PC 1 and PC 2 differentiated lower-order facial identification and higher-order facial feature processing, the cortical areas within each PC contained brain areas associated with functions effected by the neural networks depicted in the other PCs. Thus, the middle frontal gyrus activated in all four principal components has been shown to be related to inhibition [98], successful recognition of previ- ously studied items [99], and successful error detection, Regr with data lighting the Figure 5 ession of the test scores p functional releva lots of expression coefficients of th nce of e 14 th su ese bje PCs c of PC1 ts high- response inhibition, interference resolution, and behavio- Regression plots of expression coefficients of PC1 ral conflict resolution associated with completing the with data of the test scores of the 14 subjects high- lighting the functional relevance of these PCs. Stroop Color-Word task [100]. Moreover, the PC repre- senting a basic visual function, PC1 included activation of the right anterior cingulate which has been shown to ferent cognitive functions work together to play a role in severely impact the processing of emotional expressions distinct neural networks, but, as well, further the notion [101,102]. Apparent also in PC1, the lingual gyrus and that the medial pre-frontal cortex is the junction point cuneus have been implicated in the higher-order process- where different visual, attentive, emotional and higher ing of emotional expression [103] and the lingual gyrus order cognitive processes come together in order to allow has been implicated in motional information [104]. Inter- for a subjective reaction towards the exterior world [26]. estingly the anterior cingulate has also been thought to The medial-pre-frontal integration of own cognitive con- cognitively monitor the control of response conflict in cepts, implicit affective schemata and empathy-based information processing [105] and in the regulation of anticipation of the possible reactions of important other cognitive and emotional processing [106]. In a case study, individuals allows for an effective selection and adaptive Steeves et al. [61] demonstrated that an intact fusiform planning of one's own actions. The face as the most signif- face area was sufficient for identifying faces, but a lesion in the occipital face area prevented the patient from higher-order processing of faces such as identity, gender, or emotion. Out results are also in accordance with the idea that emo- tion recognition may play a role in the lower order process of face identification, but this role is limited at best and functions mainly as a stepping stone for higher order rec- ognition of emotional expressions. [107]. Although the lingual gyrus in PC1 and the fusiform face area in PC2 are located closely anatomically, our analysis suggests that they are involved in distinct neural networks related to different subfunctions, face identification and detection of emotional facial expressions, repectively, of facial processing. Nevertheless, it is possible that the subjects recognized the emotions implicitly also in the DETECT EARRING condition. However, the cognitive instruction in this study was that during the RECOGNIZE and SHARE Regr with data lighting the Figure 6 ession of the test scores p functional releva lots of expression coefficients of th nce of e 14 th su ese bje PCs c of PC12 ts high- EMOTION conditions the subjects had to appraise the Regression plots of expression coefficients of PC12 emotions explicitly. In fact, this difference resulted in with data of the test scores of the 14 subjects high- lighting the functional relevance of these PCs. additional brain structures involved in the explicit processing conditions. Page 10 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 The involvement of closely adjacent cortical areas in dif- References 1. Leppanen JM, Moulson MC, Vogel-Farley VK, Nelson CA: An ERP ferent subfunctions related to the detection of emotional study of emotional face processing in the adult and infant face expressions highlights the difficulty of assigning brain. Child Develop 2007, 78:232-245. exclusive functional relevance to a particular area and sug- 2. Adolphs R: Neural systems for recognizing emotion. Curr Opin Neurobiol 2002, 12:169-177. gests the multi-functional nature of brain areas which par- 3. Williams KR, Wishart JG, Pitcairn TK, Willis DS: Emotion recogni- ticipate in multiple, interconnected neural networks tion by children with Down syndrome: Investigation of spe- cific impairments and error patterns. Am J Ment Retard 2005, performing lower-order as well as higher-order neural 110:378-392. processing. 4. Blair RJR: Facial expressions, their communicatory functions and neuro-cognitive substrates. Phil Trans R Soc Lond 2003, 358:561-572. Conclusion 5. Eimer M, Homes A: An ERP study on the time course of emo- Implemented in this study using PCA, employment of a tional face processing. NeuroReport 2002, 13:1-5. network analysis helps to elucidate the coordination of 6. Ruby R, Decety J: What you believe versus what you think they believe: a neuroimaging study of conceptual perspective-tak- multiple cortical areas in brain functions. We are able to ing. Eur J Neurosci 2003, 17:2475-2480. identify the participation of multiple neural networks in 7. Fonagy P, Gergely G, Jurist EL: Affect Regulation, Mentalization, and the Development of the Self. John Wiley & Sons. New York; processing highly differentiated cognitive aspects of emo- tion. Ultimately, placing cortical areas within the context 8. Decety J, Lamm C: Human empathy through the lens of social of a particular neural network may be the key to defining neuroscience. ScientificWorldJournal 2006, 6:1146-1163. 9. Preston SD, de Waal FBM: Empathy: its ultimate and proximate the functional relevance of individual cortical areas. This bases. Behav Brain Sci 2002, 25:1-72. more sensitive approach gives us a better picture of the 10. Iacoboni M, Molnar-Szakacs I, Gallese V, Buccino G, Mazziotta JC, Riz- constituent components involved in the processes of rec- zolatti G: Grasping the intentions of others with one's own mirror neuron system. PLoS Biology 2005, 3:e79. 0529–0535 ognizing and empathizing with emotional facial expres- 11. Jackson PL, Brunet E, Meltzoff AN, Decety J: Empathy examined sions by discerning networks of activity, rather than through the neural mechanisms involved in imagining how I feel versus how you feel pain. Neuropsychologia 2005, 44:752-761. simply defining areas of activity [28,29]. This in turn gives 12. Calder AJ, Young AW: Understanding the recognition of facial us a better idea of how the functional connectivity of con- identitiy and facial expression. Nat Neurosci 2005, 6:641-651. stituent regions work together in order to allow a person 13. Haxby JV, Gobbini MI, Furey ML, Ishai A, Schouten JL, Pietrini P: Dis- tributed and overlapping representations of faces and to recognize and process the facial expressions of another. objects in ventral temporal cortex. Science 2001, 293:2425-2430. 14. Posamentier MT, Abdi H: Processing faces and facial expres- Competing interests sions. Neuropsychol Rev 2003, 13:113-143. The authors declare that they have no competing interests. 15. Sinha P, Balas B, Ostrovsky Y, Russell R: Face recognition by humans: Nineteen results all computer vision researchers should know about. IEEE 2006, 94:1948-1962. Authors' contributions 16. Crocker V, McDonald S: Recognition of emotion from facial JSN performed the statistical analysis and drafted the expression following traumatic brain injury. Brain Inj 2005, manuscript. DS participated in the design of the study, 19:787-799. 17. Ashley V, Vuilleumier P, Swick D: Time course and specificity of programmed the stimuli and performed the acquisition of event-related poetentials to emotional expressions. Neurore- the fMRI data. SF participated in the fMRI experiment and port 2003, 15:211-216. 18. Blau VC, Maurer U, Tottenham N, McCandliss BD: The face-spe- the statistical analysis. RS participated in the design of the cific N170 component is modulated by emotional facial study and performed the psychological investigations. MF expression. Behav Brain Funct 2007, 3:7. participated in the design of the study and contributed to 19. Liu J, Harris A, Kanwisher N: Stages of processing in face percep- tion: an MEG study. Neuroscience 2002, 5:910-916. the interpretation of the data and drafting of the manu- 20. Kaufmann J, Schweinberger SR: Expression influences the recog- script. HJW participated in the design of the study and nition of familiar faces. Perception 2004, 33:399-408. participated in the acquisition of the fMRI data. NPA con- 21. Gallese V, Keysers C, Rizzolatti G: A unifying view of the basis of social cognition. Trends Cogn Sci 2004, 8(9):396-403. tributed to the design of the study and participated in the 22. Uddin LQ, Molnar-Szakacs I, Zaidel E, Iacoboni M: fTMS to the interpretation of the data and drafting of the manuscript. right inferior parietal lobule disrupts self-other discrimina- tion. Soc Cogn Affect Neurosci 2007, 1:65-71. JM performed the statistical analysis and participated in 23. Schütz-Bosbach S, Mancini B, Aglioti SM, Haggard P: Self and other the interpretation of the data. RJS participated in the in the human motor system. Curr Biol 2006, 16:1830-1834. design of the study, contributed to the statistical analysis, 24. Decety J, Grezes J: The power of simulation: Imagining one's own and other's behavior. Brain Res 2006, 12:4-14. the data interpretation, and drafting the manuscript. All 25. Seitz RJ, Nickel J, Azari NP: Functional modularity of the medial authors read and approved the final manuscript. prefrontal cortex: involvement in human empathy. Neuropsy- chology 2006, 20:743-751. 26. Decety J, Jackson PL: The functional architecture of Human Acknowledgements Empathy. Behav Cogn Neurosci Rev 2004, 3:71-100. The authors thank E. Rädisch for expert technical assistance during the 27. Horwitz B: Data analysis paradigms for metabolic-flow data: fMRI experiment. Combining neural modeling and functional neuroimaging. Human Brain Mapp 1994, 2:112-122. 28. Alexander GE, Moeller JR: Application of the subprofile model to functional imaging in neuropsychiatric disorders: A prin- cipal component approach to modeling brain function in dis- ease. Human Brain Mapp 1994, 2:79-94. Page 11 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 29. Friston KJ, Holmes AP, Worsely KJ, Poline J-P, Frith CD, Frackowiak 56. Puce A, Allison T, Asgari M, Gore JC, McCarthy G: Differential sen- RSJ: Statistical parametric maps in functional imaging: A gen- sitivity of human visual cortex to faces, letterstrings, and tex- eralized linear approach. Human Brain Mapp 1994, 2:189-210. tures: A functional magnetic resonance imaging study. J 30. McIntosh AR: Towards a network theory of cognition. Neural Neurosci 1996, 16:5205-5215. Netw 2000, 13:861-870. 57. Rossion B, Caldara R, Seghier M, Schuller A-M, Lazeyras F, Mayer E: 31. Seitz RJ, Knorr U, Azari NP, Weder B: Cerebral networks in sen- A network of occipito-temporal face-sensitive areas besides sorimotor distrubances. Brain Res Bull 2001, 54:299-305. the right middle fusiform gyrus is necessary for normal face 32. Oldfield RC: The assessment and analysis of handedness: the processing. Brain 2003, 126:2381-2395. Edinburgh inventory. Neuropsychologia 1971, 9:97-113. 58. Rotshtein P, Henson RNA, Treves A, Driver J, Dolan RJ: Morphing 33. Bagby RM, Parker JDA, Taylor GJ: The Twenty-Item Toronto Marilyn into Maggie dissociates physical and identity face Alexithymia Scale-I. Item selection and cross-validation of representations in the brain. Nat Neuroscience 2005, 8:107-113. the factor structure. J Psychosomat Res 1994, 38:23-32. 59. Schiltz C, Sorger B, Caldara R, Ahmed F, Mayer E, Goebel R, Rossion 34. Bagby RM, Taylor GJ, Parker JDA: The Twenty-Item Toronto B: Impaired face discrimination in acquired prosopagnosia is Alexithymia Scale-II. Convergent, Discriminant, and Con- associated with abnormal response to individual faces in the current Validity. J Psychosomat Res 1994, 38:33-40. right middle fusiform gyrus. Cereb Cortex 2006, 16(4):574-586. 35. Franz M, Popp K, Schaefer R, Sitte W, Schneider C, Hardt J, Decker 60. Sergent J, Ohta S, MacDonald B: Functional neuroanatomy of O, Braehler E: Alexithymia in the German general population. face and object processing. Brain 1992, 115:15-36. Soc Psychiatry Psych Epidemiol 2008, 43:54-62. 61. Steeves JK, Culham JC, Duchaine BC, Pratesi CC, Valyear KF, Schin- 36. Hautzinger M, Bailer M, Worall H, Keller F: BDI. Beck-Depres- dler I, Humphrey GK, Milner AD, Goodale MA: The fusiform face sions-Inventar. In Huber Bern Göttingen Toronto Seattle; 1995. area is not sufficient for face recognition: Evidence from a 37. Behr M, Becker M: SEE. Skalen zum Erleben von Emotionen. patient with dense prosopagnosia and no occipital face area. In Manual Göttingen: Hogrefe; 2004. Neuropsychologia 2006, 44:594-609. 38. Ekman P, Friesen WV: Pictures of Facial Affect [slides]. Palo 62. Yovel G, Kanwisher N: The neural basis of the behavioral face- Alto, CA: Consulting Psychologists Press; 1996. inversion effect. Curr Biol 2005, 15:2256-2262. 39. Esslen M, Pascual-Marqui RD, Hell D, Kochi K, Lehmann D: Brain 63. Allison T, Giner H, McCarthy G, Nore AC, Puce A, Luby M, Spencer areas and time course of emotional processing. Neuroimage DD: Face recognition in human extrastriate cortex. J Neuro- 2004, 21:1189-203. physiol 1994, 71:821-825. 40. Franz M, Schaefer R, Schneider C, Sitte W, Bachor J: Visual event- 64. Winston JS, Henson RNA, Fine-Goulden MR, Dolan RJ: fMRI-adap- related potentials in subjects with alexithymia: modified tation reveals dissociable neural representations of identity processing of emotional aversive information. Am J Psychiat and expression in face perception. J Neurophysiol 2004, 2004, 161:728-735. 92:1830-1839. 41. Kleiser R, Wittsack H-J, Bütefisch CM, Jörgens S, Seitz RJ: Functional 65. Takamura M: Prosopagnosia: A look at the laterality and spe- activation within the PI-DWI mismatch region in recovery cificity issues using evidence from neuropsychology and neu- from hemiparetic stroke: preliminary observations. Neuroim- rophysiology. The Harvard Brain 1996, Spring:9-13. age 2005, 24:515-523. 66. Ramnani N, Owen AM: Anterior prefrontal cortex: Insights into 42. Jackson J: A User's Guide to Principal Components. John Wiley function from anatomy and neuroimaging. Nat Rev Neurosci & Sons: New York; 1991. 2004, 3:184-194. 43. Talairach J, Tournoux P: Co-planar stereotaxic atlas of the 67. Sugiura M, Friston KJ, Willmes K, Shah NJ, Zilles K, Fink GR: Analysis human brain. New York: Thieme; 1998. of intersubject variability in activation: an application to the 44. Brett M: The MNI brain and the Talairach atlas. 14-2-2002. Ref incidental episodic retrieval during recognition test. Human Type: Internet Communication . Brain Mapp 2007, 28:49-58. 45. Gauthier : The fusiform 'face area' is part of a network that 68. Azari NP, Missimer J, Seitz RJ: Religious experience and emotion: processes faces at the individual level. J Cogn Neurosci 2000, Evidence for distinctive neural patterns. Int J Psychol Religion 12:495-504. 2005, 15:263-281. 46. George N, Dolan RJ, Fink GR, Baylis GC, Russell C, Driver J: Con- 69. Mobbs D, Garrett AS, Menon V, Rose FE, Bellugi U, Reiss AL: Anom- trast polarity and face recognition in the human fusiform alous brain activation during face and gaze processing in Wil- gyrus. Nat Neuroscience 1999, 2:574-580. liams syndrome. Neurology 2004, 62:2070-2076. 47. Hadjikhani N, Gelder B: Neural basis of prosopagnosia. Human 70. Luks TL, Simpson GV: Preparatory deployment of attention to Brain Mapp 2002, 16:176-182. motion activates higher-order motion-processing brain 48. Halgren E, Dale AM, Sereno MI, Tootell R, Marinkovic K, Rosen BR: regions. Neuroimage 2004, 22:1515-1522. Location of human face-selective cortex with respect to reti- 71. Gilbert SJ, Spengler S, Simons JS, Steele JD, Lawrie SM, Frith CD, Bur- notopic areas. Human Brain Mapp 1999, 7:29-37. gess PW: Functional specialization within rostral prefrontal 49. Ishai A, Ungerleider LG, Haxby JV: Distributed neural systems for cortex (area 10): a meta-analysis. J Cogn Neurosci 2006, the generation of visual images. Neuron 2000, 28:979-990. 18:932-948. 50. Kanwisher N, McDermott J, Chun MM: The fusiform face area: A 72. Kanwisher N, Tong F, Nakayama K: The effect of face inversion module in human extrastriate cortex specialized for face on the human fusiform face area. Cognition 1998, 68:1-11. perception. J Neurosci 1997, 17:4302-4311. 73. Dolan RJ, Fletcher P, Morris J, Kapur N, Deakin JF, Frith CD: Neural 51. Lerner Y, Hendler T, Ben-Bashat D, Harel M, Malach R: A hierarchi- activation during covert processing of positive emotional cal axis of object processing stages in the human visual cor- facial expressions. Neuroimage 1996, 4:194-200. tex. Cereb Cortex 2001, 11:287-297. 74. Hoffman EA, Haxby JV: Distinct representations of eye gaze and 52. Nakamura K, Kawashima R, Sato N, Nakamura , Sugiura M, Kato T, identity in the distributed human neural system for face per- Hatano K, Ito K, Fukuda H, Schormann T, Zilles K: Functional ception. Nat Neurosci 2000, 3:80-84. delineation of the human occipitoltemporal areas related to 75. Pessoa L, McKenna E, Gutierrez E, Ungerleider LG: Neural process- face and scene processing; a PET study. Brain 2000, ing of emotional faces requires attention. Proc Natl Acad Sci 123:1903-1912. 2002, 99:11458-11463. 53. Narumoto J, Okada T, Sadato N, Fukui K, Yonekura Y: Attention to 76. Nakamura K, Kawashima R, Ito K, Sugiura M, Kato T, Nakamura A, emotion modulates fMRI activity in human right superior Hatano K, Nagumo S, Kubota K, Fukuda H, Kojima S: Activation of temporal sulcus. Cogn Brain Res 2001, 12(2):225-231. the right inferior frontal cortex during assessment of facial 54. Onitsuka T, Shenton ME, Kasai K, Nestor PG, Toner SK, Kikinis R, emotion. J Neurophysiol 1999, 82:1610-1614. Jolesz FA: Fusiform gyrus volume reduction and facial recog- 77. Powell EW, Leman RB: Connections of the nucelus accumbens. nition in chronic schizophrenia. Arch Gen Psychiatry 2003, Brain Res 1976, 105:389-403. 60(4):349-355. 78. Roberts AC, Tomic DL, Parkinson CH, Roeling TA, Cutter D, Rob- 55. Pizzagalli DA, Lehmann D, Hendrick AM, Regard M, Pascual-Marqui bins TW, Everitt BJ: Forebrain connectivity of the prefrontal RD, Davidson RJ: Affective judgments of faces modulate early cortex in the marmoset monkey (Callithrix jacchus): an activity (~160 ms) within the fusiform gyri. Neuroimage 2002, anterograde and retrograde tract-tracing study. J Comp Neu- 16:663-677. rol 2007, 502:86-112. Page 12 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 79. Tekin S, Cummings JL: Frontal-subcortical neuronal circuits and 103. Fu CHY, Williams SCR, Brammer MJ, Suckling J, Kim J, Cleare AJ, clinical neuropsychiatry: An update. J Psychosomat Res 2002, Walsh ND, Mitterschiffthaler MT, Andrew CM, Pich EM, Bullmore 53:647-654. ET: Neural resonses to happy facial expressions in major 80. Siffert J, Allen JC: Late effects of therapy of thalamic and depression following antidepressant treatment. Am J Psych hypothalamic tumors in childhood: Vascular, neurobehavio- 2006, 164:599-607. ral and neoplastic. Pediatr Neurosurg 2000, 33:105-111. 104. Servos P, Osu R, Santi A, Kawato M: The neural substrates of bio- 81. Corbetta M, Kincade JM, Ollinger JM, McAvoy MP, Shulman GL: Vol- logical motion perception: An fMRI study. Cereb Cortex 2002, untary orienting is dissociated from target detection in 12:772-782. human posterior parietal cortex. Neuroscience 2000, 3:292-297. 105. Barch DM, Braver TS, Akbudak E, Conturo' T, Ollinger J, Snyder A: 82. Corbetta M, Shulman GL: Human cortical mechanisms of visual Anterior cingulate cortex and response conflict: Effects of attention during orienting and search. Phil Trans R Soc Lond response modality and processing domain. Cereb Cortex 2001, 1998, 353:1353-1362. 11:837-848. 83. Hariri AR, Bookheimer SY, Mazziotta JC: Modulatory emotional 106. Bush G, Luu P, Posner MI: Cognitive and emotional influences in responses: effect of a neurocortical network on the limbic anterior cingulate cortex. Trends Cogn Sci 2000, 4(6):215-222. system. Neuroreport 2000, 11:43-48. 107. Fichtenholtz HM, Dean HL, Dillon DG, Yamasaki H, McCarthy G, 84. Tamm L, Menon V, Reiss A: Parietal attentional system aberra- LaBar KS: Emotion-attention network interactions during a tions during target detection in adolescents with attention visual oddball task. Cogn Brain Res 2004, 20:67-80. deficit hyperactivity disorder: Event-related fMRI evidence. Am J Psych 2006, 163:1033-1043. 85. Garavan H, Ross TJ, Stein EA: Right hemispheric dominance of inhibitory control: an event-related functional MRI study. Proc Natl Acad Sci USA 1999, 96:8301-8306. 86. Decety J, Chaminade T, Grezes J, Meltzoff AN: A PET exploration of the neural mechanisms involved in reciprocal imitation. Neuroimage 2001, 15:265-272. 87. Frith U, Frith CD: Development and neurophysiology of men- talizing. Phil Trans R Soc Lond 2003, 358:459-473. 88. Farrer C, Franck N, Georgieff N, Frith CD, Decety J, Jeannerod M: Modulating the experience of agency: a positron emission tomography study. Neuroimage 2003, 18:324-333. 89. Völlm BA, Taylor AN, Richardson P, Corcoran R, Stirling J, McKie S, Deakin JF, Elliot R: Neuronal correlates of theory of mind and empathy: A functional magnetic resonance imaging study in a nonverbal task. Neuroimage 2005, 29:90-98. 90. Blanke O, Mohr C, Michel CM, Pascual-Leone A, Brugger P, Seeck M, Landis T, Thut G: Linking out-of-body experience and self processing to mental own-body-imagery at the temporopa- rietal junction. J Neurosci 2005, 25:550-557. 91. Karnath H, Himmelbach M, Rorden C: The subcortical anatomy of human spatial neglect: putamen, caudate nucleus and pul- vinar. Brain 2002, 125:350-360. 92. Hedden T, Gabrieli JDE: The ebb and flow of attention in the human brain. Nat Neurosci 2006, 9:863-865. 93. Saxe R, Kanwisher N: People thinking about thinking people: The role of the temporo-parietal junction in "theory of mind". Neuroimage 2003, 19:1835-1842. 94. Carr L, Iacoboni M, Dubeau M-C, Mazziota JC, Lenzi GL: Neural mechanisms of empathy in humans: A relay from neural sys- tems for imitation to limbic areas. Proc Natl Acad Sci USA 2003, 100:5497-5592. 95. Singer T, Seymour B, O'Doherty J, Kaube H, Dolan RJ, Frith CD: Empathy for pain involves the affective but not sensory com- ponents of pain. Science 2004, 303(5661):1157-1162. 96. Haxby JV, Hoffman EA, Gobbini MI: The distributed human neu- ral system for face perception. TICS 2000, 4:223-233. 97. Schweinsburg AD, Paulus MP, Barlett VC, Killeen LA, Caldwell LC, Pulido C, Brown SA, Tapert SR: An fMRI study of response inhi- bition in youths with a family history of alcoholism. Ann NY Acad Sci 2004, 1021:391-394. 98. Iidaka T, Matsumoto A, Nogawa J, Yamamoto Y, Sadato N: Fron- toparietal network involved in successful retrieval from epi- Publish with Bio Med Central and every sodic memory. Spatial and temporal analyses using fMRI and scientist can read your work free of charge ERP. Cereb Cortex 2006, 16:1349-1360. 99. McIntosh AR, Rajah MN, Lobaugh N: Interactions of prefrontal "BioMed Central will be the most significant development for cortex in relation to awareness in sensory learning. Science disseminating the results of biomedical researc h in our lifetime." 1999, 284:1531-1533. Sir Paul Nurse, Cancer Research UK 100. Adleman NE, Menon V, Blasey CM, White CD, Warsofsky IS, Glover GH, Reiss AL: A developmental fMRI study of the stroop color- Your research papers will be: word task. Neuroimage 2002, 16:61-75. available free of charge to the entire biomedical community 101. Lane RD, Rieman EM, Axelrod B, Yun LS, Holmes A, Schwartz GE: Neural correlates of levels of emotional awareness. Evidence peer reviewed and published immediately upon acceptance of an interaction between emotion and attention in the ante- cited in PubMed and archived on PubMed Central rior cingulate cortex. J Cogn Neurosci 1998, 10:525-535. 102. Schäfer R, Popp K, Jörgens S, Lindenberg R, Franz M, Seitz RJ: Alex- yours — you keep the copyright ithymia-like disorder in right anterior cingulate infarction. BioMedcentral Neurocase 2007, 13:201-208. Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp Page 13 of 13 (page number not for citation purposes) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Behavioral and Brain Functions Springer Journals

On the neural networks of empathy: A principal component analysis of an fMRI study

Download PDF
13 pages

Loading next page...
 
/lp/springer-journals/on-the-neural-networks-of-empathy-a-principal-component-analysis-of-an-Vr6WT5d88j

References (124)

Publisher
Springer Journals
Copyright
Copyright © 2008 by Nomi et al; licensee BioMed Central Ltd.
Subject
Biomedicine; Neurosciences; Neurology; Behavioral Therapy; Psychiatry
eISSN
1744-9081
DOI
10.1186/1744-9081-4-41
pmid
18798977
Publisher site
See Article on Publisher Site

Abstract

Background: Human emotional expressions serve an important communicatory role allowing the rapid transmission of valence information among individuals. We aimed at exploring the neural networks mediating the recognition of and empathy with human facial expressions of emotion. Methods: A principal component analysis was applied to event-related functional magnetic imaging (fMRI) data of 14 right-handed healthy volunteers (29 +/- 6 years). During scanning, subjects viewed happy, sad and neutral face expressions in the following conditions: emotion recognition, empathizing with emotion, and a control condition of simple object detection. Functionally relevant principal components (PCs) were identified by planned comparisons at an alpha level of p < 0.001. Results: Four PCs revealed significant differences in variance patterns of the conditions, thereby revealing distinct neural networks: mediating facial identification (PC 1), identification of an expressed emotion (PC 2), attention to an expressed emotion (PC 12), and sense of an emotional state (PC 27). Conclusion: Our findings further the notion that the appraisal of human facial expressions involves multiple neural circuits that process highly differentiated cognitive aspects of emotion. Introduction on correctly recognizing and reacting to rapid fluctuations Human emotional facial expressions contain information in the emotional states of others [4,5]. This capability may which is essential for social interaction and communica- have played a key role in our ability to survive and evolve tion [1-3]. Social interaction and communication depend [6]. Page 1 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 In the research field of psychological childhood develop- Each psychological state may be associated with distinct ment the competence to detect, share, and utilise cogni- neural networks containing cortical areas which interact tive patterns and emotional states of the other was within and across neutral networks. The interactions fos- conceptualized as mentalizing [7]. The important aspect is ter an exponentially complex environment within which that the emotion expressed in someone else's face is not the processes driving emphatic responses occur. While the only detected but also valuated from the subjective per- processes have been explored to some extent, much work spective of the observer. In fact, the appraisal of and the is needed to reveal the precise nature of the complex inter- resonance with an emotion observed in somebody else is action among the constituents of empathy. central to the concept of empathy [8-11]. Accordingly, empathy has been suggested to involve dis- We can acquire two important sources of information tinct, distributed neural networks. Such distributed net- upon perceiving a face: the identification and emotional works demand the application of a network analysis, expression of the individual [12-15]. Support for distinct which subjects voxels of the image matrix to a multivariate networks of facial identification and emotional recogni- rather than a univariate analysis. This feature allows a net- tion have been provided by studies of those with trau- work analysis to overcome two significant limitations of matic brain injuries [16], electroencephalograph (EEG) univariate analyses that are based on categorical compari- studies [17,18], magnetoencephalography (MEG) studies sons: they are unable to distinguish regional networks [19], and also by studies examining the influence of emo- because the constituent voxels may not all change at the tional expression on familiar faces [20]. EEG studies have defined level of significance and may also show areas of also hypothesized that these two networks operate in a activation which are unrelated to the phenomenon being parallel system of recognition in which emotional effects studied [27]. and structural facial features are simultaneously identified through distinct neural networks [8,9]. The network analysis applied to the blood oxygen level dependent (BOLD) signals in this functional magnetic When inferring the emotional expression of another, we resonance imaging (fMRI) study is a principal component are automatically compelled to compare our assessment analysis (PCA), which decomposes the image matrix into of their emotional state with our own. A key component statistically uncorrelated components. Each component of perspective taking and emphatic experiences is the abil- represents a distinct neural network, and the extreme ity to compare the emotional state of oneself with another voxel values of a component image, its nodes. Thus, the [8-11,21]. However, a distinct separation between first component images map the functional connectivity of and third-person experiences is necessary [8,22]. People constituent regions activated during neural stimulation. are normally able to correctly attribute emotional states Previous studies have proposed and verified the hypothe- and actions to the proper individual, whether they are our sis of functional connectivity in regions-of-interest [28] own or someone else's [23]. Confusing our emotional and voxel-based analyses [29-31]. In contrast to regions state with another would vitiate the function of empathy that show enhanced metabolic or blood flow levels corre- and cause unnecessary emotional distress and anxiety lated with mental states [31], the networks deduced by [24]. PCA incorporate no a priori assumptions regarding the neural stimulation. Therefore, a number of distinct psychological states func- tioning in concert may be necessary in order for a proper Virtually unexplored until now, the neural networks asso- emphatic experience to occur [25]. They may include the ciated with recognizing and empathizing with human evaluation of emotion of the self, evaluation of another's facial expressions of emotions are here described using emotion, comparing those emotions, and anticipating PCA. and reacting to our own or another's emotion, among others. In accord with this line of thought, Decety and We attempt to elucidate the functional neural networks Jackson [26] have proposed three major functional com- central to the processes of recognizing and empathizing ponents of emphatic experiences. First, the actions of with emotional facial stimuli recorded in fMRI acquisi- another person automatically enact a psychological state tions of healthy subjects. in oneself which mirror those actions and create a repre- sentative state which incorporates the perceptions of both Statistical testing identified four principal components the self and another. Secondly, there must be a distinct (PC) as relevant neural networks. The correlation of the separation between the perception of the self and the subjects' scores of emotional experience with the PC's fur- other person. Finally, the ability to cognitively assume the ther supported their functional relevance. Our analysis perception of another person while being aware of self/ shows the coordinated action of brain areas involved in other separation is important. the processing of visual perception and emotional Page 2 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 appraisal underlying the facial expressions of human Visual stimuli emotions. These regions, including pre-frontal control From photographs of facial affect [38], we selected for areas and occipital visual processing areas presumably presentation those with happy, sad and neutral facial constitute nodes of the networks implicated in appraising expressions found to be correctly identified in more than other people's emotion in their facial expressions. 90 percent of raters; 14 happy, 14 sad, and 14 neutral faces were used. Only faces which had been found to be cor- Methods rectly identified in more than 90 percent of raters [38] Subjects were used to ensure that the subjects internally generated Fourteen healthy, right-handed subjects (28.6 +/- 5.5 the corresponding emotion. This corresponded to a simi- years; 7 men, 7 women) participated in this study. The lar approach of a recent study [39]. As control stimuli we subjects had normal or corrected to normal vision. Right- produced a number of scrambled images from these pho- handedness was assessed using the Oldfield's question- tographs equal to the number of intact faces. All images naire [32]. In addition, the subjects' emotional compe- were digitized and controlled for luminescence. tence was tested with the German, 20-item version of the Experimental task design Toronto Alexithymia Scale (TAS-20) [33-35]. None of the participating subjects were classified as alexithymic (mean Faces were presented for durations ranging between 300 TAS-20 sum score 34.14 +/- 6.26, range 23). Subjects were and 500 ms, since this short presentation time is sufficient also evaluated with the Beck Depression Inventory [36] for conscious visual perception avoiding habituation [40]. and the scales of emotional experience (SEE) [37]. The presentation time of the faces was jittered to enhance the detection of the stimulation-related BOLD activity changes in this event-related fMRI study. Thereafter, Sc Figure 1 hematic illustration of the experimental design Schematic illustration of the experimental design. Page 3 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 scrambled faces were presented for durations between 11 subject was co-registered to the mean image of the func- and 12 s (Figure 1). This long second stimulation period tional images and also normalized to the MNI-space. was chosen to provide the subjects sufficient time to engage in appraising the emotional facial expression seen For each of the three instruction sets, the happy, sad and in the first stimulation interval and to allow changs of skin neutral face presentations were modelled as well as the conductance to occur [40]. The faces were presented in corresponding scrambled faces. The models employed the random order on a laptop connected to a projector (LCD haemodynamic response function provided by SPM2. data projector, VPL-S500E, Sony, Toyko, Japan) on a Data were temporally filtered using a Gaussian low-pass screen which was placed approximately 50 cm from the filter of 4 s and a high-pass filter of 100 s. All data were mirror in the head coil. Immediately before each fMRI scaled to the grand mean. Realignment parameters as scan the subjects were instructed to view the faces in a determined in the realignment step were used as con- mind set according to one of the following cognitive founding covariates. The duration of all events was mod- instructions: a) identify the emotion expressed in the faces elled with 4 s for face presentations and 8 s for the delay (RECOGNIZE), b) empathize with the emotion expressed period. The repeated condition images of the 18 experi- in the faces (SHARE EMOTION), c) count the earrings in mental conditions were averaged for each of the 14 sub- the faces shown (control condition: DETECT EARRINGS). jects, yielding a total of 252 averaged condition images. Each instruction was given twice in random order across Data were modelled using the canonical haemodynamic the subjects, yielding six separate scans per subject. Thus, response function provided by SPM2. The duration of all the visual stimuli were identical in the different experi- events was modelled explicitly with 4 s for face presenta- mental conditions, but the visual information to be proc- tions and 8 s for the delay period. To identify the brain essed differed according to the instructions. Since each areas related to viewing the faces a comparison with view- condition was repeated twice, 28 happy, 28 sad and 28 ing the scrambled faces was calculated. Only areas with a neutral faces were presented. The presentation was done p < 0.05 corrected at cluster level with a cluster threshold in random order across the subjects to counteract possible > 20 voxels were accepted (Table 1). sequence or habituation effects. After each condition, the subjects were debriefed about how well they could per- The PCA employed in house software of which some form the task. modules were adapted from SPM2. Extracerebral voxels were excluded from the analysis using a mask derived Functional magnetic resonance imaging from the gray matter component yielded by segmentation The subjects lay supine in the MRI scanner and viewed the of the high resolution anatomical 3D image volume into faces on a screen via a mirror which was fixed to the head gray matter, white matter and cerebrospinal fluid using coil. Image presentation was controlled by a TTL-stimulus the segmentation module of SPM2. Voxel values of the coming from the MRI-scanner as described in detail else- segmented image ranged between 0 and 1; the mask where [41]. Scanning was performed on a Siemens Vision included only those exceeding 0.35, excluding most of 1.5 T scanner (Erlangen, Germany) using an EPI-GE white matter. Calculation of the residual matrix is the first sequence: TR = 5 s, TE = 66 ms, flip angle = 90°. The whole step. From a matrix whose rows corresponded to the 252 brain was covered by 30 transaxial slices oriented parallel conditons and columns to the 180 thousands voxels in a to the bi-commissural plane with in-plane resolution of single image volume the mean voxel value of each row 3.125 × 3.125 mm, slice thickness of 4 mm, and interslice was subtracted from each element as was the mean voxel gap of 0.4 mm. Each acquisition consisted of 255 vol- value of each column subtracted from each element. umes. The first 3 volumes of each session did not enter the Thereafter, the grand mean of all voxel values in the orig- analysis. A high-resolution 3D T1-weighted image (TR = inal matrix was added to each element. The result of this 40 ms, TE = 5 ms, flip angle = 40°) consisting of 180 sag- normalization procedure is the residual matrix for which ittal slices with in plane resolution of 1.0 × 1.0 mm was the row, column and grand means vanish. Using the sin- also acquired for each subject. gular value decomposition implemented in Matlab, the residual matrix was then decomposed into 252 compo- Data analysis nents, consisting of an image, an expression coefficient, Image data were analysed with SPM2 (Wellcome Depart- and an eigenvalue for each component. The eigenvalue ment of Cognitive Neurology, London, UK; http:// was proportional to the square root of the fraction of var- www.fil.ion.ucl.ac.uk/spm). Images were slice-time cor- iance described by each component, while the expression rected, realigned, normalized to the template created in coefficients described the amount that each subject and the Montreal Neurology Institute (MNI), and spatially condition contributed to the component. The principal smoothed with a 10 × 10 × 10 mm Gaussian filter. The components (PCs) were ranked according to the propor- normalization step resampled the images to a voxel size of tion of variance that each component explains, i.e, PC 1 2 × 2 × 2 mm. The anatomical T1-weighted image of each explains the greatest amount of variance. The expression Page 4 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 Table 1: Cerebral activations related to viewing emotional face expressions as compared with viewing scrambled faces Anatomical location Coordinates Brodmann area Cognitive instruction (Task) x y z Recognize Empathize Object detection Lingual gyrus R 28 -72 -10 BA 18 + Fusiform gyrus R 36 -69 -10 BA 19 + + Cuneus R 28 -78 16 BA 31 + + Superior frontal gyrus L -3 22 52 BA 6 + + + Inferior frontal gyrus L -49 2 20 BA 44 + + + Inferior frontal gyrus R 55 8 16 BA 44 + + + Middle frontal gyrus R 43 29 30 BA 9 + + Supramarginal gyrus L -46 -49 28 BA 42 + Parietal operculum R 50 -24 10 BA 40 + Sup. temporal gyrus R 34 16 -6 BA 38 + Hypothalamus L -3 -6 -11 + Cerebellum L -14 -60 -22 + Activations in categorical analysis with SPM2 at p < 0.05 FDR with clusters > 20 voxels. The coordinates (x, y, z; mm) of the peak activations (+) are listed in stereotactic space [43]. L refers to left hemisphere; R refers to right hemisphere. coefficients and voxel values of a PC were orthonormal Results and their orthogonality reflected the statistical independ- Before the fMRI experiment the subjects' capacity to expe- ence of the PCs. The PC image displays the degree to rience emotions was tested with the German, 20-item ver- which the voxels covaried in each PC, their voxel values sion of the Toronto Alexithymia Scale (TAS-20 [35]). It (loadings) ranged between -1 and 1. was found that the 14 participating subjects had a normal mean TAS-20 sum score (34.1 +/- 6.3, range 23). This In order to provide a neurophysiological interpretation of indicated that each of the subjects had a high capacity of the components, statistical tests, e.g. unpaired t-tests and introspection and emotional awareness. In the fMRI ses- tests of correlation (Pearson) were applied to the expres- sion the subjects stated that they could readily identify the sion coefficients. The formal criteria for relevant PCs were: seen emotion and generate the corresponding emotion (1) the statistical tests identified the condition differenti- internally as instructed. They detected 92 +/- 0.2 percent ating PCs at a significance level of p < 0.001, and (2) the of the faces wearing earrings. PCs fulfilled the Guttman-Kaiser criterion, the most com- mon retention criterion in PCA [42] in which PCs associ- The categorical analysis showed that viewing the emo- ated with eigenvalues of the covariance matrix larger in tional face expressions as compared with viewing the magnitude than the average of all eigenvalues are scrambled faces resulted in activations of right visual cor- retained, implying in this analysis that they ranked among tical areas and bilaterally the inferior frontal and superior the first 61 PCs. The voxels describing the nodes of a neu- frontal gyrus (Table 1). Recognizing emotional facial ral network associated with a relevant PC image volume expressions showed the most extensive activation pattern st fulfilled the conditions that the voxel values lie in the 1 involving also the hypothalamus, the left supramarginal th percentile or the 99 percentile of the volume's voxel gyrus, cortical areas at the right temporal parietal junction, value distribution, and that the voxels belong to clusters and the left hypothalamus (Table 1). Empathizing with of greater than 50 voxels. the seen emotion as compared with object detection (con- trol condition) resulted in one activation area which The anatomical locations of the peak activations and of occurred in the left inferior frontal gyrus. Note, that no the coordinates of the maximal PC loadings of the signif- activation occurred in the anterior prefrontal or orbitof- icant PCs are reported in Talairach space [43]. A freely dis- rontal cortex. tributed Matlab script [44] effected the transformation from MNI space. The network analysis revealed that out of the total of 61 retained PCs four differentiated the experimental condi- Correlation of the expression coefficients of the signifi- tions as found by formal statistical testing (Table 2); for cant PCs with the TAS-20, the Beck Depression Inventory, the 11 statistical tests described below, the probability and the SEE scales was conducted using a Pearson two- threshold corrected for multiple comparisons is p < 0.001. tailed correlation (SPSS for Windows, Version 12.0.1.). PC1 explained 16.8 percent of the variance and distin- The significance level of the correlations was set to p < guished between viewing the happy, sad and neutral faces 0.01. from viewing the scrambled faces during RECOGNIZE, Page 5 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 Table 2: Cerebral circuits in processing of emotional face expressions Coordinates Anatomical location Broadmann area Cluster size XY Z PC1: Face identification (16.8%) Positive loadings 4 53 4 Superior frontal gyrus R BA 10 248 -4 48 -2 Anterior cingulate L BA 32 81 1 31 -7 Anterior cingulate R BA 32 76 38 26 45 Middle frontal gyrus L BA 8 66 -50 -63 31 Angular gyrus L BA 39 53 44 -61 28 Supramarginal gyrus R BA 39 396 Negative loadings -21 -77 -10 Lingual gyrus L BA 18 306 28 -72 -10 Lingual gyrus R BA 18 314 34 -70 15 Cuneus R BA 31 428 -6 -75 44 Precuneus L BA 7 54 PC2: Identification of expressed emotion (4.7%) Positive loadings -30 -74 -10 Fusiform gyrus L BA 19 158 36 -69 -13 Fusiform gyrus R BA 19 252 40 -87 8 Middle occipital gyrus R BA 19 60 6 3 59 Superior frontal gyrus L BA 6 87 -49 2 20 Inferior frontal gyrus L BA 44 65 -3 -6 -11 Hypothalumus L 50 4 -6 -11 Hypothalumus R 63 Negative loadings -5 -73 46 Precuneus L BA 7 51 2 -63 52 Precuneus R BA 7 71 -63 -10 0 Superior temporal gyrus L BA 22 66 -2 4 6 Thalamus L 259 7 -3 6 Thalamus 300 -14 -30 -22 Pons L 80 50 -60 -27 Neocerebellum R 71 PC 12: Attention to expressed emotion (1.6%) Positive loadings 4 -86 17 Cuneus R BA 18 230 54 8 36 Middle frontal gyrus R BA 9 115 55 17 16 Inferior frontal gyrus R BA 44 56 -51 -35 -2 Middle temporal gyrus L BA 22 64 59 -56 3 Middle temporal gyrus R BA 22 119 8-8 12 Thalamus R 377 Negative loadings -12 -77 48 Precuneus L BA 7 117 8 -69 48 Precuneus R BA 7 68 -42 -48 56 Inferior parietal lobule L BA 40 128 -32 60 6 Middle frontal gyrus L BA 10 246 43 29 30 Middle frontal gyrus R BA 9 132 PC 27: Sense of emotional state (0.9%) Positive loadings -48 -57 -6 Middle occipital gyrus L BA 37 68 -28 52 21 Middle frontal gyrus L BA 10 153 -550 Caudate L 66 -13 -18 -25 Pons L 131 -1 -47 -36 Medulla oblongata 192 Negative loadings -42 -59 -17 Fusiform gyrus L BA 37 58 -46 -49 28 Supramarginal gyrus L BA 42 86 51 -34 50 Inferior parietal lobule R BA 40 54 55 -19 16 Parietal operculum R BA 40 53 -55 -40 8 Superior temporal gyrus L BA 22 162 38 8 -26 Superior temporal gyrus R BA 38 59 Shown are the relevant principal components (PC) and their designations. Indicated in the parentheses beside each PC entry is the amount of th variance accounted for by that PC. The coordinates (x, y, z; mm) are listed in stereotactic space [43]. Only those voxels with values lying in the 99 st percentile (positive loadings) or the 1 percentile (negative loadings) and in cluster with sizes greater than 50 voxels were included. L refers to left hemisphere; R refers to right hemisphere. Page 6 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 SHARE EMOTION and DETECT EARRINGS. Accordingly, subjects exhibited score values suggestive of depression, PC1 represented a neural network associated with face this correlation was obtained in the normal range of the identification. The areas with the positive loadings Beck's depression inventory. Nevertheless, a more nega- included the right dorsolateral and superior frontal cortex, tive emotional experience was related to recognizing neu- the left anterior cingulate, and bilaterally the inferior pari- tral faces. Further, the TAS-20 scores correlated negatively etal region. The areas with the negative loading included with PC2 expression coefficients computed for viewing bilaterally the lingual gyrus, the precuneus, and the happy and neutral faces in the control condition DETECT cuneus, which are areas involved in higher order process- EARRINGS. Note, that the TAS-20 classified all subjects as ing of visual information (Table 2, Figures 2 and 3). highly emotionally sensitive. Thus, this correlation sug- gests that the more sensitive to processing of emotion our PC2 explained 4.7 percent of the variance and differenti- subjects were, the more they were so during implicit ated viewing the happy, sad, and neutral faces from view- processing of faces. The SEE-scale score values related to ing scrambled faces during the RECOGNIZE and SHARE experience of emotional control correlated negatively EMOTION conditions (Table 2, Figures 2 and 3). There- with PC 1 expression coefficients computed for viewing of fore, PC2 was expected to depict a neural network associ- sad faces in the RECOGNIZE condition (Figure 5). A sim- ated with the identification of an expressed emotion. In ilar correlation was found for PC27 expression coeffi- fact, the areas with positive loadings included bilaterally cients computed for viewing of neutral faces in the the fusiform gyrus, the right middle occipital gyrus, the RECOGNIZE condition. This suggested that recognizing right superior frontal gyrus, and the left inferior frontal sad and neutral faces was most pronounced in the subjects gyrus. The areas with negative loadings included bilater- whose scores indicated relatively impaired emotional ally the precuneus, the left superior temporal gyrus, as control. Finally, the SEE-scale score values of experience of well as the thalamus, pons, and neocerebellum (Table 2). self-control correlated with PC 12 expression coefficients The relative localization of the cortical areas in the occip- computed for the SHARE EMOTION condition after view- ital cortex and the superior frontal gyrus involved in PC1 ing happy and neutral faces (Figure 6). This suggested that and PC2 is illustrated in Figure 3. subjects with a high level of self-control most strongly empathized with happy and neutral faces. Thus, viewing PC12 explained 1.6 percent of the variance and reflected sad faces may have impaired the subject's perception of the contrast of the conditions SHARE EMOTION and emotional control, while processing the happy face DETECT EARRINGS as well as the contrast of the condi- expression appears to have improved it (Table 3). In con- tions RECOGNIZE and DETECT EARRINGS during and trast, the processing of neutral faces was related to a rela- after viewing happy, sad, or neutral faces. The compari- tively negative emotion, possibly due to the ambiguous sons showed that PC12 represented a neural network character of the neutral faces (Table 3). associated with the subjects' attention to an expressed emotion. Areas involved in this network included the Discussion right cuneus, bilaterally the precuneus, middle frontal and The novel finding of this event-related fMRI study is that temporal gyrus, and left inferior parietal lobule (Table 2). recognizing and empathizing with emotional face expres- sions engaged a widespread cortical network involving PC27 explained 0.9 percent of the variance and repre- visual areas in the temporal and occipital cortex which are sented the contrast of the conditions SHARE EMOTION known to be involved in face processing [12,45-62]. Pre- with RECOGNIZE during and after viewing happy, sad, vious fMRI studies have identified several brain regions and neutral faces. We hypothesized, therefore, that this PC that are consistently activated when perceiving facial stim- characterizes the neural network subserving the sensation uli. These regions include the fusiform gyrus, which is of the emotional state associated with empathy. The often referred to as the "fusiform face area" [45,49-55,61- regions involved were the left fusiform and middle occip- 63], a face-selective region in the occipito-temporal cortex ital gyrus, the left middle frontal gyrus, bilaterally the infe- [45-48,56-60], and the superior temporal sulcus rior parietal lobule and the superior temporal gyrus (Table [12,13,48,52]. 2, Figure 4). Subcortical structures, such as the left caudate and brain stem were also involved. Because the involved cortical areas are active within and across multiple neural networks, the difficulty of assign- Correlation of the PC expression coefficients with the ing a single function to a cortical area becomes apparent behavioral scales of emotional processing yielded the fol- [2]. In fact, disagreement over assigning functions to cor- lowing observations (Table 3). Beck's depression inven- tical areas does exist, with some arguing that function may tory subject scores correlated significantly with the PC2 be tied to elements of experiment design [64]. However, expression coefficients computed for the RECOGNIZE inability to agree over the specific function of cortical condition after viewing neutral faces. Since none of the areas would further suggest that a description of "face net- Page 7 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 B pose SPM2 in a sa in tion ing) negative loading; pre-SM positiv Figure 2 r P ain areas involved in C of s an d on the 1 ( e d in P e u cu p xn e pr e rio C g u canonical sin ession is 2 ( ttal plane sh r: fr re pon re d loadin cune tal – ne t gh gati y A, hy us, thalamus: blu e PC1 and PC2 rus g) g owing the areas involved le-sub ve loading; anterior por- : gre pothalamus: yellow – e ject M n – po Rs image of superim- ie tiv – e load- Brain areas involved in the PC1 and PC2 superim- posed on the canonical single-subject MR image of SPM2 in a sagittal plane showing the areas involved in PC1 (cuneus: red – negative loading; anterior por- tion of superior frontal gyrus: green – positive load- ing) and in PC2 (precuneus, thalamus: blue – negative loading; pre-SMA, hypothalamus: yellow – positive expression loading). Brain on the canonical single-subje an axia in the fusi ing) relative Figure 3 ar l plane showing the la eas form face area in to volved PC1 (red, negative in P C (PC2, yellow, positive load- 1 and P teral position of ct MR im C loa 2 sup d age of SPM2 in ing) erimposed activity Brain areas involved in PC1 and PC2 superimposed work" or "face pathway" may be more accurate than "face on the canonical single-subject MR image of SPM2 in region" [65]. an axial plane showing the lateral position of activity in the fusiform face area (PC2, yellow, positive load- Notably, cortical areas in the anterior lateral and medial ing) relative to PC1 (red, negative loading). Note the prefrontal cortex known to participate in manipulating medial prefrontal involvement in PC1 (green, positive load- and monitoring information and controlling behavior ing). were also involved [26,66]. Multivariate image analysis using PCA permits the characterization of different net- works that include brain areas changing brain activity in ously to be involved in attention [71], and the anterior related to a given task and brain areas contributing to the portion of the superior frontal gyrus has been implicated function without necessarily changing their activity [28- in theory of mind paradigms [26,71]. 31,67]. Specifically, we applied inferential statistical tests to identify the PCs that effectively differentiated between PC2 represented the functional neural network associated the experimental conditions [31,68]. The four thereby with detecting emotional facial expressions, since the net- identified PCs showed more cerebral areas involved than work nodes included bilaterally the fusiform gyrus, corre- in the simple task comparions. These PCs revealed corre- sponding to the so-called fusiform face area. Some studies lations with behavioral data obtained prior to the fMRI have suggested that the fusiform face area mediates the acquisitions, which highlighted their functional rele- lower order processing of simple face recognition [72], vance. while others have implicated it in higher order processing of faces at a specific level [45] including emotional detec- Functional neural networks tion [73] and identity discrimination [58,64]. Also Of the four differentiating PCs, PC1 distingished the neu- included among the nodes were the posterior portion of ral network involved in face indentification, since the the superior frontal and the inferior frontal gyrus, which involved lingual gyrus, cuneus and precuneus have been have been implicated in higher order processing of faces known to be involved in face recognition and visual involving the perception of gaze direction [74], attention processing [68,69]. The lingual gyrus has been linked to to faces [75], eye and mouth movements [56], and empa- an early stage of facial processing which occurs before spe- thy [25,76]. Thus, our analysis, relating the concerted cific identification occurs [70]. Conversely, the angular action of the fusiform and superior frontal gyrus to the gyrus and the anterior cingulate have been shown previ- detection of emotional face expressions, substantiates pre- Page 8 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 and are thought to regulate the emotions, drives, and motivated behavior [2,79,80]. The third statistically relevant PC, PC12, was related to the attentive processing of expressed emotions. Accordingly, the involved cortical areas, e.g. the precuneus, cuneus, middle frontal gyrus, and thalamus have all been linked to tasks related to controlling attention [81-83]. The infe- rior frontal gyrus has been shown to be involved in response inhibition during attention [84], while the mid- dle temporal gyrus has been implicated in attention to facial stimuli [47] and to visual attention in general [81,82]. PC27 appeared to represent a network mediating a 'feel- ing' or sense of an observed emotional state. In fact, the superior temporal gyrus has been shown to play an important role in perceiving self/other distinctions, and, importantly, in experiencing a sense of agency [85,86]. Also, the superior temporal gyrus seems to be important for "theory of mind" capabilities [87-89]. Similarly, the right inferior parietal lobule has been related to an experi- enced sense of agency [85,86], while the temporal parietal junction has been thought to be crucial to larger networks mediating spatial unity of the self and body [89,90], and attention [81,91,92]. In PC27 MR ima loading) Figure 4 volvem supe ge of SPM2 in an axial pla en rimpose t of thed left s on th uperior temporal gyrus in e canonical sin ne (blue, negative gle-subject Involvement of the left superior temporal gyrus in Activity within and across functional networks PC27 superimposed on the canonical single-subject MR image of SPM2 in an axial plane (blue, negative Empathic processes allow individuals to quickly asses the loading). emotional states and needs of other individuals while rap- idly transmitting our own experiences and needs [93]. Thus, emphatic experiences are crucial to rapid and suc- vious work showing that the conjoint activity of the fusi- cessful social interaction [94,95]. The recognition of emo- form gyrus and cortex lining the superior temporal sulcus tional states and cognitive processing of another's are related to higher order processing of facial features emotional expression are critical skillsets utilized by all [45,53,56,58,64,72-74]. We propose that these areas con- humans upon perceiving another person's face. In order stitute a system involved in facial affect processing. to properly behave in a social situation, one must under- stand the context of the situation before an appropriate PC2 also implicated subcortical structures such as the tha- action can be taken. lamus and the hypothalamus bilaterally, and the cerebel- lum. These structures belong to the anterior cingulate – The participation of the prefrontal cortex in the four rele- ventral striatum – thalamus – hypothalamus loop [76-78] vant PCs in our study not only support the view that dif- Table 3: Correlation of behavioral data vs. expression coefficients Toronto alexthymia scale-20 Beck depression inventory Experience of control of emotion Experience of self control R value P value R value P value R value P value R value P value Condition/PC Condition/PC Condition/PC Condition/PC CH 2 -0.670 0.009 RaN 2 0.667 0.009 RS 1 -0.687 0.007 EaH 12 0.757 0.002 CN 2 -0.683 0.007 RN 27 -0.675 0.008 EaN 12 0.778 0.001 C = DETECT EARRINGS (Control), E = SHARE EMOTION, R = RECOGNIZE, H = Happy, S = Sad, N = Neutral, aH = after Happy, aN = after Neutral Page 9 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 icant carrier of emotional information is an important source for these evaluative and appraisal functions. The four neural networks involved in the recognition of faces and the cognitive control of emotions capture the functions that this study aimed to explore. Our results support numerous studies that hypothesize distinct neu- ral networks for recognizing facial features and emotional expressions [11,13-17,96,97]. In particular, although PC 1 and PC 2 differentiated lower-order facial identification and higher-order facial feature processing, the cortical areas within each PC contained brain areas associated with functions effected by the neural networks depicted in the other PCs. Thus, the middle frontal gyrus activated in all four principal components has been shown to be related to inhibition [98], successful recognition of previ- ously studied items [99], and successful error detection, Regr with data lighting the Figure 5 ession of the test scores p functional releva lots of expression coefficients of th nce of e 14 th su ese bje PCs c of PC1 ts high- response inhibition, interference resolution, and behavio- Regression plots of expression coefficients of PC1 ral conflict resolution associated with completing the with data of the test scores of the 14 subjects high- lighting the functional relevance of these PCs. Stroop Color-Word task [100]. Moreover, the PC repre- senting a basic visual function, PC1 included activation of the right anterior cingulate which has been shown to ferent cognitive functions work together to play a role in severely impact the processing of emotional expressions distinct neural networks, but, as well, further the notion [101,102]. Apparent also in PC1, the lingual gyrus and that the medial pre-frontal cortex is the junction point cuneus have been implicated in the higher-order process- where different visual, attentive, emotional and higher ing of emotional expression [103] and the lingual gyrus order cognitive processes come together in order to allow has been implicated in motional information [104]. Inter- for a subjective reaction towards the exterior world [26]. estingly the anterior cingulate has also been thought to The medial-pre-frontal integration of own cognitive con- cognitively monitor the control of response conflict in cepts, implicit affective schemata and empathy-based information processing [105] and in the regulation of anticipation of the possible reactions of important other cognitive and emotional processing [106]. In a case study, individuals allows for an effective selection and adaptive Steeves et al. [61] demonstrated that an intact fusiform planning of one's own actions. The face as the most signif- face area was sufficient for identifying faces, but a lesion in the occipital face area prevented the patient from higher-order processing of faces such as identity, gender, or emotion. Out results are also in accordance with the idea that emo- tion recognition may play a role in the lower order process of face identification, but this role is limited at best and functions mainly as a stepping stone for higher order rec- ognition of emotional expressions. [107]. Although the lingual gyrus in PC1 and the fusiform face area in PC2 are located closely anatomically, our analysis suggests that they are involved in distinct neural networks related to different subfunctions, face identification and detection of emotional facial expressions, repectively, of facial processing. Nevertheless, it is possible that the subjects recognized the emotions implicitly also in the DETECT EARRING condition. However, the cognitive instruction in this study was that during the RECOGNIZE and SHARE Regr with data lighting the Figure 6 ession of the test scores p functional releva lots of expression coefficients of th nce of e 14 th su ese bje PCs c of PC12 ts high- EMOTION conditions the subjects had to appraise the Regression plots of expression coefficients of PC12 emotions explicitly. In fact, this difference resulted in with data of the test scores of the 14 subjects high- lighting the functional relevance of these PCs. additional brain structures involved in the explicit processing conditions. Page 10 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 The involvement of closely adjacent cortical areas in dif- References 1. Leppanen JM, Moulson MC, Vogel-Farley VK, Nelson CA: An ERP ferent subfunctions related to the detection of emotional study of emotional face processing in the adult and infant face expressions highlights the difficulty of assigning brain. Child Develop 2007, 78:232-245. exclusive functional relevance to a particular area and sug- 2. Adolphs R: Neural systems for recognizing emotion. Curr Opin Neurobiol 2002, 12:169-177. gests the multi-functional nature of brain areas which par- 3. Williams KR, Wishart JG, Pitcairn TK, Willis DS: Emotion recogni- ticipate in multiple, interconnected neural networks tion by children with Down syndrome: Investigation of spe- cific impairments and error patterns. Am J Ment Retard 2005, performing lower-order as well as higher-order neural 110:378-392. processing. 4. Blair RJR: Facial expressions, their communicatory functions and neuro-cognitive substrates. Phil Trans R Soc Lond 2003, 358:561-572. Conclusion 5. Eimer M, Homes A: An ERP study on the time course of emo- Implemented in this study using PCA, employment of a tional face processing. NeuroReport 2002, 13:1-5. network analysis helps to elucidate the coordination of 6. Ruby R, Decety J: What you believe versus what you think they believe: a neuroimaging study of conceptual perspective-tak- multiple cortical areas in brain functions. We are able to ing. Eur J Neurosci 2003, 17:2475-2480. identify the participation of multiple neural networks in 7. Fonagy P, Gergely G, Jurist EL: Affect Regulation, Mentalization, and the Development of the Self. John Wiley & Sons. New York; processing highly differentiated cognitive aspects of emo- tion. Ultimately, placing cortical areas within the context 8. Decety J, Lamm C: Human empathy through the lens of social of a particular neural network may be the key to defining neuroscience. ScientificWorldJournal 2006, 6:1146-1163. 9. Preston SD, de Waal FBM: Empathy: its ultimate and proximate the functional relevance of individual cortical areas. This bases. Behav Brain Sci 2002, 25:1-72. more sensitive approach gives us a better picture of the 10. Iacoboni M, Molnar-Szakacs I, Gallese V, Buccino G, Mazziotta JC, Riz- constituent components involved in the processes of rec- zolatti G: Grasping the intentions of others with one's own mirror neuron system. PLoS Biology 2005, 3:e79. 0529–0535 ognizing and empathizing with emotional facial expres- 11. Jackson PL, Brunet E, Meltzoff AN, Decety J: Empathy examined sions by discerning networks of activity, rather than through the neural mechanisms involved in imagining how I feel versus how you feel pain. Neuropsychologia 2005, 44:752-761. simply defining areas of activity [28,29]. This in turn gives 12. Calder AJ, Young AW: Understanding the recognition of facial us a better idea of how the functional connectivity of con- identitiy and facial expression. Nat Neurosci 2005, 6:641-651. stituent regions work together in order to allow a person 13. Haxby JV, Gobbini MI, Furey ML, Ishai A, Schouten JL, Pietrini P: Dis- tributed and overlapping representations of faces and to recognize and process the facial expressions of another. objects in ventral temporal cortex. Science 2001, 293:2425-2430. 14. Posamentier MT, Abdi H: Processing faces and facial expres- Competing interests sions. Neuropsychol Rev 2003, 13:113-143. The authors declare that they have no competing interests. 15. Sinha P, Balas B, Ostrovsky Y, Russell R: Face recognition by humans: Nineteen results all computer vision researchers should know about. IEEE 2006, 94:1948-1962. Authors' contributions 16. Crocker V, McDonald S: Recognition of emotion from facial JSN performed the statistical analysis and drafted the expression following traumatic brain injury. Brain Inj 2005, manuscript. DS participated in the design of the study, 19:787-799. 17. Ashley V, Vuilleumier P, Swick D: Time course and specificity of programmed the stimuli and performed the acquisition of event-related poetentials to emotional expressions. Neurore- the fMRI data. SF participated in the fMRI experiment and port 2003, 15:211-216. 18. Blau VC, Maurer U, Tottenham N, McCandliss BD: The face-spe- the statistical analysis. RS participated in the design of the cific N170 component is modulated by emotional facial study and performed the psychological investigations. MF expression. Behav Brain Funct 2007, 3:7. participated in the design of the study and contributed to 19. Liu J, Harris A, Kanwisher N: Stages of processing in face percep- tion: an MEG study. Neuroscience 2002, 5:910-916. the interpretation of the data and drafting of the manu- 20. Kaufmann J, Schweinberger SR: Expression influences the recog- script. HJW participated in the design of the study and nition of familiar faces. Perception 2004, 33:399-408. participated in the acquisition of the fMRI data. NPA con- 21. Gallese V, Keysers C, Rizzolatti G: A unifying view of the basis of social cognition. Trends Cogn Sci 2004, 8(9):396-403. tributed to the design of the study and participated in the 22. Uddin LQ, Molnar-Szakacs I, Zaidel E, Iacoboni M: fTMS to the interpretation of the data and drafting of the manuscript. right inferior parietal lobule disrupts self-other discrimina- tion. Soc Cogn Affect Neurosci 2007, 1:65-71. JM performed the statistical analysis and participated in 23. Schütz-Bosbach S, Mancini B, Aglioti SM, Haggard P: Self and other the interpretation of the data. RJS participated in the in the human motor system. Curr Biol 2006, 16:1830-1834. design of the study, contributed to the statistical analysis, 24. Decety J, Grezes J: The power of simulation: Imagining one's own and other's behavior. Brain Res 2006, 12:4-14. the data interpretation, and drafting the manuscript. All 25. Seitz RJ, Nickel J, Azari NP: Functional modularity of the medial authors read and approved the final manuscript. prefrontal cortex: involvement in human empathy. Neuropsy- chology 2006, 20:743-751. 26. Decety J, Jackson PL: The functional architecture of Human Acknowledgements Empathy. Behav Cogn Neurosci Rev 2004, 3:71-100. The authors thank E. Rädisch for expert technical assistance during the 27. Horwitz B: Data analysis paradigms for metabolic-flow data: fMRI experiment. Combining neural modeling and functional neuroimaging. Human Brain Mapp 1994, 2:112-122. 28. Alexander GE, Moeller JR: Application of the subprofile model to functional imaging in neuropsychiatric disorders: A prin- cipal component approach to modeling brain function in dis- ease. Human Brain Mapp 1994, 2:79-94. Page 11 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 29. Friston KJ, Holmes AP, Worsely KJ, Poline J-P, Frith CD, Frackowiak 56. Puce A, Allison T, Asgari M, Gore JC, McCarthy G: Differential sen- RSJ: Statistical parametric maps in functional imaging: A gen- sitivity of human visual cortex to faces, letterstrings, and tex- eralized linear approach. Human Brain Mapp 1994, 2:189-210. tures: A functional magnetic resonance imaging study. J 30. McIntosh AR: Towards a network theory of cognition. Neural Neurosci 1996, 16:5205-5215. Netw 2000, 13:861-870. 57. Rossion B, Caldara R, Seghier M, Schuller A-M, Lazeyras F, Mayer E: 31. Seitz RJ, Knorr U, Azari NP, Weder B: Cerebral networks in sen- A network of occipito-temporal face-sensitive areas besides sorimotor distrubances. Brain Res Bull 2001, 54:299-305. the right middle fusiform gyrus is necessary for normal face 32. Oldfield RC: The assessment and analysis of handedness: the processing. Brain 2003, 126:2381-2395. Edinburgh inventory. Neuropsychologia 1971, 9:97-113. 58. Rotshtein P, Henson RNA, Treves A, Driver J, Dolan RJ: Morphing 33. Bagby RM, Parker JDA, Taylor GJ: The Twenty-Item Toronto Marilyn into Maggie dissociates physical and identity face Alexithymia Scale-I. Item selection and cross-validation of representations in the brain. Nat Neuroscience 2005, 8:107-113. the factor structure. J Psychosomat Res 1994, 38:23-32. 59. Schiltz C, Sorger B, Caldara R, Ahmed F, Mayer E, Goebel R, Rossion 34. Bagby RM, Taylor GJ, Parker JDA: The Twenty-Item Toronto B: Impaired face discrimination in acquired prosopagnosia is Alexithymia Scale-II. Convergent, Discriminant, and Con- associated with abnormal response to individual faces in the current Validity. J Psychosomat Res 1994, 38:33-40. right middle fusiform gyrus. Cereb Cortex 2006, 16(4):574-586. 35. Franz M, Popp K, Schaefer R, Sitte W, Schneider C, Hardt J, Decker 60. Sergent J, Ohta S, MacDonald B: Functional neuroanatomy of O, Braehler E: Alexithymia in the German general population. face and object processing. Brain 1992, 115:15-36. Soc Psychiatry Psych Epidemiol 2008, 43:54-62. 61. Steeves JK, Culham JC, Duchaine BC, Pratesi CC, Valyear KF, Schin- 36. Hautzinger M, Bailer M, Worall H, Keller F: BDI. Beck-Depres- dler I, Humphrey GK, Milner AD, Goodale MA: The fusiform face sions-Inventar. In Huber Bern Göttingen Toronto Seattle; 1995. area is not sufficient for face recognition: Evidence from a 37. Behr M, Becker M: SEE. Skalen zum Erleben von Emotionen. patient with dense prosopagnosia and no occipital face area. In Manual Göttingen: Hogrefe; 2004. Neuropsychologia 2006, 44:594-609. 38. Ekman P, Friesen WV: Pictures of Facial Affect [slides]. Palo 62. Yovel G, Kanwisher N: The neural basis of the behavioral face- Alto, CA: Consulting Psychologists Press; 1996. inversion effect. Curr Biol 2005, 15:2256-2262. 39. Esslen M, Pascual-Marqui RD, Hell D, Kochi K, Lehmann D: Brain 63. Allison T, Giner H, McCarthy G, Nore AC, Puce A, Luby M, Spencer areas and time course of emotional processing. Neuroimage DD: Face recognition in human extrastriate cortex. J Neuro- 2004, 21:1189-203. physiol 1994, 71:821-825. 40. Franz M, Schaefer R, Schneider C, Sitte W, Bachor J: Visual event- 64. Winston JS, Henson RNA, Fine-Goulden MR, Dolan RJ: fMRI-adap- related potentials in subjects with alexithymia: modified tation reveals dissociable neural representations of identity processing of emotional aversive information. Am J Psychiat and expression in face perception. J Neurophysiol 2004, 2004, 161:728-735. 92:1830-1839. 41. Kleiser R, Wittsack H-J, Bütefisch CM, Jörgens S, Seitz RJ: Functional 65. Takamura M: Prosopagnosia: A look at the laterality and spe- activation within the PI-DWI mismatch region in recovery cificity issues using evidence from neuropsychology and neu- from hemiparetic stroke: preliminary observations. Neuroim- rophysiology. The Harvard Brain 1996, Spring:9-13. age 2005, 24:515-523. 66. Ramnani N, Owen AM: Anterior prefrontal cortex: Insights into 42. Jackson J: A User's Guide to Principal Components. John Wiley function from anatomy and neuroimaging. Nat Rev Neurosci & Sons: New York; 1991. 2004, 3:184-194. 43. Talairach J, Tournoux P: Co-planar stereotaxic atlas of the 67. Sugiura M, Friston KJ, Willmes K, Shah NJ, Zilles K, Fink GR: Analysis human brain. New York: Thieme; 1998. of intersubject variability in activation: an application to the 44. Brett M: The MNI brain and the Talairach atlas. 14-2-2002. Ref incidental episodic retrieval during recognition test. Human Type: Internet Communication . Brain Mapp 2007, 28:49-58. 45. Gauthier : The fusiform 'face area' is part of a network that 68. Azari NP, Missimer J, Seitz RJ: Religious experience and emotion: processes faces at the individual level. J Cogn Neurosci 2000, Evidence for distinctive neural patterns. Int J Psychol Religion 12:495-504. 2005, 15:263-281. 46. George N, Dolan RJ, Fink GR, Baylis GC, Russell C, Driver J: Con- 69. Mobbs D, Garrett AS, Menon V, Rose FE, Bellugi U, Reiss AL: Anom- trast polarity and face recognition in the human fusiform alous brain activation during face and gaze processing in Wil- gyrus. Nat Neuroscience 1999, 2:574-580. liams syndrome. Neurology 2004, 62:2070-2076. 47. Hadjikhani N, Gelder B: Neural basis of prosopagnosia. Human 70. Luks TL, Simpson GV: Preparatory deployment of attention to Brain Mapp 2002, 16:176-182. motion activates higher-order motion-processing brain 48. Halgren E, Dale AM, Sereno MI, Tootell R, Marinkovic K, Rosen BR: regions. Neuroimage 2004, 22:1515-1522. Location of human face-selective cortex with respect to reti- 71. Gilbert SJ, Spengler S, Simons JS, Steele JD, Lawrie SM, Frith CD, Bur- notopic areas. Human Brain Mapp 1999, 7:29-37. gess PW: Functional specialization within rostral prefrontal 49. Ishai A, Ungerleider LG, Haxby JV: Distributed neural systems for cortex (area 10): a meta-analysis. J Cogn Neurosci 2006, the generation of visual images. Neuron 2000, 28:979-990. 18:932-948. 50. Kanwisher N, McDermott J, Chun MM: The fusiform face area: A 72. Kanwisher N, Tong F, Nakayama K: The effect of face inversion module in human extrastriate cortex specialized for face on the human fusiform face area. Cognition 1998, 68:1-11. perception. J Neurosci 1997, 17:4302-4311. 73. Dolan RJ, Fletcher P, Morris J, Kapur N, Deakin JF, Frith CD: Neural 51. Lerner Y, Hendler T, Ben-Bashat D, Harel M, Malach R: A hierarchi- activation during covert processing of positive emotional cal axis of object processing stages in the human visual cor- facial expressions. Neuroimage 1996, 4:194-200. tex. Cereb Cortex 2001, 11:287-297. 74. Hoffman EA, Haxby JV: Distinct representations of eye gaze and 52. Nakamura K, Kawashima R, Sato N, Nakamura , Sugiura M, Kato T, identity in the distributed human neural system for face per- Hatano K, Ito K, Fukuda H, Schormann T, Zilles K: Functional ception. Nat Neurosci 2000, 3:80-84. delineation of the human occipitoltemporal areas related to 75. Pessoa L, McKenna E, Gutierrez E, Ungerleider LG: Neural process- face and scene processing; a PET study. Brain 2000, ing of emotional faces requires attention. Proc Natl Acad Sci 123:1903-1912. 2002, 99:11458-11463. 53. Narumoto J, Okada T, Sadato N, Fukui K, Yonekura Y: Attention to 76. Nakamura K, Kawashima R, Ito K, Sugiura M, Kato T, Nakamura A, emotion modulates fMRI activity in human right superior Hatano K, Nagumo S, Kubota K, Fukuda H, Kojima S: Activation of temporal sulcus. Cogn Brain Res 2001, 12(2):225-231. the right inferior frontal cortex during assessment of facial 54. Onitsuka T, Shenton ME, Kasai K, Nestor PG, Toner SK, Kikinis R, emotion. J Neurophysiol 1999, 82:1610-1614. Jolesz FA: Fusiform gyrus volume reduction and facial recog- 77. Powell EW, Leman RB: Connections of the nucelus accumbens. nition in chronic schizophrenia. Arch Gen Psychiatry 2003, Brain Res 1976, 105:389-403. 60(4):349-355. 78. Roberts AC, Tomic DL, Parkinson CH, Roeling TA, Cutter D, Rob- 55. Pizzagalli DA, Lehmann D, Hendrick AM, Regard M, Pascual-Marqui bins TW, Everitt BJ: Forebrain connectivity of the prefrontal RD, Davidson RJ: Affective judgments of faces modulate early cortex in the marmoset monkey (Callithrix jacchus): an activity (~160 ms) within the fusiform gyri. Neuroimage 2002, anterograde and retrograde tract-tracing study. J Comp Neu- 16:663-677. rol 2007, 502:86-112. Page 12 of 13 (page number not for citation purposes) Behavioral and Brain Functions 2008, 4:41 http://www.behavioralandbrainfunctions.com/content/4/1/41 79. Tekin S, Cummings JL: Frontal-subcortical neuronal circuits and 103. Fu CHY, Williams SCR, Brammer MJ, Suckling J, Kim J, Cleare AJ, clinical neuropsychiatry: An update. J Psychosomat Res 2002, Walsh ND, Mitterschiffthaler MT, Andrew CM, Pich EM, Bullmore 53:647-654. ET: Neural resonses to happy facial expressions in major 80. Siffert J, Allen JC: Late effects of therapy of thalamic and depression following antidepressant treatment. Am J Psych hypothalamic tumors in childhood: Vascular, neurobehavio- 2006, 164:599-607. ral and neoplastic. Pediatr Neurosurg 2000, 33:105-111. 104. Servos P, Osu R, Santi A, Kawato M: The neural substrates of bio- 81. Corbetta M, Kincade JM, Ollinger JM, McAvoy MP, Shulman GL: Vol- logical motion perception: An fMRI study. Cereb Cortex 2002, untary orienting is dissociated from target detection in 12:772-782. human posterior parietal cortex. Neuroscience 2000, 3:292-297. 105. Barch DM, Braver TS, Akbudak E, Conturo' T, Ollinger J, Snyder A: 82. Corbetta M, Shulman GL: Human cortical mechanisms of visual Anterior cingulate cortex and response conflict: Effects of attention during orienting and search. Phil Trans R Soc Lond response modality and processing domain. Cereb Cortex 2001, 1998, 353:1353-1362. 11:837-848. 83. Hariri AR, Bookheimer SY, Mazziotta JC: Modulatory emotional 106. Bush G, Luu P, Posner MI: Cognitive and emotional influences in responses: effect of a neurocortical network on the limbic anterior cingulate cortex. Trends Cogn Sci 2000, 4(6):215-222. system. Neuroreport 2000, 11:43-48. 107. Fichtenholtz HM, Dean HL, Dillon DG, Yamasaki H, McCarthy G, 84. Tamm L, Menon V, Reiss A: Parietal attentional system aberra- LaBar KS: Emotion-attention network interactions during a tions during target detection in adolescents with attention visual oddball task. Cogn Brain Res 2004, 20:67-80. deficit hyperactivity disorder: Event-related fMRI evidence. Am J Psych 2006, 163:1033-1043. 85. Garavan H, Ross TJ, Stein EA: Right hemispheric dominance of inhibitory control: an event-related functional MRI study. Proc Natl Acad Sci USA 1999, 96:8301-8306. 86. Decety J, Chaminade T, Grezes J, Meltzoff AN: A PET exploration of the neural mechanisms involved in reciprocal imitation. Neuroimage 2001, 15:265-272. 87. Frith U, Frith CD: Development and neurophysiology of men- talizing. Phil Trans R Soc Lond 2003, 358:459-473. 88. Farrer C, Franck N, Georgieff N, Frith CD, Decety J, Jeannerod M: Modulating the experience of agency: a positron emission tomography study. Neuroimage 2003, 18:324-333. 89. Völlm BA, Taylor AN, Richardson P, Corcoran R, Stirling J, McKie S, Deakin JF, Elliot R: Neuronal correlates of theory of mind and empathy: A functional magnetic resonance imaging study in a nonverbal task. Neuroimage 2005, 29:90-98. 90. Blanke O, Mohr C, Michel CM, Pascual-Leone A, Brugger P, Seeck M, Landis T, Thut G: Linking out-of-body experience and self processing to mental own-body-imagery at the temporopa- rietal junction. J Neurosci 2005, 25:550-557. 91. Karnath H, Himmelbach M, Rorden C: The subcortical anatomy of human spatial neglect: putamen, caudate nucleus and pul- vinar. Brain 2002, 125:350-360. 92. Hedden T, Gabrieli JDE: The ebb and flow of attention in the human brain. Nat Neurosci 2006, 9:863-865. 93. Saxe R, Kanwisher N: People thinking about thinking people: The role of the temporo-parietal junction in "theory of mind". Neuroimage 2003, 19:1835-1842. 94. Carr L, Iacoboni M, Dubeau M-C, Mazziota JC, Lenzi GL: Neural mechanisms of empathy in humans: A relay from neural sys- tems for imitation to limbic areas. Proc Natl Acad Sci USA 2003, 100:5497-5592. 95. Singer T, Seymour B, O'Doherty J, Kaube H, Dolan RJ, Frith CD: Empathy for pain involves the affective but not sensory com- ponents of pain. Science 2004, 303(5661):1157-1162. 96. Haxby JV, Hoffman EA, Gobbini MI: The distributed human neu- ral system for face perception. TICS 2000, 4:223-233. 97. Schweinsburg AD, Paulus MP, Barlett VC, Killeen LA, Caldwell LC, Pulido C, Brown SA, Tapert SR: An fMRI study of response inhi- bition in youths with a family history of alcoholism. Ann NY Acad Sci 2004, 1021:391-394. 98. Iidaka T, Matsumoto A, Nogawa J, Yamamoto Y, Sadato N: Fron- toparietal network involved in successful retrieval from epi- Publish with Bio Med Central and every sodic memory. Spatial and temporal analyses using fMRI and scientist can read your work free of charge ERP. Cereb Cortex 2006, 16:1349-1360. 99. McIntosh AR, Rajah MN, Lobaugh N: Interactions of prefrontal "BioMed Central will be the most significant development for cortex in relation to awareness in sensory learning. Science disseminating the results of biomedical researc h in our lifetime." 1999, 284:1531-1533. Sir Paul Nurse, Cancer Research UK 100. Adleman NE, Menon V, Blasey CM, White CD, Warsofsky IS, Glover GH, Reiss AL: A developmental fMRI study of the stroop color- Your research papers will be: word task. Neuroimage 2002, 16:61-75. available free of charge to the entire biomedical community 101. Lane RD, Rieman EM, Axelrod B, Yun LS, Holmes A, Schwartz GE: Neural correlates of levels of emotional awareness. Evidence peer reviewed and published immediately upon acceptance of an interaction between emotion and attention in the ante- cited in PubMed and archived on PubMed Central rior cingulate cortex. J Cogn Neurosci 1998, 10:525-535. 102. Schäfer R, Popp K, Jörgens S, Lindenberg R, Franz M, Seitz RJ: Alex- yours — you keep the copyright ithymia-like disorder in right anterior cingulate infarction. BioMedcentral Neurocase 2007, 13:201-208. Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp Page 13 of 13 (page number not for citation purposes)

Journal

Behavioral and Brain Functions – Springer Journals

Published: Sep 17, 2008

There are no references for this article.