Abstract

Negative emotional signals are known to influence task performance, but so far, investigations have focused on how emotion interacts with perceptual processes by mobilizing attentional resources. The attention-independent effects of negative emotional signals are less well understood. Here, we show that threat signals trigger defensive responses independently of what observers pay attention to. Participants were scanned using functional magnetic resonance imaging while watching short video clips of threatening actions and performed either color or emotion judgments. Seeing threatening actions interfered with performance in both tasks. Amygdala activation reflected both stimulus and task conditions. In contrast, threat stimuli prompted a constant activity in a network underlying reflexive defensive behavior (periaqueductal gray, hypothalamus, and premotor cortex). Threat stimuli also disrupted ongoing behavior and provoked motor conflict in prefrontal regions during both tasks. The present results are consistent with the view that emotions trigger adaptive action tendencies independently of task settings.

Introduction

Adaptive behavior entails attending to the task at hand while remaining able to promptly detect and react to relevant or unpredictable threat signals in the environment. How does the brain achieve a balance between these 2? There is now clear evidence that salient stimuli influence the course of perceptual processes. Emotional signal processing, notably that of threatening facial and bodily expressions, is prioritized (Hansen and Hansen 1988; Ohman et al. 2001; Fox and Damjanovic 2006; Tamietto et al. 2007; Becker 2009; Zeelenberg and Bocanegra 2010). Animal and human studies support the view that the amygdala increases vigilance and facilitates perceptual processing (Whalen 1998; Davis and Whalen 2001; Vuilleumier et al. 2004) and is particularly sensitive to threatening stimuli (LeDoux 1995; Morris et al. 1996; Whalen 1998; Phelps and LeDoux 2005; Hoffman et al. 2007). In line with this, several researchers have argued that the amygdala operates relatively independently of whether or not the stimulus is attended to (Vuilleumier et al. 2001; Ohman 2002; Dolan and Vuilleumier 2003). This does not exclude top-down influences as high task demands have been shown to reduce amygdala responses (Pessoa et al. 2002; Bishop et al. 2007; Mitchell et al. 2007; Silvert et al. 2007; Lim et al. 2008; Pessoa 2008).

In addition to facilitating perceptual processing, threat signals also trigger physiological reflexes and influence ongoing behavior (Davis et al. 1993; Ohman et al. 1995; Hamm et al. 1997; Lang et al. 1998; Panksepp 1998; Williams et al. 2005; Pessoa 2009). The underlying rationale seems to be that threatening stimuli enhance responsiveness in regions related to defensive behavior. In monkeys, premotor neurons of the region polysensory zone (PZ) respond to looming stimuli and the electrical stimulation of the same neurons produces defensive movements (Cooke and Graziano 2004; Graziano and Cooke 2006). In humans, observing threatening actions (as compared with neutral or joyful actions) increases activity in regions involved in action preparation: premotor cortex (PM), pre-supplementary motor area (pre-SMA), and inferior frontal gyrus (IFG) (de Gelder et al. 2004; Grosbras and Paus 2006; Grèzes et al. 2007; Pichon et al. 2008, 2009). Also, exposure to anger signals increases activity in the amygdala and the hypothalamus (Pichon et al. 2008, 2009), 2 nuclei that are part of subcorticocortical networks that interface with motor and autonomic systems important for the emotional experience of fear and rage (Bard 1928; Brown et al. 1969; Siegel and Edinger 1983; LeDoux 1995; Panksepp 1998; Canteras 2002; Barbas et al. 2003; Sewards and Sewards 2003; Adams 2006).

But while the effects of threatening stimuli on the attentional demands in perception have already been explored, it remains largely unknown whether attention influences activity in action and defense-related brain areas. Amygdala response to threat can be altered by a high task load, but other brain regions may react relatively independently and continue to support adaptive behavior and action preparation whatever the task at hand. If so, motor-related regions may react differently to task conditions and attentional demands than will the amygdala. Here, we tested this hypothesis using functional magnetic resonance imaging (fMRI). Participants watched movies of threatening and of neutral bodily actions. In one condition, they were requested to name the color of a dot that appeared very briefly on the actor's upper body (color-naming task) while in the other condition they named the emotional expression of the actor (emotion-naming task). The motivation to use a demanding color-naming task was to isolate threat-responsive regions independently of the task requirements. As stimuli, we used dynamic actions expressing fear and anger that were previously shown to elicit strong activations in subcortical and cortical regions important for preparation of defensive behavior (de Gelder et al. 2004; de Gelder 2006; Grèzes et al. 2006; Grosbras and Paus 2006; Pichon et al. 2008, 2009).

Materials and Methods

Participants

Sixteen right-handed volunteers (8 females; mean age = 25.6 years ± SD = 8; and 8 males; mean age = 23.5 years ± 2.6) with no neurological or psychiatric history participated in the imaging study. All provided written informed consent according to institutional guidelines of the local research ethics committee and were paid for their participation.

Stimuli

Seventy-one full-light 3-s videos (23 fear, 24 anger, and 24 neutral) were used for the present experiment. Videos were selected from a larger set of stimuli based on the recognition performance obtained in a pilot study and were previously used (Grezes et al. 2007; Pichon et al. 2008; Pichon et al. 2009). One fear movie was dropped because of frequent misclassification. A group of 12 semiprofessional actors (6 males), all graduates from a professional acting school, were hired in order to account for variability in acting style. They were paid for their services. Under professional direction, they enacted different scenarios corresponding to fearful, angry, and neutral situations. The fear script instructed the actors to open the door and face a threat (human or nonhuman, such as a snake). The anger version of this scenario also consisted of opening the door but at the same time reacting to someone or something that made them angry. We used these 2 emotions because, for the observer, perceiving a fearful expression or an angry expression directed at oneself signals a potential threat (Dimberg 1986; Schupp et al. 2004; Strauss et al. 2005). Therefore, in order to create the impression for the observer to be involved in the scene, the recordings were made with the camera positioned directly facing the door viewed from the outside. For the neutral scenario, the actors were required to open the same door in a neutral way. The actors’ performance was repeated until deemed satisfactory by the director. Importantly, faces were blurred such that only information from the body was available. All video clips were further computer edited to insert a small color dot (red, green, or blue, visual angle = 0.3°) that was briefly flashed during 50 ms at random times in the second half of the movie. To minimize shifts in spatial attention between tasks, the location of the color dot was carefully chosen so that it always fell on the actor's upper body. Colors were randomized across emotional expressions.

Design and fMRI Procedure

We aimed at comparing the neural responses induced by attended or unattended threatening and neutral actions. To manipulate attention, we used 2 tasks (see Fig. 1). During the emotion-naming task, subjects attended video clips of body expressions and were asked to judge whether the action was neutral, signaled fear, or signaled anger. In the color-naming task, subjects were requested to detect a color dot appearing briefly (50 ms) during the video clip and to report whether it was red, green, or blue. Importantly, the video materials including the color dots were the same in both conditions such that actions were to be ignored in the color-naming task and colors were to be ignored in the emotion-naming task.

Figure 1.

Experimental design and stimulus examples. Subjects viewed a video sequence showing an actor expressing a threatening or neutral action and a color dot appearing briefly for 50 ms onto the actor's upper body. Depending on the task, subjects categorized the emotion or the color of the dot.

Figure 1.

Experimental design and stimulus examples. Subjects viewed a video sequence showing an actor expressing a threatening or neutral action and a color dot appearing briefly for 50 ms onto the actor's upper body. Depending on the task, subjects categorized the emotion or the color of the dot.

Figure 2.

Behavioral results. (A) RTs plotted as a function of attention and stimuli expressions showed that threatening stimuli were recognized more slowly than neutral ones. (B) This effect was cumulative in time since RTs were increased when the previous trial contained a threatening stimulus. (C) Mean quantification of movement across expressions was estimated by quantifying, for each video clip, the variation of light luminance between pairs of frames for each pixel. According to this quantification, threat and neutral movies did not differ significantly. (D) Confusion matrix showing that all expressions were clearly recognized above chance during the emotion-naming task (percentages displayed take into account omitted responses that are not displayed here). (E) Mean emotional intensity scores across expressions. Subjects rated threat movies as subjectively more emotionally intense than neutral ones. Error bars represent standard error of the mean. The asterisks denote levels of statistical significance (*P < 0.05, ***P < 0.001).

Figure 2.

Behavioral results. (A) RTs plotted as a function of attention and stimuli expressions showed that threatening stimuli were recognized more slowly than neutral ones. (B) This effect was cumulative in time since RTs were increased when the previous trial contained a threatening stimulus. (C) Mean quantification of movement across expressions was estimated by quantifying, for each video clip, the variation of light luminance between pairs of frames for each pixel. According to this quantification, threat and neutral movies did not differ significantly. (D) Confusion matrix showing that all expressions were clearly recognized above chance during the emotion-naming task (percentages displayed take into account omitted responses that are not displayed here). (E) Mean emotional intensity scores across expressions. Subjects rated threat movies as subjectively more emotionally intense than neutral ones. Error bars represent standard error of the mean. The asterisks denote levels of statistical significance (*P < 0.05, ***P < 0.001).

The experiment was divided into 2 successive scanning runs of 21 min each. Within each run, stimuli were blocked by task and blocks alternated between series of attention to emotion or attention to color tasks. At the beginning of each block, subjects were instructed by a text on the screen lasting 2 s whether they had to recognize emotions or detect colors. Stimuli and null events (5 s) were randomly mixed within blocks. Each task block contained 6 events (including nulls). After each stimulus presentation, subjects were instructed by a response screen (fear/anger/neutral or red/green/blue) to push the corresponding button using a response pad placed in their right hand. Subjects had a delay of 2 s to give their answer. The order of responses was randomized between trials to avoid motor anticipation-related effects. A total of 36 blocks per task was presented (142 video clips + 74 null events per task). Stimuli were back projected onto a screen positioned behind the subject's head and viewed through a mirror attached to the head coil. The stimulus was centered on the display screen and subtended 10.8° of visual angle vertically and 7.3° horizontally.

fMRI Data Acquisition

Gradient-echo T2*-weighted transverse echo-planar images (EPI) with blood oxygenation level–dependent (BOLD) contrast were acquired with a 3-T Siemens Magnetom Trio scanner (Siemens, Erlangen, Germany). Participants used earplugs to attenuate scanner noise, and padding was used to reduce head movements. Each volume contained 32 axial slices (repetition time [TR] = 2000 ms, echo time [TE] = 30 ms, 3.5 mm thickness without gap yielding isotropic voxels of 3.5 mm3, flip angle = 90°, field of view [FOV] = 224 mm, resolution = 64 × 64) acquired in an interleaved manner. An automatic shimming procedure was performed to minimize inhomogeneities of the static magnetic field. We collected a total of 1270 functional volumes for each subject as well as high-resolution T1-weighted anatomical images (TR = 2250 ms, TE = 2.6 ms, slice thickness = 1 mm, 192 sagittal slices, flip angle = 9°, FOV = 256 mm, resolution = 256 × 256). We administered the behavioral protocol using Presentation software (www.neurobs.com).

fMRI Image Processing

Image processing was carried out using SPM2 (Wellcome Department of Imaging Neuroscience; see www.fil.ion.ucl.ac.uk/spm) implemented in MATLAB (Mathworks Inc., Sherborn, MA). The first 5 volumes of each scanning run were discarded to avoid T1 saturation effects. The remaining 1260 functional images were reoriented to the AC–PC line, corrected for differences in slice acquisition time using the middle slice in time as reference, spatially realigned to the first volume by rigid body transformation, spatially normalized to the standard Montreal Neurological Institute (MNI) EPI template to allow group analysis, resampled to an isotropic voxel size of 2 mm, and spatially smoothed with an isotropic 8-mm full width at half-maximum Gaussian kernel (Friston et al. 1995).

Behavioral Analysis

Reaction times (RTs) and accuracy were analyzed by means of repeated measures analyses of variance (ANOVAs) with task (emotion and color) and expression (threat and neutral) as within-subject factors. Two-tailed paired t-tests were used for comparisons between experimental conditions. We considered 3 significance levels: 0.05, 0.01, and 0.001.

fMRI Image Analysis

A 2-stage general linear model (GLM) was used to examine the effect sizes of each condition and compare them to the group level. The statistical analyses were also carried out using SPM2.

At the subject level, we performed standard GLM analyses where task-related signal changes were modeled separately for each subject. Fear and anger expressions were collapsed into a single regressor to estimate threat-specific effects. Beforehand, we used the behavioral data to ensure that the recognition of fear and anger did neither differ in terms of main effects of task or expression (repeated measures ANOVA, P > 0.4 and P > 0.08) nor was there an interaction between both factors although the P value was close to significance (P = 0.056). Additional paired t-tests showed that the recognition rates between anger and fear did differ neither in the color task (T1,15 = −0.68, P = 0.503) nor in the emotion task (T1,15 = 1.8, P = 0.084). For RTs, neither main effects of task or expression nor their interaction reached significance (all P values > 0.35). We modeled 5 conditions for each session: 4 conditions modeled the occurrence of threat and neutral stimuli within each task (emotion and color). The fifth condition was used to model the instruction screen preceding each block. For each condition, a covariate was calculated by convolving the onset of each event with a canonical hemodynamic response function over a duration that encompassed the stimulation (3 s) and the response (2 s) periods. Six additional covariates were modeled per session, corresponding to the temporal derivatives of the realignment parameters (the difference between scans in the estimation of the 3 rigid body translations and the 3 rotations determined from initial spatial registration) in order to capture residual movement-related artifacts. A last constant covariate represented the session-specific mean over scans. The model included a high-pass filter with a standard cutoff period of 128 s to account for low-frequency noise of the scanner. Effects at each brain voxel were estimated using a least squares algorithm to produce condition-specific images of parameter estimates for each experimental condition.

At the group level, we performed 2-way repeated measures ANOVA with the factors task (emotion and color) and expression (threat and neutral). In this way, the variance estimates at the group level incorporated appropriately weighted within-subject and between-subject variance effects. A nonsphericity correction was applied for variance differences across conditions. After model estimation, we calculated the following contrasts:

  1. To isolate threat-driven responses independently of task, we performed a conjunction analysis (null hypothesis) between the 2 following contrasts: emotion-naming task (threat vs. neutral) and color-naming task (threat vs. neutral).

  2. Within threat-responsive regions revealed in 1, we performed correlation analyses with the slowing in response times observed when subjects processed threat stimuli in both tasks. We used the magnitude of the contrast of threat versus neutral stimuli (estimated for each subject at the first level) and the difference in response time between threat and neutral conditions. These correlations were calculated separately for emotion- and color-naming tasks (see Supplementary Table S3).

  3. We then estimated brain regions showing a task-by-expression interaction, that is, regions that showed a greater differential response between threat and neutral stimuli in the emotion-naming task as compared with the color-naming task (see Fig. 4 and Table 2, inclusive masking procedure using the simple effect of naming threatening vs. neutral actions).

  4. To isolate responses showing an additive effect of threat and task, we estimated a conjunction (see Table 3) between the following contrasts: emotion-naming task (threat vs. neutral), color-naming task (threat vs. neutral), threat stimuli (emotion-naming vs. color-naming task), and neutral stimuli (emotion-naming vs. color-naming task).

  5. We calculated the main effects specific to each task to ensure they were associated with consistent BOLD responses within brain regions engaged in emotion- and color-naming tasks (see Fig. 5, Supplementary Figs S1 and S2 and Supplementary Tables S1 and S2).

Table 1

Conjunction analysis of brain regions showing an effect of threat (vs. neutral) in both tasks

R/L Anatomical region MNI coordinates
 
Z value Size in voxels 
  x y z   
pre-SMA −2 24 54 3.21a 
Inferior frontal sulcus (BA44) −44 12 32 3.92a 227 
L and R Anterior insula ±38 30 −6 3.17/3.3a 6/2 
PM 52 52 3.28a 11 
Thalamus −10 3.1a 
Hypothalamus dorsal 10 −8 −4 3.03b 88 
PAG −26 −4 2.63b ↓ 88 
Fusiform gyrus −40 −44 −22 4a 96 
Fusiform gyrus 44 −48 −20 3.65a 29 
STS, posterior part 46 −36 −2 5.17a ↓ 1912 
L and R STS, posterior part ±50 −50 4.76/4.85a ↓ 1067/↓ 1912 
L and R Middle temporal gyrus (MTV5/EBA) ±48 −66 5.28/7.16a 1067/1912 
Occipital pole −16 −104 3.37b 20 
R/L Anatomical region MNI coordinates
 
Z value Size in voxels 
  x y z   
pre-SMA −2 24 54 3.21a 
Inferior frontal sulcus (BA44) −44 12 32 3.92a 227 
L and R Anterior insula ±38 30 −6 3.17/3.3a 6/2 
PM 52 52 3.28a 11 
Thalamus −10 3.1a 
Hypothalamus dorsal 10 −8 −4 3.03b 88 
PAG −26 −4 2.63b ↓ 88 
Fusiform gyrus −40 −44 −22 4a 96 
Fusiform gyrus 44 −48 −20 3.65a 29 
STS, posterior part 46 −36 −2 5.17a ↓ 1912 
L and R STS, posterior part ±50 −50 4.76/4.85a ↓ 1067/↓ 1912 
L and R Middle temporal gyrus (MTV5/EBA) ±48 −66 5.28/7.16a 1067/1912 
Occipital pole −16 −104 3.37b 20 
a

Indicates activation peaks that survive false discovery rate correction (P < 0.05) with a height threshold of P < 0.001 (uncorrected). MTV5, middle temporal V5 complex; EBA, extrastriate body area.

b

Indicates peaks in a priori regions (amygdala, hypothalamus, and PAG) that survive SVC (FWE P < 0.05) at P < 0.005 uncorrected. Subpeaks in clusters marked with ↓

Table 2

Brain regions showing a significant interaction for attended threat

R/L Anatomical region MNI coordinates
 
Z value Size in voxels 
  x y z   
Medial superior frontal gyrus (dmPFC BA9) −6 64 28 4.2a 249 
Posterior orbital gyrus −28 22 −20 3.91 49 
IFG (BA45) 50 32 −2 3.2 
Anterior hippocampus (extending to amygdala) −16 −14 −16 5.44a 118 
Anterior hippocampus (80%)/amygdala (50%) 32 −8 −22 3.9 38 
Pulvinar 14 −30 3.7 27 
Thalamus −4 −16 3.23 
Parahippocampal gyrus −12 −34 −4 3.35 16 
Temporoparietal junction, supramarginal gyrus −56 −40 24 3.31 
Fusiform gyrus 38 −56 −12 3.31 
Fusiform gyrus 40 −58 −10 3.14 
STS 50 −16 −10 3.22 
STS 46 −28 3.17 
STS 62 −32 3.18 
R/L Anatomical region MNI coordinates
 
Z value Size in voxels 
  x y z   
Medial superior frontal gyrus (dmPFC BA9) −6 64 28 4.2a 249 
Posterior orbital gyrus −28 22 −20 3.91 49 
IFG (BA45) 50 32 −2 3.2 
Anterior hippocampus (extending to amygdala) −16 −14 −16 5.44a 118 
Anterior hippocampus (80%)/amygdala (50%) 32 −8 −22 3.9 38 
Pulvinar 14 −30 3.7 27 
Thalamus −4 −16 3.23 
Parahippocampal gyrus −12 −34 −4 3.35 16 
Temporoparietal junction, supramarginal gyrus −56 −40 24 3.31 
Fusiform gyrus 38 −56 −12 3.31 
Fusiform gyrus 40 −58 −10 3.14 
STS 50 −16 −10 3.22 
STS 46 −28 3.17 
STS 62 −32 3.18 
a

Indicates activation peaks that survive false discovery rate correction (P < 0.05) with a height threshold of P < 0.001 (uncorrected). dmPFC, dorsomedial prefrontal cortex.

Table 3

Brain regions showing an additive effect of threat and the emotion recognition task

R/L Anatomical region MNI coordinates
 
Z value Size in voxels 
  x y z   
Lateral orbitofrontal cortex (BA47) 38 30 −6 
IFG (dorsal BA45) −54 22 24 3.39 19 
Temporoparietal junction, supramarginal gyrus 54 −34 24 3.91a ↓ 1352 
L and R Fusiform gyrus (fusiform body area) ±40 −42 −24 3.98/3.65a 33/27 
STS, posterior part 48 −32 −4 4.98a ↓ 1352 
STS, posterior part −58 −62 4.12a 270 
L and R STS, posterior part ±50 −50 3.86/4.85a ↓ 270/↓ 1352 
STS, horizontal segment 52 −60 5.43a 1352 
Middle temporal gyrus (MTV5/EBA) 52 −74 3.32a ↓ 1352 
Inferior occipital gyrus −44 −86 −6 4.24a 28 
R/L Anatomical region MNI coordinates
 
Z value Size in voxels 
  x y z   
Lateral orbitofrontal cortex (BA47) 38 30 −6 
IFG (dorsal BA45) −54 22 24 3.39 19 
Temporoparietal junction, supramarginal gyrus 54 −34 24 3.91a ↓ 1352 
L and R Fusiform gyrus (fusiform body area) ±40 −42 −24 3.98/3.65a 33/27 
STS, posterior part 48 −32 −4 4.98a ↓ 1352 
STS, posterior part −58 −62 4.12a 270 
L and R STS, posterior part ±50 −50 3.86/4.85a ↓ 270/↓ 1352 
STS, horizontal segment 52 −60 5.43a 1352 
Middle temporal gyrus (MTV5/EBA) 52 −74 3.32a ↓ 1352 
Inferior occipital gyrus −44 −86 −6 4.24a 28 
a

Indicates activation peaks that survive false discovery rate correction (P < 0.05) with a height threshold of P < 0.001 (uncorrected). Subpeaks in clusters marked with ↓. MTV5, middle temporal V5 complex; EBA, extrastriate body area.

Figure 3.

Statistical map representing threat-responsive regions, irrespective of the task performed by subjects. (A) On the left, group (n = 16) average activations and percent signal change at local maxima in the PAG extending to posterior medial hypothalamus (HYP) and right lateral PM. (B) On the right, group average activations and percent signal change at local maxima in pre-SMA, left IFG BA44, and anterior insula for the same contrast across tasks. The lower scatterplots show that during exposure to threat, anterior insula (right scatterplot) was significantly correlated with slowing in response times in both tasks, whereas activity in lateral PM, PAG, or HYP was not (left scatterplot). This correlation was also true in the demanding color-naming task for the pre-SMA and left BA44 (see Supplementary Table S3). Activations were rendered on sections of the ICBM-152 average T1-weighted brain (SPM(t), thresholded at P < 0.005 uncorrected for illustrative purpose only, error bars represent 95% confidence interval, and emotion- and color-naming tasks represented, respectively, by the red solid line and blue dotted line).

Figure 3.

Statistical map representing threat-responsive regions, irrespective of the task performed by subjects. (A) On the left, group (n = 16) average activations and percent signal change at local maxima in the PAG extending to posterior medial hypothalamus (HYP) and right lateral PM. (B) On the right, group average activations and percent signal change at local maxima in pre-SMA, left IFG BA44, and anterior insula for the same contrast across tasks. The lower scatterplots show that during exposure to threat, anterior insula (right scatterplot) was significantly correlated with slowing in response times in both tasks, whereas activity in lateral PM, PAG, or HYP was not (left scatterplot). This correlation was also true in the demanding color-naming task for the pre-SMA and left BA44 (see Supplementary Table S3). Activations were rendered on sections of the ICBM-152 average T1-weighted brain (SPM(t), thresholded at P < 0.005 uncorrected for illustrative purpose only, error bars represent 95% confidence interval, and emotion- and color-naming tasks represented, respectively, by the red solid line and blue dotted line).

Figure 4.

Attentional modulation in hippocampus and amygdala. Group average activation and percent signal change showing (A) the interaction for attended threat in the right anterior hippocampus/posterior amygdala and (B) the bilateral amygdala response to threat in the emotion-naming task. (C) Contrasting threat to neutral stimuli in the color-naming task yielded no significant response in bilateral hippocampus or amygdala, even using a liberal threshold of P = 0.005 uncorrected as shown in the present contrast. Plots were computed on local activation peaks and use the same conventions as in Figure 3; SPM(t) thresholded at P < 0.001 uncorrected (<0.005 for C).

Figure 4.

Attentional modulation in hippocampus and amygdala. Group average activation and percent signal change showing (A) the interaction for attended threat in the right anterior hippocampus/posterior amygdala and (B) the bilateral amygdala response to threat in the emotion-naming task. (C) Contrasting threat to neutral stimuli in the color-naming task yielded no significant response in bilateral hippocampus or amygdala, even using a liberal threshold of P = 0.005 uncorrected as shown in the present contrast. Plots were computed on local activation peaks and use the same conventions as in Figure 3; SPM(t) thresholded at P < 0.001 uncorrected (<0.005 for C).

Figure 5.

Effect of attention in the fusiform gyrus. Although visual stimulation was kept identical between tasks, activations in the fusiform gyrus, as shown here, reflected the attentional and perceptual requirements of each task. (A) During the body emotion-naming task, we observed bilateral activation in fusiform body area (Peelen and Downing 2005). The parameter estimates confirmed that activity in fusiform body area was higher during emotion- (red solid line) than color-naming task (blue dotted line). (B) During color-naming task, bilateral activations were detected in the human homologue of V4, known to play a crucial role in color perception and visual attention, especially during demanding tasks (Zeki et al. 1991; Gallant et al. 2000). The parameter estimates confirmed that activity in this regions was higher during color-naming than body emotion-naming task (both SPMs superimposed on axial sections of the ICBM-152 average T1-weighted brain; SPM(t) thresholded at P < 0.001 (uncorrected), plots of percent signal change used the same conventions as in Fig. 3).

Figure 5.

Effect of attention in the fusiform gyrus. Although visual stimulation was kept identical between tasks, activations in the fusiform gyrus, as shown here, reflected the attentional and perceptual requirements of each task. (A) During the body emotion-naming task, we observed bilateral activation in fusiform body area (Peelen and Downing 2005). The parameter estimates confirmed that activity in fusiform body area was higher during emotion- (red solid line) than color-naming task (blue dotted line). (B) During color-naming task, bilateral activations were detected in the human homologue of V4, known to play a crucial role in color perception and visual attention, especially during demanding tasks (Zeki et al. 1991; Gallant et al. 2000). The parameter estimates confirmed that activity in this regions was higher during color-naming than body emotion-naming task (both SPMs superimposed on axial sections of the ICBM-152 average T1-weighted brain; SPM(t) thresholded at P < 0.001 (uncorrected), plots of percent signal change used the same conventions as in Fig. 3).

For all statistical maps, we report activations that survived the threshold of T > 3.28 (P < 0.001, uncorrected) and indicate peaks that survived false discovery rate correction (Genovese et al. 2002). We also inspected all contrasts with a liberal threshold of P = 0.005 in a priori regions of interest (amygdala, hypothalamus, and periaqueductal gray [PAG]) and reported these brain areas when they survived small volume correction (SVC) (familywise error [FWE], P < 0.05) (peaks from the meta-analysis of Kober et al. 2008, [±2 −30 −6] for PAG and [±10 −6 −6] for hypothalamus, SVC 5 mm). Illustrations of maps were overlaid on the ICBM-152 brain template. Anatomical labeling was performed with reference to the atlas of Duvernoy (1999) and the anatomy toolbox (Eickhoff et al. 2005). Surface rendering of statistical maps and estimation of Brodmann areas (BAs) were carried out using Caret (Van Essen et al. 2001) and the PALS-B12 atlas (Van Essen 2005), an average brain atlas derived from structural MRI volumes of 12 normal young adults that were adjusted to the ICBM-152 space. Finally, to ensure that threat-related responses were not led by fear or anger only, we estimated a new GLM where fear and anger responses were modeled separately and compared using post hoc t-tests fear and anger responses in the above regions of interest (see Figs 3 and 4).

Results

Quantitative and Qualitative Analyses of Video Clips

In order to control for quantitative differences in movement between the anger, fear, and neutral movies, we estimated the amount of movement per video clip by quantifying the variation of light intensity (luminance) between pairs of frames for each pixel. For each frame, these absolute differences were averaged across pixels that scored higher than 10 (on a scale reaching a maximum of 255), a value which corresponds to the noise level of the camera. These estimations were then averaged for each movie, and the resulting scores were used to test the hypothesis of a difference in movement between expressions. Mean quantification of movement for threat and neutral movies (see Fig. 2C) were, respectively, 40.69 ± 6.18 and 40.03 ± 4.82 (40.88 ± 7.56 for fear and 41.12 ± 6.72 for anger). No significant differences were detected between threat and neutral movies or between fear, anger, and neutral movies (all P values > 0.4).

To assess potential differences in emotional intensity between fear and anger movies, each movie was also rated by a different group of 39 subjects using a graded Likert scale with extremities labeled “low” and “high” (27 females, mean age = 22.63 ± 2.47 years; and 12 males, mean age = 21.45 ± 2.07 years). Subjects could slide a mouse cursor along this scale and the scores collected ranged from 0 to 100. Mean estimations of intensity for threat and neutral movies (see Fig. 2E) were, respectively, 47.1 ± 12.9 and 12.3 ± 19 (48.1 ± 13.2 for fear and 46.2 ± 13.6 for anger). A repeated measure ANOVA revealed a significant difference between fear, anger, and neutral expressions (F2,74 = 99.18, P < 0.001, Greenhouse–Geisser sphericity correction). Post hoc paired t-tests corrected for multiple comparisons showed that whereas fear and anger movies were rated similarly (T1,37 = 1.59, P = 0.36), they were perceived as more intense than neutral movies (respectively, T1,37 = 10.51, P < 0.001, and T1,37 = 10, P < 0.001).

Behavioral Results

RTs and accuracy were analyzed by means of repeated measures ANOVAs with task (emotion and color) and expression (threat and neutral) as within-subject factors. RTs showed that threatening stimuli influenced performance in both emotion- and color-naming tasks. Subjects’ responses were slower for threatening stimuli compared with neutral ones (main effect: P < 0.05; simple effect in the emotion-naming task: P = 0.11; simple effect in the color task: P < 0.05; Fig. 2A). We also observed a cumulative effect across trials since response time also increased when the preceding trial was a threatening stimulus as compared with a neutral one (P < 0.05; Fig. 2B). This effect is comparable to the interference observed in emotional variants of the Stroop task, which show that naming a word, a color, or performing a lexical decision is slower when the underlying stimulus is emotional (Mathews 1990; Fox 1993; Algom et al. 2004). This increase in response latency provides an index of the extent to which the task-irrelevant emotional stimulus was processed (Okon-Singer et al. 2007). Accuracy showed that participants performed well in both emotion- (88.5 ± 4.7%) and color-naming tasks (77.3 ± 20.2%), although the color-naming task was more difficult (P < 0.05). Additional paired t-tests showed that the recognition rates between anger and fear did differ neither in the color task (T1,15 = −0.68, P = 0.503) nor in the emotion task (T1,15 = 1.8, P = 0.084), although this almost significant result may suggest that anger movies tend to be slightly better identified than fearful movies (86.2% vs. 81.1%). For RTs, there were no significant differences between fear and anger in both tasks (all P values > 0.22).

Brain Regions Responsive to Threat

To identify brain regions responsive to threat, we searched for all voxels that responded to the presentation of threatening versus neutral actions both in the emotion-naming and the color-naming tasks. Independent of task requirements, exposure to threatening body signals produced increased activity in the PAG and the posterior medial hypothalamus, subcortical structures that play a considerable role in autonomic reflexes and integrated defensive behaviors (Carrive 1993; Panksepp 1998; McNaughton and Corr 2004; Mobbs et al. 2007). Coordinates of the present PAG activation were similar to those reported previously for imminent danger (xyz: [3 −25 −7] in Mobbs et al. 2007; xyz: [2 −26 −4] here). Activity also increased in cortical regions including right lateral PM, bilateral anterior insula, pre-SMA, and left IFG BA44 (see Fig. 3, the full list of activations is provided in Table 1). The same network was revealed at P < 0.005 uncorrected when modeling correct trials only. We also tested whether activity in these regions was modulated by the task using ANOVA with beta values associated with each maximum peak. Only the right anterior insula (F1,15 = 15.881, P = 0.001) and the hypothalamus (F1,15 = 13.761, P = 0.002) showed a main effect of task, with higher activity during the emotion-naming task for the right anterior insula and higher activity during the color-naming task for the posterior hypothalamus. We could not detect any main effect of task or any task-by-threat interaction in IFG BA44 (all P values > 0.65), pre-SMA (P values > 0.18), left anterior insula (P values > 0.36), right PM (P values > 0.63), and PAG (P values > 0.59). Finally, to ensure that these responses were not led by fear or anger only, we estimated a new GLM where fear and anger responses were modeled separately and compared using post hoc t-tests fear and anger responses in the above regions of interest. We observed no difference between fear and anger. P values were, respectively for the emotion- and color-naming tasks, P = 0.21 and P = 0.22 in PAG, P = 0.58 and P = 0.57 in right hypothalamus, P = 0.77 and P = 0.66 in right pre-SMA, and P = 0.07 and P = 0.51 in left BA44. Only right PM response was stronger for anger as compared with fearful stimuli in both tasks (P < 0.001 and P = 0.015).

Within the above brain areas, we then searched for a neural signature of the behavioral response pattern. We performed correlation analyses between magnitude of activity within above threat-responsive regions independent of task demands and response time slowing observed during exposure to threat stimuli as compared with neutral ones (Fig. 3A,B). Trials where subjects made an error were not taken into account for the calculation of this RT index. Correlations were performed separately for each task using the contrast (threat vs. neutral) specific to emotion- or color-naming tasks. Mean contrast values were extracted for each subject using a 5-mm sphere centered onto the group mean coordinates of the above regions (pre-SMA, bilateral anterior insula, left BA44, PAG, hypothalamus, and right lateral PM, see Table 1). We observed significant positive Spearman rho correlations during the color-naming task in left BA44 (r = 0.494, P = 0.026), pre-SMA (r = 0.694, P = 0.001), left (r = 0.667, P = 0.002), and right anterior insula (r = 0.794, P < 0.001) and during the emotion-naming task in left anterior insula only (r = 0.579, P = 0.009), with a trend toward significance in pre-SMA (r = 0.374, P = 0.1). The same results were found when modeling correct trials only. In contrast, PAG, hypothalamus, and right lateral PM were not correlated with behavioral interference effects (see Supplementary Table S3 for full correlation results).

Brain Regions Responsive to Threat during the Emotion-Naming Task

As expected from the literature (Winston et al. 2003), amygdala activity increased when subjects perceived threatening stimuli (vs. neutral) and were requested to name the emotion (Fig. 4B). This effect was not observed in the color-naming task (Fig. 4C). Therefore, we identified voxels that showed a task-by-threat interaction (Fig. 4A), that is to say which responded more to threatening stimuli than neutral ones, in the emotion-naming task as compared with the color-naming task (inclusive masking procedure of the interaction with the simple effect emotion-naming threat vs. emotion-naming neutral and a threshold of P < 0.001 uncorrected). This interaction revealed increased bilateral anterior hippocampus response to threat during the emotion-naming task only (left: [−16 −14 −16], Z = 5.44, and right: [32 −8 −22], Z = 3.9). Careful examination using the anatomy toolbox (Eickhoff et al. 2005) established that a portion of each cluster extended to the posterior part of the amygdala. This illustrated the tendency toward interaction in the parameter estimates of the amygdala, that is, a difference between threatening and neutral stimuli only detected in the emotion-naming task (see Fig. 4 and Table 2 for full results). Moreover, we identified brain areas that showed additive effects of threat and emotion-naming task, such that responses to threat stimuli were elevated relative to neutral actions and were increased in the emotion-naming task relative to the color-naming task (see Table 3 for full results). We detected, among other areas, the superior temporal sulcus (STS), the fusiform gyrus, and the IFG. Again, we performed post hoc t-tests to ensure that responses to threat observed in amygdala and hippocampus during the emotion-naming task were not led either by fear or anger. No significant difference between fear and anger was observed in right hippocampus (P = 0.27 and P = 0.77) and left (P = 0.14 and P = 0.77) or right amygdala (P = 0.16 and P = 0.95).

Effectiveness of Task Manipulations

To further assess the validity of our experimental manipulation, we established that both tasks effectively influenced the activity of the specific brain areas, which play a key role in body and color perception, respectively (see Supplementary Figs S1 and S2). On the one hand, the contrast between emotion- versus color-naming tasks revealed increased activity in the fusiform body area (see Fig. 5A, and Supplementary Table S1 for full results), which is indeed known to selectively respond to perception of whole bodies, body parts, and schematic depictions of body shapes (Peelen and Downing 2005; Schwarzlose et al. 2005; Peelen et al. 2006). On the other hand, the contrast between situations where subjects attended to colors versus emotions also revealed significant responses in the fusiform gyrus but in a more posterior area than the peak observed in the emotion-naming task (see Fig. 5B, and Supplementary Table S2 for full results). This location corresponds to the location of the human visual area V4 (xyzTAL: [20 −66 −4] and [−26 −68 −8], corresponding to MNI coordinates [20 −68 −9] and [−26 −70 −14], from Zeki et al. 1991) important for color perception (Lueck et al. 1989; McKeefry and Zeki 1997; Gallant et al. 2000). Thus, both tasks met our experimental requirements as they induced task-specific modulations in regions involved in body or color perception.

Discussion

The present study explored the neural bases of the ability to carry out a task while still being able to react to threatening signals in the environment. We compared behavioral performance and brain activation in a color-naming and an emotion-naming condition. We identified 2 subcorticocortical networks. The first, which includes the PM, the hypothalamus, and the PAG, is impervious to attentional influence from task demands. Its reactivity to threatening stimuli is the same whether or not the subjects attend to the affective content of actions and this irrespective of the changes in amygdala's activity that were modulated by the task. We suggest that this network plays a role in reflexive defensive actions. In addition, threatening stimuli impact subject's behavioral responses in both tasks and provoke motor conflict in the first network's prefrontal areas. This is reflected by a correlation between the magnitude of activity within these regions and response time slowing between threatening and neutral signals. The second network, which includes the amygdala and regions of the temporal cortex such as the fusiform gyrus and the STS, is susceptible to influence from task demands and emotional factors as it primarily responds when subjects attend to actions and their affective content.

Threat-Related Subcorticocortical Network Independent of Tasks

The set of brain areas responsive to threatening expressions but unaffected by the type of attention (either the emotion- or color-naming task) consists of the subcortical structures PAG, hypothalamus, and thalamus and of a cortical premotor area at the border between ventral and dorsal PM (PMv/PMd). The observed subcortical areas are well known for their role in emotional reactivity (Bard 1928; Hess and Akert 1955; Brown et al. 1969; Siegel and Edinger 1983; Blanchard and Blanchard 1988; Carrive 1993; Panksepp 1998; Canteras 2002; Sewards and Sewards 2003; Adams 2006) and are part of the defensive fear system described in mammals and humans (Panksepp 1998; McNaughton and Corr 2004). In animals, electrical and chemical stimulation of the hypothalamus and PAG elicit vegetative responses as well as typical defensive reactions ranging from stereotypical flight to attack (Brown et al. 1969; Bandler and Shipley 1994; Sewards and Sewards 2003). Consistent with this, lesion of these structures severely reduces defensive behavior (Blanchard et al. 1981; Canteras et al. 1997). In humans, stimulation of the posterior medial hypothalamus, with a location corresponding to the present peak of activity, results in aggressive behavior (Bejjani et al. 2002). Moreover, 2 neuroimaging studies have shown that, as the distance between observer and source of threat decreases and threat becomes more imminent, activity in the PAG increases (Mobbs et al. 2007; Mobbs et al. 2009). Taken together, the recruitment of these subcortical areas during attended as well as unattended threat suggests that salient threat signals autonomously trigger neural structures involved in reflexive defensive responses (de Gelder et al. 2004; de Gelder 2006; Grèzes et al. 2007; Pichon et al. 2008, 2009; Bannerman et al. 2010).

Also of interest is the activity observed in cortical areas, notably at the PMv/PMd border. The PM is known to be implicated in external stimulus-driven actions (Passingham 1993; Hoshi and Tanji 2004) and motor preparation (Hoshi and Tanji 2004). Electrophysiological studies in monkeys report that the PZ of F4, at the PMv/PMd border, hosts neurons that respond to the perception of tactile and approaching visual stimuli, and electrical stimulation of this area triggers characteristic defensive movements (Cooke and Graziano 2004; Graziano and Cooke 2006). Interestingly, a recent fMRI study in humans has identified a polysensory region in the PM responding to both visual and tactile stimuli near the peak observed here (Nakashita et al. 2008). Our coordinates (zMNI = 52) correspond to the PMv/PMd border (range, z = 40–56, see Tomassini et al. 2007) and are similar to those reported in previous research using threatening faces or actions (z = 46 in Whalen et al. 2001; z = 40 in Grèzes et al. 2007; z = 52 in Pichon et al. 2008; z = 52 in Pichon et al. 2009). As suggested earlier, perceiving threat in others automatically triggers action preparation (de Gelder et al. 2004; de Gelder 2006; Grèzes et al. 2007; Pichon et al. 2008, 2009). We now demonstrate that this motor preparation mechanism is resistant to attentional control and remains functional when one is engaged in an unrelated task.

Conflict between Reflexive and Intentional Actions

Completing a challenging task requires one to disregard potentially distracting information including emotional signals (Pochon et al. 2002) as illustrated by the present color-naming task. Yet, when danger looms, the brain must remain able to detect it and disengage from the ongoing task in order to allocate resources to cope with the new situation (Corbetta et al. 2008). Automatically triggered defensive actions may thus potentially conflict with the subject's intentional response to the current task. If a threatening stimulus acts as a distracter and interferes with both emotion- and color-naming tasks, one expects this to be reflected in participants' performances. This is indeed the case. Response times were slower for threatening action stimuli as compared with neutral ones even during the color-naming task. This indicates that the emotional content of the stimuli is sufficiently threatening to divert resources toward their processing and therefore to provoke an interference with ongoing tasks (Lang et al. 1998; Panksepp 1998; Pessoa 2009).

To find the neural signature of interference between reflexive and task-related actions, we searched for threat-responsive regions in the brain network unaffected by attentional control, where activity correlated with the increase in response latency measured for threatening actions. As described before, this behavioral measure provides an index of the extent to which the task-irrelevant emotional stimulus is processed. Slower response time was correlated with the magnitude of activity in the pre-SMA, left IFG BA44, and bilateral anterior insula during the demanding color-naming task and with the magnitude of activity in left anterior insula only during the emotion-naming task.

The pre-SMA has been implicated in voluntary actions that are “self-initiated” or driven by “internal goals” and participates in complex cognitive control, such as alternation of motor plans, and motor selection and preparation (e.g., Passingham 1993; Picard and Strick 1996; Deiber et al. 1999; Lau et al. 2004; Nachev et al. 2008). Importantly, it also plays a critical role in situations of response conflict by exerting control over voluntary actions and suppression of competing prepotent actions (Isoda and Hikosaka 2007; Nachev et al. 2007; Chen et al. 2009). In monkeys, electrical microstimulation in the pre-SMA induces a switch between automatic incorrect responses to slower correct responses (Isoda and Hikosaka 2007). Importantly the present context, Lee et al. (2008) found more activity in the pre-SMA when participants produced facial expressions discordant with observed emotions (e.g., smiling in response to angry expressions). The pre-SMA may therefore play a role in overcoming the primary emotional motor response induced by perceived emotions.

The anterior part of the insula/frontal operculum was also shown to be implicated in interference resolution and task control (Badre and Wagner 2005; Wager et al. 2005; Nee et al. 2007; Higo et al. 2011), particularly when the conflicting information is emotion laden (Levens and Phelps 2010). Finally, according to Brass and von Cramon (2004), activity in the posterior inferior frontal sulcus (which corresponds to IFG BA44 here) allows one to selectively attend to specific information while ignoring irrelevant information (xyzMNI [−41 18 26] in Brass and von Cramon 2004; present coordinates [−44 12 32]). Similarly, Koechlin et al. (2003) and Koechlin and Hyafil (2007) propose that BA44 is involved in action selection as a function of the immediate context ([−44 8 20] in Koechlin et al. 2003; [−45 15 30] in Kouneiher et al. 2009), while Rushworth et al. (2005) suggest that it underlies the selection of appropriate stimulus–response association among alternative possibilities. In the present context, we suggest that the activations of pre-SMA, anterior insula/frontal operculum, and IFG BA44 may underlie interference resolution between automatic and task-instructed actions (Chen et al. 2009; Nelson et al. 2009).

Emotional Signals and the Amygdala

Activity in amygdala was triggered during visual presentation of threatening stimuli specifically during the emotion-naming task. However, the level of activity was lower when the subject's attention was not directed to the emotional content but to the color of visually presented dots. The anterior hippocampus, which sends contextual input to the amygdala (Phelps and LeDoux 2005), also displayed a similar pattern of interaction. Several experiments have shown that the amygdala can respond independently of selective attention (Vuilleumier et al. 2001; Anderson et al. 2003; Winston et al. 2003), while others have shown that task demands and degree of covert attention influence the extent to which emotionally salient stimuli such as fearful faces are processed (Pessoa et al. 2002; Bishop et al. 2007; Mitchell et al. 2007; Silvert et al. 2007; Lim et al. 2008). It has been suggested that the discrepancy between studies is related to the fact that high-load tasks as compared with low-load tasks (e.g., gender discrimination) compromise amygdala responses to unattended threat signals (Pessoa 2005). The present results are in line with the proposition by Lavie (2005) proposition that tasks involving high but not low perceptual load (Sinke et al. 2010) weaken the perceptual processing of potentially interfering distracters, here the emotional stimuli.

The finding that the subcorticocortical network described above is independent of amygdala activation raises the question how emotional information reaches that network. First, anatomical studies in macaque monkeys have shown that the PAG receives cortical projections from the STS and the PM (An et al. 1998). Monkey data also suggest that different sites of the prefrontal cortex, including ventral insula and orbital area 12o and caudal 12l (area 47/12 in humans, see Monchi et al. 2001), have direct connections with the PAG (An et al. 1998). These latter prefrontal areas also project to the hypothalamus (Ongur et al. 1998; Barbas et al. 2003). In addition, the STS is connected to the PM (Luppino et al. 2001). All these brain areas show an increased response during the perception of threatening signals and may therefore be critical for sustaining emotional processing independently of the amygdala.

Second, it is proposed that fear-induced defensive behavior relies on a hierarchically organized subcortical circuit consisting of the amygdala, hypothalamus, and PAG. Indeed, while defensive behavior triggered by amygdala stimulation requires the integrity of both the hypothalamus and the PAG, defensive behavior triggered by hypothalamus or PAG stimulation does not depend on the integrity of the amygdala (Fernandez de Molina and Hunsperger 1962). Amygdala lesions performed in adult monkeys indicate that this structure is important for the normal acquisition of a fear-potentiated startle reflex, but not for its retention and expression (Antoniadis et al. 2007). Also, PAG inactivation impairs acquisition of unconditioned freezing indicating that PAG participates in relaying aversive information to the amygdala (Johansen et al. 2010). The hierarchical organization of the subcortical defensive system could explain why monkeys with neonatal amygdala lesions display nonadaptive fearful behaviors during dyadic social interactions (Prather et al. 2001). This suggests that if the amygdala is not essential for the expression of defensive behavior, it may nevertheless serve to refine reflexive actions elaborated in hierarchically lower regions such as the PAG and the hypothalamus (Panksepp 1998) and relate these to learned behavioral contexts (Prather et al. 2001).

Conclusion

The present study shows that threat signals interfere with ongoing behavior and trigger responses independently of attention in a subcorticocortical network related to defensive behavior (PAG, hypothalamus, and PM) and in a network underlying interference resolution (pre-SMA, IFG BA44, and anterior insula/frontal operculum). In contrast to the amygdala, where activity can be modulated by task demands, response to threat in this network is independent from attentional control. We suggest that the identified brain network sustains motor vigilance and supports reflexive defensive behaviors that evolved to cope with threat. In this sense, the present results are consistent with the view that, at their core, emotions are essentially action tendencies that represent “efficient modes of adaptation to changing environmental demands” (Frijda 1986; Davidson 1993; Levenson 2003; Low et al. 2008).

Funding

Human Frontier Science program (HFSP-RGP0054/2004-C) and the European Union research funding NEST program (FP6-2005-NEST-Path Imp 043403). The Cognitive Neuroscience Laboratory is supported by the Bettencourt Schueller Foundation.

Supplementary Material

Supplementary material can be found at http://www.cercor.oxfordjournals.org/.

We are grateful to Michèle Chadwick and Alexi Hervais-Adelman for reading the manuscript, to Charlotte Sinke for subject recruiting and assistance in scanning, and to Sven Gijsen and France Maloumian for skillful technical assistance. Conflict of Interest: None of the authors have a financial interest related to this work.

References

Adams
DB
Brain mechanisms of aggressive behavior: an updated review
Neurosci Biobehav Rev
 , 
2006
, vol. 
30
 (pg. 
304
-
318
)
Algom
D
Chajut
E
Lev
S
A rational look at the emotional stroop phenomenon: a generic slowdown, not a stroop effect
J Exp Psychol Gen
 , 
2004
, vol. 
133
 (pg. 
323
-
338
)
An
X
Bandler
R
Ongür
D
Price
JL
Prefrontal cortical projections to longitudinal columns in the midbrain periaqueductal gray in macaque monkeys
J Comp Neurol
 , 
1998
, vol. 
401
 (pg. 
455
-
479
)
Anderson
AK
Christoff
K
Panitz
D
De Rosa
E
Gabrieli
JDE
Neural correlates of the automatic processing of threat facial signals
J Neurosci
 , 
2003
, vol. 
23
 (pg. 
5627
-
5633
)
Antoniadis
EA
Winslow
JT
Davis
M
Amaral
DG
Role of the primate amygdala in fear-potentiated startle: effects of chronic lesions in the rhesus monkey
J Neurosci
 , 
2007
, vol. 
27
 (pg. 
7386
-
7396
)
Badre
D
Wagner
AD
Frontal lobe mechanisms that resolve proactive interference
Cereb Cortex
 , 
2005
, vol. 
15
 (pg. 
2003
-
2012
)
Bandler
R
Shipley
MT
Columnar organization in the midbrain periaqueductal gray: modules for emotional expression?
Trends Neurosci
 , 
1994
, vol. 
17
 (pg. 
379
-
389
)
Bannerman
RL
Milders
M
Sahraie
A
Attentional cueing: fearful body postures capture attention with saccades
J Vis
 , 
2010
, vol. 
10
 pg. 
23
 
Barbas
H
Saha
S
Rempel-Clower
N
Ghashghaei
HT
Serial pathways from primate prefrontal cortex to autonomic areas may influence emotional expression
BMC Neurosci
 , 
2003
, vol. 
4
 pg. 
25
 
Bard
P
A diencephalic mechanism for the expression of rage with special reference to the sympathetic nervous system
Am J Physiol
 , 
1928
, vol. 
84
 (pg. 
490
-
515
)
Becker
MW
Panic search: fear produces efficient visual search for nonthreatening objects
Psychol Sci
 , 
2009
, vol. 
20
 (pg. 
435
-
437
)
Bejjani
BP
Houeto
JL
Hariz
M
Yelnik
J
Mesnage
V
Bonnet
AM
Pidoux
B
Dormont
D
Cornu
P
Agid
Y
Aggressive behavior induced by intraoperative stimulation in the triangle of Sano
Neurology
 , 
2002
, vol. 
59
 (pg. 
1425
-
1427
)
Bishop
SJ
Jenkins
R
Lawrence
AD
Neural processing of fearful faces: effects of anxiety are gated by perceptual capacity limitations
Cereb Cortex
 , 
2007
, vol. 
17
 (pg. 
1595
-
1603
)
Blanchard
DC
Blanchard
RJ
Ethoexperimental approaches to the biology of emotion
Annu Rev Psychol
 , 
1988
, vol. 
39
 (pg. 
43
-
68
)
Blanchard
DC
Williams
G
Lee
EMC
Blanchard
RJ
Taming of wild Rattus norvegicus by lesions of the mesencephalic central gray
Physiol Psychol
 , 
1981
, vol. 
9
 (pg. 
157
-
163
)
Brass
M
von Cramon
DY
Decomposing components of task preparation with functional magnetic resonance imaging
J Cogn Neurosci
 , 
2004
, vol. 
16
 (pg. 
609
-
620
)
Brown
JL
Hunsperger
RW
Rosvold
HE
Defence, attack, and flight elicited by electrical stimulation of the hypothalamus of the cat
Exp Brain Res
 , 
1969
, vol. 
8
 (pg. 
113
-
129
)
Canteras
NS
The medial hypothalamic defensive system: hodological organization and functional implications
Pharmacol Biochem Behav
 , 
2002
, vol. 
71
 (pg. 
481
-
491
)
Canteras
NS
Chiavegatto
S
Valle
LE
Swanson
LW
Severe reduction of rat defensive behavior to a predator by discrete hypothalamic chemical lesions
Brain Res Bull
 , 
1997
, vol. 
44
 (pg. 
297
-
305
)
Carrive
P
The periaqueductal gray and defensive behavior: functional representation and neuronal organization
Behav Brain Res
 , 
1993
, vol. 
58
 (pg. 
27
-
47
)
Chen
CY
Muggleton
NG
Tzeng
OJL
Hung
DL
Juan
CH
Control of prepotent responses by the superior medial frontal cortex
Neuroimage
 , 
2009
, vol. 
44
 (pg. 
537
-
545
)
Cooke
DF
Graziano
MS
Sensorimotor integration in the precentral gyrus: polysensory neurons and defensive movements
J Neurophysiol
 , 
2004
, vol. 
91
 (pg. 
1648
-
1660
)
Corbetta
M
Patel
G
Shulman
GL
The reorienting system of the human brain: from environment to theory of mind
Neuron
 , 
2008
, vol. 
58
 (pg. 
306
-
324
)
Davidson
RJ
Lewis
M
Haviland
J
The neuropsychology of emotion and affective style
Handbook of emotion
 , 
1993
New York
Guilford Press
(pg. 
143
-
154
)
Davis
M
Falls
WA
Campeau
S
Kim
M
Fear-potentiated startle: a neural and pharmacological analysis
Behav Brain Res
 , 
1993
, vol. 
58
 (pg. 
175
-
198
)
Davis
M
Whalen
PJ
The amygdala: vigilance and emotion
Mol Psychiatry
 , 
2001
, vol. 
6
 (pg. 
13
-
34
)
de Gelder
B
Towards the neurobiology of emotional body language
Nat Rev Neurosci
 , 
2006
, vol. 
7
 (pg. 
242
-
249
)
de Gelder
B
Snyder
J
Greve
D
Gerard
G
Hadjikhani
N
Fear fosters flight: a mechanism for fear contagion when perceiving emotion expressed by a whole body
Proc Natl Acad Sci U S A
 , 
2004
, vol. 
101
 (pg. 
16701
-
16706
)
Deiber
MP
Honda
M
Ibanez
V
Sadato
N
Hallett
M
Mesial motor areas in self-initiated versus externally triggered movements examined with fMRI: effect of movement type and rate
J Neurophysiol
 , 
1999
, vol. 
81
 (pg. 
3065
-
3077
)
Dimberg
U
Facial expressions as excitatory and inhibitory stimuli for conditioned autonomic responses
Biol Psychol
 , 
1986
, vol. 
22
 (pg. 
37
-
57
)
Dolan
RJ
Vuilleumier
P
Amygdala automaticity in emotional processing
Ann N Y Acad Sci
 , 
2003
, vol. 
985
 (pg. 
348
-
355
)
Duvernoy
HM
The human brain. Surface, three-dimensional sectional anatomy with MRI, and blood supply
 , 
1999
Vienna (Austria)
Springer-Verlag
Eickhoff
SB
Stephan
KE
Mohlberg
H
Grefkes
C
Fink
GR
Amunts
K
Zilles
K
A new SPM toolbox for combining probabilistic cytoarchitectonic maps and functional imaging data
Neuroimage
 , 
2005
, vol. 
25
 (pg. 
1325
-
1335
)
Fernandez de Molina
A
Hunsperger
RW
Organization of the subcortical system governing defence and flight reactions in the cat
J Physiol
 , 
1962
, vol. 
160
 (pg. 
200
-
213
)
Fox
E
Attentional bias in anxiety: selective or not?
Behav Res Ther
 , 
1993
, vol. 
31
 (pg. 
487
-
493
)
Fox
E
Damjanovic
L
The eyes are sufficient to produce a threat superiority effect
Emotion
 , 
2006
, vol. 
6
 (pg. 
534
-
539
)
Frijda
NH
The emotions
 , 
1986
Cambridge (UK)
Cambridge University Press
Friston
KJ
Holmes
AP
Worsley
KJ
Poline
JB
Frith
CD
Frackowiak
RSJ
Statistical parametric maps in functional imaging: a general linear approach
Hum Brain Mapp
 , 
1995
, vol. 
2
 (pg. 
189
-
210
)
Gallant
JL
Shoup
RE
Mazer
JA
A human extrastriate area functionally homologous to macaque V4
Neuron
 , 
2000
, vol. 
27
 (pg. 
227
-
235
)
Genovese
CR
Lazar
NA
Nichols
T
Thresholding of statistical maps in functional neuroimaging using the false discovery rate
Neuroimage
 , 
2002
, vol. 
15
 (pg. 
870
-
878
)
Graziano
MS
Cooke
DF
Parieto-frontal interactions, personal space, and defensive behavior
Neuropsychologia
 , 
2006
, vol. 
44
 (pg. 
845
-
859
)
Grèzes
J
Berthoz
S
Passingham
RE
Amygdala activation when one is the target of deceit: did he lie to you or to someone else?
Neuroimage
 , 
2006
, vol. 
30
 (pg. 
601
-
608
)
Grèzes
J
Pichon
S
de Gelder
B
Perceiving fear in dynamic body expressions
Neuroimage
 , 
2007
, vol. 
35
 (pg. 
959
-
967
)
Grosbras
MH
Paus
T
Brain networks involved in viewing angry hands or faces
Cereb Cortex
 , 
2006
, vol. 
16
 (pg. 
1087
-
1096
)
Hamm
AO
Cuthbert
BN
Globisch
J
Vaitl
D
Fear and the startle reflex: blink modulation and autonomic response patterns in animal and mutilation fearful subjects
Psychophysiology
 , 
1997
, vol. 
34
 (pg. 
97
-
107
)
Hansen
CH
Hansen
RD
Finding the face in the crowd: an anger superiority effect
J Pers Soc Psychol
 , 
1988
, vol. 
54
 (pg. 
917
-
924
)
Hess
WR
Akert
K
Experimental data on role of hypothalamus in mechanism of emotional behavior
AMA Arch Neurol Psychiatry
 , 
1955
, vol. 
73
 (pg. 
127
-
129
)
Higo
T
Mars
RB
Boorman
ED
Buch
ER
Rushworth
MFS
Distributed and causal influence of frontal operculum in task control
Proc Natl Acad Sci U S A
 , 
2011
, vol. 
108
 (pg. 
4230
-
4235
)
Hoffman
KL
Gothard
KM
Schmid
MC
Logothetis
NK
Facial-expression and gaze-selective responses in the monkey amygdala
Curr Biol
 , 
2007
, vol. 
17
 (pg. 
766
-
772
)
Hoshi
E
Tanji
J
Functional specialization in dorsal and ventral premotor areas
Prog Brain Res
 , 
2004
, vol. 
143
 (pg. 
507
-
511
)
Isoda
M
Hikosaka
O
Switching from automatic to controlled action by monkey medial frontal cortex
Nat Neurosci
 , 
2007
, vol. 
10
 (pg. 
240
-
248
)
Johansen
JP
Tarpley
JW
LeDoux
JE
Blair
HT
Neural substrates for expectation-modulated fear learning in the amygdala and periaqueductal gray
Nat Neurosci
 , 
2010
, vol. 
13
 (pg. 
979
-
986
)
Kober
H
Barrett
LF
Joseph
J
Bliss-Moreau
E
Lindquist
K
Wager
TD
Functional grouping and cortical-subcortical interactions in emotion: a meta-analysis of neuroimaging studies
Neuroimage
 , 
2008
, vol. 
42
 (pg. 
998
-
1031
)
Koechlin
E
Hyafil
A
Anterior prefrontal function and the limits of human decision-making
Science
 , 
2007
, vol. 
318
 (pg. 
594
-
598
)
Koechlin
E
Ody
C
Kouneiher
F
The architecture of cognitive control in the human prefrontal cortex
Science
 , 
2003
, vol. 
302
 (pg. 
1181
-
1185
)
Kouneiher
F
Charron
S
Koechlin
E
Motivation and cognitive control in the human prefrontal cortex
Nat Neurosci
 , 
2009
, vol. 
12
 (pg. 
939
-
945
)
Lang
PJ
Bradley
MM
Cuthbert
BN
Emotion, motivation, and anxiety: brain mechanisms and psychophysiology
Biol Psychiatry
 , 
1998
, vol. 
44
 (pg. 
1248
-
1263
)
Lau
HC
Rogers
RD
Haggard
P
Passingham
RE
Attention to intention
Science (New York, N.Y.)
 , 
2004
, vol. 
303
 (pg. 
1208
-
1210
)
Lavie
N
Distracted and confused? Selective attention under load
Trends Cogn Sci
 , 
2005
, vol. 
9
 (pg. 
75
-
82
)
LeDoux
JE
Emotion: clues from the brain
Annu Rev Psychol
 , 
1995
, vol. 
46
 (pg. 
209
-
235
)
Lee
TW
Dolan
RJ
Critchley
HD
Controlling emotional expression: behavioral and neural correlates of nonimitative emotional responses
Cereb Cortex
 , 
2008
, vol. 
18
 (pg. 
104
-
113
)
Levens
SM
Phelps
EA
Insula and orbital frontal cortex activity underlying emotion interference resolution in working memory
J Cogn Neurosci
 , 
2010
, vol. 
22
 (pg. 
2790
-
2803
)
Levenson
RW
Blood, sweat, and fears: the autonomic architecture of emotion
Ann N Y Acad Sci
 , 
2003
, vol. 
1000
 (pg. 
348
-
366
)
Lim
SL
Padmala
S
Pessoa
L
Affective learning modulates spatial competition during low-load attentional conditions
Neuropsychologia
 , 
2008
, vol. 
46
 (pg. 
1267
-
1278
)
Low
A
Lang
PJ
Smith
JC
Bradley
MM
Both predator and prey: emotional arousal in threat and reward
Psychol Sci
 , 
2008
, vol. 
19
 (pg. 
865
-
873
)
Lueck
CJ
Zeki
S
Friston
KJ
Deiber
MP
Cope
P
Cunningham
VJ
Lammertsma
AA
Kennard
C
Frackowiak
RS
The colour centre in the cerebral cortex of man
Nature
 , 
1989
, vol. 
340
 (pg. 
386
-
389
)
Luppino
G
Calzavara
R
Rozzi
S
Matelli
M
Projections from the superior temporal sulcus to the agranular frontal cortex in the macaque
Eur J Neurosci
 , 
2001
, vol. 
14
 (pg. 
1035
-
1040
)
Mathews
A
Why worry? The cognitive function of anxiety
Behav Res Ther
 , 
1990
, vol. 
28
 (pg. 
455
-
468
)
McKeefry
DJ
Zeki
S
The position and topography of the human colour centre as revealed by functional magnetic resonance imaging
Brain
 , 
1997
, vol. 
120
 (pg. 
2229
-
2242
)
McNaughton
N
Corr
PJ
A two-dimensional neuropsychology of defense: fear/anxiety and defensive distance
Neurosci Biobehav Rev
 , 
2004
, vol. 
28
 (pg. 
285
-
305
)
Mitchell
DGV
Nakic
M
Fridberg
D
Kamel
N
Pine
DS
Blair
RJR
The impact of processing load on emotion
Neuroimage
 , 
2007
, vol. 
34
 (pg. 
1299
-
1309
)
Mobbs
D
Marchant
JL
Hassabis
D
Seymour
B
Tan
G
Gray
M
Petrovic
P
Dolan
RJ
Frith
CD
From threat to fear: the neural organization of defensive fear systems in humans
J Neurosci
 , 
2009
, vol. 
29
 (pg. 
12236
-
12243
)
Mobbs
D
Petrovic
P
Marchant
JL
Hassabis
D
Weiskopf
N
Seymour
B
Dolan
RJ
Frith
CD
When fear is near: threat imminence elicits prefrontal-periaqueductal gray shifts in humans
Science (New York, N.Y.)
 , 
2007
, vol. 
317
 (pg. 
1079
-
1083
)
Monchi
O
Petrides
M
Petre
V
Worsley
K
Dagher
A
Wisconsin Card Sorting revisited: distinct neural circuits participating in different stages of the task identified by event-related functional magnetic resonance imaging
J Neurosci
 , 
2001
, vol. 
21
 (pg. 
7733
-
7741
)
Morris
JS
Frith
CD
Perrett
DI
Rowland
D
Young
AW
Calder
AJ
Dolan
RJ
A differential neural response in the human amygdala to fearful and happy facial expressions
Nature
 , 
1996
, vol. 
383
 (pg. 
812
-
815
)
Nachev
P
Kennard
C
Husain
M
Functional role of the supplementary and pre-supplementary motor areas
Nat Rev Neurosci
 , 
2008
, vol. 
9
 (pg. 
856
-
869
)
Nachev
P
Wydell
H
O'neill
K
Husain
M
Kennard
C
The role of the pre-supplementary motor area in the control of action
Neuroimage
 , 
2007
, vol. 
36
 
Suppl. 2
(pg. 
T155
-
T163
)
Nakashita
S
Saito
DN
Kochiyama
T
Honda
M
Tanabe
HC
Sadato
N
Tactile-visual integration in the posterior parietal cortex: a functional magnetic resonance imaging study
Brain Res Bull
 , 
2008
, vol. 
75
 (pg. 
513
-
525
)
Nee
DE
Wager
TD
Jonides
J
Interference resolution: insights from a meta-analysis of neuroimaging tasks
Cogn Affect Behav Neurosci
 , 
2007
, vol. 
7
 (pg. 
1
-
17
)
Nelson
JK
Reuter-Lorenz
PA
Persson
J
Sylvester
CY
Jonides
J
Mapping interference resolution across task domains: a shared control process in left inferior frontal gyrus
Brain Res
 , 
2009
, vol. 
1256
 (pg. 
92
-
100
)
Ohman
A
Automaticity and the amygdala: nonconscious responses to emotional faces
Curr Dir Psychol Sci
 , 
2002
, vol. 
11
 (pg. 
62
-
66
)
Ohman
A
Esteves
F
Soares
JJF
Preparedness and preattentive associative learning: electrodermal conditioning to masked stimuli
J Psychophysiol
 , 
1995
, vol. 
9
 (pg. 
99
-
108
)
Ohman
A
Lundqvist
D
Esteves
F
The face in the crowd revisited: a threat advantage with schematic stimuli
J Pers Soc Psychol
 , 
2001
, vol. 
80
 (pg. 
381
-
396
)
Okon-Singer
H
Tzelgov
J
Henik
A
Distinguishing between automaticity and attention in the processing of emotionally significant stimuli
Emotion
 , 
2007
, vol. 
7
 (pg. 
147
-
157
)
Ongur
D
An
X
Price
JL
Prefrontal cortical projections to the hypothalamus in macaque monkeys
J Comp Neurol
 , 
1998
, vol. 
401
 (pg. 
480
-
505
)
Panksepp
J
Affective neuroscience: the foundation of human and animal emotions
 , 
1998
New York
Oxford University Press
Passingham
RE
The frontal lobes and voluntary action
 , 
1993
Oxford
Oxford University Press
Peelen
MV
Downing
PE
Selectivity for the human body in the fusiform gyrus
J Neurophysiol
 , 
2005
, vol. 
93
 (pg. 
603
-
608
)
Peelen
MV
Wiggett
AJ
Downing
PE
Patterns of fMRI activity dissociate overlapping functional brain areas that respond to biological motion
Neuron
 , 
2006
, vol. 
49
 (pg. 
815
-
822
)
Pessoa
L
To what extent are emotional visual stimuli processed without attention and awareness?
Curr Opin Neurobiol
 , 
2005
, vol. 
15
 (pg. 
188
-
196
)
Pessoa
L
On the relationship between emotion and cognition
Nat Rev Neurosci
 , 
2008
, vol. 
9
 (pg. 
148
-
158
)
Pessoa
L
How do emotion and motivation direct executive control?
Trends Cogn Sci
 , 
2009
, vol. 
13
 (pg. 
160
-
166
)
Pessoa
L
McKenna
M
Gutierrez
E
Ungerleider
LG
Neural processing of emotional faces requires attention
Proc.Natl.Acad.Sci.U.S.A
 , 
2002
, vol. 
99
 (pg. 
11458
-
11463
)
Phelps
EA
LeDoux
JE
Contributions of the amygdala to emotion processing: from animal models to human behavior
Neuron
 , 
2005
, vol. 
48
 (pg. 
175
-
187
)
Picard
N
Strick
PL
Motor areas of the medial wall: a review of their location and functional activation
Cereb Cortex
 , 
1996
, vol. 
6
 (pg. 
342
-
353
)
Pichon
S
de Gelder
B
Grèzes
J
Emotional modulation of visual and motor areas by dynamic body expressions of anger
Soc Neurosci
 , 
2008
, vol. 
3
 (pg. 
199
-
212
)
Pichon
S
de Gelder
B
Grèzes
J
Two different faces of threat. Comparing the neural systems for recognizing fear and anger in dynamic body expressions
Neuroimage
 , 
2009
, vol. 
47
 (pg. 
1873
-
1883
)
Pochon
JB
Levy
R
Fossati
P
Lehericy
S
Poline
JB
Pillon
B
Le Bihan
D
Dubois
B
The neural system that bridges reward and cognition in humans: an fMRI study
Proc Natl Acad Sci U S A
 , 
2002
, vol. 
99
 (pg. 
5669
-
5674
)
Prather
MD
Lavenex
P
Mauldin-Jourdain
ML
Mason
WA
Capitanio
JP
Mendoza
SP
Amaral
DG
Increased social fear and decreased fear of objects in monkeys with neonatal amygdala lesions
Neuroscience
 , 
2001
, vol. 
106
 (pg. 
653
-
658
)
Rushworth
MFS
Buckley
MJ
Gough
PM
Alexander
IH
Kyriazis
D
McDonald
KR
Passingham
RE
Attentional selection and action selection in the ventral and orbital prefrontal cortex
J Neurosci
 , 
2005
, vol. 
25
 (pg. 
11628
-
11636
)
Schupp
HT
Ohman
A
Junghofer
M
Weike
AI
Stockburger
J
Hamm
AO
The facilitated processing of threatening faces: an ERP analysis
Emotion
 , 
2004
, vol. 
4
 (pg. 
189
-
200
)
Schwarzlose
RF
Baker
CI
Kanwisher
N
Separate face and body selectivity on the fusiform gyrus
J Neurosci
 , 
2005
, vol. 
25
 (pg. 
11055
-
11059
)
Sewards
TV
Sewards
MA
Representations of motivational drives in mesial cortex, medial thalamus, hypothalamus and midbrain
Brain Res Bull
 , 
2003
, vol. 
61
 (pg. 
25
-
49
)
Siegel
A
Edinger
HM
Role of the limbic system in hypothalamically elicited attack behavior
Neurosci Biobehav Rev
 , 
1983
, vol. 
7
 (pg. 
395
-
407
)
Silvert
L
Lepsien
J
Fragopanagos
N
Goolsby
B
Kiss
M
Taylor
JG
Raymond
JE
Shapiro
KL
Eimer
M
Nobre
AC
Influence of attentional demands on the processing of emotional facial expressions in the amygdala
Neuroimage
 , 
2007
, vol. 
38
 (pg. 
357
-
366
)
Sinke
CBA
Sorger
B
Goebel
R
de Gelder
B
Tease or threat? Judging social interactions from bodily expressions
Neuroimage
 , 
2010
, vol. 
49
 (pg. 
1717
-
1727
)
Strauss
MM
Makris
N
Aharon
I
Vangel
MG
Goodman
J
Kennedy
DN
Gasic
GP
Breiter
HC
fMRI of sensitization to angry faces
Neuroimage
 , 
2005
, vol. 
26
 (pg. 
389
-
413
)
Tamietto
M
Geminiani
G
Genero
R
de Gelder
B
Seeing fearful body language overcomes attentional deficits in patients with neglect
J Cogn Neurosci
 , 
2007
, vol. 
19
 (pg. 
445
-
454
)
Tomassini
V
Jbabdi
S
Klein
JC
Behrens
TEJ
Pozzilli
C
Matthews
PM
Rushworth
MFS
Johansen-Berg
H
Diffusion-weighted imaging tractography-based parcellation of the human lateral premotor cortex identifies dorsal and ventral subregions with anatomical and functional specializations
J Neurosci
 , 
2007
, vol. 
27
 (pg. 
10259
-
10269
)
Van Essen
DC
A population-average, landmark-and surface-based (PALS) atlas of human cerebral cortex
Neuroimage
 , 
2005
, vol. 
28
 (pg. 
635
-
662
)
Van Essen
DC
Drury
HA
Dickson
J
Harwell
J
Hanlon
D
Anderson
CH
An integrated software suite for surface-based analyses of cerebral cortex
J Am Med Inform Assoc
 , 
2001
, vol. 
8
 pg. 
443
 
Vuilleumier
P
Armony
JL
Driver
J
Dolan
RJ
Effects of attention and emotion on face processing in the human brain: an event-related fMRI study
Neuron
 , 
2001
, vol. 
30
 (pg. 
829
-
841
)
Vuilleumier
P
Richardson
MP
Armony
JL
Driver
J
Dolan
RJ
Distant influences of amygdala lesion on visual cortical activation during emotional face processing
Nat Neurosci
 , 
2004
, vol. 
7
 (pg. 
1271
-
1278
)
Wager
TD
Sylvester
CY
Lacey
SC
Nee
DE
Franklin
M
Jonides
J
Common and unique components of response inhibition revealed by fMRI
Neuroimage
 , 
2005
, vol. 
27
 (pg. 
323
-
340
)
Whalen
PJ
Fear, vigilance, and ambiguity: initial neuroimaging studies of the human amygdala
Curr Dir Psychol Sci
 , 
1998
, vol. 
7
 (pg. 
177
-
188
)
Whalen
PJ
Shin
LM
McInerney
SC
Fischer
H
Wright
CI
Rauch
SL
A functional MRI study of human amygdala responses to facial expressions of fear versus anger
Emotion
 , 
2001
, vol. 
1
 (pg. 
70
-
83
)
Williams
LM
Das
P
Liddell
B
Olivieri
G
Peduto
A
Brammer
MJ
Gordon
E
BOLD, sweat and fears: fMRI and skin conductance distinguish facial fear signals
Neuroreport
 , 
2005
, vol. 
16
 (pg. 
49
-
52
)
Winston
JS
O'Doherty
J
Dolan
RJ
Common and distinct neural responses during direct and incidental processing of multiple facial emotions
Neuroimage
 , 
2003
, vol. 
20
 (pg. 
84
-
97
)
Zeelenberg
R
Bocanegra
BR
Auditory emotional cues enhance visual perception
Cognition
 , 
2010
, vol. 
115
 (pg. 
202
-
206
)
Zeki
S
Watson
JD
Lueck
CJ
Friston
KJ
Kennard
C
Frackowiak
RS
A direct demonstration of functional specialization in human visual cortex
J Neurosci
 , 
1991
, vol. 
11
 (pg. 
641
-
649
)