Multiple mechanisms of consciousness: the neural correlates of emotional awareness

Abstract

Emotional stimuli, including facial expressions, are thought to gain rapid and privileged access to processing resources in the brain. Despite this access, we are conscious of only a fraction of the myriad of emotion-related cues we face everyday. It remains unclear, therefore, what the relationship is between activity in neural regions associated with emotional representation and the phenomenological experience of emotional awareness. We used functional magnetic resonance imaging and binocular rivalry to delineate the neural correlates of awareness of conflicting emotional expressions in humans. Behaviorally, fearful faces were significantly more likely to be perceived than disgusted or neutral faces. Functionally, increased activity was observed in regions associated with facial expression processing, including the amygdala and fusiform gyrus during emotional awareness. In contrast, awareness of neutral faces and suppression of fearful faces were associated with increased activity in dorsolateral prefrontal and inferior parietal cortices. The amygdala showed increased functional connectivity with ventral visual system regions during fear awareness and increased connectivity with perigenual prefrontal cortex (pgPFC; Brodmann's area 32/10) when fear was suppressed. Despite being prioritized for awareness, emotional items were associated with reduced activity in areas considered critical for consciousness. Contributions to consciousness from bottom-up and top-down neural regions may be additive, such that increased activity in specialized regions within the extended ventral visual system may reduce demands on a frontoparietal system important for awareness. The possibility is raised that interactions between pgPFC and the amygdala, previously implicated in extinction, may also influence whether or not an emotional stimulus is accessible to consciousness.

Figure 1.

The binocular rivalry task. An illustration of the experimental scenario with a sample stimulus in composite form, involving superimposed red and blue facial expressions and deconstructed into its constituent parts. Participants wore red/blue three-dimensional glasses to filter retinal input. In this example, the left eye receives input from the blue fearful facial expression, while at the same time a red disgusted facial expression is input to the right eye. Note that input from each eye goes to both hemispheres of the brain. This manipulation often allows one of the expressions to reach awareness while the other is suppressed from consciousness.

Figure 2.

Fear is prioritized for awareness. Behavioral results demonstrated that participants were significantly more likely to be conscious of fearful facial expressions than both disgusted (t₍₁₅₎ = 2.26, p = 0.039) and neutral (t₍₁₅₎ = 2.83, p = 0.013) expressions when paired with a competing facial expression. Error bars depict the SEM between subjects ( = p < 0.05).

Figure 3.

Activity in both bottom-up and top-down neural regions modulated as a function of emotional awareness. a–c, The contrast of fear awareness (perceived fear minus unperceived fear) revealed greater activity in the right amygdala (p < 0.01, a) when fear was perceived and greater activation of the left perigenual prefrontal cortex (b) when fear was present but unperceived. The contrast of disgust awareness (perceived disgust minus unperceived, but present, disgust) showed greater activation in left fusiform gyrus (c) when disgust was perceived (p < 0.01). d, e, The main effect of perceived emotion revealed greater activity during the awareness of fearful and disgusted faces versus neutral ones in left inferior occipital lobe (d) and the left fusiform gyrus (e). The y-axis depicts percentage signal change (% sig. change) in BOLD response; error bars depict SEM between subjects (*p < 0.001, ^¥p < 0.01, p < 0.05).

Figure 4.

Perigenual prefrontal cortex activity associated with awareness of neutral relative to emotional stimuli. a, b, Images depicting the main effect of perceived emotion revealed greater activation in the right (a) and left (b) perigenual prefrontal cortex when a neutral expression was perceived relative to when either a fearful or disgusted expression was perceived (*p < 0.001, ^¥p < 0.01). The y-axis depicts percentage signal change (% sig. change) in BOLD response; error bars depict the SEM between subjects.

Figure 5.

Enhanced activity in a frontoparietal network associated with awareness of neutral relative to emotional stimuli. a–d, Images depicting the main effect of perceived emotion revealed greater activation in the right (a) and left (b) middle frontal gyrus and the left (c) and right (d) inferior parietal cortices when a neutral relative to fearful or disgusted expression was perceived (*p < 0.001, p < 0.05). e, f, The results of the contrast of emotion suppression [FN_N + DN_N] > neutral only condition (NN) revealed greater activation in the right (e) middle frontal gyrus and the left (f) middle frontal gyrus during emotion suppression. The y-axis depicts percentage signal change in BOLD response; error bars depict the SEM between subjects.

Figure 6.

Connectivity with the amygdala varies as a function of awareness. a–c, Context-dependent connectivity analysis revealing areas with a significant modulation of functional connectivity with the right amygdala in the presence of perceived fearful trials relative to unperceived fearful trials (p < 0.01). Areas in orange and yellow show enhanced functional connectivity during perceived fearful trials relative to unperceived fearful trials: right superior temporal cortex (a) and right middle temporal cortex (b). Regions in blue show reduced functional connectivity with the right amygdala during perceived fearful trials relative to unperceived fearful trials: left perigenual prefrontal cortex (c).