Emotion modulates visual processing May 10, 2007Posted by Johan in Emotion, Face Perception, Neuroscience, Social Neuroscience.
Affective neuroscientists generally argue that emotion is present in every aspect of human cognition: memory, attention, executive functioning and perception are all assumed to be mediated by emotional states. Antonio Damasio famously went beyond this and argued that emotional states play a crucial causal role in executive functioning. According to his somatic marker hypothesis, decision-making in the individual situation is guided by internal simulations of each possible course of action. The simulation of each possible action produces an emotional reaction, and deciding what to do is then a simple matter of choosing what feels best.
While Damasio’s work on executive functioning has perhaps been heavy on theory and light on direct evidence, researchers looking at the role of affect in perception have made some headway with a more inductive approach. Vuilleumier et al (2004) obtained direct evidence that the amygdala (the brain’s fear centre) can modulate activity in the fusiform gyrus (aka the fusiform face area) and visual cortex.
Vuilleumier et al’s (2004) experimental design deserves some attention. Research in cognitive neuroscience is guided mainly by lesion studies, where the behavioural deficits of brain-damaged patients is tied to the location of their injury, and functional neuroimaging studies, where brain activation in healthy participants is measured. Vuilleumier et al (2004) managed to incorporate both these elements into a single study by comparing the functional neuroimaging activation patterns in patients with brain damage to that of healthy controls.
The patients in the study suffered from medial temporal lobe sclerosis, a hardening of the cerebral tissue often caused by epilepsy. Patients who had damage to the hippocampus (which plays a crucial role in consolidating new memories) and the amygdala were separated from patients who had damage to the hippocampus alone, thus producing three groups: patients with hippocampal-amygdala damage, patients with hippocampal damage, and healthy controls.
The three groups of participants viewed fearful and neutral facial expressions, which were either relevant or irrelevant to the task at hand. The figure below outlines the behavioural task.
Participants were asked to say whether pairs of houses or faces were the same or different. In one block, they had to focus on the vertical pair while ignoring the horizontal pair, and vice versa. Thus, in each trial the participants would see two pictures of houses and two pictures of faces, which could be neutral or fearful. Because the participants were instructed to attend to either the horizontal or the vertical pair, one pair of houses or faces in each trial were task-irrelevant. The figure below shows the fearful > neutral face subtraction, meaning that the areas that are lit up here were activated by fearful faces beyond the activation produced by neutral faces.
Both the hippocampus-only patients and the healthy controls showed increased activation in visual cortex and the fusiform gyrus to the fearful faces as compared to the neutral faces (see D and E in the figure, respectively), but this difference did not appear in the amygdala-hippocampus patients (F in the figure).
To further make the point that the amygdala caused the enhanced processing, Vuilleumier et al (2004) also looked at which of the amygdalae that had been affected by sclerosis (no participants had bilateral damage), to obtain a within-groups measure. In the patients with amygdala and hippocampal lesions, left amygdala damage abolished the usual increased activation in the left fusiform gyrus, and right amygdala damage abolished increased activation in the right fusiform gyrus. This implies that while each amygdala may project diffusely to much of the cortex, it may be limited to its own hemisphere.
In addition to the areas involved in visual processing, amygdala damage also reduced the typical fearful-neutral differential activation in a range of other areas, including the left medial hypothalamus, the retrosplenial cortex, the hippocampal region, the superior temporal sulcus, and the rostral cingulate cortex. Once again, this emphasises the point that the amygdala is capable of modulating processing in large parts of the brain.
While Vuilleumier et al’s (2004) design may seem convoluted at first, with its combined functional imaging and lesion study methods, the design allows the researchers to conclude with some confidence that the increased activation in visual cortex and the fusiform gyrus in response to fearful faces is caused by the amygdala (or the hippocampus, the critic might argue). It is quite striking that the tiny amygdala – it is fittingly named after the greek term for almond – can produce measurable effects on regions as distant as visual cortex.
So in closing, the amygdala appears to enhance sensory processing of emotional stimuli. This has long been a tenet of affective neuroscientists, following case studies of amygdala-lesioned patients who were selectively impaired at detecting fearful impressions, but the Vuilleumier et al (2004) study is the first I know of to provide direct evidence that the amygdala plays this role in perception.
Vuilleumier, P., Richardson, M.P., Armony, J.L., Driver, J., & Dolan, R.J. (2004). Distant influences of amygdala lesion on visual cortical activation during emotional face processing. Nature Neuroscience, 7, 1271-1278.