Facial expression recognition: A meta-analytic review of theoretical models and neuroimaging evidence

Discrimination of facial expressions is an elementary function of the human brain. While the way emotions are represented in the brain has long been debated, common and specific neural representations in recognition of facial expressions are also complicated. To examine brain organizations and asymmetry on discrete and dimensional facial emotions, we conducted an activation likelihood estimation meta-analysis and meta-analytic connectivity modelling on 141 studies with a total of 3138 participants. We found consistent engagement of the amygdala and a common set of brain networks across discrete and dimensional emotions. The left-hemisphere dominance of the amygdala and AI across categories of facial expression, but category-specific lateralization of the vmPFC, suggesting a flexibly asymmetrical neural representations of facial expression recognition. These results converge to characteristic activation and connectivity patterns across discrete and dimensional emotion categories in recognition of facial expressions. Our findings provide the first quantitatively meta-analytic brain network-based evidence supportive of the psychological constructionist hypothesis in facial expression recognition.