Abstract
Humans are sensitive to others' gaze direction. Evidence from a behavior study using continuous flash suppression (CFS) suggests that faces with direct and averted gaze are processed differently before they reach conscious awareness (Stein et al. 2011). An ERP study showed that invisible faces with direct gaze elicited significantly larger negative deflections at 200, 250, and 300 ms over the parietofrontal electrodes (Yokoyama et al. 2013). These results suggest that faces with direct gaze are preferentially processed in the brain unconsciously. In the current fMRI study we investigated whether brain regions involved in face and gaze processing were sensitive to gaze information rendered invisible through CFS. With a slow event-related design, four types of stimuli were presented to observers: visible and invisible faces with direct and averted gaze. The four conditions were shown in random order in each scan session, and observers were instructed to press a button to indicate when they saw a face. To minimize the physical difference between the two visibility conditions and to mimic the perception when faces leak from CFS in the invisible conditions, the same kind of noises were blended in the same eye with faces in the visible trials. Data from observers who saw any face during invisible trials were excluded from analysis. Results show that although signals from invisible faces were significantly reduced relative to visible faces, a number of brain regions, including the OFA, FFA, and the amygdala, responded differentially to invisible faces with direct vs. averted gazes. STS also responded moderately, but more variably, to invisible faces. There was a tendency for the invisible faces with the direct gaze to elicit stronger responses than invisible faces with an averted gaze; a tendency not observed for visible faces. Our results provide the neural correlates for the unconscious processing of gaze information.
Meeting abstract presented at VSS 2014