Abstract
Background: Evoked potential studies of familiarity effects on face processing have lead to divergent conclusions, with some suggesting that they emerge as early as 140-170ms, and others later, around 250-400ms. However, those reports rely on group averages of mean or peak amplitude within a priori time windows. Objective: Our goal was to determine the temporal dynamics of the familiarity effect at the level of the individual subject, using a moving window analysis that did not make assumptions regarding specific time windows. Methods: 10 observers judged the pleasantness of faces presented in a screen. A trial started with a fixation point (2700-2900ms) followed by an anonymous or a familiar (internet celebrity pictures) face (100ms) and a mask (300ms). Results contrasting anonymous versus familiar faces were analyzed using a point-by-point bootstrap analysis of single subjects (Figure 1). Results: Significant differences between familiar and unfamiliar faces were most consistently found between: 80-120ms in O1, O2, Oz; 120-140ms in Fz, Cz, Pz; 140-160ms in P8, 160-180ms in P7; 300-360ms in Fz, Cz, Pz (Figure 2). As the earliest differences were observed in occipital electrodes, these may reflect differences in stimulus properties between the anonymous and famous images. However, differences at 120-180ms appeared at times and in locations where face-selective responses occur (the N170 in P8/7 and VPP in Pz/Cz) and not in occipital electrodes, suggesting that they reflected modulation of face-processing by familiarity. Less consistent differences observed after 300ms in central electrodes coincide with the P300 component, which has been associated to processing novel stimuli. Conclusion: This approach suggests that familiarity effects on face processing emerge around 120ms and evolve dynamically across the scalp, probably reflecting the activation of a widespread face perception network. Central derivations manifest such influence first, followed by temporo-occipital locations in the right and finally in the left hemisphere.
Meeting abstract presented at VSS 2013