Two critical and functionally distinct stages of face and body perception

Abstract

Cortical regions that respond preferentially to particular object categories, such as faces and bodies, are essential for visual perception of these object categories. However, precisely when these regions play a causal role in recognition of their preferred categories is unclear. Here we addressed this question using transcranial magnetic stimulation (TMS). Across a series of experiments, TMS was delivered over the functionally localized right occipital face area (rOFA) or right extrastriate body area (rEBA) at different latencies, up to 150 ms, after stimulus onset while adult human participants performed delayed match-to-sample tasks on face and body stimuli. Results showed that TMS disrupted task performance during two temporally distinct time periods after stimulus onset, the first at 40/50 ms and the second at 100/110 ms. These two time periods exhibited functionally distinct patterns of impairment: TMS delivered during the early time period (at 40/50 ms) disrupted task performance for both preferred (faces at rOFA and bodies at rEBA) and nonpreferred (bodies at rOFA and faces at rEBA) categories. In contrast, TMS delivered during the later time period (at 100/110 ms) disrupted task performance for the preferred category only of each area (faces at rOFA and bodies at rEBA). These results indicate that category-selective cortical regions are critical for two functionally distinct stages of visual object recognition: an early, presumably preparatory stage that is not category selective occurring almost immediately after stimulus onset, followed by a later stage of category-specific perceptual processing.

Figure 1.

Locations in one participant of the rOFA in red (faces − objects) and the rEBA in blue (bodies − objects).

Figure 2.

Timeline of the experimental trial procedure from experiments 1 and 6. In experiments 1–5, participants judged whether the probe stimulus was the same or different from the prime stimulus. In experiment 6, participants judged the gaze direction of the stimulus.

Figure 3.

Results from experiments 1–3 (error bars denote SEs). An asterisk denotes a significant difference in Bonferroni's corrected tests. A, dTMS delivered over rOFA disrupted face recognition at 40/50 and 100/110 ms. B, dTMS delivered over rEBA disrupted body recognition at 40/50 and 100/110 ms. C, dTMS delivered over the rOFA and rEBA during the early window disrupted face and body recognition. dTMS delivered during the late window disrupted recognition of faces over the rOFA and bodies over the rEBA only.

Figure 4.

Results from experiments 4–6 (error bars denote SEs). An asterisk denotes a significant difference in Bonferroni's corrected tests. A, dTMS delivered over the rOFA disrupted face recognition at 40/50, 60/70, and 100/110 ms when we added temporal jitter. B, dTMS delivered over the rOFA disrupted body recognition at 40/50 and 60/70 ms when we added temporal jitter. C, dTMS delivered over the rOFA disrupted gaze discrimination at 40/50 and 100/110 ms.