Imaging Object-Scene Integration in Visible and Invisible Natural Scenes

Abstract

Integrating objects with their context is a key step in the interpretation of complex visual scenes. Humans can do this very quickly, yet the brain mechanisms that mediate this ability are not yet understood. Here, we used functional Magnetic Resonance Imaging (fMRI) to measure brain activity while participants viewed visual scenes depicting a person performing an action with an object that was either congruent or incongruent with the scene. Univariate and multivariate analyses revealed different activity for congruent compared to incongruent scenes in the lateral occipital complex, inferior temporal cortex, parahippocampal cortex, and prefrontal cortex, in line with existing models of scene processing. Importantly, and in contrast to previous studies, these activations could not be explained by task-induced conflicts. A secondary goal of this study was to examine whether object-context integration could occur in the absence of awareness, by comparing brain activity elicited by congruent vs. incongruent scenes that were suppressed from awareness using visual masking. We found no evidence for brain activity differentiating between congruent and incongruent invisible scenes. Overall, our results provide novel support for the roles of PHC and PFC in conscious object-context integration which cannot be explained by either low-level differences or task demands. Yet they further suggest that activity in these regions is decreased by visual masking to the point of becoming undetectable with our fMRI protocol.