Abstract
It is well known that neurons in the temporal lobe classify objects, such as faces, and it is generally assumed that the activity of such neurons is necessary for conscious awareness of these objects. However, object categorization may also occur unconsciously, as has been shown by the selective activation of object selective neurons by masked objects. So what distinguishes conscious from unconscious object recognition? We constructed schematic images containing objects such as faces and houses while keeping local retinal stimulation between conditions identical. Using a dichoptic fusion paradigm, we manipulated stimulus visibility such that objects were either visible or not visible. Confirming earlier results, we found that both consciously perceived and non-perceived objects result in category specific BOLD activation. Critically however, we show that only consciously seen objects show a distinct neural signature of figure-ground segregation in early and midlevel visual areas, which is completely absent when objects are not seen. Although counterintuitive, this implies that consciousness is more intimately related to processes of figure-ground segregation and perceptual organization than to object categorization. We propose that figure-ground segregation is a prerequisite for visual awareness, and that both phenomena share part of their neural correlate, which is recurrent processing in visual cortex.