Published in Annual Meeting, Neuroscience, (406.01), Nov 2013.
The visual areas of the primate cerebral cortex provide distinct representations of the visual world, each with a distinct function and topographic representation. Neurons in primary visual cortex respond selectively to orientation and spatial frequency, whereas neurons in inferotemporal and lateral occipital areas respond selectively to complex objects. But the areas in between, in particular V2 and V4, have been more difficult to differentiate on functional grounds. In previous work (Freeman, Ziemba et al., Nature Neurosci, 2013), we showed that, in both macaque and human, synthetic stimuli whose higher-order statistical properties emulated natural texture images yielded differential responses in V2, but not V1. We have now examined responses to these stimuli and to photographs of objects and scenes, to better distinguish the properties of V2 from those of the areas that provide its input and receive its output. We used fMRI to measure BOLD responses while human observers viewed rapidly presented images, while performing an attentionally-diverting task at fixation. We measured BOLD responses while presenting stimuli drawn from different stimulus classes in alternating blocks. We used three classes: photographs of scenes and objects, texture images matched to the higher-order statistics of these photographs (specifically, the spatially-averaged distribution and correlations of model V1 outputs, Portilla & Simoncelli, 2000), and noise images matched only to the overall orientation and frequency content of the photographs. As we found previously, fMRI responses to textures were larger than those to noise in V2, V3, and to some extent V4, but not in V1, and not in inferotemporal and lateral occipital areas thought to be selective for complex objects. In contrast, responses to photographs were larger than those to noise in those higher areas, as well as in V2-V4, but not in V1. The differentiation between V1 and V2, however, was less robust than for the more controlled, homogenous textures. Finally, comparing responses to photographs and textures revealed differential activity in higher areas, but little or no differential responses in V1-V4. Together, these three comparisons offer a simple functional account of the visual cortical cascade. Whereas V1 encodes basic spectral properties, V2, V3, and to some extent V4 represent the higher-order statistics of textures. Downstream areas capture the kinds of global structures that are unique to images of natural scenes and objects.