It has long been debated whether attention alters the categorical selectivity in regions such as the fusiform face area (FFA) and the visual word form area (VWFA). We addressed this issue by examining whether the spatial pattern of neural representations for certain stimulus categories in these regions would change under different attention conditions. Faces, Chinese characters, and textures were presented in a block design fMRI experiment where participants in different runs attended to the stimuli under different conditions of attention. After localizing regions of interest (ROIs) in FFA and VWFA using general linear models, we performed spatial pattern analyses to examine both within- and cross-condition classification in these ROIs. The within-condition results replicated previous findings showing significant classification accuracy reduction when there was less attention compared with more attention. Critically, cross-condition classification in both FFA and VWFA revealed significantly above-chance accuracy for all stimulus categories, suggesting similar spatial neural representations across different attention conditions. Further strengthening this conclusion, when the contrast-to-noise ratio (CNR) of the signals was adjusted to increase signal strength, cross-condition classification accuracy for faces in FFA and for Chinese characters in VWFA improved significantly, even approaching within-condition accuracy. This indicates that attention does not modulate the spatial pattern of neural representations involved in category selectivity, but only changes the signal strength relative to the noise level.
Copyright © 2012 Elsevier Ltd. All rights reserved.