Purchase this article with an account.
Caterina Gratton, Kartik Sreenivasan, Michael Silver, Mark D'Esposito; Effects of feature-based attention on voxel tuning curves for individual faces. Journal of Vision 2012;12(9):915. doi: 10.1167/12.9.915.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Although it is known that attention changes the profile of neural population responses to low-level features, the influence of attention on the representations of complex categories such as faces are less understood. Here, we recorded fMRI responses to individual members of a set of morphed faces, constructed voxel-based tuning curves along the face morph axis, and examined how these tuning curves were modulated by feature-based attention directed to one of the faces. Analyses were conducted in voxels in inferior temporal and occipital cortex that responded preferentially to faces compared to scenes. Voxel-based tuning curves (Serences et al., 2009) were generated by measuring the responses of individual voxels to each of six faces along a morph continuum (F1 to F6). Cross-validation using independent data sets revealed that the majority of tuning curves exhibited significant selectivity for individual faces. Furthermore, tuning curves showed distinct characteristics in different areas: posterior face areas (FFA, OFA, STS) had significant tuning for individual faces, while more anterior face areas in temporal cortex showed more categorical tuning for faces (F1-3 vs. F4-6). In the attention task, individuals were cued to attend to just one of a pair of superimposed faces (F1 and F6) and to detect morphs of the attended face to a new face that was orthogonal to the F1-F6 continuum. Attending to one of the faces selectively enhanced responses to the superimposed face pair in voxels previously defined as preferring the attended face. These findings show that fMRI can be used to classify individual face preferences in single voxels. Furthermore, they demonstrate that directing attention to individual faces selectively modulates responses in individual voxels that are selective for facial identity. In addition, we are employing coherency analysis to identify regions that interact with voxels that are tuned for individual faces.
Meeting abstract presented at VSS 2012
This PDF is available to Subscribers Only