Purchase this article with an account.
Pinglei Bao, Xiaomin Yue, Bosco S. Tjan; BOLD signal response functions for object and face processing in noise. Journal of Vision 2008;8(6):46. doi: 10.1167/8.6.46.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Tjan, Lestou, and Kourtzi (2006, J Neurophysiol) postulated that the log-log slope of signal response function (BOLD signal amplitude vs. signal-to-noise ratio (SNR) of the stimulus) is related to the intrinsic uncertainty or feature invariance of a visual area, and therefore indicates the ordering of the information processing in the cortex. Here, we examined the generality of this conjecture using two classes of stimuli, tested separately: common objects and faces. In the scanner, subjects decided which of the two noisy sample images matched the category (for objects) or gender (for faces) of the target image. For each scan, four levels of image SNR were tested in a rapid event-related design. The mean luminance and RMS contrast of the images were kept constant. To measure the log-log slope of signal response functions across the cortical surface without using predefined regions of interest (ROIs), we defined, for each voxel, a “floating” ROI of 15 mm in diameter on the cortical surface, from which we determined the hemodynamic response functions (HRF) for each stimulus SNR using deconvolution. We then estimated the peak amplitudes of the HRFs and computed the log-log slope of the signal response function. We found that for both objects and faces, the log-log slope of the signal response function increased orderly from low- to high-level visual areas, consistent with Tjan et al. We also found that “islands” of slope maxima developed in the anterior regions of the visual cortex, where slopes were high. The locations, but not the values, of the maxima were consistent across both faces and objects, with some corresponding to well-known regions: LO, pFs, and FFA. Assuming the conjecture of Tjan et al, these islands would represent different branches of the visual processing hierarchy.
This PDF is available to Subscribers Only