Purchase this article with an account.
Lars Strother, Sarah Shomstein, Marlene Behrmann; Evidence from fMRI for structural non-selectivity in texture segregation. Journal of Vision 2007;7(9):974. doi: https://doi.org/10.1167/7.9.974.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Texture segregation is a fundamental process in visual perception. The present study employed rapid event-related functional magnetic resonance imaging (fMRI) to investigate neural mechanisms involved in the perception of texture-defined regions. Observers viewed arrays of Gabor elements arranged on a polar grid and were instructed to indicate the location of a target region that differed in global structure from the remainder of the array. Local orientation determined three types of structure in this region: (a) concentric curvilinear structure, (b) randomly arranged elements or (c) an intermediate level of perturbed curvilinear structure. These same properties also characterized regions surrounding the target. Target salience was determined behaviorally both prior to and during fMRI acquisition and was confirmed to be a function of the structure of the target region relative to its surroundings.
We used retinotopic mapping to compare BOLD responses in early visual areas (V1, V2, V3/VP, V4/V8) to those obtained in higher visual areas. We found strong evidence for correlations between BOLD signal and target salience in V4/V8 but not in earlier visual areas. The strongest correlations were observed for attention-related areas in parietal cortex (e.g. SPL and TPJ). Our findings suggest differential sensitivity to the structure of a target relative to its surroundings. Whereas some visual areas (e.g. V4/V8 and LO) exhibited responses consistent with the acquired behavioral data, this consistency was observed to a lesser degree in earlier retinotopic visual areas. Interestingly, the observed fMRI responses do not seem to correspond to the specific structure of a target region but to the salience of this region in a particular context. Our findings have implications for the neural systems involved in mapping visual salience irrespective of the basic visual properties that give rise to the perceived global structure.
This PDF is available to Subscribers Only