Purchase this article with an account.
Daniel Berman, Dirk B. Walther; Differential Selectivity for Spatial Frequencies in Anterior and Posterior PPA. Journal of Vision 2014;14(10):1081. doi: 10.1167/14.10.1081.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The Parahippocampal Place Area (PPA) has been reported to consist of functionally distinct subregions: The anterior region is more related to high-level representations of scene content, whereas the posterior region is driven more by low-level stimulus properties (Baldassano et. al., 2013). How do these distinct regions differ in their sensitivity to spatial frequencies? On the one hand, PPA is sensitive to global image properties such as openness and naturalness (Park et al. 2011) or physical size and clutter (Park et al. in press), which are mainly determined by low spatial frequencies. On the other hand, PPA has been reported to respond more strongly to high than low spatial frequencies (Rajimehr et. al., 2011). In studies like these, spatial frequency content of stimuli is typically confounded with semantic information. Here we test spatial frequency sensitivity of PPA in an fMRI experiment with stimuli designed to be devoid of any coherent structure. We presented subjects with a series of visual patterns with narrowly defined spatial frequency bands centered around 0.2 to 18.1 cycles per degree and random phase. We found that anterior PPA was not significantly activated by these unstructured, semantically meaningless stimuli. Posterior PPA, however, was significantly activated. Moreover, the BOLD signal in posterior PPA was modulated by spatial frequencies. In contrast to the findings by Rajimehr et. al. we found that low spatial frequencies activated posterior PPA more strongly than high spatial frequencies in 13 out of 14 subjects. In order to explore the effect of semantic information on spatial frequency selectivity we are performing a follow-up study using frequency-filtered images of natural scenes. We will present results from a univariate analysis as well as decoding of scene categories from activity patterns in anterior and posterior PPA, separately for high-pass and low-pass filtered images.
Meeting abstract presented at VSS 2014
This PDF is available to Subscribers Only