July 2013
Volume 13, Issue 9
Free
Vision Sciences Society Annual Meeting Abstract  |   July 2013
Investigating the spatial frequency content of cortical feedback using fMRI
Author Affiliations
  • Lars Muckli
    Centre for Cognitive Neuroimaging (CCNi), Research Institute of Neuroscience and Psychology, College of Medical, Veterinary and Life Sciences, University of Glasgow
  • Lucy S Petro
    Centre for Cognitive Neuroimaging (CCNi), Research Institute of Neuroscience and Psychology, College of Medical, Veterinary and Life Sciences, University of Glasgow
  • Hans-Christian Rath
    Centre for Cognitive Neuroimaging (CCNi), Research Institute of Neuroscience and Psychology, College of Medical, Veterinary and Life Sciences, University of Glasgow
  • Fraser W Smith
    Centre for Cognitive Neuroimaging (CCNi), Research Institute of Neuroscience and Psychology, College of Medical, Veterinary and Life Sciences, University of Glasgow
Journal of Vision July 2013, Vol.13, 1067. doi:https://doi.org/10.1167/13.9.1067
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Lars Muckli, Lucy S Petro, Hans-Christian Rath, Fraser W Smith; Investigating the spatial frequency content of cortical feedback using fMRI. Journal of Vision 2013;13(9):1067. https://doi.org/10.1167/13.9.1067.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

One aim in contemporary neuroscience is to understand the information processing in cortical feedback to early visual areas, such as its spatial frequency content. It has been shown that V1 can generalise between photographs filtered at different spatial frequencies, and even between line drawings and complete natural scenes (Walther et al., 2011). However, it remains unknown how feedback to V1 generalises between spatial frequencies, i.e. when V1 receives no feedforward scene information. Using functional magnetic resonance imaging, we identified regions of V1 responding to a white occluder presented over the lower right quadrant of spatial-frequency filtered versions of centrally-presented natural scenes. Activation patterns from this "non-stimulated" (and thus feedback-receiving) region of V1 were entered into a multivariate pattern classification analysis (Smith & Muckli, 2010). In the control condition, subjects were shown the scene in the lower right quadrant (henceforth the ‘feedforward’ condition). We performed two analyses, firstly in both feedforward and feedback conditions, we classified between two different images presented at either high (HSF) or low spatial frequencies (LSF). Secondly, again in both feedforward and feedback conditions, we cross-classified across spatial frequencies i.e. training the classifier on low-spatial frequency images and testing on high-spatial frequency images, and vice versa. The first analysis revealed the "non-stimulated" region of V1 responding to the occluder carries complex information that can discern the surrounding context, as we were able to classify above chance between the two HSF and two LSF scenes, although as expected not as accurately as in the feedforward condition. We were also able to cross-classify in the feedforward condition. However, the missing cross-classification of feedback indicates a lack of generalisation from high to low spatial frequencies, at least in the parameter space measured so far.

Meeting abstract presented at VSS 2013

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×