August 2016
Volume 16, Issue 12
Open Access
Vision Sciences Society Annual Meeting Abstract  |   September 2016
Investigating cortical feedback of objects and background scene to foveal and peripheral V1 using fMRI
Author Affiliations
  • Matthew Bennett
    Centre for Cognitive Neuroimaging, University of Glasgow
  • Lucy Petro
    Centre for Cognitive Neuroimaging, University of Glasgow
  • Lars Muckli
    Centre for Cognitive Neuroimaging, University of Glasgow
Journal of Vision September 2016, Vol.16, 568. doi:https://doi.org/10.1167/16.12.568
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Matthew Bennett, Lucy Petro, Lars Muckli; Investigating cortical feedback of objects and background scene to foveal and peripheral V1 using fMRI. Journal of Vision 2016;16(12):568. https://doi.org/10.1167/16.12.568.

      Download citation file:


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

      ×
  • Supplements
Abstract

When discriminating abstract objects isolated in peripheral visual field, object information can be detected in non-stimulated foveal but not peripheral cortex (Williams et al., 2008). Conversely, complex natural scene information can be detected in non-stimulated peripheral cortex (Smith & Muckli, 2010; Muckli et al. 2015). This demonstrates cortical feedback contributions to object and scene representation in V1. Whether object and scene feedback are automatically directed to foveal versus peripheral V1, or if this is due to varying stimulus complexity/naturalism remains unknown. We addressed this question using computer-generated images of objects embedded in naturalistic scenes. Nine subjects underwent block-design fMRI (3T). Eight naturalistic computer-generated images variously combined objects and background scenes (Figure 1). The central portion and upper-right quadrant were occluded, preventing feed-forward stimulation of the ROIs, allowing isolation of feedback signals. Subjects discriminated either the scene or object. Multivoxel patterns (TR 2s; voxel-size 2mm3) from the ROIs were entered into support vector machine (SVM) classifiers. Bootstrapping determined above-chance group-level classification. In fovea (Figure 2), we classified object presence regardless of task (55.7%, 53.8%, collapsing task: 56.5%). Classifying object identity was only possible when collapsing task (55.9%). Also in fovea, we classified scene information during the scene task (56.7%) and when collapsing task (59.5%) but not during object task (52.2%). Therefore, the non-stimulated fovea contains object and task-dependent scene information. In periphery (Figure 3), no object information was detected. We classified scene information during object task (57.4%) and when collapsing task (55.3%), but not during scene task (54.6%). Therefore the non-stimulated periphery contains only scene information. Our data suggest specialisation in cortical feedback to V1 cortex: scene information is fed-back diffusely to foveal and peripheral V1, whereas object feedback is directed to foveal cortex – possibly for high resolution scrutiny (Williams et al., 2008; Levy et al., 2001).

Meeting abstract presented at VSS 2016

×
×

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.

×