Decoding the reference frame of spatial vision by means of fMRI population receptive field mapping

Abstract

We perceive a stable visual world across eye movements, despite the drastic retinal image changes these movements produce. To explain the spatial stability of our perception, previous studies suggested that the brain encodes the location of attended visual stimuli in external coordinates using a spatiotopic reference frame. Such a spatiotopic reference frame is inconsistent with the well-established retinotopic organization of the visual system. Here, we used ultra-high-field (7T) fMRI single voxel population receptive field mapping to probe the reference frame of spatial vision, by independently manipulating both gaze direction and spatial attention. In separate runs, participants directed their attention to either a bar stimulus that systematically traversed the visual field, or to a fixation stimulus placed at one of three distinct horizontal screen positions. Both fixation and bar stimuli contained oriented noise textures on which participants performed an equally demanding discrimination task. We found that estimated population receptive fields in all cortical visual field maps shift with the eye, irrespective of how spatial attention is deployed. This is consistent with a fundamentally retinotopic reference frame for spatial visual processing. We next reasoned that a spatiotopic reference frame could be computed by gain-field operations at the level of entire visual areas rather than at the level of individual voxels. Using Bayesian decoding of stimulus location from BOLD response patterns in individual visual areas, we found that decoded stimulus locations also shift with the eye. This result holds for all visual areas and irrespective of the deployment of spatial attention. We conclude that spatial vision is encoded in a retinotopic reference frame throughout the visual hierarchy.