Abstract
Retinal disparity is a major source of information about the depth structure of an object or a scene around the fixation point. Veridical shape, egocentric distance, or their time derivatives indicating shape changes and motion in depth relative to the head, are not directly given by the retinal disparity field. Headcentric motion in depth is given by the changing target vergence, which depends on the changes of retinal disparity and eye vergence. Here we investigated with fMRI which cortical areas responded to changing retinal disparity or changing headcentric disparity. We used wide-field flow patterns to simulate forward motion through a cloud of dots. We decoupled the optic flow from the disparity information by projecting the motion of each point onto a virtual Vieth-Muller torus. This attributes to each point the same horizontal retinal disparity, while the motion of the point as seen from the cyclopic eye remains as if the point moved in 3D space. By motion in depth of the Vieth-Muller projection surface - independent of the fixation point- we decoupled retinocentric and headcentric (horizontal) disparity. Thus, we manipulated independently the contributions of retinal- and headcentric change in disparity and optic flow to the perceived motion in depth. In eight subjects, changing vergence and changing retinal disparity evoked strong acitvity in early visual areas. Optic flow sensitive areas pMST, pVIP, V3A, and V6[sup]+[/sup] showed strong responses to changes in retinal disparity, but only V3A and V6[sup]+[/sup] showed strong responses to headcentric changes in disparity. BOLD responses to the headcentric not the retinocentric changing disparity were consistent with the perceived speed of changing distance.
Meeting abstract presented at VSS 2013