Estimating retinal fixation location using fMRI

Abstract

Purpose. We are developing a method for retinotopic mapping of the visual cortex in people with central field loss. Retinotopic mapping based on polar coordinates uses rotating wedges and expanding rings, which assume stable foveal fixation. However, people with central field loss usually adopt a non-foveal retinal location for fixation and that location is typically unknown. Therefore, instead of using the polar-coordinate based mapping method, which requires a known origin, we developed a set of phase-coded mapping stimuli in Cartesian coordinates that will also allow us to estimate the fixation locations relative to the unknown fovea location. The current study aims to validate this new mapping procedure in observers with normal vision. Method. High-resolution structural and functional MR images were acquired on a Siemens 3T system. Two mapping stimuli were used. Both were 2 deg wide checkered bars, which extended across the full-display. One was a vertical bar moving across the screen from left to right. The other was a horizontal bar that moved from top to bottom. In both cases, the bar traversed the display 4 times, each pass taking 72 s. Since the subjects fixated on a dot at the center of the display, the bars should cross the fixation at the mid-point of the 72-second period. This time point was compared with the time point when the bars cross the fovea, which was estimated by deriving the switch of V1 activation from one hemisphere to the other hemisphere produced by the vertical bar, and from the lower bank to the upper bank of the calcarine sulcus produced by the horizontal bar. Results. With this new method, the estimated fovea locations were within 2 deg from the actual locations. Conclusions. Our new stimuli allow us to estimate the fixation location with acceptable precision. This new procedure may be useful for retinotopic mapping on people with central field loss.