Motion-Dependent Filling-In of Spatiotemporal Information at the Blind Spot

Abstract

We usually do not notice the blind spot, a receptor-free region on the retina. Stimuli extending through the blind spot appear filled in. However, if an object does not reach through but ends in the blind spot, it is perceived as "cut off" at the boundary. Here we show that even when there is no corresponding stimulation at opposing edges of the blind spot, well known motion-induced position shifts also extend into the blind spot and elicit a dynamic filling-in process that allows spatial structure to be extrapolated into the blind spot. We presented observers with sinusoidal gratings that drifted into or out of the blind spot, or flickered in counterphase. Gratings moving into the blind spot were perceived to be longer than those moving out of the blind spot or flickering, revealing motion-dependent filling-in. Further, observers could perceive more of a grating's spatial structure inside the blind spot than would be predicted from simple filling-in of luminance information from the blind spot edge. This is evidence for a dynamic filling-in process that uses spatiotemporal information from the motion system to extrapolate visual percepts into the scotoma of the blind spot. Our findings also provide further support for the notion that an explicit spatial shift of topographic representations contributes to motion-induced position illusions.

Fig 1. Blind spot measurements.

Results from the blind spot measurements are shown for 5 observers.

Fig 2. Experiment 1.

A Stimuli in Experiment 1. A sinusoidal grating was shown, drifting into or out of the blind spot, or flickering in counterphase. In the other interval, a flickering grating was shown in the fellow eye. Its length was varied from trial to trial, and observers judged which interval contained the “longer” grating. B Aggregate psychometric functions for 4 observers (8 eyes), showing cumulative Gaussian fits to responses as a function of the length difference in the fellow eye vs. the blind spot eye. The PSE for flickering gratings was defined as zero (no elongation or shortening). Gratings drifting into the blind spot (blue curve) appeared longer. C Means of PSEs from psychometric functions fitted to individual observers’ data (and SEM) for inward and outward drifting gratings.

Fig 3. Experiment 2.

A Stimuli in Experiment 2. A 2D-plaid was shown at the top border of the blind spot, drifting into or out of the blind spot, or drifting sideways toward fixation. In the other interval, the plaid was shown in the fellow eye drifting sideways, with its length varied from trial to trial. Observers judged which stimulus appeared “longer”. B Aggregate psychometric functions for 4 observers (8 eyes). The PSE for sideways drifting gratings was defined as zero (no elongation or shortening). C Means of PSEs from psychometric functions fitted to individual observers’ data (and SEM) for inward and outward drifting plaids.

Fig 4. Experiment 3.

Do moving stimuli extending into the blind spot appear longer because of a spread of luminance information (A), or because of extrapolation of the grating’s structure (B)? B shows the stimuli employed in Experiment 3. Observers judged the numerosity of dark/light transitions of a grating that was flickering or drifting into or out of the blind spot. The grating’s spatial frequency was varied from trial to trial. Observers compared it to a flickering grating with a fixed spatial frequency (0.5 cycles per degree) in another interval. C Aggregate psychometric functions for 4 observers (8 eyes), showing numerosity responses as a function of spatial frequency for flickering, inward, and outward drifting gratings. D Means of PSEs (with SEMs) from individual function fits. Inward drifting gratings with a lower spatial frequency were perceptually matched to a flickering grating at 0.5 cycles per degree. This indicates that the number of perceived dark/light transitions was increased for inward drifting gratings.

Fig 5. Experiment 4.

Do moving stimuli that extend through the blind spot appear filled-in using luminance information from the blind spot borders (A), or does filling-in use information from the grating’s structure (B)? B shows a schematic of stimuli employed in Experiment 4. Observers viewed a grating reaching through the blind spot that was flickering, drifting up or down, or remaining stationary. The grating’s spatial frequency was varied from trial to trial, and observers judged its apparent density by comparing it to a flickering grating in another interval, whose spatial frequency was always fixed at 0.3 cycles per degree. C Aggregate psychometric functions for 4 observers (8 eyes), showing responses as a function of spatial frequency for flickering, upward, downward drifting, and static gratings. D Means of individual PSEs (with SEMs). Static gratings with a physically higher spatial frequency appeared perceptually matched in density to a flickering grating of 0.3 cycles per degree. In other words, a static grating’s perceptual density was grossly underestimated. Moving gratings also needed increased spatial frequency to appear perceptually matched, but were perceived more accurately than static gratings.