Pooja Nandagopal¹, Andrew J Anderson¹, Allison M Mckendrick^1,2; ¹The University of Melbourne, ²Lions Eye Institute, University of Western Australia

Motion segmentation (distinguishing a moving foreground object from its background) is thought to benefit from spatial suppression of large, high-contrast backgrounds, and has been studied in central vision. This study examined motion segmentation and spatial suppression of motion in central and peripheral vision, their correlation, and whether scaling peripheral stimuli based on cortical magnification eliminates perceptual differences between eccentricities. Ten healthy adults (mean 27.5 years, range:19–35) completed computer-based tasks at five contrast levels (12%–92%), three eccentricities: (0°, 10°, 20°) and two stimulus conditions (scaled and unscaled). In the motion segmentation task, participants identified the tilt (45° right/left) of a motion defined black-and-white textured ellipse within an oppositely moving black-and-white textured background. Segmentation threshold was the minimum stimulus exposure duration for accurate tilt discrimination. To measure motion segmentation efficiency, participants were shown the ellipse in isolation and asked to identify its motion direction (up or down). Motion segmentation efficiency was then calculated as the difference in log10 threshold values: log10(ellipse motion threshold) – log10(motion segmentation threshold). Suppression strength was measured requiring participants to identify the motion direction of textured background patch (radius = 5.3°) presented alone. Duration thresholds were measured, with suppression index calculated as the difference between log10 duration thresholds of the highest and lowest contrast stimuli for each participant. In the unscaled condition, motion segmentation thresholds increased (RM-ANOVA, main effect of eccentricity, p<0.01) and suppression index (p=0.007) decreased with eccentricity. The correlation between segmentation efficiency and suppression index weakened systematically (Pearson’s r = -0.87 at 0°, r = -0.54 at 10°, r = -0.33 at 20°). Scaling reduced segmentation thresholds (RM-ANOVA, main effect of scaling, p<0.01) and increased the suppression index (p<0.01) but did not eliminate regional differences. Therefore, our ability to segment moving foreground objects deceases with eccentricity, even when the stimulus in the periphery is enlarged.