Abstract
Goal: Visual perception is heterogeneous across the visual field. Contrast sensitivity (CS) varies as a function of radial (eccentric) and polar (isoeccentric) position. CS decreases with eccentricity. At isoeccentric locations, CS is higher on the horizontal than the vertical meridian, and is higher along the lower than the upper vertical meridian. Here, using the external noise paradigm, we assess whether variations in sensitivity with eccentricity and polar angle are due to differences in processing efficiency or internal noise, and how endogenous (voluntary) covert attention affects CS differences at these locations. Method: In each trial, participants reported the orientation of a ±45°-oriented Gabor (5 cpd) presented at one of 9 possible locations along the horizontal (left/right) and vertical (up/down) meridian, at 3 eccentricities (0, 4 and 8°). A response cue indicated the target location, which was embedded in different levels of Gaussian white noise. Using symbolic precues, endogenous attention was either focused at the upcoming target location or distributed across all locations. We computed threshold versus noise contrast functions to estimate differences in processing efficiency and internal noise as a function of eccentricity and polar angle. Results: Thresholds increased with eccentricity under low external noise, but were similar across eccentricity (0–8°) under high external noise. Conversely, CS differences at isoeccentric locations remained similar across external noise levels: thresholds were better along the horizontal than the vertical meridian, and better at the lower than the upper vertical meridian. Endogenous attention lowered thresholds across all external noise levels similarly at different eccentricities and polar angles. Conclusion: These findings demonstrate that variations in CS with eccentricity and polar angle are mediated by distinct mechanisms. Internal noise differences underlie the effect of eccentricity. However, differences across isoeccentric locations result from reduced efficiency along the vertical meridian, particularly along the upper vertical meridian, even when attended.
Acknowledgement: NIH NEI RO1-EY027401 to MC