Meridional Amblyopia Impairs Contour Integration

Abstract

Abstract: : Purpose: We studied the ability of normal and amblyopic observers to use a contour string as a cue to locate a test patch in noise. Methods: A collinear string of Gabor patches (the cue) was used as a pointer to a test patch whose orientation matched that of the string. The cue + test were presented at a random location within a 2 degree window, among a variable number of noise patches. We measured the detectability of the cue at different noise densities. For our amblyopic observers we adjusted the contrast, and the number of patches in the cueing string presented to the amblyopic eye (AE) to match the visibility of a 3-patch string presented to the non-amblyopic eye (NAE). We also measured contrast increment thresholds on a patch whose location and orientation were cued by a "matched" string. Results: Contrast increment thresholds measured in normal observers and in the NAEs of our amblyopic observers were inversely related to the visibility of the cue. Increasing spatial uncertainty and decreasing the number of patches in the cue led to higher contrast increment thresholds. The amblyopic eyes (AE) showed the same general trend with a "matched" cueing string that had lower spatial uncertainty, higher contrast and a larger number of patches. The exception occurred in two of our 4 amblyopic observers who had high astigmatism and meridional amblyopia. Despite matching the visibility of the cueing string, these observers had elevated increment thresholds for a test (and cue) oriented along their amblyopic meridians, while thresholds for tests oriented along orthogonal meridians in the AE were comparable to thresholds in the NAE eye. Conclusions: These results suggest that while some factors such as reduced spatial uncertainty, increased contrast and a greater number of patches can overcome the contour integration deficit in the AE, there remains a further impairment along an amblyopic meridian. Thus, meridional amblyopia impairs contour integration.

Keywords: astigmatism • grouping and segmentation • spatial vision