Abstract
Abstract: :
Purpose: The extent and magnitude of contour interaction for high contrast letters is well documented. However, previous work has shown no effect of contour interaction for low contrast letter targets. The aim of this study was to determine the relationship between contour interaction and target contrast. Methods: Ten visually normal subjects (mean age 24.1±4.2 years) participated with informed consent in the study. The subjects wore their full spectacle correction and all had acuities of 6/6 (20/20) or better. A computer based vision testing program (TestChart2000PRO, Thomson Software) was employed to present single optotypes with and without flanking bars, at 10 levels of contrast varying between 90% and 5%. Stimuli were presented on a standard SVGA monitor at a viewing distance of 6m.The separation between letters and flanking bars was set at 40% of the letter width, producing maximal contour interaction. Acuity thresholds and contrast ratios were determined psychometrically for each contrast level. Results: The mean crowding ratios were 1.20±0.13 at 90% target contrast and this fell to 0.99±0.23 at 5% target contrast. Crowding ratios were significantly greater than 1 (p<0.01) for all contrast levels above 40%. Between 5% and 40% target contrast the crowding ratios were not significantly different from 1, however, a substantial increase in the variability of the crowding ratios over this range was evident. Conclusions: Contour interaction produces a similar decrease in letter acuity for all target contrasts above 50%. For target contrasts in the mid range from 20–50% the magnitude of contour interaction shows a high degree of inter–subject variability. Contour interaction is absent for low target contrasts (5% & 10%). These results suggest a non–linearity may exist in the mechanisms underlying contour interaction.