Abstract
Visual discrimination can improve with practice. For stereoacuity, 80% improvement in threshold have been reported (Fendick and Westheimer, Vision Research 1983). Here, we exploit the “cue recruitment” paradigm (Haijiang et al., PNAS 2006) during perceptual learning to improve stereoacuity. We used static occlusion cues that had varied reliability as “training wheels” to guide participants' use of disparity during the construction of perceived depth, which might in principle lead to better performance on a stereoacuity task. Stimuli were textured rectangles, one above the other, with varied horizontal offset. Subjects indicated which appeared closer and were given feedback. Four subjects were screened and assigned to one of two groups, “disparity-plus-occlusion” and “disparity-only”. Half of the stimuli presented to the disparity-plus-occlusion group contained occlusion cues, and the remainder contained only disparity because the region of overlap was occluded. All stimuli for the disparity-only group contained only disparity. Each subject participated in three sessions lasting approximately 45 minutes, each containing 300 trials, on three non-successive days. In both groups, the same staircase procedure controlled the amount of disparity and monitored improvements in threshold based on responses from every other trial (always a disparity-only trial). Two of four subjects showed improvements in stereoacuity threshold on day three as compared to day one, but this improvement was recorded in one subject from each group. Thus, data collected to date do not suggest a positive correlation between improvements in stereoacuity thresholds and training with correlated cues in subjects with normal vision. Additional data may resolve this question. Furthermore, it could be clinically useful to add “training wheel” cues to accelerate improvement in visual function, for example, during vision training/orthoptics after surgical alignment of the eyes.
NIH R01 EY-013988, NIH Loan Repayment Grant to Dr. Llerena Law.