Purchase this article with an account.
Yi Gao, Alexander Baldwin, Robert Hess; Can perceptual learning alleviate the global motion direction discrimination deficit in amblyopia?. Journal of Vision 2016;16(12):1111. doi: https://doi.org/10.1167/16.12.1111.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Amblyopia causes a contrast sensitivity deficit in the amblyopic eye. However there is also evidence of additional deficits affecting functions further along the processing stream, e.g. an impairment in global motion processing that is not specific to the amblyopic eye. Perceptual learning studies on normal observers have demonstrated significant improvements in global motion tasks. On this basis, we have applied these methods in an attempt to: i) improve global motion processing in amblyopic observers, and ii) determine whether the learning effects are specific to the trained eye. We tested 5 normals and 6 amblyopes on a motion direction discrimination task. Our stimuli were a field of isotropic log-Gabors with peak spatial frequency of 3 c/deg (spatially band-pass "dots"). In each trial we first presented a stimulus with a fixed reference motion direction, and then a test stimulus with its motion direction defined as an offset from that reference. The observer responded whether the second interval's direction was clockwise or anti-clockwise relative to the first. The difficulty of the task was varied by modifying the offset angle. We measured monocular baseline thresholds for each eye (day 1), and then conducted 10 days of monocular training for 40 minutes/day (days 2-11). Half of the amblyopes trained with their amblyopic eye, half with their fellow eye. After training we then made two retest measurements for each eye (days 12 and 13). Surprisingly, we do not find the expected training effect in either our normal or our amblyopic observers. Thresholds were generally lower following training, however this difference is not statistically significant. It is possible that the critical difference between our study and those that have found large training effects is the spatially broadband nature of the stimuli used in previous studies.
Meeting abstract presented at VSS 2016
This PDF is available to Subscribers Only