Purchase this article with an account.
Matthew Cavanaugh, Michael Melnick, Ruyuan Zhang, Mariel Roberts, Anasuya Das, Duje Tadin, Marisa Carrasco, Krystel Huxlin; Residual inefficiencies of recovered vision in cortically blind fields – insights from equivalent noise analysis. Journal of Vision 2014;14(10):659. doi: 10.1167/14.10.659.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Visual perceptual training was recently shown to recover coarse motion direction discrimination abilities in subjects with cortical blindness (CB) induced by damage to the primary visual cortex (Huxlin et al., 2009). Here we investigated: (a) whether such training also improves fine direction discriminations; (b) how training alters processing in CB by measuring how fine direction discriminations are affected by external noise. Seven CB observers and two age-matched visually-intact controls underwent baseline evaluation with controlled fixation. CB observers performed 300 trials/day on a left-right global direction discrimination task (Huxlin et al., 2009). After ~6 months of coarse motion discrimination training, we assessed: (a) fine motion direction discriminations using a same-different task; (b) direction difference thresholds for random dot stimuli containing different amounts of directional noise and analyzed the resulting threshold-vs-noise (TvN) functions. For the CB observers, training improved direction range thresholds for coarse direction discrimination back to normal levels of performance. While subjects also regained some ability to perform finer direction discriminations, difference thresholds at the retrained locations were approximately 3x worse than at corresponding locations in the intact hemifields. Threshold versus Noise (TvN) functions showed that thresholds decreased at all noise levels but primarily at low levels. However, compared to TvN curves in the intact visual hemifield and those of the control observers, the retrained, blind hemifield locations still possessed higher thresholds at all noise levels. Application of the Perceptual Template Model (Dosher & Lu, 1999) and the Linear Amplifier Model (Pelli, 1981) to the global motion discrimination paradigm suggests that training primarily increased signal enhancement (reduced internal noise), and that the residual inefficiency in motion processing may reflect some combination of abnormally high internal noise and inability to exclude external noise that are not completely overcome by training in cortically blind hemifields.
Meeting abstract presented at VSS 2014
This PDF is available to Subscribers Only