Abstract
Subjects with unilateral cortical blindness (CB) report difficulties reading, navigating, and recognizing shapes despite normal visual processing in at least one visual hemifield and intact eye movements. Dilks and colleagues (2007) reported shape distortion across the blind field border in one quadrantanope, prompting us to ask if this could represent a more generalized phenomenon. To begin answering this question, we assessed target localization in 6 CB subjects. A Gabor patch (3° diameter) was presented on a touchscreen at one of 121 eccentric locations, followed by a full-screen mask. Subjects were instructed to touch the screen where the target had been presented. There were five repetitions per target location. The task was completed using static (spatial frequency [SF]: 1 cpd, temporal frequency [TF]: 0 Hz) or drifting (SF: 1 cpd, TF: 10 Hz) Gabors. Four of the CB subjects repeated the task after one year of global direction discrimination training (see Huxlin et al., 2009; Das et al., 2014). Localization accuracy was computed as the mean difference between the five touch responses and target location; reliability was computed as the standard error of that mean. Humphrey perimetry was performed to map areas of reduced luminance sensitivity (LS) in each CB subject. Compared to intact regions of the Humphrey visual field (LS ≥25 dB), CB subjects were less accurate and reliable in deeply blind regions (LS ≤3 dB), but displayed comparable accuracy and reliability in intermediate regions (LS=3-25 dB). Target type (static vs. drifting) had no effect on accuracy or reliability. Quadrantanopes maintained a general ability to distinguish upper from lower, and central from more peripheral targets in their blind field; hemianopes did not. Finally, visual discrimination training, which improved performance in CB fields, also reduced mean localization error, suggesting that discrimination training improves utilization of location information in cortically blind fields.
Meeting abstract presented at VSS 2016