Abstract
Purpose: :
Cortical visual impairment (CVI) is a leading cause of bilateral vision impairment. Since many patients with CVI cannot perform an optotype acuity test, their acuity is often measured with a grating stimulus using a preferential looking test or the visual evoked potential (VEP). The purpose of this study is to determine the relationship between VEP vernier acuity, VEP grating acuity and behavioral grating acuity.
Methods: :
Sweep VEP vernier acuity, sweep VEP grating acuity, and behavioral grating acuity were measured in 36 patients with CVI. The patients ranged in age from 3.2 to 22.7 years (mean: 12.3; SD: 5.3). For the VEP measurements, the grating stimulus was swept from 1-20 c/deg and the vernier stimulus was swept from 1-90 arc min. Each sweep took 10 seconds. Behavioral grating acuity was measured with preferential looking cards. Because the measures of vernier acuity and grating acuity have different units, the results were expressed as times worse from normal (with normal being 30 c/deg and 0.5 arc min respectively) and then converted to log units.
Results: :
All 3 measures were obtained in 28 patients. Behavioral acuity could not be assessed in 8 patients. VEP vernier acuity ranged from normal to a loss of 1.4 log units (11.3 arc min); VEP grating acuity ranged from normal to a loss of 0.8 log units (5.2 c/deg); behavioral acuity ranged from normal to a loss of 1.6 log units (0.8 c/deg). VEP grating acuity and VEP vernier acuity were significantly related (r=0.72) but the slope was 1.4, indicating that vernier acuity was 2.5 times worse than VEP grating acuity. The slope relating behavioral grating acuity loss to VEP grating acuity loss was 1.6, indicating that the behavioral acuity was much more reduced (approx 4 times worse) than the VEP grating acuity (r=0.64). The slope relating VEP vernier loss to behavioral grating loss was 0.9, indicating that behavioral acuity and VEP vernier acuity showed approximately the same amount of reduction (r=0.67). Furthermore, a Bland-Altman plot showed a linear relationship between the two grating measures with a steep slope (r=0.80; slope=1.0) indicating that the discrepancy increases as the average acuity worsens. A Bland-Altman plot showed a non-significant relationship between VEP vernier acuity and behavioral acuity (r=0.32) but with high variability (68% within+/-0.3 log units).
Conclusions: :
VEP grating acuity was significantly better than behavioral grating acuity in patients with CVI. VEP vernier acuity was more similar to the behavioral measures. VEP vernier acuity may be a better predictor of behavioral acuity than the VEP grating acuity.
Keywords: visual impairment: neuro-ophthalmological disease • visual acuity