Abstract
Purpose :
Contrast sensitivity (CS) thresholds across the visual field (VF) have asymmetries along horizontal and vertical meridians, depicted by the “Hill of Vision” (HoV). In order to optimise clinical VF testing, we aimed to determine the reason for the asymmetries. We hypothesise that the asymmetries may be due to changes in spatial summation characteristics: altered CS within complete spatial summation, differences in the critical area (Ac), and the slope of partial summation (n2).
Methods :
We measured thresholds using the Humphrey Visual Field Analyzer across Goldmann sizes I-V for 48 normal observers on the 30-2 pattern and 20 normal observers on a custom test that extended the range to 48° nasally and temporally, and 42° superiorly and inferiorly. Khuu and Kalloniatis’s dB* metric (2015, IOVS) was used to scale the dB output using a size factor to equate for spatial summation. Data were fitted using two-line segmental regression to determine Ac and n2 (where n1 = -1), and were compared along horizontal and vertical meridians.
Results :
Nasal-temporal asymmetries were most apparent beyond 15°, with larger asymmetries at greater eccentricities (p<0.0001) and with small stimulus sizes (p<0.0001). The size III stimulus was close to or outside Ac, showing asymmetries that firstly followed the same pattern as size IV-V, then diverged to become similar to size I-II in the periphery, following the Ac increase with eccentricity. CS asymmetry with size IV-V stimuli plateaued at 36-48°, corresponding to steepening and asymmetry of n2 nasally. Inferior thresholds were higher than at analogous superior locations. These were affected by eccentricity (p<0.0001) but not size, corresponding to no asymmetry in Ac or n2 vertically.
Conclusions :
Horizontal asymmetries in the HoV are explained by three components of the spatial summation function. If the stimulus size is within Ac, asymmetry is determined solely by changes in CS; outside Ac, it is dependent upon changes in CS and n2. Ac dictates the position of n2, explaining the convergence of threshold asymmetries seen with larger stimuli toward parity. Vertical asymmetries increased monotonically with eccentricity, with no effect of size, Ac or n2. To obtain CS measurements that represent contrast detection ability alone across the VF, stimuli within Ac should be used for perimetry. A Goldmann size III target is at or outside Ac in the 30-2 VF, and so is dependent upon partial summation.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.