Using high-density perimetry to explore new approaches for characterizing visual field defects

High-density threshold perimetry has found that conventional static threshold perimetry misses defects due to undersampling. However, high-density testing can be both slow and limited by normal fixational eye movements. We explored alternatives by studying displays of high-density perimetry results for angioscotomas in healthy eyes-areas of reduced sensitivity in the shadows of blood vessels. The right eyes of four healthy adults were tested with a Digital Light Ophthalmoscope that gathered retinal images while presenting visual stimuli. The images were used to infer stimulus location on each trial. Contrast thresholds for a Goldmann size III stimulus were measured at 247 locations of a 13°×19° rectangular grid, with separation 0.5°, extending from 11° to 17° horizontally and -3° to +6° vertically, covering a portion of the optic nerve head and several major blood vessels. Maps of perimetric sensitivity identified diffuse regions of reduced sensitivity near the blood vessels, but these showed moderate structure-function agreement that was only modestly improved when effects of eye position were accounted for. An innovative method termed slice display was used to locate regions of reduced sensitivity. Slice display demonstrated that many fewer trials could yield similar structure-function agreement. These results are an indication that test duration might be reduced dramatically by focusing on location of defects rather than maps of sensitivity. Such alternatives to conventional threshold perimetry have the potential to map the shape of defects without the extensive time demands of high-density threshold perimetry. Simulations illustrate how such an algorithm could operate.

Functional changes at the preferred retinal locus in subjects with bilateral central vision loss

PURPOSE

Subjects with bilateral central vision loss (CVL) use a retinal region called the preferred retinal locus (PRL) for performing various visual tasks. We probed the fixation PRL in individuals with bilateral macular disease, including age-related macular degeneration (AMD) and Stargardt disease (STGD), for localized sensitivity deficits.

METHODS

Three letter words at the critical print size were presented in the NIDEK MP-1 microperimeter to determine the fixation PRL and its radial retinal eccentricity from the residual fovea in 29 subjects with bilateral CVL. Fixation stability was defined as the median bivariate contour ellipse area (BCEA) from 3 fixation assessments. A standard 10-2 grid (68 locations, 2° apart) was used to determine central retinal sensitivity for Goldmann size II test spots. Baseline and follow-up supra-threshold screening of the fixation PRL for localized sensitivity deficits was performed using high density (0.2° or 0.3° apart) 0 dB Goldmann size II test spots. Custom MATLAB code and a dual bootstrapping algorithm were used to register test-spot locations from the baseline and follow-up tests. Locations where the 0 dB test spots were not seen on either test were labeled as micro-scotomas (MSs).

RESULTS

Median BCEA correlated poorly with the radial eccentricity of the fixation PRL. Mean (±SD) sensitivity around the PRL from 10-2 testing was 4.93 ± 4.73 dB. The average percentage of MSs was similar for patients with AMD (25.4%), STGD (20.3%), and other etiologies of CVL (27.1%).

CONCLUSIONS

The fixation PRL in subjects with bilateral CVL frequently includes local regions of sensitivity loss.