Abstract
Purpose :
The Pelli-Robson (PR) chart measures contrast sensitivity by presenting letters of decreasing contrast to determine a patient’s threshold contrast. The purpose of this study is to demonstrate the comparability of a contrast sensitivity application developed for a wearable augmented reality platform to the PR contrast sensitivity test.
Methods :
Fourteen participants with and without contrast sensitivity deficits were recruited to complete the Heru contrast sensitivity (HCS) application and PR chart test. Randomization determined which eye, OD or OS, was tested for each participant. The PR chart test was administered to participants with full refractive correction. The Heru contrast sensitivity application (HCS) was installed on a Magic Leap 1, size 2 headset, and uses a tumbling E presentation on a light background with a shrinking staircase thresholding strategy. Trial lenses were inserted into the headset to correct using the distance spherical equivalent. To improve our sampling of the lower end of the contrast sensitivity range, participants repeated HCS and PR tests using neutral density filters with 2x and 3x attenuation (ND2 and ND3). Contrast sensitivity values were compared between the two tests using Bland Altman analysis.
Results :
Our study included 14 eyes, 79% female, with an age range of 24-80 years old (mean age of 49 years). Fourteen eyes from 14 subjects were independently tested with both methods using no filter and using a neutral density filter with an attenuation factor of 2 (ND2). Eleven of those eyes were also tested with a neutral density filter with an attenuation factor of 3, but two of the results were invalid because the subject could not see the PR chart. The overall mean difference was -0.08 dB with limits of agreement from -0.7 to 0.75 dB, showing significant variability. The mean difference and total number of outliers increased with the use of a neutral density filter (-0.03 with no filter, -0.14 with ND2, and -0.08 with ND3 (Figure 1), with outliers of 7%, 21% and 22% respectively.
Conclusions :
It is possible to implement a contrast sensitivity test on a wearable device, and for the tested population the mean log contrast sensitivity matches well on average to the Pelli-Robson.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.