Abstract
Purpose :
Manual Visual Acuity (VA) testing is an important but relatively time-consuming test required for all patients in the ophthalmology outpatient clinic. It also requires a large number of trained manpower, which could be better deployed elsewhere. The purpose of this study is to develop a reliable automated Augmented Reality (AR) system for VA testing to improve workplace efficiency.
Methods :
In total, 60 patients from National University Hospital, Singapore were enrolled in a prospective observational study. Each subject underwent VA testing of both eyes with the manual Snellen chart and the AR headset. VA results without pinhole occlusion were analysed with Bland-Altman analysis and intraclass correlation. Other outcome measures included manpower and time, total cost, space needed and patient feedback on ease of use and comfort.
Results :
We included 53 out of 60 subjects for the final analysis. There was a clinically insignificant mean difference of 0.05 logMAR (p<0.05), with the AR headset underestimating vision determined by manual testing on the Snellen chart. The 95% limits of agreement were ±0.33 logMAR and intraclass correlation coefficient demonstrated good inter-rater reliability (r=0.77, p<0.05). Automated VA testing increased labour productivity by 157%, reduced direct machinery costs by 67.2% and decreased space consumption by 89.2%. The AR headset was rated mildly uncomfortable but easy to use, with no correlation determined between ease of use and age (r=-0.10, p>0.05) or education levels (r=-0.02, p>0.05).
Conclusions :
This study supports the validity of the AR headset as an alternative measurement method of VA to increase workplace efficiency. The AR headset is user-friendly for all, increasing its potential to be used as a community screening tool, and as a platform for other forms of visual testing.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.