Automated Vision Test Development and Validation
Abstract
The Optec Vision Test, originally produced in 1951 as the Armed Forces Vision Tester, is the sole device used to qualify individuals for Air Force flight duties. Although the external appearance of the device has changed since its first inception, the design of the slides used to present the visual stimuli is exactly the same as those originally produced. The goals of this effort were to evaluate proof of concept that the vision screening tests currently administered using the Optec Vision Test could be transitioned to a computer-based, automated system; to produce software for desktop displays; and to evaluate features such as user interfaces, threshold algorithms, validity of results, and screening techniques that could minimize testing time. This was a prospective study consisting of 27 individuals aged 18-40, a range that represents the Air Force flying population. There was no stated requirement for gender and there were no exclusion criteria related to visual status as subjects with both normal and non-normal visual skills were desired. Automated, computer-based vision tests were developed to assess high and low (5 Michelson) contrast visual acuity, letter contrast sensitivity at 20/25 and 20/50 acuity levels, color contrast sensitivity, and stereoacuity. The current effort demonstrates that automated vision tests produce reliable results, with coefficients of determination for repeated testing above 0.80 for many of the tasks. However, to achieve this high level of reproducibility and reduce the standard error of the threshold estimate to near asymptotic levels, 30 or more trials are often required. Successful implementation of automated (or any) vision testing in an aerospace medicine clinic requires methods of determining visual status quickly, but with high accuracy. We found that 100% sensitivity could be achieved using a fast screening method, however, at the cost of performing full threshold testing on over 30% of normal subjects.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 2016
- Accession Number
- AD1021902
Entities
People
- Alex Van Atta
- Dan Lamothe
- Darrell Rousse
- James Gaska
- Marc Winterbottom
- Steve Wright
- Steven Hadley
Organizations
- United States Air Force School of Aerospace Medicine