Purchase this article with an account.
R. A. Schuchard, M. Uslan, T. Wilson, W. LiKamWa; Evaluating the Barten SQRI for Predicting Small Visual Display Character Recognition. Invest. Ophthalmol. Vis. Sci. 2009;50(13):4725.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The prevalence of small, low contrast, visual displays (e.g., cell phones and PDAs) is increasing. These devices can obviously pose a strong text-recognition challenge for the person with vision loss. This study investigated the relationship between successful recognition of a single digit/character and a measure that combines the contrast sensitivity function of a person with vision loss with an image quality metric (Barten SQRI).
34 subjects with central vision loss performed: 1) near visual acuity with 100% contrast, 5% contrast, & dark stimuli (ETDRS & SKILL); 2) contrast sensitivity (Peli-Robson and CSF from CRT square wave stimuli); 3) PRL/fovea and scotoma characteristics (SLO); 4) reading performance (SKRead); and 5) threshold contrast with 2-up/1-down adaptive staircase for 0.25, 0.50, 1.0, 2.0, 4.0, and 8.0 degree digits/characters (CRT). The Barten SQRI was computed from each subject’s contrast sensitivity function (CSF) and the image display’s modulation transfer function.
Subjects had median (range) 100% near visual acuity of 0.42 logMAR (0.48 to 1.50), 5% visual acuity of 1.06 logMAR (0.54 to 1.54); dark acuity of 1.22 logMAR (0.84 to 1.54); contrast sensitivity of 1.20 logcontrast (0.30 to 1.55); critical print size of 1.05 logMAR (0.10 to 1.30); and maximum reading rate of 36.2 words/minute (0 to 110.3; not all subjects could read the largest SKRead text). All subjects had central visual field deficits. All subjects had decreasing visual performance in single digit recognition with decreasing digit size and in decreased visual acuity with lower contrast targets (5% and Dark visual acuity). The Barten SQRI significantly and strongly predicted the required contrast for single digit recognition (r2=0.81; p<0.01) compared to contrast sensitivity alone (r2=0.48; p<0.01). Visual acuity (100%, 5%, or dark) did not as strongly predict single digit recognition accuracy (r2=0.28 to 0.47; p<0.01) or SKRead critical print size (r2=0.33 to 0.53; p<0.01). Analysis of relationships that include subject characteristics (e.g., using fovea/PRL to read characters, central visual field characteristics, and ability to read SKRead text at the largest 1.3 logMAR text size) gives further understanding of the relationship between character recognition and the Barten SQRI.
Combined measures of individual contrast sensitivity with image display characteristics may be better outcome measures of the effect of vision loss on performance with small displays than standard visual function measures.
This PDF is available to Subscribers Only