Purchase this article with an account.
Y.-G. He, A. Chhetri, M. Bartlett, G. Mitzel, K. Zhang, B. Krenik, Y.-Z. Wang; A Handheld-Based Shape Discrimination Testing Method for Visual Function Monitoring. Invest. Ophthalmol. Vis. Sci. 2010;51(13):1834. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
With recent advances in treatments for age-related macular degeneration (AMD) and diabetic retinopathy, there is an increasing demand for remote visual function testing to detect early disease changes for timely intervention. Previous studies using PC-based testing systems showed that threshold for detecting shape distortion is significantly elevated in patients with AMD (Wang et al., 2002 IOVS, Vol. 43, p.2055). Here we evaluated a new handheld shape discrimination testing system designed for remote visual function monitoring.
The handheld shape discrimination test was implemented on the iPhone platform using a spatial 3-alternative forced-choice (3AFC) staircase paradigm. Visual stimuli were deformed circular shapes generated by modulating the radius sinusoidally. In each trial, subjects indicated by touch input which of 3 circular shapes on the iPhone screen was deformed. The PC-based test was implemented with a temporal 2AFC paradigm. A maximum likelihood fitting procedure was used to estimate radial modulation threshold. 55 normal volunteers (22 to 70 years, mean age 42±16SD, mean VA 20/20) participated in the study. Tests were conducted twice with the dominant eye. 22 eyes with retinal diseases affecting macula (VA 20/25 to 20/80 with mean of 20/40) were also tested with the handheld device.
For normal volunteers, Bland-Altman analyses of test-retest repeatability revealed that both handheld and PC testing protocols had comparable variability. The mean % modulation thresholds were 0.30±0.1SD and 0.26±0.1SD for handheld and PC protocols, respectively. While the handheld protocol generated a slightly higher estimate of threshold than PC protocol, this bias was not significant since the 95% confidence interval of mean difference included zero. The limits of agreement between the two protocols were in the order of mean threshold. Normal aging had little effect on the handheld shape discrimination performance (r=0.001; p>0.994). For the eyes with retinal diseases, the mean threshold measured by the handheld protocol was 1.19%±0.87SD, about a factor of 4 higher than normal (t=7.17; p<0.001).
The handheld shape discrimination testing method is readily accessible, intuitive to use, low-cost, comparable to PC-based testing methods, and sensitive to macular diseases. It potentially provides patients with AMD and diabetic retinopathy a new tool to monitor their visual function changes outside of the clinical setting.
This PDF is available to Subscribers Only