Purpose:
The contrast sensitivity function (CSF) characterizes functional vision, but its measurement can be time-consuming. The quick CSF method, (Lesmes et al, JOV, 2010), is a novel psychophysical method that combines Bayesian adaptive inference and a trial-to-trial information gain strategy, to efficiently estimate the full shape of the CSF. Our goal was to improve the clinical utility of the quick CSF method, apply it to assess spatial contrast sensitivity in a low vision population, and compare its results to standard clinical vision measures.
Methods:
For 21 patients referred to low vision rehabilitation, we measured the spatial contrast sensitivity function, from frequencies of .22 to 13.5 cycles per degree, in addition to Pelli-Robson contrast sensitivity, and logMAR acuity. The quick CSF algorithm was used to select the stimulus frequency and contrast presented on each trial. The full CSF was estimated from 15 quick CSF trials, from which two summary metrics were calculated: (1) the area under the log CSF (AULCSF), which provides a global contrast sensitivity measure, and (2) CSF acuity, a high frequency metric that defines the spatial frequency at which sensitivity = 2 (threshold = 50%).
Results:
The AULCSF estimates obtained with the quick CSF were correlated with Pelli-Robson sensitivity (r=.67), CSF acuity was correlated with logMAR acuity (r=-.69), but Pelli-Robson sensitivity and logMAR acuity were not correlated (r=-.14). AULCSF estimates obtained with 15 trials were the same as those obtained with 30 trials (mean difference = 2%; s.d.= 18%).
Conclusions:
This computerized, monitor-based test can sample stimuli more precisely and with greater flexibility than cards or charts, while automating the processing of patient responses. This study demonstrates that with as few as 15 trials, which lasts 1-2 minutes, the quick CSF provides a reasonably detailed assessment of visual function in people with low vision.
Keywords: contrast sensitivity • low vision • visual acuity