Purchase this article with an account.
Haiyan Zheng, Fang Hou, Luis A Lesmes, Zhong-Lin Lu; Validation of the quick CSF method with digit stimuli. Invest. Ophthalmol. Vis. Sci. 2018;59(9):1275.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The contrast sensitivity function (CSF), typically measured with sinewave gratings, provides a comprehensive assessment of functional vision. Recently, we implemented the quick CSF method (Lesmes, et al, 2010) with a 10 alternative forced-choice (10AFC) paradigm using filtered Sloan letters and greatly improved the efficiency of the test (Hou, et al, 2015). However, people in non-English speaking countries who are unfamiliar with English letters can't benefit from the more efficient quick CSF test. To address this problem, we implemented the quick CSF method using a new set of digit stimuli and conducted two psychophysics experiments to validate the new procedure.
We created a set of Sloan digits (0 ~ 9) based on the principles used to design Sloan letters. The digits were then filtered with a raised cosine filter (Chung, Legge et al. 2002), rescaled to different sizes to cover a range of spatial frequencies (0.5 ~ 16 cpd), and used as stimuli in a 10AFC digit identification task. We measured (1) the CSFs of five normal participants using both the 10AFC digit identification and 2AFC grating orientation (± 45°) identification tasks with the Psi method (Kontsevich and Tyler 1999), and (2) the CSFs of five normal participants using both the quick CSF and Psi methods with the 10AFC digit identification task.
We found that the digit CSF almost perfectly matched the grating CSF after controlling for stimulus and task differences. The root mean square error (RMSE) between the digit and grating CSFs was 0.049 ± 0.027 log units. With 150 trials, the 68.2% half width credible interval (HWCI) of digit CSF with the Psi method was 0.057 ± 0.005 log units, smaller than that of the grating CSF (0.127 ± 0.022 log units). We also found that, with the digit identification task, the CSFs obtained with the quick CSF method matched very well with those obtained with the Psi method (RMSE = 0.053 ± 0.019 log units). With 150 trials, the average 68.2% HWCI of the CSFs obtained with the quick CSF procedure was 0.037 ± 0.009 log units, which is significantly lower than that from the Psi method (0.052 ± 0.004 log units, p < 0.001).
The digit CSF provides essentially the same measure as the grating CSF. The quick CSF method combined with the digit stimuli could serve as a precise and efficient test instrument for CSF, especially for people who are unfamiliar with English letters.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.
This PDF is available to Subscribers Only