Abstract
The Contrast Sensitivity Function (CSF) is altered in many visual disorders. However, measurement of the CSF is often overlooked for lack of time (e.g. Pelli & Bex, 2013). Here, we introduce an ultra-fast psychophysical method to measure the CSF. Observers were asked to fit a truncated log-parabola (Lesmes, et al., 2010) to an on-screen rendition of the Campbell-Robson chart (Campbell & Robson, 1968; Pelli, 1987) by adjusting its four parameters in just three clicks. The chart measured 14.4×8.1 deg of visual angle. Spatial frequencies spanned 0.16 to 40 cycles per degree (cpd) from left to right, and contrasts ranged from 0.005 to 0.25 from top to bottom. Participants first clicked on the perceived peak of the CSF, providing fmax and ymax estimates. Second, they adjusted the full-width at half-maximum of the right half of a log-parabola by moving the mouse sideways, providing a β estimate. Finally, participants adjusted the height of the truncation parameter (Δ) by moving the mouse vertically. Seventy-eight participants (age: 7-36) completed three runs. In addition, an orientation discrimination task (horizontal/vertical) was used with the same participants to validate the ultra-fast CSF method. Gabor patches measured 2 degrees of visual angle and ranged from 0.5 to 30 cpd in seven logarithmic steps. A threshold for each spatial frequency was measured in 336 trials with the QUEST method (Watson & Pelli, 1983). We obtained an overall root-mean-square error of 4.53 dB between the QUEST thresholds and the ultra-fast thresholds, which compares favorably to other methods, such as qCSF (Lesmes et al., 2010). The ultra-fast CSF method still needs to be improved, but it shows promise for use with children and adults with typical vision. Its use with patient populations will have to be validated.
Meeting abstract presented at VSS 2016