Abstract
Letter optotypes are often used in the clinical evaluation of contrast sensitivity. A common assumption is that letters, which are broadband in spatial frequency content, provide the same information about visual function as grating stimuli, which are spatial frequency delimited. We evaluated this assumption using Sloan letter identification and grating orientation tasks to obtain contrast sensitivity functions (CSFs) under conditions designed to favor either the magnocellular (MC) or parvocellular (PC) pathway. CSFs were obtained from two visually normal observers using two testing paradigms: a steady-pedestal paradigm (brief stimulus presentation on a steady luminance pedestal) and a pulsed-pedestal paradigm (simultaneous brief presentation of the stimulus and luminance pedestal), thought to favor the MC and PC pathways, respectively (Leonova, et al., 2003). Additionally, CSFs were obtained under these conditions for Sloan letters that were restricted in spatial frequency content by using a cosine log filter (Peli, 1990) with a peak frequency of 2.5 c/letter. CSFs for Gabor patches were low-pass for the steady-pedestal paradigm and band-pass for the pulsed-pedestal paradigm, confirming previous findings (Leonova, et al., 2003). In comparison, CSFs for Sloan letters were low-pass in both the steady- and pulsed-pedestal paradigms. Filtered letters and Gabor patches produced similar CSFs under the steady-pedestal paradigm. For the pulsed-pedestal paradigm, however, the CSF for filtered letters was not as band-pass as that for Gabor patches. Thus, Gabor patches, letters, and filtered letters yielded similar CSFs under conditions favoring the MC pathway, but they produced different CSFs under conditions emphasizing the PC pathway, indicating that equivalence of the two optotypes depends on the visual pathway mediating sensitivity.
Supported by NIH Grant EY08301.