Abstract
Purpose: To determine if the temporal characteristics of externally added luminance noise can be manipulated to target selectively the magnocellular (MC) and parvocellular (PC) visual pathways. Methods: Letter contrast sensitivity (CS) was measured for three visually-normal subjects using briefly presented letters. CS was measured for letters of different size under four paradigms: 1) against an unchanging luminance pedestal ("steady-pedestal paradigm," known to target the MC pathway), 2) against a pulsed luminance pedestal ("pulsed-pedestal paradigm," known to target the PC pathway), 3) in a field of unchanging luminance checks (static noise), 4) in a field of randomly changing luminance checks (dynamic noise). CS was measured for letters that were high- and low-pass filtered with a range of filter cutoffs to quantify the object frequency information (cycles-per-letter) mediating letter identification, which was used as an index of the pathway mediating CS. A follow-up experiment determined the effect of target duration on the object frequencies mediating letter identification in static and dynamic noise. Results: Object frequency increased as letter size increased for each paradigm, consistent with previous results showing that subjects use different information to identify letters at different sizes. ANOVA indicated that the object frequencies measured under the static noise and steady-pedestal paradigms did not differ significantly (p > 0.05), but differed considerably from those measured under the dynamic noise (both p < 0.05) and pulsed-pedestal (both p < 0.05) paradigms. The object frequencies mediating letter identification were independent of target duration in static and dynamic noise. Conclusions: The spatiotemporal characteristics of noise can be manipulated to target the MC (static noise) and PC (dynamic noise) pathways. The ability to target these pathways at long stimulus durations in a letter identification task has potential importance in the design of future clinical CS tests.
Meeting abstract presented at VSS 2016