Annabel Wing-Yan Fan¹ (), Alex S Baldwin¹; ¹McGill

Although contrast detection models can also account for stimulus identification performance, neuroimaging and behavioural studies have found partially distinct mechanisms underlying the tasks. We examined the effect of stimulus duration on detection or identification tasks with filtered dynamic noise targets and the interaction between the temporal spectrum of the dynamic noise and the benefit of temporal integration. Detection and identification (left- or right-oblique target) thresholds were measured for one-dimensional spatially pink noise presented for 50, 200, or 1000 ms in a circular envelope. The noise was either static or dynamic. In the dynamic case, it had either a white or pink temporal spectrum. In the detection task, a noise target appeared in either the first or second interval (with the other interval blank). In the discrimination task, both intervals contained a noise stimulus, one being left- and the other right-oblique. The observer reported which interval contained the left-oblique target. Ten participants with healthy vision performed all experiment conditions. Thresholds were analysed with a three-way rm-ANOVA (task x temporal noise x noise duration). Threshold RMS contrast significantly varied depending on the task (F1,9 = 16.7, p < 0.01), temporal spectrum (F2,18 = 21.0, p < 0.0001), and duration (F2,18 = 246.1, p < 0.0001). Identification thresholds were lower than detection thresholds. Thresholds also decreased with stimulus duration. Comparing between constant and dynamic noise conditions, thresholds were highest for stimuli with white temporal spectra, followed by pink and constant noise. We find that temporal integration improves sensitivity with increasing stimulus duration, though for white noise it is likely that linear combination over short durations reduces the effective contrast. Although we find greater sensitivity in the identification task, we do not find any differences in the effects of the temporal aspects of the stimuli compared to simple detection.

Acknowledgements: This work is funded by the Vision Health Research Network of the FRQS, FRQS, Fiera Capital Awards for Diversity, Equity, and Inclusion in Health Care, and an NSERC Discovery Grant.