Abstract
Normally sighted adults can combine visual and auditory cues to location optimally by weighting each according to its reliability (Alais & Burr, Curr Biol 2004). Here we investigated the extent to which observers take into account natural variations in sensory precision across the visual field. This will provide a baseline for studying cue combination in patients with central or peripheral vision loss. Visual localization thresholds deteriorate towards the periphery, where sampling precision is limited by the sparser distribution of retinal photoreceptors. Freefield auditory localisation thresholds also deteriorate towards the periphery, but usually at a slower rate to vision. We assessed whether normally sighted adults continued to combine visual and auditory signals optimally when presented in (i) central (1-17°) and (ii) peripheral (36-53°) space. Eleven normally sighted adults (18-42yrs) were presented with blocks of audio-only (100ms white noise burst), visual-only (25ms flash from 5-50 LEDs, spanning 25 degrees) and, audio-and-visual-combined stimuli, either centrally (N=6) or peripherally (N=5). On each trial, participants indicated whether the standard (central: 1°; peripheral: 36°) or comparison (+0-16° relative to standard) was further right. Psychometric functions were fitted to these responses. Congruent bimodal trials measured improvements in sensitivity when two cues were available together (via slope), while bimodal trials with ±1° conflicts measured cue weighting (via PSE). Initial results show higher single-cue thresholds in the periphery, but improved performance with bimodal stimuli in both central and peripheral conditions. This shows that normally sighted adults combine visual and auditory estimates to improve their precision in both central and peripheral space. These data will provide a baseline for understanding whether and how patients with Stargardt’s disease (central vision loss) or retinitis pigmentosa (peripheral vision loss) learn to re-weight sensory signals following visual impairment.
Meeting abstract presented at VSS 2015