Purchase this article with an account.
Christopher W. Tyler, Lenny L. Kontsevich; Encoding perceived depth. Journal of Vision 2006;6(6):347. doi: https://doi.org/10.1167/6.6.347.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose. Our visual experience operates at the level of integrated perceptual cues, which are derived from an array of cue modalities. A classical example is perceived depth, which is derived from a variety of monocular, binocular and kinesthetic depth cues. We ask at what level in the visual processing stream perceived depth is encoded. FMRI methods are used to identify the locus in visual cortex where matched depth percepts from stereoscopic, motion, texture and shading cues generate matching signal strengths.
Methods. The stimuli consisted of arrays of bulges formed from fourth-power (flat-topped) Gaussian-like functions. The observer's task was to match the depth of the bulges from various depth cues in the one hemifield to the fixed depth from the disparity-defined bulges on the other hemifield. Cortical activations to a) depth modulation, b) cue modulation with invariant depth, were recorded by fMRI in a 3T scanner with 40 planes at 3 × 3 × 3 mm3 resolution at 3 s TR.
Results. The main site of activation by the depth modulation was in the ODS(KO) region that responds to depth structure, including kinetic contour structure (Tyler et al., 2005, NeuroImage) or adjacent retinotopic areas V3A/B. This region typically showed well-matched activations for all four depth cues, with no significant response for alternation between cues that generated matched depth percepts.
Conclusion. A region in the lateral occipital cortex of each observer was identified that exhibited the requisite behavior for the site of the generic depth representation from multimodal visual cues.
This PDF is available to Subscribers Only