Abstract
Human adults can improve their perceptual precision by averaging multiple sensory estimates (e.g., vision and touch, or depth via multiple visual cues). Strikingly, it is not until ~age 10 years that children combine multiple cues to reduce sensory uncertainty. Why perceptual benefits from sensory fusion emerge so late in childhood is currently unclear. It is possible that the neural computations by which single cues are fused develop slowly in childhood. For example due an extended process of learning how cues relate to each other (Gori et al., 2008, Curr Biol). An alternative possibility is that the required neural computations are in place, but that children are still learning how to use the resulting output adaptively to improve perception. To understand the development of sensory fusion at the level of neural representation, we combined psychophysics, retinotopic mapping, and pattern classification fMRI in a developmental study with children, using methods that were first validated with adults (Ban et al., 2012, Nat. Neurosci). In a sample of 100 children aged 6-13 years we measured perceptual depth discrimination thresholds and established that adult-like fusion of motion-parallax and disparity cues develops between the 10th and 11th year. In a subset of 27 children aged 8-13 years, we then investigated the changes in neural representations that occurred while adult-like cue integration emerged perceptually. There was clear evidence of depth cue integration in visual area V3B in older children (>10.5 years), a region that also combines visual depth cues in the mature brain. In contrast, we found no evidence for sensory fusion in the visual cortex of younger children, who also did not display perceptual cue combination in the behavioral psychophysics task (< 10.5 years). This suggests that the neural processes giving rise to fused depth representations in V3B are still developing at these younger ages.
Meeting abstract presented at VSS 2015