Abstract
Areas of the putative visual cortex are involved in combining multisensory information about object shape. In particular, the lateral occipital complex (LOC) is a known region involved in visuohaptic object recognition. Relative to adult research, there have been few studies investigating the development of crossmodal perception of visual and haptic information, and fewer still exploring the neural substrates of this ability. In the present study, BOLD fMRI was measured in children aged 7 to 8.5 years and in adults during intramodal (visual-to-visual, haptic-to-haptic) and crossmodal (visual-to-haptic, haptic-to-visual) delayed match-to-sample recognition tasks. In both children and adults, results indicated that the ventral occipitotemporal cortex followed a medial to lateral organization in which there was a visual to haptic bias in the mapping of sensory modalities used to process shape information. These areas were adjacent to (and on either side of) an overlapping bimodal region within the LOC (corresponding to LOtv). Additionally, results showed a crossmodal enhancement effect in which crossmodal matching produced greater activation than intramodal matching in the overlapping bimodal LOC region (LOtv), but not in the medial visual-preferring or lateral haptic-preferring areas of the LOC, suggesting that this bimodal region was sensitive to sensory changes. Finally, although children and adults did not differ qualitatively in the overall patterns of neural activity, children produced more widespread, but lower intensity, activation than adults. This suggests that fundamental neural support for visuohaptic object processing is present by 7 years, but requires fine-tuning to be adult-like. Finding a haptic-preferring region in the LOC suggests that the multisensory signals integrated in the LOtv likely arrive via neighboring modality-biased regions within the LOC itself. We hypothesize that these modality-biased regions transform somatosensory and visual signals to facilitate multisensory integration in the LOtv. A future step should be identifying the exact nature of those transformations.
Meeting abstract presented at VSS 2015