Abstract
Vision undergoes rapid and dramatic development early in life. These changes have been investigated extensively with behavioral techniques, but there is a dearth of direct neural studies of the developing visual system. This stands in contrast to the adult visual system where a great deal has been learned over two decades by using fMRI, including studies of retinotopy, attention, and perceptual learning. The lack of fMRI studies in infants and toddlers reflects the technical complexity of performing such experiments, including head motion, attention span and fussiness, inability to understand or follow instructions, uncomfortable apparatus, and adult-centric analysis approaches. When infants have been scanned successfully, it has typically been during sleep or sedation, preventing the kinds of cognitive tasks that would allow for comparisons to adults and for benefitting from the tremendous advances in adult cognitive neuroscience. Here we report on our attempts to re-imagine task-based fMRI procedures for early developmental populations. The resulting advances include: projector-based panoramic visual presentation, flexible experiment transitions, within-volume motion artefact detection and functional alignment. With these procedures, we have now scanned 18 participants under the age of 36 months and retained a high percentage of the functional data collected from each participant. These procedures also produce reliable evoked BOLD activity in visual cortex in response to a wide variety of visual stimuli. With these tools in hand, we are now investigating the retinotopic organization, attention networks, and plasticity of the infant visual cortex. Each of these studies presents its own unique challenges, such as the difficulty ensuring fixation during retinotopy. Nevertheless, the hope is to complement the wealth of work on visual development with a detailed understanding of early brain function.
Meeting abstract presented at VSS 2018