アイテム詳細

要旨

Using a visuo-vestibular adaptation paradigm, we measured how the richness/realism of the visual scene affects sensorimotor learning. Participants seating on a motorized wheelchair had to maintain their stretched arm fixed in space despite whole-body rotations. After a test session in which body rotations occurred without visual information, participants were exposed to a visuo-vestibular adaptation, and then tested again in a test session identical to the first one. In the adaptation phase, biased visual information about motion amplitude was provided during body rotations via a head-mounted display, so that participants learned a biased visuo-vestibular mapping. Four different types of adaptation were used (with different participants). Specifically, we manipulated orthogonally the visual information relative to the own body (avatar arm vs sphere indicating fingertip position) and that relative to the surrounding environment (virtual room vs sphere). All four types of visual environments gave rise to a significant adaptation (p<0.05). Interestingly, the amplitude of the adaptation (both relative and absolute) was significantly larger only when the room and the arm were realistically displayed. These results suggest that for a sensorimotor learning facilitation to occur, both body-related and environment-related visual information must be realistic.