Little is known about the way in which visual and proprioceptive cues for object location are combined. In order to investigate this, we measured the ability of participants to discriminate the location of a virtual target in the presence or absence of proprioceptive information. Participants viewed three virtual reference spheres using a head mounted display. The spheres lay on a circle of radius 14 cm and defined a reference plane that was located within easy reaching distance of the participant. In the proprioceptive condition, participants moved a hand-held pointer to the centre of each sphere in turn. When the pointer was fully inside the reference sphere, the sphere changed colour from red to green to indicate correct alignment. Once all three spheres had turned green, the participant clicked a mouse which made the spheres disappear to be replaced by a target sphere. A similar reaching movement was then performed to locate the centre of the target sphere. Once both the reference and target spheres had been located, the participant's task was to judge the depth of the target sphere relative to the plane defined by the previously-displayed reference spheres. For the vision-only condition, the procedure was similar but participants made their judgements without reaching towards either the reference or target spheres. Thresholds (defined as the standard deviation of the fitted cumulative Gaussian) were measured both with and without hand movements and also for two levels of transparency of the spheres. Results show that thresholds for combined proprioception and vision were significantly lower than for vision alone. In other words, proprioceptive input significantly improved depth discrimination despite vision being both sufficient to complete the task and necessary to define the proprioceptive cue.
展开▼
