Response to perturbations of visual feedback during reaching movements

Abstract

Purpose: Being able to see one's hand while reaching for an object is known to improve accuracy and speed. However, there remains some question as to how much this information contributes, and whether it is used for continuous control. A standard technique for investigating a dynamic control process is to measure its perturbation response. We applied this technique to study visual control of reaching by testing perturbations of visual feedback during movements in a virtual environment. Method: The subjects' task was to reach and touch a succession of targets in a virtual environment. Visual targets were aligned with a tabletop, which provided consistent haptic feedback. Images presented by a monitor were viewed through a horizontal mirror, using shutter glasses to present left and right stereo views. The mirror occluded subjects' view of their hand. Visual feedback was provided by a sphere representing a subject's fingertip ( a virtual finger), which was updated in real-time using an optical tracking system and markers on a subjects' finger. In a baseline condition, the virtual finger accurately tracked the location of the unseen finger throughout a trial. In perturbed trials, a small discrepancy (2cm) between the actual and virtual finger was added at variable points during a movement. Perturbation onsets were gradual to be less noticeable. Results: Across perturbed trials, hand trajectories typically showed an adjustment at the end of a movement to correct error in the initial reach. When perturbations were early in the movement, there were also corrective adjustments before approaching the target. Detectable responses were observed as early as 40% of the way along the movement path, with delays of around 200ms. Conclusion: Visual feedback from the hand is used not only to control fine adjustments at the end of a reach, but also to guide movement during the initial phase of reaching.