Parallel Specification of Visuomotor Feedback Gains during Bimanual Reaching to Independent Goals

Abstract

During goal-directed reaching, rapid visuomotor feedback processes enable the human motor system to quickly correct for errors in the trajectory of the hand that arise from motor noise and, in some cases, external perturbations. To date, these visuomotor responses, the gain of which is sensitive to features of the task and environment, have primarily been examined in the context of unimanual reaching movements toward a single target. However, many natural tasks involve moving both hands together, often to separate targets, such that errors can occur in parallel and at different spatial locations. Here, we examined the resource capacity of automatic visuomotor corrective mechanisms by comparing feedback gains during bimanual reaches, toward two targets, to feedback gains during unimanual reaches toward single targets. To investigate the sensitivity of the feedback gains and their relation to visual-spatial processing, we manipulated the widths of the targets and participants' gaze location. We found that the gain of corrective responses to cursor displacements, while strongly modulated by target width and gaze position, were only slightly reduced during bimanual control. Our results show that automatic visuomotor corrective mechanisms can efficiently operate in parallel across multiple spatial locations.

Keywords: motor control; online corrections; vision; visual perturbations.

Figure 1.

Experimental methods. (A) Experimental setup. Participants performed reaching movements in the horizontal plane while holding on to the handles of the robotic manipulandum. (B) Example bimanual nonchannel trial of experiment 1. Participants were instructed to fixate on the left or right reach target. Reach targets could be both narrow (in blue) or wide (in red). On a subset of trials, one of the hand cursors was visually displaced to the left or right after it passed under a visual occluder, requiring a correction of the movement trajectory. (C) Cursor paths from an example participant in response to a leftward (in green), zero (in gray), and rightward (in blue) shift of the left-hand cursor during bimanual reaching to narrow targets in nonchannel trials. (D) Same as C, but with reaching to wide targets (orange, leftward cursor shift; gray, no cursor shift; red, rightward cursor shift). (E) Example bimanual force channel trial of experiment 1. Participants’ hand movements were constrained along a straight line from start to target position, allowing us to measure the forces applied into the virtual wall of the channel (depicted by the black dashed lines). In cursor perturbation trials, the cursor automatically moved back to this line 250 ms after the perturbation. (F) Example bimanual force channel trial of experiment 2. Participants were instructed to fixate on a central fixation target. On a subset of trials, a single or both hand cursors were visually displaced to the left or right.

Figure 2.

Visuomotor responses in experiment 1. (A) Raw forces measured in channel trials in response to a leftward (–3 cm; in green and orange) and rightward (3 cm; in blue and red) displacement of the visual cursor during reaches to narrow (top row) and wide targets (bottom row) of the same example participant as in Fig. 1. (B) Difference in force responses to leftward and rightward cursor perturbations during reaches to narrow (in blue) and wide targets (in red), averaged across participants. Blue and red shaded areas indicate ±1 SEM. The black vertical line indicates the average onset of the corrective response (see Methods). The gray shaded area indicates the 180- to 230-ms interval across which the force differences were averaged to obtain a single measure of the strength of the response. (C) Mean force differences averaged across the 180- to 230-ms interval following the cursor perturbation. Error bars represent ±1 SEM. (D) Mean force differences at the nonperturbed hand in bimanual conditions.

Figure 3.

Visuomotor responses in experiment 2. Bars represent the mean force differences at a single hand averaged across the 180- to 230-ms interval following the cursor perturbation. Error bars represent ±1 SEM. Target sizes in parentheses indicate the size of the target of the other hand.