Abstract
Previous research has shown that the brain uses sophisticated representations of both sensory and motor uncertainty to optimize task performance. Knowledge of sensory uncertainty is used to near-optimally combine sensory information, while knowledge of motor precision can be used to plan movements that optimize expected gain. We conducted an experiment to examine how the brain combines its knowledge of sensory certainty and motor precision to perform a task that depended on both. Participants reached to the center of a 2D, isotropic, Gaussian distribution of dots within a 1200 ms time-limit. Participants traded off the amount of sensory and motor variability by selecting the reach initiation time. Specifically, before reach initiation the number of visible dots increased with time, thereby improving the certainty of the participant's estimate of the true target location. Once the reach was initiated, no new dots appeared and the remaining trial time was available for the reach. However, speed-accuracy trade-offs in motor control make early reaches (much remaining time) precise and late reaches (little remaining time) imprecise. Based on each participant's visual-only and motor-only target acquisition performance, we computed an “ideal reacher” that selected a reach initiation time with minimal predicted reach endpoint errors from the true target location. We found that people selected reach initiation times similar to those of the “ideal reacher”, even in the absence of performance feedback. Further, when we manipulated the overall quality of the visual target stimuli, people shifted their reach initiation timing much like the “ideal reacher”. We conclude that the brain jointly represents time-dependent sensory and motor variability relationships, and uses this knowledge to maximize performance in sensorimotor tasks.
Supported by NIH EY015261-01.