Visuomotor adaptation influences perceptual decision-making

Abstract

Accurately predicting the sensory consequences of movement commands using a “forward model” and combining them with actual feedback relayed via sensory systems is thought to yield better perceptual estimates of limb and environmental state than those possible with sensory feedback alone. However, sensory predictions are accurate only if the forward model accurately reflects the properties of the body, the environment and the relationship between the two. This accuracy is maintained via adaptation, in which the forward model is updated based on differences between the predicted and actual sensory consequences of action. Here we investigate how updating sensory predictions through motor adaptation influences perceptual processing and subsequent decisions based on the outcome of those perceptual processes. Twelve young, healthy subjects adapted to a 10 degree visuomotor rotation while using a stylus to trace a trajectory displayed on a computer screen in front of them. Subjects showed clear within- and between-trace adaptation such that the applied rotation was effectively canceled with training. Subsequently, subjects were asked to report the color of a target that moved on the screen among several distractors while they also moved their hand, which was not directly visible. Target motion was either random, matched to that of the hand, or along a path that was rotated by 10 degrees relative to hand motion. Surprisingly, subjects were much more accurate in reporting target color if it moved along the rotated path compared to veridical or random motion. When between-trace adaptation was prevented by means of a secondary cognitive task, subjects showed greater accuracy in reporting the color of a target moving along the rotated path for a few initial trials, reflecting learning that occurred within the last trace of the adaptation block. However, with time subjects shifted towards more accurately reporting the color of the target moving along the actual hand path. These results demonstrate the strong influence of adaptation and updating of predictive mechanisms that adaptation entails on perception and decisions based on perceptual processing. Perceptual decisions can be clearly modified by manipulating the relationship between movement commands and their sensory consequences, and the degree of modification may depend on how strongly the adapted state is reinforced.