Flexible planning of corrective responses for double-step reduction in the number of potential targets

Ryoji Onagawa,Kazutoshi Kudo

doi:10.1038/s41598-021-86325-9

Abstract

Humans are often required to plan/execute movements in the presence of multiple motor targets simultaneously. Under such situations, it is widely confirmed that humans frequently initiate movements towards the weighted average direction of distinct motor plans toward each potential target. However, in situations where the potential targets change in a step-by-step manner, the strategy to proceed towards the weighted average direction at each time could be sub-optimal in light of the costs of the corrective response. Herein, we tested the sensorimotor strategy followed during a step-by-step reduction of potential goals. To test the hypothesis, we compared the corrective responses when the number of targets went from three to two, and when the number of targets went from three to one at the same time. As the results, weak corrections were confirmed when the number of targets was reduced from three to two. Moreover, the corrective responses when the number of targets went from three to two was smaller than the average behavior estimated from the corrective responses when the number of targets went from three to one at the same time. This pattern of corrective responses reflects the suppression of unnecessary corrections that generate noise and cost to the control system. These results suggest that the corrective responses are flexibly modulated depending on the necessity, and cannot be explained by weighted average behavior.

Highlights

Humans are often required to plan/execute movements in the presence of multiple motor targets simultaneously
An individual is simultaneously presented with multiple potential targets and is required to launch a movement before knowing the final target location, which is presented after movement onset[1,2,3,10] or after passing through a given spatial threshold[4,6,8]
There was no significant difference between the models (Visual-averaging vs. Motor-averaging; t[13] = 1.342, d = 0.421, pbonf = 0.382). These results suggest that the corrective responses between the 1st and 2nd threshold in double step trials deviate from the averaging behavior, and the observed corrective responses are weaker than the averaging behavior

Summary

Introduction

Humans are often required to plan/execute movements in the presence of multiple motor targets simultaneously. If motor planning and execution for a step-by-step target revealing follows the weighted average behavior, as observed in previous studies, an individual would have to correct their in-flight movement according to changes in potential targets In such situations, correcting an ongoing movement every time that the number of potential goals is decreased can be a suboptimal control strategy, since the correction of in-flight movement demands costs and incorporates noise into the sensorimotor system[11]. In the single-step trials, a final target was determined when the cursor moved 2 cm away from the start position (first threshold, yellow frame circles) These trials were set up to estimate corrective responses in accordance with the averaging behavior. An ongoing movement is at each phase of target information updates (i.e., decreases in the number of potential targets) could be effective for clarifying the sensorimotor control policy under uncertainty about task goals

Methods

Results

Conclusion