Reduced Brain Complexity During the Execution of a Sequential Arm Movement Following an Augmented-Feedback-Based Practice: A Multiscale Entropy Analysis.

Motor learning is accompanied by adaptive changes in neural processing, yet the extent to which feedback-based practice alters the complexity of cortical activity during movement execution remains unclear. The present study examined learning-associated changes in brain complexity during the execution of a sequential arm movement following augmented-feedback-based practice. Thirty-eight healthy young adults practiced a sequential arm movement with augmented feedback across five sessions. Cortical activity was recorded with electroencephalography during the first and last practice sessions. Multiscale entropy was used to quantify brain complexity across 50 time scales. Behavioral measures indicated improved movement accuracy and reduced dual-task costs across practice. At the cortical level, a significant global decrease in entropy was observed from the first to the last practice session. Further analyses revealed that the differences were confined to coarse temporal window and midline frontal electrodes Fz and FCz. In addition, changes in coarse-scale entropy at FCz were negatively correlated with reductions in absolute error. Together, the present findings suggest that augmented-feedback-based practice is accompanied by a refinement of cortical complexity during movement execution, particularly in large-scale frontal dynamics. Multiscale entropy may therefore provide a sensitive framework for capturing learning-related neural adaptations that complement conventional electroencephalographic measures and advance the understanding of motor learning at the systems level.