Affiliations: University of Toronto, Toronto, Ontario, Canada
Correspondence:
[*]
Address for correspondence: Luc Tremblay, 55 Harbord St., University of Toronto, Toronto, Ontario, M5S 2W6, Canada. E-mail: [email protected]
Abstract: Background:Robotic guidance has been shown to facilitate motor skill acquisition, through altered sensorimotor control, in neurologically impaired and healthy populations. Objective:To determine if robot-guided practice and online visual feedback availability primarily influences movement planning or online control mechanisms. Methods:In this two-experiment study, participants first performed a pre-test involving reaches with or without vision, to obtain baseline measures. In both experiments, participants then underwent an acquisition phase where they either actively followed robot-guided trajectories or trained unassisted. Only in the second experiment, robot-guided or unassisted acquisition was performed either with or without online vision. Following acquisition, all participants completed a post-test that was the same as the pre-test. Planning and online control mechanisms were assessed through endpoint error and kinematic analyses. Results:The robot-guided and unassisted groups generally exhibited comparable changes in endpoint accuracy and precision. Kinematic analyses revealed that only participants who practiced with the robot exhibited significantly reduced the proportion of movement time spent during the limb deceleration phase (i.e., time after peak velocity). This was true regardless of online visual feedback availability during training. Conclusion:The influence of robot-assisted motor skill acquisition is best explained by improved motor planning processes.
Keywords: Motor control, robot, upper-limb, assistance, rehabilitation
