Accelerating motor adaptation by virtual reality based modulation of error memories

• Published in: IEEE International Conference on Rehabilitation Robotics pp. 623 - 629
• Main Authors: Ballester, Belen Rubio, Serra Oliva, Laura, Duff, Armin, Verschure, Paul
• Format: Conference Proceeding Journal Article
• Language: English
• Published: IEEE 01.08.2015
• Subjects: Acceleration; Adaptation; Conferences; error feedback; Errors; History; memory of errors; Modulation; Motor adaptation; Motors; Noise; Patients; Perturbation; rehabilitation; Sensitivity; Virtual reality; Visual; Visualization
• ISSN: 1945-7898; 1945-7901
• DOI: 10.1109/ICORR.2015.7281270

When exposed to visual perturbations, the motor system rapidly learns to reduce errors through adaptation of future motor commands. However, in cerebellar and stroke patients with proprioceptive impairments, motor adaptation rates are significantly slower. A recent study suggests that adaptation rates may be modulated by the stability of perturbations and the history of errors. We hypothesize that reducing the visual magnitude of directional error in a reaching task should increase the speed of motor adaptation to a perturbation. We developed a method for the modulation of error magnitudes in virtual reality (VR), and we conducted two experiments exploring its effects on adaptation. 34 healthy subjects used a chair-mounted arm support to perform reaching movements towards a target while learning to compensate for visuomotor rotations. We observed that diminishing the visually perceived variability of visual directional error through VR accelerates adaptation. This effect showed persistence into trials when virtual error reduction was no longer present confirming the modulatory role of error memory. These findings support the hypothesis that the brain may keep a history of recent errors and differentially adjusts learning rates relative to the frequency of occurrence of errors. We predict that exposure to a distribution of visual motor errors with low variability and centered around task-relevant error values accelerates motor adaptation in patients with cerebellar degeneration.