Title: The role of afferent vs. efferent signals in implicit sensorimotor adaptation
Thesis Supervisor: Dr. Romeo Chua
Committee members: Dr. J. Timothy Inglis, Dr. Hyosub Kim
Defence Chair: Dr Jean-Sébastien Blouin
Abstract: Humans continuously adjust their movements to adapt to changes in both external surroundings (e.g., walking on a slippery surface) or internal conditions (e.g., fatigue). Whereas we may be consciously aware of external changes as they are often sudden or obvious, internal changes are usually slow and gradual, provoking motor adaptation that can occur implicitly. The clamped visual feedback task where participants reach towards targets on a visual display while observing task-irrelevant rotated cursor feedback has been developed to isolate the implicit component of adaptation. Despite awareness of the manipulation and instructions to ignore the visual feedback, participants’ reach direction gradually drifts in the opposite direction of the cursor, eventually reaching an upper bound. It has been hypothesized that the perturbed visual information biases where the participant feels their hand to be, and adaptation persists until the perceived hand position re-aligns with the target. While proprioception has been proposed as a critical sensory information source contributing to the perceived hand position (and therefore driving adaptation), the role of the sensory prediction derived from the efference copy has largely been neglected. Current protocols do not allow for the separation of afferent and efferent sensory contributions; therefore, their individual contributions towards overall implicit adaptation behaviour are poorly understood.
The purpose of this study was to examine the effect of dissociating the afferent and efferent information available during implicit sensorimotor adaptation. I investigated whether the misaligned visual feedback during the clamped visual-motor rotation task could alter our perceived hand position by biasing our efferent, motor-related prediction of where our hand should be. This was examined through an isometric aiming task. Instead of performing actual reaching movements, participants were instructed to move a visual cursor towards various targets by applying horizontal force manually to an immovable handle. During perturbation trials, the cursor followed an invariant path rotated relative to the target. Participants were instructed to ignore this task-irrelevant cursor feedback and continue to “reach” towards the target. We found that participants implicitly adapted in a task where they were required to generate isometric forces to move an error-clamped rotated cursor on a visual display instead of performing actual reaches to the target. Furthermore, we observed faster learning rates and greater overall adaptation than in a typical clamped reaching paradigm. This finding was confirmed in a secondary experiment where participants performed actual reaching movements and demonstrated significantly less adaptation. Although the isometric task did not remove proprioception itself, it decoupled this feedback source from the task. Thus, these findings suggest that while afferent proprioceptive feedback of hand position at the target location most likely plays a role in adaptation behavior, it may not be necessary to induce adaptation. The results of this study have important implications on current computational models of implicit sensorimotor adaptation.