| Abstract No.: | B-D2128 |
| Country: | Canada |
| | |
| Title: | HOW IS LEARNING GENERALIZED? |
| | |
| Authors/Affiliations: | 1 Amaris Siegel*; 1 Denise Henriques;
1 York University, Toronto, ON, Canada
|
| | |
| Content: | Objectives: Motor learning requires a remapping of the relationship between vision and motor output. Intermanual transfer is a form of generalization where learning a new visuomotor mapping with one arm affects the performance of the opposite arm. After adapting to altered visual feedback with an unseen hand, it is unclear how intermanual transfer occurs. Does transfer occur because the brain is learning new cursor mechanics, which are constant for each hand? If so, then we predict that bimanual transfer should occur when subjects learn to reach with a cursor representing their hand and not an image of their hand.
Methods and Materials: Subjects reached to 10 radial targets 10 cm from the central start position on a vertical screen. One group of subjects performed the reaching task with a rotated view of cursor motion representing their unseen hand. Another group of subjects saw a rotated view of their hand while they performed the same task; these movements were captured using a camera, and displayed in real time. The two groups were further divided so that the magnitude of the rotation was either 45° or 105° counter-clockwise (CCW). During training, subjects reached with their right hand under one of the four different conditions for 200 trials. Following training, subjects reached to the same targets under the same viewing condition but with the opposite (left) hand for 30 trials. Deviations from a straight path were measured using maximum deviation perpendicular from the movement path.
Results: To measure intermanual transfer, we compared the maximum deviation for the first 10 reaches with the right hand with those of the untrained left hand. When subjects reached with the cursor representing their hand, the maximum deviation was significantly smaller for the untrained hand for both the 45° rotation (p < .001) and the 105° rotation (p = .003): the percentage drop in maximum deviation from training to testing was about 57% and 36% respectively. When subjects reached with a view of the hand rotated 45°, the maximum deviations for initial reaches did not differ across the two hands (p > .05), and only marginally, but not significantly (20% drop) for the 105° rotated hand view.
Conclusion: When participants view a cursor representing their unseen hand, learning is generalized, but when the rotated hand is visible, learning does not seem to generalize to the same extent. Our results suggest that intermanual transfer may occur because the brain is learning an internal model of the cursor, rather than the arm motor system.
|
| | |
| Back |
|
