Generalization of visuomotor adaptation across spatial reference frames
Date
2015
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Motor learning is a neurological process in which movement practice or
experience leads to a change in motor behavior. Prism adaptation (PA) is an early
form of one type of motor learning, in which motor patterns change due to a
displacement of visual information. During PA people perform a visuomotor task
(e.g., reaching or throwing to a target) while wearing prism lenses over the eyes.
Initial performance errors, occurring in the direction of the prism shift, are corrected
through trial and error practice. When the prisms are subsequently removed, errors
occur in the opposite direction and are known as aftereffects. Aftereffects indicate
that the adaptation has been stored by the central nervous system.
PA has been shown to generalize (i.e., transfer to untrained contexts) in some
cases, but not all. In addition, PA has been shown to improve the symptoms of some
patients with the neuropathology known as neglect, a disorder of spatial
representations in which patients fail to detect stimuli in the contralesional hemispace.
Neglect can occur in allocentric (world-centered) or egocentric (self-centered) spatial
reference frames or both. Interestingly however, most intervention studies using PA
treatment have not evaluated its efficacy differentially with respect to these.
In order for PA treatment to be beneficial there must be adequate
generalization to the reference frames (e.g., allocentric, egocentric) affected by the
disorder. To determine how PA generalizes with respect to these spatial reference
frames, healthy participants adapted to rightward displacing prisms by throwing a ball
at a target while in either a seated or side-lying position. Following adaptation
participants rotated to the alternate position and were tested for aftereffects. The
rotation decoupled the allocentric and egocentric reference frames, and the direction of
the aftereffects was used to determine the reference frame of PA generalization.
During PA internal models of motor control are modified in response to a
visual sensory prediction error that may be represented in either allocentric or
egocentric coordinates, or both. Therefore aftereffects could have appeared along the
same axis as the initial visual displacement (allocentric generalization), along the axis
perpendicular to this (egocentric generalization), or in the region between these two
axes (mixed generalization).
Results showed that when participants adapted their throwing to prisms while
in a seated position, significant aftereffects appeared when side-lying, and they were
expressed egocentrically. This egocentric generalization suggests PA may only be
effective for treating egocentric forms of neglect. Surprisingly however, participants
who adapted while lying on their side showed no significant aftereffects, in either
reference frame, when tested in the seated position (i.e., the adaptation did not transfer
from side-lying to seated). This lack of transfer suggests that adaptation during sidelying
throwing was context specific, and this may have been due to the novelty of the
throwing position.