Friday, December 5, 2008

Motor adaptation and motor learning in neurorehabilitation


Amy Bastian Understanding sensorimotor adaptation and learning for rehabilitation. Curr Opin Neurol 2008; 21: 628-633.

Motor adaptation and motor learning are critical for flexibility. Adaptation refers to trial to trial modification of movements based on error feedback, especially change in the pattern or force of direction oof the movement. Once adapted patients cannot revert to prior behavior unless they deadapt. It occurs over tens to hundreds of movements and over minutes to hours, and can occur in all types of movements including reaching, balancing, eye movements and walking. Adaptation may have a rapid and slow phase. Sensory prediction error is the difference between predicted and observed outcome of the movement. They are used to calibrate internal representations of body dynamics and the environment and recalibrate.

Brain may also alter movements to minimize the "costs" including energy demands, fatigue, and others especially during walkingand reaching adaptation.

Lesions are especially important in the cerebellum that decreased trial by trial improvement during adaptation and reduces stored aftereffects. Basal ganglia damage due to PD or HD leaves adaptation intact largely. Cerebral damage slows but does not abolish adaptation with reaching, and does not impair split belt treadmill walking.

Rehabilitation with robots or treadmills are used . In a reaching task, adaptation was rehabilitated more if perturbation (robotic arm) was introduced gradually rather than at full strength. Adaptation after stroke with neglect was long lasting (5 days) in prism task of reaching. Aftereffects can be substantial. After effects can also prove the ability to normalize an action. Learning via repeated adaptation has been less studied than single session learning.

Motor learning is formation of a new motor pattern that occurs via long term practice. One form is tied to adaptation. Motor learning can be blocked by rTMS.

No comments: