Distinct Motor Learning Mechanisms for Training Locomotion

Hugo W. Moser Res Inst Kennedy Krieger, Baltimore MD

Investigators

Linked publications & trials

Abstract

? DESCRIPTION (provided by applicant): Walking ability is a prognostic factor for quality of life. Many gait disorders are caused by neurological damage, with stroke being among the most common. Patients with cerebral stroke typically have difficulty walking and require locomotor rehabilitation. The ability to learn a new walking pattern through rehabilitation depends on a suite of learning mechanisms including instructive, reinforcement and adaptive learning. In instructive learning, patients are instructed how to move (typical of classic rehabilitation strategies). In reinforcement learning, patients independently discover the correct movements and these actions are reinforced with reward. In adaptive learning, a patient's movements are perturbed and the patient must adapt to these perturbations to successfully move. However, little is known about how these learning mechanisms interact. Our long-term goal is to understand how to optimally combine these three motor learning mechanisms to design locomotor rehabilitation strategies for patients with gait disorders. The proposed project utilize a custom-designed split-belt treadmill/visual-feedback system that will allow for the evaluation of instructive, reinforcement, and adaptive learning. Instructive and reinforcement learning will be provided with the visual display that provides feedback about the person's walking, while adaptive learning will be provided by driving the two belts of the split-belt treadmill at differen speeds. Aim 1 investigates how these three learning mechanisms interact when learning new walking patterns in healthy adults. We predict that subjects will be able to learn by two mechanisms simultaneously (e.g. instruction while adapting to the split-belts) without interference. Based on previous work, we also predict that reinforcing recently learned actions will increase retention of the new walking pattern. Aim 2 focuses on how to enhance adaptive learning with instructive and reinforcement learning for patients with cerebral stroke. Recent work has shown that patients with cerebral stroke can be rehabilitated via adaptive learning with a split-belt treadmill and that their improved walking pattern is present months after training. However for most of these patients, rehabilitation is incomplete. We predict that the addition of instructive learning during and reinforcement learning following adaptive learning techniques will help guide patients to a better walking pattern with improved retention. The results of the proposed research will advance the understanding of the multiple motor learning mechanisms used for locomotion. This information will lay the foundation for designing optimal rehabilitation strategies for patients with gait disorders.