Effects of Physical Guidance on Motor Control and Learning During Human Walking.

Abstract

Physical guidance is often used in rehabilitation when teaching patients to re-learn movements. However, the effects of guidance on motor learning of complex skills are not clear. The overall goal of this dissertation is to determine how physical guidance affects the neural control and motor learning of human walking. In the first experiment, I studied the effects of manual assistance on kinematics and muscle activation during body-weight supported treadmill training in subjects with incomplete spinal cord injury. I found that kinematics and muscle activity did not substantially change when subjects were given manual assistance. Manual assistance allowed subjects with spinal cord injury to train at faster speeds and made muscle activation patterns more similar to those in able-bodied subjects. In the next set of experiments, I used a novel treadmill-mounted balance beam (beam-mill) to study learning of walking balance in neurologically intact human subjects. Subjects practiced walking on the beam-mill with different types of physical assistance and were compared to those that practiced without assistance. In the second experiment, physical assistance was provided with a spring-based stabilization device. Results showed that error-reducing physical assistance hindered learning of the unassisted task. In the third experiment, I investigated whether augmenting error during practice would enhance learning of beam-walking since movement errors drive learning. Two groups of subjects practiced with a destabilization device that had springs with medium or high negative stiffness. Another group walked on a narrower beam to augment error, but with more similar dynamics to the evaluation task. Subjects that practiced unassisted had greater performance gains than those that practiced with error augmentation. However, practicing on the narrow beam had the best performance gains of the error augmentation groups. In the last experiment, subjects practiced with a device that permitted normal movement variability but minimized catastrophic error (i.e. stepping off the beam). Subjects that practiced with this device had very small performance gains, demonstrating that catastrophic errors are important for learning walking balance (if they can be made safely). Overall, these studies support using physical guidance during gait rehabilitation but emphasize that task specificity should be maintained during practice.