| Description |
This dissertation explores robotic technologies to generate gait at self-selected speed for application in gait rehabilitation. We present two major research thrusts to improve the current gait rehabilitation for patients with walking impairments. The first thrust includes two studies. The first study aims at demonstrating the potentials of an advanced robotic locomotion interface called the Treadport for gait rehabilitation of spinal cord injury patients. The outcome of this study implies that standard treadmills may impose some constraints on the patients motions that can be overcome on the Treadport. The second study focuses on creating a realistic walking experience on the Treadport. A new controller is proposed for the Treadport based on the major factors influencing walking experience on a locomotion interface. When combined with the users volition, the controller enables the user to naturally self-select their walking speeds as they would when walking over ground. The second thrust includes three studies. The first study presents the design and fabrication of an Underactuated WEarable Arm swing Rehabilitator called the UWEAR that aims at integrating arm swing in gait rehabilitation. The UWEAR is designed to be used with a body-weight-support treadmill. It is backdrivable, capable of assisting the users arm swing in the sagittal plane, and it has unhindered kinematics in the remaining degrees of freedom. The experimental results show the UWEAR’s ability to induce arm swing in its users under various conditions. The second study presents a comprehensive look at the effects of a variety of walking conditions on arm-swing patterns during walking. The results describe the effects of surface slope, walking speed, and physical characteristics on arm-swing patterns in healthy individuals. Finally, the third study proposes a novel method for generating proper arm-swing trajectories in real-time using only measurements of the angular velocity of a person’s thighs, to be used during gait rehabilitation. The proposed method generates smooth trajectories that have high correlations with the actual measured arm trajectories of the healthy individuals. The method is verified on gait data sets gathered from patients with Parkinson disease, and even their pathological thigh trajectories result in proper arm-swing trajectories. |
