Chair: Dr. Chandramouli Krishnan

Zoom registration: https://umich.zoom.us/meeting/register/-O2KxTBkRny7lrYL585zIA
PASSCODE: 099355

Abstract
Stroke is one of the leading causes of adult-onset disability in the United States. Restitution of motor control and improved independence in daily life following a stroke is mediated by high dosages of intensive, task-specific therapy. However, adequate therapy is often limited by the high costs of outpatient therapy sessions and reduced third party coverage. As a result, many stroke survivors must reduce their therapy dosage and do not achieve their full recovery potential. Rehabilitation devices, such as robots, present a unique opportunity to increase dosage by automating the repetitive features of rehabilitation, tracking patient progression, integrating gaming interfaces to increase patient engagement, and expanding in-home therapy. Unfortunately, many of these devices are highly complex, motorized systems that are too expensive and bulky for clinical or in-home use. Therefore, it is pertinent that researchers investigate ways to decrease the cost of rehabilitation devices while maintaining their utility in rehabilitation. Hence, in this dissertation, we designed, developed, and examined low-cost devices for stroke rehabilitation. We developed and examined three low-cost devices for stroke rehabilitation. The first device was SepaRRo, a semi-passive planar upper extremity robot for stroke rehabilitation. SepaRRo uses controllable brakes to provide training forces to the user’s end-effector (i.e., hand) during targeted reaching that can either resist and steer their motion. Steering forces from SepaRRo have two potential applications in rehabilitation: guide a user onto a path to assist in learning a motor skill or steer a user away from a path so that they alter their muscle coordination. In this dissertation, we examined the ability of SepaRRo’s steering forces to help users learn a motor skill and examined how steering forces could alter muscle coordination during functional resistance training in stroke survivors. The second device we developed and examined in this dissertation was the Hand eMBot, a self-powered passive robot for hand rehabilitation. Here, self-power refers to a class of stroke rehabilitation robots that use power from the less-impaired limb to assist the more-impaired limb. The Hand eMBot uses this principle to assist the more-impaired hand with finger flexion and extension. The Hand eMBot includes three coupling transmissions that couple the motion of the (i) thumbs, (ii) index fingers, (iii) all remaining fingers, thereby allowing the self-power to assist in gross and dexterous digit motion. We demonstrated the ability of the Hand eMBot to reflect motion to the opposite limb, alter corticospinal excitability, and improve functional usage of the more-impaired hand in stroke survivors. The third and final device we examined was the NewGait®, a commercially available passive elastic exosuit for post-stroke gait rehabilitation. The NewGait® system features a variety of attachable elastic bands that span the lower extremity joints whose stretching and relaxing during the gait cycle apply assistive torques to the wearer. In this chapter, we examined the NewGait’s ability to supplement a post-stroke gait intervention and how it compared to a similar commercially available device. Collectively, this dissertation will set a foundation of work for a new generation of low-cost devices for stroke rehabilitation that could increase the accessibility of cutting-edge technologies to patient populations and improve recovery outcomes.