NSF Award
1347496
Stroke occurs in 2.9% of the US adult population and there are nearly 800,000 new or recurring<br/>cases annually. Neuromuscular impairment is a common consequence, though one from which recovery<br/>is possible with appropriate therapy. Though most healthcare plans cover post-stroke therapy,<br/>insurance reimbursement caps typically limit it to a few weeks, despite a growing body of<br/>evidence that therapeutic gains are possible months and even years after a stroke. Adding<br/>the fact that patients generally prefer out-of-clinic care, there is a compelling case for<br/>delivery of physical therapy in the home setting. The new product that will result from this<br/>project will increase dosing of meaningful exercises that promote recovery and minimize co-morbidities<br/>such as pain, swelling, spasticity and (shoulder) subluxation.<br/>This project?s outputs respond to other trends in the US healthcare system, including the<br/>dwindling supply of qualified care providers, the movement toward evidence based medicine<br/>and the economic requirement for increased self-care. The ultimate goal of our Virtual Coach<br/>technology is to leverage scarce human-based care with computer based solutions at substantially<br/>lower cost. The self-adjusting goal-setting capability developed in the proposed research<br/>will enable professional therapists to extend their reach to patients.<br/><br/>One significant challenge in computer-based coaching is devising algorithms that affect human<br/>behavior in an intended way. It is difficult to predict how people will react to commands,<br/>instructions, prompts and rewards. Some respond well to a nurturing "atta boy!" while others<br/>respond better to a drill sergeant?s "that?s not good enough!" Moreover, not only is there<br/>variability among people, but also within a particular individual from day-to-day. To be a<br/>viable, pervasive solution, Virtual Coaches will need to be responsive to the psychology<br/>of the people they serve. Pursuit of our objectives, especially the exploratory research objective,<br/>addresses this challenge in the context of motivating people to adhere to an exercise regimen.<br/>A second challenge is designing exercise software factoring in the heterogeneous presentation<br/>of individuals with neuromuscular impairments. Such patients have varying levels of motor<br/>control, cognitive ability and co-morbidities associated with their impairments. Off the<br/>shelf exercise packages do not account for their limitations. Our translational research objective<br/>is to overcome this challenge by leveraging research conducted on impaired patients and an<br/>already developed software architecture for coaching them through movements that promote long<br/>term recovery.
