This study has 2 parts: In one part of this study, people with stroke will either play a custom designed computer game for stroke rehabilitation called Duck Duck Punch or an off the shelf computer game with their weaker arm 3 times per week for 6 weeks. Evaluations will determine whether or not one computer game improved arm movement more than the other. In the second part of the study, people with stroke, caregivers of people with stroke and stroke rehabilitation therapists will meet in several focus groups to design a useful and informative Duck Duck Punch performance report.

Stroke is a problem nationally, but especially in the southeastern USA, a region known as the "stroke belt" where stroke incidence is high and age of stroke onset is low. The vast majority, >75%, of stroke survivors experience paresis of one arm/hand that does not resolve acutely. Long-term arm movement impairment restricts independence with self-care and vocational activities, increases caregiver burden and reduces quality of life. Although rehabilitation improves outcomes, systematic financial pressures increasingly limit its duration. Unfortunately, this is happening at a time when strong evidence is emerging that traditional therapy programs do not provide adequate amounts of movement practice needed for motor recovery. Thus, there is a need for innovative technology to augment traditional stroke rehabilitation programs in a way that can provide the necessary movement practice within the constraints of current rehabilitation practice.

To meet this need, the Principal Investigators developed a prototype Kinect-based post-stroke rehabilitation game called Duck Duck Punch (DDP). While maintaining the appeal of a game, DDP has a therapeutic focus because its unique design elicits an arm motor recovery process consistent with evidence-based stroke rehabilitation principles. The player moves his/her physical arm to control an avatar arm to reach and "punch" virtual ducks. Custom features allow tailoring of game difficulty to match a player's impairment level so that the player seeks to accomplish optimally challenging movement goals. By design, the avatar does not respond to atypical arm motions, which encourages the player to trial and error a variety of motions until implicitly learning the more normal strategy. Thus, unlike most commercially available "off the shelf" games, success requires "therapist approved" healthy arm motions. Success motivates continued game play for extended practice of healthy motions. Therapists can integrate DDP into in-clinic or in-home therapies for additional quasi-supervised movement practice and receive a performance report that quantifies and monitors progress toward recovery goals. Further development of this report will enable its integration into a billable rehabilitation program.

The Investigators licensed DDP and formed a company, Recovr, which has received investment funding for initial start-up and market research. Of note, DDP has also received FDA 510(k) Clearance to "support physical rehabilitation of adults in the clinic and at home via performance of therapist-assigned reach exercises for the upper extremities." In a funded NIH/NIGMS pilot project, the investigators established the technical merit and feasibility of DDP as a tool to augment inpatient, outpatient and home-based stroke rehabilitation by increasing therapist- and patient-directed movement practice opportunities. Very promising results motivated the current project that seeks to test the functional viability of DDP and determine its commercial potential.

Inclusion Criteria:

• experienced unilateral hemispheric ischemic or hemorrhagic stroke at least 3 months but no more than 7 years prior
• exhibit voluntarily shoulder flexion of the affected arm ≥30° with simultaneous elbow extension ≥20°. The investigators reason that persons at this motor ability level have residual arm activation and enough ability to engage in treatment-related reaching movements elicited by the computer games
• baseline FMA-UE score of at least 19 points but no more than 52 points (out of 60 points) based on previously published research by this study's investigators in which categories were defined based on post-stroke UE motor impairment
• passive range of motion in affected shoulder, elbow and wrist within 20 degrees of normal values
• 21-90 years of age
• a caregiver or friend who is willing to assist with the set up and operation of the computer game throughout the 6 week intervention.

Exclusion Criteria:

• lesion in brainstem or cerebellum because lesions in these locations my interfere with the visual-perceptual and cognitive skills needed for motor re-learning as is expected to occur as a result of the intervention
• presence of other neurological disease that may impair motor skills (e.g., Parkinson's Disease)
• pain in the affected arm that interferes with reaching movements
• significant cognitive impairment, defined as Montreal Cognitive Assessment score < 22
• orthopedic condition or impaired corrected vision that alters the kinematics of reaching
• unable to travel to the UE Motor Function Laboratory in Charleston, South Carolina 4 times (pre-, mid- post- and retention testing).

At the conclusion of the study, de-identified data will be available to other researchers and may be obtained by emailing the PI

Adverse event data were collected throughout the duration of the study a total of 1 year, and 11 months.

Blinded raters perform scoring of all assessments