BACKGROUND: Early after stroke, upper limb rehabilitation often fails to restore arm function or track progress effectively. We investigated the feasibility, safety, and preliminary efficacy of a protocolized, high-dose arm training program, and piloted a novel sensor-based kinematic assessment to capture movement quality. METHODS: Fifty-four inpatients with upper limb impairment 2-14 days post-stroke were recruited. Participants were stratified by impairment to MindPod® (n=23) or Bimanual Arm Trainer (BAT, n=20), each receiving up to 45 minutes/day of protocolized Video Gaming Technology (VGT) during hospitalization. Feasibility outcomes included session completion, physiologic response, and patient-reported experience. Preliminary efficacy was measured with the Fugl Meyer Upper Extremity (FMUE) assessment. The kinematic pilot study included five patients post stroke, five healthy controls, and five patients pre/post intervention. Metrics included peak velocity, total movement time, smoothness (normal velocity peaks-NVPs), and trunk displacement. RESULTS: VGT was safe with minimal adverse events (.01% of total sessions) and small physiological changes (baseline systolic Blood Pressure changes >15% in 7.8% of the sessions). The 45-minute training target was not achieved (mean time on task reached 29.9 minutes) yet most participants found VGT enjoyable (93%). Both groups approached or exceeded the FMUE Minimally Clinical Important Difference (MCID=4-13): BAT M=14.2 (SE=1.6); MindPod® M=5.1 (SE=2.8).

The kinematic assessment was well tolerated. Healthy participants completed tasks faster than stroke participants (Finger-to-Nose, p=.006; Apple, p=.028). Participants with stroke showed more instability, with higher NVPs compared to the average amount expressed by healthy participants (FTN p=.012; A3 p=.068; Apple p=.012) Post intervention, stroke participants improved FMUE scores by 7.2 points (meeting MCID) and demonstrated reduced NVPs and task time. Changes in FMUE correlated strongly with apple task metrics (Spearman’s>0.70) and moderately with trunk measures during FTN. CONCLUSIONS: A standardized, protocolized approach to early arm rehabilitation is safe, feasible, and well-tolerated in the early post-stroke recovery period, supporting future efficacy studies. The novel kinematic assessment was sensitive to movement quality changes and may provide a valuable tool for monitoring impairment reduction. Larger studies including joint rotations and range of motion are needed to refine movement quality assessment and guide optimal upper limb rehabilitation.