Reem M. Azar

Purpose: Robot-assisted therapy predominantly targets motor impairment, although post-stroke motor and cognitive impairments co-occur. We report user experience in a longitudinal deployment of a Rehab Community-based Affordable Robot Exercise System (RehabCARES) for delivering personalized motor/cognitive training. Methods: Six adults with chronic stroke (FM-UE 20-61; MoCA 22-28) completed 12 therapy visits with THEAbot, a reconfigurable 1-DoF robot. Rehab CARES also included a haptic/biometric dynamic difficulty adjustment controller and eight games characterized by cognitive processes and motor/cognitive demand. We report gaming preference, usage, and user experience (pain, exertion, usability, affect, workload) across therapy. Results: User experience varied, but generally, self-reported pain was low, perceived exertion was light to moderate, enjoyment was strong and stable, and sense of control increased while excitement plateaued. Workload showed decreasing frustration and effort. Game selection during therapy was partially consistent with motor/cognitive clinical profiles; preferences and perceptual abilities may have influenced choices. Usability scores exceeded acceptability at all time points, peaking at the final visit (86 +/- 4.5 out of 100). Conclusions: RehabCARES provided personalized motor/cognitive training to a heterogeneous cohort in a community-based setting, delivering a positive user experience. Hardware reconfigurability and motor/cognitive game variety likely enabled this success.

Background: Following stroke, motor and cognitive impairments often limit functional independence. Personalized rehabilitation requires comprehensive clinical assessments, which are costly and time-consuming. Imaging scans are routinely collected, but their predictive value in determining post-stroke function remains unclear. Objectives: To determine the ability of clinical assessments, lesion features with medical chart information, and a combined clinical and biomedical model to predict kinematic outcomes from a robot-based tracking task in chronic stroke individuals. Methods: Eighteen participants (70±6 years) completed a robot-based tracking task and a battery of motor and cognitive assessments. Three elastic net regression models were developed: 1) clinical assessments only; 2) lesion metrics and chart information only; and 3) a combined model. Models were evaluated using leave-one-subject-out cross-validation, bootstrapping, and permutation testing with false discovery rate correction. Robot-based outcomes included tracking error, normalized distance, velocity error response, and delay. Results: The clinical assessment model significantly predicted tracking error (R²=0.45), delay (R²=0.22), velocity error response (R²=0.18), and normalized distance (R²=0.12). The combined model significantly predicted tracking error (R²=0.37), delay (R²=0.19), and velocity error response (R²=0.11). The imaging model did not significantly predict any outcome. Box and Blocks Test and Color Trails 1 were stable features. The right dorsal attention network was a stable predictor in several combined models. Conclusion: Clinical assessments outperformed lesion features in predicting tracking performance after chronic stroke. Attention- and motor-related measures were the strongest predictors, highlighting the importance of cognitive-motor interactions in function. A larger sample and more granular lesion features may improve prediction from imaging biomarkers.