Robot enhanced stroke therapy optimizes rehabilitation (RESTORE): a pilot study

Abstract

Background: Robotic rehabilitation after stroke provides the potential to increase and carefully control dosage of therapy. Only a small number of studies, however, have examined robotic therapy in the first few weeks post-stroke. In this study we designed robotic upper extremity therapy tasks for the bilateral Kinarm Exoskeleton Lab and piloted them in individuals with subacute stroke. Pilot testing was focused mainly on the feasibility of implementing these new tasks, although we recorded a number of standardized outcome measures before and after training.

Methods: Our team developed 9 robotic therapy tasks to incorporate feedback, intensity, challenge, and subject engagement as well as addressing both unimanual and bimanual arm activities. Subacute stroke participants were assigned to a robotic therapy (N = 9) or control group (N = 10) in a matched-group manner. The robotic therapy group completed 1-h of robotic therapy per day for 10 days in addition to standard therapy. The control group participated only in standard of care therapy. Clinical and robotic assessments were completed prior to and following the intervention. Clinical assessments included the Fugl-Meyer Assessment of Upper Extremity (FMA UE), Action Research Arm Test (ARAT) and Functional Independence Measure (FIM). Robotic assessments of upper limb sensorimotor function included a Visually Guided Reaching task and an Arm Position Matching task, among others. Paired sample t-tests were used to compare initial and final robotic therapy scores as well as pre- and post-clinical and robotic assessments.

Results: Participants with subacute stroke (39.8 days post-stroke) completed the pilot study. Minimal adverse events occurred during the intervention and adding 1 h of robotic therapy was feasible. Clinical and robotic scores did not significantly differ between groups at baseline. Scores on the FMA UE, ARAT, FIM, and Visually Guided Reaching improved significantly in the robotic therapy group following completion of the robotic intervention. However, only FIM and Arm Position Match improved over the same time in the control group.

Conclusions: The Kinarm therapy tasks have the potential to improve outcomes in subacute stroke. Future studies are necessary to quantify the benefits of this robot-based therapy in a larger cohort.

Trial registration: ClinicalTrials.gov, NCT04201613, Registered 17 December 2019-Retrospectively Registered, https://clinicaltrials.gov/ct2/show/NCT04201613 .

Keywords: Robotic rehabilitation; Robotics; Stroke rehabilitation; Subacute stroke.

Fig. 1

The Kinarm Exoskeleton. A frontal view of the Kinarm Exoskeleton

Fig. 2

Visualization of robotic therapy tasks. Visual examples of some of the tasks used within the pilot study. In these tasks, and all other tasks, participants are not able to see their arms. These visuals have been added to assist in understanding how the subjects interact with the virtual environment. Yellow arrows indicate the direction of hand movement in the reaching tasks. Black arrows indicate the direction of the forces applied during the task, when applicable

Fig. 3

Study flow diagram. Participants in the robot group received 10 sessions of robotic therapy in addition to standard of care for 10 days, whereas those in the control group received only standard of care for 10 days prior to the follow-up assessments

Fig. 4

Robotic therapy task scores over time. The task scores of each participant over the course of the robotic intervention. Each panel indicates scores for a specific task. An open circle indicates that the participant completed the task on that day of the intervention. No circle is present if the participant did not complete the task on that specific day. Red inlays indicate a level-up on that day and black inlays indicate a level-down. The bold black line indicates average scores. Asterisks beside the task title indicates a significant improvement in score from day 1 to day 10. (* = p < 0.05 ** = p < 0.01)