Evaluation of HEXORR Tone Assistance Mode Against Spring Assistance

Robots are a promising tool for rehabilitation, and research suggests combining assistance with subject participation to maintain motivation and engagement. In this study, we compare two different types of robotic assistance for hand rehabilitation: Tone, a novel method designed to directly compensate for the user's hypertonia, and Spring, a commonly used method simulating a spring-like force attracting the subject to the target position. Eight chronic stroke subjects played a Gate Game, where they guide a ball through a sweeping, on-screen gate. Each subject was exposed to 5 modes: unassisted, assisted by the Spring and Tone modes, and "delay" modes where the assistance was provided after a period without assistance. Compared to unassisted trials, subjects produced less positive work and more negative work with all types of assistance but also reached a larger ROM and scored higher on the game. Notably, subjects produced more positive work overall with Tone compared to Spring. Further, subjects produced positive work over a significantly shorter range with Spring compared to Unassisted, while there was no difference between Tone and Unassisted. While both modes were effective at enabling the training of a larger workspace, these results suggest Tone assistance maintains a greater level of subject engagement and participation, likely making it a useful tool for hand rehabilitation.