Clinical effectiveness and patient perceptions of an ILC mediated by ES system using a robotic workstation

Background An inability to perform tasks involving reaching is a common problem following stroke. Evidence supports the use of robotic therapy and electrical stimulation (ES) to reduce upper limb impairments, but current systems may not encourage maximal voluntary contribution from the patient. Iterative Learning Control (ILC) is a technique developed for robotic processes which repeatedly perform a task in order to sequentially improving accuracy. The aim of this study was to test the feasibility of applying ILC mediated by ES to stroke rehabilitation using a robotic workstation, and to measure patients’ perceptions of the system. Methods 5 hemiplegic stroke patients used the workstation (see Figure 1) to track 2 dimensional trajectories, over 18 1 hour sessions within a 3 month period. Outcome measures taken prior to and after the intervention consisted of the Fugl-Meyer Assessment (FMA) and the Action Research Arm Test (ARAT), isometric force and unassisted tracking error. During the intervention patients performed a range of tracking tasks in which their remaining voluntary activity was augmented by ES. Patients were asked questions in each session and undertook a semi structured interview (including Likert style and open questions) subsequent to the study.ResultsCompliance was excellent. Statistically significant improvements were measured (p=0.05) in: FMA motor score (see Table 1), accuracy of unassisted tracking, and in isometric force. The Likert responses showed that all patients found the intervention enjoyable and would have liked further sessions (see Table 2). Patients’ responses to the open questions on the system effectiveness were classified as physical or functional (see Table 3). Conclusion This study demonstrated the acceptability and feasibility of using ILC mediated by ES for upper limb stroke rehabilitation. Future work will address issues raised in the interview as well as investigating using ES for the wrist and hand