Maintenance of Gains from Robot-assisted Stroke Therapy

Charles G. Burgar, MD; Peter S. Lum, PhD; Peggy Shor, BS OTR; Matra Majmundar, OTR TYC; H.F. Machiel Van der Loos, PhD

Objectives: The aging veteran population is at risk for cerebrovascular disease and there is a need for more efficient and effective rehabilitation interventions. This project assessed the therapeutic efficacy of robot-assisted movement in subjects with chronic stroke. We previously reported significantly more improvement in upper limb (UL) motor control immediately following robot-assisted therapy, compared to conventional therapy. This report presents the 6-month follow-up results and an analysis of pain and functional outcome data.

Methods: Subjects with chronic hemiplegia (> 6 months post-stroke) were randomly assigned to a robot or control group. All received 24 one-hour UL therapy sessions over a two-month period. Robot-assisted movements were practiced by 13 subjects with the aid of MIME™, an investigational device based on a Puma 560 robot arm. Unilateral exercise of the limb contralateral to the stroke progressed from passive range-of-motion (ROM), to active-assisted ROM, then to resisted movements constrained to predetermined trajectories. During bimanual exercise, a digitizer sensed the subject's movement of the ipsilateral limb and the robot moved the contralateral limb to the mirror-image position. Control subjects (14) received treatment based on Neuro-Developmental Therapy, targeting UL function, and non-contact exposure to the robot during target tracking tasks. All subjects were evaluated pre- and post-treatment with clinical and biomechanical measures.

Results: The proximal UL scores on the Fugl-Meyer (FM) exam improved more in the robot group, compared to controls, during and immediately after completing treatment, then were maintained. Control subjects continued to improve and the difference was no longer significant at the 6-month follow-up. Functional Independence Measure (FIM) upper body scores did not change during treatment but were significantly higher in the robot group at 6 months. Modest trends reflecting improvement in the distal UL FM, Barthel, and FM pain scores were observed in both groups but differences were non-significant.

Conclusions: Assisted-movement UL therapy using MIME in this chronic stroke population promoted safe recovery of motor function that was maintained at follow-up. The mechanisms producing post-treatment FIM gains are not yet identified. Use of MIME, a low compliance robotic manipulator, did not produce adverse changes in the FM pain scores. Subjectively, the device was well tolerated. Studies in patients 1-20 weeks post stroke are in progress.

Acknowledgment: This work was supported by VA Rehab R&D Merit Review Grant #B2056RA.