Towards a bio-inspired control strategy for bilateral stroke rehabilitation based on motor control principles

Conventional robotic devices used for robot aided neurorehabilitation focus on unilateral training of the upper extremity post neurologic injury such as stroke, cerebral palsy etc. In this paper, we seek to establish adherence of bimanual motions of healthy and stroke subjects to the higher laws governing principles of symmetric and asymmetric division of labor in bimanual activities. As a first pass, we propose three metrics to capture these principles. We identify that the Euclidean and velocity difference metrics were able to capture key aspects of key bilateral motion control principles for the bilateral tasks studied, but may fall short if the tasks were different and asymmetric. We propose a control strategy to use the metrics to support task-oriented, bilateral stroke rehabilitation. Future work requires the development of metrics that more fully capture the motor control principles irrespective of tasks and the implementation of a bilateral controller.