A neurally inspired robotic control algorithm for gait rehabilitation in hemiplegic stroke patients

Cerebrovascular accident or stroke is one of the major brain impairments that affects numerous people globally. After a unilateral stroke, sensory motor damages contralateral to the brain lesion occur in many patients. As a result, gait remains impaired and asymmetric. This paper describes and simulates a novel closed loop algorithm designed for the control of a lower limb exoskeleton for post-stroke rehabilitation. The algorithm has been developed to control a lower limb exoskeleton including actuators for the hip and knee joints, and feedback sensors for the measure of joint angular excursions. It has been designed to control and correct the gait cycle of the affected leg using kinematics information from the unaffected one. In particular, a probabilistic particle filter like algorithm has been used at the top-level control to modulate gait velocity and the joint angular excursions. Simulation results show that the algorithm is able to correct and control velocity of the affected side restoring phase synchronization between the legs.