Development of a hybrid robotic system based on an adaptive and associative assistance for rehabilitation of reaching movement after stroke

Abstract

Stroke causes irreversible neurological damage. Depending on the location and the size of this brain injury, different body functions could result affected. One of the most common consequences is motor impairments. The level of motor impairment affectation varies between post-stroke subjects, but often, it hampers the execution of most activities of daily living. Consequently, the quality of life of the stroke population is severely decreased. The rehabilitation of the upper-limb motor functions has gained special attention in the scientific community due the poor reported prognosis of post-stroke patients for recovering normal upper-extremity function after standard rehabilitation therapy. Driven by the advance of technology and the design of new rehabilitation methods, the use of robot devices, functional electrical stimulation and brain-computer interfaces as a neuromodulation system is proposed as a novel and promising rehabilitation tools. Although the uses of these technologies present potential benefits with respect to standard rehabilitation methods, there still are some milestones to be addressed for the consolidation of these methods and techniques in clinical settings. Mentioned evidences reflect the motivation for this dissertation. This thesis presents the development and validation of a hybrid robotic system based on an adaptive and associative assistance for rehabilitation of reaching movements in post-stroke subjects. The hybrid concept refers the combined use of robotic devices with functional electrical stimulation. Adaptive feature states a tailored assistance according to the users’ motor residual capabilities, while the associative term denotes a precise pairing between the users’ motor intent and the peripheral hybrid assistance. The development of the hybrid platform comprised the following tasks: 1. The identification of the current challenges for hybrid robotic system, considering twofold perspectives: technological and clinical. The hybrid systems submitted in literature were critically reviewed for such purpose. These identified features will lead the subsequent development and method framed in this work. 2. The development and validation of a hybrid robotic system, combining a mechanical exoskeleton with functional electrical stimulation to assist the execution of functional reaching movements. Several subsystems are integrated within the hybrid platform, which interact each other to cooperatively complement the rehabilitation task. Complementary, the implementation of a controller based on functional electrical stimulation to dynamically adjust the level of assistance is addressed. The controller is conceived to tackle one of the main limitations when using electrical stimulation, i.e. the highly nonlinear and time-varying muscle response. An experimental procedure was conducted with healthy and post-stroke patients to corroborate the technical feasibility and the usability evaluation of the system. 3. The implementation of an associative strategy within the hybrid platform. Three different strategies based on electroencephalography and electromyography signals were analytically compared. The main idea is to provide a precise temporal association between the hybrid assistance delivered at the periphery (arm muscles) and the users’ own intention to move and to configure a feasible clinical setup to be use in real rehabilitation scenarios. 4. Carry out a comprehensive pilot clinical intervention considering a small cohort of patient with post-stroke patients to evaluate the different proposed concepts and assess the feasibility of using the hybrid system in rehabilitation settings. In summary, the works here presented prove the feasibility of using the hybrid robotic system as a rehabilitative tool with post-stroke subjects. Moreover, it is demonstrated the adaptive controller is able to adjust the level of assistance to achieve successful tracking movement with the affected arm. Remarkably, the accurate association in time between motor cortex activation, represented through the motor-related cortical potential measured with electroencephalography, and the supplied hybrid assistance during the execution of functional (multidegree of freedom) reaching movement facilitate distributed cortical plasticity. These results encourage the validation of the overall hybrid concept in a large clinical trial including an increased number of patients with a control group, in order to achieve more robust clinical results and confirm the presented herein.