Recovery of motor function in patients victims of stroke by brain-machine interface training

Abstract

Stroke is a significant cause of disability worldwide, especially in developing regions such as Latin America and, in particular, Brazil. Currently do not exist effective treatments for stroke-related motor deficits, and although promising studies using non-invasive brain-machine interfaces (BMI) have been conducted, few studies have analyzed brain electrical activity during BMI in stroke victims. Previous articles from our research group have proven motor rehabilitation for stroke affecting the upper limb, based on non-invasive BMI (such as EEG and MEG). The recovery presented was more evident when the patient maintained some residual strength proximal in the arm, indicating that the recruitment of intact circuits could play an important role in motor learning. However, BMI systems lack improvement for better integration with visual and sensory-motor signals, in a feedback system. The objective of the proposed project represents a new strategy in the management of stroke: to induce motor recovery of the non-functioning hand in chronic phase stroke patients by inducing cortical plasticity through training in non-invasive BMI based on signs of brain electrical activity in a virtual reality (VR) environment and feedback system for control of hybrid actuators (exoskeleton and functional electrical stimulation, FES). The intention is to develop models through pattern recognition to program BMI devices and neuroprothesis, which will allow us to make important advances in this field, developing new therapeutic techniques based on EEG signals (hardware and software), promoting functional recovery in patients with high degree of motor incapacity. Currently there are no similar methods available as treatment established in the health unic system or in the private system of Brazil. (AU)