Towards Upper- and Lower-Limb Rehabilitation of Post-Stroke Patients through Brain-Computer Interface, Cerebrospinal Stimulation and Robotic Devices
摘要
This work presents the proposal of a complete system for upper- and lower-limb rehabilitation for post-stroke patients using a Brain-Computer Interface (BCI) based on Motor Imagery (MI), cerebrospinal Direct Current Stimulation (csDCS), a robotic glove, and a robotic monocycle. Common Spatial Pattern (CSP), Filter-Bank CSP (FBCSP) and Riemannian Geometry (RG) were used as feature for a classifier (Linear Discriminant Analysis – LDA) for MI recognition. Experiments were conducted with two post-stroke patients, which conducted 3-week sessions, improving their lower-limb movements. An evaluation of their upper-limb progress is being conducted right now.