A Decoding Method Using Riemannian Local Linear Feature Construction for a Lower-Limb Motor Imagery Brain-Computer Interface System

摘要

Recently, motor imagery brain-computer interfaces (BCIs) have been developed for use inmotor function assistance and rehabilitation engineering. In particular, lower-limb motor imageryBCI systems are receiving increasing attention in the field of motor rehabilitation, because thesesystems could accurately and rapidly identify a patient's lower-limb movement intention, whichcould improve the practicability of the motor rehabilitation. In this study, a novel lower-limb BCIsystem combining visual stimulation, auditory stimulation, functional electrical stimulation, andproprioceptive stimulation was designed to assist patients in lower-limb rehabilitation training. Inaddition, the Riemannian local linear feature construction (RLLFC) algorithm is proposed to improvethe performance of decoding by using unsupervised basis learning and representation weight calculationin the motor imagery BCI system. Three in-house experiment were performed to demonstratethe effectiveness of the proposed system in comparison with other state-of-the-art methods. Theexperimental results indicate that the proposed system can learn low-dimensional features andcorrectly characterize the relationship between the testing trial and its k-nearest neighbors.

关键词