A study to implement a Brain-Computer Interface (BCI) based on Sensorimotor Rhythms

Abstract

This academic article describes an offline research aiming to implement a Brain-Computer Interface (BCI) system based on the manipulation of Sensorimotor Rhythms (SMR) acquired from the Electroencephalograph (EEG). The specific frequencies for each volunteer were found with the analysis of the Distinction-Sensitive Learning Vector Quantization (DSLVQ) algorithm. Extracting features was performed using the Power Samples method whereas the pattern recognition with the Linear Discriminant Analysis (LDA) classification. The results reached 100% correct for a databank (DataSet I) already validated and 80% correct for the voluntary of this research (PAT14).