Single-trial connectivity estimation for classification of motor imagery data

Abstract

We show that single-trial MI classification is possible with connectivity measures extracted from VAR models, and that a BCI could potentially utilize such measures.