Effective connectivity between different brain regions during a working memory (WM) task is assessed using time-varying dynamic Bayesian networks (TV -DBN) and EEG signals.In -Degree connectivity to frontal electrodes in theta band showed higher values during execution and manipulation periods of a low demand WM task compared to those of a more difficult one. In contrast, stronger frontal connectivity manifested during the maintenance period, as a result of increased task difficulty. Utilizing the connectivity values among a few EEG channels and TV-DBN, which is a computationally efficient algorithm, we propose a new brain computer interface (BCI) system

Keywords