Evaluation of a proposal for sustained attention training through BCI with an estimate of effective connectivity.

Many assistive and rehabilitation Brain Computer Interface (BCI) systems have been developed by using brain imaging techniques through electroencephalogram (EEG). This study aims to estimate the effective connectivity of subjects that received attention training through a BCI-based game system. For this purpose, mathematical methods and computational tools were used here to evaluate the effect of using the BCI for sustained attention training, programmed with different levels of difficulty for two groups of subjects. As a result, it was possible to identify differences in connectivity patterns between both groups, mainly the flow outflow characteristic in the left temporal region of subjects who performed the training using the game programmed with greater difficulty level. With this preliminary result, we concluded that the proposed strategy for sustained attention training by applying different speed control in a BCI-based game resulted in the generation of different patterns, capable of distinguishing the users in the two groups. Our approach and findings can contribute to improving the development of EEG-based BCI technologies in identifying patterns related to the state of attention.