Effects of Emotional Stability on Success in Learning to Control a Brain–Computer Interface

Abstract

Learning to control a “brain–computer interface” (BCI) system was studied using specially developed training. Training included performance of slow cyclic movements with the right and left hands (the durations of each movement corresponded to the durations of the hand motor imagery when working with the BCI), imagining these movements, and also sitting calmly before movements started. Control of the BCI was tested before and after training. The Eysenck personality questionnaire was used to assess the level of emotional stability (neuroticism). Comparison of the accuracy of classifi cation of the state of the brain during motor imagery before and after training showed that the success in learning to control the BCI depended on subjects’ scores on the neuroticism scale. After training, the separability of brain states on motor imagery with the left and right hands increased signifi cantly in users with low, but not in those with high, neuroticism; furthermore, separability of brain states in the latter decreased signifi cantly during right hand motor imagery and at rest. The subjective diffi culty of the imaginary (but not real) movements was greater when scores on the neuroticism scale were higher. These results provide evidence of the need to consider the emotional stability of users when developing methods of training to BCI control.