INFLUENCE OF EMOTIONAL STABILITY ON SUCCESSFULNESS OF LEARNING TO CONTROL BRAIN-COMPUTER INTERFACE

Learning of brain -computer interface (BCI) control by means movement training was studied. The training included slow cyclic movements of the right and the left hands (the duration of each movement corresponded to the duration of imagination of the hand movements during BCI control), the imagery of these movements (each of the conditions for 3 min) and the sitting quietly before the movements (6 min). Before the training and after it subjects performed BCI control. Neuroticism was estimated in all of the subjects by Eysenck personality questionnaire. The comparison of the values of the classification accuracy of brain states during motor imagery before and after the training showed that the success of BCI control learning depends on the performance of subjects on a scale of neuroticism (emotional stability/instability). The classification accuracy of brain states during imagination of the right and the left hand was significantly increased after training in users with low levels of neuroticism (values in the range 4 to 11), but not in users with high levels (values in the range 15 to 22). Moreover, the classification accuracy between brain states during imagination of the right hand and in the rest decreased significantly after training in the users with high levels of neuroticism. The more subjective complexity of imagery (but not real) movements, the less emotional stability. The results indicate the need to consider the level of user's neuroticism when designing the methods to learn BCI control.