Investigating the effects of virtually embodied feedback on performance in motor imagery BCI-based control applications

Abstract

Control applications for patients with motor disabilities commonly use motor imagery-based brain-computer interfaces to enable their users. However, such applications usually require the user to have extensive experience with motor imagery, as discriminant neural patterns are normally cultivated through practice. This study therefore proposes a novel control protocol that provides virtually embodiable feedback, which mirrors the movement of the classified intention, during control scenarios to alleviate the need for extensive training. To verify that the proposed protocol is effective, subjects underwent two virtual reality control scenarios, both in which they used electroenchephalography to control the movement of a virtual drone. One scenario provides embodiable feedback to follow the proposed control protocol, while the other scenario does not provide any feedback. Subjects also answered a questionnaire in which they rated the scenarios on motor imagery performance, embodiment and presence. The results showed that subjects found the protocol to be helpful in improving performance. Moreover, the self-rated embodiment and scores showed significantly positive linear relationships with performance. The findings in our study provide evidence that providing embodiable feedback may be helpful for users using control applications by inducing more discriminable and enhanced neural activity, lowering the need for prior training.