Meditation as an effective BCI training protocol for controlling virtual wheeled robots

Abstract

Background: Brain-Computer Interface (BCI) systems allow the use of electroencephalographic activities to control devices. As such, it can be an important tool for disabled people since using a BCI does not involves any muscular stimulation. Despite its great potential as an assistive technology, BCI systems are yet scarcely available outside scientific contexts. The main reason is that BCI is still not accurate enough for real world situations. In this paper, we investigated if mindfulness meditation can help BCI users to perform better, particularly on controlling virtual wheeled robots. Controlling virtual wheeled robots must be seen as an important BCI research issue, since disabled people can take the best advantages from assistive wheeled robots. Methods: Case-control study with 30 subjects, meditators and non-meditators, who controlled a virtual wheeled robot using a BCI system. Results: In straight-ahead moves, the robot was 30% faster when controlled by meditators. In stop-and-go moves, the meditators controlled the robot with an accuracy of 66% while the non-meditators’ accuracy was 27%. In rectangular shape moves, meditators also performed better than non-meditators, with an accuracy of 40% against 7% of non-meditators. Conclusion: The results show that meditators performed better than non-meditators. As such, we recommend combining mindfulness meditation with standard BCI training protocols for better control of virtual wheeled robots.