Evaluation of various strategies to improve the training of a brain computer interface system

Brain-computer interfaces (BCI) allow for communication between the human brain and computers. They have been used to create systems able to generate control over a software or mechatronic device, becoming an important tool for the rehabilitation of handicapped people, especially for patients with quadriplegia. Due to the importance of BCIs, any attempt to improve their response has a great value. The work described in this paper consists of the evaluation of various stimuli applied to improve the training of a BCI based system. The device is the Emotiv® EPOC which includes a Software Development Kit (SDK) to train the system. During the usual training of the SDK; images, videos, thoughts, movements and gestures were used as stimuli to improve the users' skills for accomplishing specific actions. After the evaluation of these stimuli, they were ranked according to their effectiveness. Finally, a powered wheelchair was controlled using the Emotiv® EPOC by applying the best-ranked stimulus and an assessment protocol was used in order to evaluate the performance of users driving the wheelchair with the BCI, with satisfactory results