Shaping of neuronal activity through a Brain Computer Interface

Abstract

Neuronal responses are human actions which can be measured by an EEG, and which imply changes in waves when neurons are synchronized. This activity could be changed by principles of behaviour analysis. This research tests the efficacy of the behaviour shaping procedure to progressively change neuronal activity, so that those brain responses are adapted according to the differential reinforcement of visual feedback. The Brain Computer Interface (BCI) enables us to record the EEG in real time using signals to manage a computer or a virtual environment. In this study, participants had to move a virtual reality car to the right or left-hand side of the road, using only their EEG activity. A simultaneous mixed design with concurrent base-line A-B was applied with two types of neuronal and cognitive activities among groups. A group of four participants carried out the BCI task with visual reinforcement contingencies or feedback with a trial and error procedure; in addition, a group of five participants carried out the BCI task by a shaping procedure, using augmented visual feedback for correct responses and attenuated feedback for wrong responses. Furthermore, the shaping criteria feedback decreased progressively according to the progress of participants’ responses. The results showed a progressive control using BCI car movement in the shaping group; whereas the control group showed difficulties in learning. The parallelism between the shaping criteria and the change in EEG waves showed the possibility of influencing directly on the neural activity with behavioural procedures.