EEG system design for VR viewers and emotions recognition

BACKGROUND: Taking advantage of virtual reality today is within everyone's reach and this has led the large commercial companies and research centers to re-evaluate their methodologies. In this context the interest in proposing the Brain Computer Interfaces (BCIs) as an interpreter of the personal experience induced by virtual reality viewers is increasing more and more. OBJECTIVE: The present work aims to describe the design of an electroencephalographic system (EEG) that can easily be integrated with virtual reality viewers currently on the market. The final applications of such system are several, but our intention, inspired by Neuromarketing, wants to analyze the possibility of recognize the mental state of like and dislike. METHODS: The design process involved two phases: the first relating to the development of the hardware system that led to the analysis of techniques to obtain the most possible clean signals; the second one concerns the analysis of the acquired signals to determine the possible presence of characteristics which belong and distinguish the two mental states of like and dislike, through basic statistical analysis techniques. RESULTS: Our analysis shows that differences between the like and dislike state of mind can be found analyzing the power in the different frequencies band relative to the brain's activity classification (Theta, Alpha, Beta and Gamma): in the like case the power is slightly higher respect the dislike one. Moreover we have found through the use or logistic regression that the EEG channels F7, F8 and Fp1 are the most determinant component in the detection, along with the frequencies in the Beta-high band (20-30 Hz).