Self-paced emotional imagery-based brain computer interface system

Abstract

This paper proposed a self-paced emotional imagery based noninvasive Brain-computer interface (BCI) system. Electroencephalography (EEG) was used to observe brain phenomenon and regions during imagery positive and negative emotions. Absolute power at peak frequency of EEG bands from quantitative EEG analysis was used to create a parameter two classes of emotion by using Linear discriminant analysis (LDA) classifier. The study of brain response via EEG supports previously proposed EEG-based emotion recognition. The results showed the proposed algorithms achieved averaged accuracy rate of 53.3–83.3%. The proposed system can be used for real-time BCI. The aim is an assistive devices and emotion monitoring based on BCI that can practically use in clinical applications.