Eeg Studies Of Musical Emotion Perception

Emotive process and cognitive process are both the function of our brain. There exists a certain internal connection between music and emotion. The electroencephalogram (EEG) study of musical emotion perception is the study of emotional brain mechanism using music to evoke emotion and using EEG data analysis as the tool. It is a newly emerging interdisciplinary area that integrates psychology of music, psychology of emotion, physiology and neuroscience. In this work, we designed an experiment to collect musical emotion EEG data and mainly focused on the linear and nonlinear analysis of the data. The main contents are as follows:1. The EEG alpha (α) powers of the four music-emotional conditions of'fear, joy, happiness and sadness'were analyzed. The results showed that: The alpha power was closely related to the emotional valence and intensity; The EEG alpha power of not only the frontal area, but also the parieto-occipital area increased as the emotional intensity decreased; The EEG alpha power was related to the emotional valence and modulated by the emotional intensity.2. The EEG data of the four different conditions of'eyes-closed rest, fear music, happy music and sad music'were analyzed by a detrended fluctuation analysis (DFA) method. The results showed that: There existed a scaling behavior in EEG; There were two scaling regions with two different scaling exponents ofβ1 andβ2, which respectively represented the complexity of higher frequency activity above theαband and lower frequency activity below theαband; The activities of these two frequency bands were modulated by the musical emotional intensity; As the emotional intensity decreased, the complexity of the higher frequency activity increased and that of the lower frequency activity decreased; The complexity of the higher frequency activity was weakly correlated with the'Approach-Withdrawal'emotional model of the brain; Both the fear and sad music made certain EEG differences compared with the baseline condition of eyes-closed rest.3. We analyzed the EEG data from two conditions of'eyes-closed rest and eyes-open rest'using the above-mentioned DFA method. The results showed that each of the two scaling exponents could distinguish states of eyes-closed rest and eyes-open rest easily. Specifically, theβ1 value of eyes-closed rest condition was always significantly larger than that of eyes-open rest condition, while the value ofβ2 was the contrary.Our work shows that the change of EEG indeed can reflect the different brain states during different musical emotion perception. EEG nonlinear analysis is very important for the study of musical emotion perception. DFA is a valuable nonlinear method for EEG data analysis and it is potentially of value in the study of different brain states.