Asymmetry Of Brain Emotional State Based On EEG And VR Techniques

The brain is the highest part of the human central nervous system.In the past decade,the study of brain functional state has been a hot topic for human beings to explore the mysteries.With the rapid development of brain science,neuroscience and information science,the study of brain emotional state using Electroencephalography(EEG)data has become a more active field,among which non-implantable electrodes have been favored by researchers because of their convenient and non-invasive advantages.Emotion research using EEG characteristics can reflect the process of emotion generation and regulation better.At the same time,the development of artificial intelligence has laid a good foundation for EEG and human-computer interaction,and the exploration of the asymmetry of brain emotional state is gradually diversified.In this paper,Power Spectral Density(PSD),hemispheric asymmetry index,PhaseLocked Value(PLV),brain network index and graph theory analysis were selected as tools to study the asymmetry of brain emotional state.Twenty subjects with an average age of 23.9 years and no visual and hearing defects were selected as experimental subjects.Combining EEG and Virtual Reality(VR)technology is a relatively advanced research direction currently.It uses visual and auditory senses to interact with objects in a virtual three-dimensional environment.Twelve films that can induce positive,neutral and negative emotions are selected.Under the stimulation of different VR films,EEG signals are collected,and on this basis,subjects are asked to score the degree of emotion induced by the same stimulation material as the priority of data selection.Then the average power spectral density of each frequency band is extracted,and the asymmetry of the left and right hemispheres is observed.Calculate the reference index of hemispheric asymmetry and compare the reference index in different frequency bands and brain regions.The results showed that the energy of frontal region in positive emotion was larger than that in negative emotion in alpha band,and showed a positive and neutral negative decreasing trend,indicating that alpha band was the main characteristic band of emotional activity,and the activity of frontal region was closely related to emotion.Brain network is a result of the nonlinear dynamic behavior of neuron clusters.In this paper,we use PLV to construct the functional brain network of the left and right hemispheres under different emotions in the alpha band,and then select the appropriate threshold through the small-world index,combined with functional connection and brain network analysis.The results showed that the phase synchronization between the left hemisphere of the frontal region under positive emotion was higher than that of the right hemisphere under positive emotion,and the phase synchronization of the right hemisphere was higher than that of the left hemisphere under negative emotion,and there was significant difference.In the network index analysis,the clustering coefficient of the right hemisphere of the negative emotion is high,the characteristic path length is low,and there is a significant difference in the more effective information communication between the electrode pairs,while the positive emotion has higher information transmission efficiency in the right hemisphere.The same conclusion is drawn from the results of graph theory analysis of brain networks.It is inferred that the frontal region of the left hemisphere dominates under the positive emotion and the frontal region of the right hemisphere dominates under the negative emotion.This study mainly combines EEG and VR technology to explore hemispheric asymmetry in different emotional states,which is not only an important topic in cognitive science,neuroscience,psychology and other fields,but also closely related to our living status and quality.This paper provides a new idea for the identification of different emotions,a new method for the diagnosis,treatment of brain functional diseases and mental diseases.