Research On Eeg Differences Based On Emotion Recognition Of Left-and Right-Handers

Therefore,when used EEG signals for emotion recognition,the emotion recognition performance of right-handers may be better than left-handers,left-handed people may be at a disadvantage in emotional EEG response;when emotions were included,righthanded people EEG patterns may have greater similarity,while left-handed people may have more individual differences in EEG patterns,which means that the EEG patterns of left-and right-handers should be different in emotional EEG response.Research on EEG differences based on emotion recognition of left-and right-handers,the paper found that the EEG patterns of left-and right-handers may be different in emotional EEG response.This study is conducive to promoting related research on emotion recognition among different handers,and can help researchers better understand how emotions are processed in the brain.Handedness refers to the bias of one hand,and right-handedness is called righthanders,otherwise it is called left-handers.Left-handers are more prone to psychological problems such as anxiety and depression than right-handers,and left-handers are more prone to diseases such as autism,schizophrenia and emotional disorders.Therefore,recognizing the emotions of left-handers in time is of great significance in preventing these psychological and physical problems.Left-and right-handers have different brain structures and functions,the left-and right-handers have different brain response mechanisms when they receive the same emotional stimuli.However,consulting the literature found that most of electroencephalogram(EEG)-based emotion recognition studies were based on right-handers,and few researchers have analyzed whether there were some differences in EEG signals for emotion recognition between left-and righthanders.For the above reasons,the paper creatively proposed a study of EEG differences based on emotion recognition of left-and right-handers.This study collected the EEG signals of several students,which was stimulated by pictures,the emotion recognition as a tool was used for analyzing the differences in left-and right handed subjects’ EEG signals.The paper carried out the following work:(1)EEG data were collected from 42 subjects(21 right-handed subjects and 21 lefthanded subjects).The experimental design task of the paper was to let the two groups of subjects watched the emotion-evoked pictures(40 positive pictures and 40 negative pictures)separately,induced their positive and negative emotion,and collected the EEG signals throughout the process.Prior to the start of the experiment,subjects were recruited,the subjects who met the criteria were screened and grouped by the state-trait anxiety inventory(STAI)and the left-and right-handed questionnaire.In the experiment,subjects watched positive and negative pictures in turn,after each picture was viewed,subjects were required to score the valence and arousal according to the emotional feelings at the time.After the acquisition of the EEG signals were completed,a series of pre-processing operations were needed on the collected original EEG signals.The pre-processing operations mainly included downsampling,channel positioning,re-reference electrode settings,ophthalmology,filtering and data segmentation.In this study,independent component analysis(ICA)was used for removing the electroencephalogram artifacts of the original EEG signal,the EEG signal was filtered by a band-pass filter to filter out the four target bands needed for the study,they were: theta band(4–8 Hz),alpha(8–12 Hz),beta(12–30 Hz),and gamma(30–47 Hz).(2)Extracted the EEG features related to emotions,and established a left-and right-handed subjects emotion recognition model.In the paper,two frequency-domain features of EEG signals were extracted,namely,power spectral density and total energy.The sequential backward selection method was applied to features reduction,K-Nearest Neighbor(KNN)classifier,decision tree classifier and the "leave-one-subject-out" crossvalidation method established an emotion recognition model for left-and right-handed subjects,and used the emotional recognition model to classify positive and negative emotion,the F1 score was calculated to measure the performance of the emotion recognition model.(3)Analyzed and comprehensively discussed the emotion recognition results and F1 scores of left and right hands.The gamma band of right-handed subjects obtained the best emotion recognition result of 69.33%.The best classification of left-handed subjects was 65.30% in the gamma band.For the F1 score,both left-and right-handed subjects obtained the highest F1 scores in the gamma band,which were 71.53% and 65.96%.We found that the left-and right-handed subjects had better emotion recognition performance in the high-frequency band(beta rhythm and gamma rhythm)of the EEG signal than the low-frequency band(theta rhythm and alpha rhythm),which was consistent with previous findings of emotion recognition results based on the right-handers.However,from the results can be seen that the right-handed subjects’ best emotion recognition results and F1 scores were both about 4%-6% higher than the left-handed subjects.