EEG Study Of Emotion Regulation By Aerobic Exercise

In recent years,with the acceleration of the pace of life,social competition has become increasingly fierce and cruel,and people's negative emotions have increased day by day.Excessive negative emotions will have a great impact on people's physical and psychological,and easily lead to abnormal situations such as irritability,insomnia,memory loss,and inability to concentrate.Long-term negative emotions will seriously affect daily life and work and study,and even lead to symptoms such as decreased immunity and endocrine disorders,and even severe cases of depression.These are uncontrollable for some negative situations in daily life,but we can strengthen the regulation of negative emotions.At present,exercise has been used as an adjuvant treatment for depression.Many depressed patients have improved their condition through exercise,which shows that exercise does have a role in regulating emotions.However,most of the sports treatment of depression are limited to the detection of the scale.The use of the scale to judge the emotional changes of patients is too subjective and the observation period is too long.It is impossible to detect the emotions of the subjects in real time in order to prescribe the right medicine.This project will analyze the effects of aerobic exercise on emotional regulation from the level of electroencephalogram(EEG),and analyze the effect of aerobic exercise on emotional regulation from the EEG signals.To this end,this study will use two experiments to analyze the differences in EEG between people without exercise habits,acute aerobic exercise,andlong-term aerobic exercise in different situations,and analyze the effects of aerobic exercise on emotions from the EEG signals.The effect of regulation is to propose a new detection method for aerobic exercise to prevent and control negative emotions.The main findings and conclusions are as follows:(1)Acute aerobic exercise contrast experiment.The design task of this experiment is to let the two groups of participants watch the emotion-evoked pictures before and after the activity,and record the EEG throughout.In this experiment,80 non-exercise people are recruited as experimental subjects through the International Physical Activity Questionnaire(IPAQ),and they are randomly divided into exercise groups and control groups.The experimental process first is the subjects sit quietly for 1 minute to obtain their EEG signals as a baseline,then the subjects will be watching the pictures to induce emotions.After each picture played,the subjects would fill in the PAD-P scale to subjectively score the pictures;during the activity,The exercise group will perform a 20-minute moderate-intensity cycling exercise,while wearing a sports bracelet to measure its real-time heart rate to monitor the exercise intensity;repeat the first step again after the heart rate recovers.The original EEG is pre-processed in the frequency domain using Fourier transform.Finally,the frontal asymmetry of the EEG is calculated from the power spectrum to analyze the similarities and differences between the two groups of participants.(2)Long-term aerobic contrast exercise.This experiment recruit 40 participants with long-term exercise habits and 40 participants without exercise habits through the physical exercise rating scale.Among them,the participants of long-term exercise habits are all students in the School of Physical Education of Southwest University,all of which are maintained aerobic exercise habits for more than three years.The content of the experiment is to allow two groups of subjects to watch pictures of three valence for emotional evoking,and record their EEG throughout.After pre-processing,firstly select the EEG signal located in the frontal region channel;secondly,denoise the signal and intercept the frequency band by wavelet transform,extract features on each frequency band according to previous related research;then select SBS and ten-fold cross-validation to construct the best emotion recognition model,which is used to distinguish the differences between the two groups in emotion recognition.The main findings and conclusions are as follows:(1)Combined with the results of the two experiments,it is found that aerobic exercisehas a significant positive effect on emotion regulation.In the acute aerobic exercise experiment,the frontal asymmetry of the exercise group is significantly increased compared to the previous test after the acute aerobic exercise,but the control group has no significantly increased,indicating that the aerobic exercise has a emotion regulating effect.In the long-term aerobic exercise experiment,the emotion recognition rate of the subjects in the long-term exercise habit group is higher than that of the non-exercise group,and the emotion perception ability of the subjects in the exercise habit group is significantly higher than that of the non-exercise group.Aerobic exercise has a positive effect on emotion regulation.(2)For the acute aerobic exercise contrast experiment,it is found that under different situational stimuli,the way that acute aerobic exercise regulates emotions is different.Under negative stimulation,the energy value of the left brain in the exercise group decreased,while the energy value of the right brain increased,exercise enhanced the increase of positive emotions and suppressed the generation of negative emotions to regulate emotions;The energy values of both the left and right brains of the subjects decreased,and exercise stimulated the generation of two emotions.(3)For long-term aerobic exercise habits contrast experiments,it is found that the support vector machine(SVM)algorithm is better than the k-nearest neighbor(KNN)algorithm.Through backward feature selection and ten-fold cross-validation,the recognition rate,Receiver Operating Characteristic(ROC),and Area Under ROC Curve(AUC)of the SVM algorithm are better than the KNN algorithm.(4)For the frequency bands selected(alpha,beta),alpha and beta waves both contribute to emotion recognition,and beta waves are better than alpha waves.The positive emotion recognition rate in all band under the SVM algorithm is 73.54%,the positive emotion recognition rate in the beta band is 69%,and the positive emotion recognition rate in the alpha band is 64.42%.