EEG-based Emotional Differences Study Of Premenstrual Syndrome In Women

With the increasingly fierce and cruel competition in society,women today are facing greater pressure.Patients with premenstrual syndrome are more and more common,which seriously affect women’s quality of life and work efficiency,and have serious negative effects on women’s physical and psychological effects,interpersonal communication and reduces the quality of life and work.Serious symptoms can even lead to suicidal thoughts,resulting in family tragedies.Studies show that the severity of female premenstrual syndrome is greatly affected by emotions,negative emotions can increase the degree of female disease,and female premenstrual syndrome can also be reflected by emotional changes.Therefore,the emotional induction study of women with premenstrual syndrome is of great significance.At present,most studies on premenstrual syndrome are based on EEG signals or MRI research,but because the main characteristics of premenstrual syndrome are stress,anxiety,emotional instability,etc.,many studies on anxiety are based on ECG signals,so this paper analyzes the physiological patterns of premenstrual syndrome women and normal women based on ECG signals.In addition,in previous studies,emotional pictures or facial expressions were used as stimulating materials to study the emotional attention bias of premenstrual syndrome.However,this paper attempts to use emotional video as a stimulating material to study the emotional induction difference of premenstrual syndrome in different cycles of menstruation.The details are as follows:(1)The experimental design task of this paper is to let the two groups of subjects watch three-segment emotion-inducing video in the luteal phase and follicular phase respectively,and collect the ECG signals in the whole process.Subjects should be recruited before the start of the experiment,and the subjects should be screened and grouped by the menstrual questionnaire and the Beck Depression Scale.Those who meet the criteria can start the experiment and fill out the two-month daily problem severity record.First,the subjects sat down as baseline state.Then the subjects watched the neutral video,the positive video,and the negative video to perform the emotionally induced task.Then the subjects sat and regained their minds,and finally experimenter collected the self-rating scale.The ECG signals collected by the experiment were denoised by wavelet transform,and then the R peaks were located by adaptive sliding time window positioning algorithm.A few errors are manually corrected,and then the RR interval sequence could be calculated.Then the feature extraction of the RR interval sequence was carried out,including: traditional features,nonlinear features,and Krawtchouk moment features.Then the backward selection algorithm was used for feature extraction,and the extreme learning machine and the one-test method cross-validation was used for identification and classification,and the physiological parameters of different cycles of menstruation in the same group and the same period of menstrual in different groups were analyzed separately,and the optimal solution of the feature set corresponding to the physiological model was obtained.(2)According to the self-rating scale,we can find that the subjects are generally biased towards negative stimuli during the luteal phase,and tend to favor positive stimuli during the follicular phase.The scores of the daily problem severity scores show that the subjects are more prone to feel mental and physical discomfort before menstruation;and for all subjects of the group N and the group P,the scores of the two groups had obvious differences and premenstrual syndrome women are more likely to have mental and physical discomfort.The internal consistency of the overall daily problem severity record is excellent.The statistical analysis method is used to analyze the physiological parameters of different groups and different cycles of menstruation.The results shown that compared with the same group in different cycles of menstruation,the emotional state of subjects of group N and the group P in the luteal phase and the follicular phase can all be distinguished by the characteristics of the RR interval of the ECG,and the discrimination of the group P is better than that of the group N.In addition,when watching the positive video,the feature values in the luteal phase were larger than those in the follicular phase,indicating that women weremore likely to induce positive emotions during the follicular phase.For the different groups of subjects in the same period of menstruation,by analyzing the value differences of all features between the two groups,we found that ECG can characterize the autonomic nervous activity mechanism of subjects.For the recognition classification results,the recognition rate of N group of subjects induced by neutral video emotion in the luteal phase and follicular phase is 80%,the recognition rate induced by the positive video is 85%,and the recognition rate induced by the negative video is 78.95%;the recognition rate of P group of subjects induced by neutral video emotion in the luteal phase and follicular phase is 82.5%,the recognition rate induced by the positive video is 80%,and the recognition rate induced by the negative video is 77.5%;In addition,the recognition rate of N group subjects and P group subjects induced by neutral video in the luteal phase is 92.5%,the recognition rate of N group subjects and P group subjects induced by positive video is 75%,and the recognition rate of N group subjects and P group subjects induced by negative video is 86.84%;the recognition rate of N group subjects and P group subjects induced by neutral video in the luteal phase is 87.5%,the recognition rate of N group subjects and P group subjects induced by positive video is 80%,and the recognition rate of N group subjects and P group subjects induced by negative video is 84.21%.From which we could know that in addition to the low recognition rate of negative video and the low recognition rate of positive video-induced emotional state in the luteal phase of the two groups,the former is consistent with the results of previous studies,and the latter reason cannot be concluded now.Later studies can collect experimental data for verification and analysis,the recognition rate of other cases is high.In this paper,we studied the emotionally induced physiological patterns of women with premenstrual syndrome and normal women based on ECG signals.It was found that certain physiological parameters can be used to distinguish the group and the menstrual cycle to some extent,and have a reference value to make a diagnosis of premenstrual syndrome in the future.