A Combined Personality Self-rating Scale And EEG Study On Schizophrenics And Their First-degree Relatives

Background:Regardless of how the style of war changes in the future,mankind has always been the determining factor.From the first day of the birth of military psychology,psychological selection has always been the focus of military psychology research and application.Due to the particularity of the military service system and the recruitment system in China,the method of basing on the self-reported personality questionnaire is still the main form of psychological selection for young people in China.Although the self-reported questionnaire is one of the most recognized methods of personality detection in the world and it is widely used in large-scale psychological selection,it's detecting validity is still affected by many factors,including self-perception accuracy,social approval and so on.These shortcomings have always been the problems that the psychological selection application field is trying to solve.With the advancement of modern science and technology,the foreign military has begun to try to use cognitive neuroscience technology to make up for the lack of self-reporting methods and to improvethe accuracy of psychological selection.So the multi-quality fusion method based on physiological and psychological signals emerged as the times require.Schizophrenia is one of the mental disease which seriously affect combat effectiveness and safety in the army.With the application and development of the psychological examination of conscription in China,the incidence of schizophrenia in the army has dropped significantly,but there are still certain high-risk groups,which are at higher risk of transforming into patients in the military environment.As the national conscription psychological selection center,we established China's first psychological selection system and standards in 2006,and have been working to improve the application of psychological selection.The use of multi-quality fusion technology to improve the recognition rate of high-risk for schizophrenia is an important research direction to enhance the psychological quality of our military personnel.Subjects and methods:This study was divided into four parts.For the first three parts,first-degree relatives of schizophrenia were selected as the high-risk group(HR),patients with schizophrenia(SZ)as the disease group,and healthy people(NC)as the normal control group.In the fourth part,schizophrenia patients were selected as the disease group and healthy people as the control group.The first part collected three groups' short-term(30s)and long-term(3min)resting EEG signals.Exploring whether short-term resting EEG could be used to extract EEG differences in three groups of people.The second part collected the EEG signal when doing self-rating scale including the overall status of doing the questionnaire and doing some specific items.The third part carried out machine learning using these abnormal indicators,and came out a multi-quality fusion index system,which could be used to classify the three groups of people more accurately,thus providing a theoretical basis and technical support for improving the effectiveness of psychological screening for conscription.The fourth part used self referential task to analyze the event-related potential(ERP)and event-related desynchronization/synchronization(ERD/ERS),exploring whether theEEG abnormality when doing self-rating scales in patients with schizophrenia was caused by the self-evaluation function decline.Results:Study 1 compared resting EEG band power of SZ group(35 cases),HR group(35cases)and NC group(33 cases),and found three main results:1.Short-term(30s)EEG abnormalities could be used to distinguish three groups of people.In the eyes-open resting condition,the EEG band power differences among three groups in short-term and long-term(3min)resting state were consistent.The delta,theta and alpha band power of SZ group at midline region were significantly higher than the NC group,meanwhile there was no difference between HR group and NC group.There was no significant difference between SZ or HR group and NC group in eye-closed condition.The above results suggested that short-time eye-open resting state could induce differences in EEG power among the three groups,indicating the feasibility of short-time eye-open resting EEG detection joining the conscription psychological detection process.2.The abnormal gamma1 band power in the resting state could be used as an indicator of the high risk of schizophrenia.The power of gamma1 band in the left centro-parietal region of SZ group was significantly lower than that in NC group under the condition of 3 min-eyes-open resting state.Gamma1 band power at the left and right centro-parietal region of HR group and SZ group were equal,while which was significantly greater in the left centro-parietal region of NC group than right.The interaction between the group and hemisphere was significant,suggesting that gamma1 band power asymmetry disappeared in HR group and SZ group.The resting-state gamma1 band power in HR group was significantly negatively correlated with their schizotypal personality traits.These data indicated that the resting gamma1 band power abnormality can be used as an indicator of the high risk of schizophrenia.3.Long-term resting-state alpha and gamma1 band power change had a linear regression relationship with time,and NC group had the largest change slope,SZ group had the smallest or displays the opposite change,and HR group was between NC group and SZ group.In the eyes-open resting state,the alpha band power of the three groupsincreased with time,and the slopes were different: NC group>SZ group>HR group.In the eyes-closed resting state,NC group and HR group showed a downward trend with time,and the slope was different: NC group>HR group,but SZ group showed an increasing trend with time.In the eye-open resting state,the gamma1 band power of HR group and SZ group decreased with time,the slope: HR group>SZ group,and the gamma1 power of NC group decrease rapidly in the first 30 s resting state;In the eyes-closed resting state,all three groups had no linear relationship with time.Study 2 analyzed the task-state EEG data of the three groups and obtains four main results:1.The differences in the overall EEG band power of the three groups when doing self-rating scale were similar to that in resting state,but the scale task better induced the differences between HR group and NC group.The delta and theta band power of HR group were significantly lower than NC group and SZ group,and NC group had lower power than SZ group.In order to eliminate baseline EEG differences among three groups,covariance analysis was used with 30s-resting-state EEG band power as covariate to compare EEG differences induced by self-rating task,and we found that schizophrenia gamma1 power at midline site of HR group was significantly higher than NC group and SZ group.2.The changes of alpha and gamma1 band power showed a linear regression relationship with the self-rating task process.The alpha band power of NC group and SZ group showed an increasing trend,and the increase rate was the same,while HR group did not show a linear relationship.The power of gamma1 band in both NC group and HR group showed a decreasing trend with the task process,and the slope was different: NC group>HR group,while SZ group showed an increasing trend,suggesting that the cognitive process of SZ in the self-rating task process was very different.3.The different items of the scale can induce different EEG power in the three groups.The scores of HR group in the dimension of separation(DIT)increased,and the power of gamma1 band also increased on DIT items,suggesting the sensitivity of HR group to the DIT items.When HR group answered the forward items,the gamma1 bandpower was lower than that of the reverse items,and when SZ group answered the forward items,the power of theta and alpha was higher than that of the reverse items,suggesting that the two groups of people had different sensitivity to the forward and reverse items in the scale compared with NC group.4.SZ group answering "no" induced higher frontal theta band power than answering "yes",while HR group and NC group not,indicating the abnormalities of schizophrenia patients in retrieving,extracting and processing self-irrelevant information.Based on the results of the first two studies,study 3 came out a multi-quality fusion index system with high discrimination ability by machine learning:A total of 7056 indicators were selected to enter machine learning.These indicators then used the leave-one-out cross-validation(LOOCV)method to compare any two of the three groups,each extracting 34 indicators,and then using the support vector machine(SVM)and LOOCV methods to perform three-category verification.It is proved that the recognition rate of these indicators for the three groups of people can reach a high level of70.9%.There were a lot of indicators between SZ group and NC group,including the delta,theta and alpha band power of the resting state and the scale task state,the slope of the gamma1 power in the 30 s resting state and scale score.The indicators between HR group and NC group was concentrated on the delta and theta power when doing the self-rating scale.The indicators between HR group and SZ group included delta,theta,and alpha power in the frontal region of the eye-open resting,and the delta,theta,and alpha power of most electrodes in the task-state.Study 4 analyzed the ERP and ERD/ERS of 33 schizophrenic patients and 28 normal subjects and found three main results:1.The self-referential memory score of in SZ group decreased significantly,suggesting that their self-referential function is impaired.The amplitude of P3 component of ERP was relatively reduced,and the degree of desynchronization of low beta band(13-15H)was also decreased in SZ group,suggesting that P3 and low beta band power could be used as EEG indicators of self-referential processing.2.Negative personality words induced a greater positive slow wave(600-1000ms)inNC group when evaluating self,but not in SZ group,while evaluating others,negative personality words induced a greater positive slow wave in SZ group but not in NC group,suggesting that patients with schizophrenia had a reduced ability to distinguish self from others.Conclusions:In summary,this study obtained two main results.Firstly,verified the feasibility of30 s resting-state EEG monitoring to be added into psychological selection system for conscripts and came out a index system with a high recognition rate to classify schizophrenia patients,high-risk of schizophrenia and the normal people based on EEG band power during resting state and scale task state,which could be used to further improve effectiveness of psychological selection for conscripts.Secondly,verified that self-evaluation cognitive processing was impaired in schizophrenia patients and found the related EEG evidence.