Objective The continuous and discrete Bayesian network models of serum uric acid and related metabolic indexes were established to explore the influencing factors of serum uric acid and compare the characteristics and advantages of the results of the two networks.Methods A total of 4646 patients with serum uric acid and metabolic diseases from chronic disease monitoring in Shanxi Province in 2015 were selected.IPCB algorithm was used to establish a continuous Bayesian network of serum uric acid.Meanwhile,the above indicators were discretized,and MMHC was used to establish a discrete Bayesian network of high uric acid.Results The discrete Bayesian network found 14 edges,in which triglyceride and diastolic blood pressure abnormalities were directly related to the occurrence of high uric acid,leading to the occurrence of high uric acid.Age was an indirect factor;Age,TG,LDL,HDL,SP and DP are directly related to uric acid level.With the increase of age,TG and LDL and the decrease of HDL,uric acid level increases,while the increase of uric acid level leads to the increase of SP and DP.TC is indirectly related to uric acid.Conclusion The two network models adapt to different data types,but the continuous Bayesian network has more direct correlation factors and better overall explanatory degree.

Objective To compare the existing evaluation methods of sample's representativeness and provide reference for selection of sample representativeness evaluation methods in clinical research.Methods Simulate the target population of lung cancer patients and select samples with different sample sizes and different degrees of deviation based on the distribution of traits of lung cancer patients in China and the actual situation of sample screening in domestic clinical studies.Calculate sample representativeness using the existing evaluation methods of sample's representativeness,and calculate estimation deviation(bias).By constructing the correlation model between the measured value of each method and bias,analyze the accuracy and stability of each method.Results The overall structural variance rate RV1、RV2、C-statistic based on propensity score、SGCR and K-S distance could well measure the degrees of deviation of different samples.Under different sample sizes,the R2 of RV2 and RV1 are greater than 0.90,and R2 of C-Statistic、SGCR and K-S distance were greater than 0.80.Conclusion The overall structural variance rate is more accurate and stable because the traits weight is taken into account.In particular,RV2 can better measure the representativeness of samples with different degrees of deviation and accurately reflect the estimation deviation.However,when it is difficult to obtain the feature importance information,the reliability of the representative measurement of SGCR as well as C-statistic and K-S distance used the propensity score-based method are acceptable.

Objective The adaptive synthetic sampling(ADASYN)algorithm and category inverse proportion weighting method weighting method were used to balance the datasets,then multi-classification prediction of cognitive normal(CN),mild cognitive impairment(MCI),and Alzheimer's disease(AD)combined with classifiers were performed.Methods Data were obtained from the Alzheimer's Disease Neuroimaging Initiative(ADNI)database,which was filled in missing values by random forest(RF),and feature subsets were selected by elastic net(EN).We chose ADASYN algorithm and category inverse proportion weighting method processing the category imbalance data,and four models were constructed by combining RF and support vector machine(SVM)respectively:ADASYN-RF,ADASYN-SVM,weighted random forest(WRF),and weighted support vector machine(WSVM).We evaluated the classification performance by macro-P,macro-R,macro-F1,ACC,Kappa value and area under the receiver operating characteristics curve(AUC).Results ADASYN-RF had the best classification performance(Kappa=0.938,AUC=0.980),followed by ADASYN-SVM.The most important classification features obtained using ADASYN-RF were CDRSB,LDELTOTAL,and MMSE,which have been clinically validated.Conclusions Both the ADASYN algorithm and the category inverse proportion weighting method can assist in improving classifier performance,and the ADASYN algorithm is superior.

Objective To explore the comorbidity patterns of different gender patients with chronic obstructive pulmonary disease(COPD)and provide theoretical basis for the hierarchical management of COPD patients.Methods The patients with chronic obstructive pulmonary disease from December 2011 to May 2020 in the Second Hospital of Shanxi Medical University were selected as the study subjects.The cross-sectional diagnostic information of patients was collected to establish a comorbidity network.The community of the comorbidity network was divided by Louvain method,and the comorbidity types,combinations and prevalence rates of patients with different genders were analyzed and compared.Results Our study manifested that there were differences in the prevalence of single chronic disease and comorbidity in patients with different genders.The corresponding comorbidity network and community structure of each patient group were also different.Conclusion The comorbidity network of male patients is more concentrated and complex than that of female patients.The comorbidity pattern of male patients is more complex and diverse than that of female patients,and the relationship between diseases is closer.It verifies that the comorbidity pattern of different gender patients is different.In patients with chronic obstructive pulmonary disease,more attention should be paid to male patients in order to reduce the incidence of comorbidity.

Objective An autoregressive integrated moving average model(ARIMA)was established to predict the incidence of tuberculosis in Heze in 2022.Methods Based on the monthly registered incidence of tuberculosis patients in Heze city from 2010 to 2020,the optimal ARIMA model was established to predict the incidence in 2021 and compare with the actual value,so as to evaluate the prediction effect and predict the incidence trend in 2022.Results The incidence of tuberculosis in Heze city showed a decreasing trend year by year,with certain seasonal changes.The optimal model was ARIMA(0,1,1)(1,1,1)12,the fitting results showed that the overall prediction error rate was 2.59%and the mean absolute percentage error was 17.76%in 2021.The number of cases predicted in 2022 was 1644,which continued to show a downward trend and the epidemic situation was stable.Conclusion ARIMA(0,1,1)(1,1,1)12 model can better predict the short-term incidence trend of tuberculosis in Heze city,but it should be modified according to the changes of monitoring data to improve the prediction accuracy.

Objective To investigate the classification effect of deep neural networks in irregular time series data,and to predict the recurrence risk of 362 patients with diffuse large B-cell lymphoma(DLBCL)in a hospital in Shanxi from 2014 to 2020.Methods A total of 362 diagnosed DLBCL patients who achieved complete remission after initial chemotherapy were collected retrospectively,and the recurrence risk was predicted within the next two years.First,LASSO regression was used to screen the variables.Then a deep neural network model of irregular time series data based on GRU-ODE-Bayes was constructed and compared with some traditional models and other deep neural network models.Results Among all the models under study,the traditional models do not perform as well as the deep neural network models in classification.The GRU-ODE-Bayes model was the best,with AUC of 0.85,sensitivity of 0.84,specificity of 0.71,and G-means of 0.77.Conclusion Compared with other models,the GRU-ODE-Bayes model can predict the recurrence of DLBCL patients more accurately.It could benefit the individualized treatment for patients and decision-making for physicians.

Objective Based on the health level,the factors affecting female labor participation were understood,and the years of delayed employment for middle-aged and elderly women were estimated.Methods Using the 2011,2013,2015,and 2018(CHARLS)data,Women aged 45～69 years were selected as subjects.The CMR health measurement model was used to estimate the delayed working years of women with different characteristics.Results The average labor force participation rate for women aged 45～69 is 66.93%,with the poorer the level of health,the more likely they are to drop out of the labor market.women aged 50～69 have an additional working capacity of 3.29～3.61 years;the urban women's group aged 50～69 have an additional working capacity of 7.86～8.36 years,and the junior high school and above women's group aged 50～69 have an additional working capacity of 5.89～6.33 years.Conclusion Women in China still have great potential for healthy work to be tapped,especially for middle-aged and elderly women with good health in cities and towns and with high education.Reasonably formulating retirement policies for women of different groups and gradually realizing the retirement of men and women at the same age will help to further promote the workplace gender equality and healthy aging.

Objective To develop the satisfaction scale on built environment of hygienic city initiative and to evaluate its reliability and validity.Methods By using the qualitative research methods including policy analysis,expert consultation,personal interviews,and quantitative method with statistical analysis,the content of the scale was finally determined.Principal component analysis of exploratory factor analysis was used to construct the dimensions of the scales.The reliability and validity of the scales were evaluated with internal consistency reliability,split-half reliability,content validity,surface validity,and structural validity evaluation methods.Results A satisfaction evaluation scale comprising of 4 dimensions and 20 items was established.The Cronbach's α coefficient of the satisfaction evaluation scale was 0.91,and the Spearman-Brown splitting coefficients of the scale was 0.851.The results of confirmatory factor analysis showed that all test values were in the standard range,which means that content validity of the scale was good.Conclusion The satisfaction scale on the built environment of hygienic city initiative developed in this study has a good reliability and validity.The practical verification of the scale need to be carried out to further explore the applicability of the scale.

Objective To understand the transition of cognitive states among older adults and its influencing factors in order to provide a scientific basis for early intervention of cognitive impairment.Methods Based on ADNI data,subjects aged≥60 years and had at least one follow-up were included in this study.Cognition was divided into(normal cognition,CN),(mild cognitive impairment,MCI)and(Alzheimer's disease,AD).A multi-state Markov model was used to explore the transfer rules among the three cognitive states,and to estimate the effects of age,gender,education level,APOE4 allele and marital status.Results A total of 10073 records of 1907 subjects were included.Among them,the baseline age was 73.9±6.45 years old,and the average follow-up was 6.6 times.The intensity of transition from MCI to AD was 2.88 times of its reversal to CN(0.118 vs.0.041).For CN subjects,the probability of remaining at CN after 10 years decreased by 33.6%compared with that after the first year,while the probability of developing MCI and AD increased by 3 times and 55 times,respectively.For MCI subjects,the probability of staying at MCI after 10 years is only 0.238,while the probability of developing AD is 3.95 times that of its reversal(0.608 vs.0.154).In addition,the average residence time of CN and MCI is 18.43 years and 6.30 years,respectively.Multivariate analysis showed that male,older age,low educational level and carrying APOE4 allele increased the risk of MCI.Conclusion The cognitive function of the elderly showed a downward trend overtime.Older age,low educational level,and carrying APOE4 are risk factors for cognitive decline.

Objective This study aims to predict the coronavirus disease 2019(COVID-19)epidemic in Xi'an based on SEAIQR model and Dropout-LSTM model,and to provide a scientific basis for evaluating the effectiveness of the"dynamic zero-COVID policy".Methods Considering a large number of asymptomatic infections,the changing parameters,and control procedures,we developed a time-dependent susceptible-exposed-asymptomatic-infected-quarantined-removed(SEAIQR)model with stage-specific interventions.Considering the time-series characteristics of COVID-19 data and the nonlinear relationship between them,we constructed a deep learning Dropout-LSTM model.The data of newly confirmed cases in Xi'an from December 9th,2021 to January 31st,2022 were used to fit the model,and the data from February 1st,2022 to February 7th,2022 were used to evaluate the model performance of forecasting.We then calculated the effective reproduction number(Rt)and analyzed the sensitivity of the different measurement scenarios.Results The peak of newly confirmed cases predicted by the SEAIQR model would appear on December 26th,2021,with 176 cases,and the"dynamic zero-COVID policy"may be achieved in January 24th,2022,with R2=0.849.The Dropout-LSTM model can reflect the time-series and nonlinear characteristics of the data,and the predicted newly confirmed cases were highly consistent with the actual situation,with R2=0.937.The MAE and RMSE of the Dropout-LSTM model were lower than those of the SEAIQR model,indicating that the predicted results were more ideal.At the beginning of the outbreak,R0 was 5.63.Since the implementation of comprehensive control,Rt has shown a gradual downward trend,dropping to below 1.0 on December 27th,2021.With the reduction of effective contact rate,the early implementation of control measures and the improvement of immunity threshold,the peak of newly confirmed cases will continue to decrease.Conclusion The proposed Dropout-LSTM model forecasts the epidemic well,which can provide a reference for decision-making of the"dynamic zero-COVID policy."